diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,56034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.5886410669119337,
+  "eval_steps": 500,
+  "global_step": 8000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 7.358013336399172e-05,
+      "grad_norm": 12.875,
+      "learning_rate": 0.0,
+      "loss": 1.3441,
+      "step": 1
+    },
+    {
+      "epoch": 0.00014716026672798344,
+      "grad_norm": 15.0,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.7824,
+      "step": 2
+    },
+    {
+      "epoch": 0.00022074040009197518,
+      "grad_norm": 16.625,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 2.041,
+      "step": 3
+    },
+    {
+      "epoch": 0.0002943205334559669,
+      "grad_norm": 11.875,
+      "learning_rate": 5e-06,
+      "loss": 1.1798,
+      "step": 4
+    },
+    {
+      "epoch": 0.0003679006668199586,
+      "grad_norm": 12.125,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.3734,
+      "step": 5
+    },
+    {
+      "epoch": 0.00044148080018395036,
+      "grad_norm": 9.5,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.5999,
+      "step": 6
+    },
+    {
+      "epoch": 0.0005150609335479421,
+      "grad_norm": 7.03125,
+      "learning_rate": 1e-05,
+      "loss": 1.5471,
+      "step": 7
+    },
+    {
+      "epoch": 0.0005886410669119338,
+      "grad_norm": 4.71875,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 1.0944,
+      "step": 8
+    },
+    {
+      "epoch": 0.0006622212002759255,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.8931,
+      "step": 9
+    },
+    {
+      "epoch": 0.0007358013336399172,
+      "grad_norm": 2.375,
+      "learning_rate": 1.5e-05,
+      "loss": 1.0534,
+      "step": 10
+    },
+    {
+      "epoch": 0.0008093814670039089,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.8836,
+      "step": 11
+    },
+    {
+      "epoch": 0.0008829616003679007,
+      "grad_norm": 2.15625,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 1.1655,
+      "step": 12
+    },
+    {
+      "epoch": 0.0009565417337318924,
+      "grad_norm": 2.265625,
+      "learning_rate": 2e-05,
+      "loss": 1.7158,
+      "step": 13
+    },
+    {
+      "epoch": 0.0010301218670958842,
+      "grad_norm": 2.171875,
+      "learning_rate": 2.1666666666666667e-05,
+      "loss": 1.2426,
+      "step": 14
+    },
+    {
+      "epoch": 0.0011037020004598759,
+      "grad_norm": 1.6796875,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 1.048,
+      "step": 15
+    },
+    {
+      "epoch": 0.0011772821338238676,
+      "grad_norm": 1.875,
+      "learning_rate": 2.5e-05,
+      "loss": 1.3412,
+      "step": 16
+    },
+    {
+      "epoch": 0.0012508622671878592,
+      "grad_norm": 1.5078125,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 1.0901,
+      "step": 17
+    },
+    {
+      "epoch": 0.001324442400551851,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.8333333333333335e-05,
+      "loss": 0.9881,
+      "step": 18
+    },
+    {
+      "epoch": 0.0013980225339158428,
+      "grad_norm": 1.4609375,
+      "learning_rate": 3e-05,
+      "loss": 0.9804,
+      "step": 19
+    },
+    {
+      "epoch": 0.0014716026672798345,
+      "grad_norm": 1.5625,
+      "learning_rate": 3.1666666666666666e-05,
+      "loss": 1.3544,
+      "step": 20
+    },
+    {
+      "epoch": 0.0015451828006438262,
+      "grad_norm": 2.140625,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 0.9491,
+      "step": 21
+    },
+    {
+      "epoch": 0.0016187629340078179,
+      "grad_norm": 1.2890625,
+      "learning_rate": 3.5e-05,
+      "loss": 0.9426,
+      "step": 22
+    },
+    {
+      "epoch": 0.0016923430673718095,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 1.0448,
+      "step": 23
+    },
+    {
+      "epoch": 0.0017659232007358014,
+      "grad_norm": 1.3359375,
+      "learning_rate": 3.8333333333333334e-05,
+      "loss": 1.2337,
+      "step": 24
+    },
+    {
+      "epoch": 0.0018395033340997931,
+      "grad_norm": 1.3046875,
+      "learning_rate": 4e-05,
+      "loss": 1.5146,
+      "step": 25
+    },
+    {
+      "epoch": 0.0019130834674637848,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 1.1482,
+      "step": 26
+    },
+    {
+      "epoch": 0.0019866636008277765,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 0.8161,
+      "step": 27
+    },
+    {
+      "epoch": 0.0020602437341917684,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.5e-05,
+      "loss": 0.991,
+      "step": 28
+    },
+    {
+      "epoch": 0.00213382386755576,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 0.7794,
+      "step": 29
+    },
+    {
+      "epoch": 0.0022074040009197517,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.8333333333333334e-05,
+      "loss": 0.8139,
+      "step": 30
+    },
+    {
+      "epoch": 0.002280984134283743,
+      "grad_norm": 1.09375,
+      "learning_rate": 5e-05,
+      "loss": 0.8812,
+      "step": 31
+    },
+    {
+      "epoch": 0.002354564267647735,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.9999999329148614e-05,
+      "loss": 0.6803,
+      "step": 32
+    },
+    {
+      "epoch": 0.002428144401011727,
+      "grad_norm": 1.265625,
+      "learning_rate": 4.999999731659447e-05,
+      "loss": 0.9949,
+      "step": 33
+    },
+    {
+      "epoch": 0.0025017245343757185,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.9999993962337696e-05,
+      "loss": 0.7634,
+      "step": 34
+    },
+    {
+      "epoch": 0.0025753046677397104,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.9999989266378464e-05,
+      "loss": 0.9535,
+      "step": 35
+    },
+    {
+      "epoch": 0.002648884801103702,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.999998322871703e-05,
+      "loss": 0.9741,
+      "step": 36
+    },
+    {
+      "epoch": 0.0027224649344676937,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.999997584935371e-05,
+      "loss": 0.8228,
+      "step": 37
+    },
+    {
+      "epoch": 0.0027960450678316856,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.9999967128288905e-05,
+      "loss": 0.8874,
+      "step": 38
+    },
+    {
+      "epoch": 0.002869625201195677,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.9999957065523085e-05,
+      "loss": 0.7428,
+      "step": 39
+    },
+    {
+      "epoch": 0.002943205334559669,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.9999945661056786e-05,
+      "loss": 0.7836,
+      "step": 40
+    },
+    {
+      "epoch": 0.0030167854679236604,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.999993291489062e-05,
+      "loss": 0.9396,
+      "step": 41
+    },
+    {
+      "epoch": 0.0030903656012876523,
+      "grad_norm": 1.3359375,
+      "learning_rate": 4.9999918827025275e-05,
+      "loss": 0.9364,
+      "step": 42
+    },
+    {
+      "epoch": 0.0031639457346516442,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.99999033974615e-05,
+      "loss": 0.8486,
+      "step": 43
+    },
+    {
+      "epoch": 0.0032375258680156357,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.999988662620013e-05,
+      "loss": 1.0065,
+      "step": 44
+    },
+    {
+      "epoch": 0.0033111060013796276,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.9999868513242065e-05,
+      "loss": 1.1676,
+      "step": 45
+    },
+    {
+      "epoch": 0.003384686134743619,
+      "grad_norm": 1.4609375,
+      "learning_rate": 4.9999849058588276e-05,
+      "loss": 1.1662,
+      "step": 46
+    },
+    {
+      "epoch": 0.003458266268107611,
+      "grad_norm": 1.0,
+      "learning_rate": 4.99998282622398e-05,
+      "loss": 0.7682,
+      "step": 47
+    },
+    {
+      "epoch": 0.003531846401471603,
+      "grad_norm": 1.3359375,
+      "learning_rate": 4.999980612419777e-05,
+      "loss": 1.3762,
+      "step": 48
+    },
+    {
+      "epoch": 0.0036054265348355943,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.999978264446335e-05,
+      "loss": 1.1862,
+      "step": 49
+    },
+    {
+      "epoch": 0.0036790066681995862,
+      "grad_norm": 1.125,
+      "learning_rate": 4.9999757823037816e-05,
+      "loss": 1.0162,
+      "step": 50
+    },
+    {
+      "epoch": 0.0037525868015635777,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.999973165992251e-05,
+      "loss": 0.9144,
+      "step": 51
+    },
+    {
+      "epoch": 0.0038261669349275696,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.999970415511881e-05,
+      "loss": 1.7539,
+      "step": 52
+    },
+    {
+      "epoch": 0.0038997470682915615,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.999967530862821e-05,
+      "loss": 1.0273,
+      "step": 53
+    },
+    {
+      "epoch": 0.003973327201655553,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.999964512045226e-05,
+      "loss": 1.1683,
+      "step": 54
+    },
+    {
+      "epoch": 0.004046907335019545,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.999961359059257e-05,
+      "loss": 1.0855,
+      "step": 55
+    },
+    {
+      "epoch": 0.004120487468383537,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.999958071905082e-05,
+      "loss": 0.9039,
+      "step": 56
+    },
+    {
+      "epoch": 0.004194067601747528,
+      "grad_norm": 1.515625,
+      "learning_rate": 4.999954650582881e-05,
+      "loss": 0.8736,
+      "step": 57
+    },
+    {
+      "epoch": 0.00426764773511152,
+      "grad_norm": 1.3515625,
+      "learning_rate": 4.999951095092835e-05,
+      "loss": 0.9829,
+      "step": 58
+    },
+    {
+      "epoch": 0.004341227868475512,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.9999474054351357e-05,
+      "loss": 1.087,
+      "step": 59
+    },
+    {
+      "epoch": 0.0044148080018395035,
+      "grad_norm": 1.125,
+      "learning_rate": 4.99994358160998e-05,
+      "loss": 1.0548,
+      "step": 60
+    },
+    {
+      "epoch": 0.004488388135203495,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.9999396236175754e-05,
+      "loss": 1.2358,
+      "step": 61
+    },
+    {
+      "epoch": 0.004561968268567486,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.999935531458132e-05,
+      "loss": 0.9955,
+      "step": 62
+    },
+    {
+      "epoch": 0.004635548401931478,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.999931305131871e-05,
+      "loss": 1.0251,
+      "step": 63
+    },
+    {
+      "epoch": 0.00470912853529547,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.999926944639018e-05,
+      "loss": 1.0653,
+      "step": 64
+    },
+    {
+      "epoch": 0.004782708668659462,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.999922449979808e-05,
+      "loss": 1.0879,
+      "step": 65
+    },
+    {
+      "epoch": 0.004856288802023454,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.999917821154481e-05,
+      "loss": 1.0973,
+      "step": 66
+    },
+    {
+      "epoch": 0.004929868935387445,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.999913058163287e-05,
+      "loss": 0.893,
+      "step": 67
+    },
+    {
+      "epoch": 0.005003449068751437,
+      "grad_norm": 1.0,
+      "learning_rate": 4.99990816100648e-05,
+      "loss": 1.3085,
+      "step": 68
+    },
+    {
+      "epoch": 0.005077029202115429,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.9999031296843244e-05,
+      "loss": 0.9653,
+      "step": 69
+    },
+    {
+      "epoch": 0.005150609335479421,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.99989796419709e-05,
+      "loss": 1.3415,
+      "step": 70
+    },
+    {
+      "epoch": 0.005224189468843413,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.999892664545053e-05,
+      "loss": 1.1798,
+      "step": 71
+    },
+    {
+      "epoch": 0.005297769602207404,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.999887230728497e-05,
+      "loss": 0.9253,
+      "step": 72
+    },
+    {
+      "epoch": 0.0053713497355713955,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.9998816627477166e-05,
+      "loss": 0.7838,
+      "step": 73
+    },
+    {
+      "epoch": 0.0054449298689353874,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.999875960603008e-05,
+      "loss": 1.5514,
+      "step": 74
+    },
+    {
+      "epoch": 0.005518510002299379,
+      "grad_norm": 1.3671875,
+      "learning_rate": 4.9998701242946785e-05,
+      "loss": 1.3704,
+      "step": 75
+    },
+    {
+      "epoch": 0.005592090135663371,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.9998641538230415e-05,
+      "loss": 1.0745,
+      "step": 76
+    },
+    {
+      "epoch": 0.005665670269027362,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.999858049188417e-05,
+      "loss": 1.1878,
+      "step": 77
+    },
+    {
+      "epoch": 0.005739250402391354,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.999851810391132e-05,
+      "loss": 1.0663,
+      "step": 78
+    },
+    {
+      "epoch": 0.005812830535755346,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.9998454374315216e-05,
+      "loss": 1.0414,
+      "step": 79
+    },
+    {
+      "epoch": 0.005886410669119338,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.9998389303099284e-05,
+      "loss": 0.7892,
+      "step": 80
+    },
+    {
+      "epoch": 0.00595999080248333,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.999832289026701e-05,
+      "loss": 1.0337,
+      "step": 81
+    },
+    {
+      "epoch": 0.006033570935847321,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.999825513582197e-05,
+      "loss": 0.7221,
+      "step": 82
+    },
+    {
+      "epoch": 0.006107151069211313,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.999818603976779e-05,
+      "loss": 0.9394,
+      "step": 83
+    },
+    {
+      "epoch": 0.006180731202575305,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.999811560210817e-05,
+      "loss": 0.6104,
+      "step": 84
+    },
+    {
+      "epoch": 0.006254311335939297,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.999804382284692e-05,
+      "loss": 0.7438,
+      "step": 85
+    },
+    {
+      "epoch": 0.0063278914693032885,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.9997970701987855e-05,
+      "loss": 1.26,
+      "step": 86
+    },
+    {
+      "epoch": 0.0064014716026672795,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.999789623953493e-05,
+      "loss": 0.8338,
+      "step": 87
+    },
+    {
+      "epoch": 0.006475051736031271,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.9997820435492116e-05,
+      "loss": 1.2148,
+      "step": 88
+    },
+    {
+      "epoch": 0.006548631869395263,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.99977432898635e-05,
+      "loss": 1.1692,
+      "step": 89
+    },
+    {
+      "epoch": 0.006622212002759255,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.999766480265321e-05,
+      "loss": 0.9819,
+      "step": 90
+    },
+    {
+      "epoch": 0.006695792136123247,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.999758497386547e-05,
+      "loss": 1.4166,
+      "step": 91
+    },
+    {
+      "epoch": 0.006769372269487238,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.999750380350456e-05,
+      "loss": 0.7442,
+      "step": 92
+    },
+    {
+      "epoch": 0.00684295240285123,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.999742129157483e-05,
+      "loss": 1.0275,
+      "step": 93
+    },
+    {
+      "epoch": 0.006916532536215222,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.999733743808071e-05,
+      "loss": 0.8956,
+      "step": 94
+    },
+    {
+      "epoch": 0.006990112669579214,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.999725224302671e-05,
+      "loss": 0.8214,
+      "step": 95
+    },
+    {
+      "epoch": 0.007063692802943206,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.9997165706417395e-05,
+      "loss": 1.1712,
+      "step": 96
+    },
+    {
+      "epoch": 0.007137272936307197,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.99970778282574e-05,
+      "loss": 0.8904,
+      "step": 97
+    },
+    {
+      "epoch": 0.007210853069671189,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.9996988608551454e-05,
+      "loss": 1.0042,
+      "step": 98
+    },
+    {
+      "epoch": 0.0072844332030351806,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.999689804730435e-05,
+      "loss": 0.8162,
+      "step": 99
+    },
+    {
+      "epoch": 0.0073580133363991725,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.9996806144520936e-05,
+      "loss": 0.8645,
+      "step": 100
+    },
+    {
+      "epoch": 0.007431593469763164,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.999671290020615e-05,
+      "loss": 1.0243,
+      "step": 101
+    },
+    {
+      "epoch": 0.007505173603127155,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.999661831436499e-05,
+      "loss": 0.9468,
+      "step": 102
+    },
+    {
+      "epoch": 0.007578753736491147,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.999652238700253e-05,
+      "loss": 1.154,
+      "step": 103
+    },
+    {
+      "epoch": 0.007652333869855139,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.999642511812394e-05,
+      "loss": 1.4333,
+      "step": 104
+    },
+    {
+      "epoch": 0.007725914003219131,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.999632650773442e-05,
+      "loss": 0.7866,
+      "step": 105
+    },
+    {
+      "epoch": 0.007799494136583123,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.999622655583927e-05,
+      "loss": 0.7542,
+      "step": 106
+    },
+    {
+      "epoch": 0.007873074269947114,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.999612526244385e-05,
+      "loss": 1.2052,
+      "step": 107
+    },
+    {
+      "epoch": 0.007946654403311106,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.99960226275536e-05,
+      "loss": 1.369,
+      "step": 108
+    },
+    {
+      "epoch": 0.008020234536675098,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.9995918651174016e-05,
+      "loss": 0.9614,
+      "step": 109
+    },
+    {
+      "epoch": 0.00809381467003909,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.99958133333107e-05,
+      "loss": 1.0933,
+      "step": 110
+    },
+    {
+      "epoch": 0.008167394803403082,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.999570667396929e-05,
+      "loss": 0.9463,
+      "step": 111
+    },
+    {
+      "epoch": 0.008240974936767074,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.999559867315551e-05,
+      "loss": 1.0003,
+      "step": 112
+    },
+    {
+      "epoch": 0.008314555070131065,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.999548933087516e-05,
+      "loss": 0.7679,
+      "step": 113
+    },
+    {
+      "epoch": 0.008388135203495056,
+      "grad_norm": 1.34375,
+      "learning_rate": 4.9995378647134106e-05,
+      "loss": 1.2124,
+      "step": 114
+    },
+    {
+      "epoch": 0.008461715336859047,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.99952666219383e-05,
+      "loss": 0.9484,
+      "step": 115
+    },
+    {
+      "epoch": 0.00853529547022304,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.999515325529373e-05,
+      "loss": 0.8705,
+      "step": 116
+    },
+    {
+      "epoch": 0.008608875603587031,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.99950385472065e-05,
+      "loss": 1.4054,
+      "step": 117
+    },
+    {
+      "epoch": 0.008682455736951023,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.999492249768276e-05,
+      "loss": 0.688,
+      "step": 118
+    },
+    {
+      "epoch": 0.008756035870315015,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.999480510672874e-05,
+      "loss": 1.208,
+      "step": 119
+    },
+    {
+      "epoch": 0.008829616003679007,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.9994686374350744e-05,
+      "loss": 0.8497,
+      "step": 120
+    },
+    {
+      "epoch": 0.008903196137042999,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.9994566300555124e-05,
+      "loss": 0.9617,
+      "step": 121
+    },
+    {
+      "epoch": 0.00897677627040699,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.9994444885348344e-05,
+      "loss": 1.2366,
+      "step": 122
+    },
+    {
+      "epoch": 0.009050356403770983,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.999432212873692e-05,
+      "loss": 0.7457,
+      "step": 123
+    },
+    {
+      "epoch": 0.009123936537134973,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.999419803072743e-05,
+      "loss": 1.1876,
+      "step": 124
+    },
+    {
+      "epoch": 0.009197516670498965,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.999407259132655e-05,
+      "loss": 0.9813,
+      "step": 125
+    },
+    {
+      "epoch": 0.009271096803862957,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.9993945810540985e-05,
+      "loss": 0.7469,
+      "step": 126
+    },
+    {
+      "epoch": 0.009344676937226949,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.9993817688377566e-05,
+      "loss": 1.1302,
+      "step": 127
+    },
+    {
+      "epoch": 0.00941825707059094,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.999368822484315e-05,
+      "loss": 0.9609,
+      "step": 128
+    },
+    {
+      "epoch": 0.009491837203954932,
+      "grad_norm": 1.3125,
+      "learning_rate": 4.9993557419944696e-05,
+      "loss": 1.1844,
+      "step": 129
+    },
+    {
+      "epoch": 0.009565417337318924,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.999342527368922e-05,
+      "loss": 0.8395,
+      "step": 130
+    },
+    {
+      "epoch": 0.009638997470682916,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.999329178608382e-05,
+      "loss": 1.1009,
+      "step": 131
+    },
+    {
+      "epoch": 0.009712577604046908,
+      "grad_norm": 1.3125,
+      "learning_rate": 4.999315695713566e-05,
+      "loss": 1.1639,
+      "step": 132
+    },
+    {
+      "epoch": 0.0097861577374109,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.999302078685196e-05,
+      "loss": 1.095,
+      "step": 133
+    },
+    {
+      "epoch": 0.00985973787077489,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.999288327524004e-05,
+      "loss": 1.2376,
+      "step": 134
+    },
+    {
+      "epoch": 0.009933318004138882,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.999274442230729e-05,
+      "loss": 0.8503,
+      "step": 135
+    },
+    {
+      "epoch": 0.010006898137502874,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.9992604228061145e-05,
+      "loss": 1.1394,
+      "step": 136
+    },
+    {
+      "epoch": 0.010080478270866866,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.999246269250914e-05,
+      "loss": 0.8703,
+      "step": 137
+    },
+    {
+      "epoch": 0.010154058404230858,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.999231981565886e-05,
+      "loss": 1.0047,
+      "step": 138
+    },
+    {
+      "epoch": 0.01022763853759485,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.999217559751799e-05,
+      "loss": 1.3795,
+      "step": 139
+    },
+    {
+      "epoch": 0.010301218670958841,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.9992030038094243e-05,
+      "loss": 1.2553,
+      "step": 140
+    },
+    {
+      "epoch": 0.010374798804322833,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.999188313739546e-05,
+      "loss": 1.3212,
+      "step": 141
+    },
+    {
+      "epoch": 0.010448378937686825,
+      "grad_norm": 1.2890625,
+      "learning_rate": 4.999173489542951e-05,
+      "loss": 1.5042,
+      "step": 142
+    },
+    {
+      "epoch": 0.010521959071050817,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.999158531220434e-05,
+      "loss": 1.1514,
+      "step": 143
+    },
+    {
+      "epoch": 0.010595539204414807,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.9991434387728e-05,
+      "loss": 0.8765,
+      "step": 144
+    },
+    {
+      "epoch": 0.0106691193377788,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.999128212200858e-05,
+      "loss": 1.4513,
+      "step": 145
+    },
+    {
+      "epoch": 0.010742699471142791,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.999112851505424e-05,
+      "loss": 0.9521,
+      "step": 146
+    },
+    {
+      "epoch": 0.010816279604506783,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.999097356687324e-05,
+      "loss": 0.9354,
+      "step": 147
+    },
+    {
+      "epoch": 0.010889859737870775,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.999081727747389e-05,
+      "loss": 1.1123,
+      "step": 148
+    },
+    {
+      "epoch": 0.010963439871234767,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.999065964686458e-05,
+      "loss": 0.6834,
+      "step": 149
+    },
+    {
+      "epoch": 0.011037020004598759,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.9990500675053765e-05,
+      "loss": 1.0249,
+      "step": 150
+    },
+    {
+      "epoch": 0.01111060013796275,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.9990340362049974e-05,
+      "loss": 0.7333,
+      "step": 151
+    },
+    {
+      "epoch": 0.011184180271326742,
+      "grad_norm": 1.25,
+      "learning_rate": 4.999017870786182e-05,
+      "loss": 1.1691,
+      "step": 152
+    },
+    {
+      "epoch": 0.011257760404690733,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.9990015712497974e-05,
+      "loss": 1.3583,
+      "step": 153
+    },
+    {
+      "epoch": 0.011331340538054725,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.998985137596719e-05,
+      "loss": 1.1294,
+      "step": 154
+    },
+    {
+      "epoch": 0.011404920671418716,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.9989685698278274e-05,
+      "loss": 1.2782,
+      "step": 155
+    },
+    {
+      "epoch": 0.011478500804782708,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.998951867944013e-05,
+      "loss": 0.8186,
+      "step": 156
+    },
+    {
+      "epoch": 0.0115520809381467,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.998935031946171e-05,
+      "loss": 1.305,
+      "step": 157
+    },
+    {
+      "epoch": 0.011625661071510692,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.998918061835207e-05,
+      "loss": 1.0957,
+      "step": 158
+    },
+    {
+      "epoch": 0.011699241204874684,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.998900957612029e-05,
+      "loss": 0.7775,
+      "step": 159
+    },
+    {
+      "epoch": 0.011772821338238676,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.998883719277557e-05,
+      "loss": 0.8133,
+      "step": 160
+    },
+    {
+      "epoch": 0.011846401471602668,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.9988663468327156e-05,
+      "loss": 0.6707,
+      "step": 161
+    },
+    {
+      "epoch": 0.01191998160496666,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.998848840278437e-05,
+      "loss": 1.0147,
+      "step": 162
+    },
+    {
+      "epoch": 0.01199356173833065,
+      "grad_norm": 1.546875,
+      "learning_rate": 4.998831199615661e-05,
+      "loss": 1.0418,
+      "step": 163
+    },
+    {
+      "epoch": 0.012067141871694642,
+      "grad_norm": 1.0,
+      "learning_rate": 4.9988134248453333e-05,
+      "loss": 1.0659,
+      "step": 164
+    },
+    {
+      "epoch": 0.012140722005058634,
+      "grad_norm": 1.2890625,
+      "learning_rate": 4.9987955159684095e-05,
+      "loss": 1.0881,
+      "step": 165
+    },
+    {
+      "epoch": 0.012214302138422626,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.9987774729858494e-05,
+      "loss": 0.8431,
+      "step": 166
+    },
+    {
+      "epoch": 0.012287882271786617,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.998759295898622e-05,
+      "loss": 0.7996,
+      "step": 167
+    },
+    {
+      "epoch": 0.01236146240515061,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.9987409847077033e-05,
+      "loss": 1.1523,
+      "step": 168
+    },
+    {
+      "epoch": 0.012435042538514601,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.9987225394140744e-05,
+      "loss": 1.0186,
+      "step": 169
+    },
+    {
+      "epoch": 0.012508622671878593,
+      "grad_norm": 1.484375,
+      "learning_rate": 4.9987039600187266e-05,
+      "loss": 1.361,
+      "step": 170
+    },
+    {
+      "epoch": 0.012582202805242585,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.9986852465226574e-05,
+      "loss": 0.9546,
+      "step": 171
+    },
+    {
+      "epoch": 0.012655782938606577,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.99866639892687e-05,
+      "loss": 1.074,
+      "step": 172
+    },
+    {
+      "epoch": 0.012729363071970567,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.998647417232375e-05,
+      "loss": 0.7568,
+      "step": 173
+    },
+    {
+      "epoch": 0.012802943205334559,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.998628301440194e-05,
+      "loss": 1.0731,
+      "step": 174
+    },
+    {
+      "epoch": 0.012876523338698551,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.9986090515513506e-05,
+      "loss": 0.8271,
+      "step": 175
+    },
+    {
+      "epoch": 0.012950103472062543,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.9985896675668784e-05,
+      "loss": 1.154,
+      "step": 176
+    },
+    {
+      "epoch": 0.013023683605426535,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.998570149487819e-05,
+      "loss": 0.9304,
+      "step": 177
+    },
+    {
+      "epoch": 0.013097263738790527,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.998550497315218e-05,
+      "loss": 1.0314,
+      "step": 178
+    },
+    {
+      "epoch": 0.013170843872154519,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.99853071105013e-05,
+      "loss": 0.7165,
+      "step": 179
+    },
+    {
+      "epoch": 0.01324442400551851,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.998510790693619e-05,
+      "loss": 0.7941,
+      "step": 180
+    },
+    {
+      "epoch": 0.013318004138882502,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.9984907362467525e-05,
+      "loss": 0.9811,
+      "step": 181
+    },
+    {
+      "epoch": 0.013391584272246494,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.9984705477106076e-05,
+      "loss": 0.9994,
+      "step": 182
+    },
+    {
+      "epoch": 0.013465164405610484,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.9984502250862666e-05,
+      "loss": 0.9321,
+      "step": 183
+    },
+    {
+      "epoch": 0.013538744538974476,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.998429768374822e-05,
+      "loss": 0.8844,
+      "step": 184
+    },
+    {
+      "epoch": 0.013612324672338468,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.998409177577369e-05,
+      "loss": 1.1324,
+      "step": 185
+    },
+    {
+      "epoch": 0.01368590480570246,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.998388452695015e-05,
+      "loss": 0.7917,
+      "step": 186
+    },
+    {
+      "epoch": 0.013759484939066452,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.998367593728872e-05,
+      "loss": 0.8042,
+      "step": 187
+    },
+    {
+      "epoch": 0.013833065072430444,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.998346600680059e-05,
+      "loss": 1.1884,
+      "step": 188
+    },
+    {
+      "epoch": 0.013906645205794436,
+      "grad_norm": 1.328125,
+      "learning_rate": 4.998325473549702e-05,
+      "loss": 1.6683,
+      "step": 189
+    },
+    {
+      "epoch": 0.013980225339158428,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.998304212338936e-05,
+      "loss": 1.213,
+      "step": 190
+    },
+    {
+      "epoch": 0.01405380547252242,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.998282817048902e-05,
+      "loss": 0.8588,
+      "step": 191
+    },
+    {
+      "epoch": 0.014127385605886411,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.998261287680747e-05,
+      "loss": 1.2776,
+      "step": 192
+    },
+    {
+      "epoch": 0.014200965739250402,
+      "grad_norm": 0.875,
+      "learning_rate": 4.9982396242356265e-05,
+      "loss": 0.7973,
+      "step": 193
+    },
+    {
+      "epoch": 0.014274545872614394,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.998217826714705e-05,
+      "loss": 0.9769,
+      "step": 194
+    },
+    {
+      "epoch": 0.014348126005978385,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.9981958951191507e-05,
+      "loss": 1.0079,
+      "step": 195
+    },
+    {
+      "epoch": 0.014421706139342377,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.9981738294501416e-05,
+      "loss": 1.121,
+      "step": 196
+    },
+    {
+      "epoch": 0.01449528627270637,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.998151629708861e-05,
+      "loss": 1.2495,
+      "step": 197
+    },
+    {
+      "epoch": 0.014568866406070361,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.998129295896502e-05,
+      "loss": 0.8233,
+      "step": 198
+    },
+    {
+      "epoch": 0.014642446539434353,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.998106828014261e-05,
+      "loss": 1.0031,
+      "step": 199
+    },
+    {
+      "epoch": 0.014716026672798345,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.998084226063344e-05,
+      "loss": 0.7791,
+      "step": 200
+    },
+    {
+      "epoch": 0.014789606806162337,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.998061490044966e-05,
+      "loss": 0.8824,
+      "step": 201
+    },
+    {
+      "epoch": 0.014863186939526329,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.998038619960346e-05,
+      "loss": 1.125,
+      "step": 202
+    },
+    {
+      "epoch": 0.014936767072890319,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.998015615810711e-05,
+      "loss": 0.828,
+      "step": 203
+    },
+    {
+      "epoch": 0.01501034720625431,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.997992477597295e-05,
+      "loss": 0.9252,
+      "step": 204
+    },
+    {
+      "epoch": 0.015083927339618303,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.9979692053213425e-05,
+      "loss": 1.188,
+      "step": 205
+    },
+    {
+      "epoch": 0.015157507472982295,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.9979457989841e-05,
+      "loss": 0.9181,
+      "step": 206
+    },
+    {
+      "epoch": 0.015231087606346286,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.9979222585868245e-05,
+      "loss": 0.6557,
+      "step": 207
+    },
+    {
+      "epoch": 0.015304667739710278,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.997898584130779e-05,
+      "loss": 1.1323,
+      "step": 208
+    },
+    {
+      "epoch": 0.01537824787307427,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.9978747756172345e-05,
+      "loss": 0.9684,
+      "step": 209
+    },
+    {
+      "epoch": 0.015451828006438262,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.9978508330474686e-05,
+      "loss": 0.9946,
+      "step": 210
+    },
+    {
+      "epoch": 0.015525408139802254,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.9978267564227666e-05,
+      "loss": 0.995,
+      "step": 211
+    },
+    {
+      "epoch": 0.015598988273166246,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.99780254574442e-05,
+      "loss": 0.756,
+      "step": 212
+    },
+    {
+      "epoch": 0.015672568406530238,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.997778201013729e-05,
+      "loss": 0.9518,
+      "step": 213
+    },
+    {
+      "epoch": 0.015746148539894228,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.997753722231998e-05,
+      "loss": 1.4175,
+      "step": 214
+    },
+    {
+      "epoch": 0.01581972867325822,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.997729109400544e-05,
+      "loss": 0.6918,
+      "step": 215
+    },
+    {
+      "epoch": 0.015893308806622212,
+      "grad_norm": 1.265625,
+      "learning_rate": 4.9977043625206853e-05,
+      "loss": 1.2282,
+      "step": 216
+    },
+    {
+      "epoch": 0.015966888939986205,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.997679481593751e-05,
+      "loss": 0.6038,
+      "step": 217
+    },
+    {
+      "epoch": 0.016040469073350196,
+      "grad_norm": 1.265625,
+      "learning_rate": 4.9976544666210765e-05,
+      "loss": 1.4195,
+      "step": 218
+    },
+    {
+      "epoch": 0.016114049206714186,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.997629317604005e-05,
+      "loss": 1.161,
+      "step": 219
+    },
+    {
+      "epoch": 0.01618762934007818,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.997604034543885e-05,
+      "loss": 1.6678,
+      "step": 220
+    },
+    {
+      "epoch": 0.01626120947344217,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.997578617442073e-05,
+      "loss": 1.1499,
+      "step": 221
+    },
+    {
+      "epoch": 0.016334789606806163,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.9975530662999344e-05,
+      "loss": 0.8027,
+      "step": 222
+    },
+    {
+      "epoch": 0.016408369740170153,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.997527381118839e-05,
+      "loss": 1.3696,
+      "step": 223
+    },
+    {
+      "epoch": 0.016481949873534147,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.997501561900167e-05,
+      "loss": 1.175,
+      "step": 224
+    },
+    {
+      "epoch": 0.016555530006898137,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.997475608645304e-05,
+      "loss": 1.0295,
+      "step": 225
+    },
+    {
+      "epoch": 0.01662911014026213,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.9974495213556414e-05,
+      "loss": 0.6764,
+      "step": 226
+    },
+    {
+      "epoch": 0.01670269027362612,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.9974233000325806e-05,
+      "loss": 0.9065,
+      "step": 227
+    },
+    {
+      "epoch": 0.01677627040699011,
+      "grad_norm": 1.125,
+      "learning_rate": 4.9973969446775275e-05,
+      "loss": 1.1367,
+      "step": 228
+    },
+    {
+      "epoch": 0.016849850540354105,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.9973704552918974e-05,
+      "loss": 1.4926,
+      "step": 229
+    },
+    {
+      "epoch": 0.016923430673718095,
+      "grad_norm": 1.3359375,
+      "learning_rate": 4.997343831877112e-05,
+      "loss": 1.0596,
+      "step": 230
+    },
+    {
+      "epoch": 0.01699701080708209,
+      "grad_norm": 1.3203125,
+      "learning_rate": 4.9973170744346e-05,
+      "loss": 1.6111,
+      "step": 231
+    },
+    {
+      "epoch": 0.01707059094044608,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.997290182965797e-05,
+      "loss": 0.8489,
+      "step": 232
+    },
+    {
+      "epoch": 0.017144171073810072,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.997263157472147e-05,
+      "loss": 0.8698,
+      "step": 233
+    },
+    {
+      "epoch": 0.017217751207174063,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.9972359979551e-05,
+      "loss": 1.535,
+      "step": 234
+    },
+    {
+      "epoch": 0.017291331340538056,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.997208704416113e-05,
+      "loss": 0.9235,
+      "step": 235
+    },
+    {
+      "epoch": 0.017364911473902046,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.9971812768566527e-05,
+      "loss": 0.9751,
+      "step": 236
+    },
+    {
+      "epoch": 0.017438491607266036,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.997153715278189e-05,
+      "loss": 1.4104,
+      "step": 237
+    },
+    {
+      "epoch": 0.01751207174063003,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.9971260196822015e-05,
+      "loss": 1.7236,
+      "step": 238
+    },
+    {
+      "epoch": 0.01758565187399402,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.9970981900701776e-05,
+      "loss": 1.0166,
+      "step": 239
+    },
+    {
+      "epoch": 0.017659232007358014,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.9970702264436095e-05,
+      "loss": 0.9333,
+      "step": 240
+    },
+    {
+      "epoch": 0.017732812140722004,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.9970421288039994e-05,
+      "loss": 1.1386,
+      "step": 241
+    },
+    {
+      "epoch": 0.017806392274085998,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.997013897152854e-05,
+      "loss": 1.3691,
+      "step": 242
+    },
+    {
+      "epoch": 0.017879972407449988,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.996985531491688e-05,
+      "loss": 0.8539,
+      "step": 243
+    },
+    {
+      "epoch": 0.01795355254081398,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.996957031822026e-05,
+      "loss": 0.7394,
+      "step": 244
+    },
+    {
+      "epoch": 0.01802713267417797,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.996928398145396e-05,
+      "loss": 0.9116,
+      "step": 245
+    },
+    {
+      "epoch": 0.018100712807541965,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.996899630463335e-05,
+      "loss": 1.235,
+      "step": 246
+    },
+    {
+      "epoch": 0.018174292940905955,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.996870728777386e-05,
+      "loss": 0.906,
+      "step": 247
+    },
+    {
+      "epoch": 0.018247873074269946,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.9968416930891016e-05,
+      "loss": 0.5867,
+      "step": 248
+    },
+    {
+      "epoch": 0.01832145320763394,
+      "grad_norm": 1.375,
+      "learning_rate": 4.9968125234000396e-05,
+      "loss": 0.7491,
+      "step": 249
+    },
+    {
+      "epoch": 0.01839503334099793,
+      "grad_norm": 1.375,
+      "learning_rate": 4.9967832197117645e-05,
+      "loss": 1.0552,
+      "step": 250
+    },
+    {
+      "epoch": 0.018468613474361923,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.99675378202585e-05,
+      "loss": 1.0264,
+      "step": 251
+    },
+    {
+      "epoch": 0.018542193607725913,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.996724210343876e-05,
+      "loss": 1.0846,
+      "step": 252
+    },
+    {
+      "epoch": 0.018615773741089907,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.9966945046674294e-05,
+      "loss": 1.0058,
+      "step": 253
+    },
+    {
+      "epoch": 0.018689353874453897,
+      "grad_norm": 1.28125,
+      "learning_rate": 4.996664664998104e-05,
+      "loss": 1.09,
+      "step": 254
+    },
+    {
+      "epoch": 0.01876293400781789,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.9966346913375016e-05,
+      "loss": 0.8106,
+      "step": 255
+    },
+    {
+      "epoch": 0.01883651414118188,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.996604583687231e-05,
+      "loss": 1.338,
+      "step": 256
+    },
+    {
+      "epoch": 0.01891009427454587,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.9965743420489076e-05,
+      "loss": 0.845,
+      "step": 257
+    },
+    {
+      "epoch": 0.018983674407909865,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.996543966424155e-05,
+      "loss": 1.1945,
+      "step": 258
+    },
+    {
+      "epoch": 0.019057254541273855,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.996513456814602e-05,
+      "loss": 0.8249,
+      "step": 259
+    },
+    {
+      "epoch": 0.01913083467463785,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.996482813221888e-05,
+      "loss": 1.2986,
+      "step": 260
+    },
+    {
+      "epoch": 0.01920441480800184,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.996452035647656e-05,
+      "loss": 0.6122,
+      "step": 261
+    },
+    {
+      "epoch": 0.019277994941365832,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.996421124093559e-05,
+      "loss": 0.8912,
+      "step": 262
+    },
+    {
+      "epoch": 0.019351575074729822,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.9963900785612546e-05,
+      "loss": 1.1519,
+      "step": 263
+    },
+    {
+      "epoch": 0.019425155208093816,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.99635889905241e-05,
+      "loss": 0.8285,
+      "step": 264
+    },
+    {
+      "epoch": 0.019498735341457806,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.996327585568699e-05,
+      "loss": 1.0632,
+      "step": 265
+    },
+    {
+      "epoch": 0.0195723154748218,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.9962961381118005e-05,
+      "loss": 1.0233,
+      "step": 266
+    },
+    {
+      "epoch": 0.01964589560818579,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.996264556683403e-05,
+      "loss": 0.9842,
+      "step": 267
+    },
+    {
+      "epoch": 0.01971947574154978,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.996232841285202e-05,
+      "loss": 1.4804,
+      "step": 268
+    },
+    {
+      "epoch": 0.019793055874913774,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.996200991918899e-05,
+      "loss": 1.5742,
+      "step": 269
+    },
+    {
+      "epoch": 0.019866636008277764,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.9961690085862035e-05,
+      "loss": 0.7653,
+      "step": 270
+    },
+    {
+      "epoch": 0.019940216141641758,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.996136891288832e-05,
+      "loss": 0.8671,
+      "step": 271
+    },
+    {
+      "epoch": 0.020013796275005748,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.9961046400285075e-05,
+      "loss": 1.0324,
+      "step": 272
+    },
+    {
+      "epoch": 0.02008737640836974,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.9960722548069624e-05,
+      "loss": 0.728,
+      "step": 273
+    },
+    {
+      "epoch": 0.02016095654173373,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.996039735625932e-05,
+      "loss": 0.7373,
+      "step": 274
+    },
+    {
+      "epoch": 0.020234536675097725,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.996007082487165e-05,
+      "loss": 0.9377,
+      "step": 275
+    },
+    {
+      "epoch": 0.020308116808461715,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.9959742953924125e-05,
+      "loss": 1.3766,
+      "step": 276
+    },
+    {
+      "epoch": 0.020381696941825705,
+      "grad_norm": 1.0,
+      "learning_rate": 4.995941374343432e-05,
+      "loss": 0.9286,
+      "step": 277
+    },
+    {
+      "epoch": 0.0204552770751897,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.9959083193419934e-05,
+      "loss": 1.1085,
+      "step": 278
+    },
+    {
+      "epoch": 0.02052885720855369,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.995875130389869e-05,
+      "loss": 0.7682,
+      "step": 279
+    },
+    {
+      "epoch": 0.020602437341917683,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.995841807488841e-05,
+      "loss": 0.9412,
+      "step": 280
+    },
+    {
+      "epoch": 0.020676017475281673,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.995808350640697e-05,
+      "loss": 1.086,
+      "step": 281
+    },
+    {
+      "epoch": 0.020749597608645667,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.995774759847232e-05,
+      "loss": 1.0259,
+      "step": 282
+    },
+    {
+      "epoch": 0.020823177742009657,
+      "grad_norm": 1.2890625,
+      "learning_rate": 4.995741035110249e-05,
+      "loss": 1.4607,
+      "step": 283
+    },
+    {
+      "epoch": 0.02089675787537365,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.9957071764315596e-05,
+      "loss": 0.8889,
+      "step": 284
+    },
+    {
+      "epoch": 0.02097033800873764,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.99567318381298e-05,
+      "loss": 1.0494,
+      "step": 285
+    },
+    {
+      "epoch": 0.021043918142101634,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.995639057256334e-05,
+      "loss": 1.1224,
+      "step": 286
+    },
+    {
+      "epoch": 0.021117498275465624,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.995604796763453e-05,
+      "loss": 1.0876,
+      "step": 287
+    },
+    {
+      "epoch": 0.021191078408829615,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.995570402336176e-05,
+      "loss": 0.8649,
+      "step": 288
+    },
+    {
+      "epoch": 0.021264658542193608,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.995535873976349e-05,
+      "loss": 0.5695,
+      "step": 289
+    },
+    {
+      "epoch": 0.0213382386755576,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.9955012116858256e-05,
+      "loss": 0.8223,
+      "step": 290
+    },
+    {
+      "epoch": 0.021411818808921592,
+      "grad_norm": 1.125,
+      "learning_rate": 4.995466415466465e-05,
+      "loss": 1.0022,
+      "step": 291
+    },
+    {
+      "epoch": 0.021485398942285582,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.9954314853201355e-05,
+      "loss": 0.894,
+      "step": 292
+    },
+    {
+      "epoch": 0.021558979075649576,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.995396421248712e-05,
+      "loss": 0.9251,
+      "step": 293
+    },
+    {
+      "epoch": 0.021632559209013566,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.9953612232540734e-05,
+      "loss": 0.7351,
+      "step": 294
+    },
+    {
+      "epoch": 0.02170613934237756,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.9953258913381126e-05,
+      "loss": 1.2922,
+      "step": 295
+    },
+    {
+      "epoch": 0.02177971947574155,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.9952904255027246e-05,
+      "loss": 0.5528,
+      "step": 296
+    },
+    {
+      "epoch": 0.02185329960910554,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.995254825749812e-05,
+      "loss": 0.9348,
+      "step": 297
+    },
+    {
+      "epoch": 0.021926879742469534,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.9952190920812856e-05,
+      "loss": 0.7843,
+      "step": 298
+    },
+    {
+      "epoch": 0.022000459875833524,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.995183224499064e-05,
+      "loss": 1.1061,
+      "step": 299
+    },
+    {
+      "epoch": 0.022074040009197517,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.99514722300507e-05,
+      "loss": 0.8966,
+      "step": 300
+    },
+    {
+      "epoch": 0.022147620142561508,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.995111087601239e-05,
+      "loss": 0.7591,
+      "step": 301
+    },
+    {
+      "epoch": 0.0222212002759255,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.995074818289507e-05,
+      "loss": 0.8156,
+      "step": 302
+    },
+    {
+      "epoch": 0.02229478040928949,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.995038415071823e-05,
+      "loss": 0.7609,
+      "step": 303
+    },
+    {
+      "epoch": 0.022368360542653485,
+      "grad_norm": 1.5234375,
+      "learning_rate": 4.995001877950139e-05,
+      "loss": 1.1763,
+      "step": 304
+    },
+    {
+      "epoch": 0.022441940676017475,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.994965206926417e-05,
+      "loss": 1.0027,
+      "step": 305
+    },
+    {
+      "epoch": 0.022515520809381465,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.994928402002625e-05,
+      "loss": 1.2505,
+      "step": 306
+    },
+    {
+      "epoch": 0.02258910094274546,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.994891463180737e-05,
+      "loss": 1.0475,
+      "step": 307
+    },
+    {
+      "epoch": 0.02266268107610945,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.994854390462737e-05,
+      "loss": 1.2729,
+      "step": 308
+    },
+    {
+      "epoch": 0.022736261209473443,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.994817183850614e-05,
+      "loss": 0.6231,
+      "step": 309
+    },
+    {
+      "epoch": 0.022809841342837433,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.994779843346365e-05,
+      "loss": 1.3261,
+      "step": 310
+    },
+    {
+      "epoch": 0.022883421476201427,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.994742368951993e-05,
+      "loss": 1.1003,
+      "step": 311
+    },
+    {
+      "epoch": 0.022957001609565417,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.9947047606695104e-05,
+      "loss": 1.0592,
+      "step": 312
+    },
+    {
+      "epoch": 0.02303058174292941,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.994667018500935e-05,
+      "loss": 0.8715,
+      "step": 313
+    },
+    {
+      "epoch": 0.0231041618762934,
+      "grad_norm": 1.453125,
+      "learning_rate": 4.9946291424482927e-05,
+      "loss": 1.2742,
+      "step": 314
+    },
+    {
+      "epoch": 0.023177742009657394,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.994591132513616e-05,
+      "loss": 1.0977,
+      "step": 315
+    },
+    {
+      "epoch": 0.023251322143021384,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.994552988698945e-05,
+      "loss": 0.8418,
+      "step": 316
+    },
+    {
+      "epoch": 0.023324902276385374,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.9945147110063264e-05,
+      "loss": 1.0759,
+      "step": 317
+    },
+    {
+      "epoch": 0.023398482409749368,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.994476299437814e-05,
+      "loss": 0.8482,
+      "step": 318
+    },
+    {
+      "epoch": 0.023472062543113358,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.994437753995471e-05,
+      "loss": 0.8143,
+      "step": 319
+    },
+    {
+      "epoch": 0.023545642676477352,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.994399074681364e-05,
+      "loss": 0.8498,
+      "step": 320
+    },
+    {
+      "epoch": 0.023619222809841342,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.9943602614975714e-05,
+      "loss": 0.9454,
+      "step": 321
+    },
+    {
+      "epoch": 0.023692802943205336,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.994321314446174e-05,
+      "loss": 0.7395,
+      "step": 322
+    },
+    {
+      "epoch": 0.023766383076569326,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.9942822335292624e-05,
+      "loss": 0.9949,
+      "step": 323
+    },
+    {
+      "epoch": 0.02383996320993332,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.9942430187489354e-05,
+      "loss": 0.7036,
+      "step": 324
+    },
+    {
+      "epoch": 0.02391354334329731,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.994203670107295e-05,
+      "loss": 1.0528,
+      "step": 325
+    },
+    {
+      "epoch": 0.0239871234766613,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.994164187606456e-05,
+      "loss": 0.9755,
+      "step": 326
+    },
+    {
+      "epoch": 0.024060703610025293,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.9941245712485354e-05,
+      "loss": 0.9171,
+      "step": 327
+    },
+    {
+      "epoch": 0.024134283743389284,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.994084821035659e-05,
+      "loss": 1.0351,
+      "step": 328
+    },
+    {
+      "epoch": 0.024207863876753277,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.994044936969961e-05,
+      "loss": 0.8692,
+      "step": 329
+    },
+    {
+      "epoch": 0.024281444010117267,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.9940049190535833e-05,
+      "loss": 1.2068,
+      "step": 330
+    },
+    {
+      "epoch": 0.02435502414348126,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.9939647672886714e-05,
+      "loss": 0.896,
+      "step": 331
+    },
+    {
+      "epoch": 0.02442860427684525,
+      "grad_norm": 1.3046875,
+      "learning_rate": 4.99392448167738e-05,
+      "loss": 1.3046,
+      "step": 332
+    },
+    {
+      "epoch": 0.024502184410209245,
+      "grad_norm": 0.875,
+      "learning_rate": 4.993884062221873e-05,
+      "loss": 0.6141,
+      "step": 333
+    },
+    {
+      "epoch": 0.024575764543573235,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.9938435089243187e-05,
+      "loss": 0.8297,
+      "step": 334
+    },
+    {
+      "epoch": 0.02464934467693723,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.993802821786893e-05,
+      "loss": 1.2474,
+      "step": 335
+    },
+    {
+      "epoch": 0.02472292481030122,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.99376200081178e-05,
+      "loss": 1.4081,
+      "step": 336
+    },
+    {
+      "epoch": 0.02479650494366521,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.99372104600117e-05,
+      "loss": 0.8162,
+      "step": 337
+    },
+    {
+      "epoch": 0.024870085077029203,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.9936799573572626e-05,
+      "loss": 0.7549,
+      "step": 338
+    },
+    {
+      "epoch": 0.024943665210393193,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.9936387348822604e-05,
+      "loss": 0.8515,
+      "step": 339
+    },
+    {
+      "epoch": 0.025017245343757186,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.993597378578378e-05,
+      "loss": 0.4967,
+      "step": 340
+    },
+    {
+      "epoch": 0.025090825477121177,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.993555888447834e-05,
+      "loss": 0.7248,
+      "step": 341
+    },
+    {
+      "epoch": 0.02516440561048517,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.9935142644928545e-05,
+      "loss": 0.8738,
+      "step": 342
+    },
+    {
+      "epoch": 0.02523798574384916,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.993472506715675e-05,
+      "loss": 1.2474,
+      "step": 343
+    },
+    {
+      "epoch": 0.025311565877213154,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.993430615118535e-05,
+      "loss": 0.834,
+      "step": 344
+    },
+    {
+      "epoch": 0.025385146010577144,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.993388589703684e-05,
+      "loss": 1.1228,
+      "step": 345
+    },
+    {
+      "epoch": 0.025458726143941134,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.993346430473376e-05,
+      "loss": 0.8792,
+      "step": 346
+    },
+    {
+      "epoch": 0.025532306277305128,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.993304137429874e-05,
+      "loss": 0.9142,
+      "step": 347
+    },
+    {
+      "epoch": 0.025605886410669118,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.9932617105754486e-05,
+      "loss": 1.0518,
+      "step": 348
+    },
+    {
+      "epoch": 0.02567946654403311,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.993219149912376e-05,
+      "loss": 1.2483,
+      "step": 349
+    },
+    {
+      "epoch": 0.025753046677397102,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.993176455442941e-05,
+      "loss": 1.0214,
+      "step": 350
+    },
+    {
+      "epoch": 0.025826626810761096,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.993133627169435e-05,
+      "loss": 1.2017,
+      "step": 351
+    },
+    {
+      "epoch": 0.025900206944125086,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.993090665094156e-05,
+      "loss": 0.7452,
+      "step": 352
+    },
+    {
+      "epoch": 0.02597378707748908,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.993047569219408e-05,
+      "loss": 1.2199,
+      "step": 353
+    },
+    {
+      "epoch": 0.02604736721085307,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.993004339547508e-05,
+      "loss": 0.9061,
+      "step": 354
+    },
+    {
+      "epoch": 0.026120947344217063,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.992960976080772e-05,
+      "loss": 0.9567,
+      "step": 355
+    },
+    {
+      "epoch": 0.026194527477581053,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.9929174788215296e-05,
+      "loss": 0.8923,
+      "step": 356
+    },
+    {
+      "epoch": 0.026268107610945043,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.992873847772115e-05,
+      "loss": 0.8817,
+      "step": 357
+    },
+    {
+      "epoch": 0.026341687744309037,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.992830082934869e-05,
+      "loss": 0.6113,
+      "step": 358
+    },
+    {
+      "epoch": 0.026415267877673027,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.9927861843121403e-05,
+      "loss": 1.0466,
+      "step": 359
+    },
+    {
+      "epoch": 0.02648884801103702,
+      "grad_norm": 1.28125,
+      "learning_rate": 4.9927421519062864e-05,
+      "loss": 1.1488,
+      "step": 360
+    },
+    {
+      "epoch": 0.02656242814440101,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.9926979857196686e-05,
+      "loss": 0.8077,
+      "step": 361
+    },
+    {
+      "epoch": 0.026636008277765005,
+      "grad_norm": 1.265625,
+      "learning_rate": 4.992653685754658e-05,
+      "loss": 1.7968,
+      "step": 362
+    },
+    {
+      "epoch": 0.026709588411128995,
+      "grad_norm": 1.0,
+      "learning_rate": 4.992609252013632e-05,
+      "loss": 1.1667,
+      "step": 363
+    },
+    {
+      "epoch": 0.02678316854449299,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.9925646844989756e-05,
+      "loss": 0.9725,
+      "step": 364
+    },
+    {
+      "epoch": 0.02685674867785698,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.99251998321308e-05,
+      "loss": 1.125,
+      "step": 365
+    },
+    {
+      "epoch": 0.02693032881122097,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.992475148158344e-05,
+      "loss": 1.1913,
+      "step": 366
+    },
+    {
+      "epoch": 0.027003908944584962,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.992430179337176e-05,
+      "loss": 1.0423,
+      "step": 367
+    },
+    {
+      "epoch": 0.027077489077948953,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.9923850767519865e-05,
+      "loss": 0.9187,
+      "step": 368
+    },
+    {
+      "epoch": 0.027151069211312946,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.992339840405198e-05,
+      "loss": 1.6005,
+      "step": 369
+    },
+    {
+      "epoch": 0.027224649344676936,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.9922944702992375e-05,
+      "loss": 0.6701,
+      "step": 370
+    },
+    {
+      "epoch": 0.02729822947804093,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.9922489664365405e-05,
+      "loss": 1.3002,
+      "step": 371
+    },
+    {
+      "epoch": 0.02737180961140492,
+      "grad_norm": 1.125,
+      "learning_rate": 4.992203328819548e-05,
+      "loss": 0.8571,
+      "step": 372
+    },
+    {
+      "epoch": 0.027445389744768914,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.9921575574507095e-05,
+      "loss": 0.9139,
+      "step": 373
+    },
+    {
+      "epoch": 0.027518969878132904,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.992111652332483e-05,
+      "loss": 1.1773,
+      "step": 374
+    },
+    {
+      "epoch": 0.027592550011496894,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.99206561346733e-05,
+      "loss": 1.0986,
+      "step": 375
+    },
+    {
+      "epoch": 0.027666130144860888,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.992019440857724e-05,
+      "loss": 1.3481,
+      "step": 376
+    },
+    {
+      "epoch": 0.027739710278224878,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.9919731345061396e-05,
+      "loss": 1.1635,
+      "step": 377
+    },
+    {
+      "epoch": 0.02781329041158887,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.9919266944150643e-05,
+      "loss": 1.1497,
+      "step": 378
+    },
+    {
+      "epoch": 0.02788687054495286,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.9918801205869904e-05,
+      "loss": 1.1652,
+      "step": 379
+    },
+    {
+      "epoch": 0.027960450678316855,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.991833413024416e-05,
+      "loss": 0.95,
+      "step": 380
+    },
+    {
+      "epoch": 0.028034030811680846,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.99178657172985e-05,
+      "loss": 0.8521,
+      "step": 381
+    },
+    {
+      "epoch": 0.02810761094504484,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.991739596705804e-05,
+      "loss": 1.1388,
+      "step": 382
+    },
+    {
+      "epoch": 0.02818119107840883,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.9916924879548e-05,
+      "loss": 0.9586,
+      "step": 383
+    },
+    {
+      "epoch": 0.028254771211772823,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.991645245479367e-05,
+      "loss": 0.8291,
+      "step": 384
+    },
+    {
+      "epoch": 0.028328351345136813,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.9915978692820395e-05,
+      "loss": 1.2366,
+      "step": 385
+    },
+    {
+      "epoch": 0.028401931478500803,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.99155035936536e-05,
+      "loss": 0.766,
+      "step": 386
+    },
+    {
+      "epoch": 0.028475511611864797,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.991502715731879e-05,
+      "loss": 1.0383,
+      "step": 387
+    },
+    {
+      "epoch": 0.028549091745228787,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.991454938384153e-05,
+      "loss": 0.8598,
+      "step": 388
+    },
+    {
+      "epoch": 0.02862267187859278,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.991407027324746e-05,
+      "loss": 0.9006,
+      "step": 389
+    },
+    {
+      "epoch": 0.02869625201195677,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.9913589825562294e-05,
+      "loss": 0.7649,
+      "step": 390
+    },
+    {
+      "epoch": 0.028769832145320764,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.991310804081182e-05,
+      "loss": 1.4923,
+      "step": 391
+    },
+    {
+      "epoch": 0.028843412278684755,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.991262491902189e-05,
+      "loss": 1.223,
+      "step": 392
+    },
+    {
+      "epoch": 0.02891699241204875,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.991214046021843e-05,
+      "loss": 0.8991,
+      "step": 393
+    },
+    {
+      "epoch": 0.02899057254541274,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.991165466442744e-05,
+      "loss": 0.9163,
+      "step": 394
+    },
+    {
+      "epoch": 0.02906415267877673,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.9911167531675006e-05,
+      "loss": 0.7929,
+      "step": 395
+    },
+    {
+      "epoch": 0.029137732812140722,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.9910679061987266e-05,
+      "loss": 1.0313,
+      "step": 396
+    },
+    {
+      "epoch": 0.029211312945504712,
+      "grad_norm": 1.28125,
+      "learning_rate": 4.9910189255390414e-05,
+      "loss": 0.9564,
+      "step": 397
+    },
+    {
+      "epoch": 0.029284893078868706,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.990969811191076e-05,
+      "loss": 0.674,
+      "step": 398
+    },
+    {
+      "epoch": 0.029358473212232696,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.990920563157466e-05,
+      "loss": 0.7931,
+      "step": 399
+    },
+    {
+      "epoch": 0.02943205334559669,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.990871181440854e-05,
+      "loss": 0.898,
+      "step": 400
+    },
+    {
+      "epoch": 0.02950563347896068,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.99082166604389e-05,
+      "loss": 1.0065,
+      "step": 401
+    },
+    {
+      "epoch": 0.029579213612324674,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.9907720169692315e-05,
+      "loss": 1.0143,
+      "step": 402
+    },
+    {
+      "epoch": 0.029652793745688664,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.990722234219544e-05,
+      "loss": 1.067,
+      "step": 403
+    },
+    {
+      "epoch": 0.029726373879052657,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.9906723177974976e-05,
+      "loss": 0.8685,
+      "step": 404
+    },
+    {
+      "epoch": 0.029799954012416648,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.990622267705772e-05,
+      "loss": 0.858,
+      "step": 405
+    },
+    {
+      "epoch": 0.029873534145780638,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.990572083947054e-05,
+      "loss": 0.9264,
+      "step": 406
+    },
+    {
+      "epoch": 0.02994711427914463,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.990521766524037e-05,
+      "loss": 1.0102,
+      "step": 407
+    },
+    {
+      "epoch": 0.03002069441250862,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.9904713154394197e-05,
+      "loss": 1.2105,
+      "step": 408
+    },
+    {
+      "epoch": 0.030094274545872615,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.9904207306959104e-05,
+      "loss": 1.1406,
+      "step": 409
+    },
+    {
+      "epoch": 0.030167854679236605,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.990370012296225e-05,
+      "loss": 0.9657,
+      "step": 410
+    },
+    {
+      "epoch": 0.0302414348126006,
+      "grad_norm": 1.0,
+      "learning_rate": 4.990319160243084e-05,
+      "loss": 1.0093,
+      "step": 411
+    },
+    {
+      "epoch": 0.03031501494596459,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.9902681745392186e-05,
+      "loss": 0.8834,
+      "step": 412
+    },
+    {
+      "epoch": 0.030388595079328583,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.990217055187362e-05,
+      "loss": 1.045,
+      "step": 413
+    },
+    {
+      "epoch": 0.030462175212692573,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.990165802190261e-05,
+      "loss": 1.1901,
+      "step": 414
+    },
+    {
+      "epoch": 0.030535755346056563,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.990114415550663e-05,
+      "loss": 1.0724,
+      "step": 415
+    },
+    {
+      "epoch": 0.030609335479420557,
+      "grad_norm": 1.4140625,
+      "learning_rate": 4.990062895271329e-05,
+      "loss": 0.7789,
+      "step": 416
+    },
+    {
+      "epoch": 0.030682915612784547,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.9900112413550216e-05,
+      "loss": 1.0514,
+      "step": 417
+    },
+    {
+      "epoch": 0.03075649574614854,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.9899594538045136e-05,
+      "loss": 0.9318,
+      "step": 418
+    },
+    {
+      "epoch": 0.03083007587951253,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.989907532622585e-05,
+      "loss": 0.7865,
+      "step": 419
+    },
+    {
+      "epoch": 0.030903656012876524,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.989855477812022e-05,
+      "loss": 1.2269,
+      "step": 420
+    },
+    {
+      "epoch": 0.030977236146240514,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.989803289375618e-05,
+      "loss": 1.0722,
+      "step": 421
+    },
+    {
+      "epoch": 0.031050816279604508,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.989750967316174e-05,
+      "loss": 0.9473,
+      "step": 422
+    },
+    {
+      "epoch": 0.0311243964129685,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.989698511636498e-05,
+      "loss": 1.2507,
+      "step": 423
+    },
+    {
+      "epoch": 0.031197976546332492,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.989645922339406e-05,
+      "loss": 0.9462,
+      "step": 424
+    },
+    {
+      "epoch": 0.03127155667969648,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.9895931994277187e-05,
+      "loss": 1.3416,
+      "step": 425
+    },
+    {
+      "epoch": 0.031345136813060476,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.989540342904267e-05,
+      "loss": 0.8029,
+      "step": 426
+    },
+    {
+      "epoch": 0.031418716946424466,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.989487352771887e-05,
+      "loss": 1.3004,
+      "step": 427
+    },
+    {
+      "epoch": 0.031492297079788456,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.9894342290334227e-05,
+      "loss": 0.7786,
+      "step": 428
+    },
+    {
+      "epoch": 0.031565877213152446,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.989380971691725e-05,
+      "loss": 0.834,
+      "step": 429
+    },
+    {
+      "epoch": 0.03163945734651644,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.989327580749653e-05,
+      "loss": 0.6176,
+      "step": 430
+    },
+    {
+      "epoch": 0.031713037479880433,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.989274056210071e-05,
+      "loss": 1.259,
+      "step": 431
+    },
+    {
+      "epoch": 0.031786617613244424,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.989220398075852e-05,
+      "loss": 0.9209,
+      "step": 432
+    },
+    {
+      "epoch": 0.031860197746608414,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.9891666063498756e-05,
+      "loss": 0.8342,
+      "step": 433
+    },
+    {
+      "epoch": 0.03193377787997241,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.989112681035029e-05,
+      "loss": 0.747,
+      "step": 434
+    },
+    {
+      "epoch": 0.0320073580133364,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.9890586221342064e-05,
+      "loss": 1.0977,
+      "step": 435
+    },
+    {
+      "epoch": 0.03208093814670039,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.989004429650308e-05,
+      "loss": 0.8333,
+      "step": 436
+    },
+    {
+      "epoch": 0.03215451828006438,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.9889501035862426e-05,
+      "loss": 1.3032,
+      "step": 437
+    },
+    {
+      "epoch": 0.03222809841342837,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.9888956439449274e-05,
+      "loss": 0.6674,
+      "step": 438
+    },
+    {
+      "epoch": 0.03230167854679237,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.9888410507292825e-05,
+      "loss": 0.7147,
+      "step": 439
+    },
+    {
+      "epoch": 0.03237525868015636,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.988786323942241e-05,
+      "loss": 0.9657,
+      "step": 440
+    },
+    {
+      "epoch": 0.03244883881352035,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.988731463586737e-05,
+      "loss": 1.1865,
+      "step": 441
+    },
+    {
+      "epoch": 0.03252241894688434,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.988676469665715e-05,
+      "loss": 0.7875,
+      "step": 442
+    },
+    {
+      "epoch": 0.032595999080248336,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.988621342182128e-05,
+      "loss": 1.2024,
+      "step": 443
+    },
+    {
+      "epoch": 0.032669579213612326,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.9885660811389347e-05,
+      "loss": 1.3675,
+      "step": 444
+    },
+    {
+      "epoch": 0.03274315934697632,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.988510686539099e-05,
+      "loss": 1.0933,
+      "step": 445
+    },
+    {
+      "epoch": 0.03281673948034031,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.9884551583855954e-05,
+      "loss": 0.7227,
+      "step": 446
+    },
+    {
+      "epoch": 0.0328903196137043,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.9883994966814034e-05,
+      "loss": 0.6531,
+      "step": 447
+    },
+    {
+      "epoch": 0.032963899747068294,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.98834370142951e-05,
+      "loss": 0.8958,
+      "step": 448
+    },
+    {
+      "epoch": 0.033037479880432284,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.9882877726329106e-05,
+      "loss": 0.6407,
+      "step": 449
+    },
+    {
+      "epoch": 0.033111060013796274,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.988231710294606e-05,
+      "loss": 0.8751,
+      "step": 450
+    },
+    {
+      "epoch": 0.033184640147160265,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.9881755144176044e-05,
+      "loss": 0.7451,
+      "step": 451
+    },
+    {
+      "epoch": 0.03325822028052426,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.988119185004923e-05,
+      "loss": 0.927,
+      "step": 452
+    },
+    {
+      "epoch": 0.03333180041388825,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.988062722059585e-05,
+      "loss": 1.1373,
+      "step": 453
+    },
+    {
+      "epoch": 0.03340538054725224,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.988006125584619e-05,
+      "loss": 1.0013,
+      "step": 454
+    },
+    {
+      "epoch": 0.03347896068061623,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.987949395583064e-05,
+      "loss": 0.8468,
+      "step": 455
+    },
+    {
+      "epoch": 0.03355254081398022,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.987892532057964e-05,
+      "loss": 0.9814,
+      "step": 456
+    },
+    {
+      "epoch": 0.03362612094734422,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.98783553501237e-05,
+      "loss": 1.2678,
+      "step": 457
+    },
+    {
+      "epoch": 0.03369970108070821,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.987778404449342e-05,
+      "loss": 0.8017,
+      "step": 458
+    },
+    {
+      "epoch": 0.0337732812140722,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.987721140371946e-05,
+      "loss": 1.1828,
+      "step": 459
+    },
+    {
+      "epoch": 0.03384686134743619,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.987663742783255e-05,
+      "loss": 1.1213,
+      "step": 460
+    },
+    {
+      "epoch": 0.03392044148080019,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.98760621168635e-05,
+      "loss": 0.8114,
+      "step": 461
+    },
+    {
+      "epoch": 0.03399402161416418,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.987548547084317e-05,
+      "loss": 0.9586,
+      "step": 462
+    },
+    {
+      "epoch": 0.03406760174752817,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.9874907489802526e-05,
+      "loss": 1.0378,
+      "step": 463
+    },
+    {
+      "epoch": 0.03414118188089216,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.987432817377258e-05,
+      "loss": 1.0166,
+      "step": 464
+    },
+    {
+      "epoch": 0.03421476201425615,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.987374752278441e-05,
+      "loss": 1.1897,
+      "step": 465
+    },
+    {
+      "epoch": 0.034288342147620145,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.987316553686921e-05,
+      "loss": 0.9125,
+      "step": 466
+    },
+    {
+      "epoch": 0.034361922280984135,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.9872582216058174e-05,
+      "loss": 0.8171,
+      "step": 467
+    },
+    {
+      "epoch": 0.034435502414348125,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.987199756038263e-05,
+      "loss": 1.4177,
+      "step": 468
+    },
+    {
+      "epoch": 0.034509082547712115,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.987141156987396e-05,
+      "loss": 1.0405,
+      "step": 469
+    },
+    {
+      "epoch": 0.03458266268107611,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.987082424456361e-05,
+      "loss": 1.4172,
+      "step": 470
+    },
+    {
+      "epoch": 0.0346562428144401,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.9870235584483096e-05,
+      "loss": 0.5733,
+      "step": 471
+    },
+    {
+      "epoch": 0.03472982294780409,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.9869645589664e-05,
+      "loss": 1.2018,
+      "step": 472
+    },
+    {
+      "epoch": 0.03480340308116808,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.9869054260138015e-05,
+      "loss": 0.7661,
+      "step": 473
+    },
+    {
+      "epoch": 0.03487698321453207,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.986846159593685e-05,
+      "loss": 0.9603,
+      "step": 474
+    },
+    {
+      "epoch": 0.03495056334789607,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.986786759709232e-05,
+      "loss": 1.0224,
+      "step": 475
+    },
+    {
+      "epoch": 0.03502414348126006,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.986727226363631e-05,
+      "loss": 1.3507,
+      "step": 476
+    },
+    {
+      "epoch": 0.03509772361462405,
+      "grad_norm": 1.265625,
+      "learning_rate": 4.986667559560075e-05,
+      "loss": 1.1848,
+      "step": 477
+    },
+    {
+      "epoch": 0.03517130374798804,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.98660775930177e-05,
+      "loss": 0.8361,
+      "step": 478
+    },
+    {
+      "epoch": 0.03524488388135204,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.986547825591922e-05,
+      "loss": 1.3683,
+      "step": 479
+    },
+    {
+      "epoch": 0.03531846401471603,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.986487758433748e-05,
+      "loss": 0.673,
+      "step": 480
+    },
+    {
+      "epoch": 0.03539204414808002,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.986427557830473e-05,
+      "loss": 0.9089,
+      "step": 481
+    },
+    {
+      "epoch": 0.03546562428144401,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.9863672237853274e-05,
+      "loss": 0.6803,
+      "step": 482
+    },
+    {
+      "epoch": 0.035539204414808005,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.986306756301548e-05,
+      "loss": 0.9858,
+      "step": 483
+    },
+    {
+      "epoch": 0.035612784548171995,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.986246155382381e-05,
+      "loss": 0.9134,
+      "step": 484
+    },
+    {
+      "epoch": 0.035686364681535986,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.98618542103108e-05,
+      "loss": 1.0081,
+      "step": 485
+    },
+    {
+      "epoch": 0.035759944814899976,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.986124553250903e-05,
+      "loss": 1.3159,
+      "step": 486
+    },
+    {
+      "epoch": 0.035833524948263966,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.986063552045116e-05,
+      "loss": 1.2957,
+      "step": 487
+    },
+    {
+      "epoch": 0.03590710508162796,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.9860024174169936e-05,
+      "loss": 0.8686,
+      "step": 488
+    },
+    {
+      "epoch": 0.03598068521499195,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.985941149369817e-05,
+      "loss": 0.875,
+      "step": 489
+    },
+    {
+      "epoch": 0.03605426534835594,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.9858797479068746e-05,
+      "loss": 0.9396,
+      "step": 490
+    },
+    {
+      "epoch": 0.036127845481719933,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.985818213031461e-05,
+      "loss": 0.9777,
+      "step": 491
+    },
+    {
+      "epoch": 0.03620142561508393,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.985756544746879e-05,
+      "loss": 1.4404,
+      "step": 492
+    },
+    {
+      "epoch": 0.03627500574844792,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.985694743056438e-05,
+      "loss": 0.8215,
+      "step": 493
+    },
+    {
+      "epoch": 0.03634858588181191,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.985632807963456e-05,
+      "loss": 1.0838,
+      "step": 494
+    },
+    {
+      "epoch": 0.0364221660151759,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.9855707394712546e-05,
+      "loss": 0.9544,
+      "step": 495
+    },
+    {
+      "epoch": 0.03649574614853989,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.9855085375831665e-05,
+      "loss": 1.1742,
+      "step": 496
+    },
+    {
+      "epoch": 0.03656932628190389,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.98544620230253e-05,
+      "loss": 1.3149,
+      "step": 497
+    },
+    {
+      "epoch": 0.03664290641526788,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.98538373363269e-05,
+      "loss": 0.9338,
+      "step": 498
+    },
+    {
+      "epoch": 0.03671648654863187,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.985321131576999e-05,
+      "loss": 1.1112,
+      "step": 499
+    },
+    {
+      "epoch": 0.03679006668199586,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.985258396138817e-05,
+      "loss": 1.1511,
+      "step": 500
+    },
+    {
+      "epoch": 0.036863646815359856,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.985195527321511e-05,
+      "loss": 0.9077,
+      "step": 501
+    },
+    {
+      "epoch": 0.036937226948723846,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.985132525128455e-05,
+      "loss": 0.801,
+      "step": 502
+    },
+    {
+      "epoch": 0.037010807082087836,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.9850693895630305e-05,
+      "loss": 1.2966,
+      "step": 503
+    },
+    {
+      "epoch": 0.037084387215451826,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.985006120628625e-05,
+      "loss": 0.6195,
+      "step": 504
+    },
+    {
+      "epoch": 0.03715796734881582,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.984942718328635e-05,
+      "loss": 1.0418,
+      "step": 505
+    },
+    {
+      "epoch": 0.037231547482179814,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.984879182666462e-05,
+      "loss": 1.0322,
+      "step": 506
+    },
+    {
+      "epoch": 0.037305127615543804,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.984815513645516e-05,
+      "loss": 1.0583,
+      "step": 507
+    },
+    {
+      "epoch": 0.037378707748907794,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.984751711269216e-05,
+      "loss": 0.9566,
+      "step": 508
+    },
+    {
+      "epoch": 0.037452287882271784,
+      "grad_norm": 1.3359375,
+      "learning_rate": 4.9846877755409836e-05,
+      "loss": 1.5125,
+      "step": 509
+    },
+    {
+      "epoch": 0.03752586801563578,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.984623706464252e-05,
+      "loss": 0.9958,
+      "step": 510
+    },
+    {
+      "epoch": 0.03759944814899977,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.984559504042459e-05,
+      "loss": 0.9692,
+      "step": 511
+    },
+    {
+      "epoch": 0.03767302828236376,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.984495168279049e-05,
+      "loss": 0.8629,
+      "step": 512
+    },
+    {
+      "epoch": 0.03774660841572775,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.984430699177477e-05,
+      "loss": 0.7285,
+      "step": 513
+    },
+    {
+      "epoch": 0.03782018854909174,
+      "grad_norm": 0.875,
+      "learning_rate": 4.984366096741201e-05,
+      "loss": 0.6445,
+      "step": 514
+    },
+    {
+      "epoch": 0.03789376868245574,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.984301360973689e-05,
+      "loss": 0.963,
+      "step": 515
+    },
+    {
+      "epoch": 0.03796734881581973,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.984236491878415e-05,
+      "loss": 0.9509,
+      "step": 516
+    },
+    {
+      "epoch": 0.03804092894918372,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.984171489458861e-05,
+      "loss": 0.7679,
+      "step": 517
+    },
+    {
+      "epoch": 0.03811450908254771,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.984106353718515e-05,
+      "loss": 1.0823,
+      "step": 518
+    },
+    {
+      "epoch": 0.03818808921591171,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.984041084660872e-05,
+      "loss": 0.7272,
+      "step": 519
+    },
+    {
+      "epoch": 0.0382616693492757,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.983975682289437e-05,
+      "loss": 0.9714,
+      "step": 520
+    },
+    {
+      "epoch": 0.03833524948263969,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.9839101466077173e-05,
+      "loss": 1.0269,
+      "step": 521
+    },
+    {
+      "epoch": 0.03840882961600368,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.983844477619233e-05,
+      "loss": 0.883,
+      "step": 522
+    },
+    {
+      "epoch": 0.03848240974936767,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.983778675327506e-05,
+      "loss": 0.7904,
+      "step": 523
+    },
+    {
+      "epoch": 0.038555989882731664,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.9837127397360684e-05,
+      "loss": 1.0432,
+      "step": 524
+    },
+    {
+      "epoch": 0.038629570016095655,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.98364667084846e-05,
+      "loss": 1.0571,
+      "step": 525
+    },
+    {
+      "epoch": 0.038703150149459645,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.983580468668225e-05,
+      "loss": 0.7834,
+      "step": 526
+    },
+    {
+      "epoch": 0.038776730282823635,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.9835141331989175e-05,
+      "loss": 0.9083,
+      "step": 527
+    },
+    {
+      "epoch": 0.03885031041618763,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.983447664444097e-05,
+      "loss": 1.0536,
+      "step": 528
+    },
+    {
+      "epoch": 0.03892389054955162,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.983381062407331e-05,
+      "loss": 1.0827,
+      "step": 529
+    },
+    {
+      "epoch": 0.03899747068291561,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.9833143270921936e-05,
+      "loss": 0.891,
+      "step": 530
+    },
+    {
+      "epoch": 0.0390710508162796,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.983247458502267e-05,
+      "loss": 0.8906,
+      "step": 531
+    },
+    {
+      "epoch": 0.0391446309496436,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.983180456641139e-05,
+      "loss": 0.7901,
+      "step": 532
+    },
+    {
+      "epoch": 0.03921821108300759,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.983113321512406e-05,
+      "loss": 0.8367,
+      "step": 533
+    },
+    {
+      "epoch": 0.03929179121637158,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.983046053119671e-05,
+      "loss": 1.4232,
+      "step": 534
+    },
+    {
+      "epoch": 0.03936537134973557,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.982978651466544e-05,
+      "loss": 0.9988,
+      "step": 535
+    },
+    {
+      "epoch": 0.03943895148309956,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.982911116556643e-05,
+      "loss": 1.3246,
+      "step": 536
+    },
+    {
+      "epoch": 0.03951253161646356,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.9828434483935915e-05,
+      "loss": 0.6563,
+      "step": 537
+    },
+    {
+      "epoch": 0.03958611174982755,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.9827756469810216e-05,
+      "loss": 1.0454,
+      "step": 538
+    },
+    {
+      "epoch": 0.03965969188319154,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.982707712322572e-05,
+      "loss": 0.6676,
+      "step": 539
+    },
+    {
+      "epoch": 0.03973327201655553,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.982639644421889e-05,
+      "loss": 0.8891,
+      "step": 540
+    },
+    {
+      "epoch": 0.039806852149919525,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.982571443282625e-05,
+      "loss": 0.9849,
+      "step": 541
+    },
+    {
+      "epoch": 0.039880432283283515,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.982503108908441e-05,
+      "loss": 0.9774,
+      "step": 542
+    },
+    {
+      "epoch": 0.039954012416647505,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.982434641303003e-05,
+      "loss": 0.8891,
+      "step": 543
+    },
+    {
+      "epoch": 0.040027592550011495,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.982366040469988e-05,
+      "loss": 0.7822,
+      "step": 544
+    },
+    {
+      "epoch": 0.040101172683375486,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.982297306413075e-05,
+      "loss": 1.4514,
+      "step": 545
+    },
+    {
+      "epoch": 0.04017475281673948,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.982228439135954e-05,
+      "loss": 0.6972,
+      "step": 546
+    },
+    {
+      "epoch": 0.04024833295010347,
+      "grad_norm": 1.125,
+      "learning_rate": 4.982159438642321e-05,
+      "loss": 0.8759,
+      "step": 547
+    },
+    {
+      "epoch": 0.04032191308346746,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.982090304935879e-05,
+      "loss": 0.6461,
+      "step": 548
+    },
+    {
+      "epoch": 0.04039549321683145,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.982021038020338e-05,
+      "loss": 0.6901,
+      "step": 549
+    },
+    {
+      "epoch": 0.04046907335019545,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.981951637899417e-05,
+      "loss": 1.034,
+      "step": 550
+    },
+    {
+      "epoch": 0.04054265348355944,
+      "grad_norm": 1.34375,
+      "learning_rate": 4.9818821045768384e-05,
+      "loss": 1.7883,
+      "step": 551
+    },
+    {
+      "epoch": 0.04061623361692343,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.981812438056335e-05,
+      "loss": 0.6118,
+      "step": 552
+    },
+    {
+      "epoch": 0.04068981375028742,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.9817426383416456e-05,
+      "loss": 1.4013,
+      "step": 553
+    },
+    {
+      "epoch": 0.04076339388365141,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.981672705436516e-05,
+      "loss": 0.8169,
+      "step": 554
+    },
+    {
+      "epoch": 0.04083697401701541,
+      "grad_norm": 1.3046875,
+      "learning_rate": 4.981602639344699e-05,
+      "loss": 1.2825,
+      "step": 555
+    },
+    {
+      "epoch": 0.0409105541503794,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.9815324400699564e-05,
+      "loss": 0.9921,
+      "step": 556
+    },
+    {
+      "epoch": 0.04098413428374339,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.981462107616054e-05,
+      "loss": 0.7899,
+      "step": 557
+    },
+    {
+      "epoch": 0.04105771441710738,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.981391641986768e-05,
+      "loss": 0.8169,
+      "step": 558
+    },
+    {
+      "epoch": 0.041131294550471376,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.981321043185878e-05,
+      "loss": 0.7177,
+      "step": 559
+    },
+    {
+      "epoch": 0.041204874683835366,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.9812503112171746e-05,
+      "loss": 0.6038,
+      "step": 560
+    },
+    {
+      "epoch": 0.041278454817199356,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.981179446084454e-05,
+      "loss": 0.6604,
+      "step": 561
+    },
+    {
+      "epoch": 0.041352034950563346,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.9811084477915184e-05,
+      "loss": 0.7858,
+      "step": 562
+    },
+    {
+      "epoch": 0.041425615083927336,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.9810373163421786e-05,
+      "loss": 0.6443,
+      "step": 563
+    },
+    {
+      "epoch": 0.04149919521729133,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.980966051740252e-05,
+      "loss": 0.6786,
+      "step": 564
+    },
+    {
+      "epoch": 0.041572775350655324,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.980894653989563e-05,
+      "loss": 1.1585,
+      "step": 565
+    },
+    {
+      "epoch": 0.041646355484019314,
+      "grad_norm": 1.0,
+      "learning_rate": 4.980823123093945e-05,
+      "loss": 0.679,
+      "step": 566
+    },
+    {
+      "epoch": 0.041719935617383304,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.980751459057234e-05,
+      "loss": 1.1809,
+      "step": 567
+    },
+    {
+      "epoch": 0.0417935157507473,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.9806796618832786e-05,
+      "loss": 0.8619,
+      "step": 568
+    },
+    {
+      "epoch": 0.04186709588411129,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.98060773157593e-05,
+      "loss": 0.9854,
+      "step": 569
+    },
+    {
+      "epoch": 0.04194067601747528,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.980535668139051e-05,
+      "loss": 0.9795,
+      "step": 570
+    },
+    {
+      "epoch": 0.04201425615083927,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.980463471576507e-05,
+      "loss": 1.1124,
+      "step": 571
+    },
+    {
+      "epoch": 0.04208783628420327,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.9803911418921736e-05,
+      "loss": 0.9321,
+      "step": 572
+    },
+    {
+      "epoch": 0.04216141641756726,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.980318679089933e-05,
+      "loss": 0.8999,
+      "step": 573
+    },
+    {
+      "epoch": 0.04223499655093125,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.980246083173672e-05,
+      "loss": 0.7831,
+      "step": 574
+    },
+    {
+      "epoch": 0.04230857668429524,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.98017335414729e-05,
+      "loss": 0.9732,
+      "step": 575
+    },
+    {
+      "epoch": 0.04238215681765923,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.980100492014687e-05,
+      "loss": 0.7804,
+      "step": 576
+    },
+    {
+      "epoch": 0.042455736951023226,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.9800274967797755e-05,
+      "loss": 1.0128,
+      "step": 577
+    },
+    {
+      "epoch": 0.042529317084387216,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.979954368446472e-05,
+      "loss": 1.0641,
+      "step": 578
+    },
+    {
+      "epoch": 0.04260289721775121,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.9798811070187025e-05,
+      "loss": 1.2838,
+      "step": 579
+    },
+    {
+      "epoch": 0.0426764773511152,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.979807712500397e-05,
+      "loss": 0.8511,
+      "step": 580
+    },
+    {
+      "epoch": 0.042750057484479194,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.979734184895495e-05,
+      "loss": 0.8227,
+      "step": 581
+    },
+    {
+      "epoch": 0.042823637617843184,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.979660524207943e-05,
+      "loss": 0.6334,
+      "step": 582
+    },
+    {
+      "epoch": 0.042897217751207174,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.9795867304416945e-05,
+      "loss": 1.0858,
+      "step": 583
+    },
+    {
+      "epoch": 0.042970797884571164,
+      "grad_norm": 1.0,
+      "learning_rate": 4.9795128036007096e-05,
+      "loss": 0.9474,
+      "step": 584
+    },
+    {
+      "epoch": 0.043044378017935155,
+      "grad_norm": 0.75,
+      "learning_rate": 4.9794387436889555e-05,
+      "loss": 0.5507,
+      "step": 585
+    },
+    {
+      "epoch": 0.04311795815129915,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.979364550710407e-05,
+      "loss": 0.7823,
+      "step": 586
+    },
+    {
+      "epoch": 0.04319153828466314,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.979290224669045e-05,
+      "loss": 0.8691,
+      "step": 587
+    },
+    {
+      "epoch": 0.04326511841802713,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.97921576556886e-05,
+      "loss": 0.795,
+      "step": 588
+    },
+    {
+      "epoch": 0.04333869855139112,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.979141173413848e-05,
+      "loss": 0.9836,
+      "step": 589
+    },
+    {
+      "epoch": 0.04341227868475512,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.9790664482080114e-05,
+      "loss": 1.2645,
+      "step": 590
+    },
+    {
+      "epoch": 0.04348585881811911,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.978991589955361e-05,
+      "loss": 0.7881,
+      "step": 591
+    },
+    {
+      "epoch": 0.0435594389514831,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.978916598659913e-05,
+      "loss": 1.2378,
+      "step": 592
+    },
+    {
+      "epoch": 0.04363301908484709,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.978841474325694e-05,
+      "loss": 1.2396,
+      "step": 593
+    },
+    {
+      "epoch": 0.04370659921821108,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.978766216956735e-05,
+      "loss": 0.9337,
+      "step": 594
+    },
+    {
+      "epoch": 0.04378017935157508,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.9786908265570746e-05,
+      "loss": 0.7447,
+      "step": 595
+    },
+    {
+      "epoch": 0.04385375948493907,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.978615303130759e-05,
+      "loss": 0.862,
+      "step": 596
+    },
+    {
+      "epoch": 0.04392733961830306,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.9785396466818426e-05,
+      "loss": 1.2706,
+      "step": 597
+    },
+    {
+      "epoch": 0.04400091975166705,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.9784638572143835e-05,
+      "loss": 1.3343,
+      "step": 598
+    },
+    {
+      "epoch": 0.044074499885031045,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.9783879347324503e-05,
+      "loss": 1.4976,
+      "step": 599
+    },
+    {
+      "epoch": 0.044148080018395035,
+      "grad_norm": 0.875,
+      "learning_rate": 4.978311879240118e-05,
+      "loss": 0.8082,
+      "step": 600
+    },
+    {
+      "epoch": 0.044221660151759025,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.9782356907414686e-05,
+      "loss": 0.9675,
+      "step": 601
+    },
+    {
+      "epoch": 0.044295240285123015,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.9781593692405896e-05,
+      "loss": 0.8128,
+      "step": 602
+    },
+    {
+      "epoch": 0.044368820418487005,
+      "grad_norm": 1.125,
+      "learning_rate": 4.978082914741577e-05,
+      "loss": 1.0967,
+      "step": 603
+    },
+    {
+      "epoch": 0.044442400551851,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.978006327248537e-05,
+      "loss": 0.8955,
+      "step": 604
+    },
+    {
+      "epoch": 0.04451598068521499,
+      "grad_norm": 2.078125,
+      "learning_rate": 4.977929606765576e-05,
+      "loss": 1.1202,
+      "step": 605
+    },
+    {
+      "epoch": 0.04458956081857898,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.977852753296814e-05,
+      "loss": 0.8542,
+      "step": 606
+    },
+    {
+      "epoch": 0.04466314095194297,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.9777757668463744e-05,
+      "loss": 0.7782,
+      "step": 607
+    },
+    {
+      "epoch": 0.04473672108530697,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.9776986474183894e-05,
+      "loss": 0.8192,
+      "step": 608
+    },
+    {
+      "epoch": 0.04481030121867096,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.977621395016998e-05,
+      "loss": 1.1688,
+      "step": 609
+    },
+    {
+      "epoch": 0.04488388135203495,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.9775440096463447e-05,
+      "loss": 0.7741,
+      "step": 610
+    },
+    {
+      "epoch": 0.04495746148539894,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.977466491310585e-05,
+      "loss": 1.3333,
+      "step": 611
+    },
+    {
+      "epoch": 0.04503104161876293,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.977388840013877e-05,
+      "loss": 0.7828,
+      "step": 612
+    },
+    {
+      "epoch": 0.04510462175212693,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.97731105576039e-05,
+      "loss": 0.9337,
+      "step": 613
+    },
+    {
+      "epoch": 0.04517820188549092,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.977233138554297e-05,
+      "loss": 0.9152,
+      "step": 614
+    },
+    {
+      "epoch": 0.04525178201885491,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.977155088399781e-05,
+      "loss": 0.6864,
+      "step": 615
+    },
+    {
+      "epoch": 0.0453253621522189,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.977076905301029e-05,
+      "loss": 0.883,
+      "step": 616
+    },
+    {
+      "epoch": 0.045398942285582895,
+      "grad_norm": 1.0,
+      "learning_rate": 4.9769985892622393e-05,
+      "loss": 0.9834,
+      "step": 617
+    },
+    {
+      "epoch": 0.045472522418946885,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.976920140287613e-05,
+      "loss": 0.9779,
+      "step": 618
+    },
+    {
+      "epoch": 0.045546102552310876,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.976841558381361e-05,
+      "loss": 0.8692,
+      "step": 619
+    },
+    {
+      "epoch": 0.045619682685674866,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.976762843547701e-05,
+      "loss": 1.6504,
+      "step": 620
+    },
+    {
+      "epoch": 0.04569326281903886,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.976683995790856e-05,
+      "loss": 0.8912,
+      "step": 621
+    },
+    {
+      "epoch": 0.04576684295240285,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.97660501511506e-05,
+      "loss": 1.1648,
+      "step": 622
+    },
+    {
+      "epoch": 0.04584042308576684,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.97652590152455e-05,
+      "loss": 0.9922,
+      "step": 623
+    },
+    {
+      "epoch": 0.04591400321913083,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.976446655023572e-05,
+      "loss": 0.9775,
+      "step": 624
+    },
+    {
+      "epoch": 0.045987583352494824,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.9763672756163804e-05,
+      "loss": 1.387,
+      "step": 625
+    },
+    {
+      "epoch": 0.04606116348585882,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.976287763307234e-05,
+      "loss": 0.9029,
+      "step": 626
+    },
+    {
+      "epoch": 0.04613474361922281,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.976208118100399e-05,
+      "loss": 0.7147,
+      "step": 627
+    },
+    {
+      "epoch": 0.0462083237525868,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.976128340000153e-05,
+      "loss": 1.1444,
+      "step": 628
+    },
+    {
+      "epoch": 0.04628190388595079,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.976048429010775e-05,
+      "loss": 0.9068,
+      "step": 629
+    },
+    {
+      "epoch": 0.04635548401931479,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.9759683851365545e-05,
+      "loss": 0.7561,
+      "step": 630
+    },
+    {
+      "epoch": 0.04642906415267878,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.975888208381787e-05,
+      "loss": 0.6867,
+      "step": 631
+    },
+    {
+      "epoch": 0.04650264428604277,
+      "grad_norm": 1.375,
+      "learning_rate": 4.975807898750776e-05,
+      "loss": 1.1213,
+      "step": 632
+    },
+    {
+      "epoch": 0.04657622441940676,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.975727456247831e-05,
+      "loss": 0.7334,
+      "step": 633
+    },
+    {
+      "epoch": 0.04664980455277075,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.9756468808772696e-05,
+      "loss": 0.6628,
+      "step": 634
+    },
+    {
+      "epoch": 0.046723384686134746,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.975566172643415e-05,
+      "loss": 0.7996,
+      "step": 635
+    },
+    {
+      "epoch": 0.046796964819498736,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.975485331550601e-05,
+      "loss": 0.6378,
+      "step": 636
+    },
+    {
+      "epoch": 0.046870544952862726,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.9754043576031636e-05,
+      "loss": 0.9027,
+      "step": 637
+    },
+    {
+      "epoch": 0.046944125086226716,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.9753232508054506e-05,
+      "loss": 1.0751,
+      "step": 638
+    },
+    {
+      "epoch": 0.047017705219590714,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.9752420111618136e-05,
+      "loss": 0.7651,
+      "step": 639
+    },
+    {
+      "epoch": 0.047091285352954704,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.975160638676612e-05,
+      "loss": 1.1114,
+      "step": 640
+    },
+    {
+      "epoch": 0.047164865486318694,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.9750791333542154e-05,
+      "loss": 1.0256,
+      "step": 641
+    },
+    {
+      "epoch": 0.047238445619682684,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.974997495198996e-05,
+      "loss": 0.6675,
+      "step": 642
+    },
+    {
+      "epoch": 0.047312025753046674,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.9749157242153354e-05,
+      "loss": 1.1,
+      "step": 643
+    },
+    {
+      "epoch": 0.04738560588641067,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.974833820407622e-05,
+      "loss": 1.2962,
+      "step": 644
+    },
+    {
+      "epoch": 0.04745918601977466,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.974751783780253e-05,
+      "loss": 1.1137,
+      "step": 645
+    },
+    {
+      "epoch": 0.04753276615313865,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.974669614337628e-05,
+      "loss": 0.8745,
+      "step": 646
+    },
+    {
+      "epoch": 0.04760634628650264,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.9745873120841603e-05,
+      "loss": 1.1793,
+      "step": 647
+    },
+    {
+      "epoch": 0.04767992641986664,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.9745048770242645e-05,
+      "loss": 1.2452,
+      "step": 648
+    },
+    {
+      "epoch": 0.04775350655323063,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.974422309162366e-05,
+      "loss": 1.0531,
+      "step": 649
+    },
+    {
+      "epoch": 0.04782708668659462,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.974339608502896e-05,
+      "loss": 0.6883,
+      "step": 650
+    },
+    {
+      "epoch": 0.04790066681995861,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.974256775050292e-05,
+      "loss": 1.2653,
+      "step": 651
+    },
+    {
+      "epoch": 0.0479742469533226,
+      "grad_norm": 1.125,
+      "learning_rate": 4.9741738088090006e-05,
+      "loss": 1.1366,
+      "step": 652
+    },
+    {
+      "epoch": 0.0480478270866866,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.974090709783474e-05,
+      "loss": 1.3409,
+      "step": 653
+    },
+    {
+      "epoch": 0.04812140722005059,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.974007477978171e-05,
+      "loss": 1.1575,
+      "step": 654
+    },
+    {
+      "epoch": 0.04819498735341458,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.97392411339756e-05,
+      "loss": 1.6028,
+      "step": 655
+    },
+    {
+      "epoch": 0.04826856748677857,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.973840616046115e-05,
+      "loss": 0.7437,
+      "step": 656
+    },
+    {
+      "epoch": 0.048342147620142564,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.973756985928316e-05,
+      "loss": 0.8768,
+      "step": 657
+    },
+    {
+      "epoch": 0.048415727753506554,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.973673223048652e-05,
+      "loss": 0.885,
+      "step": 658
+    },
+    {
+      "epoch": 0.048489307886870545,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.973589327411617e-05,
+      "loss": 1.4515,
+      "step": 659
+    },
+    {
+      "epoch": 0.048562888020234535,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.9735052990217165e-05,
+      "loss": 1.393,
+      "step": 660
+    },
+    {
+      "epoch": 0.048636468153598525,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.973421137883458e-05,
+      "loss": 1.3093,
+      "step": 661
+    },
+    {
+      "epoch": 0.04871004828696252,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.9733368440013585e-05,
+      "loss": 0.8184,
+      "step": 662
+    },
+    {
+      "epoch": 0.04878362842032651,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.973252417379941e-05,
+      "loss": 0.8445,
+      "step": 663
+    },
+    {
+      "epoch": 0.0488572085536905,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.9731678580237386e-05,
+      "loss": 0.6878,
+      "step": 664
+    },
+    {
+      "epoch": 0.04893078868705449,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.9730831659372886e-05,
+      "loss": 0.7854,
+      "step": 665
+    },
+    {
+      "epoch": 0.04900436882041849,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.9729983411251355e-05,
+      "loss": 0.8962,
+      "step": 666
+    },
+    {
+      "epoch": 0.04907794895378248,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.972913383591833e-05,
+      "loss": 1.0992,
+      "step": 667
+    },
+    {
+      "epoch": 0.04915152908714647,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.972828293341939e-05,
+      "loss": 0.7466,
+      "step": 668
+    },
+    {
+      "epoch": 0.04922510922051046,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.9727430703800214e-05,
+      "loss": 1.155,
+      "step": 669
+    },
+    {
+      "epoch": 0.04929868935387446,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.972657714710653e-05,
+      "loss": 0.9673,
+      "step": 670
+    },
+    {
+      "epoch": 0.04937226948723845,
+      "grad_norm": 1.3046875,
+      "learning_rate": 4.972572226338416e-05,
+      "loss": 1.392,
+      "step": 671
+    },
+    {
+      "epoch": 0.04944584962060244,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.9724866052678974e-05,
+      "loss": 1.004,
+      "step": 672
+    },
+    {
+      "epoch": 0.04951942975396643,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.972400851503693e-05,
+      "loss": 0.6966,
+      "step": 673
+    },
+    {
+      "epoch": 0.04959300988733042,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.972314965050404e-05,
+      "loss": 1.0663,
+      "step": 674
+    },
+    {
+      "epoch": 0.049666590020694415,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.97222894591264e-05,
+      "loss": 0.9235,
+      "step": 675
+    },
+    {
+      "epoch": 0.049740170154058405,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.972142794095019e-05,
+      "loss": 0.9954,
+      "step": 676
+    },
+    {
+      "epoch": 0.049813750287422395,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.972056509602163e-05,
+      "loss": 1.1419,
+      "step": 677
+    },
+    {
+      "epoch": 0.049887330420786385,
+      "grad_norm": 1.125,
+      "learning_rate": 4.971970092438702e-05,
+      "loss": 0.7516,
+      "step": 678
+    },
+    {
+      "epoch": 0.04996091055415038,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.9718835426092766e-05,
+      "loss": 1.015,
+      "step": 679
+    },
+    {
+      "epoch": 0.05003449068751437,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.97179686011853e-05,
+      "loss": 0.5791,
+      "step": 680
+    },
+    {
+      "epoch": 0.05010807082087836,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.971710044971114e-05,
+      "loss": 1.0074,
+      "step": 681
+    },
+    {
+      "epoch": 0.05018165095424235,
+      "grad_norm": 1.265625,
+      "learning_rate": 4.971623097171688e-05,
+      "loss": 0.9841,
+      "step": 682
+    },
+    {
+      "epoch": 0.05025523108760634,
+      "grad_norm": 1.125,
+      "learning_rate": 4.971536016724919e-05,
+      "loss": 0.8017,
+      "step": 683
+    },
+    {
+      "epoch": 0.05032881122097034,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.9714488036354803e-05,
+      "loss": 1.0498,
+      "step": 684
+    },
+    {
+      "epoch": 0.05040239135433433,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.971361457908053e-05,
+      "loss": 1.3636,
+      "step": 685
+    },
+    {
+      "epoch": 0.05047597148769832,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.971273979547322e-05,
+      "loss": 0.8792,
+      "step": 686
+    },
+    {
+      "epoch": 0.05054955162106231,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.9711863685579855e-05,
+      "loss": 0.9422,
+      "step": 687
+    },
+    {
+      "epoch": 0.05062313175442631,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.9710986249447436e-05,
+      "loss": 0.8948,
+      "step": 688
+    },
+    {
+      "epoch": 0.0506967118877903,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.9710107487123054e-05,
+      "loss": 0.8737,
+      "step": 689
+    },
+    {
+      "epoch": 0.05077029202115429,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.970922739865388e-05,
+      "loss": 1.5617,
+      "step": 690
+    },
+    {
+      "epoch": 0.05084387215451828,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.9708345984087137e-05,
+      "loss": 0.7197,
+      "step": 691
+    },
+    {
+      "epoch": 0.05091745228788227,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.970746324347013e-05,
+      "loss": 1.0125,
+      "step": 692
+    },
+    {
+      "epoch": 0.050991032421246266,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.9706579176850246e-05,
+      "loss": 0.742,
+      "step": 693
+    },
+    {
+      "epoch": 0.051064612554610256,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.970569378427491e-05,
+      "loss": 0.6562,
+      "step": 694
+    },
+    {
+      "epoch": 0.051138192687974246,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.9704807065791656e-05,
+      "loss": 1.3955,
+      "step": 695
+    },
+    {
+      "epoch": 0.051211772821338236,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.970391902144806e-05,
+      "loss": 0.8316,
+      "step": 696
+    },
+    {
+      "epoch": 0.05128535295470223,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.9703029651291806e-05,
+      "loss": 0.9463,
+      "step": 697
+    },
+    {
+      "epoch": 0.05135893308806622,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.97021389553706e-05,
+      "loss": 1.069,
+      "step": 698
+    },
+    {
+      "epoch": 0.051432513221430214,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.970124693373225e-05,
+      "loss": 0.939,
+      "step": 699
+    },
+    {
+      "epoch": 0.051506093354794204,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.9700353586424634e-05,
+      "loss": 0.7969,
+      "step": 700
+    },
+    {
+      "epoch": 0.051579673488158194,
+      "grad_norm": 1.25,
+      "learning_rate": 4.969945891349569e-05,
+      "loss": 1.1327,
+      "step": 701
+    },
+    {
+      "epoch": 0.05165325362152219,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.969856291499344e-05,
+      "loss": 1.6125,
+      "step": 702
+    },
+    {
+      "epoch": 0.05172683375488618,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.9697665590965964e-05,
+      "loss": 0.6791,
+      "step": 703
+    },
+    {
+      "epoch": 0.05180041388825017,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.969676694146143e-05,
+      "loss": 0.8961,
+      "step": 704
+    },
+    {
+      "epoch": 0.05187399402161416,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.9695866966528046e-05,
+      "loss": 0.89,
+      "step": 705
+    },
+    {
+      "epoch": 0.05194757415497816,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.9694965666214136e-05,
+      "loss": 1.1385,
+      "step": 706
+    },
+    {
+      "epoch": 0.05202115428834215,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.969406304056806e-05,
+      "loss": 0.8156,
+      "step": 707
+    },
+    {
+      "epoch": 0.05209473442170614,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.969315908963826e-05,
+      "loss": 0.7964,
+      "step": 708
+    },
+    {
+      "epoch": 0.05216831455507013,
+      "grad_norm": 1.0,
+      "learning_rate": 4.969225381347325e-05,
+      "loss": 1.0184,
+      "step": 709
+    },
+    {
+      "epoch": 0.052241894688434126,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.969134721212161e-05,
+      "loss": 0.986,
+      "step": 710
+    },
+    {
+      "epoch": 0.052315474821798116,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.9690439285632006e-05,
+      "loss": 1.3072,
+      "step": 711
+    },
+    {
+      "epoch": 0.052389054955162107,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.9689530034053165e-05,
+      "loss": 1.2645,
+      "step": 712
+    },
+    {
+      "epoch": 0.0524626350885261,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.968861945743387e-05,
+      "loss": 1.2966,
+      "step": 713
+    },
+    {
+      "epoch": 0.05253621522189009,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.968770755582301e-05,
+      "loss": 0.8053,
+      "step": 714
+    },
+    {
+      "epoch": 0.052609795355254084,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.96867943292695e-05,
+      "loss": 0.9755,
+      "step": 715
+    },
+    {
+      "epoch": 0.052683375488618074,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.968587977782237e-05,
+      "loss": 0.8545,
+      "step": 716
+    },
+    {
+      "epoch": 0.052756955621982064,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.96849639015307e-05,
+      "loss": 0.998,
+      "step": 717
+    },
+    {
+      "epoch": 0.052830535755346054,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.9684046700443636e-05,
+      "loss": 1.0482,
+      "step": 718
+    },
+    {
+      "epoch": 0.05290411588871005,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.968312817461041e-05,
+      "loss": 1.1994,
+      "step": 719
+    },
+    {
+      "epoch": 0.05297769602207404,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.968220832408031e-05,
+      "loss": 0.9737,
+      "step": 720
+    },
+    {
+      "epoch": 0.05305127615543803,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.968128714890272e-05,
+      "loss": 1.1198,
+      "step": 721
+    },
+    {
+      "epoch": 0.05312485628880202,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.9680364649127054e-05,
+      "loss": 0.7691,
+      "step": 722
+    },
+    {
+      "epoch": 0.05319843642216601,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.967944082480284e-05,
+      "loss": 1.0819,
+      "step": 723
+    },
+    {
+      "epoch": 0.05327201655553001,
+      "grad_norm": 1.421875,
+      "learning_rate": 4.967851567597964e-05,
+      "loss": 1.0186,
+      "step": 724
+    },
+    {
+      "epoch": 0.053345596688894,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.967758920270712e-05,
+      "loss": 1.0298,
+      "step": 725
+    },
+    {
+      "epoch": 0.05341917682225799,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.9676661405035e-05,
+      "loss": 0.9255,
+      "step": 726
+    },
+    {
+      "epoch": 0.05349275695562198,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.9675732283013064e-05,
+      "loss": 1.2375,
+      "step": 727
+    },
+    {
+      "epoch": 0.05356633708898598,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.967480183669118e-05,
+      "loss": 1.0639,
+      "step": 728
+    },
+    {
+      "epoch": 0.05363991722234997,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.9673870066119294e-05,
+      "loss": 1.2414,
+      "step": 729
+    },
+    {
+      "epoch": 0.05371349735571396,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.9672936971347394e-05,
+      "loss": 1.0286,
+      "step": 730
+    },
+    {
+      "epoch": 0.05378707748907795,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.967200255242558e-05,
+      "loss": 1.2123,
+      "step": 731
+    },
+    {
+      "epoch": 0.05386065762244194,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.9671066809403976e-05,
+      "loss": 0.9672,
+      "step": 732
+    },
+    {
+      "epoch": 0.053934237755805935,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.967012974233282e-05,
+      "loss": 0.9823,
+      "step": 733
+    },
+    {
+      "epoch": 0.054007817889169925,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.966919135126239e-05,
+      "loss": 0.7568,
+      "step": 734
+    },
+    {
+      "epoch": 0.054081398022533915,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.9668251636243065e-05,
+      "loss": 1.4507,
+      "step": 735
+    },
+    {
+      "epoch": 0.054154978155897905,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.9667310597325255e-05,
+      "loss": 1.4083,
+      "step": 736
+    },
+    {
+      "epoch": 0.0542285582892619,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.966636823455948e-05,
+      "loss": 0.9924,
+      "step": 737
+    },
+    {
+      "epoch": 0.05430213842262589,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.9665424547996306e-05,
+      "loss": 0.9729,
+      "step": 738
+    },
+    {
+      "epoch": 0.05437571855598988,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.966447953768639e-05,
+      "loss": 0.9148,
+      "step": 739
+    },
+    {
+      "epoch": 0.05444929868935387,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.966353320368044e-05,
+      "loss": 1.2248,
+      "step": 740
+    },
+    {
+      "epoch": 0.05452287882271786,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.9662585546029246e-05,
+      "loss": 1.2476,
+      "step": 741
+    },
+    {
+      "epoch": 0.05459645895608186,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.9661636564783664e-05,
+      "loss": 0.703,
+      "step": 742
+    },
+    {
+      "epoch": 0.05467003908944585,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.966068625999463e-05,
+      "loss": 0.9597,
+      "step": 743
+    },
+    {
+      "epoch": 0.05474361922280984,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.965973463171314e-05,
+      "loss": 0.8979,
+      "step": 744
+    },
+    {
+      "epoch": 0.05481719935617383,
+      "grad_norm": 0.875,
+      "learning_rate": 4.9658781679990265e-05,
+      "loss": 0.8228,
+      "step": 745
+    },
+    {
+      "epoch": 0.05489077948953783,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.965782740487715e-05,
+      "loss": 0.7687,
+      "step": 746
+    },
+    {
+      "epoch": 0.05496435962290182,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.965687180642501e-05,
+      "loss": 0.897,
+      "step": 747
+    },
+    {
+      "epoch": 0.05503793975626581,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.965591488468514e-05,
+      "loss": 0.9762,
+      "step": 748
+    },
+    {
+      "epoch": 0.0551115198896298,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.9654956639708885e-05,
+      "loss": 1.0854,
+      "step": 749
+    },
+    {
+      "epoch": 0.05518510002299379,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.965399707154766e-05,
+      "loss": 0.8447,
+      "step": 750
+    },
+    {
+      "epoch": 0.055258680156357785,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.965303618025299e-05,
+      "loss": 1.0321,
+      "step": 751
+    },
+    {
+      "epoch": 0.055332260289721776,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.9652073965876425e-05,
+      "loss": 0.9771,
+      "step": 752
+    },
+    {
+      "epoch": 0.055405840423085766,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.9651110428469614e-05,
+      "loss": 0.6425,
+      "step": 753
+    },
+    {
+      "epoch": 0.055479420556449756,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.9650145568084266e-05,
+      "loss": 0.84,
+      "step": 754
+    },
+    {
+      "epoch": 0.05555300068981375,
+      "grad_norm": 1.0,
+      "learning_rate": 4.964917938477216e-05,
+      "loss": 1.1714,
+      "step": 755
+    },
+    {
+      "epoch": 0.05562658082317774,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.964821187858516e-05,
+      "loss": 0.7757,
+      "step": 756
+    },
+    {
+      "epoch": 0.05570016095654173,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.964724304957518e-05,
+      "loss": 0.9814,
+      "step": 757
+    },
+    {
+      "epoch": 0.05577374108990572,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.964627289779421e-05,
+      "loss": 0.7757,
+      "step": 758
+    },
+    {
+      "epoch": 0.05584732122326972,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.964530142329433e-05,
+      "loss": 0.6509,
+      "step": 759
+    },
+    {
+      "epoch": 0.05592090135663371,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.964432862612767e-05,
+      "loss": 0.8458,
+      "step": 760
+    },
+    {
+      "epoch": 0.0559944814899977,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.964335450634644e-05,
+      "loss": 1.7253,
+      "step": 761
+    },
+    {
+      "epoch": 0.05606806162336169,
+      "grad_norm": 1.3203125,
+      "learning_rate": 4.964237906400291e-05,
+      "loss": 1.1215,
+      "step": 762
+    },
+    {
+      "epoch": 0.05614164175672568,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.964140229914944e-05,
+      "loss": 0.8251,
+      "step": 763
+    },
+    {
+      "epoch": 0.05621522189008968,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.964042421183846e-05,
+      "loss": 1.0398,
+      "step": 764
+    },
+    {
+      "epoch": 0.05628880202345367,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.9639444802122445e-05,
+      "loss": 0.9634,
+      "step": 765
+    },
+    {
+      "epoch": 0.05636238215681766,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.963846407005397e-05,
+      "loss": 1.187,
+      "step": 766
+    },
+    {
+      "epoch": 0.05643596229018165,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.963748201568565e-05,
+      "loss": 0.861,
+      "step": 767
+    },
+    {
+      "epoch": 0.056509542423545646,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.963649863907022e-05,
+      "loss": 0.7786,
+      "step": 768
+    },
+    {
+      "epoch": 0.056583122556909636,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.9635513940260434e-05,
+      "loss": 1.0423,
+      "step": 769
+    },
+    {
+      "epoch": 0.056656702690273626,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.9634527919309145e-05,
+      "loss": 1.2678,
+      "step": 770
+    },
+    {
+      "epoch": 0.056730282823637616,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.9633540576269265e-05,
+      "loss": 0.9998,
+      "step": 771
+    },
+    {
+      "epoch": 0.05680386295700161,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.96325519111938e-05,
+      "loss": 0.6784,
+      "step": 772
+    },
+    {
+      "epoch": 0.056877443090365604,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.963156192413579e-05,
+      "loss": 0.9524,
+      "step": 773
+    },
+    {
+      "epoch": 0.056951023223729594,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.963057061514838e-05,
+      "loss": 1.2812,
+      "step": 774
+    },
+    {
+      "epoch": 0.057024603357093584,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.962957798428475e-05,
+      "loss": 0.8375,
+      "step": 775
+    },
+    {
+      "epoch": 0.057098183490457574,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.96285840315982e-05,
+      "loss": 0.6929,
+      "step": 776
+    },
+    {
+      "epoch": 0.05717176362382157,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.962758875714206e-05,
+      "loss": 0.6131,
+      "step": 777
+    },
+    {
+      "epoch": 0.05724534375718556,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.962659216096976e-05,
+      "loss": 1.079,
+      "step": 778
+    },
+    {
+      "epoch": 0.05731892389054955,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.9625594243134755e-05,
+      "loss": 0.8023,
+      "step": 779
+    },
+    {
+      "epoch": 0.05739250402391354,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.962459500369062e-05,
+      "loss": 0.6319,
+      "step": 780
+    },
+    {
+      "epoch": 0.05746608415727753,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.962359444269098e-05,
+      "loss": 0.9587,
+      "step": 781
+    },
+    {
+      "epoch": 0.05753966429064153,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.962259256018954e-05,
+      "loss": 1.3477,
+      "step": 782
+    },
+    {
+      "epoch": 0.05761324442400552,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.962158935624006e-05,
+      "loss": 1.0746,
+      "step": 783
+    },
+    {
+      "epoch": 0.05768682455736951,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.962058483089638e-05,
+      "loss": 0.9957,
+      "step": 784
+    },
+    {
+      "epoch": 0.0577604046907335,
+      "grad_norm": 1.40625,
+      "learning_rate": 4.9619578984212415e-05,
+      "loss": 1.3374,
+      "step": 785
+    },
+    {
+      "epoch": 0.0578339848240975,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.961857181624214e-05,
+      "loss": 1.3233,
+      "step": 786
+    },
+    {
+      "epoch": 0.05790756495746149,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.961756332703962e-05,
+      "loss": 1.5237,
+      "step": 787
+    },
+    {
+      "epoch": 0.05798114509082548,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.9616553516658974e-05,
+      "loss": 1.0261,
+      "step": 788
+    },
+    {
+      "epoch": 0.05805472522418947,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.961554238515439e-05,
+      "loss": 0.7153,
+      "step": 789
+    },
+    {
+      "epoch": 0.05812830535755346,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.961452993258015e-05,
+      "loss": 1.025,
+      "step": 790
+    },
+    {
+      "epoch": 0.058201885490917454,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.961351615899057e-05,
+      "loss": 1.0908,
+      "step": 791
+    },
+    {
+      "epoch": 0.058275465624281444,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.961250106444007e-05,
+      "loss": 1.0561,
+      "step": 792
+    },
+    {
+      "epoch": 0.058349045757645435,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.9611484648983106e-05,
+      "loss": 1.1467,
+      "step": 793
+    },
+    {
+      "epoch": 0.058422625891009425,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.961046691267427e-05,
+      "loss": 1.1308,
+      "step": 794
+    },
+    {
+      "epoch": 0.05849620602437342,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.960944785556814e-05,
+      "loss": 1.2358,
+      "step": 795
+    },
+    {
+      "epoch": 0.05856978615773741,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.960842747771943e-05,
+      "loss": 0.8898,
+      "step": 796
+    },
+    {
+      "epoch": 0.0586433662911014,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.96074057791829e-05,
+      "loss": 1.0806,
+      "step": 797
+    },
+    {
+      "epoch": 0.05871694642446539,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.9606382760013374e-05,
+      "loss": 0.9684,
+      "step": 798
+    },
+    {
+      "epoch": 0.05879052655782939,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.960535842026577e-05,
+      "loss": 1.5033,
+      "step": 799
+    },
+    {
+      "epoch": 0.05886410669119338,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.960433275999504e-05,
+      "loss": 0.816,
+      "step": 800
+    },
+    {
+      "epoch": 0.05893768682455737,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.960330577925625e-05,
+      "loss": 1.0746,
+      "step": 801
+    },
+    {
+      "epoch": 0.05901126695792136,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.9602277478104496e-05,
+      "loss": 1.2213,
+      "step": 802
+    },
+    {
+      "epoch": 0.05908484709128535,
+      "grad_norm": 1.0,
+      "learning_rate": 4.960124785659499e-05,
+      "loss": 1.298,
+      "step": 803
+    },
+    {
+      "epoch": 0.05915842722464935,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.960021691478297e-05,
+      "loss": 0.7752,
+      "step": 804
+    },
+    {
+      "epoch": 0.05923200735801334,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.959918465272377e-05,
+      "loss": 0.9438,
+      "step": 805
+    },
+    {
+      "epoch": 0.05930558749137733,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.95981510704728e-05,
+      "loss": 0.7657,
+      "step": 806
+    },
+    {
+      "epoch": 0.05937916762474132,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.959711616808551e-05,
+      "loss": 1.2213,
+      "step": 807
+    },
+    {
+      "epoch": 0.059452747758105315,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.959607994561746e-05,
+      "loss": 1.065,
+      "step": 808
+    },
+    {
+      "epoch": 0.059526327891469305,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.9595042403124254e-05,
+      "loss": 0.8997,
+      "step": 809
+    },
+    {
+      "epoch": 0.059599908024833295,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.959400354066157e-05,
+      "loss": 0.8904,
+      "step": 810
+    },
+    {
+      "epoch": 0.059673488158197285,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.959296335828517e-05,
+      "loss": 0.7214,
+      "step": 811
+    },
+    {
+      "epoch": 0.059747068291561276,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.959192185605088e-05,
+      "loss": 0.8603,
+      "step": 812
+    },
+    {
+      "epoch": 0.05982064842492527,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.959087903401459e-05,
+      "loss": 1.0214,
+      "step": 813
+    },
+    {
+      "epoch": 0.05989422855828926,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.958983489223227e-05,
+      "loss": 1.1437,
+      "step": 814
+    },
+    {
+      "epoch": 0.05996780869165325,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.958878943075995e-05,
+      "loss": 1.3009,
+      "step": 815
+    },
+    {
+      "epoch": 0.06004138882501724,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.9587742649653746e-05,
+      "loss": 1.1063,
+      "step": 816
+    },
+    {
+      "epoch": 0.06011496895838124,
+      "grad_norm": 1.8125,
+      "learning_rate": 4.958669454896983e-05,
+      "loss": 0.7659,
+      "step": 817
+    },
+    {
+      "epoch": 0.06018854909174523,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.9585645128764456e-05,
+      "loss": 1.0277,
+      "step": 818
+    },
+    {
+      "epoch": 0.06026212922510922,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.958459438909395e-05,
+      "loss": 0.6037,
+      "step": 819
+    },
+    {
+      "epoch": 0.06033570935847321,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.9583542330014694e-05,
+      "loss": 0.9885,
+      "step": 820
+    },
+    {
+      "epoch": 0.0604092894918372,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.958248895158315e-05,
+      "loss": 1.1207,
+      "step": 821
+    },
+    {
+      "epoch": 0.0604828696252012,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.9581434253855855e-05,
+      "loss": 0.9179,
+      "step": 822
+    },
+    {
+      "epoch": 0.06055644975856519,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.958037823688941e-05,
+      "loss": 1.0005,
+      "step": 823
+    },
+    {
+      "epoch": 0.06063002989192918,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.95793209007405e-05,
+      "loss": 1.3325,
+      "step": 824
+    },
+    {
+      "epoch": 0.06070361002529317,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.957826224546585e-05,
+      "loss": 1.1585,
+      "step": 825
+    },
+    {
+      "epoch": 0.060777190158657166,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.9577202271122286e-05,
+      "loss": 0.8979,
+      "step": 826
+    },
+    {
+      "epoch": 0.060850770292021156,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.95761409777667e-05,
+      "loss": 0.8202,
+      "step": 827
+    },
+    {
+      "epoch": 0.060924350425385146,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.9575078365456043e-05,
+      "loss": 1.1022,
+      "step": 828
+    },
+    {
+      "epoch": 0.060997930558749136,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.957401443424735e-05,
+      "loss": 0.763,
+      "step": 829
+    },
+    {
+      "epoch": 0.061071510692113126,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.957294918419772e-05,
+      "loss": 1.3115,
+      "step": 830
+    },
+    {
+      "epoch": 0.06114509082547712,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.957188261536431e-05,
+      "loss": 0.6945,
+      "step": 831
+    },
+    {
+      "epoch": 0.061218670958841113,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.9570814727804374e-05,
+      "loss": 1.0005,
+      "step": 832
+    },
+    {
+      "epoch": 0.061292251092205104,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.956974552157522e-05,
+      "loss": 0.7747,
+      "step": 833
+    },
+    {
+      "epoch": 0.061365831225569094,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.9568674996734224e-05,
+      "loss": 0.9603,
+      "step": 834
+    },
+    {
+      "epoch": 0.06143941135893309,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.9567603153338847e-05,
+      "loss": 1.3527,
+      "step": 835
+    },
+    {
+      "epoch": 0.06151299149229708,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.956652999144661e-05,
+      "loss": 0.7923,
+      "step": 836
+    },
+    {
+      "epoch": 0.06158657162566107,
+      "grad_norm": 1.0,
+      "learning_rate": 4.956545551111511e-05,
+      "loss": 1.0774,
+      "step": 837
+    },
+    {
+      "epoch": 0.06166015175902506,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.956437971240201e-05,
+      "loss": 0.9355,
+      "step": 838
+    },
+    {
+      "epoch": 0.06173373189238905,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.9563302595365045e-05,
+      "loss": 0.8825,
+      "step": 839
+    },
+    {
+      "epoch": 0.06180731202575305,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.9562224160062024e-05,
+      "loss": 0.7766,
+      "step": 840
+    },
+    {
+      "epoch": 0.06188089215911704,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.956114440655083e-05,
+      "loss": 0.9358,
+      "step": 841
+    },
+    {
+      "epoch": 0.06195447229248103,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.9560063334889396e-05,
+      "loss": 0.8375,
+      "step": 842
+    },
+    {
+      "epoch": 0.06202805242584502,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.955898094513576e-05,
+      "loss": 1.0116,
+      "step": 843
+    },
+    {
+      "epoch": 0.062101632559209016,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.955789723734799e-05,
+      "loss": 0.9917,
+      "step": 844
+    },
+    {
+      "epoch": 0.062175212692573006,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.955681221158426e-05,
+      "loss": 1.3232,
+      "step": 845
+    },
+    {
+      "epoch": 0.062248792825937,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.955572586790281e-05,
+      "loss": 0.8649,
+      "step": 846
+    },
+    {
+      "epoch": 0.06232237295930099,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.9554638206361924e-05,
+      "loss": 0.8878,
+      "step": 847
+    },
+    {
+      "epoch": 0.062395953092664984,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.955354922701998e-05,
+      "loss": 1.0983,
+      "step": 848
+    },
+    {
+      "epoch": 0.062469533226028974,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.955245892993543e-05,
+      "loss": 1.2533,
+      "step": 849
+    },
+    {
+      "epoch": 0.06254311335939296,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.955136731516679e-05,
+      "loss": 0.9845,
+      "step": 850
+    },
+    {
+      "epoch": 0.06261669349275696,
+      "grad_norm": 1.28125,
+      "learning_rate": 4.955027438277262e-05,
+      "loss": 1.934,
+      "step": 851
+    },
+    {
+      "epoch": 0.06269027362612095,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.9549180132811603e-05,
+      "loss": 0.8889,
+      "step": 852
+    },
+    {
+      "epoch": 0.06276385375948494,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.954808456534245e-05,
+      "loss": 0.9598,
+      "step": 853
+    },
+    {
+      "epoch": 0.06283743389284893,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.954698768042396e-05,
+      "loss": 1.1668,
+      "step": 854
+    },
+    {
+      "epoch": 0.06291101402621292,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.954588947811502e-05,
+      "loss": 0.716,
+      "step": 855
+    },
+    {
+      "epoch": 0.06298459415957691,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.9544789958474535e-05,
+      "loss": 0.8584,
+      "step": 856
+    },
+    {
+      "epoch": 0.0630581742929409,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.954368912156154e-05,
+      "loss": 0.6127,
+      "step": 857
+    },
+    {
+      "epoch": 0.06313175442630489,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.954258696743511e-05,
+      "loss": 1.0402,
+      "step": 858
+    },
+    {
+      "epoch": 0.06320533455966888,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.954148349615439e-05,
+      "loss": 0.9507,
+      "step": 859
+    },
+    {
+      "epoch": 0.06327891469303289,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.954037870777859e-05,
+      "loss": 1.3577,
+      "step": 860
+    },
+    {
+      "epoch": 0.06335249482639688,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.9539272602367025e-05,
+      "loss": 1.1217,
+      "step": 861
+    },
+    {
+      "epoch": 0.06342607495976087,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.953816517997905e-05,
+      "loss": 0.8124,
+      "step": 862
+    },
+    {
+      "epoch": 0.06349965509312486,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.953705644067409e-05,
+      "loss": 0.9313,
+      "step": 863
+    },
+    {
+      "epoch": 0.06357323522648885,
+      "grad_norm": 1.25,
+      "learning_rate": 4.953594638451166e-05,
+      "loss": 0.9996,
+      "step": 864
+    },
+    {
+      "epoch": 0.06364681535985284,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.953483501155133e-05,
+      "loss": 0.6392,
+      "step": 865
+    },
+    {
+      "epoch": 0.06372039549321683,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.9533722321852737e-05,
+      "loss": 1.9765,
+      "step": 866
+    },
+    {
+      "epoch": 0.06379397562658082,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.953260831547561e-05,
+      "loss": 1.0975,
+      "step": 867
+    },
+    {
+      "epoch": 0.06386755575994482,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.953149299247973e-05,
+      "loss": 0.8411,
+      "step": 868
+    },
+    {
+      "epoch": 0.06394113589330881,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.953037635292495e-05,
+      "loss": 0.6156,
+      "step": 869
+    },
+    {
+      "epoch": 0.0640147160266728,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.952925839687121e-05,
+      "loss": 0.6623,
+      "step": 870
+    },
+    {
+      "epoch": 0.06408829616003679,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.952813912437849e-05,
+      "loss": 1.2113,
+      "step": 871
+    },
+    {
+      "epoch": 0.06416187629340078,
+      "grad_norm": 1.3046875,
+      "learning_rate": 4.9527018535506885e-05,
+      "loss": 0.7784,
+      "step": 872
+    },
+    {
+      "epoch": 0.06423545642676477,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.952589663031651e-05,
+      "loss": 0.9496,
+      "step": 873
+    },
+    {
+      "epoch": 0.06430903656012876,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.9524773408867584e-05,
+      "loss": 0.7851,
+      "step": 874
+    },
+    {
+      "epoch": 0.06438261669349275,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.95236488712204e-05,
+      "loss": 0.578,
+      "step": 875
+    },
+    {
+      "epoch": 0.06445619682685674,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.9522523017435294e-05,
+      "loss": 1.0817,
+      "step": 876
+    },
+    {
+      "epoch": 0.06452977696022075,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.952139584757269e-05,
+      "loss": 1.3044,
+      "step": 877
+    },
+    {
+      "epoch": 0.06460335709358474,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.9520267361693095e-05,
+      "loss": 1.0066,
+      "step": 878
+    },
+    {
+      "epoch": 0.06467693722694873,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.951913755985705e-05,
+      "loss": 1.0606,
+      "step": 879
+    },
+    {
+      "epoch": 0.06475051736031272,
+      "grad_norm": 1.125,
+      "learning_rate": 4.951800644212522e-05,
+      "loss": 1.3807,
+      "step": 880
+    },
+    {
+      "epoch": 0.06482409749367671,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.951687400855828e-05,
+      "loss": 0.8976,
+      "step": 881
+    },
+    {
+      "epoch": 0.0648976776270407,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.951574025921701e-05,
+      "loss": 1.037,
+      "step": 882
+    },
+    {
+      "epoch": 0.06497125776040469,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.9514605194162275e-05,
+      "loss": 0.7599,
+      "step": 883
+    },
+    {
+      "epoch": 0.06504483789376868,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.951346881345498e-05,
+      "loss": 0.6583,
+      "step": 884
+    },
+    {
+      "epoch": 0.06511841802713267,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.951233111715611e-05,
+      "loss": 0.9101,
+      "step": 885
+    },
+    {
+      "epoch": 0.06519199816049667,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.951119210532673e-05,
+      "loss": 0.8401,
+      "step": 886
+    },
+    {
+      "epoch": 0.06526557829386066,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.951005177802796e-05,
+      "loss": 0.8865,
+      "step": 887
+    },
+    {
+      "epoch": 0.06533915842722465,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.950891013532101e-05,
+      "loss": 0.9345,
+      "step": 888
+    },
+    {
+      "epoch": 0.06541273856058864,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.950776717726715e-05,
+      "loss": 1.1009,
+      "step": 889
+    },
+    {
+      "epoch": 0.06548631869395263,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.95066229039277e-05,
+      "loss": 1.0859,
+      "step": 890
+    },
+    {
+      "epoch": 0.06555989882731662,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.950547731536409e-05,
+      "loss": 0.9128,
+      "step": 891
+    },
+    {
+      "epoch": 0.06563347896068061,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.9504330411637794e-05,
+      "loss": 0.8637,
+      "step": 892
+    },
+    {
+      "epoch": 0.0657070590940446,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.950318219281037e-05,
+      "loss": 1.0806,
+      "step": 893
+    },
+    {
+      "epoch": 0.0657806392274086,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.950203265894344e-05,
+      "loss": 0.6772,
+      "step": 894
+    },
+    {
+      "epoch": 0.0658542193607726,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.950088181009869e-05,
+      "loss": 0.7715,
+      "step": 895
+    },
+    {
+      "epoch": 0.06592779949413659,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.949972964633789e-05,
+      "loss": 1.4542,
+      "step": 896
+    },
+    {
+      "epoch": 0.06600137962750058,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.949857616772288e-05,
+      "loss": 1.1663,
+      "step": 897
+    },
+    {
+      "epoch": 0.06607495976086457,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.949742137431555e-05,
+      "loss": 0.7811,
+      "step": 898
+    },
+    {
+      "epoch": 0.06614853989422856,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.949626526617789e-05,
+      "loss": 0.9057,
+      "step": 899
+    },
+    {
+      "epoch": 0.06622212002759255,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.9495107843371934e-05,
+      "loss": 0.9346,
+      "step": 900
+    },
+    {
+      "epoch": 0.06629570016095654,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.949394910595981e-05,
+      "loss": 0.9646,
+      "step": 901
+    },
+    {
+      "epoch": 0.06636928029432053,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.9492789054003695e-05,
+      "loss": 0.945,
+      "step": 902
+    },
+    {
+      "epoch": 0.06644286042768452,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.9491627687565855e-05,
+      "loss": 0.7129,
+      "step": 903
+    },
+    {
+      "epoch": 0.06651644056104852,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.949046500670862e-05,
+      "loss": 0.8891,
+      "step": 904
+    },
+    {
+      "epoch": 0.06659002069441251,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.948930101149437e-05,
+      "loss": 0.7404,
+      "step": 905
+    },
+    {
+      "epoch": 0.0666636008277765,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.94881357019856e-05,
+      "loss": 1.1495,
+      "step": 906
+    },
+    {
+      "epoch": 0.0667371809611405,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.9486969078244835e-05,
+      "loss": 1.0328,
+      "step": 907
+    },
+    {
+      "epoch": 0.06681076109450448,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.9485801140334687e-05,
+      "loss": 0.9204,
+      "step": 908
+    },
+    {
+      "epoch": 0.06688434122786847,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.948463188831784e-05,
+      "loss": 0.9084,
+      "step": 909
+    },
+    {
+      "epoch": 0.06695792136123246,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.9483461322257054e-05,
+      "loss": 1.5526,
+      "step": 910
+    },
+    {
+      "epoch": 0.06703150149459645,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.948228944221513e-05,
+      "loss": 0.904,
+      "step": 911
+    },
+    {
+      "epoch": 0.06710508162796044,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.9481116248254976e-05,
+      "loss": 1.0075,
+      "step": 912
+    },
+    {
+      "epoch": 0.06717866176132445,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.947994174043955e-05,
+      "loss": 0.7878,
+      "step": 913
+    },
+    {
+      "epoch": 0.06725224189468844,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.947876591883189e-05,
+      "loss": 1.3461,
+      "step": 914
+    },
+    {
+      "epoch": 0.06732582202805243,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.947758878349509e-05,
+      "loss": 0.9798,
+      "step": 915
+    },
+    {
+      "epoch": 0.06739940216141642,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.947641033449234e-05,
+      "loss": 0.7173,
+      "step": 916
+    },
+    {
+      "epoch": 0.06747298229478041,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.947523057188686e-05,
+      "loss": 0.8946,
+      "step": 917
+    },
+    {
+      "epoch": 0.0675465624281444,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.9474049495742006e-05,
+      "loss": 1.2353,
+      "step": 918
+    },
+    {
+      "epoch": 0.06762014256150839,
+      "grad_norm": 0.875,
+      "learning_rate": 4.947286710612112e-05,
+      "loss": 0.956,
+      "step": 919
+    },
+    {
+      "epoch": 0.06769372269487238,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.9471683403087695e-05,
+      "loss": 1.2927,
+      "step": 920
+    },
+    {
+      "epoch": 0.06776730282823637,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.9470498386705236e-05,
+      "loss": 1.1632,
+      "step": 921
+    },
+    {
+      "epoch": 0.06784088296160037,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.9469312057037346e-05,
+      "loss": 1.243,
+      "step": 922
+    },
+    {
+      "epoch": 0.06791446309496436,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.94681244141477e-05,
+      "loss": 0.7535,
+      "step": 923
+    },
+    {
+      "epoch": 0.06798804322832835,
+      "grad_norm": 1.0,
+      "learning_rate": 4.946693545810002e-05,
+      "loss": 1.3773,
+      "step": 924
+    },
+    {
+      "epoch": 0.06806162336169234,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.946574518895813e-05,
+      "loss": 1.0478,
+      "step": 925
+    },
+    {
+      "epoch": 0.06813520349505633,
+      "grad_norm": 1.390625,
+      "learning_rate": 4.94645536067859e-05,
+      "loss": 2.1322,
+      "step": 926
+    },
+    {
+      "epoch": 0.06820878362842032,
+      "grad_norm": 1.2890625,
+      "learning_rate": 4.94633607116473e-05,
+      "loss": 1.1893,
+      "step": 927
+    },
+    {
+      "epoch": 0.06828236376178431,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.946216650360632e-05,
+      "loss": 1.2635,
+      "step": 928
+    },
+    {
+      "epoch": 0.0683559438951483,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.9460970982727074e-05,
+      "loss": 1.1346,
+      "step": 929
+    },
+    {
+      "epoch": 0.0684295240285123,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.945977414907371e-05,
+      "loss": 1.0327,
+      "step": 930
+    },
+    {
+      "epoch": 0.0685031041618763,
+      "grad_norm": 1.28125,
+      "learning_rate": 4.9458576002710474e-05,
+      "loss": 0.8676,
+      "step": 931
+    },
+    {
+      "epoch": 0.06857668429524029,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.945737654370165e-05,
+      "loss": 0.7823,
+      "step": 932
+    },
+    {
+      "epoch": 0.06865026442860428,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.9456175772111624e-05,
+      "loss": 1.0325,
+      "step": 933
+    },
+    {
+      "epoch": 0.06872384456196827,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.9454973688004835e-05,
+      "loss": 1.1964,
+      "step": 934
+    },
+    {
+      "epoch": 0.06879742469533226,
+      "grad_norm": 1.0,
+      "learning_rate": 4.94537702914458e-05,
+      "loss": 0.9395,
+      "step": 935
+    },
+    {
+      "epoch": 0.06887100482869625,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.94525655824991e-05,
+      "loss": 1.1721,
+      "step": 936
+    },
+    {
+      "epoch": 0.06894458496206024,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.945135956122938e-05,
+      "loss": 0.8203,
+      "step": 937
+    },
+    {
+      "epoch": 0.06901816509542423,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.945015222770139e-05,
+      "loss": 0.9418,
+      "step": 938
+    },
+    {
+      "epoch": 0.06909174522878822,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.94489435819799e-05,
+      "loss": 0.9211,
+      "step": 939
+    },
+    {
+      "epoch": 0.06916532536215222,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.944773362412979e-05,
+      "loss": 0.974,
+      "step": 940
+    },
+    {
+      "epoch": 0.06923890549551621,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.944652235421599e-05,
+      "loss": 1.1142,
+      "step": 941
+    },
+    {
+      "epoch": 0.0693124856288802,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.94453097723035e-05,
+      "loss": 1.0195,
+      "step": 942
+    },
+    {
+      "epoch": 0.0693860657622442,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.944409587845741e-05,
+      "loss": 1.0767,
+      "step": 943
+    },
+    {
+      "epoch": 0.06945964589560819,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.9442880672742866e-05,
+      "loss": 1.1185,
+      "step": 944
+    },
+    {
+      "epoch": 0.06953322602897218,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.944166415522509e-05,
+      "loss": 1.3641,
+      "step": 945
+    },
+    {
+      "epoch": 0.06960680616233617,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.944044632596935e-05,
+      "loss": 0.9978,
+      "step": 946
+    },
+    {
+      "epoch": 0.06968038629570016,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.9439227185041015e-05,
+      "loss": 1.0883,
+      "step": 947
+    },
+    {
+      "epoch": 0.06975396642906415,
+      "grad_norm": 1.2890625,
+      "learning_rate": 4.943800673250553e-05,
+      "loss": 1.2315,
+      "step": 948
+    },
+    {
+      "epoch": 0.06982754656242815,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.943678496842837e-05,
+      "loss": 0.934,
+      "step": 949
+    },
+    {
+      "epoch": 0.06990112669579214,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.9435561892875114e-05,
+      "loss": 0.6817,
+      "step": 950
+    },
+    {
+      "epoch": 0.06997470682915613,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.943433750591141e-05,
+      "loss": 0.9523,
+      "step": 951
+    },
+    {
+      "epoch": 0.07004828696252012,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.9433111807602957e-05,
+      "loss": 0.6998,
+      "step": 952
+    },
+    {
+      "epoch": 0.07012186709588411,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.943188479801554e-05,
+      "loss": 1.1778,
+      "step": 953
+    },
+    {
+      "epoch": 0.0701954472292481,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.9430656477215016e-05,
+      "loss": 0.8764,
+      "step": 954
+    },
+    {
+      "epoch": 0.07026902736261209,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.94294268452673e-05,
+      "loss": 1.3627,
+      "step": 955
+    },
+    {
+      "epoch": 0.07034260749597608,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.942819590223838e-05,
+      "loss": 0.9452,
+      "step": 956
+    },
+    {
+      "epoch": 0.07041618762934007,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.942696364819433e-05,
+      "loss": 1.3631,
+      "step": 957
+    },
+    {
+      "epoch": 0.07048976776270408,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.942573008320128e-05,
+      "loss": 1.4654,
+      "step": 958
+    },
+    {
+      "epoch": 0.07056334789606807,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.942449520732543e-05,
+      "loss": 0.801,
+      "step": 959
+    },
+    {
+      "epoch": 0.07063692802943206,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.942325902063305e-05,
+      "loss": 0.868,
+      "step": 960
+    },
+    {
+      "epoch": 0.07071050816279605,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.9422021523190486e-05,
+      "loss": 0.9592,
+      "step": 961
+    },
+    {
+      "epoch": 0.07078408829616004,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.9420782715064154e-05,
+      "loss": 0.941,
+      "step": 962
+    },
+    {
+      "epoch": 0.07085766842952403,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.9419542596320544e-05,
+      "loss": 1.2685,
+      "step": 963
+    },
+    {
+      "epoch": 0.07093124856288802,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.94183011670262e-05,
+      "loss": 0.8669,
+      "step": 964
+    },
+    {
+      "epoch": 0.071004828696252,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.941705842724775e-05,
+      "loss": 0.8033,
+      "step": 965
+    },
+    {
+      "epoch": 0.07107840882961601,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.941581437705189e-05,
+      "loss": 0.7802,
+      "step": 966
+    },
+    {
+      "epoch": 0.07115198896298,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.9414569016505394e-05,
+      "loss": 0.8719,
+      "step": 967
+    },
+    {
+      "epoch": 0.07122556909634399,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.941332234567509e-05,
+      "loss": 1.311,
+      "step": 968
+    },
+    {
+      "epoch": 0.07129914922970798,
+      "grad_norm": 0.875,
+      "learning_rate": 4.941207436462788e-05,
+      "loss": 1.0031,
+      "step": 969
+    },
+    {
+      "epoch": 0.07137272936307197,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.9410825073430756e-05,
+      "loss": 1.2397,
+      "step": 970
+    },
+    {
+      "epoch": 0.07144630949643596,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.940957447215075e-05,
+      "loss": 1.4965,
+      "step": 971
+    },
+    {
+      "epoch": 0.07151988962979995,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.9408322560854985e-05,
+      "loss": 1.0148,
+      "step": 972
+    },
+    {
+      "epoch": 0.07159346976316394,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.940706933961065e-05,
+      "loss": 0.6665,
+      "step": 973
+    },
+    {
+      "epoch": 0.07166704989652793,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.9405814808485e-05,
+      "loss": 0.7528,
+      "step": 974
+    },
+    {
+      "epoch": 0.07174063002989194,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.940455896754537e-05,
+      "loss": 0.9434,
+      "step": 975
+    },
+    {
+      "epoch": 0.07181421016325593,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.940330181685915e-05,
+      "loss": 0.7931,
+      "step": 976
+    },
+    {
+      "epoch": 0.07188779029661992,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.9402043356493816e-05,
+      "loss": 1.1085,
+      "step": 977
+    },
+    {
+      "epoch": 0.0719613704299839,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.9400783586516895e-05,
+      "loss": 0.6421,
+      "step": 978
+    },
+    {
+      "epoch": 0.0720349505633479,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.939952250699602e-05,
+      "loss": 0.7732,
+      "step": 979
+    },
+    {
+      "epoch": 0.07210853069671189,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.939826011799885e-05,
+      "loss": 0.713,
+      "step": 980
+    },
+    {
+      "epoch": 0.07218211083007588,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.9396996419593136e-05,
+      "loss": 0.6547,
+      "step": 981
+    },
+    {
+      "epoch": 0.07225569096343987,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.939573141184671e-05,
+      "loss": 0.7736,
+      "step": 982
+    },
+    {
+      "epoch": 0.07232927109680386,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.9394465094827446e-05,
+      "loss": 1.1406,
+      "step": 983
+    },
+    {
+      "epoch": 0.07240285123016786,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.9393197468603324e-05,
+      "loss": 1.4124,
+      "step": 984
+    },
+    {
+      "epoch": 0.07247643136353185,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.939192853324236e-05,
+      "loss": 1.4157,
+      "step": 985
+    },
+    {
+      "epoch": 0.07255001149689584,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.9390658288812675e-05,
+      "loss": 0.9396,
+      "step": 986
+    },
+    {
+      "epoch": 0.07262359163025983,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.938938673538242e-05,
+      "loss": 0.6932,
+      "step": 987
+    },
+    {
+      "epoch": 0.07269717176362382,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.9388113873019835e-05,
+      "loss": 1.0153,
+      "step": 988
+    },
+    {
+      "epoch": 0.07277075189698781,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.938683970179325e-05,
+      "loss": 1.0883,
+      "step": 989
+    },
+    {
+      "epoch": 0.0728443320303518,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.938556422177104e-05,
+      "loss": 0.9006,
+      "step": 990
+    },
+    {
+      "epoch": 0.07291791216371579,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.9384287433021646e-05,
+      "loss": 0.8341,
+      "step": 991
+    },
+    {
+      "epoch": 0.07299149229707978,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.9383009335613606e-05,
+      "loss": 1.3068,
+      "step": 992
+    },
+    {
+      "epoch": 0.07306507243044379,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.938172992961551e-05,
+      "loss": 1.0953,
+      "step": 993
+    },
+    {
+      "epoch": 0.07313865256380778,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.938044921509602e-05,
+      "loss": 0.9872,
+      "step": 994
+    },
+    {
+      "epoch": 0.07321223269717177,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.9379167192123864e-05,
+      "loss": 1.1377,
+      "step": 995
+    },
+    {
+      "epoch": 0.07328581283053576,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.937788386076786e-05,
+      "loss": 0.9147,
+      "step": 996
+    },
+    {
+      "epoch": 0.07335939296389975,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.9376599221096865e-05,
+      "loss": 1.6023,
+      "step": 997
+    },
+    {
+      "epoch": 0.07343297309726374,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.937531327317983e-05,
+      "loss": 1.0771,
+      "step": 998
+    },
+    {
+      "epoch": 0.07350655323062773,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.9374026017085776e-05,
+      "loss": 0.7791,
+      "step": 999
+    },
+    {
+      "epoch": 0.07358013336399172,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.9372737452883774e-05,
+      "loss": 1.045,
+      "step": 1000
+    },
+    {
+      "epoch": 0.07365371349735571,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.937144758064299e-05,
+      "loss": 0.5552,
+      "step": 1001
+    },
+    {
+      "epoch": 0.07372729363071971,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.937015640043264e-05,
+      "loss": 0.6764,
+      "step": 1002
+    },
+    {
+      "epoch": 0.0738008737640837,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.9368863912322036e-05,
+      "loss": 1.0907,
+      "step": 1003
+    },
+    {
+      "epoch": 0.07387445389744769,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.936757011638052e-05,
+      "loss": 0.7189,
+      "step": 1004
+    },
+    {
+      "epoch": 0.07394803403081168,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.936627501267755e-05,
+      "loss": 0.6947,
+      "step": 1005
+    },
+    {
+      "epoch": 0.07402161416417567,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.936497860128262e-05,
+      "loss": 0.9126,
+      "step": 1006
+    },
+    {
+      "epoch": 0.07409519429753966,
+      "grad_norm": 0.75,
+      "learning_rate": 4.93636808822653e-05,
+      "loss": 0.8022,
+      "step": 1007
+    },
+    {
+      "epoch": 0.07416877443090365,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.936238185569525e-05,
+      "loss": 0.8789,
+      "step": 1008
+    },
+    {
+      "epoch": 0.07424235456426764,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.9361081521642176e-05,
+      "loss": 1.1579,
+      "step": 1009
+    },
+    {
+      "epoch": 0.07431593469763163,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.935977988017587e-05,
+      "loss": 1.1439,
+      "step": 1010
+    },
+    {
+      "epoch": 0.07438951483099564,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.935847693136619e-05,
+      "loss": 1.4934,
+      "step": 1011
+    },
+    {
+      "epoch": 0.07446309496435963,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.935717267528305e-05,
+      "loss": 0.7336,
+      "step": 1012
+    },
+    {
+      "epoch": 0.07453667509772362,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.935586711199647e-05,
+      "loss": 0.7628,
+      "step": 1013
+    },
+    {
+      "epoch": 0.07461025523108761,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.9354560241576494e-05,
+      "loss": 0.9268,
+      "step": 1014
+    },
+    {
+      "epoch": 0.0746838353644516,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.935325206409327e-05,
+      "loss": 0.8797,
+      "step": 1015
+    },
+    {
+      "epoch": 0.07475741549781559,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.935194257961701e-05,
+      "loss": 0.6649,
+      "step": 1016
+    },
+    {
+      "epoch": 0.07483099563117958,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.935063178821798e-05,
+      "loss": 1.018,
+      "step": 1017
+    },
+    {
+      "epoch": 0.07490457576454357,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.934931968996654e-05,
+      "loss": 1.0185,
+      "step": 1018
+    },
+    {
+      "epoch": 0.07497815589790756,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.9348006284933094e-05,
+      "loss": 1.1034,
+      "step": 1019
+    },
+    {
+      "epoch": 0.07505173603127156,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.934669157318814e-05,
+      "loss": 1.3453,
+      "step": 1020
+    },
+    {
+      "epoch": 0.07512531616463555,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.9345375554802246e-05,
+      "loss": 1.0863,
+      "step": 1021
+    },
+    {
+      "epoch": 0.07519889629799954,
+      "grad_norm": 1.328125,
+      "learning_rate": 4.9344058229846015e-05,
+      "loss": 1.431,
+      "step": 1022
+    },
+    {
+      "epoch": 0.07527247643136353,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.934273959839015e-05,
+      "loss": 1.0453,
+      "step": 1023
+    },
+    {
+      "epoch": 0.07534605656472752,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.9341419660505444e-05,
+      "loss": 1.4968,
+      "step": 1024
+    },
+    {
+      "epoch": 0.07541963669809151,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.9340098416262714e-05,
+      "loss": 1.0828,
+      "step": 1025
+    },
+    {
+      "epoch": 0.0754932168314555,
+      "grad_norm": 0.875,
+      "learning_rate": 4.9338775865732874e-05,
+      "loss": 0.8269,
+      "step": 1026
+    },
+    {
+      "epoch": 0.0755667969648195,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.93374520089869e-05,
+      "loss": 1.1307,
+      "step": 1027
+    },
+    {
+      "epoch": 0.07564037709818348,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.9336126846095846e-05,
+      "loss": 0.7401,
+      "step": 1028
+    },
+    {
+      "epoch": 0.07571395723154749,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.933480037713083e-05,
+      "loss": 1.185,
+      "step": 1029
+    },
+    {
+      "epoch": 0.07578753736491148,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.9333472602163035e-05,
+      "loss": 0.6656,
+      "step": 1030
+    },
+    {
+      "epoch": 0.07586111749827547,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.933214352126373e-05,
+      "loss": 1.0686,
+      "step": 1031
+    },
+    {
+      "epoch": 0.07593469763163946,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.933081313450423e-05,
+      "loss": 0.6938,
+      "step": 1032
+    },
+    {
+      "epoch": 0.07600827776500345,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.932948144195596e-05,
+      "loss": 0.8716,
+      "step": 1033
+    },
+    {
+      "epoch": 0.07608185789836744,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.9328148443690356e-05,
+      "loss": 1.202,
+      "step": 1034
+    },
+    {
+      "epoch": 0.07615543803173143,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.9326814139778985e-05,
+      "loss": 1.1178,
+      "step": 1035
+    },
+    {
+      "epoch": 0.07622901816509542,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.932547853029344e-05,
+      "loss": 0.6975,
+      "step": 1036
+    },
+    {
+      "epoch": 0.07630259829845941,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.932414161530541e-05,
+      "loss": 0.7133,
+      "step": 1037
+    },
+    {
+      "epoch": 0.07637617843182341,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.932280339488664e-05,
+      "loss": 1.4606,
+      "step": 1038
+    },
+    {
+      "epoch": 0.0764497585651874,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.9321463869108954e-05,
+      "loss": 0.9715,
+      "step": 1039
+    },
+    {
+      "epoch": 0.0765233386985514,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.932012303804423e-05,
+      "loss": 1.0028,
+      "step": 1040
+    },
+    {
+      "epoch": 0.07659691883191538,
+      "grad_norm": 0.875,
+      "learning_rate": 4.931878090176445e-05,
+      "loss": 0.9707,
+      "step": 1041
+    },
+    {
+      "epoch": 0.07667049896527937,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.931743746034162e-05,
+      "loss": 0.8805,
+      "step": 1042
+    },
+    {
+      "epoch": 0.07674407909864336,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.931609271384785e-05,
+      "loss": 0.7408,
+      "step": 1043
+    },
+    {
+      "epoch": 0.07681765923200735,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.931474666235531e-05,
+      "loss": 1.489,
+      "step": 1044
+    },
+    {
+      "epoch": 0.07689123936537134,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.931339930593625e-05,
+      "loss": 0.636,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07696481949873533,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.931205064466297e-05,
+      "loss": 1.081,
+      "step": 1046
+    },
+    {
+      "epoch": 0.07703839963209934,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.931070067860785e-05,
+      "loss": 0.7789,
+      "step": 1047
+    },
+    {
+      "epoch": 0.07711197976546333,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.930934940784333e-05,
+      "loss": 0.7309,
+      "step": 1048
+    },
+    {
+      "epoch": 0.07718555989882732,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.930799683244195e-05,
+      "loss": 1.0883,
+      "step": 1049
+    },
+    {
+      "epoch": 0.07725914003219131,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.930664295247629e-05,
+      "loss": 1.576,
+      "step": 1050
+    },
+    {
+      "epoch": 0.0773327201655553,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.930528776801901e-05,
+      "loss": 0.8766,
+      "step": 1051
+    },
+    {
+      "epoch": 0.07740630029891929,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.9303931279142844e-05,
+      "loss": 0.705,
+      "step": 1052
+    },
+    {
+      "epoch": 0.07747988043228328,
+      "grad_norm": 1.6953125,
+      "learning_rate": 4.9302573485920587e-05,
+      "loss": 0.6259,
+      "step": 1053
+    },
+    {
+      "epoch": 0.07755346056564727,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.930121438842512e-05,
+      "loss": 1.1955,
+      "step": 1054
+    },
+    {
+      "epoch": 0.07762704069901127,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.9299853986729365e-05,
+      "loss": 0.6949,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07770062083237526,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.929849228090635e-05,
+      "loss": 0.8881,
+      "step": 1056
+    },
+    {
+      "epoch": 0.07777420096573925,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.929712927102914e-05,
+      "loss": 0.8625,
+      "step": 1057
+    },
+    {
+      "epoch": 0.07784778109910324,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.92957649571709e-05,
+      "loss": 1.0278,
+      "step": 1058
+    },
+    {
+      "epoch": 0.07792136123246723,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.929439933940484e-05,
+      "loss": 0.8735,
+      "step": 1059
+    },
+    {
+      "epoch": 0.07799494136583122,
+      "grad_norm": 0.875,
+      "learning_rate": 4.929303241780425e-05,
+      "loss": 0.8721,
+      "step": 1060
+    },
+    {
+      "epoch": 0.07806852149919521,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.92916641924425e-05,
+      "loss": 0.7614,
+      "step": 1061
+    },
+    {
+      "epoch": 0.0781421016325592,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.929029466339301e-05,
+      "loss": 1.2894,
+      "step": 1062
+    },
+    {
+      "epoch": 0.0782156817659232,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.928892383072928e-05,
+      "loss": 0.9365,
+      "step": 1063
+    },
+    {
+      "epoch": 0.0782892618992872,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.9287551694524894e-05,
+      "loss": 1.2815,
+      "step": 1064
+    },
+    {
+      "epoch": 0.07836284203265119,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.928617825485347e-05,
+      "loss": 0.7414,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07843642216601518,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.928480351178873e-05,
+      "loss": 0.8891,
+      "step": 1066
+    },
+    {
+      "epoch": 0.07851000229937917,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.928342746540446e-05,
+      "loss": 0.8198,
+      "step": 1067
+    },
+    {
+      "epoch": 0.07858358243274316,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.92820501157745e-05,
+      "loss": 0.7424,
+      "step": 1068
+    },
+    {
+      "epoch": 0.07865716256610715,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.928067146297277e-05,
+      "loss": 1.1136,
+      "step": 1069
+    },
+    {
+      "epoch": 0.07873074269947114,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.927929150707326e-05,
+      "loss": 1.0194,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07880432283283513,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.927791024815004e-05,
+      "loss": 0.828,
+      "step": 1071
+    },
+    {
+      "epoch": 0.07887790296619912,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.927652768627722e-05,
+      "loss": 1.1475,
+      "step": 1072
+    },
+    {
+      "epoch": 0.07895148309956312,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.9275143821529025e-05,
+      "loss": 1.2907,
+      "step": 1073
+    },
+    {
+      "epoch": 0.07902506323292711,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.927375865397969e-05,
+      "loss": 0.7493,
+      "step": 1074
+    },
+    {
+      "epoch": 0.0790986433662911,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.9272372183703594e-05,
+      "loss": 0.8228,
+      "step": 1075
+    },
+    {
+      "epoch": 0.0791722234996551,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.9270984410775125e-05,
+      "loss": 0.7071,
+      "step": 1076
+    },
+    {
+      "epoch": 0.07924580363301909,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.926959533526876e-05,
+      "loss": 0.98,
+      "step": 1077
+    },
+    {
+      "epoch": 0.07931938376638308,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.926820495725905e-05,
+      "loss": 0.968,
+      "step": 1078
+    },
+    {
+      "epoch": 0.07939296389974707,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.926681327682061e-05,
+      "loss": 1.0755,
+      "step": 1079
+    },
+    {
+      "epoch": 0.07946654403311106,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.926542029402815e-05,
+      "loss": 1.4966,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07954012416647505,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.9264026008956403e-05,
+      "loss": 0.8471,
+      "step": 1081
+    },
+    {
+      "epoch": 0.07961370429983905,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.926263042168021e-05,
+      "loss": 0.5744,
+      "step": 1082
+    },
+    {
+      "epoch": 0.07968728443320304,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.926123353227447e-05,
+      "loss": 1.1622,
+      "step": 1083
+    },
+    {
+      "epoch": 0.07976086456656703,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.925983534081416e-05,
+      "loss": 0.9022,
+      "step": 1084
+    },
+    {
+      "epoch": 0.07983444469993102,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.9258435847374286e-05,
+      "loss": 0.7833,
+      "step": 1085
+    },
+    {
+      "epoch": 0.07990802483329501,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.925703505202999e-05,
+      "loss": 0.8629,
+      "step": 1086
+    },
+    {
+      "epoch": 0.079981604966659,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.925563295485644e-05,
+      "loss": 1.7629,
+      "step": 1087
+    },
+    {
+      "epoch": 0.08005518510002299,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.925422955592887e-05,
+      "loss": 1.1017,
+      "step": 1088
+    },
+    {
+      "epoch": 0.08012876523338698,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.9252824855322624e-05,
+      "loss": 1.2518,
+      "step": 1089
+    },
+    {
+      "epoch": 0.08020234536675097,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.925141885311306e-05,
+      "loss": 1.0385,
+      "step": 1090
+    },
+    {
+      "epoch": 0.08027592550011498,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.9250011549375664e-05,
+      "loss": 0.9763,
+      "step": 1091
+    },
+    {
+      "epoch": 0.08034950563347897,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.924860294418594e-05,
+      "loss": 1.0588,
+      "step": 1092
+    },
+    {
+      "epoch": 0.08042308576684296,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.92471930376195e-05,
+      "loss": 0.6802,
+      "step": 1093
+    },
+    {
+      "epoch": 0.08049666590020695,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.924578182975201e-05,
+      "loss": 0.8913,
+      "step": 1094
+    },
+    {
+      "epoch": 0.08057024603357094,
+      "grad_norm": 2.5,
+      "learning_rate": 4.9244369320659204e-05,
+      "loss": 1.4806,
+      "step": 1095
+    },
+    {
+      "epoch": 0.08064382616693493,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.9242955510416877e-05,
+      "loss": 1.084,
+      "step": 1096
+    },
+    {
+      "epoch": 0.08071740630029892,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.924154039910092e-05,
+      "loss": 0.9423,
+      "step": 1097
+    },
+    {
+      "epoch": 0.0807909864336629,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.924012398678728e-05,
+      "loss": 1.2512,
+      "step": 1098
+    },
+    {
+      "epoch": 0.0808645665670269,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.923870627355196e-05,
+      "loss": 0.8654,
+      "step": 1099
+    },
+    {
+      "epoch": 0.0809381467003909,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.923728725947106e-05,
+      "loss": 0.7415,
+      "step": 1100
+    },
+    {
+      "epoch": 0.08101172683375489,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.923586694462073e-05,
+      "loss": 0.9244,
+      "step": 1101
+    },
+    {
+      "epoch": 0.08108530696711888,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.9234445329077207e-05,
+      "loss": 0.7599,
+      "step": 1102
+    },
+    {
+      "epoch": 0.08115888710048287,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.9233022412916766e-05,
+      "loss": 0.9252,
+      "step": 1103
+    },
+    {
+      "epoch": 0.08123246723384686,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.923159819621578e-05,
+      "loss": 0.7523,
+      "step": 1104
+    },
+    {
+      "epoch": 0.08130604736721085,
+      "grad_norm": 0.875,
+      "learning_rate": 4.9230172679050686e-05,
+      "loss": 1.0046,
+      "step": 1105
+    },
+    {
+      "epoch": 0.08137962750057484,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.922874586149799e-05,
+      "loss": 0.7878,
+      "step": 1106
+    },
+    {
+      "epoch": 0.08145320763393883,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.922731774363426e-05,
+      "loss": 0.9076,
+      "step": 1107
+    },
+    {
+      "epoch": 0.08152678776730282,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.9225888325536154e-05,
+      "loss": 1.5259,
+      "step": 1108
+    },
+    {
+      "epoch": 0.08160036790066683,
+      "grad_norm": 0.75,
+      "learning_rate": 4.922445760728037e-05,
+      "loss": 0.7926,
+      "step": 1109
+    },
+    {
+      "epoch": 0.08167394803403082,
+      "grad_norm": 1.0,
+      "learning_rate": 4.9223025588943704e-05,
+      "loss": 0.9363,
+      "step": 1110
+    },
+    {
+      "epoch": 0.0817475281673948,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.922159227060301e-05,
+      "loss": 0.8192,
+      "step": 1111
+    },
+    {
+      "epoch": 0.0818211083007588,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.92201576523352e-05,
+      "loss": 1.3752,
+      "step": 1112
+    },
+    {
+      "epoch": 0.08189468843412279,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.9218721734217274e-05,
+      "loss": 0.7666,
+      "step": 1113
+    },
+    {
+      "epoch": 0.08196826856748678,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.9217284516326304e-05,
+      "loss": 1.1508,
+      "step": 1114
+    },
+    {
+      "epoch": 0.08204184870085077,
+      "grad_norm": 3.3125,
+      "learning_rate": 4.921584599873941e-05,
+      "loss": 0.8317,
+      "step": 1115
+    },
+    {
+      "epoch": 0.08211542883421476,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.9214406181533795e-05,
+      "loss": 1.1477,
+      "step": 1116
+    },
+    {
+      "epoch": 0.08218900896757875,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.921296506478674e-05,
+      "loss": 0.7427,
+      "step": 1117
+    },
+    {
+      "epoch": 0.08226258910094275,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.921152264857557e-05,
+      "loss": 0.7214,
+      "step": 1118
+    },
+    {
+      "epoch": 0.08233616923430674,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.921007893297772e-05,
+      "loss": 0.8582,
+      "step": 1119
+    },
+    {
+      "epoch": 0.08240974936767073,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.920863391807066e-05,
+      "loss": 0.7496,
+      "step": 1120
+    },
+    {
+      "epoch": 0.08248332950103472,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.920718760393194e-05,
+      "loss": 0.6417,
+      "step": 1121
+    },
+    {
+      "epoch": 0.08255690963439871,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.920573999063918e-05,
+      "loss": 1.1525,
+      "step": 1122
+    },
+    {
+      "epoch": 0.0826304897677627,
+      "grad_norm": 1.28125,
+      "learning_rate": 4.920429107827007e-05,
+      "loss": 1.1709,
+      "step": 1123
+    },
+    {
+      "epoch": 0.08270406990112669,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.9202840866902374e-05,
+      "loss": 0.9924,
+      "step": 1124
+    },
+    {
+      "epoch": 0.08277765003449068,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.9201389356613925e-05,
+      "loss": 1.2457,
+      "step": 1125
+    },
+    {
+      "epoch": 0.08285123016785467,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.919993654748262e-05,
+      "loss": 0.7907,
+      "step": 1126
+    },
+    {
+      "epoch": 0.08292481030121868,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.919848243958642e-05,
+      "loss": 0.8108,
+      "step": 1127
+    },
+    {
+      "epoch": 0.08299839043458267,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.9197027033003376e-05,
+      "loss": 0.9923,
+      "step": 1128
+    },
+    {
+      "epoch": 0.08307197056794666,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.9195570327811595e-05,
+      "loss": 0.747,
+      "step": 1129
+    },
+    {
+      "epoch": 0.08314555070131065,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.9194112324089256e-05,
+      "loss": 0.7146,
+      "step": 1130
+    },
+    {
+      "epoch": 0.08321913083467464,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.91926530219146e-05,
+      "loss": 0.8503,
+      "step": 1131
+    },
+    {
+      "epoch": 0.08329271096803863,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.919119242136595e-05,
+      "loss": 0.8781,
+      "step": 1132
+    },
+    {
+      "epoch": 0.08336629110140262,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.918973052252169e-05,
+      "loss": 0.7988,
+      "step": 1133
+    },
+    {
+      "epoch": 0.08343987123476661,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.918826732546029e-05,
+      "loss": 0.923,
+      "step": 1134
+    },
+    {
+      "epoch": 0.0835134513681306,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.918680283026026e-05,
+      "loss": 0.6523,
+      "step": 1135
+    },
+    {
+      "epoch": 0.0835870315014946,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.9185337037000215e-05,
+      "loss": 1.0275,
+      "step": 1136
+    },
+    {
+      "epoch": 0.08366061163485859,
+      "grad_norm": 1.25,
+      "learning_rate": 4.9183869945758795e-05,
+      "loss": 1.8992,
+      "step": 1137
+    },
+    {
+      "epoch": 0.08373419176822258,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.918240155661477e-05,
+      "loss": 1.0567,
+      "step": 1138
+    },
+    {
+      "epoch": 0.08380777190158657,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.9180931869646904e-05,
+      "loss": 0.8504,
+      "step": 1139
+    },
+    {
+      "epoch": 0.08388135203495056,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.917946088493412e-05,
+      "loss": 1.3163,
+      "step": 1140
+    },
+    {
+      "epoch": 0.08395493216831455,
+      "grad_norm": 0.875,
+      "learning_rate": 4.917798860255533e-05,
+      "loss": 0.9608,
+      "step": 1141
+    },
+    {
+      "epoch": 0.08402851230167854,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.917651502258955e-05,
+      "loss": 0.8979,
+      "step": 1142
+    },
+    {
+      "epoch": 0.08410209243504253,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.917504014511587e-05,
+      "loss": 0.7785,
+      "step": 1143
+    },
+    {
+      "epoch": 0.08417567256840654,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.917356397021346e-05,
+      "loss": 0.8761,
+      "step": 1144
+    },
+    {
+      "epoch": 0.08424925270177053,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.917208649796152e-05,
+      "loss": 0.8724,
+      "step": 1145
+    },
+    {
+      "epoch": 0.08432283283513452,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.9170607728439355e-05,
+      "loss": 0.8965,
+      "step": 1146
+    },
+    {
+      "epoch": 0.08439641296849851,
+      "grad_norm": 0.875,
+      "learning_rate": 4.916912766172632e-05,
+      "loss": 0.82,
+      "step": 1147
+    },
+    {
+      "epoch": 0.0844699931018625,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.916764629790186e-05,
+      "loss": 0.9645,
+      "step": 1148
+    },
+    {
+      "epoch": 0.08454357323522649,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.916616363704547e-05,
+      "loss": 0.9855,
+      "step": 1149
+    },
+    {
+      "epoch": 0.08461715336859048,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.916467967923671e-05,
+      "loss": 0.7346,
+      "step": 1150
+    },
+    {
+      "epoch": 0.08469073350195447,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.916319442455524e-05,
+      "loss": 0.8516,
+      "step": 1151
+    },
+    {
+      "epoch": 0.08476431363531846,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.916170787308076e-05,
+      "loss": 0.9306,
+      "step": 1152
+    },
+    {
+      "epoch": 0.08483789376868246,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.9160220024893064e-05,
+      "loss": 1.1613,
+      "step": 1153
+    },
+    {
+      "epoch": 0.08491147390204645,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.915873088007198e-05,
+      "loss": 0.7963,
+      "step": 1154
+    },
+    {
+      "epoch": 0.08498505403541044,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.9157240438697446e-05,
+      "loss": 0.9718,
+      "step": 1155
+    },
+    {
+      "epoch": 0.08505863416877443,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.915574870084944e-05,
+      "loss": 1.3683,
+      "step": 1156
+    },
+    {
+      "epoch": 0.08513221430213842,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.9154255666608026e-05,
+      "loss": 0.5786,
+      "step": 1157
+    },
+    {
+      "epoch": 0.08520579443550241,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.915276133605333e-05,
+      "loss": 0.793,
+      "step": 1158
+    },
+    {
+      "epoch": 0.0852793745688664,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.915126570926555e-05,
+      "loss": 1.0918,
+      "step": 1159
+    },
+    {
+      "epoch": 0.0853529547022304,
+      "grad_norm": 3.828125,
+      "learning_rate": 4.914976878632496e-05,
+      "loss": 1.3251,
+      "step": 1160
+    },
+    {
+      "epoch": 0.08542653483559438,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.914827056731188e-05,
+      "loss": 0.9405,
+      "step": 1161
+    },
+    {
+      "epoch": 0.08550011496895839,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.914677105230674e-05,
+      "loss": 0.6488,
+      "step": 1162
+    },
+    {
+      "epoch": 0.08557369510232238,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.914527024139e-05,
+      "loss": 1.388,
+      "step": 1163
+    },
+    {
+      "epoch": 0.08564727523568637,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.914376813464221e-05,
+      "loss": 0.9075,
+      "step": 1164
+    },
+    {
+      "epoch": 0.08572085536905036,
+      "grad_norm": 0.75,
+      "learning_rate": 4.914226473214399e-05,
+      "loss": 0.6648,
+      "step": 1165
+    },
+    {
+      "epoch": 0.08579443550241435,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.9140760033976016e-05,
+      "loss": 1.1211,
+      "step": 1166
+    },
+    {
+      "epoch": 0.08586801563577834,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.913925404021905e-05,
+      "loss": 0.6689,
+      "step": 1167
+    },
+    {
+      "epoch": 0.08594159576914233,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.9137746750953907e-05,
+      "loss": 0.8045,
+      "step": 1168
+    },
+    {
+      "epoch": 0.08601517590250632,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.9136238166261485e-05,
+      "loss": 0.9566,
+      "step": 1169
+    },
+    {
+      "epoch": 0.08608875603587031,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.913472828622275e-05,
+      "loss": 1.16,
+      "step": 1170
+    },
+    {
+      "epoch": 0.08616233616923431,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.9133217110918726e-05,
+      "loss": 1.3695,
+      "step": 1171
+    },
+    {
+      "epoch": 0.0862359163025983,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.913170464043053e-05,
+      "loss": 0.8043,
+      "step": 1172
+    },
+    {
+      "epoch": 0.0863094964359623,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.913019087483932e-05,
+      "loss": 1.3014,
+      "step": 1173
+    },
+    {
+      "epoch": 0.08638307656932628,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.9128675814226346e-05,
+      "loss": 1.0793,
+      "step": 1174
+    },
+    {
+      "epoch": 0.08645665670269027,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.912715945867291e-05,
+      "loss": 1.1457,
+      "step": 1175
+    },
+    {
+      "epoch": 0.08653023683605426,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.9125641808260395e-05,
+      "loss": 0.7501,
+      "step": 1176
+    },
+    {
+      "epoch": 0.08660381696941825,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.9124122863070255e-05,
+      "loss": 0.9969,
+      "step": 1177
+    },
+    {
+      "epoch": 0.08667739710278224,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.9122602623184e-05,
+      "loss": 0.8683,
+      "step": 1178
+    },
+    {
+      "epoch": 0.08675097723614623,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.912108108868322e-05,
+      "loss": 0.7663,
+      "step": 1179
+    },
+    {
+      "epoch": 0.08682455736951024,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.911955825964958e-05,
+      "loss": 0.8899,
+      "step": 1180
+    },
+    {
+      "epoch": 0.08689813750287423,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.91180341361648e-05,
+      "loss": 0.7079,
+      "step": 1181
+    },
+    {
+      "epoch": 0.08697171763623822,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.91165087183107e-05,
+      "loss": 0.8011,
+      "step": 1182
+    },
+    {
+      "epoch": 0.08704529776960221,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.911498200616911e-05,
+      "loss": 1.0859,
+      "step": 1183
+    },
+    {
+      "epoch": 0.0871188779029662,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.911345399982198e-05,
+      "loss": 1.1429,
+      "step": 1184
+    },
+    {
+      "epoch": 0.08719245803633019,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.911192469935132e-05,
+      "loss": 0.8551,
+      "step": 1185
+    },
+    {
+      "epoch": 0.08726603816969418,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.9110394104839206e-05,
+      "loss": 0.9391,
+      "step": 1186
+    },
+    {
+      "epoch": 0.08733961830305817,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.910886221636777e-05,
+      "loss": 1.4532,
+      "step": 1187
+    },
+    {
+      "epoch": 0.08741319843642216,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.9107329034019245e-05,
+      "loss": 0.7465,
+      "step": 1188
+    },
+    {
+      "epoch": 0.08748677856978616,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.910579455787589e-05,
+      "loss": 1.4572,
+      "step": 1189
+    },
+    {
+      "epoch": 0.08756035870315015,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.910425878802008e-05,
+      "loss": 0.8336,
+      "step": 1190
+    },
+    {
+      "epoch": 0.08763393883651414,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.9102721724534233e-05,
+      "loss": 0.9518,
+      "step": 1191
+    },
+    {
+      "epoch": 0.08770751896987813,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.910118336750083e-05,
+      "loss": 1.0194,
+      "step": 1192
+    },
+    {
+      "epoch": 0.08778109910324212,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.909964371700243e-05,
+      "loss": 1.04,
+      "step": 1193
+    },
+    {
+      "epoch": 0.08785467923660611,
+      "grad_norm": 1.3125,
+      "learning_rate": 4.909810277312168e-05,
+      "loss": 1.8463,
+      "step": 1194
+    },
+    {
+      "epoch": 0.0879282593699701,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.9096560535941264e-05,
+      "loss": 0.7999,
+      "step": 1195
+    },
+    {
+      "epoch": 0.0880018395033341,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.9095017005543956e-05,
+      "loss": 1.28,
+      "step": 1196
+    },
+    {
+      "epoch": 0.08807541963669809,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.909347218201259e-05,
+      "loss": 1.0784,
+      "step": 1197
+    },
+    {
+      "epoch": 0.08814899977006209,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.9091926065430084e-05,
+      "loss": 0.7425,
+      "step": 1198
+    },
+    {
+      "epoch": 0.08822257990342608,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.90903786558794e-05,
+      "loss": 0.8728,
+      "step": 1199
+    },
+    {
+      "epoch": 0.08829616003679007,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.9088829953443605e-05,
+      "loss": 1.4336,
+      "step": 1200
+    },
+    {
+      "epoch": 0.08836974017015406,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.9087279958205794e-05,
+      "loss": 0.9031,
+      "step": 1201
+    },
+    {
+      "epoch": 0.08844332030351805,
+      "grad_norm": 0.75,
+      "learning_rate": 4.908572867024917e-05,
+      "loss": 0.7032,
+      "step": 1202
+    },
+    {
+      "epoch": 0.08851690043688204,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.9084176089656975e-05,
+      "loss": 0.9688,
+      "step": 1203
+    },
+    {
+      "epoch": 0.08859048057024603,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.908262221651254e-05,
+      "loss": 1.2545,
+      "step": 1204
+    },
+    {
+      "epoch": 0.08866406070361002,
+      "grad_norm": 1.0,
+      "learning_rate": 4.9081067050899257e-05,
+      "loss": 1.2364,
+      "step": 1205
+    },
+    {
+      "epoch": 0.08873764083697401,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.9079510592900583e-05,
+      "loss": 1.2256,
+      "step": 1206
+    },
+    {
+      "epoch": 0.08881122097033801,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.907795284260006e-05,
+      "loss": 1.3518,
+      "step": 1207
+    },
+    {
+      "epoch": 0.088884801103702,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.907639380008129e-05,
+      "loss": 1.0418,
+      "step": 1208
+    },
+    {
+      "epoch": 0.088958381237066,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.9074833465427925e-05,
+      "loss": 0.8984,
+      "step": 1209
+    },
+    {
+      "epoch": 0.08903196137042999,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.9073271838723734e-05,
+      "loss": 0.8314,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08910554150379398,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.90717089200525e-05,
+      "loss": 0.9828,
+      "step": 1211
+    },
+    {
+      "epoch": 0.08917912163715797,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.907014470949812e-05,
+      "loss": 0.8529,
+      "step": 1212
+    },
+    {
+      "epoch": 0.08925270177052196,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.9068579207144525e-05,
+      "loss": 1.2085,
+      "step": 1213
+    },
+    {
+      "epoch": 0.08932628190388595,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.906701241307575e-05,
+      "loss": 1.207,
+      "step": 1214
+    },
+    {
+      "epoch": 0.08939986203724994,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.9065444327375876e-05,
+      "loss": 1.1959,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08947344217061394,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.9063874950129066e-05,
+      "loss": 1.0856,
+      "step": 1216
+    },
+    {
+      "epoch": 0.08954702230397793,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.9062304281419525e-05,
+      "loss": 0.9216,
+      "step": 1217
+    },
+    {
+      "epoch": 0.08962060243734192,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.906073232133157e-05,
+      "loss": 1.2706,
+      "step": 1218
+    },
+    {
+      "epoch": 0.08969418257070591,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.905915906994955e-05,
+      "loss": 0.9168,
+      "step": 1219
+    },
+    {
+      "epoch": 0.0897677627040699,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.9057584527357894e-05,
+      "loss": 1.1207,
+      "step": 1220
+    },
+    {
+      "epoch": 0.08984134283743389,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.905600869364113e-05,
+      "loss": 0.8928,
+      "step": 1221
+    },
+    {
+      "epoch": 0.08991492297079788,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.905443156888381e-05,
+      "loss": 1.1832,
+      "step": 1222
+    },
+    {
+      "epoch": 0.08998850310416187,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.905285315317058e-05,
+      "loss": 0.9636,
+      "step": 1223
+    },
+    {
+      "epoch": 0.09006208323752586,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.905127344658615e-05,
+      "loss": 0.7163,
+      "step": 1224
+    },
+    {
+      "epoch": 0.09013566337088987,
+      "grad_norm": 1.3046875,
+      "learning_rate": 4.90496924492153e-05,
+      "loss": 1.7359,
+      "step": 1225
+    },
+    {
+      "epoch": 0.09020924350425386,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.904811016114288e-05,
+      "loss": 1.1708,
+      "step": 1226
+    },
+    {
+      "epoch": 0.09028282363761785,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.9046526582453814e-05,
+      "loss": 0.8912,
+      "step": 1227
+    },
+    {
+      "epoch": 0.09035640377098184,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.904494171323307e-05,
+      "loss": 0.7695,
+      "step": 1228
+    },
+    {
+      "epoch": 0.09042998390434583,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.904335555356573e-05,
+      "loss": 0.9219,
+      "step": 1229
+    },
+    {
+      "epoch": 0.09050356403770982,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.9041768103536904e-05,
+      "loss": 0.8511,
+      "step": 1230
+    },
+    {
+      "epoch": 0.0905771441710738,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.90401793632318e-05,
+      "loss": 0.8241,
+      "step": 1231
+    },
+    {
+      "epoch": 0.0906507243044378,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.903858933273566e-05,
+      "loss": 1.0765,
+      "step": 1232
+    },
+    {
+      "epoch": 0.0907243044378018,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.903699801213385e-05,
+      "loss": 1.3338,
+      "step": 1233
+    },
+    {
+      "epoch": 0.09079788457116579,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.9035405401511745e-05,
+      "loss": 1.0775,
+      "step": 1234
+    },
+    {
+      "epoch": 0.09087146470452978,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.9033811500954836e-05,
+      "loss": 0.9106,
+      "step": 1235
+    },
+    {
+      "epoch": 0.09094504483789377,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.903221631054865e-05,
+      "loss": 0.6733,
+      "step": 1236
+    },
+    {
+      "epoch": 0.09101862497125776,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.903061983037881e-05,
+      "loss": 0.9866,
+      "step": 1237
+    },
+    {
+      "epoch": 0.09109220510462175,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.902902206053099e-05,
+      "loss": 1.0376,
+      "step": 1238
+    },
+    {
+      "epoch": 0.09116578523798574,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.902742300109094e-05,
+      "loss": 0.9408,
+      "step": 1239
+    },
+    {
+      "epoch": 0.09123936537134973,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.902582265214448e-05,
+      "loss": 1.5036,
+      "step": 1240
+    },
+    {
+      "epoch": 0.09131294550471372,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.9024221013777494e-05,
+      "loss": 1.1578,
+      "step": 1241
+    },
+    {
+      "epoch": 0.09138652563807773,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.902261808607594e-05,
+      "loss": 0.8581,
+      "step": 1242
+    },
+    {
+      "epoch": 0.09146010577144172,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.902101386912585e-05,
+      "loss": 0.941,
+      "step": 1243
+    },
+    {
+      "epoch": 0.0915336859048057,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.9019408363013305e-05,
+      "loss": 0.9933,
+      "step": 1244
+    },
+    {
+      "epoch": 0.0916072660381697,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.9017801567824494e-05,
+      "loss": 0.9652,
+      "step": 1245
+    },
+    {
+      "epoch": 0.09168084617153369,
+      "grad_norm": 1.0,
+      "learning_rate": 4.901619348364563e-05,
+      "loss": 1.0716,
+      "step": 1246
+    },
+    {
+      "epoch": 0.09175442630489768,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.901458411056302e-05,
+      "loss": 1.1948,
+      "step": 1247
+    },
+    {
+      "epoch": 0.09182800643826167,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.901297344866304e-05,
+      "loss": 0.9243,
+      "step": 1248
+    },
+    {
+      "epoch": 0.09190158657162566,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.901136149803213e-05,
+      "loss": 1.1386,
+      "step": 1249
+    },
+    {
+      "epoch": 0.09197516670498965,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.900974825875679e-05,
+      "loss": 1.0933,
+      "step": 1250
+    },
+    {
+      "epoch": 0.09204874683835365,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.900813373092362e-05,
+      "loss": 0.8543,
+      "step": 1251
+    },
+    {
+      "epoch": 0.09212232697171764,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.900651791461926e-05,
+      "loss": 0.9306,
+      "step": 1252
+    },
+    {
+      "epoch": 0.09219590710508163,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.900490080993042e-05,
+      "loss": 0.8872,
+      "step": 1253
+    },
+    {
+      "epoch": 0.09226948723844562,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.900328241694389e-05,
+      "loss": 1.4837,
+      "step": 1254
+    },
+    {
+      "epoch": 0.09234306737180961,
+      "grad_norm": 0.75,
+      "learning_rate": 4.900166273574653e-05,
+      "loss": 0.508,
+      "step": 1255
+    },
+    {
+      "epoch": 0.0924166475051736,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.9000041766425264e-05,
+      "loss": 0.8884,
+      "step": 1256
+    },
+    {
+      "epoch": 0.09249022763853759,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.8998419509067086e-05,
+      "loss": 0.776,
+      "step": 1257
+    },
+    {
+      "epoch": 0.09256380777190158,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.899679596375906e-05,
+      "loss": 0.9784,
+      "step": 1258
+    },
+    {
+      "epoch": 0.09263738790526557,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.899517113058831e-05,
+      "loss": 1.8613,
+      "step": 1259
+    },
+    {
+      "epoch": 0.09271096803862958,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.899354500964206e-05,
+      "loss": 1.2688,
+      "step": 1260
+    },
+    {
+      "epoch": 0.09278454817199357,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.8991917601007566e-05,
+      "loss": 0.7221,
+      "step": 1261
+    },
+    {
+      "epoch": 0.09285812830535756,
+      "grad_norm": 1.9296875,
+      "learning_rate": 4.899028890477216e-05,
+      "loss": 0.9086,
+      "step": 1262
+    },
+    {
+      "epoch": 0.09293170843872155,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.898865892102326e-05,
+      "loss": 0.8697,
+      "step": 1263
+    },
+    {
+      "epoch": 0.09300528857208554,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.898702764984835e-05,
+      "loss": 1.2328,
+      "step": 1264
+    },
+    {
+      "epoch": 0.09307886870544953,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.898539509133496e-05,
+      "loss": 1.3106,
+      "step": 1265
+    },
+    {
+      "epoch": 0.09315244883881352,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.898376124557073e-05,
+      "loss": 0.9223,
+      "step": 1266
+    },
+    {
+      "epoch": 0.09322602897217751,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.898212611264333e-05,
+      "loss": 1.0953,
+      "step": 1267
+    },
+    {
+      "epoch": 0.0932996091055415,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.898048969264051e-05,
+      "loss": 0.8959,
+      "step": 1268
+    },
+    {
+      "epoch": 0.0933731892389055,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.897885198565011e-05,
+      "loss": 0.7999,
+      "step": 1269
+    },
+    {
+      "epoch": 0.09344676937226949,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.8977212991760014e-05,
+      "loss": 0.9548,
+      "step": 1270
+    },
+    {
+      "epoch": 0.09352034950563348,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.897557271105817e-05,
+      "loss": 0.6872,
+      "step": 1271
+    },
+    {
+      "epoch": 0.09359392963899747,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.897393114363264e-05,
+      "loss": 0.8207,
+      "step": 1272
+    },
+    {
+      "epoch": 0.09366750977236146,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.8972288289571494e-05,
+      "loss": 1.1634,
+      "step": 1273
+    },
+    {
+      "epoch": 0.09374108990572545,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.8970644148962916e-05,
+      "loss": 0.9048,
+      "step": 1274
+    },
+    {
+      "epoch": 0.09381467003908944,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.8968998721895145e-05,
+      "loss": 1.0288,
+      "step": 1275
+    },
+    {
+      "epoch": 0.09388825017245343,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.896735200845647e-05,
+      "loss": 0.8646,
+      "step": 1276
+    },
+    {
+      "epoch": 0.09396183030581742,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.896570400873529e-05,
+      "loss": 0.87,
+      "step": 1277
+    },
+    {
+      "epoch": 0.09403541043918143,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.896405472282004e-05,
+      "loss": 1.0347,
+      "step": 1278
+    },
+    {
+      "epoch": 0.09410899057254542,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.8962404150799236e-05,
+      "loss": 0.841,
+      "step": 1279
+    },
+    {
+      "epoch": 0.09418257070590941,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.896075229276146e-05,
+      "loss": 0.8673,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0942561508392734,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.8959099148795365e-05,
+      "loss": 0.9358,
+      "step": 1281
+    },
+    {
+      "epoch": 0.09432973097263739,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.895744471898967e-05,
+      "loss": 0.6897,
+      "step": 1282
+    },
+    {
+      "epoch": 0.09440331110600138,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.895578900343316e-05,
+      "loss": 0.806,
+      "step": 1283
+    },
+    {
+      "epoch": 0.09447689123936537,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.89541320022147e-05,
+      "loss": 0.9246,
+      "step": 1284
+    },
+    {
+      "epoch": 0.09455047137272936,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.895247371542323e-05,
+      "loss": 1.5952,
+      "step": 1285
+    },
+    {
+      "epoch": 0.09462405150609335,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.8950814143147725e-05,
+      "loss": 1.1463,
+      "step": 1286
+    },
+    {
+      "epoch": 0.09469763163945735,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.894915328547727e-05,
+      "loss": 0.7581,
+      "step": 1287
+    },
+    {
+      "epoch": 0.09477121177282134,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.894749114250098e-05,
+      "loss": 0.9762,
+      "step": 1288
+    },
+    {
+      "epoch": 0.09484479190618533,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.894582771430808e-05,
+      "loss": 0.6978,
+      "step": 1289
+    },
+    {
+      "epoch": 0.09491837203954932,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.8944163000987834e-05,
+      "loss": 1.0474,
+      "step": 1290
+    },
+    {
+      "epoch": 0.09499195217291331,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.894249700262958e-05,
+      "loss": 0.8466,
+      "step": 1291
+    },
+    {
+      "epoch": 0.0950655323062773,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.894082971932274e-05,
+      "loss": 1.1338,
+      "step": 1292
+    },
+    {
+      "epoch": 0.0951391124396413,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.893916115115678e-05,
+      "loss": 1.1831,
+      "step": 1293
+    },
+    {
+      "epoch": 0.09521269257300528,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.893749129822125e-05,
+      "loss": 1.0496,
+      "step": 1294
+    },
+    {
+      "epoch": 0.09528627270636927,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.893582016060578e-05,
+      "loss": 0.7504,
+      "step": 1295
+    },
+    {
+      "epoch": 0.09535985283973328,
+      "grad_norm": 0.75,
+      "learning_rate": 4.893414773840005e-05,
+      "loss": 0.673,
+      "step": 1296
+    },
+    {
+      "epoch": 0.09543343297309727,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.893247403169382e-05,
+      "loss": 1.3115,
+      "step": 1297
+    },
+    {
+      "epoch": 0.09550701310646126,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.89307990405769e-05,
+      "loss": 0.915,
+      "step": 1298
+    },
+    {
+      "epoch": 0.09558059323982525,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.8929122765139206e-05,
+      "loss": 0.7835,
+      "step": 1299
+    },
+    {
+      "epoch": 0.09565417337318924,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.892744520547069e-05,
+      "loss": 1.058,
+      "step": 1300
+    },
+    {
+      "epoch": 0.09572775350655323,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.8925766361661376e-05,
+      "loss": 1.0108,
+      "step": 1301
+    },
+    {
+      "epoch": 0.09580133363991722,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.892408623380137e-05,
+      "loss": 0.8173,
+      "step": 1302
+    },
+    {
+      "epoch": 0.09587491377328121,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.892240482198084e-05,
+      "loss": 0.7766,
+      "step": 1303
+    },
+    {
+      "epoch": 0.0959484939066452,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.892072212629003e-05,
+      "loss": 0.6871,
+      "step": 1304
+    },
+    {
+      "epoch": 0.0960220740400092,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.8919038146819244e-05,
+      "loss": 1.2853,
+      "step": 1305
+    },
+    {
+      "epoch": 0.0960956541733732,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.891735288365886e-05,
+      "loss": 1.4199,
+      "step": 1306
+    },
+    {
+      "epoch": 0.09616923430673718,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.8915666336899315e-05,
+      "loss": 1.5016,
+      "step": 1307
+    },
+    {
+      "epoch": 0.09624281444010117,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.8913978506631134e-05,
+      "loss": 0.9186,
+      "step": 1308
+    },
+    {
+      "epoch": 0.09631639457346516,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.8912289392944885e-05,
+      "loss": 0.9134,
+      "step": 1309
+    },
+    {
+      "epoch": 0.09638997470682915,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.8910598995931236e-05,
+      "loss": 0.7541,
+      "step": 1310
+    },
+    {
+      "epoch": 0.09646355484019314,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.890890731568089e-05,
+      "loss": 1.0797,
+      "step": 1311
+    },
+    {
+      "epoch": 0.09653713497355713,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.890721435228465e-05,
+      "loss": 1.0187,
+      "step": 1312
+    },
+    {
+      "epoch": 0.09661071510692112,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.8905520105833375e-05,
+      "loss": 1.4667,
+      "step": 1313
+    },
+    {
+      "epoch": 0.09668429524028513,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.8903824576417986e-05,
+      "loss": 0.9393,
+      "step": 1314
+    },
+    {
+      "epoch": 0.09675787537364912,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.890212776412948e-05,
+      "loss": 1.2802,
+      "step": 1315
+    },
+    {
+      "epoch": 0.09683145550701311,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.8900429669058914e-05,
+      "loss": 0.9362,
+      "step": 1316
+    },
+    {
+      "epoch": 0.0969050356403771,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.889873029129743e-05,
+      "loss": 1.23,
+      "step": 1317
+    },
+    {
+      "epoch": 0.09697861577374109,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.8897029630936244e-05,
+      "loss": 0.7691,
+      "step": 1318
+    },
+    {
+      "epoch": 0.09705219590710508,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.88953276880666e-05,
+      "loss": 0.739,
+      "step": 1319
+    },
+    {
+      "epoch": 0.09712577604046907,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.8893624462779854e-05,
+      "loss": 0.9665,
+      "step": 1320
+    },
+    {
+      "epoch": 0.09719935617383306,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.889191995516741e-05,
+      "loss": 1.2012,
+      "step": 1321
+    },
+    {
+      "epoch": 0.09727293630719705,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.8890214165320746e-05,
+      "loss": 0.9751,
+      "step": 1322
+    },
+    {
+      "epoch": 0.09734651644056105,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.888850709333141e-05,
+      "loss": 1.0713,
+      "step": 1323
+    },
+    {
+      "epoch": 0.09742009657392504,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.888679873929103e-05,
+      "loss": 0.8512,
+      "step": 1324
+    },
+    {
+      "epoch": 0.09749367670728903,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.888508910329126e-05,
+      "loss": 1.0594,
+      "step": 1325
+    },
+    {
+      "epoch": 0.09756725684065302,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.888337818542388e-05,
+      "loss": 1.2818,
+      "step": 1326
+    },
+    {
+      "epoch": 0.09764083697401701,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.88816659857807e-05,
+      "loss": 1.1116,
+      "step": 1327
+    },
+    {
+      "epoch": 0.097714417107381,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.887995250445361e-05,
+      "loss": 0.6879,
+      "step": 1328
+    },
+    {
+      "epoch": 0.097787997240745,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.8878237741534584e-05,
+      "loss": 1.156,
+      "step": 1329
+    },
+    {
+      "epoch": 0.09786157737410899,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.8876521697115627e-05,
+      "loss": 0.971,
+      "step": 1330
+    },
+    {
+      "epoch": 0.09793515750747299,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.887480437128885e-05,
+      "loss": 0.8463,
+      "step": 1331
+    },
+    {
+      "epoch": 0.09800873764083698,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.8873085764146406e-05,
+      "loss": 1.2374,
+      "step": 1332
+    },
+    {
+      "epoch": 0.09808231777420097,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.887136587578055e-05,
+      "loss": 1.104,
+      "step": 1333
+    },
+    {
+      "epoch": 0.09815589790756496,
+      "grad_norm": 1.0,
+      "learning_rate": 4.886964470628357e-05,
+      "loss": 0.9857,
+      "step": 1334
+    },
+    {
+      "epoch": 0.09822947804092895,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.886792225574784e-05,
+      "loss": 0.6217,
+      "step": 1335
+    },
+    {
+      "epoch": 0.09830305817429294,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.88661985242658e-05,
+      "loss": 0.7217,
+      "step": 1336
+    },
+    {
+      "epoch": 0.09837663830765693,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.8864473511929975e-05,
+      "loss": 1.8299,
+      "step": 1337
+    },
+    {
+      "epoch": 0.09845021844102092,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.8862747218832924e-05,
+      "loss": 0.8521,
+      "step": 1338
+    },
+    {
+      "epoch": 0.09852379857438491,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.8861019645067296e-05,
+      "loss": 0.7501,
+      "step": 1339
+    },
+    {
+      "epoch": 0.09859737870774891,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.8859290790725816e-05,
+      "loss": 0.9327,
+      "step": 1340
+    },
+    {
+      "epoch": 0.0986709588411129,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.885756065590126e-05,
+      "loss": 1.3556,
+      "step": 1341
+    },
+    {
+      "epoch": 0.0987445389744769,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.8855829240686486e-05,
+      "loss": 0.8993,
+      "step": 1342
+    },
+    {
+      "epoch": 0.09881811910784088,
+      "grad_norm": 1.0,
+      "learning_rate": 4.885409654517441e-05,
+      "loss": 0.963,
+      "step": 1343
+    },
+    {
+      "epoch": 0.09889169924120488,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.885236256945803e-05,
+      "loss": 1.1411,
+      "step": 1344
+    },
+    {
+      "epoch": 0.09896527937456887,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.8850627313630405e-05,
+      "loss": 0.8609,
+      "step": 1345
+    },
+    {
+      "epoch": 0.09903885950793286,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.884889077778465e-05,
+      "loss": 0.8643,
+      "step": 1346
+    },
+    {
+      "epoch": 0.09911243964129685,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.8847152962013974e-05,
+      "loss": 0.8406,
+      "step": 1347
+    },
+    {
+      "epoch": 0.09918601977466084,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.884541386641165e-05,
+      "loss": 0.9565,
+      "step": 1348
+    },
+    {
+      "epoch": 0.09925959990802484,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.8843673491070984e-05,
+      "loss": 0.8559,
+      "step": 1349
+    },
+    {
+      "epoch": 0.09933318004138883,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.884193183608541e-05,
+      "loss": 1.0358,
+      "step": 1350
+    },
+    {
+      "epoch": 0.09940676017475282,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.8840188901548375e-05,
+      "loss": 1.3706,
+      "step": 1351
+    },
+    {
+      "epoch": 0.09948034030811681,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.883844468755344e-05,
+      "loss": 0.9408,
+      "step": 1352
+    },
+    {
+      "epoch": 0.0995539204414808,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.88366991941942e-05,
+      "loss": 0.655,
+      "step": 1353
+    },
+    {
+      "epoch": 0.09962750057484479,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.883495242156433e-05,
+      "loss": 0.9964,
+      "step": 1354
+    },
+    {
+      "epoch": 0.09970108070820878,
+      "grad_norm": 1.6640625,
+      "learning_rate": 4.8833204369757586e-05,
+      "loss": 1.0287,
+      "step": 1355
+    },
+    {
+      "epoch": 0.09977466084157277,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.883145503886778e-05,
+      "loss": 1.0197,
+      "step": 1356
+    },
+    {
+      "epoch": 0.09984824097493676,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.882970442898879e-05,
+      "loss": 1.3703,
+      "step": 1357
+    },
+    {
+      "epoch": 0.09992182110830077,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.8827952540214564e-05,
+      "loss": 0.9072,
+      "step": 1358
+    },
+    {
+      "epoch": 0.09999540124166476,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.8826199372639136e-05,
+      "loss": 0.6855,
+      "step": 1359
+    },
+    {
+      "epoch": 0.10006898137502875,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.8824444926356593e-05,
+      "loss": 0.9797,
+      "step": 1360
+    },
+    {
+      "epoch": 0.10014256150839274,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.882268920146109e-05,
+      "loss": 0.7094,
+      "step": 1361
+    },
+    {
+      "epoch": 0.10021614164175673,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.882093219804684e-05,
+      "loss": 1.0087,
+      "step": 1362
+    },
+    {
+      "epoch": 0.10028972177512072,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.881917391620816e-05,
+      "loss": 0.6034,
+      "step": 1363
+    },
+    {
+      "epoch": 0.1003633019084847,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.8817414356039406e-05,
+      "loss": 0.937,
+      "step": 1364
+    },
+    {
+      "epoch": 0.1004368820418487,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.8815653517635e-05,
+      "loss": 0.7939,
+      "step": 1365
+    },
+    {
+      "epoch": 0.10051046217521269,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.881389140108946e-05,
+      "loss": 0.91,
+      "step": 1366
+    },
+    {
+      "epoch": 0.10058404230857669,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.8812128006497335e-05,
+      "loss": 0.9494,
+      "step": 1367
+    },
+    {
+      "epoch": 0.10065762244194068,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.881036333395329e-05,
+      "loss": 1.3114,
+      "step": 1368
+    },
+    {
+      "epoch": 0.10073120257530467,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.8808597383552e-05,
+      "loss": 0.831,
+      "step": 1369
+    },
+    {
+      "epoch": 0.10080478270866866,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.880683015538827e-05,
+      "loss": 1.3189,
+      "step": 1370
+    },
+    {
+      "epoch": 0.10087836284203265,
+      "grad_norm": 1.3125,
+      "learning_rate": 4.8805061649556924e-05,
+      "loss": 1.0953,
+      "step": 1371
+    },
+    {
+      "epoch": 0.10095194297539664,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.8803291866152876e-05,
+      "loss": 1.2109,
+      "step": 1372
+    },
+    {
+      "epoch": 0.10102552310876063,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.880152080527112e-05,
+      "loss": 0.7837,
+      "step": 1373
+    },
+    {
+      "epoch": 0.10109910324212462,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.8799748467006694e-05,
+      "loss": 1.2422,
+      "step": 1374
+    },
+    {
+      "epoch": 0.10117268337548861,
+      "grad_norm": 1.3203125,
+      "learning_rate": 4.879797485145472e-05,
+      "loss": 1.377,
+      "step": 1375
+    },
+    {
+      "epoch": 0.10124626350885262,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.879619995871038e-05,
+      "loss": 1.1502,
+      "step": 1376
+    },
+    {
+      "epoch": 0.1013198436422166,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.8794423788868934e-05,
+      "loss": 1.0014,
+      "step": 1377
+    },
+    {
+      "epoch": 0.1013934237755806,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.879264634202571e-05,
+      "loss": 0.7864,
+      "step": 1378
+    },
+    {
+      "epoch": 0.10146700390894459,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.879086761827609e-05,
+      "loss": 0.9298,
+      "step": 1379
+    },
+    {
+      "epoch": 0.10154058404230858,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.878908761771554e-05,
+      "loss": 0.9901,
+      "step": 1380
+    },
+    {
+      "epoch": 0.10161416417567257,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.8787306340439587e-05,
+      "loss": 1.1022,
+      "step": 1381
+    },
+    {
+      "epoch": 0.10168774430903656,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.8785523786543836e-05,
+      "loss": 1.1433,
+      "step": 1382
+    },
+    {
+      "epoch": 0.10176132444240055,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.878373995612394e-05,
+      "loss": 0.9623,
+      "step": 1383
+    },
+    {
+      "epoch": 0.10183490457576454,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.878195484927565e-05,
+      "loss": 0.8162,
+      "step": 1384
+    },
+    {
+      "epoch": 0.10190848470912854,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.8780168466094757e-05,
+      "loss": 1.0115,
+      "step": 1385
+    },
+    {
+      "epoch": 0.10198206484249253,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.877838080667714e-05,
+      "loss": 0.7371,
+      "step": 1386
+    },
+    {
+      "epoch": 0.10205564497585652,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.877659187111873e-05,
+      "loss": 0.659,
+      "step": 1387
+    },
+    {
+      "epoch": 0.10212922510922051,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.877480165951555e-05,
+      "loss": 1.0697,
+      "step": 1388
+    },
+    {
+      "epoch": 0.1022028052425845,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.877301017196366e-05,
+      "loss": 0.9541,
+      "step": 1389
+    },
+    {
+      "epoch": 0.10227638537594849,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.877121740855922e-05,
+      "loss": 0.7799,
+      "step": 1390
+    },
+    {
+      "epoch": 0.10234996550931248,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.876942336939844e-05,
+      "loss": 0.7024,
+      "step": 1391
+    },
+    {
+      "epoch": 0.10242354564267647,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.87676280545776e-05,
+      "loss": 0.9268,
+      "step": 1392
+    },
+    {
+      "epoch": 0.10249712577604046,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.876583146419305e-05,
+      "loss": 0.7755,
+      "step": 1393
+    },
+    {
+      "epoch": 0.10257070590940447,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.8764033598341214e-05,
+      "loss": 0.6314,
+      "step": 1394
+    },
+    {
+      "epoch": 0.10264428604276846,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.876223445711858e-05,
+      "loss": 1.1173,
+      "step": 1395
+    },
+    {
+      "epoch": 0.10271786617613245,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.87604340406217e-05,
+      "loss": 0.6412,
+      "step": 1396
+    },
+    {
+      "epoch": 0.10279144630949644,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.875863234894721e-05,
+      "loss": 0.816,
+      "step": 1397
+    },
+    {
+      "epoch": 0.10286502644286043,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.8756829382191796e-05,
+      "loss": 0.7737,
+      "step": 1398
+    },
+    {
+      "epoch": 0.10293860657622442,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.87550251404522e-05,
+      "loss": 0.8287,
+      "step": 1399
+    },
+    {
+      "epoch": 0.10301218670958841,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.8753219623825296e-05,
+      "loss": 0.8514,
+      "step": 1400
+    },
+    {
+      "epoch": 0.1030857668429524,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.875141283240794e-05,
+      "loss": 1.509,
+      "step": 1401
+    },
+    {
+      "epoch": 0.10315934697631639,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.874960476629713e-05,
+      "loss": 1.1237,
+      "step": 1402
+    },
+    {
+      "epoch": 0.10323292710968039,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.874779542558988e-05,
+      "loss": 0.9198,
+      "step": 1403
+    },
+    {
+      "epoch": 0.10330650724304438,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.8745984810383316e-05,
+      "loss": 0.9819,
+      "step": 1404
+    },
+    {
+      "epoch": 0.10338008737640837,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.874417292077458e-05,
+      "loss": 0.886,
+      "step": 1405
+    },
+    {
+      "epoch": 0.10345366750977236,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.874235975686095e-05,
+      "loss": 1.3282,
+      "step": 1406
+    },
+    {
+      "epoch": 0.10352724764313635,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.874054531873969e-05,
+      "loss": 1.0641,
+      "step": 1407
+    },
+    {
+      "epoch": 0.10360082777650034,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.873872960650822e-05,
+      "loss": 0.8754,
+      "step": 1408
+    },
+    {
+      "epoch": 0.10367440790986433,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.873691262026396e-05,
+      "loss": 1.068,
+      "step": 1409
+    },
+    {
+      "epoch": 0.10374798804322832,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.873509436010444e-05,
+      "loss": 1.0116,
+      "step": 1410
+    },
+    {
+      "epoch": 0.10382156817659231,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.873327482612723e-05,
+      "loss": 0.9322,
+      "step": 1411
+    },
+    {
+      "epoch": 0.10389514830995632,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.873145401842999e-05,
+      "loss": 1.0501,
+      "step": 1412
+    },
+    {
+      "epoch": 0.10396872844332031,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.872963193711042e-05,
+      "loss": 0.9781,
+      "step": 1413
+    },
+    {
+      "epoch": 0.1040423085766843,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.872780858226634e-05,
+      "loss": 0.9769,
+      "step": 1414
+    },
+    {
+      "epoch": 0.10411588871004829,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.872598395399558e-05,
+      "loss": 0.9307,
+      "step": 1415
+    },
+    {
+      "epoch": 0.10418946884341228,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.872415805239607e-05,
+      "loss": 0.9257,
+      "step": 1416
+    },
+    {
+      "epoch": 0.10426304897677627,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.872233087756582e-05,
+      "loss": 1.0647,
+      "step": 1417
+    },
+    {
+      "epoch": 0.10433662911014026,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.872050242960287e-05,
+      "loss": 1.3918,
+      "step": 1418
+    },
+    {
+      "epoch": 0.10441020924350425,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.871867270860535e-05,
+      "loss": 0.8001,
+      "step": 1419
+    },
+    {
+      "epoch": 0.10448378937686825,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.8716841714671465e-05,
+      "loss": 0.587,
+      "step": 1420
+    },
+    {
+      "epoch": 0.10455736951023224,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.871500944789949e-05,
+      "loss": 0.7017,
+      "step": 1421
+    },
+    {
+      "epoch": 0.10463094964359623,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.871317590838774e-05,
+      "loss": 0.853,
+      "step": 1422
+    },
+    {
+      "epoch": 0.10470452977696022,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.871134109623463e-05,
+      "loss": 1.0274,
+      "step": 1423
+    },
+    {
+      "epoch": 0.10477810991032421,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.8709505011538634e-05,
+      "loss": 0.9413,
+      "step": 1424
+    },
+    {
+      "epoch": 0.1048516900436882,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.870766765439827e-05,
+      "loss": 0.7312,
+      "step": 1425
+    },
+    {
+      "epoch": 0.1049252701770522,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.870582902491218e-05,
+      "loss": 0.8642,
+      "step": 1426
+    },
+    {
+      "epoch": 0.10499885031041618,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.8703989123179004e-05,
+      "loss": 0.7064,
+      "step": 1427
+    },
+    {
+      "epoch": 0.10507243044378017,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.87021479492975e-05,
+      "loss": 1.2552,
+      "step": 1428
+    },
+    {
+      "epoch": 0.10514601057714418,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.87003055033665e-05,
+      "loss": 1.1853,
+      "step": 1429
+    },
+    {
+      "epoch": 0.10521959071050817,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.869846178548486e-05,
+      "loss": 1.0046,
+      "step": 1430
+    },
+    {
+      "epoch": 0.10529317084387216,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.869661679575153e-05,
+      "loss": 0.7922,
+      "step": 1431
+    },
+    {
+      "epoch": 0.10536675097723615,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.8694770534265535e-05,
+      "loss": 0.9383,
+      "step": 1432
+    },
+    {
+      "epoch": 0.10544033111060014,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.8692923001125956e-05,
+      "loss": 1.0083,
+      "step": 1433
+    },
+    {
+      "epoch": 0.10551391124396413,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.8691074196431956e-05,
+      "loss": 1.2046,
+      "step": 1434
+    },
+    {
+      "epoch": 0.10558749137732812,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.868922412028275e-05,
+      "loss": 0.8876,
+      "step": 1435
+    },
+    {
+      "epoch": 0.10566107151069211,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.868737277277762e-05,
+      "loss": 0.9315,
+      "step": 1436
+    },
+    {
+      "epoch": 0.1057346516440561,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.868552015401594e-05,
+      "loss": 0.8065,
+      "step": 1437
+    },
+    {
+      "epoch": 0.1058082317774201,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.868366626409713e-05,
+      "loss": 0.7333,
+      "step": 1438
+    },
+    {
+      "epoch": 0.1058818119107841,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.868181110312068e-05,
+      "loss": 1.2215,
+      "step": 1439
+    },
+    {
+      "epoch": 0.10595539204414808,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.867995467118616e-05,
+      "loss": 1.3123,
+      "step": 1440
+    },
+    {
+      "epoch": 0.10602897217751207,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.867809696839319e-05,
+      "loss": 1.0686,
+      "step": 1441
+    },
+    {
+      "epoch": 0.10610255231087606,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.867623799484148e-05,
+      "loss": 1.177,
+      "step": 1442
+    },
+    {
+      "epoch": 0.10617613244424005,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.867437775063079e-05,
+      "loss": 0.6957,
+      "step": 1443
+    },
+    {
+      "epoch": 0.10624971257760404,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.8672516235860975e-05,
+      "loss": 0.8737,
+      "step": 1444
+    },
+    {
+      "epoch": 0.10632329271096803,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.867065345063192e-05,
+      "loss": 1.1496,
+      "step": 1445
+    },
+    {
+      "epoch": 0.10639687284433202,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.86687893950436e-05,
+      "loss": 1.2304,
+      "step": 1446
+    },
+    {
+      "epoch": 0.10647045297769603,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.866692406919605e-05,
+      "loss": 1.3171,
+      "step": 1447
+    },
+    {
+      "epoch": 0.10654403311106002,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.866505747318939e-05,
+      "loss": 1.1928,
+      "step": 1448
+    },
+    {
+      "epoch": 0.10661761324442401,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.8663189607123796e-05,
+      "loss": 1.0115,
+      "step": 1449
+    },
+    {
+      "epoch": 0.106691193377788,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.866132047109951e-05,
+      "loss": 0.7637,
+      "step": 1450
+    },
+    {
+      "epoch": 0.10676477351115199,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.865945006521684e-05,
+      "loss": 0.7481,
+      "step": 1451
+    },
+    {
+      "epoch": 0.10683835364451598,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.8657578389576175e-05,
+      "loss": 0.8649,
+      "step": 1452
+    },
+    {
+      "epoch": 0.10691193377787997,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.865570544427795e-05,
+      "loss": 1.1992,
+      "step": 1453
+    },
+    {
+      "epoch": 0.10698551391124396,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.8653831229422705e-05,
+      "loss": 0.7719,
+      "step": 1454
+    },
+    {
+      "epoch": 0.10705909404460795,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.865195574511101e-05,
+      "loss": 1.3225,
+      "step": 1455
+    },
+    {
+      "epoch": 0.10713267417797195,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.8650078991443523e-05,
+      "loss": 0.9803,
+      "step": 1456
+    },
+    {
+      "epoch": 0.10720625431133594,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.8648200968520965e-05,
+      "loss": 1.5307,
+      "step": 1457
+    },
+    {
+      "epoch": 0.10727983444469993,
+      "grad_norm": 1.125,
+      "learning_rate": 4.864632167644413e-05,
+      "loss": 1.355,
+      "step": 1458
+    },
+    {
+      "epoch": 0.10735341457806392,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.864444111531386e-05,
+      "loss": 1.348,
+      "step": 1459
+    },
+    {
+      "epoch": 0.10742699471142791,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.86425592852311e-05,
+      "loss": 1.061,
+      "step": 1460
+    },
+    {
+      "epoch": 0.1075005748447919,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.8640676186296844e-05,
+      "loss": 1.1889,
+      "step": 1461
+    },
+    {
+      "epoch": 0.1075741549781559,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.863879181861215e-05,
+      "loss": 1.0199,
+      "step": 1462
+    },
+    {
+      "epoch": 0.10764773511151988,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.8636906182278134e-05,
+      "loss": 1.0298,
+      "step": 1463
+    },
+    {
+      "epoch": 0.10772131524488388,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.863501927739601e-05,
+      "loss": 1.0112,
+      "step": 1464
+    },
+    {
+      "epoch": 0.10779489537824788,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.863313110406704e-05,
+      "loss": 1.2276,
+      "step": 1465
+    },
+    {
+      "epoch": 0.10786847551161187,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.863124166239257e-05,
+      "loss": 0.9624,
+      "step": 1466
+    },
+    {
+      "epoch": 0.10794205564497586,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.862935095247398e-05,
+      "loss": 1.678,
+      "step": 1467
+    },
+    {
+      "epoch": 0.10801563577833985,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.862745897441277e-05,
+      "loss": 0.8602,
+      "step": 1468
+    },
+    {
+      "epoch": 0.10808921591170384,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.862556572831045e-05,
+      "loss": 0.8751,
+      "step": 1469
+    },
+    {
+      "epoch": 0.10816279604506783,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.862367121426865e-05,
+      "loss": 0.8047,
+      "step": 1470
+    },
+    {
+      "epoch": 0.10823637617843182,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.862177543238903e-05,
+      "loss": 0.9325,
+      "step": 1471
+    },
+    {
+      "epoch": 0.10830995631179581,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.861987838277333e-05,
+      "loss": 1.3588,
+      "step": 1472
+    },
+    {
+      "epoch": 0.1083835364451598,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.861798006552338e-05,
+      "loss": 1.3279,
+      "step": 1473
+    },
+    {
+      "epoch": 0.1084571165785238,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.861608048074104e-05,
+      "loss": 0.7374,
+      "step": 1474
+    },
+    {
+      "epoch": 0.1085306967118878,
+      "grad_norm": 1.34375,
+      "learning_rate": 4.8614179628528265e-05,
+      "loss": 1.157,
+      "step": 1475
+    },
+    {
+      "epoch": 0.10860427684525178,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.861227750898708e-05,
+      "loss": 0.8502,
+      "step": 1476
+    },
+    {
+      "epoch": 0.10867785697861577,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.861037412221955e-05,
+      "loss": 0.9861,
+      "step": 1477
+    },
+    {
+      "epoch": 0.10875143711197977,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.860846946832783e-05,
+      "loss": 0.9201,
+      "step": 1478
+    },
+    {
+      "epoch": 0.10882501724534376,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.860656354741415e-05,
+      "loss": 1.1797,
+      "step": 1479
+    },
+    {
+      "epoch": 0.10889859737870775,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.860465635958079e-05,
+      "loss": 1.3831,
+      "step": 1480
+    },
+    {
+      "epoch": 0.10897217751207174,
+      "grad_norm": 1.296875,
+      "learning_rate": 4.860274790493011e-05,
+      "loss": 1.5445,
+      "step": 1481
+    },
+    {
+      "epoch": 0.10904575764543573,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.860083818356452e-05,
+      "loss": 0.8569,
+      "step": 1482
+    },
+    {
+      "epoch": 0.10911933777879973,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.859892719558653e-05,
+      "loss": 1.044,
+      "step": 1483
+    },
+    {
+      "epoch": 0.10919291791216372,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.859701494109868e-05,
+      "loss": 0.5431,
+      "step": 1484
+    },
+    {
+      "epoch": 0.10926649804552771,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.859510142020362e-05,
+      "loss": 0.6719,
+      "step": 1485
+    },
+    {
+      "epoch": 0.1093400781788917,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.859318663300402e-05,
+      "loss": 0.8481,
+      "step": 1486
+    },
+    {
+      "epoch": 0.10941365831225569,
+      "grad_norm": 0.75,
+      "learning_rate": 4.859127057960266e-05,
+      "loss": 0.7634,
+      "step": 1487
+    },
+    {
+      "epoch": 0.10948723844561968,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.858935326010237e-05,
+      "loss": 1.5947,
+      "step": 1488
+    },
+    {
+      "epoch": 0.10956081857898367,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.8587434674606036e-05,
+      "loss": 0.7334,
+      "step": 1489
+    },
+    {
+      "epoch": 0.10963439871234766,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.858551482321663e-05,
+      "loss": 1.0087,
+      "step": 1490
+    },
+    {
+      "epoch": 0.10970797884571165,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.85835937060372e-05,
+      "loss": 1.0311,
+      "step": 1491
+    },
+    {
+      "epoch": 0.10978155897907566,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.858167132317083e-05,
+      "loss": 0.7996,
+      "step": 1492
+    },
+    {
+      "epoch": 0.10985513911243965,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.857974767472071e-05,
+      "loss": 0.6179,
+      "step": 1493
+    },
+    {
+      "epoch": 0.10992871924580364,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.857782276079006e-05,
+      "loss": 1.0871,
+      "step": 1494
+    },
+    {
+      "epoch": 0.11000229937916763,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.8575896581482195e-05,
+      "loss": 0.9696,
+      "step": 1495
+    },
+    {
+      "epoch": 0.11007587951253162,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.8573969136900495e-05,
+      "loss": 0.7804,
+      "step": 1496
+    },
+    {
+      "epoch": 0.1101494596458956,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.857204042714839e-05,
+      "loss": 0.8615,
+      "step": 1497
+    },
+    {
+      "epoch": 0.1102230397792596,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.8570110452329395e-05,
+      "loss": 1.0255,
+      "step": 1498
+    },
+    {
+      "epoch": 0.11029661991262359,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.85681792125471e-05,
+      "loss": 0.7892,
+      "step": 1499
+    },
+    {
+      "epoch": 0.11037020004598758,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.856624670790513e-05,
+      "loss": 0.7113,
+      "step": 1500
+    },
+    {
+      "epoch": 0.11044378017935158,
+      "grad_norm": 0.875,
+      "learning_rate": 4.856431293850722e-05,
+      "loss": 0.9847,
+      "step": 1501
+    },
+    {
+      "epoch": 0.11051736031271557,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.856237790445714e-05,
+      "loss": 0.6973,
+      "step": 1502
+    },
+    {
+      "epoch": 0.11059094044607956,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.856044160585872e-05,
+      "loss": 0.7668,
+      "step": 1503
+    },
+    {
+      "epoch": 0.11066452057944355,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.855850404281592e-05,
+      "loss": 0.6934,
+      "step": 1504
+    },
+    {
+      "epoch": 0.11073810071280754,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.85565652154327e-05,
+      "loss": 0.8594,
+      "step": 1505
+    },
+    {
+      "epoch": 0.11081168084617153,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.8554625123813116e-05,
+      "loss": 0.8429,
+      "step": 1506
+    },
+    {
+      "epoch": 0.11088526097953552,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.855268376806129e-05,
+      "loss": 0.9555,
+      "step": 1507
+    },
+    {
+      "epoch": 0.11095884111289951,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.8550741148281406e-05,
+      "loss": 1.3677,
+      "step": 1508
+    },
+    {
+      "epoch": 0.11103242124626352,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.854879726457773e-05,
+      "loss": 1.0277,
+      "step": 1509
+    },
+    {
+      "epoch": 0.1111060013796275,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.854685211705458e-05,
+      "loss": 0.9829,
+      "step": 1510
+    },
+    {
+      "epoch": 0.1111795815129915,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.854490570581635e-05,
+      "loss": 1.2195,
+      "step": 1511
+    },
+    {
+      "epoch": 0.11125316164635549,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.854295803096751e-05,
+      "loss": 0.6009,
+      "step": 1512
+    },
+    {
+      "epoch": 0.11132674177971948,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.854100909261256e-05,
+      "loss": 1.1862,
+      "step": 1513
+    },
+    {
+      "epoch": 0.11140032191308347,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.853905889085613e-05,
+      "loss": 1.0422,
+      "step": 1514
+    },
+    {
+      "epoch": 0.11147390204644746,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.8537107425802865e-05,
+      "loss": 0.7051,
+      "step": 1515
+    },
+    {
+      "epoch": 0.11154748217981145,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.85351546975575e-05,
+      "loss": 0.6874,
+      "step": 1516
+    },
+    {
+      "epoch": 0.11162106231317544,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.853320070622484e-05,
+      "loss": 1.0476,
+      "step": 1517
+    },
+    {
+      "epoch": 0.11169464244653944,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.853124545190973e-05,
+      "loss": 0.7274,
+      "step": 1518
+    },
+    {
+      "epoch": 0.11176822257990343,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.852928893471713e-05,
+      "loss": 0.8882,
+      "step": 1519
+    },
+    {
+      "epoch": 0.11184180271326742,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.8527331154752035e-05,
+      "loss": 1.0539,
+      "step": 1520
+    },
+    {
+      "epoch": 0.11191538284663141,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.852537211211951e-05,
+      "loss": 0.9637,
+      "step": 1521
+    },
+    {
+      "epoch": 0.1119889629799954,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.8523411806924704e-05,
+      "loss": 1.4332,
+      "step": 1522
+    },
+    {
+      "epoch": 0.11206254311335939,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.852145023927281e-05,
+      "loss": 0.7953,
+      "step": 1523
+    },
+    {
+      "epoch": 0.11213612324672338,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.85194874092691e-05,
+      "loss": 0.8669,
+      "step": 1524
+    },
+    {
+      "epoch": 0.11220970338008737,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.851752331701894e-05,
+      "loss": 0.9211,
+      "step": 1525
+    },
+    {
+      "epoch": 0.11228328351345136,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.851555796262771e-05,
+      "loss": 1.3201,
+      "step": 1526
+    },
+    {
+      "epoch": 0.11235686364681537,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.85135913462009e-05,
+      "loss": 0.8635,
+      "step": 1527
+    },
+    {
+      "epoch": 0.11243044378017936,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.851162346784406e-05,
+      "loss": 0.9325,
+      "step": 1528
+    },
+    {
+      "epoch": 0.11250402391354335,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.850965432766279e-05,
+      "loss": 1.066,
+      "step": 1529
+    },
+    {
+      "epoch": 0.11257760404690734,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.850768392576277e-05,
+      "loss": 1.2734,
+      "step": 1530
+    },
+    {
+      "epoch": 0.11265118418027133,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.850571226224976e-05,
+      "loss": 0.8032,
+      "step": 1531
+    },
+    {
+      "epoch": 0.11272476431363532,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.850373933722957e-05,
+      "loss": 0.7884,
+      "step": 1532
+    },
+    {
+      "epoch": 0.11279834444699931,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.8501765150808085e-05,
+      "loss": 0.8609,
+      "step": 1533
+    },
+    {
+      "epoch": 0.1128719245803633,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.849978970309125e-05,
+      "loss": 0.841,
+      "step": 1534
+    },
+    {
+      "epoch": 0.11294550471372729,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.8497812994185075e-05,
+      "loss": 0.8201,
+      "step": 1535
+    },
+    {
+      "epoch": 0.11301908484709129,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.8495835024195665e-05,
+      "loss": 0.8334,
+      "step": 1536
+    },
+    {
+      "epoch": 0.11309266498045528,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.8493855793229174e-05,
+      "loss": 1.2785,
+      "step": 1537
+    },
+    {
+      "epoch": 0.11316624511381927,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.8491875301391806e-05,
+      "loss": 0.6987,
+      "step": 1538
+    },
+    {
+      "epoch": 0.11323982524718326,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.8489893548789874e-05,
+      "loss": 1.2081,
+      "step": 1539
+    },
+    {
+      "epoch": 0.11331340538054725,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.848791053552971e-05,
+      "loss": 0.9546,
+      "step": 1540
+    },
+    {
+      "epoch": 0.11338698551391124,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.848592626171775e-05,
+      "loss": 0.7666,
+      "step": 1541
+    },
+    {
+      "epoch": 0.11346056564727523,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.84839407274605e-05,
+      "loss": 0.8671,
+      "step": 1542
+    },
+    {
+      "epoch": 0.11353414578063922,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.848195393286449e-05,
+      "loss": 1.3558,
+      "step": 1543
+    },
+    {
+      "epoch": 0.11360772591400321,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.8479965878036373e-05,
+      "loss": 0.8777,
+      "step": 1544
+    },
+    {
+      "epoch": 0.11368130604736722,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.8477976563082824e-05,
+      "loss": 0.9407,
+      "step": 1545
+    },
+    {
+      "epoch": 0.11375488618073121,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.847598598811062e-05,
+      "loss": 1.3234,
+      "step": 1546
+    },
+    {
+      "epoch": 0.1138284663140952,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.8473994153226594e-05,
+      "loss": 1.0208,
+      "step": 1547
+    },
+    {
+      "epoch": 0.11390204644745919,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.847200105853763e-05,
+      "loss": 1.0776,
+      "step": 1548
+    },
+    {
+      "epoch": 0.11397562658082318,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.8470006704150705e-05,
+      "loss": 1.1107,
+      "step": 1549
+    },
+    {
+      "epoch": 0.11404920671418717,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.846801109017285e-05,
+      "loss": 0.8734,
+      "step": 1550
+    },
+    {
+      "epoch": 0.11412278684755116,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.846601421671116e-05,
+      "loss": 1.2276,
+      "step": 1551
+    },
+    {
+      "epoch": 0.11419636698091515,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.846401608387282e-05,
+      "loss": 1.2157,
+      "step": 1552
+    },
+    {
+      "epoch": 0.11426994711427914,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.846201669176505e-05,
+      "loss": 0.8714,
+      "step": 1553
+    },
+    {
+      "epoch": 0.11434352724764314,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.846001604049515e-05,
+      "loss": 1.3024,
+      "step": 1554
+    },
+    {
+      "epoch": 0.11441710738100713,
+      "grad_norm": 1.5546875,
+      "learning_rate": 4.8458014130170506e-05,
+      "loss": 0.8927,
+      "step": 1555
+    },
+    {
+      "epoch": 0.11449068751437112,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.845601096089855e-05,
+      "loss": 1.1201,
+      "step": 1556
+    },
+    {
+      "epoch": 0.11456426764773511,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.845400653278678e-05,
+      "loss": 1.013,
+      "step": 1557
+    },
+    {
+      "epoch": 0.1146378477810991,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.845200084594278e-05,
+      "loss": 0.7813,
+      "step": 1558
+    },
+    {
+      "epoch": 0.1147114279144631,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.8449993900474187e-05,
+      "loss": 1.1068,
+      "step": 1559
+    },
+    {
+      "epoch": 0.11478500804782708,
+      "grad_norm": 1.0,
+      "learning_rate": 4.844798569648872e-05,
+      "loss": 1.1226,
+      "step": 1560
+    },
+    {
+      "epoch": 0.11485858818119107,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.844597623409414e-05,
+      "loss": 0.816,
+      "step": 1561
+    },
+    {
+      "epoch": 0.11493216831455506,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.84439655133983e-05,
+      "loss": 0.908,
+      "step": 1562
+    },
+    {
+      "epoch": 0.11500574844791907,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.8441953534509116e-05,
+      "loss": 1.1246,
+      "step": 1563
+    },
+    {
+      "epoch": 0.11507932858128306,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.843994029753456e-05,
+      "loss": 1.2163,
+      "step": 1564
+    },
+    {
+      "epoch": 0.11515290871464705,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.843792580258267e-05,
+      "loss": 1.0191,
+      "step": 1565
+    },
+    {
+      "epoch": 0.11522648884801104,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.843591004976158e-05,
+      "loss": 0.8801,
+      "step": 1566
+    },
+    {
+      "epoch": 0.11530006898137503,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.843389303917946e-05,
+      "loss": 1.2238,
+      "step": 1567
+    },
+    {
+      "epoch": 0.11537364911473902,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.843187477094456e-05,
+      "loss": 0.8479,
+      "step": 1568
+    },
+    {
+      "epoch": 0.11544722924810301,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.84298552451652e-05,
+      "loss": 1.1632,
+      "step": 1569
+    },
+    {
+      "epoch": 0.115520809381467,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.8427834461949764e-05,
+      "loss": 1.2675,
+      "step": 1570
+    },
+    {
+      "epoch": 0.11559438951483099,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.842581242140669e-05,
+      "loss": 0.6874,
+      "step": 1571
+    },
+    {
+      "epoch": 0.115667969648195,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.842378912364452e-05,
+      "loss": 1.2296,
+      "step": 1572
+    },
+    {
+      "epoch": 0.11574154978155898,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.842176456877182e-05,
+      "loss": 1.1198,
+      "step": 1573
+    },
+    {
+      "epoch": 0.11581512991492297,
+      "grad_norm": 1.5,
+      "learning_rate": 4.841973875689726e-05,
+      "loss": 0.8049,
+      "step": 1574
+    },
+    {
+      "epoch": 0.11588871004828696,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.841771168812955e-05,
+      "loss": 0.8588,
+      "step": 1575
+    },
+    {
+      "epoch": 0.11596229018165095,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.841568336257749e-05,
+      "loss": 0.7734,
+      "step": 1576
+    },
+    {
+      "epoch": 0.11603587031501494,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.8413653780349925e-05,
+      "loss": 0.997,
+      "step": 1577
+    },
+    {
+      "epoch": 0.11610945044837893,
+      "grad_norm": 0.875,
+      "learning_rate": 4.841162294155578e-05,
+      "loss": 0.8062,
+      "step": 1578
+    },
+    {
+      "epoch": 0.11618303058174292,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.840959084630406e-05,
+      "loss": 0.8249,
+      "step": 1579
+    },
+    {
+      "epoch": 0.11625661071510691,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.84075574947038e-05,
+      "loss": 0.7494,
+      "step": 1580
+    },
+    {
+      "epoch": 0.11633019084847092,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.840552288686415e-05,
+      "loss": 0.6936,
+      "step": 1581
+    },
+    {
+      "epoch": 0.11640377098183491,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.8403487022894294e-05,
+      "loss": 0.7006,
+      "step": 1582
+    },
+    {
+      "epoch": 0.1164773511151989,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.840144990290349e-05,
+      "loss": 0.9448,
+      "step": 1583
+    },
+    {
+      "epoch": 0.11655093124856289,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.839941152700107e-05,
+      "loss": 1.0342,
+      "step": 1584
+    },
+    {
+      "epoch": 0.11662451138192688,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.839737189529643e-05,
+      "loss": 0.959,
+      "step": 1585
+    },
+    {
+      "epoch": 0.11669809151529087,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.839533100789903e-05,
+      "loss": 0.8676,
+      "step": 1586
+    },
+    {
+      "epoch": 0.11677167164865486,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.83932888649184e-05,
+      "loss": 0.7941,
+      "step": 1587
+    },
+    {
+      "epoch": 0.11684525178201885,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.839124546646414e-05,
+      "loss": 0.6636,
+      "step": 1588
+    },
+    {
+      "epoch": 0.11691883191538284,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.838920081264592e-05,
+      "loss": 1.021,
+      "step": 1589
+    },
+    {
+      "epoch": 0.11699241204874684,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.8387154903573464e-05,
+      "loss": 1.5042,
+      "step": 1590
+    },
+    {
+      "epoch": 0.11706599218211083,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.838510773935657e-05,
+      "loss": 1.0714,
+      "step": 1591
+    },
+    {
+      "epoch": 0.11713957231547482,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.8383059320105124e-05,
+      "loss": 0.6976,
+      "step": 1592
+    },
+    {
+      "epoch": 0.11721315244883881,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.8381009645929044e-05,
+      "loss": 0.8119,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1172867325822028,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.837895871693834e-05,
+      "loss": 0.7397,
+      "step": 1594
+    },
+    {
+      "epoch": 0.1173603127155668,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.837690653324307e-05,
+      "loss": 1.279,
+      "step": 1595
+    },
+    {
+      "epoch": 0.11743389284893078,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.8374853094953384e-05,
+      "loss": 0.7486,
+      "step": 1596
+    },
+    {
+      "epoch": 0.11750747298229477,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.837279840217948e-05,
+      "loss": 0.7038,
+      "step": 1597
+    },
+    {
+      "epoch": 0.11758105311565878,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.837074245503164e-05,
+      "loss": 0.9153,
+      "step": 1598
+    },
+    {
+      "epoch": 0.11765463324902277,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.836868525362018e-05,
+      "loss": 1.0461,
+      "step": 1599
+    },
+    {
+      "epoch": 0.11772821338238676,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.836662679805553e-05,
+      "loss": 0.798,
+      "step": 1600
+    },
+    {
+      "epoch": 0.11780179351575075,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.836456708844814e-05,
+      "loss": 1.1684,
+      "step": 1601
+    },
+    {
+      "epoch": 0.11787537364911474,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.8362506124908574e-05,
+      "loss": 0.8785,
+      "step": 1602
+    },
+    {
+      "epoch": 0.11794895378247873,
+      "grad_norm": 1.0,
+      "learning_rate": 4.8360443907547423e-05,
+      "loss": 1.3117,
+      "step": 1603
+    },
+    {
+      "epoch": 0.11802253391584272,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.835838043647538e-05,
+      "loss": 0.9618,
+      "step": 1604
+    },
+    {
+      "epoch": 0.11809611404920671,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.8356315711803166e-05,
+      "loss": 0.9472,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1181696941825707,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.835424973364161e-05,
+      "loss": 0.805,
+      "step": 1606
+    },
+    {
+      "epoch": 0.1182432743159347,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.835218250210157e-05,
+      "loss": 0.8101,
+      "step": 1607
+    },
+    {
+      "epoch": 0.1183168544492987,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.8350114017294015e-05,
+      "loss": 1.0567,
+      "step": 1608
+    },
+    {
+      "epoch": 0.11839043458266268,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.834804427932993e-05,
+      "loss": 0.9547,
+      "step": 1609
+    },
+    {
+      "epoch": 0.11846401471602667,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.8345973288320414e-05,
+      "loss": 0.9241,
+      "step": 1610
+    },
+    {
+      "epoch": 0.11853759484939067,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.8343901044376604e-05,
+      "loss": 1.2618,
+      "step": 1611
+    },
+    {
+      "epoch": 0.11861117498275466,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.834182754760972e-05,
+      "loss": 1.0232,
+      "step": 1612
+    },
+    {
+      "epoch": 0.11868475511611865,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.8339752798131034e-05,
+      "loss": 1.1661,
+      "step": 1613
+    },
+    {
+      "epoch": 0.11875833524948264,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.8337676796051895e-05,
+      "loss": 0.6603,
+      "step": 1614
+    },
+    {
+      "epoch": 0.11883191538284663,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.8335599541483724e-05,
+      "loss": 0.8218,
+      "step": 1615
+    },
+    {
+      "epoch": 0.11890549551621063,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.8333521034538e-05,
+      "loss": 0.8218,
+      "step": 1616
+    },
+    {
+      "epoch": 0.11897907564957462,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.833144127532627e-05,
+      "loss": 0.8746,
+      "step": 1617
+    },
+    {
+      "epoch": 0.11905265578293861,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.8329360263960155e-05,
+      "loss": 1.3788,
+      "step": 1618
+    },
+    {
+      "epoch": 0.1191262359163026,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.832727800055134e-05,
+      "loss": 0.7577,
+      "step": 1619
+    },
+    {
+      "epoch": 0.11919981604966659,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.8325194485211575e-05,
+      "loss": 1.0625,
+      "step": 1620
+    },
+    {
+      "epoch": 0.11927339618303058,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.8323109718052675e-05,
+      "loss": 1.0207,
+      "step": 1621
+    },
+    {
+      "epoch": 0.11934697631639457,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.832102369918652e-05,
+      "loss": 1.2307,
+      "step": 1622
+    },
+    {
+      "epoch": 0.11942055644975856,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.8318936428725083e-05,
+      "loss": 0.7842,
+      "step": 1623
+    },
+    {
+      "epoch": 0.11949413658312255,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.831684790678036e-05,
+      "loss": 1.0629,
+      "step": 1624
+    },
+    {
+      "epoch": 0.11956771671648656,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.831475813346445e-05,
+      "loss": 0.7109,
+      "step": 1625
+    },
+    {
+      "epoch": 0.11964129684985055,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.831266710888952e-05,
+      "loss": 0.6695,
+      "step": 1626
+    },
+    {
+      "epoch": 0.11971487698321454,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.8310574833167763e-05,
+      "loss": 1.0373,
+      "step": 1627
+    },
+    {
+      "epoch": 0.11978845711657853,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.8308481306411493e-05,
+      "loss": 0.4827,
+      "step": 1628
+    },
+    {
+      "epoch": 0.11986203724994252,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.830638652873305e-05,
+      "loss": 0.8619,
+      "step": 1629
+    },
+    {
+      "epoch": 0.1199356173833065,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.8304290500244855e-05,
+      "loss": 0.8509,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1200091975166705,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.830219322105941e-05,
+      "loss": 0.8503,
+      "step": 1631
+    },
+    {
+      "epoch": 0.12008277765003449,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.830009469128927e-05,
+      "loss": 1.3459,
+      "step": 1632
+    },
+    {
+      "epoch": 0.12015635778339848,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.829799491104705e-05,
+      "loss": 0.8609,
+      "step": 1633
+    },
+    {
+      "epoch": 0.12022993791676248,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.829589388044545e-05,
+      "loss": 1.1231,
+      "step": 1634
+    },
+    {
+      "epoch": 0.12030351805012647,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.829379159959722e-05,
+      "loss": 0.9082,
+      "step": 1635
+    },
+    {
+      "epoch": 0.12037709818349046,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.8291688068615196e-05,
+      "loss": 1.1797,
+      "step": 1636
+    },
+    {
+      "epoch": 0.12045067831685445,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.828958328761226e-05,
+      "loss": 0.9844,
+      "step": 1637
+    },
+    {
+      "epoch": 0.12052425845021844,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.828747725670138e-05,
+      "loss": 0.9034,
+      "step": 1638
+    },
+    {
+      "epoch": 0.12059783858358243,
+      "grad_norm": 1.0,
+      "learning_rate": 4.828536997599559e-05,
+      "loss": 1.5587,
+      "step": 1639
+    },
+    {
+      "epoch": 0.12067141871694642,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.828326144560795e-05,
+      "loss": 1.0669,
+      "step": 1640
+    },
+    {
+      "epoch": 0.12074499885031041,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.8281151665651664e-05,
+      "loss": 0.9971,
+      "step": 1641
+    },
+    {
+      "epoch": 0.1208185789836744,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.8279040636239934e-05,
+      "loss": 0.8416,
+      "step": 1642
+    },
+    {
+      "epoch": 0.1208921591170384,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.8276928357486056e-05,
+      "loss": 0.9918,
+      "step": 1643
+    },
+    {
+      "epoch": 0.1209657392504024,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.82748148295034e-05,
+      "loss": 0.9947,
+      "step": 1644
+    },
+    {
+      "epoch": 0.12103931938376639,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.8272700052405396e-05,
+      "loss": 0.8783,
+      "step": 1645
+    },
+    {
+      "epoch": 0.12111289951713038,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.827058402630553e-05,
+      "loss": 1.3422,
+      "step": 1646
+    },
+    {
+      "epoch": 0.12118647965049437,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.826846675131738e-05,
+      "loss": 0.8693,
+      "step": 1647
+    },
+    {
+      "epoch": 0.12126005978385836,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.826634822755456e-05,
+      "loss": 1.0177,
+      "step": 1648
+    },
+    {
+      "epoch": 0.12133363991722235,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.826422845513077e-05,
+      "loss": 1.4612,
+      "step": 1649
+    },
+    {
+      "epoch": 0.12140722005058634,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.826210743415979e-05,
+      "loss": 1.0388,
+      "step": 1650
+    },
+    {
+      "epoch": 0.12148080018395033,
+      "grad_norm": 1.0,
+      "learning_rate": 4.825998516475543e-05,
+      "loss": 1.1511,
+      "step": 1651
+    },
+    {
+      "epoch": 0.12155438031731433,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.8257861647031607e-05,
+      "loss": 1.3444,
+      "step": 1652
+    },
+    {
+      "epoch": 0.12162796045067832,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.825573688110227e-05,
+      "loss": 1.3203,
+      "step": 1653
+    },
+    {
+      "epoch": 0.12170154058404231,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.825361086708146e-05,
+      "loss": 0.9841,
+      "step": 1654
+    },
+    {
+      "epoch": 0.1217751207174063,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.825148360508328e-05,
+      "loss": 0.7331,
+      "step": 1655
+    },
+    {
+      "epoch": 0.12184870085077029,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.824935509522188e-05,
+      "loss": 0.7633,
+      "step": 1656
+    },
+    {
+      "epoch": 0.12192228098413428,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.824722533761151e-05,
+      "loss": 0.944,
+      "step": 1657
+    },
+    {
+      "epoch": 0.12199586111749827,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.824509433236647e-05,
+      "loss": 1.0866,
+      "step": 1658
+    },
+    {
+      "epoch": 0.12206944125086226,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.8242962079601115e-05,
+      "loss": 1.2952,
+      "step": 1659
+    },
+    {
+      "epoch": 0.12214302138422625,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.8240828579429886e-05,
+      "loss": 1.3609,
+      "step": 1660
+    },
+    {
+      "epoch": 0.12221660151759026,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.823869383196728e-05,
+      "loss": 0.7113,
+      "step": 1661
+    },
+    {
+      "epoch": 0.12229018165095425,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.823655783732787e-05,
+      "loss": 1.1002,
+      "step": 1662
+    },
+    {
+      "epoch": 0.12236376178431824,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.8234420595626286e-05,
+      "loss": 0.9645,
+      "step": 1663
+    },
+    {
+      "epoch": 0.12243734191768223,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.823228210697723e-05,
+      "loss": 1.3537,
+      "step": 1664
+    },
+    {
+      "epoch": 0.12251092205104622,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.823014237149548e-05,
+      "loss": 1.0574,
+      "step": 1665
+    },
+    {
+      "epoch": 0.12258450218441021,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.8228001389295864e-05,
+      "loss": 0.9434,
+      "step": 1666
+    },
+    {
+      "epoch": 0.1226580823177742,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.822585916049328e-05,
+      "loss": 0.8898,
+      "step": 1667
+    },
+    {
+      "epoch": 0.12273166245113819,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.8223715685202706e-05,
+      "loss": 1.2811,
+      "step": 1668
+    },
+    {
+      "epoch": 0.12280524258450218,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.8221570963539175e-05,
+      "loss": 0.6959,
+      "step": 1669
+    },
+    {
+      "epoch": 0.12287882271786618,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.821942499561779e-05,
+      "loss": 0.7751,
+      "step": 1670
+    },
+    {
+      "epoch": 0.12295240285123017,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.8217277781553716e-05,
+      "loss": 0.8852,
+      "step": 1671
+    },
+    {
+      "epoch": 0.12302598298459416,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.82151293214622e-05,
+      "loss": 1.1222,
+      "step": 1672
+    },
+    {
+      "epoch": 0.12309956311795815,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.8212979615458534e-05,
+      "loss": 0.9906,
+      "step": 1673
+    },
+    {
+      "epoch": 0.12317314325132214,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.8210828663658106e-05,
+      "loss": 0.9114,
+      "step": 1674
+    },
+    {
+      "epoch": 0.12324672338468613,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.8208676466176325e-05,
+      "loss": 0.7124,
+      "step": 1675
+    },
+    {
+      "epoch": 0.12332030351805012,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.820652302312873e-05,
+      "loss": 0.8773,
+      "step": 1676
+    },
+    {
+      "epoch": 0.12339388365141411,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.820436833463087e-05,
+      "loss": 0.712,
+      "step": 1677
+    },
+    {
+      "epoch": 0.1234674637847781,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.820221240079838e-05,
+      "loss": 0.8774,
+      "step": 1678
+    },
+    {
+      "epoch": 0.12354104391814211,
+      "grad_norm": 0.75,
+      "learning_rate": 4.820005522174699e-05,
+      "loss": 0.8815,
+      "step": 1679
+    },
+    {
+      "epoch": 0.1236146240515061,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.8197896797592444e-05,
+      "loss": 1.206,
+      "step": 1680
+    },
+    {
+      "epoch": 0.12368820418487009,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.819573712845059e-05,
+      "loss": 1.4705,
+      "step": 1681
+    },
+    {
+      "epoch": 0.12376178431823408,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.819357621443734e-05,
+      "loss": 0.8625,
+      "step": 1682
+    },
+    {
+      "epoch": 0.12383536445159807,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.819141405566866e-05,
+      "loss": 0.8718,
+      "step": 1683
+    },
+    {
+      "epoch": 0.12390894458496206,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.818925065226059e-05,
+      "loss": 0.7995,
+      "step": 1684
+    },
+    {
+      "epoch": 0.12398252471832605,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.818708600432923e-05,
+      "loss": 0.8796,
+      "step": 1685
+    },
+    {
+      "epoch": 0.12405610485169004,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.818492011199076e-05,
+      "loss": 1.2425,
+      "step": 1686
+    },
+    {
+      "epoch": 0.12412968498505403,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.8182752975361425e-05,
+      "loss": 0.6448,
+      "step": 1687
+    },
+    {
+      "epoch": 0.12420326511841803,
+      "grad_norm": 1.0,
+      "learning_rate": 4.818058459455752e-05,
+      "loss": 1.1012,
+      "step": 1688
+    },
+    {
+      "epoch": 0.12427684525178202,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.817841496969542e-05,
+      "loss": 1.0217,
+      "step": 1689
+    },
+    {
+      "epoch": 0.12435042538514601,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.8176244100891566e-05,
+      "loss": 0.9944,
+      "step": 1690
+    },
+    {
+      "epoch": 0.12442400551851,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.817407198826247e-05,
+      "loss": 0.8699,
+      "step": 1691
+    },
+    {
+      "epoch": 0.124497585651874,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.8171898631924695e-05,
+      "loss": 0.7512,
+      "step": 1692
+    },
+    {
+      "epoch": 0.12457116578523798,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.816972403199489e-05,
+      "loss": 0.8817,
+      "step": 1693
+    },
+    {
+      "epoch": 0.12464474591860197,
+      "grad_norm": 1.0,
+      "learning_rate": 4.816754818858975e-05,
+      "loss": 1.0115,
+      "step": 1694
+    },
+    {
+      "epoch": 0.12471832605196596,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.8165371101826064e-05,
+      "loss": 0.8853,
+      "step": 1695
+    },
+    {
+      "epoch": 0.12479190618532997,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.8163192771820655e-05,
+      "loss": 1.1001,
+      "step": 1696
+    },
+    {
+      "epoch": 0.12486548631869396,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.816101319869045e-05,
+      "loss": 0.6952,
+      "step": 1697
+    },
+    {
+      "epoch": 0.12493906645205795,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.8158832382552406e-05,
+      "loss": 0.6278,
+      "step": 1698
+    },
+    {
+      "epoch": 0.12501264658542194,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.8156650323523566e-05,
+      "loss": 0.7689,
+      "step": 1699
+    },
+    {
+      "epoch": 0.12508622671878591,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.8154467021721046e-05,
+      "loss": 0.9663,
+      "step": 1700
+    },
+    {
+      "epoch": 0.12515980685214992,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.815228247726201e-05,
+      "loss": 1.3419,
+      "step": 1701
+    },
+    {
+      "epoch": 0.12523338698551392,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.8150096690263704e-05,
+      "loss": 1.0555,
+      "step": 1702
+    },
+    {
+      "epoch": 0.1253069671188779,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.814790966084343e-05,
+      "loss": 0.8885,
+      "step": 1703
+    },
+    {
+      "epoch": 0.1253805472522419,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.814572138911857e-05,
+      "loss": 0.733,
+      "step": 1704
+    },
+    {
+      "epoch": 0.12545412738560588,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.8143531875206546e-05,
+      "loss": 1.1117,
+      "step": 1705
+    },
+    {
+      "epoch": 0.12552770751896988,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.814134111922489e-05,
+      "loss": 0.9932,
+      "step": 1706
+    },
+    {
+      "epoch": 0.12560128765233386,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.813914912129116e-05,
+      "loss": 1.3604,
+      "step": 1707
+    },
+    {
+      "epoch": 0.12567486778569786,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.8136955881523004e-05,
+      "loss": 0.8963,
+      "step": 1708
+    },
+    {
+      "epoch": 0.12574844791906184,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.813476140003812e-05,
+      "loss": 0.7431,
+      "step": 1709
+    },
+    {
+      "epoch": 0.12582202805242584,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.813256567695429e-05,
+      "loss": 1.0601,
+      "step": 1710
+    },
+    {
+      "epoch": 0.12589560818578985,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.813036871238935e-05,
+      "loss": 0.9191,
+      "step": 1711
+    },
+    {
+      "epoch": 0.12596918831915382,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.8128170506461215e-05,
+      "loss": 1.2271,
+      "step": 1712
+    },
+    {
+      "epoch": 0.12604276845251783,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.812597105928784e-05,
+      "loss": 1.1664,
+      "step": 1713
+    },
+    {
+      "epoch": 0.1261163485858818,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.8123770370987284e-05,
+      "loss": 1.2332,
+      "step": 1714
+    },
+    {
+      "epoch": 0.1261899287192458,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.8121568441677656e-05,
+      "loss": 0.8885,
+      "step": 1715
+    },
+    {
+      "epoch": 0.12626350885260978,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.8119365271477105e-05,
+      "loss": 0.9694,
+      "step": 1716
+    },
+    {
+      "epoch": 0.1263370889859738,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.811716086050389e-05,
+      "loss": 1.0888,
+      "step": 1717
+    },
+    {
+      "epoch": 0.12641066911933777,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.8114955208876325e-05,
+      "loss": 0.9996,
+      "step": 1718
+    },
+    {
+      "epoch": 0.12648424925270177,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.811274831671275e-05,
+      "loss": 1.2635,
+      "step": 1719
+    },
+    {
+      "epoch": 0.12655782938606577,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.811054018413165e-05,
+      "loss": 0.9658,
+      "step": 1720
+    },
+    {
+      "epoch": 0.12663140951942975,
+      "grad_norm": 1.0,
+      "learning_rate": 4.810833081125149e-05,
+      "loss": 1.4049,
+      "step": 1721
+    },
+    {
+      "epoch": 0.12670498965279375,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.810612019819087e-05,
+      "loss": 0.8666,
+      "step": 1722
+    },
+    {
+      "epoch": 0.12677856978615773,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.8103908345068416e-05,
+      "loss": 0.773,
+      "step": 1723
+    },
+    {
+      "epoch": 0.12685214991952173,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.8101695252002846e-05,
+      "loss": 0.7949,
+      "step": 1724
+    },
+    {
+      "epoch": 0.1269257300528857,
+      "grad_norm": 1.4375,
+      "learning_rate": 4.809948091911292e-05,
+      "loss": 0.8907,
+      "step": 1725
+    },
+    {
+      "epoch": 0.12699931018624971,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.8097265346517474e-05,
+      "loss": 0.8715,
+      "step": 1726
+    },
+    {
+      "epoch": 0.1270728903196137,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.809504853433543e-05,
+      "loss": 1.3757,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1271464704529777,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.809283048268575e-05,
+      "loss": 1.327,
+      "step": 1728
+    },
+    {
+      "epoch": 0.1272200505863417,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.809061119168747e-05,
+      "loss": 0.8474,
+      "step": 1729
+    },
+    {
+      "epoch": 0.12729363071970567,
+      "grad_norm": 0.75,
+      "learning_rate": 4.80883906614597e-05,
+      "loss": 0.6874,
+      "step": 1730
+    },
+    {
+      "epoch": 0.12736721085306968,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.808616889212162e-05,
+      "loss": 0.9943,
+      "step": 1731
+    },
+    {
+      "epoch": 0.12744079098643366,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.808394588379245e-05,
+      "loss": 1.1231,
+      "step": 1732
+    },
+    {
+      "epoch": 0.12751437111979766,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.80817216365915e-05,
+      "loss": 1.0542,
+      "step": 1733
+    },
+    {
+      "epoch": 0.12758795125316164,
+      "grad_norm": 1.0,
+      "learning_rate": 4.807949615063816e-05,
+      "loss": 0.9884,
+      "step": 1734
+    },
+    {
+      "epoch": 0.12766153138652564,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.807726942605184e-05,
+      "loss": 0.8242,
+      "step": 1735
+    },
+    {
+      "epoch": 0.12773511151988964,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.8075041462952066e-05,
+      "loss": 0.9871,
+      "step": 1736
+    },
+    {
+      "epoch": 0.12780869165325362,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.807281226145839e-05,
+      "loss": 0.5976,
+      "step": 1737
+    },
+    {
+      "epoch": 0.12788227178661762,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.807058182169046e-05,
+      "loss": 0.8272,
+      "step": 1738
+    },
+    {
+      "epoch": 0.1279558519199816,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.8068350143767985e-05,
+      "loss": 1.5913,
+      "step": 1739
+    },
+    {
+      "epoch": 0.1280294320533456,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.806611722781073e-05,
+      "loss": 0.8145,
+      "step": 1740
+    },
+    {
+      "epoch": 0.12810301218670958,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.8063883073938515e-05,
+      "loss": 1.3143,
+      "step": 1741
+    },
+    {
+      "epoch": 0.12817659232007358,
+      "grad_norm": 1.3359375,
+      "learning_rate": 4.8061647682271266e-05,
+      "loss": 0.7281,
+      "step": 1742
+    },
+    {
+      "epoch": 0.12825017245343756,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.805941105292894e-05,
+      "loss": 1.1643,
+      "step": 1743
+    },
+    {
+      "epoch": 0.12832375258680156,
+      "grad_norm": 1.0,
+      "learning_rate": 4.805717318603158e-05,
+      "loss": 1.0296,
+      "step": 1744
+    },
+    {
+      "epoch": 0.12839733272016557,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.8054934081699275e-05,
+      "loss": 0.8484,
+      "step": 1745
+    },
+    {
+      "epoch": 0.12847091285352955,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.8052693740052214e-05,
+      "loss": 1.1198,
+      "step": 1746
+    },
+    {
+      "epoch": 0.12854449298689355,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.8050452161210616e-05,
+      "loss": 0.9856,
+      "step": 1747
+    },
+    {
+      "epoch": 0.12861807312025753,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.804820934529478e-05,
+      "loss": 0.7506,
+      "step": 1748
+    },
+    {
+      "epoch": 0.12869165325362153,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.8045965292425085e-05,
+      "loss": 1.194,
+      "step": 1749
+    },
+    {
+      "epoch": 0.1287652333869855,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.804372000272196e-05,
+      "loss": 0.8777,
+      "step": 1750
+    },
+    {
+      "epoch": 0.1288388135203495,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.804147347630591e-05,
+      "loss": 0.9836,
+      "step": 1751
+    },
+    {
+      "epoch": 0.1289123936537135,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.80392257132975e-05,
+      "loss": 0.7982,
+      "step": 1752
+    },
+    {
+      "epoch": 0.1289859737870775,
+      "grad_norm": 1.125,
+      "learning_rate": 4.803697671381735e-05,
+      "loss": 1.4899,
+      "step": 1753
+    },
+    {
+      "epoch": 0.1290595539204415,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.8034726477986175e-05,
+      "loss": 0.7501,
+      "step": 1754
+    },
+    {
+      "epoch": 0.12913313405380547,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.8032475005924734e-05,
+      "loss": 1.5302,
+      "step": 1755
+    },
+    {
+      "epoch": 0.12920671418716947,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.8030222297753867e-05,
+      "loss": 0.7802,
+      "step": 1756
+    },
+    {
+      "epoch": 0.12928029432053345,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.802796835359447e-05,
+      "loss": 0.7931,
+      "step": 1757
+    },
+    {
+      "epoch": 0.12935387445389745,
+      "grad_norm": 0.875,
+      "learning_rate": 4.80257131735675e-05,
+      "loss": 0.7262,
+      "step": 1758
+    },
+    {
+      "epoch": 0.12942745458726143,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.8023456757793986e-05,
+      "loss": 1.3814,
+      "step": 1759
+    },
+    {
+      "epoch": 0.12950103472062544,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.802119910639504e-05,
+      "loss": 0.7002,
+      "step": 1760
+    },
+    {
+      "epoch": 0.1295746148539894,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.801894021949183e-05,
+      "loss": 1.6328,
+      "step": 1761
+    },
+    {
+      "epoch": 0.12964819498735342,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.801668009720556e-05,
+      "loss": 0.9086,
+      "step": 1762
+    },
+    {
+      "epoch": 0.12972177512071742,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.801441873965754e-05,
+      "loss": 0.9095,
+      "step": 1763
+    },
+    {
+      "epoch": 0.1297953552540814,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.801215614696915e-05,
+      "loss": 0.8858,
+      "step": 1764
+    },
+    {
+      "epoch": 0.1298689353874454,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.800989231926178e-05,
+      "loss": 0.9846,
+      "step": 1765
+    },
+    {
+      "epoch": 0.12994251552080938,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.8007627256656965e-05,
+      "loss": 1.1344,
+      "step": 1766
+    },
+    {
+      "epoch": 0.13001609565417338,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.8005360959276255e-05,
+      "loss": 0.721,
+      "step": 1767
+    },
+    {
+      "epoch": 0.13008967578753736,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.8003093427241266e-05,
+      "loss": 0.9318,
+      "step": 1768
+    },
+    {
+      "epoch": 0.13016325592090136,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.800082466067369e-05,
+      "loss": 1.6782,
+      "step": 1769
+    },
+    {
+      "epoch": 0.13023683605426534,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.799855465969531e-05,
+      "loss": 0.799,
+      "step": 1770
+    },
+    {
+      "epoch": 0.13031041618762934,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.7996283424427935e-05,
+      "loss": 1.383,
+      "step": 1771
+    },
+    {
+      "epoch": 0.13038399632099335,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.7994010954993465e-05,
+      "loss": 1.0181,
+      "step": 1772
+    },
+    {
+      "epoch": 0.13045757645435732,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.7991737251513855e-05,
+      "loss": 0.9318,
+      "step": 1773
+    },
+    {
+      "epoch": 0.13053115658772133,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.798946231411113e-05,
+      "loss": 1.2828,
+      "step": 1774
+    },
+    {
+      "epoch": 0.1306047367210853,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.798718614290739e-05,
+      "loss": 0.8909,
+      "step": 1775
+    },
+    {
+      "epoch": 0.1306783168544493,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.798490873802478e-05,
+      "loss": 0.9724,
+      "step": 1776
+    },
+    {
+      "epoch": 0.13075189698781328,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.798263009958553e-05,
+      "loss": 0.7469,
+      "step": 1777
+    },
+    {
+      "epoch": 0.13082547712117729,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.7980350227711935e-05,
+      "loss": 0.9373,
+      "step": 1778
+    },
+    {
+      "epoch": 0.13089905725454126,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.7978069122526334e-05,
+      "loss": 0.8338,
+      "step": 1779
+    },
+    {
+      "epoch": 0.13097263738790527,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.797578678415118e-05,
+      "loss": 0.6374,
+      "step": 1780
+    },
+    {
+      "epoch": 0.13104621752126927,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.797350321270894e-05,
+      "loss": 0.9759,
+      "step": 1781
+    },
+    {
+      "epoch": 0.13111979765463325,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.7971218408322175e-05,
+      "loss": 1.3983,
+      "step": 1782
+    },
+    {
+      "epoch": 0.13119337778799725,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.7968932371113495e-05,
+      "loss": 1.0449,
+      "step": 1783
+    },
+    {
+      "epoch": 0.13126695792136123,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.7966645101205596e-05,
+      "loss": 0.7516,
+      "step": 1784
+    },
+    {
+      "epoch": 0.13134053805472523,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.7964356598721245e-05,
+      "loss": 0.7642,
+      "step": 1785
+    },
+    {
+      "epoch": 0.1314141181880892,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.796206686378324e-05,
+      "loss": 0.8153,
+      "step": 1786
+    },
+    {
+      "epoch": 0.1314876983214532,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.7959775896514484e-05,
+      "loss": 1.0266,
+      "step": 1787
+    },
+    {
+      "epoch": 0.1315612784548172,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.795748369703791e-05,
+      "loss": 0.9504,
+      "step": 1788
+    },
+    {
+      "epoch": 0.1316348585881812,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.795519026547655e-05,
+      "loss": 0.7396,
+      "step": 1789
+    },
+    {
+      "epoch": 0.1317084387215452,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.7952895601953484e-05,
+      "loss": 0.885,
+      "step": 1790
+    },
+    {
+      "epoch": 0.13178201885490917,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.795059970659187e-05,
+      "loss": 0.8995,
+      "step": 1791
+    },
+    {
+      "epoch": 0.13185559898827318,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.7948302579514906e-05,
+      "loss": 0.7972,
+      "step": 1792
+    },
+    {
+      "epoch": 0.13192917912163715,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.794600422084589e-05,
+      "loss": 0.7332,
+      "step": 1793
+    },
+    {
+      "epoch": 0.13200275925500116,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.7943704630708163e-05,
+      "loss": 1.256,
+      "step": 1794
+    },
+    {
+      "epoch": 0.13207633938836513,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.794140380922515e-05,
+      "loss": 1.0226,
+      "step": 1795
+    },
+    {
+      "epoch": 0.13214991952172914,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.793910175652032e-05,
+      "loss": 0.8569,
+      "step": 1796
+    },
+    {
+      "epoch": 0.1322234996550931,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.793679847271723e-05,
+      "loss": 1.2505,
+      "step": 1797
+    },
+    {
+      "epoch": 0.13229707978845712,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.7934493957939485e-05,
+      "loss": 0.8825,
+      "step": 1798
+    },
+    {
+      "epoch": 0.13237065992182112,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.793218821231077e-05,
+      "loss": 0.9187,
+      "step": 1799
+    },
+    {
+      "epoch": 0.1324442400551851,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.792988123595482e-05,
+      "loss": 1.1352,
+      "step": 1800
+    },
+    {
+      "epoch": 0.1325178201885491,
+      "grad_norm": 1.0,
+      "learning_rate": 4.7927573028995453e-05,
+      "loss": 1.1833,
+      "step": 1801
+    },
+    {
+      "epoch": 0.13259140032191308,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.792526359155655e-05,
+      "loss": 0.7917,
+      "step": 1802
+    },
+    {
+      "epoch": 0.13266498045527708,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.792295292376204e-05,
+      "loss": 0.9438,
+      "step": 1803
+    },
+    {
+      "epoch": 0.13273856058864106,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.792064102573595e-05,
+      "loss": 0.6914,
+      "step": 1804
+    },
+    {
+      "epoch": 0.13281214072200506,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.7918327897602344e-05,
+      "loss": 0.9107,
+      "step": 1805
+    },
+    {
+      "epoch": 0.13288572085536904,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.791601353948537e-05,
+      "loss": 1.0726,
+      "step": 1806
+    },
+    {
+      "epoch": 0.13295930098873304,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.791369795150923e-05,
+      "loss": 0.9249,
+      "step": 1807
+    },
+    {
+      "epoch": 0.13303288112209705,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.7911381133798197e-05,
+      "loss": 0.9071,
+      "step": 1808
+    },
+    {
+      "epoch": 0.13310646125546102,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.790906308647661e-05,
+      "loss": 1.2928,
+      "step": 1809
+    },
+    {
+      "epoch": 0.13318004138882503,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.7906743809668885e-05,
+      "loss": 1.1867,
+      "step": 1810
+    },
+    {
+      "epoch": 0.133253621522189,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.790442330349948e-05,
+      "loss": 0.6382,
+      "step": 1811
+    },
+    {
+      "epoch": 0.133327201655553,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.7902101568092935e-05,
+      "loss": 1.0395,
+      "step": 1812
+    },
+    {
+      "epoch": 0.13340078178891698,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.789977860357385e-05,
+      "loss": 1.1669,
+      "step": 1813
+    },
+    {
+      "epoch": 0.133474361922281,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.789745441006691e-05,
+      "loss": 1.3609,
+      "step": 1814
+    },
+    {
+      "epoch": 0.13354794205564496,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.7895128987696834e-05,
+      "loss": 0.9618,
+      "step": 1815
+    },
+    {
+      "epoch": 0.13362152218900897,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.7892802336588425e-05,
+      "loss": 0.74,
+      "step": 1816
+    },
+    {
+      "epoch": 0.13369510232237297,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.789047445686656e-05,
+      "loss": 0.9331,
+      "step": 1817
+    },
+    {
+      "epoch": 0.13376868245573695,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.788814534865615e-05,
+      "loss": 0.9501,
+      "step": 1818
+    },
+    {
+      "epoch": 0.13384226258910095,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.788581501208222e-05,
+      "loss": 1.0475,
+      "step": 1819
+    },
+    {
+      "epoch": 0.13391584272246493,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.788348344726983e-05,
+      "loss": 0.8627,
+      "step": 1820
+    },
+    {
+      "epoch": 0.13398942285582893,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.7881150654344093e-05,
+      "loss": 0.8419,
+      "step": 1821
+    },
+    {
+      "epoch": 0.1340630029891929,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.787881663343022e-05,
+      "loss": 0.8597,
+      "step": 1822
+    },
+    {
+      "epoch": 0.1341365831225569,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.787648138465347e-05,
+      "loss": 1.136,
+      "step": 1823
+    },
+    {
+      "epoch": 0.1342101632559209,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.7874144908139175e-05,
+      "loss": 0.7524,
+      "step": 1824
+    },
+    {
+      "epoch": 0.1342837433892849,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.787180720401272e-05,
+      "loss": 0.9548,
+      "step": 1825
+    },
+    {
+      "epoch": 0.1343573235226489,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.7869468272399574e-05,
+      "loss": 0.8754,
+      "step": 1826
+    },
+    {
+      "epoch": 0.13443090365601287,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.7867128113425265e-05,
+      "loss": 0.8539,
+      "step": 1827
+    },
+    {
+      "epoch": 0.13450448378937688,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.786478672721537e-05,
+      "loss": 0.7618,
+      "step": 1828
+    },
+    {
+      "epoch": 0.13457806392274085,
+      "grad_norm": 1.0,
+      "learning_rate": 4.7862444113895565e-05,
+      "loss": 0.9876,
+      "step": 1829
+    },
+    {
+      "epoch": 0.13465164405610486,
+      "grad_norm": 0.75,
+      "learning_rate": 4.786010027359156e-05,
+      "loss": 0.8211,
+      "step": 1830
+    },
+    {
+      "epoch": 0.13472522418946883,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.785775520642916e-05,
+      "loss": 1.0766,
+      "step": 1831
+    },
+    {
+      "epoch": 0.13479880432283284,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.785540891253419e-05,
+      "loss": 1.0317,
+      "step": 1832
+    },
+    {
+      "epoch": 0.13487238445619681,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.78530613920326e-05,
+      "loss": 1.0577,
+      "step": 1833
+    },
+    {
+      "epoch": 0.13494596458956082,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.785071264505038e-05,
+      "loss": 0.6501,
+      "step": 1834
+    },
+    {
+      "epoch": 0.13501954472292482,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.784836267171356e-05,
+      "loss": 0.8578,
+      "step": 1835
+    },
+    {
+      "epoch": 0.1350931248562888,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.784601147214828e-05,
+      "loss": 1.4335,
+      "step": 1836
+    },
+    {
+      "epoch": 0.1351667049896528,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.78436590464807e-05,
+      "loss": 0.7884,
+      "step": 1837
+    },
+    {
+      "epoch": 0.13524028512301678,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.7841305394837096e-05,
+      "loss": 0.9455,
+      "step": 1838
+    },
+    {
+      "epoch": 0.13531386525638078,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.7838950517343774e-05,
+      "loss": 0.9894,
+      "step": 1839
+    },
+    {
+      "epoch": 0.13538744538974476,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.78365944141271e-05,
+      "loss": 0.7096,
+      "step": 1840
+    },
+    {
+      "epoch": 0.13546102552310876,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.783423708531355e-05,
+      "loss": 0.9349,
+      "step": 1841
+    },
+    {
+      "epoch": 0.13553460565647274,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.783187853102962e-05,
+      "loss": 0.9142,
+      "step": 1842
+    },
+    {
+      "epoch": 0.13560818578983674,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.7829518751401893e-05,
+      "loss": 0.7758,
+      "step": 1843
+    },
+    {
+      "epoch": 0.13568176592320075,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.7827157746557e-05,
+      "loss": 1.3119,
+      "step": 1844
+    },
+    {
+      "epoch": 0.13575534605656472,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.782479551662168e-05,
+      "loss": 1.2775,
+      "step": 1845
+    },
+    {
+      "epoch": 0.13582892618992873,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.7822432061722693e-05,
+      "loss": 0.9439,
+      "step": 1846
+    },
+    {
+      "epoch": 0.1359025063232927,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.7820067381986885e-05,
+      "loss": 1.2515,
+      "step": 1847
+    },
+    {
+      "epoch": 0.1359760864566567,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.7817701477541154e-05,
+      "loss": 0.8253,
+      "step": 1848
+    },
+    {
+      "epoch": 0.13604966659002068,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.781533434851249e-05,
+      "loss": 0.6957,
+      "step": 1849
+    },
+    {
+      "epoch": 0.1361232467233847,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.7812965995027915e-05,
+      "loss": 1.0547,
+      "step": 1850
+    },
+    {
+      "epoch": 0.13619682685674867,
+      "grad_norm": 1.3359375,
+      "learning_rate": 4.7810596417214543e-05,
+      "loss": 1.3689,
+      "step": 1851
+    },
+    {
+      "epoch": 0.13627040699011267,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.780822561519955e-05,
+      "loss": 1.0853,
+      "step": 1852
+    },
+    {
+      "epoch": 0.13634398712347667,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.780585358911016e-05,
+      "loss": 0.9258,
+      "step": 1853
+    },
+    {
+      "epoch": 0.13641756725684065,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.780348033907369e-05,
+      "loss": 0.8984,
+      "step": 1854
+    },
+    {
+      "epoch": 0.13649114739020465,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.78011058652175e-05,
+      "loss": 1.2386,
+      "step": 1855
+    },
+    {
+      "epoch": 0.13656472752356863,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.7798730167669016e-05,
+      "loss": 0.9408,
+      "step": 1856
+    },
+    {
+      "epoch": 0.13663830765693263,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.7796353246555746e-05,
+      "loss": 0.6505,
+      "step": 1857
+    },
+    {
+      "epoch": 0.1367118877902966,
+      "grad_norm": 1.25,
+      "learning_rate": 4.779397510200525e-05,
+      "loss": 1.2214,
+      "step": 1858
+    },
+    {
+      "epoch": 0.13678546792366061,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.7791595734145164e-05,
+      "loss": 0.976,
+      "step": 1859
+    },
+    {
+      "epoch": 0.1368590480570246,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.7789215143103195e-05,
+      "loss": 0.8904,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1369326281903886,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.778683332900708e-05,
+      "loss": 1.0398,
+      "step": 1861
+    },
+    {
+      "epoch": 0.1370062083237526,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.7784450291984664e-05,
+      "loss": 0.9263,
+      "step": 1862
+    },
+    {
+      "epoch": 0.13707978845711657,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.778206603216383e-05,
+      "loss": 0.9906,
+      "step": 1863
+    },
+    {
+      "epoch": 0.13715336859048058,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.777968054967254e-05,
+      "loss": 0.9936,
+      "step": 1864
+    },
+    {
+      "epoch": 0.13722694872384456,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.777729384463882e-05,
+      "loss": 0.7839,
+      "step": 1865
+    },
+    {
+      "epoch": 0.13730052885720856,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.777490591719076e-05,
+      "loss": 0.9634,
+      "step": 1866
+    },
+    {
+      "epoch": 0.13737410899057254,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.777251676745652e-05,
+      "loss": 0.9631,
+      "step": 1867
+    },
+    {
+      "epoch": 0.13744768912393654,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.7770126395564315e-05,
+      "loss": 0.595,
+      "step": 1868
+    },
+    {
+      "epoch": 0.13752126925730052,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.776773480164243e-05,
+      "loss": 0.869,
+      "step": 1869
+    },
+    {
+      "epoch": 0.13759484939066452,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.776534198581922e-05,
+      "loss": 0.8312,
+      "step": 1870
+    },
+    {
+      "epoch": 0.13766842952402852,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.77629479482231e-05,
+      "loss": 0.9036,
+      "step": 1871
+    },
+    {
+      "epoch": 0.1377420096573925,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.776055268898256e-05,
+      "loss": 1.1004,
+      "step": 1872
+    },
+    {
+      "epoch": 0.1378155897907565,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.7758156208226156e-05,
+      "loss": 0.7739,
+      "step": 1873
+    },
+    {
+      "epoch": 0.13788916992412048,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.775575850608248e-05,
+      "loss": 0.9294,
+      "step": 1874
+    },
+    {
+      "epoch": 0.13796275005748448,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.775335958268023e-05,
+      "loss": 0.7431,
+      "step": 1875
+    },
+    {
+      "epoch": 0.13803633019084846,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.7750959438148146e-05,
+      "loss": 0.9908,
+      "step": 1876
+    },
+    {
+      "epoch": 0.13810991032421246,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.7748558072615034e-05,
+      "loss": 0.9226,
+      "step": 1877
+    },
+    {
+      "epoch": 0.13818349045757644,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.7746155486209784e-05,
+      "loss": 1.1121,
+      "step": 1878
+    },
+    {
+      "epoch": 0.13825707059094045,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.774375167906132e-05,
+      "loss": 1.1865,
+      "step": 1879
+    },
+    {
+      "epoch": 0.13833065072430445,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.7741346651298665e-05,
+      "loss": 0.825,
+      "step": 1880
+    },
+    {
+      "epoch": 0.13840423085766843,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.773894040305089e-05,
+      "loss": 0.6942,
+      "step": 1881
+    },
+    {
+      "epoch": 0.13847781099103243,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.7736532934447134e-05,
+      "loss": 0.8538,
+      "step": 1882
+    },
+    {
+      "epoch": 0.1385513911243964,
+      "grad_norm": 0.6171875,
+      "learning_rate": 4.77341242456166e-05,
+      "loss": 0.5544,
+      "step": 1883
+    },
+    {
+      "epoch": 0.1386249712577604,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.773171433668855e-05,
+      "loss": 0.9721,
+      "step": 1884
+    },
+    {
+      "epoch": 0.1386985513911244,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.772930320779232e-05,
+      "loss": 0.7255,
+      "step": 1885
+    },
+    {
+      "epoch": 0.1387721315244884,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.772689085905733e-05,
+      "loss": 0.6076,
+      "step": 1886
+    },
+    {
+      "epoch": 0.13884571165785237,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.7724477290613024e-05,
+      "loss": 1.1983,
+      "step": 1887
+    },
+    {
+      "epoch": 0.13891929179121637,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.772206250258894e-05,
+      "loss": 1.039,
+      "step": 1888
+    },
+    {
+      "epoch": 0.13899287192458037,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.771964649511469e-05,
+      "loss": 0.7667,
+      "step": 1889
+    },
+    {
+      "epoch": 0.13906645205794435,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.771722926831991e-05,
+      "loss": 0.7998,
+      "step": 1890
+    },
+    {
+      "epoch": 0.13914003219130835,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.771481082233434e-05,
+      "loss": 1.4976,
+      "step": 1891
+    },
+    {
+      "epoch": 0.13921361232467233,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.771239115728779e-05,
+      "loss": 0.8863,
+      "step": 1892
+    },
+    {
+      "epoch": 0.13928719245803634,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.7709970273310095e-05,
+      "loss": 0.7846,
+      "step": 1893
+    },
+    {
+      "epoch": 0.1393607725914003,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.770754817053119e-05,
+      "loss": 0.9603,
+      "step": 1894
+    },
+    {
+      "epoch": 0.13943435272476432,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.7705124849081063e-05,
+      "loss": 1.2508,
+      "step": 1895
+    },
+    {
+      "epoch": 0.1395079328581283,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.7702700309089776e-05,
+      "loss": 0.6949,
+      "step": 1896
+    },
+    {
+      "epoch": 0.1395815129914923,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.770027455068743e-05,
+      "loss": 1.0184,
+      "step": 1897
+    },
+    {
+      "epoch": 0.1396550931248563,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.7697847574004234e-05,
+      "loss": 1.1851,
+      "step": 1898
+    },
+    {
+      "epoch": 0.13972867325822028,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.7695419379170426e-05,
+      "loss": 1.2688,
+      "step": 1899
+    },
+    {
+      "epoch": 0.13980225339158428,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.7692989966316324e-05,
+      "loss": 0.8647,
+      "step": 1900
+    },
+    {
+      "epoch": 0.13987583352494826,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.769055933557231e-05,
+      "loss": 1.0615,
+      "step": 1901
+    },
+    {
+      "epoch": 0.13994941365831226,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.7688127487068836e-05,
+      "loss": 1.0232,
+      "step": 1902
+    },
+    {
+      "epoch": 0.14002299379167624,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.768569442093641e-05,
+      "loss": 1.0361,
+      "step": 1903
+    },
+    {
+      "epoch": 0.14009657392504024,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.7683260137305615e-05,
+      "loss": 0.9985,
+      "step": 1904
+    },
+    {
+      "epoch": 0.14017015405840422,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.7680824636307086e-05,
+      "loss": 1.0015,
+      "step": 1905
+    },
+    {
+      "epoch": 0.14024373419176822,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.767838791807154e-05,
+      "loss": 0.8507,
+      "step": 1906
+    },
+    {
+      "epoch": 0.14031731432513223,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.767594998272974e-05,
+      "loss": 0.9166,
+      "step": 1907
+    },
+    {
+      "epoch": 0.1403908944584962,
+      "grad_norm": 0.875,
+      "learning_rate": 4.7673510830412546e-05,
+      "loss": 1.0552,
+      "step": 1908
+    },
+    {
+      "epoch": 0.1404644745918602,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.7671070461250846e-05,
+      "loss": 1.0193,
+      "step": 1909
+    },
+    {
+      "epoch": 0.14053805472522418,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.766862887537561e-05,
+      "loss": 1.2958,
+      "step": 1910
+    },
+    {
+      "epoch": 0.14061163485858819,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.766618607291787e-05,
+      "loss": 0.8358,
+      "step": 1911
+    },
+    {
+      "epoch": 0.14068521499195216,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.766374205400875e-05,
+      "loss": 1.1922,
+      "step": 1912
+    },
+    {
+      "epoch": 0.14075879512531617,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.766129681877939e-05,
+      "loss": 1.3769,
+      "step": 1913
+    },
+    {
+      "epoch": 0.14083237525868014,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.765885036736104e-05,
+      "loss": 0.8358,
+      "step": 1914
+    },
+    {
+      "epoch": 0.14090595539204415,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.765640269988497e-05,
+      "loss": 1.398,
+      "step": 1915
+    },
+    {
+      "epoch": 0.14097953552540815,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.7653953816482576e-05,
+      "loss": 1.0017,
+      "step": 1916
+    },
+    {
+      "epoch": 0.14105311565877213,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.7651503717285265e-05,
+      "loss": 0.8333,
+      "step": 1917
+    },
+    {
+      "epoch": 0.14112669579213613,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.764905240242452e-05,
+      "loss": 0.8153,
+      "step": 1918
+    },
+    {
+      "epoch": 0.1412002759255001,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.764659987203192e-05,
+      "loss": 1.147,
+      "step": 1919
+    },
+    {
+      "epoch": 0.1412738560588641,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.764414612623907e-05,
+      "loss": 0.7301,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1413474361922281,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.764169116517768e-05,
+      "loss": 0.8859,
+      "step": 1921
+    },
+    {
+      "epoch": 0.1414210163255921,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.763923498897948e-05,
+      "loss": 0.6014,
+      "step": 1922
+    },
+    {
+      "epoch": 0.1414945964589561,
+      "grad_norm": 1.3125,
+      "learning_rate": 4.7636777597776306e-05,
+      "loss": 1.2059,
+      "step": 1923
+    },
+    {
+      "epoch": 0.14156817659232007,
+      "grad_norm": 0.875,
+      "learning_rate": 4.763431899170002e-05,
+      "loss": 0.791,
+      "step": 1924
+    },
+    {
+      "epoch": 0.14164175672568408,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.76318591708826e-05,
+      "loss": 0.7207,
+      "step": 1925
+    },
+    {
+      "epoch": 0.14171533685904805,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.7629398135456035e-05,
+      "loss": 0.8129,
+      "step": 1926
+    },
+    {
+      "epoch": 0.14178891699241206,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.7626935885552406e-05,
+      "loss": 0.8824,
+      "step": 1927
+    },
+    {
+      "epoch": 0.14186249712577603,
+      "grad_norm": 1.25,
+      "learning_rate": 4.7624472421303876e-05,
+      "loss": 1.1402,
+      "step": 1928
+    },
+    {
+      "epoch": 0.14193607725914004,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.7622007742842635e-05,
+      "loss": 0.9439,
+      "step": 1929
+    },
+    {
+      "epoch": 0.142009657392504,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.7619541850300976e-05,
+      "loss": 0.9133,
+      "step": 1930
+    },
+    {
+      "epoch": 0.14208323752586802,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.761707474381122e-05,
+      "loss": 1.0942,
+      "step": 1931
+    },
+    {
+      "epoch": 0.14215681765923202,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.7614606423505773e-05,
+      "loss": 1.0403,
+      "step": 1932
+    },
+    {
+      "epoch": 0.142230397792596,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.761213688951712e-05,
+      "loss": 0.9698,
+      "step": 1933
+    },
+    {
+      "epoch": 0.14230397792596,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.760966614197779e-05,
+      "loss": 0.8362,
+      "step": 1934
+    },
+    {
+      "epoch": 0.14237755805932398,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.760719418102038e-05,
+      "loss": 1.0727,
+      "step": 1935
+    },
+    {
+      "epoch": 0.14245113819268798,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.760472100677755e-05,
+      "loss": 0.9647,
+      "step": 1936
+    },
+    {
+      "epoch": 0.14252471832605196,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.760224661938205e-05,
+      "loss": 0.9848,
+      "step": 1937
+    },
+    {
+      "epoch": 0.14259829845941596,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.759977101896666e-05,
+      "loss": 1.0689,
+      "step": 1938
+    },
+    {
+      "epoch": 0.14267187859277994,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.7597294205664244e-05,
+      "loss": 1.0485,
+      "step": 1939
+    },
+    {
+      "epoch": 0.14274545872614394,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.759481617960772e-05,
+      "loss": 1.0691,
+      "step": 1940
+    },
+    {
+      "epoch": 0.14281903885950795,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.75923369409301e-05,
+      "loss": 0.9305,
+      "step": 1941
+    },
+    {
+      "epoch": 0.14289261899287192,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.7589856489764414e-05,
+      "loss": 1.23,
+      "step": 1942
+    },
+    {
+      "epoch": 0.14296619912623593,
+      "grad_norm": 1.125,
+      "learning_rate": 4.758737482624381e-05,
+      "loss": 1.4511,
+      "step": 1943
+    },
+    {
+      "epoch": 0.1430397792595999,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.758489195050145e-05,
+      "loss": 0.85,
+      "step": 1944
+    },
+    {
+      "epoch": 0.1431133593929639,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.758240786267061e-05,
+      "loss": 0.8774,
+      "step": 1945
+    },
+    {
+      "epoch": 0.14318693952632788,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.757992256288458e-05,
+      "loss": 0.9928,
+      "step": 1946
+    },
+    {
+      "epoch": 0.1432605196596919,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.7577436051276754e-05,
+      "loss": 1.0241,
+      "step": 1947
+    },
+    {
+      "epoch": 0.14333409979305586,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.757494832798057e-05,
+      "loss": 0.9499,
+      "step": 1948
+    },
+    {
+      "epoch": 0.14340767992641987,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.7572459393129567e-05,
+      "loss": 1.0906,
+      "step": 1949
+    },
+    {
+      "epoch": 0.14348126005978387,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.7569969246857285e-05,
+      "loss": 0.8805,
+      "step": 1950
+    },
+    {
+      "epoch": 0.14355484019314785,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.756747788929739e-05,
+      "loss": 0.9583,
+      "step": 1951
+    },
+    {
+      "epoch": 0.14362842032651185,
+      "grad_norm": 1.0,
+      "learning_rate": 4.7564985320583575e-05,
+      "loss": 1.0505,
+      "step": 1952
+    },
+    {
+      "epoch": 0.14370200045987583,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.756249154084963e-05,
+      "loss": 0.9484,
+      "step": 1953
+    },
+    {
+      "epoch": 0.14377558059323983,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.755999655022937e-05,
+      "loss": 1.1815,
+      "step": 1954
+    },
+    {
+      "epoch": 0.1438491607266038,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.75575003488567e-05,
+      "loss": 0.9117,
+      "step": 1955
+    },
+    {
+      "epoch": 0.1439227408599678,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.75550029368656e-05,
+      "loss": 0.7588,
+      "step": 1956
+    },
+    {
+      "epoch": 0.1439963209933318,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.755250431439009e-05,
+      "loss": 0.6512,
+      "step": 1957
+    },
+    {
+      "epoch": 0.1440699011266958,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.7550004481564266e-05,
+      "loss": 1.0559,
+      "step": 1958
+    },
+    {
+      "epoch": 0.1441434812600598,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.754750343852229e-05,
+      "loss": 0.932,
+      "step": 1959
+    },
+    {
+      "epoch": 0.14421706139342377,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.75450011853984e-05,
+      "loss": 0.8623,
+      "step": 1960
+    },
+    {
+      "epoch": 0.14429064152678778,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.7542497722326874e-05,
+      "loss": 0.6626,
+      "step": 1961
+    },
+    {
+      "epoch": 0.14436422166015175,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.753999304944207e-05,
+      "loss": 1.1838,
+      "step": 1962
+    },
+    {
+      "epoch": 0.14443780179351576,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.753748716687841e-05,
+      "loss": 0.9884,
+      "step": 1963
+    },
+    {
+      "epoch": 0.14451138192687973,
+      "grad_norm": 0.75,
+      "learning_rate": 4.753498007477038e-05,
+      "loss": 0.8274,
+      "step": 1964
+    },
+    {
+      "epoch": 0.14458496206024374,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.7532471773252535e-05,
+      "loss": 1.5581,
+      "step": 1965
+    },
+    {
+      "epoch": 0.14465854219360771,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.7529962262459484e-05,
+      "loss": 0.856,
+      "step": 1966
+    },
+    {
+      "epoch": 0.14473212232697172,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.752745154252591e-05,
+      "loss": 0.8135,
+      "step": 1967
+    },
+    {
+      "epoch": 0.14480570246033572,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.752493961358657e-05,
+      "loss": 1.2011,
+      "step": 1968
+    },
+    {
+      "epoch": 0.1448792825936997,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.752242647577626e-05,
+      "loss": 0.7709,
+      "step": 1969
+    },
+    {
+      "epoch": 0.1449528627270637,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.751991212922986e-05,
+      "loss": 0.8145,
+      "step": 1970
+    },
+    {
+      "epoch": 0.14502644286042768,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.751739657408231e-05,
+      "loss": 1.091,
+      "step": 1971
+    },
+    {
+      "epoch": 0.14510002299379168,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.751487981046861e-05,
+      "loss": 1.2284,
+      "step": 1972
+    },
+    {
+      "epoch": 0.14517360312715566,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.751236183852385e-05,
+      "loss": 1.9465,
+      "step": 1973
+    },
+    {
+      "epoch": 0.14524718326051966,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.750984265838313e-05,
+      "loss": 0.9955,
+      "step": 1974
+    },
+    {
+      "epoch": 0.14532076339388364,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.750732227018168e-05,
+      "loss": 1.0679,
+      "step": 1975
+    },
+    {
+      "epoch": 0.14539434352724764,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.750480067405476e-05,
+      "loss": 1.0622,
+      "step": 1976
+    },
+    {
+      "epoch": 0.14546792366061165,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.750227787013768e-05,
+      "loss": 1.7516,
+      "step": 1977
+    },
+    {
+      "epoch": 0.14554150379397562,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.749975385856586e-05,
+      "loss": 0.8107,
+      "step": 1978
+    },
+    {
+      "epoch": 0.14561508392733963,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.749722863947473e-05,
+      "loss": 0.5357,
+      "step": 1979
+    },
+    {
+      "epoch": 0.1456886640607036,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.749470221299984e-05,
+      "loss": 0.8121,
+      "step": 1980
+    },
+    {
+      "epoch": 0.1457622441940676,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.7492174579276774e-05,
+      "loss": 0.9148,
+      "step": 1981
+    },
+    {
+      "epoch": 0.14583582432743158,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.748964573844118e-05,
+      "loss": 1.3749,
+      "step": 1982
+    },
+    {
+      "epoch": 0.1459094044607956,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.748711569062877e-05,
+      "loss": 0.7866,
+      "step": 1983
+    },
+    {
+      "epoch": 0.14598298459415956,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.748458443597533e-05,
+      "loss": 0.9804,
+      "step": 1984
+    },
+    {
+      "epoch": 0.14605656472752357,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.748205197461671e-05,
+      "loss": 0.8009,
+      "step": 1985
+    },
+    {
+      "epoch": 0.14613014486088757,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.747951830668884e-05,
+      "loss": 1.013,
+      "step": 1986
+    },
+    {
+      "epoch": 0.14620372499425155,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.747698343232766e-05,
+      "loss": 1.1268,
+      "step": 1987
+    },
+    {
+      "epoch": 0.14627730512761555,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.7474447351669244e-05,
+      "loss": 1.0835,
+      "step": 1988
+    },
+    {
+      "epoch": 0.14635088526097953,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.7471910064849685e-05,
+      "loss": 1.0355,
+      "step": 1989
+    },
+    {
+      "epoch": 0.14642446539434353,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.746937157200515e-05,
+      "loss": 1.3579,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1464980455277075,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.746683187327189e-05,
+      "loss": 1.1034,
+      "step": 1991
+    },
+    {
+      "epoch": 0.14657162566107151,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.746429096878619e-05,
+      "loss": 0.7161,
+      "step": 1992
+    },
+    {
+      "epoch": 0.1466452057944355,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.746174885868443e-05,
+      "loss": 1.0823,
+      "step": 1993
+    },
+    {
+      "epoch": 0.1467187859277995,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.7459205543103026e-05,
+      "loss": 0.7984,
+      "step": 1994
+    },
+    {
+      "epoch": 0.1467923660611635,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.745666102217848e-05,
+      "loss": 0.8845,
+      "step": 1995
+    },
+    {
+      "epoch": 0.14686594619452747,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.745411529604736e-05,
+      "loss": 1.076,
+      "step": 1996
+    },
+    {
+      "epoch": 0.14693952632789148,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.745156836484627e-05,
+      "loss": 1.0053,
+      "step": 1997
+    },
+    {
+      "epoch": 0.14701310646125546,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.744902022871192e-05,
+      "loss": 0.7948,
+      "step": 1998
+    },
+    {
+      "epoch": 0.14708668659461946,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.744647088778105e-05,
+      "loss": 1.0067,
+      "step": 1999
+    },
+    {
+      "epoch": 0.14716026672798344,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.7443920342190485e-05,
+      "loss": 1.0513,
+      "step": 2000
+    },
+    {
+      "epoch": 0.14723384686134744,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.744136859207711e-05,
+      "loss": 0.7567,
+      "step": 2001
+    },
+    {
+      "epoch": 0.14730742699471142,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.743881563757786e-05,
+      "loss": 0.9063,
+      "step": 2002
+    },
+    {
+      "epoch": 0.14738100712807542,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.7436261478829756e-05,
+      "loss": 1.4177,
+      "step": 2003
+    },
+    {
+      "epoch": 0.14745458726143942,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.743370611596988e-05,
+      "loss": 0.6505,
+      "step": 2004
+    },
+    {
+      "epoch": 0.1475281673948034,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.7431149549135364e-05,
+      "loss": 0.6567,
+      "step": 2005
+    },
+    {
+      "epoch": 0.1476017475281674,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.742859177846342e-05,
+      "loss": 0.9314,
+      "step": 2006
+    },
+    {
+      "epoch": 0.14767532766153138,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.7426032804091315e-05,
+      "loss": 0.8187,
+      "step": 2007
+    },
+    {
+      "epoch": 0.14774890779489538,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.742347262615639e-05,
+      "loss": 1.0127,
+      "step": 2008
+    },
+    {
+      "epoch": 0.14782248792825936,
+      "grad_norm": 1.5390625,
+      "learning_rate": 4.742091124479604e-05,
+      "loss": 0.562,
+      "step": 2009
+    },
+    {
+      "epoch": 0.14789606806162336,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.741834866014773e-05,
+      "loss": 1.4887,
+      "step": 2010
+    },
+    {
+      "epoch": 0.14796964819498734,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.741578487234899e-05,
+      "loss": 1.1488,
+      "step": 2011
+    },
+    {
+      "epoch": 0.14804322832835135,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.7413219881537404e-05,
+      "loss": 1.2171,
+      "step": 2012
+    },
+    {
+      "epoch": 0.14811680846171535,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.741065368785066e-05,
+      "loss": 1.8913,
+      "step": 2013
+    },
+    {
+      "epoch": 0.14819038859507933,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.740808629142645e-05,
+      "loss": 0.7099,
+      "step": 2014
+    },
+    {
+      "epoch": 0.14826396872844333,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.740551769240256e-05,
+      "loss": 1.0766,
+      "step": 2015
+    },
+    {
+      "epoch": 0.1483375488618073,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.7402947890916865e-05,
+      "loss": 0.7768,
+      "step": 2016
+    },
+    {
+      "epoch": 0.1484111289951713,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.740037688710727e-05,
+      "loss": 1.2981,
+      "step": 2017
+    },
+    {
+      "epoch": 0.14848470912853529,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.739780468111175e-05,
+      "loss": 1.0663,
+      "step": 2018
+    },
+    {
+      "epoch": 0.1485582892618993,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.739523127306837e-05,
+      "loss": 0.8822,
+      "step": 2019
+    },
+    {
+      "epoch": 0.14863186939526327,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.739265666311521e-05,
+      "loss": 0.7141,
+      "step": 2020
+    },
+    {
+      "epoch": 0.14870544952862727,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.739008085139046e-05,
+      "loss": 1.0347,
+      "step": 2021
+    },
+    {
+      "epoch": 0.14877902966199127,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.738750383803237e-05,
+      "loss": 1.0854,
+      "step": 2022
+    },
+    {
+      "epoch": 0.14885260979535525,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.738492562317923e-05,
+      "loss": 0.8448,
+      "step": 2023
+    },
+    {
+      "epoch": 0.14892618992871925,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.7382346206969405e-05,
+      "loss": 1.1947,
+      "step": 2024
+    },
+    {
+      "epoch": 0.14899977006208323,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.737976558954135e-05,
+      "loss": 0.8409,
+      "step": 2025
+    },
+    {
+      "epoch": 0.14907335019544724,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.737718377103353e-05,
+      "loss": 0.8744,
+      "step": 2026
+    },
+    {
+      "epoch": 0.1491469303288112,
+      "grad_norm": 2.09375,
+      "learning_rate": 4.737460075158452e-05,
+      "loss": 0.8834,
+      "step": 2027
+    },
+    {
+      "epoch": 0.14922051046217522,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.737201653133295e-05,
+      "loss": 0.7441,
+      "step": 2028
+    },
+    {
+      "epoch": 0.1492940905955392,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.736943111041752e-05,
+      "loss": 1.0531,
+      "step": 2029
+    },
+    {
+      "epoch": 0.1493676707289032,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.736684448897696e-05,
+      "loss": 1.1605,
+      "step": 2030
+    },
+    {
+      "epoch": 0.1494412508622672,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.7364256667150095e-05,
+      "loss": 0.8815,
+      "step": 2031
+    },
+    {
+      "epoch": 0.14951483099563118,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.736166764507583e-05,
+      "loss": 1.0073,
+      "step": 2032
+    },
+    {
+      "epoch": 0.14958841112899518,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.7359077422893094e-05,
+      "loss": 0.9168,
+      "step": 2033
+    },
+    {
+      "epoch": 0.14966199126235916,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.73564860007409e-05,
+      "loss": 1.0928,
+      "step": 2034
+    },
+    {
+      "epoch": 0.14973557139572316,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.7353893378758326e-05,
+      "loss": 0.9621,
+      "step": 2035
+    },
+    {
+      "epoch": 0.14980915152908714,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.7351299557084515e-05,
+      "loss": 1.1565,
+      "step": 2036
+    },
+    {
+      "epoch": 0.14988273166245114,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.734870453585867e-05,
+      "loss": 0.9596,
+      "step": 2037
+    },
+    {
+      "epoch": 0.14995631179581512,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.734610831522007e-05,
+      "loss": 0.8283,
+      "step": 2038
+    },
+    {
+      "epoch": 0.15002989192917912,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.7343510895308044e-05,
+      "loss": 0.6312,
+      "step": 2039
+    },
+    {
+      "epoch": 0.15010347206254313,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.734091227626198e-05,
+      "loss": 1.0671,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1501770521959071,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.733831245822136e-05,
+      "loss": 1.1031,
+      "step": 2041
+    },
+    {
+      "epoch": 0.1502506323292711,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.733571144132569e-05,
+      "loss": 0.8168,
+      "step": 2042
+    },
+    {
+      "epoch": 0.15032421246263508,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.733310922571458e-05,
+      "loss": 1.0221,
+      "step": 2043
+    },
+    {
+      "epoch": 0.15039779259599909,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.733050581152768e-05,
+      "loss": 0.8938,
+      "step": 2044
+    },
+    {
+      "epoch": 0.15047137272936306,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.732790119890471e-05,
+      "loss": 1.0355,
+      "step": 2045
+    },
+    {
+      "epoch": 0.15054495286272707,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.732529538798545e-05,
+      "loss": 0.6696,
+      "step": 2046
+    },
+    {
+      "epoch": 0.15061853299609104,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.7322688378909754e-05,
+      "loss": 0.8844,
+      "step": 2047
+    },
+    {
+      "epoch": 0.15069211312945505,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.7320080171817536e-05,
+      "loss": 0.9323,
+      "step": 2048
+    },
+    {
+      "epoch": 0.15076569326281905,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.731747076684877e-05,
+      "loss": 1.1546,
+      "step": 2049
+    },
+    {
+      "epoch": 0.15083927339618303,
+      "grad_norm": 2.0625,
+      "learning_rate": 4.731486016414351e-05,
+      "loss": 0.8309,
+      "step": 2050
+    },
+    {
+      "epoch": 0.15091285352954703,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.731224836384184e-05,
+      "loss": 1.1214,
+      "step": 2051
+    },
+    {
+      "epoch": 0.150986433662911,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.730963536608394e-05,
+      "loss": 1.1951,
+      "step": 2052
+    },
+    {
+      "epoch": 0.151060013796275,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.7307021171010054e-05,
+      "loss": 0.8857,
+      "step": 2053
+    },
+    {
+      "epoch": 0.151133593929639,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.7304405778760466e-05,
+      "loss": 1.3764,
+      "step": 2054
+    },
+    {
+      "epoch": 0.151207174063003,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.7301789189475556e-05,
+      "loss": 1.0744,
+      "step": 2055
+    },
+    {
+      "epoch": 0.15128075419636697,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.729917140329574e-05,
+      "loss": 1.1497,
+      "step": 2056
+    },
+    {
+      "epoch": 0.15135433432973097,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.7296552420361505e-05,
+      "loss": 1.4336,
+      "step": 2057
+    },
+    {
+      "epoch": 0.15142791446309498,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.7293932240813424e-05,
+      "loss": 0.8087,
+      "step": 2058
+    },
+    {
+      "epoch": 0.15150149459645895,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.7291310864792104e-05,
+      "loss": 0.7935,
+      "step": 2059
+    },
+    {
+      "epoch": 0.15157507472982296,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.728868829243823e-05,
+      "loss": 0.5489,
+      "step": 2060
+    },
+    {
+      "epoch": 0.15164865486318693,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.728606452389255e-05,
+      "loss": 0.9969,
+      "step": 2061
+    },
+    {
+      "epoch": 0.15172223499655094,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.7283439559295884e-05,
+      "loss": 0.8286,
+      "step": 2062
+    },
+    {
+      "epoch": 0.1517958151299149,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.72808133987891e-05,
+      "loss": 0.9153,
+      "step": 2063
+    },
+    {
+      "epoch": 0.15186939526327892,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.727818604251315e-05,
+      "loss": 0.955,
+      "step": 2064
+    },
+    {
+      "epoch": 0.1519429753966429,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.7275557490609026e-05,
+      "loss": 0.8011,
+      "step": 2065
+    },
+    {
+      "epoch": 0.1520165555300069,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.727292774321781e-05,
+      "loss": 1.0804,
+      "step": 2066
+    },
+    {
+      "epoch": 0.1520901356633709,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.727029680048063e-05,
+      "loss": 0.7592,
+      "step": 2067
+    },
+    {
+      "epoch": 0.15216371579673488,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.726766466253867e-05,
+      "loss": 1.0333,
+      "step": 2068
+    },
+    {
+      "epoch": 0.15223729593009888,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.7265031329533215e-05,
+      "loss": 1.2509,
+      "step": 2069
+    },
+    {
+      "epoch": 0.15231087606346286,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.726239680160559e-05,
+      "loss": 1.1099,
+      "step": 2070
+    },
+    {
+      "epoch": 0.15238445619682686,
+      "grad_norm": 0.875,
+      "learning_rate": 4.7259761078897166e-05,
+      "loss": 0.9803,
+      "step": 2071
+    },
+    {
+      "epoch": 0.15245803633019084,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.72571241615494e-05,
+      "loss": 0.9686,
+      "step": 2072
+    },
+    {
+      "epoch": 0.15253161646355484,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.7254486049703834e-05,
+      "loss": 0.9403,
+      "step": 2073
+    },
+    {
+      "epoch": 0.15260519659691882,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.725184674350203e-05,
+      "loss": 1.0394,
+      "step": 2074
+    },
+    {
+      "epoch": 0.15267877673028282,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.724920624308563e-05,
+      "loss": 0.6761,
+      "step": 2075
+    },
+    {
+      "epoch": 0.15275235686364683,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.724656454859636e-05,
+      "loss": 1.0326,
+      "step": 2076
+    },
+    {
+      "epoch": 0.1528259369970108,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.7243921660175996e-05,
+      "loss": 0.7864,
+      "step": 2077
+    },
+    {
+      "epoch": 0.1528995171303748,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.724127757796636e-05,
+      "loss": 0.9623,
+      "step": 2078
+    },
+    {
+      "epoch": 0.15297309726373878,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.7238632302109364e-05,
+      "loss": 1.1846,
+      "step": 2079
+    },
+    {
+      "epoch": 0.1530466773971028,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.7235985832746976e-05,
+      "loss": 0.8224,
+      "step": 2080
+    },
+    {
+      "epoch": 0.15312025753046676,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.723333817002123e-05,
+      "loss": 0.9363,
+      "step": 2081
+    },
+    {
+      "epoch": 0.15319383766383077,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.723068931407422e-05,
+      "loss": 0.9264,
+      "step": 2082
+    },
+    {
+      "epoch": 0.15326741779719474,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.72280392650481e-05,
+      "loss": 1.0241,
+      "step": 2083
+    },
+    {
+      "epoch": 0.15334099793055875,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.722538802308508e-05,
+      "loss": 1.3347,
+      "step": 2084
+    },
+    {
+      "epoch": 0.15341457806392275,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.722273558832748e-05,
+      "loss": 1.1008,
+      "step": 2085
+    },
+    {
+      "epoch": 0.15348815819728673,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.7220081960917625e-05,
+      "loss": 0.9284,
+      "step": 2086
+    },
+    {
+      "epoch": 0.15356173833065073,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.721742714099795e-05,
+      "loss": 1.1102,
+      "step": 2087
+    },
+    {
+      "epoch": 0.1536353184640147,
+      "grad_norm": 1.5859375,
+      "learning_rate": 4.721477112871091e-05,
+      "loss": 1.1702,
+      "step": 2088
+    },
+    {
+      "epoch": 0.1537088985973787,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.721211392419907e-05,
+      "loss": 0.9287,
+      "step": 2089
+    },
+    {
+      "epoch": 0.1537824787307427,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.720945552760503e-05,
+      "loss": 1.1566,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1538560588641067,
+      "grad_norm": 1.25,
+      "learning_rate": 4.720679593907145e-05,
+      "loss": 1.5028,
+      "step": 2091
+    },
+    {
+      "epoch": 0.15392963899747067,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.720413515874108e-05,
+      "loss": 0.6543,
+      "step": 2092
+    },
+    {
+      "epoch": 0.15400321913083467,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.7201473186756716e-05,
+      "loss": 0.9032,
+      "step": 2093
+    },
+    {
+      "epoch": 0.15407679926419868,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.719881002326121e-05,
+      "loss": 0.7149,
+      "step": 2094
+    },
+    {
+      "epoch": 0.15415037939756265,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.7196145668397504e-05,
+      "loss": 0.7797,
+      "step": 2095
+    },
+    {
+      "epoch": 0.15422395953092666,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.719348012230859e-05,
+      "loss": 0.6621,
+      "step": 2096
+    },
+    {
+      "epoch": 0.15429753966429063,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.7190813385137503e-05,
+      "loss": 1.0652,
+      "step": 2097
+    },
+    {
+      "epoch": 0.15437111979765464,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.718814545702738e-05,
+      "loss": 0.9207,
+      "step": 2098
+    },
+    {
+      "epoch": 0.15444469993101861,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.7185476338121395e-05,
+      "loss": 1.0707,
+      "step": 2099
+    },
+    {
+      "epoch": 0.15451828006438262,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.71828060285628e-05,
+      "loss": 0.8185,
+      "step": 2100
+    },
+    {
+      "epoch": 0.15459186019774662,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.7180134528494903e-05,
+      "loss": 1.5909,
+      "step": 2101
+    },
+    {
+      "epoch": 0.1546654403311106,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.717746183806108e-05,
+      "loss": 0.7151,
+      "step": 2102
+    },
+    {
+      "epoch": 0.1547390204644746,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.717478795740477e-05,
+      "loss": 0.6298,
+      "step": 2103
+    },
+    {
+      "epoch": 0.15481260059783858,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.717211288666946e-05,
+      "loss": 1.4797,
+      "step": 2104
+    },
+    {
+      "epoch": 0.15488618073120258,
+      "grad_norm": 1.5078125,
+      "learning_rate": 4.716943662599873e-05,
+      "loss": 0.8167,
+      "step": 2105
+    },
+    {
+      "epoch": 0.15495976086456656,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.716675917553622e-05,
+      "loss": 0.8225,
+      "step": 2106
+    },
+    {
+      "epoch": 0.15503334099793056,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.71640805354256e-05,
+      "loss": 0.7856,
+      "step": 2107
+    },
+    {
+      "epoch": 0.15510692113129454,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.7161400705810646e-05,
+      "loss": 0.7681,
+      "step": 2108
+    },
+    {
+      "epoch": 0.15518050126465854,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.7158719686835176e-05,
+      "loss": 1.1255,
+      "step": 2109
+    },
+    {
+      "epoch": 0.15525408139802255,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.715603747864307e-05,
+      "loss": 0.6637,
+      "step": 2110
+    },
+    {
+      "epoch": 0.15532766153138652,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.715335408137827e-05,
+      "loss": 0.8472,
+      "step": 2111
+    },
+    {
+      "epoch": 0.15540124166475053,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.715066949518481e-05,
+      "loss": 0.8604,
+      "step": 2112
+    },
+    {
+      "epoch": 0.1554748217981145,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.7147983720206755e-05,
+      "loss": 0.9236,
+      "step": 2113
+    },
+    {
+      "epoch": 0.1555484019314785,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.714529675658824e-05,
+      "loss": 1.0879,
+      "step": 2114
+    },
+    {
+      "epoch": 0.15562198206484248,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.714260860447348e-05,
+      "loss": 0.7562,
+      "step": 2115
+    },
+    {
+      "epoch": 0.1556955621982065,
+      "grad_norm": 0.875,
+      "learning_rate": 4.713991926400673e-05,
+      "loss": 1.0826,
+      "step": 2116
+    },
+    {
+      "epoch": 0.15576914233157046,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.713722873533234e-05,
+      "loss": 1.4371,
+      "step": 2117
+    },
+    {
+      "epoch": 0.15584272246493447,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.713453701859468e-05,
+      "loss": 1.0721,
+      "step": 2118
+    },
+    {
+      "epoch": 0.15591630259829847,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.713184411393824e-05,
+      "loss": 0.9858,
+      "step": 2119
+    },
+    {
+      "epoch": 0.15598988273166245,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.712915002150752e-05,
+      "loss": 1.2844,
+      "step": 2120
+    },
+    {
+      "epoch": 0.15606346286502645,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.712645474144711e-05,
+      "loss": 0.9151,
+      "step": 2121
+    },
+    {
+      "epoch": 0.15613704299839043,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.712375827390167e-05,
+      "loss": 1.0682,
+      "step": 2122
+    },
+    {
+      "epoch": 0.15621062313175443,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.712106061901591e-05,
+      "loss": 1.1453,
+      "step": 2123
+    },
+    {
+      "epoch": 0.1562842032651184,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.711836177693461e-05,
+      "loss": 0.9225,
+      "step": 2124
+    },
+    {
+      "epoch": 0.15635778339848241,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.7115661747802604e-05,
+      "loss": 1.0966,
+      "step": 2125
+    },
+    {
+      "epoch": 0.1564313635318464,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.71129605317648e-05,
+      "loss": 0.9089,
+      "step": 2126
+    },
+    {
+      "epoch": 0.1565049436652104,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.711025812896618e-05,
+      "loss": 0.6613,
+      "step": 2127
+    },
+    {
+      "epoch": 0.1565785237985744,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.710755453955176e-05,
+      "loss": 0.9499,
+      "step": 2128
+    },
+    {
+      "epoch": 0.15665210393193837,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.710484976366664e-05,
+      "loss": 0.948,
+      "step": 2129
+    },
+    {
+      "epoch": 0.15672568406530238,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.710214380145599e-05,
+      "loss": 0.7208,
+      "step": 2130
+    },
+    {
+      "epoch": 0.15679926419866635,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.709943665306502e-05,
+      "loss": 0.6862,
+      "step": 2131
+    },
+    {
+      "epoch": 0.15687284433203036,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.7096728318639025e-05,
+      "loss": 1.1438,
+      "step": 2132
+    },
+    {
+      "epoch": 0.15694642446539434,
+      "grad_norm": 1.3828125,
+      "learning_rate": 4.7094018798323365e-05,
+      "loss": 1.1915,
+      "step": 2133
+    },
+    {
+      "epoch": 0.15702000459875834,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.709130809226344e-05,
+      "loss": 1.2771,
+      "step": 2134
+    },
+    {
+      "epoch": 0.15709358473212232,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.7088596200604735e-05,
+      "loss": 0.7715,
+      "step": 2135
+    },
+    {
+      "epoch": 0.15716716486548632,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.708588312349279e-05,
+      "loss": 1.2178,
+      "step": 2136
+    },
+    {
+      "epoch": 0.15724074499885032,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.708316886107321e-05,
+      "loss": 0.8211,
+      "step": 2137
+    },
+    {
+      "epoch": 0.1573143251322143,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.708045341349168e-05,
+      "loss": 0.8783,
+      "step": 2138
+    },
+    {
+      "epoch": 0.1573879052655783,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.70777367808939e-05,
+      "loss": 0.951,
+      "step": 2139
+    },
+    {
+      "epoch": 0.15746148539894228,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.70750189634257e-05,
+      "loss": 0.792,
+      "step": 2140
+    },
+    {
+      "epoch": 0.15753506553230628,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.7072299961232915e-05,
+      "loss": 0.6977,
+      "step": 2141
+    },
+    {
+      "epoch": 0.15760864566567026,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.7069579774461485e-05,
+      "loss": 0.7599,
+      "step": 2142
+    },
+    {
+      "epoch": 0.15768222579903426,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.70668584032574e-05,
+      "loss": 1.1872,
+      "step": 2143
+    },
+    {
+      "epoch": 0.15775580593239824,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.7064135847766686e-05,
+      "loss": 0.9709,
+      "step": 2144
+    },
+    {
+      "epoch": 0.15782938606576224,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.706141210813549e-05,
+      "loss": 0.9034,
+      "step": 2145
+    },
+    {
+      "epoch": 0.15790296619912625,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.705868718450996e-05,
+      "loss": 0.9864,
+      "step": 2146
+    },
+    {
+      "epoch": 0.15797654633249023,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.7055961077036364e-05,
+      "loss": 1.5484,
+      "step": 2147
+    },
+    {
+      "epoch": 0.15805012646585423,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.7053233785860996e-05,
+      "loss": 0.9305,
+      "step": 2148
+    },
+    {
+      "epoch": 0.1581237065992182,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.705050531113021e-05,
+      "loss": 0.8664,
+      "step": 2149
+    },
+    {
+      "epoch": 0.1581972867325822,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.7047775652990464e-05,
+      "loss": 1.0787,
+      "step": 2150
+    },
+    {
+      "epoch": 0.15827086686594619,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.704504481158823e-05,
+      "loss": 0.9074,
+      "step": 2151
+    },
+    {
+      "epoch": 0.1583444469993102,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.7042312787070084e-05,
+      "loss": 0.9109,
+      "step": 2152
+    },
+    {
+      "epoch": 0.15841802713267417,
+      "grad_norm": 1.2734375,
+      "learning_rate": 4.703957957958265e-05,
+      "loss": 1.3144,
+      "step": 2153
+    },
+    {
+      "epoch": 0.15849160726603817,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.70368451892726e-05,
+      "loss": 0.7123,
+      "step": 2154
+    },
+    {
+      "epoch": 0.15856518739940217,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.7034109616286694e-05,
+      "loss": 0.9552,
+      "step": 2155
+    },
+    {
+      "epoch": 0.15863876753276615,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.7031372860771735e-05,
+      "loss": 0.937,
+      "step": 2156
+    },
+    {
+      "epoch": 0.15871234766613015,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.70286349228746e-05,
+      "loss": 0.739,
+      "step": 2157
+    },
+    {
+      "epoch": 0.15878592779949413,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.702589580274225e-05,
+      "loss": 1.2628,
+      "step": 2158
+    },
+    {
+      "epoch": 0.15885950793285813,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.7023155500521666e-05,
+      "loss": 1.3584,
+      "step": 2159
+    },
+    {
+      "epoch": 0.1589330880662221,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.7020414016359926e-05,
+      "loss": 0.9563,
+      "step": 2160
+    },
+    {
+      "epoch": 0.15900666819958612,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.701767135040414e-05,
+      "loss": 0.9553,
+      "step": 2161
+    },
+    {
+      "epoch": 0.1590802483329501,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.701492750280154e-05,
+      "loss": 1.3369,
+      "step": 2162
+    },
+    {
+      "epoch": 0.1591538284663141,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.701218247369935e-05,
+      "loss": 1.0824,
+      "step": 2163
+    },
+    {
+      "epoch": 0.1592274085996781,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.7009436263244914e-05,
+      "loss": 1.254,
+      "step": 2164
+    },
+    {
+      "epoch": 0.15930098873304208,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.700668887158559e-05,
+      "loss": 1.0145,
+      "step": 2165
+    },
+    {
+      "epoch": 0.15937456886640608,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.700394029886884e-05,
+      "loss": 0.7849,
+      "step": 2166
+    },
+    {
+      "epoch": 0.15944814899977006,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.700119054524218e-05,
+      "loss": 0.9497,
+      "step": 2167
+    },
+    {
+      "epoch": 0.15952172913313406,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.699843961085317e-05,
+      "loss": 0.784,
+      "step": 2168
+    },
+    {
+      "epoch": 0.15959530926649804,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.699568749584946e-05,
+      "loss": 0.8025,
+      "step": 2169
+    },
+    {
+      "epoch": 0.15966888939986204,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.6992934200378744e-05,
+      "loss": 0.8684,
+      "step": 2170
+    },
+    {
+      "epoch": 0.15974246953322602,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.69901797245888e-05,
+      "loss": 1.0372,
+      "step": 2171
+    },
+    {
+      "epoch": 0.15981604966659002,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.6987424068627425e-05,
+      "loss": 0.7471,
+      "step": 2172
+    },
+    {
+      "epoch": 0.15988962979995403,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.6984667232642546e-05,
+      "loss": 1.122,
+      "step": 2173
+    },
+    {
+      "epoch": 0.159963209933318,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.698190921678208e-05,
+      "loss": 1.0807,
+      "step": 2174
+    },
+    {
+      "epoch": 0.160036790066682,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.697915002119408e-05,
+      "loss": 1.258,
+      "step": 2175
+    },
+    {
+      "epoch": 0.16011037020004598,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.697638964602661e-05,
+      "loss": 1.2759,
+      "step": 2176
+    },
+    {
+      "epoch": 0.16018395033340999,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.6973628091427805e-05,
+      "loss": 0.8642,
+      "step": 2177
+    },
+    {
+      "epoch": 0.16025753046677396,
+      "grad_norm": 0.875,
+      "learning_rate": 4.697086535754589e-05,
+      "loss": 0.838,
+      "step": 2178
+    },
+    {
+      "epoch": 0.16033111060013797,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.6968101444529136e-05,
+      "loss": 1.2733,
+      "step": 2179
+    },
+    {
+      "epoch": 0.16040469073350194,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.696533635252586e-05,
+      "loss": 0.7795,
+      "step": 2180
+    },
+    {
+      "epoch": 0.16047827086686595,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.6962570081684464e-05,
+      "loss": 0.9203,
+      "step": 2181
+    },
+    {
+      "epoch": 0.16055185100022995,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.695980263215342e-05,
+      "loss": 0.853,
+      "step": 2182
+    },
+    {
+      "epoch": 0.16062543113359393,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.695703400408124e-05,
+      "loss": 1.5444,
+      "step": 2183
+    },
+    {
+      "epoch": 0.16069901126695793,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.695426419761652e-05,
+      "loss": 1.1323,
+      "step": 2184
+    },
+    {
+      "epoch": 0.1607725914003219,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.6951493212907905e-05,
+      "loss": 1.099,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1608461715336859,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.694872105010412e-05,
+      "loss": 1.1622,
+      "step": 2186
+    },
+    {
+      "epoch": 0.1609197516670499,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.694594770935391e-05,
+      "loss": 1.0416,
+      "step": 2187
+    },
+    {
+      "epoch": 0.1609933318004139,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.694317319080615e-05,
+      "loss": 1.1845,
+      "step": 2188
+    },
+    {
+      "epoch": 0.16106691193377787,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.694039749460973e-05,
+      "loss": 0.9836,
+      "step": 2189
+    },
+    {
+      "epoch": 0.16114049206714187,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.6937620620913617e-05,
+      "loss": 0.7475,
+      "step": 2190
+    },
+    {
+      "epoch": 0.16121407220050588,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.693484256986683e-05,
+      "loss": 1.038,
+      "step": 2191
+    },
+    {
+      "epoch": 0.16128765233386985,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.693206334161848e-05,
+      "loss": 1.0436,
+      "step": 2192
+    },
+    {
+      "epoch": 0.16136123246723386,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.692928293631772e-05,
+      "loss": 0.791,
+      "step": 2193
+    },
+    {
+      "epoch": 0.16143481260059783,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.692650135411375e-05,
+      "loss": 1.2259,
+      "step": 2194
+    },
+    {
+      "epoch": 0.16150839273396184,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.692371859515587e-05,
+      "loss": 0.9201,
+      "step": 2195
+    },
+    {
+      "epoch": 0.1615819728673258,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.692093465959342e-05,
+      "loss": 1.0345,
+      "step": 2196
+    },
+    {
+      "epoch": 0.16165555300068982,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.691814954757582e-05,
+      "loss": 0.6074,
+      "step": 2197
+    },
+    {
+      "epoch": 0.1617291331340538,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.691536325925252e-05,
+      "loss": 1.0246,
+      "step": 2198
+    },
+    {
+      "epoch": 0.1618027132674178,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.691257579477306e-05,
+      "loss": 1.0069,
+      "step": 2199
+    },
+    {
+      "epoch": 0.1618762934007818,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.690978715428705e-05,
+      "loss": 1.6378,
+      "step": 2200
+    },
+    {
+      "epoch": 0.16194987353414578,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.690699733794416e-05,
+      "loss": 0.8153,
+      "step": 2201
+    },
+    {
+      "epoch": 0.16202345366750978,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.690420634589408e-05,
+      "loss": 0.9973,
+      "step": 2202
+    },
+    {
+      "epoch": 0.16209703380087376,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.690141417828663e-05,
+      "loss": 0.9972,
+      "step": 2203
+    },
+    {
+      "epoch": 0.16217061393423776,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.689862083527164e-05,
+      "loss": 0.7335,
+      "step": 2204
+    },
+    {
+      "epoch": 0.16224419406760174,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.689582631699903e-05,
+      "loss": 0.6909,
+      "step": 2205
+    },
+    {
+      "epoch": 0.16231777420096574,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.689303062361878e-05,
+      "loss": 0.6839,
+      "step": 2206
+    },
+    {
+      "epoch": 0.16239135433432972,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.689023375528092e-05,
+      "loss": 0.9808,
+      "step": 2207
+    },
+    {
+      "epoch": 0.16246493446769372,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.688743571213557e-05,
+      "loss": 1.6107,
+      "step": 2208
+    },
+    {
+      "epoch": 0.16253851460105773,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.688463649433288e-05,
+      "loss": 1.2718,
+      "step": 2209
+    },
+    {
+      "epoch": 0.1626120947344217,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.688183610202308e-05,
+      "loss": 0.9009,
+      "step": 2210
+    },
+    {
+      "epoch": 0.1626856748677857,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.687903453535647e-05,
+      "loss": 1.0113,
+      "step": 2211
+    },
+    {
+      "epoch": 0.16275925500114968,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.687623179448339e-05,
+      "loss": 1.2858,
+      "step": 2212
+    },
+    {
+      "epoch": 0.1628328351345137,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.6873427879554274e-05,
+      "loss": 1.2891,
+      "step": 2213
+    },
+    {
+      "epoch": 0.16290641526787766,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.687062279071961e-05,
+      "loss": 1.1064,
+      "step": 2214
+    },
+    {
+      "epoch": 0.16297999540124167,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.686781652812992e-05,
+      "loss": 1.3416,
+      "step": 2215
+    },
+    {
+      "epoch": 0.16305357553460564,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.686500909193581e-05,
+      "loss": 0.8285,
+      "step": 2216
+    },
+    {
+      "epoch": 0.16312715566796965,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.686220048228796e-05,
+      "loss": 1.0324,
+      "step": 2217
+    },
+    {
+      "epoch": 0.16320073580133365,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.685939069933711e-05,
+      "loss": 1.0747,
+      "step": 2218
+    },
+    {
+      "epoch": 0.16327431593469763,
+      "grad_norm": 1.0,
+      "learning_rate": 4.6856579743234044e-05,
+      "loss": 1.0746,
+      "step": 2219
+    },
+    {
+      "epoch": 0.16334789606806163,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.685376761412963e-05,
+      "loss": 1.3434,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1634214762014256,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.6850954312174775e-05,
+      "loss": 0.602,
+      "step": 2221
+    },
+    {
+      "epoch": 0.1634950563347896,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.684813983752048e-05,
+      "loss": 0.9007,
+      "step": 2222
+    },
+    {
+      "epoch": 0.1635686364681536,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.684532419031778e-05,
+      "loss": 0.6229,
+      "step": 2223
+    },
+    {
+      "epoch": 0.1636422166015176,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.684250737071779e-05,
+      "loss": 0.7356,
+      "step": 2224
+    },
+    {
+      "epoch": 0.16371579673488157,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.683968937887169e-05,
+      "loss": 0.75,
+      "step": 2225
+    },
+    {
+      "epoch": 0.16378937686824557,
+      "grad_norm": 1.265625,
+      "learning_rate": 4.6836870214930704e-05,
+      "loss": 1.3397,
+      "step": 2226
+    },
+    {
+      "epoch": 0.16386295700160958,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.683404987904615e-05,
+      "loss": 1.1249,
+      "step": 2227
+    },
+    {
+      "epoch": 0.16393653713497355,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.683122837136937e-05,
+      "loss": 1.0647,
+      "step": 2228
+    },
+    {
+      "epoch": 0.16401011726833756,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.682840569205179e-05,
+      "loss": 0.8404,
+      "step": 2229
+    },
+    {
+      "epoch": 0.16408369740170153,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.6825581841244916e-05,
+      "loss": 1.2112,
+      "step": 2230
+    },
+    {
+      "epoch": 0.16415727753506554,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.6822756819100275e-05,
+      "loss": 1.0361,
+      "step": 2231
+    },
+    {
+      "epoch": 0.16423085766842951,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.68199306257695e-05,
+      "loss": 1.0725,
+      "step": 2232
+    },
+    {
+      "epoch": 0.16430443780179352,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.681710326140426e-05,
+      "loss": 1.0729,
+      "step": 2233
+    },
+    {
+      "epoch": 0.1643780179351575,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.6814274726156296e-05,
+      "loss": 0.9688,
+      "step": 2234
+    },
+    {
+      "epoch": 0.1644515980685215,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.68114450201774e-05,
+      "loss": 1.3399,
+      "step": 2235
+    },
+    {
+      "epoch": 0.1645251782018855,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.680861414361945e-05,
+      "loss": 1.3039,
+      "step": 2236
+    },
+    {
+      "epoch": 0.16459875833524948,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.680578209663438e-05,
+      "loss": 1.2941,
+      "step": 2237
+    },
+    {
+      "epoch": 0.16467233846861348,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.680294887937416e-05,
+      "loss": 0.8394,
+      "step": 2238
+    },
+    {
+      "epoch": 0.16474591860197746,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.680011449199085e-05,
+      "loss": 1.136,
+      "step": 2239
+    },
+    {
+      "epoch": 0.16481949873534146,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.679727893463658e-05,
+      "loss": 1.0401,
+      "step": 2240
+    },
+    {
+      "epoch": 0.16489307886870544,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.679444220746352e-05,
+      "loss": 0.6675,
+      "step": 2241
+    },
+    {
+      "epoch": 0.16496665900206944,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.679160431062391e-05,
+      "loss": 0.9519,
+      "step": 2242
+    },
+    {
+      "epoch": 0.16504023913543342,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.678876524427004e-05,
+      "loss": 0.9513,
+      "step": 2243
+    },
+    {
+      "epoch": 0.16511381926879742,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.6785925008554305e-05,
+      "loss": 0.7941,
+      "step": 2244
+    },
+    {
+      "epoch": 0.16518739940216143,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.678308360362912e-05,
+      "loss": 1.0033,
+      "step": 2245
+    },
+    {
+      "epoch": 0.1652609795355254,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.6780241029646975e-05,
+      "loss": 1.5162,
+      "step": 2246
+    },
+    {
+      "epoch": 0.1653345596688894,
+      "grad_norm": 1.3359375,
+      "learning_rate": 4.677739728676044e-05,
+      "loss": 1.1994,
+      "step": 2247
+    },
+    {
+      "epoch": 0.16540813980225338,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.677455237512212e-05,
+      "loss": 1.0858,
+      "step": 2248
+    },
+    {
+      "epoch": 0.1654817199356174,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.6771706294884696e-05,
+      "loss": 0.7822,
+      "step": 2249
+    },
+    {
+      "epoch": 0.16555530006898136,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.6768859046200924e-05,
+      "loss": 1.2687,
+      "step": 2250
+    },
+    {
+      "epoch": 0.16562888020234537,
+      "grad_norm": 0.875,
+      "learning_rate": 4.67660106292236e-05,
+      "loss": 1.0894,
+      "step": 2251
+    },
+    {
+      "epoch": 0.16570246033570935,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.6763161044105595e-05,
+      "loss": 0.7779,
+      "step": 2252
+    },
+    {
+      "epoch": 0.16577604046907335,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.6760310290999844e-05,
+      "loss": 1.0385,
+      "step": 2253
+    },
+    {
+      "epoch": 0.16584962060243735,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.6757458370059336e-05,
+      "loss": 1.1762,
+      "step": 2254
+    },
+    {
+      "epoch": 0.16592320073580133,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.6754605281437134e-05,
+      "loss": 0.7289,
+      "step": 2255
+    },
+    {
+      "epoch": 0.16599678086916533,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.675175102528635e-05,
+      "loss": 0.7508,
+      "step": 2256
+    },
+    {
+      "epoch": 0.1660703610025293,
+      "grad_norm": 1.125,
+      "learning_rate": 4.674889560176018e-05,
+      "loss": 1.3737,
+      "step": 2257
+    },
+    {
+      "epoch": 0.1661439411358933,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.674603901101186e-05,
+      "loss": 0.8059,
+      "step": 2258
+    },
+    {
+      "epoch": 0.1662175212692573,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.674318125319469e-05,
+      "loss": 0.723,
+      "step": 2259
+    },
+    {
+      "epoch": 0.1662911014026213,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.674032232846205e-05,
+      "loss": 0.841,
+      "step": 2260
+    },
+    {
+      "epoch": 0.16636468153598527,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.6737462236967374e-05,
+      "loss": 0.9784,
+      "step": 2261
+    },
+    {
+      "epoch": 0.16643826166934927,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.6734600978864164e-05,
+      "loss": 0.8706,
+      "step": 2262
+    },
+    {
+      "epoch": 0.16651184180271328,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.673173855430596e-05,
+      "loss": 0.9619,
+      "step": 2263
+    },
+    {
+      "epoch": 0.16658542193607725,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.67288749634464e-05,
+      "loss": 0.6256,
+      "step": 2264
+    },
+    {
+      "epoch": 0.16665900206944126,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.6726010206439155e-05,
+      "loss": 0.7546,
+      "step": 2265
+    },
+    {
+      "epoch": 0.16673258220280524,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.672314428343798e-05,
+      "loss": 0.8527,
+      "step": 2266
+    },
+    {
+      "epoch": 0.16680616233616924,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.672027719459668e-05,
+      "loss": 0.833,
+      "step": 2267
+    },
+    {
+      "epoch": 0.16687974246953322,
+      "grad_norm": 0.875,
+      "learning_rate": 4.671740894006913e-05,
+      "loss": 1.0085,
+      "step": 2268
+    },
+    {
+      "epoch": 0.16695332260289722,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.671453952000926e-05,
+      "loss": 1.1527,
+      "step": 2269
+    },
+    {
+      "epoch": 0.1670269027362612,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.671166893457106e-05,
+      "loss": 0.944,
+      "step": 2270
+    },
+    {
+      "epoch": 0.1671004828696252,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.67087971839086e-05,
+      "loss": 1.3125,
+      "step": 2271
+    },
+    {
+      "epoch": 0.1671740630029892,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.6705924268176e-05,
+      "loss": 0.7686,
+      "step": 2272
+    },
+    {
+      "epoch": 0.16724764313635318,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.670305018752744e-05,
+      "loss": 0.7776,
+      "step": 2273
+    },
+    {
+      "epoch": 0.16732122326971718,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.6700174942117165e-05,
+      "loss": 1.1477,
+      "step": 2274
+    },
+    {
+      "epoch": 0.16739480340308116,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.669729853209949e-05,
+      "loss": 1.0077,
+      "step": 2275
+    },
+    {
+      "epoch": 0.16746838353644516,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.6694420957628785e-05,
+      "loss": 0.7071,
+      "step": 2276
+    },
+    {
+      "epoch": 0.16754196366980914,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.6691542218859476e-05,
+      "loss": 1.012,
+      "step": 2277
+    },
+    {
+      "epoch": 0.16761554380317314,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.668866231594606e-05,
+      "loss": 0.8759,
+      "step": 2278
+    },
+    {
+      "epoch": 0.16768912393653712,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.668578124904312e-05,
+      "loss": 0.9433,
+      "step": 2279
+    },
+    {
+      "epoch": 0.16776270406990113,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.668289901830524e-05,
+      "loss": 0.8655,
+      "step": 2280
+    },
+    {
+      "epoch": 0.16783628420326513,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.668001562388713e-05,
+      "loss": 0.7792,
+      "step": 2281
+    },
+    {
+      "epoch": 0.1679098643366291,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.667713106594353e-05,
+      "loss": 1.2321,
+      "step": 2282
+    },
+    {
+      "epoch": 0.1679834444699931,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.667424534462925e-05,
+      "loss": 1.2005,
+      "step": 2283
+    },
+    {
+      "epoch": 0.16805702460335709,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.667135846009916e-05,
+      "loss": 1.1626,
+      "step": 2284
+    },
+    {
+      "epoch": 0.1681306047367211,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.666847041250819e-05,
+      "loss": 1.0837,
+      "step": 2285
+    },
+    {
+      "epoch": 0.16820418487008507,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.6665581202011345e-05,
+      "loss": 1.2828,
+      "step": 2286
+    },
+    {
+      "epoch": 0.16827776500344907,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.666269082876367e-05,
+      "loss": 0.7669,
+      "step": 2287
+    },
+    {
+      "epoch": 0.16835134513681307,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.665979929292029e-05,
+      "loss": 1.0711,
+      "step": 2288
+    },
+    {
+      "epoch": 0.16842492527017705,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.665690659463641e-05,
+      "loss": 0.7772,
+      "step": 2289
+    },
+    {
+      "epoch": 0.16849850540354105,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.6654012734067236e-05,
+      "loss": 0.7368,
+      "step": 2290
+    },
+    {
+      "epoch": 0.16857208553690503,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.665111771136811e-05,
+      "loss": 1.0117,
+      "step": 2291
+    },
+    {
+      "epoch": 0.16864566567026903,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.664822152669438e-05,
+      "loss": 0.8384,
+      "step": 2292
+    },
+    {
+      "epoch": 0.168719245803633,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.6645324180201494e-05,
+      "loss": 0.7722,
+      "step": 2293
+    },
+    {
+      "epoch": 0.16879282593699702,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.664242567204494e-05,
+      "loss": 0.7028,
+      "step": 2294
+    },
+    {
+      "epoch": 0.168866406070361,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.6639526002380275e-05,
+      "loss": 1.0704,
+      "step": 2295
+    },
+    {
+      "epoch": 0.168939986203725,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.663662517136312e-05,
+      "loss": 1.3884,
+      "step": 2296
+    },
+    {
+      "epoch": 0.169013566337089,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.6633723179149166e-05,
+      "loss": 0.7911,
+      "step": 2297
+    },
+    {
+      "epoch": 0.16908714647045298,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.6630820025894145e-05,
+      "loss": 1.1098,
+      "step": 2298
+    },
+    {
+      "epoch": 0.16916072660381698,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.6627915711753866e-05,
+      "loss": 0.8922,
+      "step": 2299
+    },
+    {
+      "epoch": 0.16923430673718096,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.6625010236884204e-05,
+      "loss": 1.4102,
+      "step": 2300
+    },
+    {
+      "epoch": 0.16930788687054496,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.662210360144108e-05,
+      "loss": 0.8983,
+      "step": 2301
+    },
+    {
+      "epoch": 0.16938146700390894,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.66191958055805e-05,
+      "loss": 0.8754,
+      "step": 2302
+    },
+    {
+      "epoch": 0.16945504713727294,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.6616286849458515e-05,
+      "loss": 1.2154,
+      "step": 2303
+    },
+    {
+      "epoch": 0.16952862727063692,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.6613376733231236e-05,
+      "loss": 0.8395,
+      "step": 2304
+    },
+    {
+      "epoch": 0.16960220740400092,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.661046545705485e-05,
+      "loss": 0.8515,
+      "step": 2305
+    },
+    {
+      "epoch": 0.16967578753736492,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.660755302108561e-05,
+      "loss": 0.8633,
+      "step": 2306
+    },
+    {
+      "epoch": 0.1697493676707289,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.66046394254798e-05,
+      "loss": 0.9357,
+      "step": 2307
+    },
+    {
+      "epoch": 0.1698229478040929,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.66017246703938e-05,
+      "loss": 1.0117,
+      "step": 2308
+    },
+    {
+      "epoch": 0.16989652793745688,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.6598808755984034e-05,
+      "loss": 1.0174,
+      "step": 2309
+    },
+    {
+      "epoch": 0.16997010807082089,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.6595891682407e-05,
+      "loss": 1.234,
+      "step": 2310
+    },
+    {
+      "epoch": 0.17004368820418486,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.6592973449819244e-05,
+      "loss": 0.5915,
+      "step": 2311
+    },
+    {
+      "epoch": 0.17011726833754887,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.659005405837739e-05,
+      "loss": 0.9665,
+      "step": 2312
+    },
+    {
+      "epoch": 0.17019084847091284,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.6587133508238115e-05,
+      "loss": 1.6637,
+      "step": 2313
+    },
+    {
+      "epoch": 0.17026442860427685,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.658421179955815e-05,
+      "loss": 0.8542,
+      "step": 2314
+    },
+    {
+      "epoch": 0.17033800873764085,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.6581288932494304e-05,
+      "loss": 0.7991,
+      "step": 2315
+    },
+    {
+      "epoch": 0.17041158887100483,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.657836490720345e-05,
+      "loss": 1.3094,
+      "step": 2316
+    },
+    {
+      "epoch": 0.17048516900436883,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.657543972384251e-05,
+      "loss": 0.86,
+      "step": 2317
+    },
+    {
+      "epoch": 0.1705587491377328,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.657251338256846e-05,
+      "loss": 0.7694,
+      "step": 2318
+    },
+    {
+      "epoch": 0.1706323292710968,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.656958588353836e-05,
+      "loss": 0.9706,
+      "step": 2319
+    },
+    {
+      "epoch": 0.1707059094044608,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.656665722690933e-05,
+      "loss": 0.9386,
+      "step": 2320
+    },
+    {
+      "epoch": 0.1707794895378248,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.656372741283854e-05,
+      "loss": 1.0143,
+      "step": 2321
+    },
+    {
+      "epoch": 0.17085306967118877,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.6560796441483234e-05,
+      "loss": 1.4127,
+      "step": 2322
+    },
+    {
+      "epoch": 0.17092664980455277,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.6557864313000695e-05,
+      "loss": 0.7166,
+      "step": 2323
+    },
+    {
+      "epoch": 0.17100022993791678,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.65549310275483e-05,
+      "loss": 0.9397,
+      "step": 2324
+    },
+    {
+      "epoch": 0.17107381007128075,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.6551996585283476e-05,
+      "loss": 1.0492,
+      "step": 2325
+    },
+    {
+      "epoch": 0.17114739020464476,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.654906098636369e-05,
+      "loss": 0.6942,
+      "step": 2326
+    },
+    {
+      "epoch": 0.17122097033800873,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.6546124230946505e-05,
+      "loss": 0.9402,
+      "step": 2327
+    },
+    {
+      "epoch": 0.17129455047137274,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.6543186319189526e-05,
+      "loss": 0.7261,
+      "step": 2328
+    },
+    {
+      "epoch": 0.1713681306047367,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.6540247251250424e-05,
+      "loss": 0.8986,
+      "step": 2329
+    },
+    {
+      "epoch": 0.17144171073810072,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.653730702728694e-05,
+      "loss": 0.7865,
+      "step": 2330
+    },
+    {
+      "epoch": 0.1715152908714647,
+      "grad_norm": 1.59375,
+      "learning_rate": 4.653436564745687e-05,
+      "loss": 1.1759,
+      "step": 2331
+    },
+    {
+      "epoch": 0.1715888710048287,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.653142311191806e-05,
+      "loss": 0.978,
+      "step": 2332
+    },
+    {
+      "epoch": 0.1716624511381927,
+      "grad_norm": 1.125,
+      "learning_rate": 4.652847942082844e-05,
+      "loss": 1.0943,
+      "step": 2333
+    },
+    {
+      "epoch": 0.17173603127155668,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.6525534574346e-05,
+      "loss": 0.942,
+      "step": 2334
+    },
+    {
+      "epoch": 0.17180961140492068,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.6522588572628765e-05,
+      "loss": 1.0099,
+      "step": 2335
+    },
+    {
+      "epoch": 0.17188319153828466,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.651964141583486e-05,
+      "loss": 0.7833,
+      "step": 2336
+    },
+    {
+      "epoch": 0.17195677167164866,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.6516693104122435e-05,
+      "loss": 1.1503,
+      "step": 2337
+    },
+    {
+      "epoch": 0.17203035180501264,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.6513743637649736e-05,
+      "loss": 0.6584,
+      "step": 2338
+    },
+    {
+      "epoch": 0.17210393193837664,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.651079301657505e-05,
+      "loss": 0.8908,
+      "step": 2339
+    },
+    {
+      "epoch": 0.17217751207174062,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.6507841241056735e-05,
+      "loss": 0.8444,
+      "step": 2340
+    },
+    {
+      "epoch": 0.17225109220510462,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.6504888311253196e-05,
+      "loss": 0.8007,
+      "step": 2341
+    },
+    {
+      "epoch": 0.17232467233846863,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.650193422732292e-05,
+      "loss": 0.7184,
+      "step": 2342
+    },
+    {
+      "epoch": 0.1723982524718326,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.649897898942445e-05,
+      "loss": 0.903,
+      "step": 2343
+    },
+    {
+      "epoch": 0.1724718326051966,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.649602259771638e-05,
+      "loss": 0.7455,
+      "step": 2344
+    },
+    {
+      "epoch": 0.17254541273856058,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.649306505235738e-05,
+      "loss": 0.7128,
+      "step": 2345
+    },
+    {
+      "epoch": 0.1726189928719246,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.649010635350617e-05,
+      "loss": 0.9328,
+      "step": 2346
+    },
+    {
+      "epoch": 0.17269257300528856,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.6487146501321535e-05,
+      "loss": 0.7971,
+      "step": 2347
+    },
+    {
+      "epoch": 0.17276615313865257,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.648418549596234e-05,
+      "loss": 0.9742,
+      "step": 2348
+    },
+    {
+      "epoch": 0.17283973327201654,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.648122333758749e-05,
+      "loss": 0.9682,
+      "step": 2349
+    },
+    {
+      "epoch": 0.17291331340538055,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.647826002635595e-05,
+      "loss": 0.7815,
+      "step": 2350
+    },
+    {
+      "epoch": 0.17298689353874455,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.647529556242676e-05,
+      "loss": 0.9898,
+      "step": 2351
+    },
+    {
+      "epoch": 0.17306047367210853,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.6472329945959014e-05,
+      "loss": 1.1301,
+      "step": 2352
+    },
+    {
+      "epoch": 0.17313405380547253,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.646936317711188e-05,
+      "loss": 0.8775,
+      "step": 2353
+    },
+    {
+      "epoch": 0.1732076339388365,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.6466395256044574e-05,
+      "loss": 1.4642,
+      "step": 2354
+    },
+    {
+      "epoch": 0.1732812140722005,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.646342618291638e-05,
+      "loss": 0.7355,
+      "step": 2355
+    },
+    {
+      "epoch": 0.1733547942055645,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.6460455957886646e-05,
+      "loss": 0.6268,
+      "step": 2356
+    },
+    {
+      "epoch": 0.1734283743389285,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.645748458111476e-05,
+      "loss": 1.1092,
+      "step": 2357
+    },
+    {
+      "epoch": 0.17350195447229247,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.6454512052760225e-05,
+      "loss": 1.0355,
+      "step": 2358
+    },
+    {
+      "epoch": 0.17357553460565647,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.6451538372982527e-05,
+      "loss": 1.0261,
+      "step": 2359
+    },
+    {
+      "epoch": 0.17364911473902048,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.6448563541941295e-05,
+      "loss": 1.1197,
+      "step": 2360
+    },
+    {
+      "epoch": 0.17372269487238445,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.6445587559796166e-05,
+      "loss": 0.8631,
+      "step": 2361
+    },
+    {
+      "epoch": 0.17379627500574846,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.6442610426706856e-05,
+      "loss": 1.0614,
+      "step": 2362
+    },
+    {
+      "epoch": 0.17386985513911243,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.643963214283314e-05,
+      "loss": 0.8628,
+      "step": 2363
+    },
+    {
+      "epoch": 0.17394343527247644,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.643665270833487e-05,
+      "loss": 0.8224,
+      "step": 2364
+    },
+    {
+      "epoch": 0.17401701540584041,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.643367212337193e-05,
+      "loss": 0.8321,
+      "step": 2365
+    },
+    {
+      "epoch": 0.17409059553920442,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.643069038810429e-05,
+      "loss": 0.9729,
+      "step": 2366
+    },
+    {
+      "epoch": 0.1741641756725684,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.642770750269198e-05,
+      "loss": 0.7919,
+      "step": 2367
+    },
+    {
+      "epoch": 0.1742377558059324,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.642472346729507e-05,
+      "loss": 1.251,
+      "step": 2368
+    },
+    {
+      "epoch": 0.1743113359392964,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.642173828207372e-05,
+      "loss": 1.0962,
+      "step": 2369
+    },
+    {
+      "epoch": 0.17438491607266038,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.6418751947188145e-05,
+      "loss": 1.1477,
+      "step": 2370
+    },
+    {
+      "epoch": 0.17445849620602438,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.641576446279861e-05,
+      "loss": 0.7813,
+      "step": 2371
+    },
+    {
+      "epoch": 0.17453207633938836,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.641277582906542e-05,
+      "loss": 1.1445,
+      "step": 2372
+    },
+    {
+      "epoch": 0.17460565647275236,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.640978604614902e-05,
+      "loss": 0.9483,
+      "step": 2373
+    },
+    {
+      "epoch": 0.17467923660611634,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.640679511420983e-05,
+      "loss": 0.9605,
+      "step": 2374
+    },
+    {
+      "epoch": 0.17475281673948034,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.6403803033408375e-05,
+      "loss": 0.8761,
+      "step": 2375
+    },
+    {
+      "epoch": 0.17482639687284432,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.6400809803905244e-05,
+      "loss": 0.6966,
+      "step": 2376
+    },
+    {
+      "epoch": 0.17489997700620832,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.639781542586106e-05,
+      "loss": 0.7238,
+      "step": 2377
+    },
+    {
+      "epoch": 0.17497355713957233,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.639481989943655e-05,
+      "loss": 0.8654,
+      "step": 2378
+    },
+    {
+      "epoch": 0.1750471372729363,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.6391823224792456e-05,
+      "loss": 0.9046,
+      "step": 2379
+    },
+    {
+      "epoch": 0.1751207174063003,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.638882540208962e-05,
+      "loss": 0.817,
+      "step": 2380
+    },
+    {
+      "epoch": 0.17519429753966428,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.6385826431488914e-05,
+      "loss": 1.4932,
+      "step": 2381
+    },
+    {
+      "epoch": 0.1752678776730283,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.63828263131513e-05,
+      "loss": 0.8208,
+      "step": 2382
+    },
+    {
+      "epoch": 0.17534145780639226,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.637982504723779e-05,
+      "loss": 0.847,
+      "step": 2383
+    },
+    {
+      "epoch": 0.17541503793975627,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.637682263390944e-05,
+      "loss": 0.6693,
+      "step": 2384
+    },
+    {
+      "epoch": 0.17548861807312024,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.6373819073327403e-05,
+      "loss": 0.7124,
+      "step": 2385
+    },
+    {
+      "epoch": 0.17556219820648425,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.637081436565286e-05,
+      "loss": 1.0774,
+      "step": 2386
+    },
+    {
+      "epoch": 0.17563577833984825,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.636780851104707e-05,
+      "loss": 0.9835,
+      "step": 2387
+    },
+    {
+      "epoch": 0.17570935847321223,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.636480150967136e-05,
+      "loss": 0.9992,
+      "step": 2388
+    },
+    {
+      "epoch": 0.17578293860657623,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.636179336168711e-05,
+      "loss": 1.0624,
+      "step": 2389
+    },
+    {
+      "epoch": 0.1758565187399402,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.6358784067255755e-05,
+      "loss": 0.9743,
+      "step": 2390
+    },
+    {
+      "epoch": 0.1759300988733042,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.63557736265388e-05,
+      "loss": 0.8359,
+      "step": 2391
+    },
+    {
+      "epoch": 0.1760036790066682,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.635276203969781e-05,
+      "loss": 0.8541,
+      "step": 2392
+    },
+    {
+      "epoch": 0.1760772591400322,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.634974930689441e-05,
+      "loss": 0.8042,
+      "step": 2393
+    },
+    {
+      "epoch": 0.17615083927339617,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.634673542829029e-05,
+      "loss": 0.9784,
+      "step": 2394
+    },
+    {
+      "epoch": 0.17622441940676017,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.634372040404719e-05,
+      "loss": 0.9491,
+      "step": 2395
+    },
+    {
+      "epoch": 0.17629799954012418,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.6340704234326934e-05,
+      "loss": 0.9291,
+      "step": 2396
+    },
+    {
+      "epoch": 0.17637157967348815,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.633768691929139e-05,
+      "loss": 1.245,
+      "step": 2397
+    },
+    {
+      "epoch": 0.17644515980685216,
+      "grad_norm": 0.75,
+      "learning_rate": 4.6334668459102484e-05,
+      "loss": 0.9182,
+      "step": 2398
+    },
+    {
+      "epoch": 0.17651873994021614,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.6331648853922225e-05,
+      "loss": 1.4692,
+      "step": 2399
+    },
+    {
+      "epoch": 0.17659232007358014,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.6328628103912666e-05,
+      "loss": 0.989,
+      "step": 2400
+    },
+    {
+      "epoch": 0.17666590020694412,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.632560620923591e-05,
+      "loss": 0.8725,
+      "step": 2401
+    },
+    {
+      "epoch": 0.17673948034030812,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.6322583170054146e-05,
+      "loss": 1.1465,
+      "step": 2402
+    },
+    {
+      "epoch": 0.1768130604736721,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.631955898652962e-05,
+      "loss": 0.6109,
+      "step": 2403
+    },
+    {
+      "epoch": 0.1768866406070361,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.631653365882463e-05,
+      "loss": 1.066,
+      "step": 2404
+    },
+    {
+      "epoch": 0.1769602207404001,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.6313507187101544e-05,
+      "loss": 0.8509,
+      "step": 2405
+    },
+    {
+      "epoch": 0.17703380087376408,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.631047957152278e-05,
+      "loss": 0.7434,
+      "step": 2406
+    },
+    {
+      "epoch": 0.17710738100712808,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.630745081225083e-05,
+      "loss": 1.0909,
+      "step": 2407
+    },
+    {
+      "epoch": 0.17718096114049206,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.6304420909448235e-05,
+      "loss": 1.0208,
+      "step": 2408
+    },
+    {
+      "epoch": 0.17725454127385606,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.63013898632776e-05,
+      "loss": 1.0963,
+      "step": 2409
+    },
+    {
+      "epoch": 0.17732812140722004,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.6298357673901615e-05,
+      "loss": 0.8698,
+      "step": 2410
+    },
+    {
+      "epoch": 0.17740170154058404,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.6295324341483e-05,
+      "loss": 0.945,
+      "step": 2411
+    },
+    {
+      "epoch": 0.17747528167394802,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.6292289866184546e-05,
+      "loss": 0.9404,
+      "step": 2412
+    },
+    {
+      "epoch": 0.17754886180731203,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.628925424816911e-05,
+      "loss": 0.7535,
+      "step": 2413
+    },
+    {
+      "epoch": 0.17762244194067603,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.6286217487599616e-05,
+      "loss": 0.991,
+      "step": 2414
+    },
+    {
+      "epoch": 0.17769602207404,
+      "grad_norm": 1.125,
+      "learning_rate": 4.628317958463902e-05,
+      "loss": 1.5002,
+      "step": 2415
+    },
+    {
+      "epoch": 0.177769602207404,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.628014053945038e-05,
+      "loss": 0.8467,
+      "step": 2416
+    },
+    {
+      "epoch": 0.17784318234076799,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.627710035219679e-05,
+      "loss": 0.8584,
+      "step": 2417
+    },
+    {
+      "epoch": 0.177916762474132,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.627405902304141e-05,
+      "loss": 0.6925,
+      "step": 2418
+    },
+    {
+      "epoch": 0.17799034260749597,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.627101655214746e-05,
+      "loss": 1.1767,
+      "step": 2419
+    },
+    {
+      "epoch": 0.17806392274085997,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.626797293967824e-05,
+      "loss": 1.1363,
+      "step": 2420
+    },
+    {
+      "epoch": 0.17813750287422395,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.626492818579707e-05,
+      "loss": 0.9227,
+      "step": 2421
+    },
+    {
+      "epoch": 0.17821108300758795,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.626188229066737e-05,
+      "loss": 0.6957,
+      "step": 2422
+    },
+    {
+      "epoch": 0.17828466314095195,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.62588352544526e-05,
+      "loss": 0.9482,
+      "step": 2423
+    },
+    {
+      "epoch": 0.17835824327431593,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.62557870773163e-05,
+      "loss": 0.9393,
+      "step": 2424
+    },
+    {
+      "epoch": 0.17843182340767993,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.625273775942206e-05,
+      "loss": 1.0312,
+      "step": 2425
+    },
+    {
+      "epoch": 0.1785054035410439,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.6249687300933516e-05,
+      "loss": 1.1365,
+      "step": 2426
+    },
+    {
+      "epoch": 0.17857898367440792,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.6246635702014396e-05,
+      "loss": 0.6551,
+      "step": 2427
+    },
+    {
+      "epoch": 0.1786525638077719,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.6243582962828466e-05,
+      "loss": 1.3892,
+      "step": 2428
+    },
+    {
+      "epoch": 0.1787261439411359,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.6240529083539564e-05,
+      "loss": 0.9388,
+      "step": 2429
+    },
+    {
+      "epoch": 0.17879972407449987,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.6237474064311574e-05,
+      "loss": 1.5013,
+      "step": 2430
+    },
+    {
+      "epoch": 0.17887330420786388,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.623441790530847e-05,
+      "loss": 0.9194,
+      "step": 2431
+    },
+    {
+      "epoch": 0.17894688434122788,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.6231360606694263e-05,
+      "loss": 1.0206,
+      "step": 2432
+    },
+    {
+      "epoch": 0.17902046447459186,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.622830216863303e-05,
+      "loss": 0.9787,
+      "step": 2433
+    },
+    {
+      "epoch": 0.17909404460795586,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.6225242591288914e-05,
+      "loss": 0.692,
+      "step": 2434
+    },
+    {
+      "epoch": 0.17916762474131984,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.622218187482612e-05,
+      "loss": 1.0808,
+      "step": 2435
+    },
+    {
+      "epoch": 0.17924120487468384,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.62191200194089e-05,
+      "loss": 0.9005,
+      "step": 2436
+    },
+    {
+      "epoch": 0.17931478500804782,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.62160570252016e-05,
+      "loss": 0.9334,
+      "step": 2437
+    },
+    {
+      "epoch": 0.17938836514141182,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.621299289236858e-05,
+      "loss": 0.9676,
+      "step": 2438
+    },
+    {
+      "epoch": 0.1794619452747758,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.62099276210743e-05,
+      "loss": 1.203,
+      "step": 2439
+    },
+    {
+      "epoch": 0.1795355254081398,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.620686121148326e-05,
+      "loss": 1.0788,
+      "step": 2440
+    },
+    {
+      "epoch": 0.1796091055415038,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.620379366376004e-05,
+      "loss": 0.7248,
+      "step": 2441
+    },
+    {
+      "epoch": 0.17968268567486778,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.620072497806926e-05,
+      "loss": 1.0184,
+      "step": 2442
+    },
+    {
+      "epoch": 0.17975626580823179,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.6197655154575615e-05,
+      "loss": 0.9649,
+      "step": 2443
+    },
+    {
+      "epoch": 0.17982984594159576,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.619458419344385e-05,
+      "loss": 0.9972,
+      "step": 2444
+    },
+    {
+      "epoch": 0.17990342607495977,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.619151209483878e-05,
+      "loss": 1.0228,
+      "step": 2445
+    },
+    {
+      "epoch": 0.17997700620832374,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.618843885892529e-05,
+      "loss": 0.9985,
+      "step": 2446
+    },
+    {
+      "epoch": 0.18005058634168775,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.61853644858683e-05,
+      "loss": 1.6175,
+      "step": 2447
+    },
+    {
+      "epoch": 0.18012416647505172,
+      "grad_norm": 1.25,
+      "learning_rate": 4.618228897583281e-05,
+      "loss": 1.115,
+      "step": 2448
+    },
+    {
+      "epoch": 0.18019774660841573,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.617921232898388e-05,
+      "loss": 1.3514,
+      "step": 2449
+    },
+    {
+      "epoch": 0.18027132674177973,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.617613454548663e-05,
+      "loss": 0.9386,
+      "step": 2450
+    },
+    {
+      "epoch": 0.1803449068751437,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.6173055625506236e-05,
+      "loss": 0.8495,
+      "step": 2451
+    },
+    {
+      "epoch": 0.1804184870085077,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.616997556920793e-05,
+      "loss": 0.7815,
+      "step": 2452
+    },
+    {
+      "epoch": 0.1804920671418717,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.616689437675702e-05,
+      "loss": 0.9896,
+      "step": 2453
+    },
+    {
+      "epoch": 0.1805656472752357,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.616381204831887e-05,
+      "loss": 0.9494,
+      "step": 2454
+    },
+    {
+      "epoch": 0.18063922740859967,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.61607285840589e-05,
+      "loss": 0.6253,
+      "step": 2455
+    },
+    {
+      "epoch": 0.18071280754196367,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.6157643984142595e-05,
+      "loss": 0.9087,
+      "step": 2456
+    },
+    {
+      "epoch": 0.18078638767532765,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.61545582487355e-05,
+      "loss": 1.4828,
+      "step": 2457
+    },
+    {
+      "epoch": 0.18085996780869165,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.615147137800321e-05,
+      "loss": 1.1915,
+      "step": 2458
+    },
+    {
+      "epoch": 0.18093354794205566,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.6148383372111406e-05,
+      "loss": 1.0188,
+      "step": 2459
+    },
+    {
+      "epoch": 0.18100712807541963,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.6145294231225816e-05,
+      "loss": 1.111,
+      "step": 2460
+    },
+    {
+      "epoch": 0.18108070820878364,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.614220395551222e-05,
+      "loss": 1.0745,
+      "step": 2461
+    },
+    {
+      "epoch": 0.1811542883421476,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.6139112545136466e-05,
+      "loss": 0.8347,
+      "step": 2462
+    },
+    {
+      "epoch": 0.18122786847551162,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.6136020000264466e-05,
+      "loss": 0.7947,
+      "step": 2463
+    },
+    {
+      "epoch": 0.1813014486088756,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.61329263210622e-05,
+      "loss": 0.8838,
+      "step": 2464
+    },
+    {
+      "epoch": 0.1813750287422396,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.6129831507695684e-05,
+      "loss": 0.7092,
+      "step": 2465
+    },
+    {
+      "epoch": 0.1814486088756036,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.612673556033103e-05,
+      "loss": 0.6203,
+      "step": 2466
+    },
+    {
+      "epoch": 0.18152218900896758,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.612363847913437e-05,
+      "loss": 1.364,
+      "step": 2467
+    },
+    {
+      "epoch": 0.18159576914233158,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.612054026427193e-05,
+      "loss": 0.9108,
+      "step": 2468
+    },
+    {
+      "epoch": 0.18166934927569556,
+      "grad_norm": 1.4921875,
+      "learning_rate": 4.611744091590999e-05,
+      "loss": 0.9871,
+      "step": 2469
+    },
+    {
+      "epoch": 0.18174292940905956,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.611434043421489e-05,
+      "loss": 1.2519,
+      "step": 2470
+    },
+    {
+      "epoch": 0.18181650954242354,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.6111238819353005e-05,
+      "loss": 0.969,
+      "step": 2471
+    },
+    {
+      "epoch": 0.18189008967578754,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.610813607149081e-05,
+      "loss": 0.8348,
+      "step": 2472
+    },
+    {
+      "epoch": 0.18196366980915152,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.6105032190794816e-05,
+      "loss": 1.1528,
+      "step": 2473
+    },
+    {
+      "epoch": 0.18203724994251552,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.610192717743162e-05,
+      "loss": 1.0308,
+      "step": 2474
+    },
+    {
+      "epoch": 0.18211083007587953,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.609882103156783e-05,
+      "loss": 1.0408,
+      "step": 2475
+    },
+    {
+      "epoch": 0.1821844102092435,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.6095713753370174e-05,
+      "loss": 0.7387,
+      "step": 2476
+    },
+    {
+      "epoch": 0.1822579903426075,
+      "grad_norm": 1.3515625,
+      "learning_rate": 4.609260534300541e-05,
+      "loss": 0.9955,
+      "step": 2477
+    },
+    {
+      "epoch": 0.18233157047597148,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.608949580064035e-05,
+      "loss": 0.6883,
+      "step": 2478
+    },
+    {
+      "epoch": 0.1824051506093355,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.608638512644188e-05,
+      "loss": 0.7779,
+      "step": 2479
+    },
+    {
+      "epoch": 0.18247873074269946,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.6083273320576945e-05,
+      "loss": 1.2412,
+      "step": 2480
+    },
+    {
+      "epoch": 0.18255231087606347,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.6080160383212556e-05,
+      "loss": 1.2351,
+      "step": 2481
+    },
+    {
+      "epoch": 0.18262589100942744,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.607704631451578e-05,
+      "loss": 0.96,
+      "step": 2482
+    },
+    {
+      "epoch": 0.18269947114279145,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.607393111465373e-05,
+      "loss": 0.9822,
+      "step": 2483
+    },
+    {
+      "epoch": 0.18277305127615545,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.60708147837936e-05,
+      "loss": 1.2662,
+      "step": 2484
+    },
+    {
+      "epoch": 0.18284663140951943,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.6067697322102646e-05,
+      "loss": 1.0083,
+      "step": 2485
+    },
+    {
+      "epoch": 0.18292021154288343,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.606457872974816e-05,
+      "loss": 0.9201,
+      "step": 2486
+    },
+    {
+      "epoch": 0.1829937916762474,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.606145900689751e-05,
+      "loss": 1.021,
+      "step": 2487
+    },
+    {
+      "epoch": 0.1830673718096114,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.605833815371815e-05,
+      "loss": 0.9086,
+      "step": 2488
+    },
+    {
+      "epoch": 0.1831409519429754,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.605521617037755e-05,
+      "loss": 0.643,
+      "step": 2489
+    },
+    {
+      "epoch": 0.1832145320763394,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.6052093057043264e-05,
+      "loss": 1.0791,
+      "step": 2490
+    },
+    {
+      "epoch": 0.18328811220970337,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.604896881388291e-05,
+      "loss": 0.8522,
+      "step": 2491
+    },
+    {
+      "epoch": 0.18336169234306737,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.6045843441064153e-05,
+      "loss": 0.7927,
+      "step": 2492
+    },
+    {
+      "epoch": 0.18343527247643138,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.6042716938754726e-05,
+      "loss": 0.9329,
+      "step": 2493
+    },
+    {
+      "epoch": 0.18350885260979535,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.603958930712242e-05,
+      "loss": 0.8862,
+      "step": 2494
+    },
+    {
+      "epoch": 0.18358243274315936,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.60364605463351e-05,
+      "loss": 1.2713,
+      "step": 2495
+    },
+    {
+      "epoch": 0.18365601287652333,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.603333065656068e-05,
+      "loss": 1.0224,
+      "step": 2496
+    },
+    {
+      "epoch": 0.18372959300988734,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.6030199637967126e-05,
+      "loss": 1.1144,
+      "step": 2497
+    },
+    {
+      "epoch": 0.1838031731432513,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.6027067490722475e-05,
+      "loss": 0.5707,
+      "step": 2498
+    },
+    {
+      "epoch": 0.18387675327661532,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.602393421499483e-05,
+      "loss": 0.7865,
+      "step": 2499
+    },
+    {
+      "epoch": 0.1839503334099793,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.602079981095234e-05,
+      "loss": 0.8931,
+      "step": 2500
+    },
+    {
+      "epoch": 0.1840239135433433,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.6017664278763225e-05,
+      "loss": 1.0626,
+      "step": 2501
+    },
+    {
+      "epoch": 0.1840974936767073,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.6014527618595776e-05,
+      "loss": 0.8422,
+      "step": 2502
+    },
+    {
+      "epoch": 0.18417107381007128,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.601138983061831e-05,
+      "loss": 0.9965,
+      "step": 2503
+    },
+    {
+      "epoch": 0.18424465394343528,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.600825091499924e-05,
+      "loss": 1.3074,
+      "step": 2504
+    },
+    {
+      "epoch": 0.18431823407679926,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.6005110871907024e-05,
+      "loss": 0.6125,
+      "step": 2505
+    },
+    {
+      "epoch": 0.18439181421016326,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.6001969701510186e-05,
+      "loss": 1.3469,
+      "step": 2506
+    },
+    {
+      "epoch": 0.18446539434352724,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.599882740397729e-05,
+      "loss": 1.126,
+      "step": 2507
+    },
+    {
+      "epoch": 0.18453897447689124,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.5995683979476995e-05,
+      "loss": 0.9961,
+      "step": 2508
+    },
+    {
+      "epoch": 0.18461255461025522,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.599253942817799e-05,
+      "loss": 0.789,
+      "step": 2509
+    },
+    {
+      "epoch": 0.18468613474361922,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.598939375024905e-05,
+      "loss": 0.7734,
+      "step": 2510
+    },
+    {
+      "epoch": 0.18475971487698323,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.598624694585899e-05,
+      "loss": 1.2937,
+      "step": 2511
+    },
+    {
+      "epoch": 0.1848332950103472,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.598309901517669e-05,
+      "loss": 0.9011,
+      "step": 2512
+    },
+    {
+      "epoch": 0.1849068751437112,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.59799499583711e-05,
+      "loss": 1.3486,
+      "step": 2513
+    },
+    {
+      "epoch": 0.18498045527707518,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.597679977561122e-05,
+      "loss": 0.8625,
+      "step": 2514
+    },
+    {
+      "epoch": 0.1850540354104392,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.597364846706612e-05,
+      "loss": 0.6354,
+      "step": 2515
+    },
+    {
+      "epoch": 0.18512761554380316,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.597049603290491e-05,
+      "loss": 0.649,
+      "step": 2516
+    },
+    {
+      "epoch": 0.18520119567716717,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.5967342473296794e-05,
+      "loss": 0.984,
+      "step": 2517
+    },
+    {
+      "epoch": 0.18527477581053114,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.5964187788411004e-05,
+      "loss": 0.7508,
+      "step": 2518
+    },
+    {
+      "epoch": 0.18534835594389515,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.596103197841686e-05,
+      "loss": 0.7298,
+      "step": 2519
+    },
+    {
+      "epoch": 0.18542193607725915,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.595787504348371e-05,
+      "loss": 0.8501,
+      "step": 2520
+    },
+    {
+      "epoch": 0.18549551621062313,
+      "grad_norm": 1.7421875,
+      "learning_rate": 4.5954716983780995e-05,
+      "loss": 1.4432,
+      "step": 2521
+    },
+    {
+      "epoch": 0.18556909634398713,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.59515577994782e-05,
+      "loss": 0.9085,
+      "step": 2522
+    },
+    {
+      "epoch": 0.1856426764773511,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.594839749074486e-05,
+      "loss": 1.2709,
+      "step": 2523
+    },
+    {
+      "epoch": 0.1857162566107151,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.59452360577506e-05,
+      "loss": 1.0379,
+      "step": 2524
+    },
+    {
+      "epoch": 0.1857898367440791,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.5942073500665076e-05,
+      "loss": 1.0317,
+      "step": 2525
+    },
+    {
+      "epoch": 0.1858634168774431,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.593890981965803e-05,
+      "loss": 0.8333,
+      "step": 2526
+    },
+    {
+      "epoch": 0.18593699701080707,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.593574501489923e-05,
+      "loss": 1.1169,
+      "step": 2527
+    },
+    {
+      "epoch": 0.18601057714417107,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.5932579086558545e-05,
+      "loss": 0.9457,
+      "step": 2528
+    },
+    {
+      "epoch": 0.18608415727753508,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.592941203480587e-05,
+      "loss": 1.0812,
+      "step": 2529
+    },
+    {
+      "epoch": 0.18615773741089905,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.592624385981119e-05,
+      "loss": 0.7941,
+      "step": 2530
+    },
+    {
+      "epoch": 0.18623131754426306,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.592307456174452e-05,
+      "loss": 1.1419,
+      "step": 2531
+    },
+    {
+      "epoch": 0.18630489767762703,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.591990414077596e-05,
+      "loss": 0.9335,
+      "step": 2532
+    },
+    {
+      "epoch": 0.18637847781099104,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.5916732597075653e-05,
+      "loss": 0.694,
+      "step": 2533
+    },
+    {
+      "epoch": 0.18645205794435502,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.591355993081382e-05,
+      "loss": 1.0429,
+      "step": 2534
+    },
+    {
+      "epoch": 0.18652563807771902,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.591038614216072e-05,
+      "loss": 1.0098,
+      "step": 2535
+    },
+    {
+      "epoch": 0.186599218211083,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.590721123128669e-05,
+      "loss": 0.8707,
+      "step": 2536
+    },
+    {
+      "epoch": 0.186672798344447,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.590403519836212e-05,
+      "loss": 0.9409,
+      "step": 2537
+    },
+    {
+      "epoch": 0.186746378477811,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.590085804355747e-05,
+      "loss": 1.2149,
+      "step": 2538
+    },
+    {
+      "epoch": 0.18681995861117498,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.5897679767043244e-05,
+      "loss": 0.6591,
+      "step": 2539
+    },
+    {
+      "epoch": 0.18689353874453898,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.589450036899001e-05,
+      "loss": 1.385,
+      "step": 2540
+    },
+    {
+      "epoch": 0.18696711887790296,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.5891319849568406e-05,
+      "loss": 1.0407,
+      "step": 2541
+    },
+    {
+      "epoch": 0.18704069901126696,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.5888138208949126e-05,
+      "loss": 0.8642,
+      "step": 2542
+    },
+    {
+      "epoch": 0.18711427914463094,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.5884955447302916e-05,
+      "loss": 0.9204,
+      "step": 2543
+    },
+    {
+      "epoch": 0.18718785927799494,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.58817715648006e-05,
+      "loss": 0.8527,
+      "step": 2544
+    },
+    {
+      "epoch": 0.18726143941135892,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.587858656161303e-05,
+      "loss": 1.162,
+      "step": 2545
+    },
+    {
+      "epoch": 0.18733501954472292,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.587540043791117e-05,
+      "loss": 1.3641,
+      "step": 2546
+    },
+    {
+      "epoch": 0.18740859967808693,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.587221319386599e-05,
+      "loss": 0.6579,
+      "step": 2547
+    },
+    {
+      "epoch": 0.1874821798114509,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.5869024829648555e-05,
+      "loss": 1.2344,
+      "step": 2548
+    },
+    {
+      "epoch": 0.1875557599448149,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.586583534542996e-05,
+      "loss": 0.7285,
+      "step": 2549
+    },
+    {
+      "epoch": 0.18762934007817889,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.586264474138141e-05,
+      "loss": 1.0602,
+      "step": 2550
+    },
+    {
+      "epoch": 0.1877029202115429,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.58594530176741e-05,
+      "loss": 0.9655,
+      "step": 2551
+    },
+    {
+      "epoch": 0.18777650034490687,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.585626017447936e-05,
+      "loss": 0.7523,
+      "step": 2552
+    },
+    {
+      "epoch": 0.18785008047827087,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.585306621196853e-05,
+      "loss": 1.1229,
+      "step": 2553
+    },
+    {
+      "epoch": 0.18792366061163485,
+      "grad_norm": 1.3203125,
+      "learning_rate": 4.584987113031301e-05,
+      "loss": 1.4735,
+      "step": 2554
+    },
+    {
+      "epoch": 0.18799724074499885,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.58466749296843e-05,
+      "loss": 0.7991,
+      "step": 2555
+    },
+    {
+      "epoch": 0.18807082087836285,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.5843477610253906e-05,
+      "loss": 0.8562,
+      "step": 2556
+    },
+    {
+      "epoch": 0.18814440101172683,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.584027917219345e-05,
+      "loss": 1.2277,
+      "step": 2557
+    },
+    {
+      "epoch": 0.18821798114509083,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.583707961567456e-05,
+      "loss": 0.7215,
+      "step": 2558
+    },
+    {
+      "epoch": 0.1882915612784548,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.5833878940868966e-05,
+      "loss": 0.7097,
+      "step": 2559
+    },
+    {
+      "epoch": 0.18836514141181881,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.5830677147948445e-05,
+      "loss": 0.9554,
+      "step": 2560
+    },
+    {
+      "epoch": 0.1884387215451828,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.5827474237084824e-05,
+      "loss": 0.7611,
+      "step": 2561
+    },
+    {
+      "epoch": 0.1885123016785468,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.5824270208449996e-05,
+      "loss": 1.4304,
+      "step": 2562
+    },
+    {
+      "epoch": 0.18858588181191077,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.582106506221591e-05,
+      "loss": 0.7863,
+      "step": 2563
+    },
+    {
+      "epoch": 0.18865946194527478,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.5817858798554594e-05,
+      "loss": 0.9078,
+      "step": 2564
+    },
+    {
+      "epoch": 0.18873304207863878,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.581465141763812e-05,
+      "loss": 1.2897,
+      "step": 2565
+    },
+    {
+      "epoch": 0.18880662221200276,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.5811442919638614e-05,
+      "loss": 0.8086,
+      "step": 2566
+    },
+    {
+      "epoch": 0.18888020234536676,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.580823330472827e-05,
+      "loss": 0.9277,
+      "step": 2567
+    },
+    {
+      "epoch": 0.18895378247873074,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.580502257307935e-05,
+      "loss": 0.6848,
+      "step": 2568
+    },
+    {
+      "epoch": 0.18902736261209474,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.5801810724864165e-05,
+      "loss": 0.9872,
+      "step": 2569
+    },
+    {
+      "epoch": 0.18910094274545872,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.5798597760255076e-05,
+      "loss": 1.2729,
+      "step": 2570
+    },
+    {
+      "epoch": 0.18917452287882272,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.579538367942454e-05,
+      "loss": 0.6431,
+      "step": 2571
+    },
+    {
+      "epoch": 0.1892481030121867,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.579216848254504e-05,
+      "loss": 1.1885,
+      "step": 2572
+    },
+    {
+      "epoch": 0.1893216831455507,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.578895216978912e-05,
+      "loss": 0.9254,
+      "step": 2573
+    },
+    {
+      "epoch": 0.1893952632789147,
+      "grad_norm": 0.875,
+      "learning_rate": 4.578573474132941e-05,
+      "loss": 1.0611,
+      "step": 2574
+    },
+    {
+      "epoch": 0.18946884341227868,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.578251619733858e-05,
+      "loss": 0.9559,
+      "step": 2575
+    },
+    {
+      "epoch": 0.18954242354564269,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.5779296537989344e-05,
+      "loss": 1.5196,
+      "step": 2576
+    },
+    {
+      "epoch": 0.18961600367900666,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.577607576345452e-05,
+      "loss": 1.1808,
+      "step": 2577
+    },
+    {
+      "epoch": 0.18968958381237067,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.577285387390694e-05,
+      "loss": 0.8348,
+      "step": 2578
+    },
+    {
+      "epoch": 0.18976316394573464,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.576963086951953e-05,
+      "loss": 1.0998,
+      "step": 2579
+    },
+    {
+      "epoch": 0.18983674407909865,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.576640675046526e-05,
+      "loss": 1.0317,
+      "step": 2580
+    },
+    {
+      "epoch": 0.18991032421246262,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.576318151691716e-05,
+      "loss": 0.882,
+      "step": 2581
+    },
+    {
+      "epoch": 0.18998390434582663,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.5759955169048334e-05,
+      "loss": 0.9327,
+      "step": 2582
+    },
+    {
+      "epoch": 0.19005748447919063,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.575672770703192e-05,
+      "loss": 1.2367,
+      "step": 2583
+    },
+    {
+      "epoch": 0.1901310646125546,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.575349913104113e-05,
+      "loss": 0.913,
+      "step": 2584
+    },
+    {
+      "epoch": 0.1902046447459186,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.575026944124924e-05,
+      "loss": 0.8268,
+      "step": 2585
+    },
+    {
+      "epoch": 0.1902782248792826,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.5747038637829584e-05,
+      "loss": 1.0053,
+      "step": 2586
+    },
+    {
+      "epoch": 0.1903518050126466,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.5743806720955546e-05,
+      "loss": 1.0264,
+      "step": 2587
+    },
+    {
+      "epoch": 0.19042538514601057,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.574057369080058e-05,
+      "loss": 0.8716,
+      "step": 2588
+    },
+    {
+      "epoch": 0.19049896527937457,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.5737339547538204e-05,
+      "loss": 0.9254,
+      "step": 2589
+    },
+    {
+      "epoch": 0.19057254541273855,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.5734104291341974e-05,
+      "loss": 0.8532,
+      "step": 2590
+    },
+    {
+      "epoch": 0.19064612554610255,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.5730867922385536e-05,
+      "loss": 1.2998,
+      "step": 2591
+    },
+    {
+      "epoch": 0.19071970567946656,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.572763044084256e-05,
+      "loss": 0.8121,
+      "step": 2592
+    },
+    {
+      "epoch": 0.19079328581283053,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.572439184688682e-05,
+      "loss": 0.9667,
+      "step": 2593
+    },
+    {
+      "epoch": 0.19086686594619454,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.5721152140692105e-05,
+      "loss": 1.044,
+      "step": 2594
+    },
+    {
+      "epoch": 0.1909404460795585,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.5717911322432295e-05,
+      "loss": 0.8499,
+      "step": 2595
+    },
+    {
+      "epoch": 0.19101402621292252,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.571466939228131e-05,
+      "loss": 1.106,
+      "step": 2596
+    },
+    {
+      "epoch": 0.1910876063462865,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.571142635041314e-05,
+      "loss": 0.5618,
+      "step": 2597
+    },
+    {
+      "epoch": 0.1911611864796505,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.570818219700185e-05,
+      "loss": 0.9936,
+      "step": 2598
+    },
+    {
+      "epoch": 0.19123476661301447,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.570493693222152e-05,
+      "loss": 0.603,
+      "step": 2599
+    },
+    {
+      "epoch": 0.19130834674637848,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.570169055624634e-05,
+      "loss": 0.8996,
+      "step": 2600
+    },
+    {
+      "epoch": 0.19138192687974248,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.569844306925052e-05,
+      "loss": 1.1775,
+      "step": 2601
+    },
+    {
+      "epoch": 0.19145550701310646,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.5695194471408366e-05,
+      "loss": 0.8291,
+      "step": 2602
+    },
+    {
+      "epoch": 0.19152908714647046,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.56919447628942e-05,
+      "loss": 0.9419,
+      "step": 2603
+    },
+    {
+      "epoch": 0.19160266727983444,
+      "grad_norm": 0.875,
+      "learning_rate": 4.5688693943882446e-05,
+      "loss": 1.0143,
+      "step": 2604
+    },
+    {
+      "epoch": 0.19167624741319844,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.5685442014547563e-05,
+      "loss": 1.0611,
+      "step": 2605
+    },
+    {
+      "epoch": 0.19174982754656242,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.568218897506408e-05,
+      "loss": 0.9734,
+      "step": 2606
+    },
+    {
+      "epoch": 0.19182340767992642,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.567893482560657e-05,
+      "loss": 0.8655,
+      "step": 2607
+    },
+    {
+      "epoch": 0.1918969878132904,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.567567956634969e-05,
+      "loss": 1.0493,
+      "step": 2608
+    },
+    {
+      "epoch": 0.1919705679466544,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.567242319746814e-05,
+      "loss": 1.328,
+      "step": 2609
+    },
+    {
+      "epoch": 0.1920441480800184,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.5669165719136675e-05,
+      "loss": 1.341,
+      "step": 2610
+    },
+    {
+      "epoch": 0.19211772821338238,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.566590713153013e-05,
+      "loss": 0.9836,
+      "step": 2611
+    },
+    {
+      "epoch": 0.1921913083467464,
+      "grad_norm": 1.3203125,
+      "learning_rate": 4.566264743482338e-05,
+      "loss": 0.8758,
+      "step": 2612
+    },
+    {
+      "epoch": 0.19226488848011036,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.565938662919137e-05,
+      "loss": 0.8905,
+      "step": 2613
+    },
+    {
+      "epoch": 0.19233846861347437,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.5656124714809096e-05,
+      "loss": 0.7487,
+      "step": 2614
+    },
+    {
+      "epoch": 0.19241204874683834,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.5652861691851624e-05,
+      "loss": 0.6465,
+      "step": 2615
+    },
+    {
+      "epoch": 0.19248562888020235,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.564959756049407e-05,
+      "loss": 1.4017,
+      "step": 2616
+    },
+    {
+      "epoch": 0.19255920901356632,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.564633232091162e-05,
+      "loss": 0.9095,
+      "step": 2617
+    },
+    {
+      "epoch": 0.19263278914693033,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.564306597327951e-05,
+      "loss": 0.6735,
+      "step": 2618
+    },
+    {
+      "epoch": 0.19270636928029433,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.563979851777304e-05,
+      "loss": 1.0966,
+      "step": 2619
+    },
+    {
+      "epoch": 0.1927799494136583,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.563652995456756e-05,
+      "loss": 0.9686,
+      "step": 2620
+    },
+    {
+      "epoch": 0.1928535295470223,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.5633260283838504e-05,
+      "loss": 0.862,
+      "step": 2621
+    },
+    {
+      "epoch": 0.1929271096803863,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.5629989505761325e-05,
+      "loss": 1.1745,
+      "step": 2622
+    },
+    {
+      "epoch": 0.1930006898137503,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.562671762051159e-05,
+      "loss": 0.8492,
+      "step": 2623
+    },
+    {
+      "epoch": 0.19307426994711427,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.5623444628264864e-05,
+      "loss": 1.0222,
+      "step": 2624
+    },
+    {
+      "epoch": 0.19314785008047827,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.562017052919683e-05,
+      "loss": 1.0419,
+      "step": 2625
+    },
+    {
+      "epoch": 0.19322143021384225,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.5616895323483184e-05,
+      "loss": 1.004,
+      "step": 2626
+    },
+    {
+      "epoch": 0.19329501034720625,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.56136190112997e-05,
+      "loss": 0.8283,
+      "step": 2627
+    },
+    {
+      "epoch": 0.19336859048057026,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.5610341592822224e-05,
+      "loss": 0.984,
+      "step": 2628
+    },
+    {
+      "epoch": 0.19344217061393423,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.560706306822664e-05,
+      "loss": 1.3144,
+      "step": 2629
+    },
+    {
+      "epoch": 0.19351575074729824,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.5603783437688906e-05,
+      "loss": 1.0334,
+      "step": 2630
+    },
+    {
+      "epoch": 0.1935893308806622,
+      "grad_norm": 0.625,
+      "learning_rate": 4.5600502701385026e-05,
+      "loss": 0.6328,
+      "step": 2631
+    },
+    {
+      "epoch": 0.19366291101402622,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.559722085949107e-05,
+      "loss": 0.8362,
+      "step": 2632
+    },
+    {
+      "epoch": 0.1937364911473902,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.559393791218318e-05,
+      "loss": 0.7109,
+      "step": 2633
+    },
+    {
+      "epoch": 0.1938100712807542,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.5590653859637545e-05,
+      "loss": 0.7784,
+      "step": 2634
+    },
+    {
+      "epoch": 0.19388365141411817,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.5587368702030396e-05,
+      "loss": 1.3132,
+      "step": 2635
+    },
+    {
+      "epoch": 0.19395723154748218,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.5584082439538055e-05,
+      "loss": 0.6926,
+      "step": 2636
+    },
+    {
+      "epoch": 0.19403081168084618,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.5580795072336894e-05,
+      "loss": 0.7574,
+      "step": 2637
+    },
+    {
+      "epoch": 0.19410439181421016,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.5577506600603326e-05,
+      "loss": 1.2969,
+      "step": 2638
+    },
+    {
+      "epoch": 0.19417797194757416,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.5574217024513846e-05,
+      "loss": 0.6515,
+      "step": 2639
+    },
+    {
+      "epoch": 0.19425155208093814,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.5570926344245003e-05,
+      "loss": 0.9621,
+      "step": 2640
+    },
+    {
+      "epoch": 0.19432513221430214,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.5567634559973394e-05,
+      "loss": 0.7323,
+      "step": 2641
+    },
+    {
+      "epoch": 0.19439871234766612,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.5564341671875674e-05,
+      "loss": 1.1237,
+      "step": 2642
+    },
+    {
+      "epoch": 0.19447229248103012,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.556104768012859e-05,
+      "loss": 1.3252,
+      "step": 2643
+    },
+    {
+      "epoch": 0.1945458726143941,
+      "grad_norm": 1.125,
+      "learning_rate": 4.5557752584908906e-05,
+      "loss": 1.3102,
+      "step": 2644
+    },
+    {
+      "epoch": 0.1946194527477581,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.555445638639347e-05,
+      "loss": 0.8272,
+      "step": 2645
+    },
+    {
+      "epoch": 0.1946930328811221,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.555115908475918e-05,
+      "loss": 0.8254,
+      "step": 2646
+    },
+    {
+      "epoch": 0.19476661301448608,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.5547860680183e-05,
+      "loss": 0.9639,
+      "step": 2647
+    },
+    {
+      "epoch": 0.1948401931478501,
+      "grad_norm": 2.0,
+      "learning_rate": 4.5544561172841936e-05,
+      "loss": 0.7421,
+      "step": 2648
+    },
+    {
+      "epoch": 0.19491377328121406,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.5541260562913093e-05,
+      "loss": 0.7926,
+      "step": 2649
+    },
+    {
+      "epoch": 0.19498735341457807,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.553795885057358e-05,
+      "loss": 0.8785,
+      "step": 2650
+    },
+    {
+      "epoch": 0.19506093354794204,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.553465603600062e-05,
+      "loss": 1.2337,
+      "step": 2651
+    },
+    {
+      "epoch": 0.19513451368130605,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.553135211937144e-05,
+      "loss": 0.7467,
+      "step": 2652
+    },
+    {
+      "epoch": 0.19520809381467005,
+      "grad_norm": 2.625,
+      "learning_rate": 4.552804710086338e-05,
+      "loss": 1.179,
+      "step": 2653
+    },
+    {
+      "epoch": 0.19528167394803403,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.55247409806538e-05,
+      "loss": 0.5965,
+      "step": 2654
+    },
+    {
+      "epoch": 0.19535525408139803,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.552143375892014e-05,
+      "loss": 1.0021,
+      "step": 2655
+    },
+    {
+      "epoch": 0.195428834214762,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.5518125435839884e-05,
+      "loss": 0.9996,
+      "step": 2656
+    },
+    {
+      "epoch": 0.195502414348126,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.551481601159059e-05,
+      "loss": 0.8108,
+      "step": 2657
+    },
+    {
+      "epoch": 0.19557599448149,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.551150548634987e-05,
+      "loss": 1.0242,
+      "step": 2658
+    },
+    {
+      "epoch": 0.195649574614854,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.5508193860295396e-05,
+      "loss": 0.9833,
+      "step": 2659
+    },
+    {
+      "epoch": 0.19572315474821797,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.5504881133604885e-05,
+      "loss": 1.6644,
+      "step": 2660
+    },
+    {
+      "epoch": 0.19579673488158197,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.550156730645614e-05,
+      "loss": 1.1838,
+      "step": 2661
+    },
+    {
+      "epoch": 0.19587031501494598,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.549825237902699e-05,
+      "loss": 1.1663,
+      "step": 2662
+    },
+    {
+      "epoch": 0.19594389514830995,
+      "grad_norm": 1.0,
+      "learning_rate": 4.549493635149535e-05,
+      "loss": 1.432,
+      "step": 2663
+    },
+    {
+      "epoch": 0.19601747528167396,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.549161922403919e-05,
+      "loss": 1.0644,
+      "step": 2664
+    },
+    {
+      "epoch": 0.19609105541503793,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.5488300996836534e-05,
+      "loss": 0.8674,
+      "step": 2665
+    },
+    {
+      "epoch": 0.19616463554840194,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.5484981670065454e-05,
+      "loss": 1.1175,
+      "step": 2666
+    },
+    {
+      "epoch": 0.19623821568176592,
+      "grad_norm": 1.3984375,
+      "learning_rate": 4.548166124390411e-05,
+      "loss": 0.6985,
+      "step": 2667
+    },
+    {
+      "epoch": 0.19631179581512992,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.547833971853067e-05,
+      "loss": 1.034,
+      "step": 2668
+    },
+    {
+      "epoch": 0.1963853759484939,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.547501709412343e-05,
+      "loss": 0.8328,
+      "step": 2669
+    },
+    {
+      "epoch": 0.1964589560818579,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.547169337086069e-05,
+      "loss": 1.0259,
+      "step": 2670
+    },
+    {
+      "epoch": 0.1965325362152219,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.546836854892084e-05,
+      "loss": 0.8838,
+      "step": 2671
+    },
+    {
+      "epoch": 0.19660611634858588,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.54650426284823e-05,
+      "loss": 0.8251,
+      "step": 2672
+    },
+    {
+      "epoch": 0.19667969648194988,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.5461715609723575e-05,
+      "loss": 0.7196,
+      "step": 2673
+    },
+    {
+      "epoch": 0.19675327661531386,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.5458387492823226e-05,
+      "loss": 1.1062,
+      "step": 2674
+    },
+    {
+      "epoch": 0.19682685674867786,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.545505827795986e-05,
+      "loss": 0.8082,
+      "step": 2675
+    },
+    {
+      "epoch": 0.19690043688204184,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.545172796531214e-05,
+      "loss": 0.6577,
+      "step": 2676
+    },
+    {
+      "epoch": 0.19697401701540584,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.544839655505882e-05,
+      "loss": 0.8486,
+      "step": 2677
+    },
+    {
+      "epoch": 0.19704759714876982,
+      "grad_norm": 1.328125,
+      "learning_rate": 4.544506404737868e-05,
+      "loss": 1.3335,
+      "step": 2678
+    },
+    {
+      "epoch": 0.19712117728213382,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.544173044245056e-05,
+      "loss": 1.4339,
+      "step": 2679
+    },
+    {
+      "epoch": 0.19719475741549783,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.5438395740453375e-05,
+      "loss": 0.8955,
+      "step": 2680
+    },
+    {
+      "epoch": 0.1972683375488618,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.54350599415661e-05,
+      "loss": 0.7659,
+      "step": 2681
+    },
+    {
+      "epoch": 0.1973419176822258,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.543172304596774e-05,
+      "loss": 0.847,
+      "step": 2682
+    },
+    {
+      "epoch": 0.19741549781558979,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.542838505383741e-05,
+      "loss": 1.1848,
+      "step": 2683
+    },
+    {
+      "epoch": 0.1974890779489538,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.542504596535424e-05,
+      "loss": 0.8251,
+      "step": 2684
+    },
+    {
+      "epoch": 0.19756265808231777,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.542170578069742e-05,
+      "loss": 0.772,
+      "step": 2685
+    },
+    {
+      "epoch": 0.19763623821568177,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.5418364500046226e-05,
+      "loss": 0.7522,
+      "step": 2686
+    },
+    {
+      "epoch": 0.19770981834904575,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.541502212357998e-05,
+      "loss": 0.8216,
+      "step": 2687
+    },
+    {
+      "epoch": 0.19778339848240975,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.541167865147804e-05,
+      "loss": 0.805,
+      "step": 2688
+    },
+    {
+      "epoch": 0.19785697861577375,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.540833408391987e-05,
+      "loss": 0.8098,
+      "step": 2689
+    },
+    {
+      "epoch": 0.19793055874913773,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.540498842108495e-05,
+      "loss": 0.7311,
+      "step": 2690
+    },
+    {
+      "epoch": 0.19800413888250173,
+      "grad_norm": 1.4453125,
+      "learning_rate": 4.540164166315284e-05,
+      "loss": 0.8703,
+      "step": 2691
+    },
+    {
+      "epoch": 0.1980777190158657,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.539829381030316e-05,
+      "loss": 1.0887,
+      "step": 2692
+    },
+    {
+      "epoch": 0.19815129914922971,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.539494486271558e-05,
+      "loss": 0.9607,
+      "step": 2693
+    },
+    {
+      "epoch": 0.1982248792825937,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.5391594820569827e-05,
+      "loss": 0.8763,
+      "step": 2694
+    },
+    {
+      "epoch": 0.1982984594159577,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.538824368404569e-05,
+      "loss": 1.3082,
+      "step": 2695
+    },
+    {
+      "epoch": 0.19837203954932167,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.538489145332303e-05,
+      "loss": 0.8263,
+      "step": 2696
+    },
+    {
+      "epoch": 0.19844561968268568,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.5381538128581746e-05,
+      "loss": 0.9347,
+      "step": 2697
+    },
+    {
+      "epoch": 0.19851919981604968,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.5378183710001806e-05,
+      "loss": 0.7509,
+      "step": 2698
+    },
+    {
+      "epoch": 0.19859277994941366,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.537482819776323e-05,
+      "loss": 1.0527,
+      "step": 2699
+    },
+    {
+      "epoch": 0.19866636008277766,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.537147159204611e-05,
+      "loss": 0.7255,
+      "step": 2700
+    },
+    {
+      "epoch": 0.19873994021614164,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.536811389303058e-05,
+      "loss": 0.7585,
+      "step": 2701
+    },
+    {
+      "epoch": 0.19881352034950564,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.5364755100896864e-05,
+      "loss": 0.7841,
+      "step": 2702
+    },
+    {
+      "epoch": 0.19888710048286962,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.536139521582519e-05,
+      "loss": 0.8066,
+      "step": 2703
+    },
+    {
+      "epoch": 0.19896068061623362,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.53580342379959e-05,
+      "loss": 1.1482,
+      "step": 2704
+    },
+    {
+      "epoch": 0.1990342607495976,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.535467216758936e-05,
+      "loss": 0.6796,
+      "step": 2705
+    },
+    {
+      "epoch": 0.1991078408829616,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.5351309004786e-05,
+      "loss": 0.7754,
+      "step": 2706
+    },
+    {
+      "epoch": 0.1991814210163256,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.534794474976634e-05,
+      "loss": 1.2938,
+      "step": 2707
+    },
+    {
+      "epoch": 0.19925500114968958,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.534457940271091e-05,
+      "loss": 1.0347,
+      "step": 2708
+    },
+    {
+      "epoch": 0.19932858128305359,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.534121296380033e-05,
+      "loss": 0.6704,
+      "step": 2709
+    },
+    {
+      "epoch": 0.19940216141641756,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.5337845433215266e-05,
+      "loss": 0.8379,
+      "step": 2710
+    },
+    {
+      "epoch": 0.19947574154978157,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.5334476811136455e-05,
+      "loss": 0.8135,
+      "step": 2711
+    },
+    {
+      "epoch": 0.19954932168314554,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.533110709774468e-05,
+      "loss": 1.1644,
+      "step": 2712
+    },
+    {
+      "epoch": 0.19962290181650955,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.532773629322078e-05,
+      "loss": 0.8303,
+      "step": 2713
+    },
+    {
+      "epoch": 0.19969648194987352,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.532436439774567e-05,
+      "loss": 0.8286,
+      "step": 2714
+    },
+    {
+      "epoch": 0.19977006208323753,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.532099141150031e-05,
+      "loss": 0.8525,
+      "step": 2715
+    },
+    {
+      "epoch": 0.19984364221660153,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.531761733466573e-05,
+      "loss": 0.8361,
+      "step": 2716
+    },
+    {
+      "epoch": 0.1999172223499655,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.5314242167422996e-05,
+      "loss": 0.7961,
+      "step": 2717
+    },
+    {
+      "epoch": 0.1999908024833295,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.531086590995325e-05,
+      "loss": 1.0901,
+      "step": 2718
+    },
+    {
+      "epoch": 0.2000643826166935,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.53074885624377e-05,
+      "loss": 0.9963,
+      "step": 2719
+    },
+    {
+      "epoch": 0.2001379627500575,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.5304110125057584e-05,
+      "loss": 1.4383,
+      "step": 2720
+    },
+    {
+      "epoch": 0.20021154288342147,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.5300730597994236e-05,
+      "loss": 0.8235,
+      "step": 2721
+    },
+    {
+      "epoch": 0.20028512301678547,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.5297349981429014e-05,
+      "loss": 0.6616,
+      "step": 2722
+    },
+    {
+      "epoch": 0.20035870315014945,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.529396827554335e-05,
+      "loss": 1.1743,
+      "step": 2723
+    },
+    {
+      "epoch": 0.20043228328351345,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.529058548051875e-05,
+      "loss": 1.0245,
+      "step": 2724
+    },
+    {
+      "epoch": 0.20050586341687746,
+      "grad_norm": 1.28125,
+      "learning_rate": 4.528720159653674e-05,
+      "loss": 1.3024,
+      "step": 2725
+    },
+    {
+      "epoch": 0.20057944355024143,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.528381662377895e-05,
+      "loss": 0.6621,
+      "step": 2726
+    },
+    {
+      "epoch": 0.20065302368360544,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.528043056242702e-05,
+      "loss": 1.2441,
+      "step": 2727
+    },
+    {
+      "epoch": 0.2007266038169694,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.527704341266269e-05,
+      "loss": 1.2101,
+      "step": 2728
+    },
+    {
+      "epoch": 0.20080018395033342,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.5273655174667745e-05,
+      "loss": 0.9185,
+      "step": 2729
+    },
+    {
+      "epoch": 0.2008737640836974,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.5270265848624004e-05,
+      "loss": 0.8128,
+      "step": 2730
+    },
+    {
+      "epoch": 0.2009473442170614,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.52668754347134e-05,
+      "loss": 1.1705,
+      "step": 2731
+    },
+    {
+      "epoch": 0.20102092435042537,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.526348393311786e-05,
+      "loss": 0.8057,
+      "step": 2732
+    },
+    {
+      "epoch": 0.20109450448378938,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.52600913440194e-05,
+      "loss": 1.1357,
+      "step": 2733
+    },
+    {
+      "epoch": 0.20116808461715338,
+      "grad_norm": 0.75,
+      "learning_rate": 4.525669766760011e-05,
+      "loss": 0.6603,
+      "step": 2734
+    },
+    {
+      "epoch": 0.20124166475051736,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.525330290404212e-05,
+      "loss": 0.6845,
+      "step": 2735
+    },
+    {
+      "epoch": 0.20131524488388136,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.524990705352761e-05,
+      "loss": 1.2872,
+      "step": 2736
+    },
+    {
+      "epoch": 0.20138882501724534,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.524651011623884e-05,
+      "loss": 0.9801,
+      "step": 2737
+    },
+    {
+      "epoch": 0.20146240515060934,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.524311209235811e-05,
+      "loss": 0.8503,
+      "step": 2738
+    },
+    {
+      "epoch": 0.20153598528397332,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.523971298206779e-05,
+      "loss": 0.925,
+      "step": 2739
+    },
+    {
+      "epoch": 0.20160956541733732,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.52363127855503e-05,
+      "loss": 1.5058,
+      "step": 2740
+    },
+    {
+      "epoch": 0.2016831455507013,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.523291150298813e-05,
+      "loss": 0.7355,
+      "step": 2741
+    },
+    {
+      "epoch": 0.2017567256840653,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.522950913456381e-05,
+      "loss": 0.9563,
+      "step": 2742
+    },
+    {
+      "epoch": 0.2018303058174293,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.522610568045994e-05,
+      "loss": 0.7932,
+      "step": 2743
+    },
+    {
+      "epoch": 0.20190388595079328,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.522270114085917e-05,
+      "loss": 0.8767,
+      "step": 2744
+    },
+    {
+      "epoch": 0.2019774660841573,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.5219295515944244e-05,
+      "loss": 0.9709,
+      "step": 2745
+    },
+    {
+      "epoch": 0.20205104621752126,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.5215888805897906e-05,
+      "loss": 0.8323,
+      "step": 2746
+    },
+    {
+      "epoch": 0.20212462635088527,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.5212481010903e-05,
+      "loss": 0.9393,
+      "step": 2747
+    },
+    {
+      "epoch": 0.20219820648424924,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.5209072131142415e-05,
+      "loss": 1.0489,
+      "step": 2748
+    },
+    {
+      "epoch": 0.20227178661761325,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.52056621667991e-05,
+      "loss": 0.7525,
+      "step": 2749
+    },
+    {
+      "epoch": 0.20234536675097722,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.5202251118056056e-05,
+      "loss": 1.0862,
+      "step": 2750
+    },
+    {
+      "epoch": 0.20241894688434123,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.519883898509635e-05,
+      "loss": 1.1756,
+      "step": 2751
+    },
+    {
+      "epoch": 0.20249252701770523,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.519542576810311e-05,
+      "loss": 0.9582,
+      "step": 2752
+    },
+    {
+      "epoch": 0.2025661071510692,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.519201146725951e-05,
+      "loss": 1.0155,
+      "step": 2753
+    },
+    {
+      "epoch": 0.2026396872844332,
+      "grad_norm": 1.3046875,
+      "learning_rate": 4.518859608274879e-05,
+      "loss": 1.1281,
+      "step": 2754
+    },
+    {
+      "epoch": 0.2027132674177972,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.518517961475426e-05,
+      "loss": 1.1438,
+      "step": 2755
+    },
+    {
+      "epoch": 0.2027868475511612,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.518176206345925e-05,
+      "loss": 1.2908,
+      "step": 2756
+    },
+    {
+      "epoch": 0.20286042768452517,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.51783434290472e-05,
+      "loss": 0.9362,
+      "step": 2757
+    },
+    {
+      "epoch": 0.20293400781788917,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.517492371170156e-05,
+      "loss": 1.2026,
+      "step": 2758
+    },
+    {
+      "epoch": 0.20300758795125315,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.517150291160588e-05,
+      "loss": 0.8406,
+      "step": 2759
+    },
+    {
+      "epoch": 0.20308116808461715,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.5168081028943726e-05,
+      "loss": 0.7653,
+      "step": 2760
+    },
+    {
+      "epoch": 0.20315474821798116,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.516465806389876e-05,
+      "loss": 0.8063,
+      "step": 2761
+    },
+    {
+      "epoch": 0.20322832835134513,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.5161234016654684e-05,
+      "loss": 0.928,
+      "step": 2762
+    },
+    {
+      "epoch": 0.20330190848470914,
+      "grad_norm": 1.0,
+      "learning_rate": 4.515780888739525e-05,
+      "loss": 0.7376,
+      "step": 2763
+    },
+    {
+      "epoch": 0.2033754886180731,
+      "grad_norm": 1.3125,
+      "learning_rate": 4.5154382676304295e-05,
+      "loss": 1.3377,
+      "step": 2764
+    },
+    {
+      "epoch": 0.20344906875143712,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.515095538356568e-05,
+      "loss": 1.0057,
+      "step": 2765
+    },
+    {
+      "epoch": 0.2035226488848011,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.5147527009363354e-05,
+      "loss": 0.9021,
+      "step": 2766
+    },
+    {
+      "epoch": 0.2035962290181651,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.51440975538813e-05,
+      "loss": 0.7933,
+      "step": 2767
+    },
+    {
+      "epoch": 0.20366980915152907,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.51406670173036e-05,
+      "loss": 0.995,
+      "step": 2768
+    },
+    {
+      "epoch": 0.20374338928489308,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.513723539981432e-05,
+      "loss": 1.0368,
+      "step": 2769
+    },
+    {
+      "epoch": 0.20381696941825708,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.513380270159765e-05,
+      "loss": 0.7137,
+      "step": 2770
+    },
+    {
+      "epoch": 0.20389054955162106,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.513036892283782e-05,
+      "loss": 0.912,
+      "step": 2771
+    },
+    {
+      "epoch": 0.20396412968498506,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.5126934063719113e-05,
+      "loss": 1.8423,
+      "step": 2772
+    },
+    {
+      "epoch": 0.20403770981834904,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.5123498124425864e-05,
+      "loss": 0.9329,
+      "step": 2773
+    },
+    {
+      "epoch": 0.20411128995171304,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.5120061105142486e-05,
+      "loss": 0.9621,
+      "step": 2774
+    },
+    {
+      "epoch": 0.20418487008507702,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.511662300605343e-05,
+      "loss": 1.0549,
+      "step": 2775
+    },
+    {
+      "epoch": 0.20425845021844102,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.5113183827343206e-05,
+      "loss": 1.184,
+      "step": 2776
+    },
+    {
+      "epoch": 0.204332030351805,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.510974356919639e-05,
+      "loss": 1.4337,
+      "step": 2777
+    },
+    {
+      "epoch": 0.204405610485169,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.5106302231797624e-05,
+      "loss": 1.0132,
+      "step": 2778
+    },
+    {
+      "epoch": 0.204479190618533,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.51028598153316e-05,
+      "loss": 0.9231,
+      "step": 2779
+    },
+    {
+      "epoch": 0.20455277075189698,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.509941631998305e-05,
+      "loss": 0.8495,
+      "step": 2780
+    },
+    {
+      "epoch": 0.204626350885261,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.509597174593679e-05,
+      "loss": 0.8408,
+      "step": 2781
+    },
+    {
+      "epoch": 0.20469993101862496,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.5092526093377685e-05,
+      "loss": 0.7082,
+      "step": 2782
+    },
+    {
+      "epoch": 0.20477351115198897,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.508907936249065e-05,
+      "loss": 1.2779,
+      "step": 2783
+    },
+    {
+      "epoch": 0.20484709128535294,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.508563155346067e-05,
+      "loss": 0.6749,
+      "step": 2784
+    },
+    {
+      "epoch": 0.20492067141871695,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.508218266647278e-05,
+      "loss": 1.1346,
+      "step": 2785
+    },
+    {
+      "epoch": 0.20499425155208092,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.507873270171208e-05,
+      "loss": 1.1635,
+      "step": 2786
+    },
+    {
+      "epoch": 0.20506783168544493,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.507528165936372e-05,
+      "loss": 1.3338,
+      "step": 2787
+    },
+    {
+      "epoch": 0.20514141181880893,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.50718295396129e-05,
+      "loss": 1.0394,
+      "step": 2788
+    },
+    {
+      "epoch": 0.2052149919521729,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.506837634264492e-05,
+      "loss": 1.0295,
+      "step": 2789
+    },
+    {
+      "epoch": 0.2052885720855369,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.5064922068645064e-05,
+      "loss": 1.2822,
+      "step": 2790
+    },
+    {
+      "epoch": 0.2053621522189009,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.506146671779874e-05,
+      "loss": 0.6555,
+      "step": 2791
+    },
+    {
+      "epoch": 0.2054357323522649,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.50580102902914e-05,
+      "loss": 0.721,
+      "step": 2792
+    },
+    {
+      "epoch": 0.20550931248562887,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.505455278630852e-05,
+      "loss": 0.8942,
+      "step": 2793
+    },
+    {
+      "epoch": 0.20558289261899287,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.505109420603568e-05,
+      "loss": 0.8244,
+      "step": 2794
+    },
+    {
+      "epoch": 0.20565647275235685,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.5047634549658475e-05,
+      "loss": 0.8666,
+      "step": 2795
+    },
+    {
+      "epoch": 0.20573005288572085,
+      "grad_norm": 2.109375,
+      "learning_rate": 4.5044173817362594e-05,
+      "loss": 1.1106,
+      "step": 2796
+    },
+    {
+      "epoch": 0.20580363301908486,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.5040712009333764e-05,
+      "loss": 0.9823,
+      "step": 2797
+    },
+    {
+      "epoch": 0.20587721315244883,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.5037249125757766e-05,
+      "loss": 1.3864,
+      "step": 2798
+    },
+    {
+      "epoch": 0.20595079328581284,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.503378516682046e-05,
+      "loss": 1.0485,
+      "step": 2799
+    },
+    {
+      "epoch": 0.20602437341917682,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.503032013270774e-05,
+      "loss": 0.8076,
+      "step": 2800
+    },
+    {
+      "epoch": 0.20609795355254082,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.502685402360556e-05,
+      "loss": 0.8173,
+      "step": 2801
+    },
+    {
+      "epoch": 0.2061715336859048,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.502338683969997e-05,
+      "loss": 1.0205,
+      "step": 2802
+    },
+    {
+      "epoch": 0.2062451138192688,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.5019918581177015e-05,
+      "loss": 0.8801,
+      "step": 2803
+    },
+    {
+      "epoch": 0.20631869395263278,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.5016449248222835e-05,
+      "loss": 0.8997,
+      "step": 2804
+    },
+    {
+      "epoch": 0.20639227408599678,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.501297884102363e-05,
+      "loss": 0.9744,
+      "step": 2805
+    },
+    {
+      "epoch": 0.20646585421936078,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.5009507359765666e-05,
+      "loss": 0.967,
+      "step": 2806
+    },
+    {
+      "epoch": 0.20653943435272476,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.500603480463523e-05,
+      "loss": 0.8408,
+      "step": 2807
+    },
+    {
+      "epoch": 0.20661301448608876,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.500256117581868e-05,
+      "loss": 1.3726,
+      "step": 2808
+    },
+    {
+      "epoch": 0.20668659461945274,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.499908647350246e-05,
+      "loss": 1.1918,
+      "step": 2809
+    },
+    {
+      "epoch": 0.20676017475281674,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.499561069787305e-05,
+      "loss": 1.2052,
+      "step": 2810
+    },
+    {
+      "epoch": 0.20683375488618072,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.499213384911696e-05,
+      "loss": 0.7237,
+      "step": 2811
+    },
+    {
+      "epoch": 0.20690733501954472,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.498865592742082e-05,
+      "loss": 0.76,
+      "step": 2812
+    },
+    {
+      "epoch": 0.2069809151529087,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.498517693297127e-05,
+      "loss": 0.9004,
+      "step": 2813
+    },
+    {
+      "epoch": 0.2070544952862727,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.498169686595501e-05,
+      "loss": 0.9867,
+      "step": 2814
+    },
+    {
+      "epoch": 0.2071280754196367,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.497821572655883e-05,
+      "loss": 0.9329,
+      "step": 2815
+    },
+    {
+      "epoch": 0.20720165555300069,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.497473351496955e-05,
+      "loss": 0.8902,
+      "step": 2816
+    },
+    {
+      "epoch": 0.2072752356863647,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.497125023137403e-05,
+      "loss": 0.6786,
+      "step": 2817
+    },
+    {
+      "epoch": 0.20734881581972867,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.496776587595924e-05,
+      "loss": 0.7063,
+      "step": 2818
+    },
+    {
+      "epoch": 0.20742239595309267,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.496428044891218e-05,
+      "loss": 1.05,
+      "step": 2819
+    },
+    {
+      "epoch": 0.20749597608645665,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.4960793950419884e-05,
+      "loss": 1.2781,
+      "step": 2820
+    },
+    {
+      "epoch": 0.20756955621982065,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.4957306380669475e-05,
+      "loss": 0.8069,
+      "step": 2821
+    },
+    {
+      "epoch": 0.20764313635318463,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.4953817739848134e-05,
+      "loss": 0.6579,
+      "step": 2822
+    },
+    {
+      "epoch": 0.20771671648654863,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.495032802814308e-05,
+      "loss": 1.2859,
+      "step": 2823
+    },
+    {
+      "epoch": 0.20779029661991263,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.49468372457416e-05,
+      "loss": 1.3279,
+      "step": 2824
+    },
+    {
+      "epoch": 0.2078638767532766,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.494334539283104e-05,
+      "loss": 0.6983,
+      "step": 2825
+    },
+    {
+      "epoch": 0.20793745688664061,
+      "grad_norm": 1.21875,
+      "learning_rate": 4.49398524695988e-05,
+      "loss": 0.9231,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2080110370200046,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.4936358476232346e-05,
+      "loss": 0.8578,
+      "step": 2827
+    },
+    {
+      "epoch": 0.2080846171533686,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.493286341291918e-05,
+      "loss": 1.246,
+      "step": 2828
+    },
+    {
+      "epoch": 0.20815819728673257,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.492936727984688e-05,
+      "loss": 0.9413,
+      "step": 2829
+    },
+    {
+      "epoch": 0.20823177742009658,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.492587007720308e-05,
+      "loss": 0.9834,
+      "step": 2830
+    },
+    {
+      "epoch": 0.20830535755346058,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.4922371805175475e-05,
+      "loss": 1.1971,
+      "step": 2831
+    },
+    {
+      "epoch": 0.20837893768682456,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.491887246395179e-05,
+      "loss": 1.4261,
+      "step": 2832
+    },
+    {
+      "epoch": 0.20845251782018856,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.4915372053719856e-05,
+      "loss": 1.1543,
+      "step": 2833
+    },
+    {
+      "epoch": 0.20852609795355254,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.4911870574667515e-05,
+      "loss": 0.7451,
+      "step": 2834
+    },
+    {
+      "epoch": 0.20859967808691654,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.4908368026982686e-05,
+      "loss": 0.7919,
+      "step": 2835
+    },
+    {
+      "epoch": 0.20867325822028052,
+      "grad_norm": 1.28125,
+      "learning_rate": 4.4904864410853344e-05,
+      "loss": 1.1645,
+      "step": 2836
+    },
+    {
+      "epoch": 0.20874683835364452,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.490135972646752e-05,
+      "loss": 1.2171,
+      "step": 2837
+    },
+    {
+      "epoch": 0.2088204184870085,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.489785397401332e-05,
+      "loss": 0.8273,
+      "step": 2838
+    },
+    {
+      "epoch": 0.2088939986203725,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.489434715367887e-05,
+      "loss": 1.4749,
+      "step": 2839
+    },
+    {
+      "epoch": 0.2089675787537365,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.489083926565238e-05,
+      "loss": 1.0795,
+      "step": 2840
+    },
+    {
+      "epoch": 0.20904115888710048,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.488733031012213e-05,
+      "loss": 0.9256,
+      "step": 2841
+    },
+    {
+      "epoch": 0.20911473902046449,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.4883820287276415e-05,
+      "loss": 0.688,
+      "step": 2842
+    },
+    {
+      "epoch": 0.20918831915382846,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.4880309197303615e-05,
+      "loss": 0.8873,
+      "step": 2843
+    },
+    {
+      "epoch": 0.20926189928719247,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.4876797040392185e-05,
+      "loss": 0.98,
+      "step": 2844
+    },
+    {
+      "epoch": 0.20933547942055644,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.4873283816730584e-05,
+      "loss": 0.9738,
+      "step": 2845
+    },
+    {
+      "epoch": 0.20940905955392045,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.4869769526507376e-05,
+      "loss": 1.3381,
+      "step": 2846
+    },
+    {
+      "epoch": 0.20948263968728442,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.486625416991118e-05,
+      "loss": 1.1353,
+      "step": 2847
+    },
+    {
+      "epoch": 0.20955621982064843,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.486273774713064e-05,
+      "loss": 1.043,
+      "step": 2848
+    },
+    {
+      "epoch": 0.20962979995401243,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.4859220258354475e-05,
+      "loss": 0.7113,
+      "step": 2849
+    },
+    {
+      "epoch": 0.2097033800873764,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.485570170377146e-05,
+      "loss": 0.8909,
+      "step": 2850
+    },
+    {
+      "epoch": 0.2097769602207404,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.485218208357045e-05,
+      "loss": 0.8449,
+      "step": 2851
+    },
+    {
+      "epoch": 0.2098505403541044,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.484866139794032e-05,
+      "loss": 1.045,
+      "step": 2852
+    },
+    {
+      "epoch": 0.2099241204874684,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.484513964707002e-05,
+      "loss": 0.9923,
+      "step": 2853
+    },
+    {
+      "epoch": 0.20999770062083237,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.484161683114856e-05,
+      "loss": 1.2173,
+      "step": 2854
+    },
+    {
+      "epoch": 0.21007128075419637,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.4838092950364995e-05,
+      "loss": 0.6753,
+      "step": 2855
+    },
+    {
+      "epoch": 0.21014486088756035,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.483456800490845e-05,
+      "loss": 0.8739,
+      "step": 2856
+    },
+    {
+      "epoch": 0.21021844102092435,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.483104199496811e-05,
+      "loss": 0.8887,
+      "step": 2857
+    },
+    {
+      "epoch": 0.21029202115428836,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.482751492073319e-05,
+      "loss": 0.9067,
+      "step": 2858
+    },
+    {
+      "epoch": 0.21036560128765233,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.4823986782393e-05,
+      "loss": 1.1571,
+      "step": 2859
+    },
+    {
+      "epoch": 0.21043918142101634,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.482045758013689e-05,
+      "loss": 1.4446,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2105127615543803,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.481692731415424e-05,
+      "loss": 1.0334,
+      "step": 2861
+    },
+    {
+      "epoch": 0.21058634168774432,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.481339598463454e-05,
+      "loss": 0.8166,
+      "step": 2862
+    },
+    {
+      "epoch": 0.2106599218211083,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.48098635917673e-05,
+      "loss": 0.9122,
+      "step": 2863
+    },
+    {
+      "epoch": 0.2107335019544723,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.4806330135742106e-05,
+      "loss": 1.5158,
+      "step": 2864
+    },
+    {
+      "epoch": 0.21080708208783627,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.480279561674856e-05,
+      "loss": 0.8485,
+      "step": 2865
+    },
+    {
+      "epoch": 0.21088066222120028,
+      "grad_norm": 0.875,
+      "learning_rate": 4.479926003497639e-05,
+      "loss": 0.7725,
+      "step": 2866
+    },
+    {
+      "epoch": 0.21095424235456428,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.479572339061533e-05,
+      "loss": 0.9324,
+      "step": 2867
+    },
+    {
+      "epoch": 0.21102782248792826,
+      "grad_norm": 0.875,
+      "learning_rate": 4.479218568385518e-05,
+      "loss": 1.1622,
+      "step": 2868
+    },
+    {
+      "epoch": 0.21110140262129226,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.47886469148858e-05,
+      "loss": 1.1396,
+      "step": 2869
+    },
+    {
+      "epoch": 0.21117498275465624,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.478510708389713e-05,
+      "loss": 0.841,
+      "step": 2870
+    },
+    {
+      "epoch": 0.21124856288802024,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.478156619107912e-05,
+      "loss": 0.8657,
+      "step": 2871
+    },
+    {
+      "epoch": 0.21132214302138422,
+      "grad_norm": 1.0,
+      "learning_rate": 4.477802423662182e-05,
+      "loss": 1.0363,
+      "step": 2872
+    },
+    {
+      "epoch": 0.21139572315474822,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.4774481220715317e-05,
+      "loss": 1.0469,
+      "step": 2873
+    },
+    {
+      "epoch": 0.2114693032881122,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.477093714354975e-05,
+      "loss": 0.857,
+      "step": 2874
+    },
+    {
+      "epoch": 0.2115428834214762,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.4767392005315334e-05,
+      "loss": 0.7133,
+      "step": 2875
+    },
+    {
+      "epoch": 0.2116164635548402,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.4763845806202323e-05,
+      "loss": 0.9766,
+      "step": 2876
+    },
+    {
+      "epoch": 0.21169004368820418,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.4760298546401026e-05,
+      "loss": 1.3768,
+      "step": 2877
+    },
+    {
+      "epoch": 0.2117636238215682,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.475675022610184e-05,
+      "loss": 0.6088,
+      "step": 2878
+    },
+    {
+      "epoch": 0.21183720395493216,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.475320084549518e-05,
+      "loss": 0.9634,
+      "step": 2879
+    },
+    {
+      "epoch": 0.21191078408829617,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.474965040477154e-05,
+      "loss": 0.8918,
+      "step": 2880
+    },
+    {
+      "epoch": 0.21198436422166014,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.4746098904121467e-05,
+      "loss": 1.0243,
+      "step": 2881
+    },
+    {
+      "epoch": 0.21205794435502415,
+      "grad_norm": 11.0625,
+      "learning_rate": 4.474254634373556e-05,
+      "loss": 1.5296,
+      "step": 2882
+    },
+    {
+      "epoch": 0.21213152448838812,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.473899272380447e-05,
+      "loss": 1.1332,
+      "step": 2883
+    },
+    {
+      "epoch": 0.21220510462175213,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.473543804451893e-05,
+      "loss": 1.2872,
+      "step": 2884
+    },
+    {
+      "epoch": 0.21227868475511613,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.4731882306069706e-05,
+      "loss": 0.9887,
+      "step": 2885
+    },
+    {
+      "epoch": 0.2123522648884801,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.472832550864763e-05,
+      "loss": 0.6461,
+      "step": 2886
+    },
+    {
+      "epoch": 0.2124258450218441,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.472476765244358e-05,
+      "loss": 0.9668,
+      "step": 2887
+    },
+    {
+      "epoch": 0.2124994251552081,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.47212087376485e-05,
+      "loss": 0.9317,
+      "step": 2888
+    },
+    {
+      "epoch": 0.2125730052885721,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.47176487644534e-05,
+      "loss": 0.6655,
+      "step": 2889
+    },
+    {
+      "epoch": 0.21264658542193607,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.471408773304934e-05,
+      "loss": 0.7572,
+      "step": 2890
+    },
+    {
+      "epoch": 0.21272016555530007,
+      "grad_norm": 0.75,
+      "learning_rate": 4.471052564362742e-05,
+      "loss": 0.7705,
+      "step": 2891
+    },
+    {
+      "epoch": 0.21279374568866405,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.470696249637881e-05,
+      "loss": 0.7093,
+      "step": 2892
+    },
+    {
+      "epoch": 0.21286732582202805,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.4703398291494745e-05,
+      "loss": 0.6706,
+      "step": 2893
+    },
+    {
+      "epoch": 0.21294090595539206,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.4699833029166514e-05,
+      "loss": 1.0228,
+      "step": 2894
+    },
+    {
+      "epoch": 0.21301448608875603,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.4696266709585454e-05,
+      "loss": 0.9742,
+      "step": 2895
+    },
+    {
+      "epoch": 0.21308806622212004,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.469269933294296e-05,
+      "loss": 0.7757,
+      "step": 2896
+    },
+    {
+      "epoch": 0.213161646355484,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.468913089943049e-05,
+      "loss": 0.9448,
+      "step": 2897
+    },
+    {
+      "epoch": 0.21323522648884802,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.468556140923954e-05,
+      "loss": 0.964,
+      "step": 2898
+    },
+    {
+      "epoch": 0.213308806622212,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.468199086256169e-05,
+      "loss": 0.8315,
+      "step": 2899
+    },
+    {
+      "epoch": 0.213382386755576,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.4678419259588576e-05,
+      "loss": 0.873,
+      "step": 2900
+    },
+    {
+      "epoch": 0.21345596688893997,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.467484660051187e-05,
+      "loss": 0.943,
+      "step": 2901
+    },
+    {
+      "epoch": 0.21352954702230398,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.4671272885523294e-05,
+      "loss": 0.8535,
+      "step": 2902
+    },
+    {
+      "epoch": 0.21360312715566798,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.466769811481466e-05,
+      "loss": 1.138,
+      "step": 2903
+    },
+    {
+      "epoch": 0.21367670728903196,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.466412228857782e-05,
+      "loss": 1.1713,
+      "step": 2904
+    },
+    {
+      "epoch": 0.21375028742239596,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.466054540700467e-05,
+      "loss": 0.9006,
+      "step": 2905
+    },
+    {
+      "epoch": 0.21382386755575994,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.465696747028719e-05,
+      "loss": 1.2365,
+      "step": 2906
+    },
+    {
+      "epoch": 0.21389744768912394,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.4653388478617385e-05,
+      "loss": 0.6855,
+      "step": 2907
+    },
+    {
+      "epoch": 0.21397102782248792,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.464980843218734e-05,
+      "loss": 0.9935,
+      "step": 2908
+    },
+    {
+      "epoch": 0.21404460795585192,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.46462273311892e-05,
+      "loss": 1.0645,
+      "step": 2909
+    },
+    {
+      "epoch": 0.2141181880892159,
+      "grad_norm": 0.75,
+      "learning_rate": 4.464264517581514e-05,
+      "loss": 0.875,
+      "step": 2910
+    },
+    {
+      "epoch": 0.2141917682225799,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.463906196625741e-05,
+      "loss": 1.2615,
+      "step": 2911
+    },
+    {
+      "epoch": 0.2142653483559439,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.463547770270832e-05,
+      "loss": 1.1156,
+      "step": 2912
+    },
+    {
+      "epoch": 0.21433892848930788,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.463189238536023e-05,
+      "loss": 0.681,
+      "step": 2913
+    },
+    {
+      "epoch": 0.2144125086226719,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.4628306014405554e-05,
+      "loss": 0.8803,
+      "step": 2914
+    },
+    {
+      "epoch": 0.21448608875603586,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.4624718590036763e-05,
+      "loss": 0.7558,
+      "step": 2915
+    },
+    {
+      "epoch": 0.21455966888939987,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.462113011244639e-05,
+      "loss": 0.7771,
+      "step": 2916
+    },
+    {
+      "epoch": 0.21463324902276384,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.461754058182703e-05,
+      "loss": 0.7916,
+      "step": 2917
+    },
+    {
+      "epoch": 0.21470682915612785,
+      "grad_norm": 0.671875,
+      "learning_rate": 4.461394999837131e-05,
+      "loss": 0.5213,
+      "step": 2918
+    },
+    {
+      "epoch": 0.21478040928949182,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.461035836227194e-05,
+      "loss": 1.3971,
+      "step": 2919
+    },
+    {
+      "epoch": 0.21485398942285583,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.4606765673721684e-05,
+      "loss": 0.6099,
+      "step": 2920
+    },
+    {
+      "epoch": 0.21492756955621983,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.460317193291335e-05,
+      "loss": 0.868,
+      "step": 2921
+    },
+    {
+      "epoch": 0.2150011496895838,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.459957714003979e-05,
+      "loss": 1.1265,
+      "step": 2922
+    },
+    {
+      "epoch": 0.2150747298229478,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.459598129529395e-05,
+      "loss": 0.8591,
+      "step": 2923
+    },
+    {
+      "epoch": 0.2151483099563118,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.4592384398868804e-05,
+      "loss": 1.0894,
+      "step": 2924
+    },
+    {
+      "epoch": 0.2152218900896758,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.4588786450957384e-05,
+      "loss": 1.0027,
+      "step": 2925
+    },
+    {
+      "epoch": 0.21529547022303977,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.458518745175281e-05,
+      "loss": 0.9333,
+      "step": 2926
+    },
+    {
+      "epoch": 0.21536905035640377,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.458158740144821e-05,
+      "loss": 1.0506,
+      "step": 2927
+    },
+    {
+      "epoch": 0.21544263048976775,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.45779863002368e-05,
+      "loss": 0.8345,
+      "step": 2928
+    },
+    {
+      "epoch": 0.21551621062313175,
+      "grad_norm": 1.125,
+      "learning_rate": 4.457438414831183e-05,
+      "loss": 1.1072,
+      "step": 2929
+    },
+    {
+      "epoch": 0.21558979075649576,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.4570780945866644e-05,
+      "loss": 0.861,
+      "step": 2930
+    },
+    {
+      "epoch": 0.21566337088985973,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.4567176693094606e-05,
+      "loss": 1.2429,
+      "step": 2931
+    },
+    {
+      "epoch": 0.21573695102322374,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.456357139018915e-05,
+      "loss": 0.7611,
+      "step": 2932
+    },
+    {
+      "epoch": 0.21581053115658771,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.4559965037343776e-05,
+      "loss": 0.9303,
+      "step": 2933
+    },
+    {
+      "epoch": 0.21588411128995172,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.455635763475202e-05,
+      "loss": 0.8904,
+      "step": 2934
+    },
+    {
+      "epoch": 0.2159576914233157,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.455274918260748e-05,
+      "loss": 0.9982,
+      "step": 2935
+    },
+    {
+      "epoch": 0.2160312715566797,
+      "grad_norm": 1.2890625,
+      "learning_rate": 4.4549139681103825e-05,
+      "loss": 1.307,
+      "step": 2936
+    },
+    {
+      "epoch": 0.21610485169004368,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.454552913043477e-05,
+      "loss": 0.918,
+      "step": 2937
+    },
+    {
+      "epoch": 0.21617843182340768,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.454191753079408e-05,
+      "loss": 0.8802,
+      "step": 2938
+    },
+    {
+      "epoch": 0.21625201195677168,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.4538304882375584e-05,
+      "loss": 1.3401,
+      "step": 2939
+    },
+    {
+      "epoch": 0.21632559209013566,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.453469118537317e-05,
+      "loss": 0.817,
+      "step": 2940
+    },
+    {
+      "epoch": 0.21639917222349966,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.453107643998077e-05,
+      "loss": 1.4482,
+      "step": 2941
+    },
+    {
+      "epoch": 0.21647275235686364,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.452746064639239e-05,
+      "loss": 0.9268,
+      "step": 2942
+    },
+    {
+      "epoch": 0.21654633249022764,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.452384380480208e-05,
+      "loss": 1.0061,
+      "step": 2943
+    },
+    {
+      "epoch": 0.21661991262359162,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.4520225915403945e-05,
+      "loss": 1.6814,
+      "step": 2944
+    },
+    {
+      "epoch": 0.21669349275695562,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.451660697839216e-05,
+      "loss": 0.5645,
+      "step": 2945
+    },
+    {
+      "epoch": 0.2167670728903196,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.4512986993960936e-05,
+      "loss": 1.4558,
+      "step": 2946
+    },
+    {
+      "epoch": 0.2168406530236836,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.450936596230456e-05,
+      "loss": 0.9441,
+      "step": 2947
+    },
+    {
+      "epoch": 0.2169142331570476,
+      "grad_norm": 1.0,
+      "learning_rate": 4.450574388361735e-05,
+      "loss": 1.1417,
+      "step": 2948
+    },
+    {
+      "epoch": 0.21698781329041159,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.450212075809371e-05,
+      "loss": 1.1416,
+      "step": 2949
+    },
+    {
+      "epoch": 0.2170613934237756,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.449849658592809e-05,
+      "loss": 1.0104,
+      "step": 2950
+    },
+    {
+      "epoch": 0.21713497355713957,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.4494871367314984e-05,
+      "loss": 1.3781,
+      "step": 2951
+    },
+    {
+      "epoch": 0.21720855369050357,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.4491245102448955e-05,
+      "loss": 0.6403,
+      "step": 2952
+    },
+    {
+      "epoch": 0.21728213382386755,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.4487617791524605e-05,
+      "loss": 1.4903,
+      "step": 2953
+    },
+    {
+      "epoch": 0.21735571395723155,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.4483989434736624e-05,
+      "loss": 0.6208,
+      "step": 2954
+    },
+    {
+      "epoch": 0.21742929409059553,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.448036003227972e-05,
+      "loss": 0.8564,
+      "step": 2955
+    },
+    {
+      "epoch": 0.21750287422395953,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.44767295843487e-05,
+      "loss": 0.7602,
+      "step": 2956
+    },
+    {
+      "epoch": 0.21757645435732353,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.4473098091138374e-05,
+      "loss": 0.9807,
+      "step": 2957
+    },
+    {
+      "epoch": 0.2176500344906875,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.446946555284366e-05,
+      "loss": 0.705,
+      "step": 2958
+    },
+    {
+      "epoch": 0.21772361462405151,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.446583196965951e-05,
+      "loss": 1.282,
+      "step": 2959
+    },
+    {
+      "epoch": 0.2177971947574155,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.446219734178092e-05,
+      "loss": 0.9672,
+      "step": 2960
+    },
+    {
+      "epoch": 0.2178707748907795,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.445856166940296e-05,
+      "loss": 0.8693,
+      "step": 2961
+    },
+    {
+      "epoch": 0.21794435502414347,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.445492495272073e-05,
+      "loss": 1.3718,
+      "step": 2962
+    },
+    {
+      "epoch": 0.21801793515750748,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.4451287191929436e-05,
+      "loss": 1.0069,
+      "step": 2963
+    },
+    {
+      "epoch": 0.21809151529087145,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.444764838722429e-05,
+      "loss": 1.0451,
+      "step": 2964
+    },
+    {
+      "epoch": 0.21816509542423546,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.4444008538800604e-05,
+      "loss": 1.0163,
+      "step": 2965
+    },
+    {
+      "epoch": 0.21823867555759946,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.444036764685368e-05,
+      "loss": 1.4415,
+      "step": 2966
+    },
+    {
+      "epoch": 0.21831225569096344,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.4436725711578965e-05,
+      "loss": 1.0115,
+      "step": 2967
+    },
+    {
+      "epoch": 0.21838583582432744,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.443308273317188e-05,
+      "loss": 0.8376,
+      "step": 2968
+    },
+    {
+      "epoch": 0.21845941595769142,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.442943871182795e-05,
+      "loss": 1.134,
+      "step": 2969
+    },
+    {
+      "epoch": 0.21853299609105542,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.442579364774274e-05,
+      "loss": 1.2769,
+      "step": 2970
+    },
+    {
+      "epoch": 0.2186065762244194,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.442214754111188e-05,
+      "loss": 1.0365,
+      "step": 2971
+    },
+    {
+      "epoch": 0.2186801563577834,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.441850039213104e-05,
+      "loss": 1.0726,
+      "step": 2972
+    },
+    {
+      "epoch": 0.21875373649114738,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.441485220099596e-05,
+      "loss": 1.5218,
+      "step": 2973
+    },
+    {
+      "epoch": 0.21882731662451138,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.441120296790243e-05,
+      "loss": 0.7332,
+      "step": 2974
+    },
+    {
+      "epoch": 0.21890089675787539,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.440755269304631e-05,
+      "loss": 0.9733,
+      "step": 2975
+    },
+    {
+      "epoch": 0.21897447689123936,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.440390137662348e-05,
+      "loss": 1.2675,
+      "step": 2976
+    },
+    {
+      "epoch": 0.21904805702460337,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.440024901882992e-05,
+      "loss": 1.324,
+      "step": 2977
+    },
+    {
+      "epoch": 0.21912163715796734,
+      "grad_norm": 1.578125,
+      "learning_rate": 4.439659561986164e-05,
+      "loss": 0.7713,
+      "step": 2978
+    },
+    {
+      "epoch": 0.21919521729133135,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.4392941179914696e-05,
+      "loss": 1.0236,
+      "step": 2979
+    },
+    {
+      "epoch": 0.21926879742469532,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.4389285699185235e-05,
+      "loss": 1.0041,
+      "step": 2980
+    },
+    {
+      "epoch": 0.21934237755805933,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.438562917786943e-05,
+      "loss": 0.9832,
+      "step": 2981
+    },
+    {
+      "epoch": 0.2194159576914233,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.438197161616352e-05,
+      "loss": 1.1787,
+      "step": 2982
+    },
+    {
+      "epoch": 0.2194895378247873,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.43783130142638e-05,
+      "loss": 0.8349,
+      "step": 2983
+    },
+    {
+      "epoch": 0.2195631179581513,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.437465337236662e-05,
+      "loss": 0.8361,
+      "step": 2984
+    },
+    {
+      "epoch": 0.2196366980915153,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.437099269066839e-05,
+      "loss": 0.7906,
+      "step": 2985
+    },
+    {
+      "epoch": 0.2197102782248793,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.436733096936557e-05,
+      "loss": 1.1253,
+      "step": 2986
+    },
+    {
+      "epoch": 0.21978385835824327,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.436366820865468e-05,
+      "loss": 0.9108,
+      "step": 2987
+    },
+    {
+      "epoch": 0.21985743849160727,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.436000440873228e-05,
+      "loss": 0.8993,
+      "step": 2988
+    },
+    {
+      "epoch": 0.21993101862497125,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.435633956979501e-05,
+      "loss": 0.9804,
+      "step": 2989
+    },
+    {
+      "epoch": 0.22000459875833525,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.4352673692039564e-05,
+      "loss": 0.8479,
+      "step": 2990
+    },
+    {
+      "epoch": 0.22007817889169923,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.4349006775662664e-05,
+      "loss": 1.5212,
+      "step": 2991
+    },
+    {
+      "epoch": 0.22015175902506323,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.434533882086112e-05,
+      "loss": 0.8776,
+      "step": 2992
+    },
+    {
+      "epoch": 0.22022533915842724,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.434166982783178e-05,
+      "loss": 0.7635,
+      "step": 2993
+    },
+    {
+      "epoch": 0.2202989192917912,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.433799979677155e-05,
+      "loss": 0.8248,
+      "step": 2994
+    },
+    {
+      "epoch": 0.22037249942515522,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.43343287278774e-05,
+      "loss": 0.987,
+      "step": 2995
+    },
+    {
+      "epoch": 0.2204460795585192,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.4330656621346336e-05,
+      "loss": 0.9917,
+      "step": 2996
+    },
+    {
+      "epoch": 0.2205196596918832,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.432698347737545e-05,
+      "loss": 0.8746,
+      "step": 2997
+    },
+    {
+      "epoch": 0.22059323982524717,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.432330929616185e-05,
+      "loss": 0.8718,
+      "step": 2998
+    },
+    {
+      "epoch": 0.22066681995861118,
+      "grad_norm": 1.0,
+      "learning_rate": 4.4319634077902746e-05,
+      "loss": 1.2857,
+      "step": 2999
+    },
+    {
+      "epoch": 0.22074040009197515,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.4315957822795374e-05,
+      "loss": 0.9954,
+      "step": 3000
+    },
+    {
+      "epoch": 0.22081398022533916,
+      "grad_norm": 0.61328125,
+      "learning_rate": 4.4312280531037025e-05,
+      "loss": 0.7495,
+      "step": 3001
+    },
+    {
+      "epoch": 0.22088756035870316,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.430860220282506e-05,
+      "loss": 0.7469,
+      "step": 3002
+    },
+    {
+      "epoch": 0.22096114049206714,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.430492283835688e-05,
+      "loss": 0.7839,
+      "step": 3003
+    },
+    {
+      "epoch": 0.22103472062543114,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.430124243782995e-05,
+      "loss": 1.5534,
+      "step": 3004
+    },
+    {
+      "epoch": 0.22110830075879512,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.429756100144179e-05,
+      "loss": 0.8677,
+      "step": 3005
+    },
+    {
+      "epoch": 0.22118188089215912,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.429387852938999e-05,
+      "loss": 0.7379,
+      "step": 3006
+    },
+    {
+      "epoch": 0.2212554610255231,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.4290195021872164e-05,
+      "loss": 0.9089,
+      "step": 3007
+    },
+    {
+      "epoch": 0.2213290411588871,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.428651047908601e-05,
+      "loss": 1.0036,
+      "step": 3008
+    },
+    {
+      "epoch": 0.22140262129225108,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.428282490122926e-05,
+      "loss": 0.7464,
+      "step": 3009
+    },
+    {
+      "epoch": 0.22147620142561508,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.4279138288499725e-05,
+      "loss": 0.7255,
+      "step": 3010
+    },
+    {
+      "epoch": 0.2215497815589791,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.427545064109524e-05,
+      "loss": 0.6455,
+      "step": 3011
+    },
+    {
+      "epoch": 0.22162336169234306,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.427176195921373e-05,
+      "loss": 0.6021,
+      "step": 3012
+    },
+    {
+      "epoch": 0.22169694182570707,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.426807224305316e-05,
+      "loss": 1.094,
+      "step": 3013
+    },
+    {
+      "epoch": 0.22177052195907104,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.426438149281154e-05,
+      "loss": 1.2464,
+      "step": 3014
+    },
+    {
+      "epoch": 0.22184410209243505,
+      "grad_norm": 0.875,
+      "learning_rate": 4.4260689708686945e-05,
+      "loss": 1.2478,
+      "step": 3015
+    },
+    {
+      "epoch": 0.22191768222579902,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.425699689087752e-05,
+      "loss": 0.8136,
+      "step": 3016
+    },
+    {
+      "epoch": 0.22199126235916303,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.4253303039581436e-05,
+      "loss": 1.1139,
+      "step": 3017
+    },
+    {
+      "epoch": 0.22206484249252703,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.424960815499695e-05,
+      "loss": 0.7272,
+      "step": 3018
+    },
+    {
+      "epoch": 0.222138422625891,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.424591223732235e-05,
+      "loss": 0.525,
+      "step": 3019
+    },
+    {
+      "epoch": 0.222212002759255,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.4242215286755987e-05,
+      "loss": 0.6138,
+      "step": 3020
+    },
+    {
+      "epoch": 0.222285582892619,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.4238517303496276e-05,
+      "loss": 0.9417,
+      "step": 3021
+    },
+    {
+      "epoch": 0.222359163025983,
+      "grad_norm": 1.2265625,
+      "learning_rate": 4.423481828774168e-05,
+      "loss": 1.2537,
+      "step": 3022
+    },
+    {
+      "epoch": 0.22243274315934697,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.4231118239690714e-05,
+      "loss": 0.7667,
+      "step": 3023
+    },
+    {
+      "epoch": 0.22250632329271097,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.422741715954195e-05,
+      "loss": 0.8733,
+      "step": 3024
+    },
+    {
+      "epoch": 0.22257990342607495,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.422371504749403e-05,
+      "loss": 0.9773,
+      "step": 3025
+    },
+    {
+      "epoch": 0.22265348355943895,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.422001190374563e-05,
+      "loss": 1.087,
+      "step": 3026
+    },
+    {
+      "epoch": 0.22272706369280296,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.421630772849549e-05,
+      "loss": 0.5611,
+      "step": 3027
+    },
+    {
+      "epoch": 0.22280064382616693,
+      "grad_norm": 0.875,
+      "learning_rate": 4.42126025219424e-05,
+      "loss": 1.0163,
+      "step": 3028
+    },
+    {
+      "epoch": 0.22287422395953094,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.4208896284285235e-05,
+      "loss": 1.0159,
+      "step": 3029
+    },
+    {
+      "epoch": 0.2229478040928949,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.420518901572288e-05,
+      "loss": 1.0498,
+      "step": 3030
+    },
+    {
+      "epoch": 0.22302138422625892,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.42014807164543e-05,
+      "loss": 1.3386,
+      "step": 3031
+    },
+    {
+      "epoch": 0.2230949643596229,
+      "grad_norm": 0.59765625,
+      "learning_rate": 4.419777138667853e-05,
+      "loss": 0.7489,
+      "step": 3032
+    },
+    {
+      "epoch": 0.2231685444929869,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.4194061026594616e-05,
+      "loss": 1.1325,
+      "step": 3033
+    },
+    {
+      "epoch": 0.22324212462635087,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.419034963640171e-05,
+      "loss": 1.0251,
+      "step": 3034
+    },
+    {
+      "epoch": 0.22331570475971488,
+      "grad_norm": 1.125,
+      "learning_rate": 4.418663721629897e-05,
+      "loss": 1.1174,
+      "step": 3035
+    },
+    {
+      "epoch": 0.22338928489307888,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.418292376648566e-05,
+      "loss": 0.9524,
+      "step": 3036
+    },
+    {
+      "epoch": 0.22346286502644286,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.417920928716106e-05,
+      "loss": 1.2242,
+      "step": 3037
+    },
+    {
+      "epoch": 0.22353644515980686,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.417549377852452e-05,
+      "loss": 0.919,
+      "step": 3038
+    },
+    {
+      "epoch": 0.22361002529317084,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.4171777240775444e-05,
+      "loss": 1.1967,
+      "step": 3039
+    },
+    {
+      "epoch": 0.22368360542653484,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.41680596741133e-05,
+      "loss": 0.9151,
+      "step": 3040
+    },
+    {
+      "epoch": 0.22375718555989882,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.41643410787376e-05,
+      "loss": 0.7744,
+      "step": 3041
+    },
+    {
+      "epoch": 0.22383076569326282,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.416062145484791e-05,
+      "loss": 0.6924,
+      "step": 3042
+    },
+    {
+      "epoch": 0.2239043458266268,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.4156900802643844e-05,
+      "loss": 1.6845,
+      "step": 3043
+    },
+    {
+      "epoch": 0.2239779259599908,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.41531791223251e-05,
+      "loss": 0.8964,
+      "step": 3044
+    },
+    {
+      "epoch": 0.2240515060933548,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.4149456414091404e-05,
+      "loss": 1.0024,
+      "step": 3045
+    },
+    {
+      "epoch": 0.22412508622671878,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.414573267814256e-05,
+      "loss": 0.8576,
+      "step": 3046
+    },
+    {
+      "epoch": 0.2241986663600828,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.41420079146784e-05,
+      "loss": 1.1391,
+      "step": 3047
+    },
+    {
+      "epoch": 0.22427224649344676,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.413828212389883e-05,
+      "loss": 0.7299,
+      "step": 3048
+    },
+    {
+      "epoch": 0.22434582662681077,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.4134555306003804e-05,
+      "loss": 0.9945,
+      "step": 3049
+    },
+    {
+      "epoch": 0.22441940676017474,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.4130827461193334e-05,
+      "loss": 1.0733,
+      "step": 3050
+    },
+    {
+      "epoch": 0.22449298689353875,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.412709858966749e-05,
+      "loss": 0.7298,
+      "step": 3051
+    },
+    {
+      "epoch": 0.22456656702690272,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.412336869162639e-05,
+      "loss": 0.8285,
+      "step": 3052
+    },
+    {
+      "epoch": 0.22464014716026673,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.4119637767270204e-05,
+      "loss": 0.9029,
+      "step": 3053
+    },
+    {
+      "epoch": 0.22471372729363073,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.4115905816799186e-05,
+      "loss": 1.1086,
+      "step": 3054
+    },
+    {
+      "epoch": 0.2247873074269947,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.411217284041359e-05,
+      "loss": 0.9996,
+      "step": 3055
+    },
+    {
+      "epoch": 0.2248608875603587,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.410843883831379e-05,
+      "loss": 0.8473,
+      "step": 3056
+    },
+    {
+      "epoch": 0.2249344676937227,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.4104703810700167e-05,
+      "loss": 0.9594,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2250080478270867,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.410096775777316e-05,
+      "loss": 0.7604,
+      "step": 3058
+    },
+    {
+      "epoch": 0.22508162796045067,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.4097230679733305e-05,
+      "loss": 0.959,
+      "step": 3059
+    },
+    {
+      "epoch": 0.22515520809381467,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.409349257678114e-05,
+      "loss": 0.6229,
+      "step": 3060
+    },
+    {
+      "epoch": 0.22522878822717865,
+      "grad_norm": 1.0,
+      "learning_rate": 4.4089753449117296e-05,
+      "loss": 0.9904,
+      "step": 3061
+    },
+    {
+      "epoch": 0.22530236836054265,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.408601329694244e-05,
+      "loss": 0.8742,
+      "step": 3062
+    },
+    {
+      "epoch": 0.22537594849390666,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.4082272120457294e-05,
+      "loss": 0.9492,
+      "step": 3063
+    },
+    {
+      "epoch": 0.22544952862727063,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.407852991986265e-05,
+      "loss": 1.6065,
+      "step": 3064
+    },
+    {
+      "epoch": 0.22552310876063464,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.407478669535934e-05,
+      "loss": 0.7719,
+      "step": 3065
+    },
+    {
+      "epoch": 0.22559668889399861,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.4071042447148246e-05,
+      "loss": 1.1843,
+      "step": 3066
+    },
+    {
+      "epoch": 0.22567026902736262,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.406729717543033e-05,
+      "loss": 0.9189,
+      "step": 3067
+    },
+    {
+      "epoch": 0.2257438491607266,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.406355088040659e-05,
+      "loss": 1.0585,
+      "step": 3068
+    },
+    {
+      "epoch": 0.2258174292940906,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.405980356227808e-05,
+      "loss": 1.1171,
+      "step": 3069
+    },
+    {
+      "epoch": 0.22589100942745458,
+      "grad_norm": 0.875,
+      "learning_rate": 4.4056055221245904e-05,
+      "loss": 1.3748,
+      "step": 3070
+    },
+    {
+      "epoch": 0.22596458956081858,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.4052305857511245e-05,
+      "loss": 0.9345,
+      "step": 3071
+    },
+    {
+      "epoch": 0.22603816969418258,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.404855547127531e-05,
+      "loss": 1.442,
+      "step": 3072
+    },
+    {
+      "epoch": 0.22611174982754656,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.4044804062739385e-05,
+      "loss": 1.0865,
+      "step": 3073
+    },
+    {
+      "epoch": 0.22618532996091056,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.4041051632104795e-05,
+      "loss": 0.6178,
+      "step": 3074
+    },
+    {
+      "epoch": 0.22625891009427454,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.403729817957293e-05,
+      "loss": 0.952,
+      "step": 3075
+    },
+    {
+      "epoch": 0.22633249022763854,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.403354370534522e-05,
+      "loss": 1.2966,
+      "step": 3076
+    },
+    {
+      "epoch": 0.22640607036100252,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.402978820962317e-05,
+      "loss": 1.0168,
+      "step": 3077
+    },
+    {
+      "epoch": 0.22647965049436652,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.402603169260834e-05,
+      "loss": 1.0222,
+      "step": 3078
+    },
+    {
+      "epoch": 0.2265532306277305,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.402227415450231e-05,
+      "loss": 1.119,
+      "step": 3079
+    },
+    {
+      "epoch": 0.2266268107610945,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.401851559550676e-05,
+      "loss": 0.9735,
+      "step": 3080
+    },
+    {
+      "epoch": 0.2267003908944585,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.40147560158234e-05,
+      "loss": 1.0884,
+      "step": 3081
+    },
+    {
+      "epoch": 0.22677397102782249,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.4010995415654e-05,
+      "loss": 1.2867,
+      "step": 3082
+    },
+    {
+      "epoch": 0.2268475511611865,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.400723379520038e-05,
+      "loss": 1.1063,
+      "step": 3083
+    },
+    {
+      "epoch": 0.22692113129455047,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.400347115466442e-05,
+      "loss": 1.4118,
+      "step": 3084
+    },
+    {
+      "epoch": 0.22699471142791447,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.399970749424805e-05,
+      "loss": 0.6913,
+      "step": 3085
+    },
+    {
+      "epoch": 0.22706829156127845,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.399594281415328e-05,
+      "loss": 1.0776,
+      "step": 3086
+    },
+    {
+      "epoch": 0.22714187169464245,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.3992177114582124e-05,
+      "loss": 1.0784,
+      "step": 3087
+    },
+    {
+      "epoch": 0.22721545182800643,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.39884103957367e-05,
+      "loss": 0.9732,
+      "step": 3088
+    },
+    {
+      "epoch": 0.22728903196137043,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.398464265781915e-05,
+      "loss": 0.836,
+      "step": 3089
+    },
+    {
+      "epoch": 0.22736261209473443,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.3980873901031695e-05,
+      "loss": 0.8301,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2274361922280984,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.397710412557657e-05,
+      "loss": 0.7434,
+      "step": 3091
+    },
+    {
+      "epoch": 0.22750977236146241,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.397333333165613e-05,
+      "loss": 1.0613,
+      "step": 3092
+    },
+    {
+      "epoch": 0.2275833524948264,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.3969561519472716e-05,
+      "loss": 0.7699,
+      "step": 3093
+    },
+    {
+      "epoch": 0.2276569326281904,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.3965788689228757e-05,
+      "loss": 0.7306,
+      "step": 3094
+    },
+    {
+      "epoch": 0.22773051276155437,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.3962014841126744e-05,
+      "loss": 1.2517,
+      "step": 3095
+    },
+    {
+      "epoch": 0.22780409289491838,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.3958239975369215e-05,
+      "loss": 1.1137,
+      "step": 3096
+    },
+    {
+      "epoch": 0.22787767302828235,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.3954464092158745e-05,
+      "loss": 0.8204,
+      "step": 3097
+    },
+    {
+      "epoch": 0.22795125316164636,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.395068719169799e-05,
+      "loss": 0.8983,
+      "step": 3098
+    },
+    {
+      "epoch": 0.22802483329501036,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.394690927418965e-05,
+      "loss": 0.8394,
+      "step": 3099
+    },
+    {
+      "epoch": 0.22809841342837434,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.394313033983648e-05,
+      "loss": 1.228,
+      "step": 3100
+    },
+    {
+      "epoch": 0.22817199356173834,
+      "grad_norm": 1.859375,
+      "learning_rate": 4.3939350388841274e-05,
+      "loss": 1.1662,
+      "step": 3101
+    },
+    {
+      "epoch": 0.22824557369510232,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.39355694214069e-05,
+      "loss": 0.8935,
+      "step": 3102
+    },
+    {
+      "epoch": 0.22831915382846632,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.393178743773629e-05,
+      "loss": 0.7445,
+      "step": 3103
+    },
+    {
+      "epoch": 0.2283927339618303,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.39280044380324e-05,
+      "loss": 1.2452,
+      "step": 3104
+    },
+    {
+      "epoch": 0.2284663140951943,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.392422042249826e-05,
+      "loss": 0.8325,
+      "step": 3105
+    },
+    {
+      "epoch": 0.22853989422855828,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.3920435391336956e-05,
+      "loss": 1.4204,
+      "step": 3106
+    },
+    {
+      "epoch": 0.22861347436192228,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.391664934475162e-05,
+      "loss": 0.7727,
+      "step": 3107
+    },
+    {
+      "epoch": 0.22868705449528628,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.391286228294544e-05,
+      "loss": 1.6442,
+      "step": 3108
+    },
+    {
+      "epoch": 0.22876063462865026,
+      "grad_norm": 0.75,
+      "learning_rate": 4.390907420612166e-05,
+      "loss": 0.7756,
+      "step": 3109
+    },
+    {
+      "epoch": 0.22883421476201427,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.3905285114483584e-05,
+      "loss": 1.0006,
+      "step": 3110
+    },
+    {
+      "epoch": 0.22890779489537824,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.390149500823457e-05,
+      "loss": 0.6596,
+      "step": 3111
+    },
+    {
+      "epoch": 0.22898137502874225,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.3897703887578015e-05,
+      "loss": 0.9563,
+      "step": 3112
+    },
+    {
+      "epoch": 0.22905495516210622,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.3893911752717374e-05,
+      "loss": 0.9856,
+      "step": 3113
+    },
+    {
+      "epoch": 0.22912853529547023,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.389011860385619e-05,
+      "loss": 1.0432,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2292021154288342,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.3886324441198015e-05,
+      "loss": 1.5668,
+      "step": 3115
+    },
+    {
+      "epoch": 0.2292756955621982,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.388252926494647e-05,
+      "loss": 0.9112,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2293492756955622,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.387873307530525e-05,
+      "loss": 0.948,
+      "step": 3117
+    },
+    {
+      "epoch": 0.2294228558289262,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.3874935872478086e-05,
+      "loss": 1.1635,
+      "step": 3118
+    },
+    {
+      "epoch": 0.2294964359622902,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.387113765666876e-05,
+      "loss": 0.953,
+      "step": 3119
+    },
+    {
+      "epoch": 0.22957001609565417,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.386733842808112e-05,
+      "loss": 1.118,
+      "step": 3120
+    },
+    {
+      "epoch": 0.22964359622901817,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.386353818691906e-05,
+      "loss": 0.6992,
+      "step": 3121
+    },
+    {
+      "epoch": 0.22971717636238215,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.385973693338653e-05,
+      "loss": 1.0678,
+      "step": 3122
+    },
+    {
+      "epoch": 0.22979075649574615,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.3855934667687545e-05,
+      "loss": 1.6797,
+      "step": 3123
+    },
+    {
+      "epoch": 0.22986433662911013,
+      "grad_norm": 0.75,
+      "learning_rate": 4.385213139002616e-05,
+      "loss": 0.8481,
+      "step": 3124
+    },
+    {
+      "epoch": 0.22993791676247413,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.3848327100606486e-05,
+      "loss": 0.7289,
+      "step": 3125
+    },
+    {
+      "epoch": 0.23001149689583814,
+      "grad_norm": 1.0,
+      "learning_rate": 4.38445217996327e-05,
+      "loss": 1.4812,
+      "step": 3126
+    },
+    {
+      "epoch": 0.2300850770292021,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.3840715487309015e-05,
+      "loss": 1.2339,
+      "step": 3127
+    },
+    {
+      "epoch": 0.23015865716256612,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.3836908163839715e-05,
+      "loss": 0.7561,
+      "step": 3128
+    },
+    {
+      "epoch": 0.2302322372959301,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.383309982942914e-05,
+      "loss": 0.6729,
+      "step": 3129
+    },
+    {
+      "epoch": 0.2303058174292941,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.382929048428167e-05,
+      "loss": 0.9063,
+      "step": 3130
+    },
+    {
+      "epoch": 0.23037939756265807,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.3825480128601734e-05,
+      "loss": 1.3051,
+      "step": 3131
+    },
+    {
+      "epoch": 0.23045297769602208,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.382166876259384e-05,
+      "loss": 1.0217,
+      "step": 3132
+    },
+    {
+      "epoch": 0.23052655782938605,
+      "grad_norm": 1.0,
+      "learning_rate": 4.381785638646253e-05,
+      "loss": 1.2486,
+      "step": 3133
+    },
+    {
+      "epoch": 0.23060013796275006,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.3814043000412405e-05,
+      "loss": 0.6965,
+      "step": 3134
+    },
+    {
+      "epoch": 0.23067371809611406,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.381022860464814e-05,
+      "loss": 0.8494,
+      "step": 3135
+    },
+    {
+      "epoch": 0.23074729822947804,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.380641319937442e-05,
+      "loss": 0.9846,
+      "step": 3136
+    },
+    {
+      "epoch": 0.23082087836284204,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.380259678479604e-05,
+      "loss": 0.9081,
+      "step": 3137
+    },
+    {
+      "epoch": 0.23089445849620602,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.379877936111779e-05,
+      "loss": 0.839,
+      "step": 3138
+    },
+    {
+      "epoch": 0.23096803862957002,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.379496092854456e-05,
+      "loss": 1.0511,
+      "step": 3139
+    },
+    {
+      "epoch": 0.231041618762934,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.379114148728128e-05,
+      "loss": 0.9411,
+      "step": 3140
+    },
+    {
+      "epoch": 0.231115198896298,
+      "grad_norm": 1.125,
+      "learning_rate": 4.378732103753292e-05,
+      "loss": 1.3649,
+      "step": 3141
+    },
+    {
+      "epoch": 0.23118877902966198,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.378349957950453e-05,
+      "loss": 0.7724,
+      "step": 3142
+    },
+    {
+      "epoch": 0.23126235916302598,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.37796771134012e-05,
+      "loss": 1.2308,
+      "step": 3143
+    },
+    {
+      "epoch": 0.23133593929639,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.377585363942805e-05,
+      "loss": 1.2,
+      "step": 3144
+    },
+    {
+      "epoch": 0.23140951942975396,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.377202915779032e-05,
+      "loss": 0.7315,
+      "step": 3145
+    },
+    {
+      "epoch": 0.23148309956311797,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.376820366869323e-05,
+      "loss": 0.65,
+      "step": 3146
+    },
+    {
+      "epoch": 0.23155667969648194,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.3764377172342095e-05,
+      "loss": 0.7376,
+      "step": 3147
+    },
+    {
+      "epoch": 0.23163025982984595,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.376054966894229e-05,
+      "loss": 1.3158,
+      "step": 3148
+    },
+    {
+      "epoch": 0.23170383996320992,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.37567211586992e-05,
+      "loss": 0.972,
+      "step": 3149
+    },
+    {
+      "epoch": 0.23177742009657393,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.375289164181832e-05,
+      "loss": 0.8627,
+      "step": 3150
+    },
+    {
+      "epoch": 0.2318510002299379,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.374906111850517e-05,
+      "loss": 0.9473,
+      "step": 3151
+    },
+    {
+      "epoch": 0.2319245803633019,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.374522958896532e-05,
+      "loss": 0.7579,
+      "step": 3152
+    },
+    {
+      "epoch": 0.2319981604966659,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.37413970534044e-05,
+      "loss": 0.9142,
+      "step": 3153
+    },
+    {
+      "epoch": 0.2320717406300299,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.373756351202809e-05,
+      "loss": 0.8533,
+      "step": 3154
+    },
+    {
+      "epoch": 0.2321453207633939,
+      "grad_norm": 1.0,
+      "learning_rate": 4.373372896504215e-05,
+      "loss": 1.1974,
+      "step": 3155
+    },
+    {
+      "epoch": 0.23221890089675787,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.372989341265235e-05,
+      "loss": 0.7317,
+      "step": 3156
+    },
+    {
+      "epoch": 0.23229248103012187,
+      "grad_norm": 0.75,
+      "learning_rate": 4.372605685506455e-05,
+      "loss": 0.8871,
+      "step": 3157
+    },
+    {
+      "epoch": 0.23236606116348585,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.372221929248464e-05,
+      "loss": 1.2479,
+      "step": 3158
+    },
+    {
+      "epoch": 0.23243964129684985,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.37183807251186e-05,
+      "loss": 0.7314,
+      "step": 3159
+    },
+    {
+      "epoch": 0.23251322143021383,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.3714541153172405e-05,
+      "loss": 0.7643,
+      "step": 3160
+    },
+    {
+      "epoch": 0.23258680156357783,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.371070057685214e-05,
+      "loss": 0.7784,
+      "step": 3161
+    },
+    {
+      "epoch": 0.23266038169694184,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.3706858996363906e-05,
+      "loss": 0.7401,
+      "step": 3162
+    },
+    {
+      "epoch": 0.2327339618303058,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.3703016411913886e-05,
+      "loss": 0.9218,
+      "step": 3163
+    },
+    {
+      "epoch": 0.23280754196366982,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.369917282370831e-05,
+      "loss": 0.7555,
+      "step": 3164
+    },
+    {
+      "epoch": 0.2328811220970338,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.369532823195344e-05,
+      "loss": 1.0403,
+      "step": 3165
+    },
+    {
+      "epoch": 0.2329547022303978,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.369148263685561e-05,
+      "loss": 0.6592,
+      "step": 3166
+    },
+    {
+      "epoch": 0.23302828236376177,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.36876360386212e-05,
+      "loss": 1.0878,
+      "step": 3167
+    },
+    {
+      "epoch": 0.23310186249712578,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.368378843745668e-05,
+      "loss": 0.5583,
+      "step": 3168
+    },
+    {
+      "epoch": 0.23317544263048975,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.367993983356852e-05,
+      "loss": 0.8004,
+      "step": 3169
+    },
+    {
+      "epoch": 0.23324902276385376,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.367609022716326e-05,
+      "loss": 1.3917,
+      "step": 3170
+    },
+    {
+      "epoch": 0.23332260289721776,
+      "grad_norm": 1.125,
+      "learning_rate": 4.367223961844752e-05,
+      "loss": 1.025,
+      "step": 3171
+    },
+    {
+      "epoch": 0.23339618303058174,
+      "grad_norm": 0.75,
+      "learning_rate": 4.366838800762795e-05,
+      "loss": 0.8707,
+      "step": 3172
+    },
+    {
+      "epoch": 0.23346976316394574,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.366453539491124e-05,
+      "loss": 1.419,
+      "step": 3173
+    },
+    {
+      "epoch": 0.23354334329730972,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.3660681780504176e-05,
+      "loss": 1.1744,
+      "step": 3174
+    },
+    {
+      "epoch": 0.23361692343067372,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.365682716461357e-05,
+      "loss": 0.6458,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2336905035640377,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.365297154744629e-05,
+      "loss": 0.9822,
+      "step": 3176
+    },
+    {
+      "epoch": 0.2337640836974017,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.3649114929209245e-05,
+      "loss": 0.9373,
+      "step": 3177
+    },
+    {
+      "epoch": 0.23383766383076568,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.364525731010943e-05,
+      "loss": 1.223,
+      "step": 3178
+    },
+    {
+      "epoch": 0.23391124396412968,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.3641398690353875e-05,
+      "loss": 0.9826,
+      "step": 3179
+    },
+    {
+      "epoch": 0.2339848240974937,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.363753907014966e-05,
+      "loss": 0.9722,
+      "step": 3180
+    },
+    {
+      "epoch": 0.23405840423085766,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.363367844970392e-05,
+      "loss": 0.8164,
+      "step": 3181
+    },
+    {
+      "epoch": 0.23413198436422167,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.362981682922386e-05,
+      "loss": 1.3447,
+      "step": 3182
+    },
+    {
+      "epoch": 0.23420556449758564,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.3625954208916705e-05,
+      "loss": 0.7872,
+      "step": 3183
+    },
+    {
+      "epoch": 0.23427914463094965,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.3622090588989784e-05,
+      "loss": 1.0747,
+      "step": 3184
+    },
+    {
+      "epoch": 0.23435272476431362,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.3618225969650416e-05,
+      "loss": 1.0023,
+      "step": 3185
+    },
+    {
+      "epoch": 0.23442630489767763,
+      "grad_norm": 1.140625,
+      "learning_rate": 4.361436035110604e-05,
+      "loss": 1.1984,
+      "step": 3186
+    },
+    {
+      "epoch": 0.2344998850310416,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.3610493733564095e-05,
+      "loss": 1.2764,
+      "step": 3187
+    },
+    {
+      "epoch": 0.2345734651644056,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.36066261172321e-05,
+      "loss": 1.021,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2346470452977696,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.360275750231763e-05,
+      "loss": 0.7916,
+      "step": 3189
+    },
+    {
+      "epoch": 0.2347206254311336,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.359888788902828e-05,
+      "loss": 1.236,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2347942055644976,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.3595017277571774e-05,
+      "loss": 0.7047,
+      "step": 3191
+    },
+    {
+      "epoch": 0.23486778569786157,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.359114566815579e-05,
+      "loss": 0.4961,
+      "step": 3192
+    },
+    {
+      "epoch": 0.23494136583122557,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.358727306098814e-05,
+      "loss": 0.7679,
+      "step": 3193
+    },
+    {
+      "epoch": 0.23501494596458955,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.358339945627665e-05,
+      "loss": 1.1169,
+      "step": 3194
+    },
+    {
+      "epoch": 0.23508852609795355,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.3579524854229215e-05,
+      "loss": 0.956,
+      "step": 3195
+    },
+    {
+      "epoch": 0.23516210623131756,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.3575649255053774e-05,
+      "loss": 0.903,
+      "step": 3196
+    },
+    {
+      "epoch": 0.23523568636468153,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.357177265895831e-05,
+      "loss": 0.9522,
+      "step": 3197
+    },
+    {
+      "epoch": 0.23530926649804554,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.3567895066150894e-05,
+      "loss": 0.8687,
+      "step": 3198
+    },
+    {
+      "epoch": 0.23538284663140951,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.356401647683962e-05,
+      "loss": 0.624,
+      "step": 3199
+    },
+    {
+      "epoch": 0.23545642676477352,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.356013689123264e-05,
+      "loss": 1.0651,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2355300068981375,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.355625630953817e-05,
+      "loss": 0.7108,
+      "step": 3201
+    },
+    {
+      "epoch": 0.2356035870315015,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.355237473196447e-05,
+      "loss": 0.7632,
+      "step": 3202
+    },
+    {
+      "epoch": 0.23567716716486548,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.354849215871986e-05,
+      "loss": 0.704,
+      "step": 3203
+    },
+    {
+      "epoch": 0.23575074729822948,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.354460859001272e-05,
+      "loss": 1.0499,
+      "step": 3204
+    },
+    {
+      "epoch": 0.23582432743159348,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.354072402605144e-05,
+      "loss": 0.992,
+      "step": 3205
+    },
+    {
+      "epoch": 0.23589790756495746,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.353683846704453e-05,
+      "loss": 0.788,
+      "step": 3206
+    },
+    {
+      "epoch": 0.23597148769832146,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.3532951913200516e-05,
+      "loss": 0.6209,
+      "step": 3207
+    },
+    {
+      "epoch": 0.23604506783168544,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.3529064364727974e-05,
+      "loss": 1.0249,
+      "step": 3208
+    },
+    {
+      "epoch": 0.23611864796504944,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.352517582183554e-05,
+      "loss": 1.0625,
+      "step": 3209
+    },
+    {
+      "epoch": 0.23619222809841342,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.352128628473191e-05,
+      "loss": 0.8386,
+      "step": 3210
+    },
+    {
+      "epoch": 0.23626580823177742,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.351739575362583e-05,
+      "loss": 0.976,
+      "step": 3211
+    },
+    {
+      "epoch": 0.2363393883651414,
+      "grad_norm": 0.875,
+      "learning_rate": 4.3513504228726096e-05,
+      "loss": 1.0203,
+      "step": 3212
+    },
+    {
+      "epoch": 0.2364129684985054,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.350961171024155e-05,
+      "loss": 0.9551,
+      "step": 3213
+    },
+    {
+      "epoch": 0.2364865486318694,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.350571819838111e-05,
+      "loss": 1.2153,
+      "step": 3214
+    },
+    {
+      "epoch": 0.23656012876523339,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.350182369335372e-05,
+      "loss": 0.7942,
+      "step": 3215
+    },
+    {
+      "epoch": 0.2366337088985974,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.3497928195368406e-05,
+      "loss": 1.0052,
+      "step": 3216
+    },
+    {
+      "epoch": 0.23670728903196137,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.349403170463421e-05,
+      "loss": 0.6689,
+      "step": 3217
+    },
+    {
+      "epoch": 0.23678086916532537,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.349013422136028e-05,
+      "loss": 1.0577,
+      "step": 3218
+    },
+    {
+      "epoch": 0.23685444929868935,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.348623574575575e-05,
+      "loss": 0.651,
+      "step": 3219
+    },
+    {
+      "epoch": 0.23692802943205335,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.348233627802988e-05,
+      "loss": 0.8211,
+      "step": 3220
+    },
+    {
+      "epoch": 0.23700160956541733,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.347843581839193e-05,
+      "loss": 1.2313,
+      "step": 3221
+    },
+    {
+      "epoch": 0.23707518969878133,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.3474534367051216e-05,
+      "loss": 0.632,
+      "step": 3222
+    },
+    {
+      "epoch": 0.23714876983214533,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.347063192421714e-05,
+      "loss": 1.0363,
+      "step": 3223
+    },
+    {
+      "epoch": 0.2372223499655093,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.3466728490099143e-05,
+      "loss": 1.0407,
+      "step": 3224
+    },
+    {
+      "epoch": 0.23729593009887331,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.3462824064906704e-05,
+      "loss": 0.8869,
+      "step": 3225
+    },
+    {
+      "epoch": 0.2373695102322373,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.345891864884937e-05,
+      "loss": 1.3454,
+      "step": 3226
+    },
+    {
+      "epoch": 0.2374430903656013,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.345501224213673e-05,
+      "loss": 0.8759,
+      "step": 3227
+    },
+    {
+      "epoch": 0.23751667049896527,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.3451104844978444e-05,
+      "loss": 1.0211,
+      "step": 3228
+    },
+    {
+      "epoch": 0.23759025063232928,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.34471964575842e-05,
+      "loss": 0.7893,
+      "step": 3229
+    },
+    {
+      "epoch": 0.23766383076569325,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.3443287080163777e-05,
+      "loss": 1.0153,
+      "step": 3230
+    },
+    {
+      "epoch": 0.23773741089905726,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.343937671292697e-05,
+      "loss": 1.2079,
+      "step": 3231
+    },
+    {
+      "epoch": 0.23781099103242126,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.343546535608364e-05,
+      "loss": 1.1658,
+      "step": 3232
+    },
+    {
+      "epoch": 0.23788457116578524,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.343155300984369e-05,
+      "loss": 0.6722,
+      "step": 3233
+    },
+    {
+      "epoch": 0.23795815129914924,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.342763967441712e-05,
+      "loss": 1.0483,
+      "step": 3234
+    },
+    {
+      "epoch": 0.23803173143251322,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.342372535001393e-05,
+      "loss": 0.8137,
+      "step": 3235
+    },
+    {
+      "epoch": 0.23810531156587722,
+      "grad_norm": 0.875,
+      "learning_rate": 4.34198100368442e-05,
+      "loss": 1.3357,
+      "step": 3236
+    },
+    {
+      "epoch": 0.2381788916992412,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.341589373511805e-05,
+      "loss": 0.7589,
+      "step": 3237
+    },
+    {
+      "epoch": 0.2382524718326052,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.3411976445045664e-05,
+      "loss": 0.9209,
+      "step": 3238
+    },
+    {
+      "epoch": 0.23832605196596918,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.340805816683728e-05,
+      "loss": 1.4987,
+      "step": 3239
+    },
+    {
+      "epoch": 0.23839963209933318,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.340413890070318e-05,
+      "loss": 0.5365,
+      "step": 3240
+    },
+    {
+      "epoch": 0.23847321223269718,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.340021864685371e-05,
+      "loss": 1.1462,
+      "step": 3241
+    },
+    {
+      "epoch": 0.23854679236606116,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.339629740549926e-05,
+      "loss": 0.7768,
+      "step": 3242
+    },
+    {
+      "epoch": 0.23862037249942517,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.339237517685027e-05,
+      "loss": 0.7225,
+      "step": 3243
+    },
+    {
+      "epoch": 0.23869395263278914,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.3388451961117246e-05,
+      "loss": 0.7952,
+      "step": 3244
+    },
+    {
+      "epoch": 0.23876753276615315,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.338452775851073e-05,
+      "loss": 0.8929,
+      "step": 3245
+    },
+    {
+      "epoch": 0.23884111289951712,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.338060256924134e-05,
+      "loss": 0.9899,
+      "step": 3246
+    },
+    {
+      "epoch": 0.23891469303288113,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.337667639351972e-05,
+      "loss": 1.0291,
+      "step": 3247
+    },
+    {
+      "epoch": 0.2389882731662451,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.337274923155659e-05,
+      "loss": 0.6663,
+      "step": 3248
+    },
+    {
+      "epoch": 0.2390618532996091,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.33688210835627e-05,
+      "loss": 1.2437,
+      "step": 3249
+    },
+    {
+      "epoch": 0.2391354334329731,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.336489194974889e-05,
+      "loss": 0.7705,
+      "step": 3250
+    },
+    {
+      "epoch": 0.2392090135663371,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.336096183032601e-05,
+      "loss": 1.0608,
+      "step": 3251
+    },
+    {
+      "epoch": 0.2392825936997011,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.335703072550499e-05,
+      "loss": 0.6806,
+      "step": 3252
+    },
+    {
+      "epoch": 0.23935617383306507,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.33530986354968e-05,
+      "loss": 0.68,
+      "step": 3253
+    },
+    {
+      "epoch": 0.23942975396642907,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.334916556051247e-05,
+      "loss": 1.1363,
+      "step": 3254
+    },
+    {
+      "epoch": 0.23950333409979305,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.3345231500763084e-05,
+      "loss": 0.7675,
+      "step": 3255
+    },
+    {
+      "epoch": 0.23957691423315705,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.334129645645977e-05,
+      "loss": 1.0675,
+      "step": 3256
+    },
+    {
+      "epoch": 0.23965049436652103,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.333736042781371e-05,
+      "loss": 0.6276,
+      "step": 3257
+    },
+    {
+      "epoch": 0.23972407449988503,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.333342341503616e-05,
+      "loss": 0.7948,
+      "step": 3258
+    },
+    {
+      "epoch": 0.23979765463324904,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.3329485418338403e-05,
+      "loss": 0.8094,
+      "step": 3259
+    },
+    {
+      "epoch": 0.239871234766613,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.332554643793177e-05,
+      "loss": 0.9658,
+      "step": 3260
+    },
+    {
+      "epoch": 0.23994481489997702,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.332160647402769e-05,
+      "loss": 0.9935,
+      "step": 3261
+    },
+    {
+      "epoch": 0.240018395033341,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.331766552683758e-05,
+      "loss": 1.2022,
+      "step": 3262
+    },
+    {
+      "epoch": 0.240091975166705,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.331372359657297e-05,
+      "loss": 0.731,
+      "step": 3263
+    },
+    {
+      "epoch": 0.24016555530006897,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.33097806834454e-05,
+      "loss": 0.7802,
+      "step": 3264
+    },
+    {
+      "epoch": 0.24023913543343298,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.3305836787666475e-05,
+      "loss": 0.6186,
+      "step": 3265
+    },
+    {
+      "epoch": 0.24031271556679695,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.330189190944788e-05,
+      "loss": 1.4881,
+      "step": 3266
+    },
+    {
+      "epoch": 0.24038629570016096,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.329794604900131e-05,
+      "loss": 1.1482,
+      "step": 3267
+    },
+    {
+      "epoch": 0.24045987583352496,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.329399920653852e-05,
+      "loss": 0.8388,
+      "step": 3268
+    },
+    {
+      "epoch": 0.24053345596688894,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.3290051382271356e-05,
+      "loss": 1.34,
+      "step": 3269
+    },
+    {
+      "epoch": 0.24060703610025294,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.328610257641168e-05,
+      "loss": 1.4536,
+      "step": 3270
+    },
+    {
+      "epoch": 0.24068061623361692,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.3282152789171414e-05,
+      "loss": 0.9708,
+      "step": 3271
+    },
+    {
+      "epoch": 0.24075419636698092,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.327820202076254e-05,
+      "loss": 0.7082,
+      "step": 3272
+    },
+    {
+      "epoch": 0.2408277765003449,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.3274250271397085e-05,
+      "loss": 0.8295,
+      "step": 3273
+    },
+    {
+      "epoch": 0.2409013566337089,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.327029754128714e-05,
+      "loss": 0.7234,
+      "step": 3274
+    },
+    {
+      "epoch": 0.24097493676707288,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.326634383064482e-05,
+      "loss": 1.2526,
+      "step": 3275
+    },
+    {
+      "epoch": 0.24104851690043688,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.326238913968234e-05,
+      "loss": 1.2503,
+      "step": 3276
+    },
+    {
+      "epoch": 0.2411220970338009,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.3258433468611914e-05,
+      "loss": 0.6983,
+      "step": 3277
+    },
+    {
+      "epoch": 0.24119567716716486,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.325447681764586e-05,
+      "loss": 0.8075,
+      "step": 3278
+    },
+    {
+      "epoch": 0.24126925730052887,
+      "grad_norm": 0.875,
+      "learning_rate": 4.325051918699651e-05,
+      "loss": 0.8359,
+      "step": 3279
+    },
+    {
+      "epoch": 0.24134283743389284,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.324656057687626e-05,
+      "loss": 1.1151,
+      "step": 3280
+    },
+    {
+      "epoch": 0.24141641756725685,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.324260098749757e-05,
+      "loss": 0.6284,
+      "step": 3281
+    },
+    {
+      "epoch": 0.24148999770062082,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.3238640419072945e-05,
+      "loss": 1.3291,
+      "step": 3282
+    },
+    {
+      "epoch": 0.24156357783398483,
+      "grad_norm": 0.75,
+      "learning_rate": 4.323467887181494e-05,
+      "loss": 0.7874,
+      "step": 3283
+    },
+    {
+      "epoch": 0.2416371579673488,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.323071634593615e-05,
+      "loss": 1.1782,
+      "step": 3284
+    },
+    {
+      "epoch": 0.2417107381007128,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.322675284164925e-05,
+      "loss": 0.6891,
+      "step": 3285
+    },
+    {
+      "epoch": 0.2417843182340768,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.3222788359166957e-05,
+      "loss": 0.7173,
+      "step": 3286
+    },
+    {
+      "epoch": 0.2418578983674408,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.321882289870202e-05,
+      "loss": 1.1282,
+      "step": 3287
+    },
+    {
+      "epoch": 0.2419314785008048,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.321485646046727e-05,
+      "loss": 0.969,
+      "step": 3288
+    },
+    {
+      "epoch": 0.24200505863416877,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.321088904467559e-05,
+      "loss": 0.9398,
+      "step": 3289
+    },
+    {
+      "epoch": 0.24207863876753277,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.3206920651539873e-05,
+      "loss": 0.8897,
+      "step": 3290
+    },
+    {
+      "epoch": 0.24215221890089675,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.320295128127312e-05,
+      "loss": 0.8354,
+      "step": 3291
+    },
+    {
+      "epoch": 0.24222579903426075,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.3198980934088354e-05,
+      "loss": 0.7275,
+      "step": 3292
+    },
+    {
+      "epoch": 0.24229937916762473,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.319500961019865e-05,
+      "loss": 1.0167,
+      "step": 3293
+    },
+    {
+      "epoch": 0.24237295930098873,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.319103730981715e-05,
+      "loss": 0.815,
+      "step": 3294
+    },
+    {
+      "epoch": 0.24244653943435274,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.318706403315703e-05,
+      "loss": 0.8111,
+      "step": 3295
+    },
+    {
+      "epoch": 0.2425201195677167,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.318308978043154e-05,
+      "loss": 1.174,
+      "step": 3296
+    },
+    {
+      "epoch": 0.24259369970108072,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.317911455185396e-05,
+      "loss": 1.1623,
+      "step": 3297
+    },
+    {
+      "epoch": 0.2426672798344447,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.3175138347637646e-05,
+      "loss": 0.6876,
+      "step": 3298
+    },
+    {
+      "epoch": 0.2427408599678087,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.317116116799598e-05,
+      "loss": 1.4496,
+      "step": 3299
+    },
+    {
+      "epoch": 0.24281444010117267,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.316718301314241e-05,
+      "loss": 1.355,
+      "step": 3300
+    },
+    {
+      "epoch": 0.24288802023453668,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.316320388329044e-05,
+      "loss": 0.7542,
+      "step": 3301
+    },
+    {
+      "epoch": 0.24296160036790065,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.315922377865363e-05,
+      "loss": 0.9499,
+      "step": 3302
+    },
+    {
+      "epoch": 0.24303518050126466,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.315524269944557e-05,
+      "loss": 0.5671,
+      "step": 3303
+    },
+    {
+      "epoch": 0.24310876063462866,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.3151260645879934e-05,
+      "loss": 0.6535,
+      "step": 3304
+    },
+    {
+      "epoch": 0.24318234076799264,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.3147277618170415e-05,
+      "loss": 1.1626,
+      "step": 3305
+    },
+    {
+      "epoch": 0.24325592090135664,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.3143293616530784e-05,
+      "loss": 0.9798,
+      "step": 3306
+    },
+    {
+      "epoch": 0.24332950103472062,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.3139308641174855e-05,
+      "loss": 0.9162,
+      "step": 3307
+    },
+    {
+      "epoch": 0.24340308116808462,
+      "grad_norm": 0.75,
+      "learning_rate": 4.3135322692316484e-05,
+      "loss": 0.7868,
+      "step": 3308
+    },
+    {
+      "epoch": 0.2434766613014486,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.31313357701696e-05,
+      "loss": 1.0084,
+      "step": 3309
+    },
+    {
+      "epoch": 0.2435502414348126,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.3127347874948177e-05,
+      "loss": 1.2432,
+      "step": 3310
+    },
+    {
+      "epoch": 0.24362382156817658,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.312335900686623e-05,
+      "loss": 0.8526,
+      "step": 3311
+    },
+    {
+      "epoch": 0.24369740170154058,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.311936916613782e-05,
+      "loss": 0.7488,
+      "step": 3312
+    },
+    {
+      "epoch": 0.2437709818349046,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.311537835297711e-05,
+      "loss": 0.8447,
+      "step": 3313
+    },
+    {
+      "epoch": 0.24384456196826856,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.3111386567598245e-05,
+      "loss": 1.2009,
+      "step": 3314
+    },
+    {
+      "epoch": 0.24391814210163257,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.3107393810215475e-05,
+      "loss": 0.912,
+      "step": 3315
+    },
+    {
+      "epoch": 0.24399172223499654,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.310340008104308e-05,
+      "loss": 1.3959,
+      "step": 3316
+    },
+    {
+      "epoch": 0.24406530236836055,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.30994053802954e-05,
+      "loss": 1.2222,
+      "step": 3317
+    },
+    {
+      "epoch": 0.24413888250172452,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.3095409708186804e-05,
+      "loss": 1.1964,
+      "step": 3318
+    },
+    {
+      "epoch": 0.24421246263508853,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.309141306493176e-05,
+      "loss": 1.406,
+      "step": 3319
+    },
+    {
+      "epoch": 0.2442860427684525,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.308741545074474e-05,
+      "loss": 0.6776,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2443596229018165,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.3083416865840296e-05,
+      "loss": 0.68,
+      "step": 3321
+    },
+    {
+      "epoch": 0.2444332030351805,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.307941731043302e-05,
+      "loss": 0.8754,
+      "step": 3322
+    },
+    {
+      "epoch": 0.2445067831685445,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.307541678473758e-05,
+      "loss": 0.9996,
+      "step": 3323
+    },
+    {
+      "epoch": 0.2445803633019085,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.307141528896864e-05,
+      "loss": 0.7086,
+      "step": 3324
+    },
+    {
+      "epoch": 0.24465394343527247,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.306741282334099e-05,
+      "loss": 0.9456,
+      "step": 3325
+    },
+    {
+      "epoch": 0.24472752356863647,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.306340938806941e-05,
+      "loss": 1.3534,
+      "step": 3326
+    },
+    {
+      "epoch": 0.24480110370200045,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.305940498336876e-05,
+      "loss": 0.7491,
+      "step": 3327
+    },
+    {
+      "epoch": 0.24487468383536445,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.305539960945396e-05,
+      "loss": 0.9569,
+      "step": 3328
+    },
+    {
+      "epoch": 0.24494826396872843,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.3051393266539964e-05,
+      "loss": 0.8684,
+      "step": 3329
+    },
+    {
+      "epoch": 0.24502184410209243,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.304738595484179e-05,
+      "loss": 0.9582,
+      "step": 3330
+    },
+    {
+      "epoch": 0.24509542423545644,
+      "grad_norm": 0.875,
+      "learning_rate": 4.304337767457449e-05,
+      "loss": 0.9837,
+      "step": 3331
+    },
+    {
+      "epoch": 0.24516900436882041,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.30393684259532e-05,
+      "loss": 1.0825,
+      "step": 3332
+    },
+    {
+      "epoch": 0.24524258450218442,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.3035358209193065e-05,
+      "loss": 1.397,
+      "step": 3333
+    },
+    {
+      "epoch": 0.2453161646355484,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.3031347024509336e-05,
+      "loss": 0.8771,
+      "step": 3334
+    },
+    {
+      "epoch": 0.2453897447689124,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.302733487211725e-05,
+      "loss": 0.9097,
+      "step": 3335
+    },
+    {
+      "epoch": 0.24546332490227638,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.302332175223216e-05,
+      "loss": 0.9695,
+      "step": 3336
+    },
+    {
+      "epoch": 0.24553690503564038,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.301930766506943e-05,
+      "loss": 1.063,
+      "step": 3337
+    },
+    {
+      "epoch": 0.24561048516900436,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.301529261084449e-05,
+      "loss": 0.7604,
+      "step": 3338
+    },
+    {
+      "epoch": 0.24568406530236836,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.301127658977283e-05,
+      "loss": 0.822,
+      "step": 3339
+    },
+    {
+      "epoch": 0.24575764543573236,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.3007259602069974e-05,
+      "loss": 0.8355,
+      "step": 3340
+    },
+    {
+      "epoch": 0.24583122556909634,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.3003241647951494e-05,
+      "loss": 0.7786,
+      "step": 3341
+    },
+    {
+      "epoch": 0.24590480570246034,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.299922272763305e-05,
+      "loss": 1.0491,
+      "step": 3342
+    },
+    {
+      "epoch": 0.24597838583582432,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.2995202841330305e-05,
+      "loss": 0.8744,
+      "step": 3343
+    },
+    {
+      "epoch": 0.24605196596918832,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.299118198925902e-05,
+      "loss": 1.2905,
+      "step": 3344
+    },
+    {
+      "epoch": 0.2461255461025523,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.2987160171634975e-05,
+      "loss": 0.8037,
+      "step": 3345
+    },
+    {
+      "epoch": 0.2461991262359163,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.2983137388674024e-05,
+      "loss": 0.855,
+      "step": 3346
+    },
+    {
+      "epoch": 0.24627270636928028,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.297911364059205e-05,
+      "loss": 0.9314,
+      "step": 3347
+    },
+    {
+      "epoch": 0.24634628650264428,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.2975088927605e-05,
+      "loss": 0.9432,
+      "step": 3348
+    },
+    {
+      "epoch": 0.2464198666360083,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.297106324992888e-05,
+      "loss": 1.0491,
+      "step": 3349
+    },
+    {
+      "epoch": 0.24649344676937227,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.2967036607779745e-05,
+      "loss": 1.1294,
+      "step": 3350
+    },
+    {
+      "epoch": 0.24656702690273627,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.2963009001373684e-05,
+      "loss": 1.3044,
+      "step": 3351
+    },
+    {
+      "epoch": 0.24664060703610025,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.295898043092685e-05,
+      "loss": 0.7408,
+      "step": 3352
+    },
+    {
+      "epoch": 0.24671418716946425,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.2954950896655466e-05,
+      "loss": 0.8605,
+      "step": 3353
+    },
+    {
+      "epoch": 0.24678776730282823,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.2950920398775783e-05,
+      "loss": 0.7389,
+      "step": 3354
+    },
+    {
+      "epoch": 0.24686134743619223,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.29468889375041e-05,
+      "loss": 1.1959,
+      "step": 3355
+    },
+    {
+      "epoch": 0.2469349275695562,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.2942856513056785e-05,
+      "loss": 1.1662,
+      "step": 3356
+    },
+    {
+      "epoch": 0.2470085077029202,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.293882312565025e-05,
+      "loss": 0.8959,
+      "step": 3357
+    },
+    {
+      "epoch": 0.24708208783628421,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.293478877550096e-05,
+      "loss": 0.9975,
+      "step": 3358
+    },
+    {
+      "epoch": 0.2471556679696482,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.293075346282543e-05,
+      "loss": 1.4542,
+      "step": 3359
+    },
+    {
+      "epoch": 0.2472292481030122,
+      "grad_norm": 0.875,
+      "learning_rate": 4.2926717187840225e-05,
+      "loss": 0.6137,
+      "step": 3360
+    },
+    {
+      "epoch": 0.24730282823637617,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.2922679950761975e-05,
+      "loss": 0.9547,
+      "step": 3361
+    },
+    {
+      "epoch": 0.24737640836974017,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.291864175180734e-05,
+      "loss": 0.9009,
+      "step": 3362
+    },
+    {
+      "epoch": 0.24744998850310415,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.2914602591193045e-05,
+      "loss": 0.9671,
+      "step": 3363
+    },
+    {
+      "epoch": 0.24752356863646816,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.2910562469135864e-05,
+      "loss": 0.9419,
+      "step": 3364
+    },
+    {
+      "epoch": 0.24759714876983213,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.290652138585262e-05,
+      "loss": 0.8009,
+      "step": 3365
+    },
+    {
+      "epoch": 0.24767072890319614,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.2902479341560195e-05,
+      "loss": 1.1007,
+      "step": 3366
+    },
+    {
+      "epoch": 0.24774430903656014,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.2898436336475525e-05,
+      "loss": 0.8986,
+      "step": 3367
+    },
+    {
+      "epoch": 0.24781788916992412,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.289439237081557e-05,
+      "loss": 0.6785,
+      "step": 3368
+    },
+    {
+      "epoch": 0.24789146930328812,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.2890347444797384e-05,
+      "loss": 1.1315,
+      "step": 3369
+    },
+    {
+      "epoch": 0.2479650494366521,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.2886301558638035e-05,
+      "loss": 0.7375,
+      "step": 3370
+    },
+    {
+      "epoch": 0.2480386295700161,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.288225471255467e-05,
+      "loss": 0.9889,
+      "step": 3371
+    },
+    {
+      "epoch": 0.24811220970338008,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.2878206906764466e-05,
+      "loss": 0.8403,
+      "step": 3372
+    },
+    {
+      "epoch": 0.24818578983674408,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.2874158141484665e-05,
+      "loss": 0.6619,
+      "step": 3373
+    },
+    {
+      "epoch": 0.24825936997010806,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.287010841693255e-05,
+      "loss": 0.9374,
+      "step": 3374
+    },
+    {
+      "epoch": 0.24833295010347206,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.286605773332548e-05,
+      "loss": 0.7605,
+      "step": 3375
+    },
+    {
+      "epoch": 0.24840653023683607,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.286200609088082e-05,
+      "loss": 0.8482,
+      "step": 3376
+    },
+    {
+      "epoch": 0.24848011037020004,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.2857953489816046e-05,
+      "loss": 1.5558,
+      "step": 3377
+    },
+    {
+      "epoch": 0.24855369050356405,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.285389993034863e-05,
+      "loss": 1.1048,
+      "step": 3378
+    },
+    {
+      "epoch": 0.24862727063692802,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.284984541269612e-05,
+      "loss": 1.2074,
+      "step": 3379
+    },
+    {
+      "epoch": 0.24870085077029203,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.284578993707613e-05,
+      "loss": 0.8701,
+      "step": 3380
+    },
+    {
+      "epoch": 0.248774430903656,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.284173350370629e-05,
+      "loss": 0.6837,
+      "step": 3381
+    },
+    {
+      "epoch": 0.24884801103702,
+      "grad_norm": 1.125,
+      "learning_rate": 4.2837676112804323e-05,
+      "loss": 1.0353,
+      "step": 3382
+    },
+    {
+      "epoch": 0.248921591170384,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.283361776458796e-05,
+      "loss": 0.8353,
+      "step": 3383
+    },
+    {
+      "epoch": 0.248995171303748,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.282955845927502e-05,
+      "loss": 1.0915,
+      "step": 3384
+    },
+    {
+      "epoch": 0.249068751437112,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.282549819708335e-05,
+      "loss": 0.8539,
+      "step": 3385
+    },
+    {
+      "epoch": 0.24914233157047597,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.282143697823086e-05,
+      "loss": 1.1669,
+      "step": 3386
+    },
+    {
+      "epoch": 0.24921591170383997,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.281737480293551e-05,
+      "loss": 0.9369,
+      "step": 3387
+    },
+    {
+      "epoch": 0.24928949183720395,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.2813311671415305e-05,
+      "loss": 0.7198,
+      "step": 3388
+    },
+    {
+      "epoch": 0.24936307197056795,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.28092475838883e-05,
+      "loss": 0.7981,
+      "step": 3389
+    },
+    {
+      "epoch": 0.24943665210393193,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.280518254057262e-05,
+      "loss": 1.0883,
+      "step": 3390
+    },
+    {
+      "epoch": 0.24951023223729593,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.280111654168642e-05,
+      "loss": 1.2146,
+      "step": 3391
+    },
+    {
+      "epoch": 0.24958381237065994,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.279704958744792e-05,
+      "loss": 0.9971,
+      "step": 3392
+    },
+    {
+      "epoch": 0.2496573925040239,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.279298167807538e-05,
+      "loss": 1.0907,
+      "step": 3393
+    },
+    {
+      "epoch": 0.24973097263738792,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.278891281378713e-05,
+      "loss": 0.9807,
+      "step": 3394
+    },
+    {
+      "epoch": 0.2498045527707519,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.2784842994801517e-05,
+      "loss": 1.0418,
+      "step": 3395
+    },
+    {
+      "epoch": 0.2498781329041159,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.278077222133697e-05,
+      "loss": 0.665,
+      "step": 3396
+    },
+    {
+      "epoch": 0.24995171303747987,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.277670049361197e-05,
+      "loss": 1.1062,
+      "step": 3397
+    },
+    {
+      "epoch": 0.2500252931708439,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.277262781184502e-05,
+      "loss": 0.8011,
+      "step": 3398
+    },
+    {
+      "epoch": 0.2500988733042079,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.2768554176254705e-05,
+      "loss": 0.8366,
+      "step": 3399
+    },
+    {
+      "epoch": 0.25017245343757183,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.2764479587059646e-05,
+      "loss": 0.9901,
+      "step": 3400
+    },
+    {
+      "epoch": 0.25024603357093583,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.276040404447853e-05,
+      "loss": 0.9289,
+      "step": 3401
+    },
+    {
+      "epoch": 0.25031961370429984,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.275632754873007e-05,
+      "loss": 1.0131,
+      "step": 3402
+    },
+    {
+      "epoch": 0.25039319383766384,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.275225010003304e-05,
+      "loss": 0.8746,
+      "step": 3403
+    },
+    {
+      "epoch": 0.25046677397102785,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.274817169860628e-05,
+      "loss": 1.025,
+      "step": 3404
+    },
+    {
+      "epoch": 0.2505403541043918,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.274409234466867e-05,
+      "loss": 0.9384,
+      "step": 3405
+    },
+    {
+      "epoch": 0.2506139342377558,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.2740012038439135e-05,
+      "loss": 0.8604,
+      "step": 3406
+    },
+    {
+      "epoch": 0.2506875143711198,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.273593078013666e-05,
+      "loss": 0.8112,
+      "step": 3407
+    },
+    {
+      "epoch": 0.2507610945044838,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.2731848569980275e-05,
+      "loss": 0.8045,
+      "step": 3408
+    },
+    {
+      "epoch": 0.25083467463784775,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.272776540818907e-05,
+      "loss": 0.821,
+      "step": 3409
+    },
+    {
+      "epoch": 0.25090825477121176,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.272368129498218e-05,
+      "loss": 0.8682,
+      "step": 3410
+    },
+    {
+      "epoch": 0.25098183490457576,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.271959623057879e-05,
+      "loss": 0.5253,
+      "step": 3411
+    },
+    {
+      "epoch": 0.25105541503793977,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.271551021519814e-05,
+      "loss": 1.1089,
+      "step": 3412
+    },
+    {
+      "epoch": 0.25112899517130377,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.271142324905951e-05,
+      "loss": 0.8853,
+      "step": 3413
+    },
+    {
+      "epoch": 0.2512025753046677,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.270733533238226e-05,
+      "loss": 0.9163,
+      "step": 3414
+    },
+    {
+      "epoch": 0.2512761554380317,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.270324646538576e-05,
+      "loss": 0.8709,
+      "step": 3415
+    },
+    {
+      "epoch": 0.2513497355713957,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.269915664828945e-05,
+      "loss": 1.1787,
+      "step": 3416
+    },
+    {
+      "epoch": 0.25142331570475973,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.269506588131284e-05,
+      "loss": 0.8657,
+      "step": 3417
+    },
+    {
+      "epoch": 0.2514968958381237,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.269097416467547e-05,
+      "loss": 0.9562,
+      "step": 3418
+    },
+    {
+      "epoch": 0.2515704759714877,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.268688149859692e-05,
+      "loss": 0.8756,
+      "step": 3419
+    },
+    {
+      "epoch": 0.2516440561048517,
+      "grad_norm": 0.875,
+      "learning_rate": 4.268278788329685e-05,
+      "loss": 0.7495,
+      "step": 3420
+    },
+    {
+      "epoch": 0.2517176362382157,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.267869331899495e-05,
+      "loss": 0.7501,
+      "step": 3421
+    },
+    {
+      "epoch": 0.2517912163715797,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.267459780591097e-05,
+      "loss": 1.1245,
+      "step": 3422
+    },
+    {
+      "epoch": 0.25186479650494364,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.267050134426471e-05,
+      "loss": 1.3597,
+      "step": 3423
+    },
+    {
+      "epoch": 0.25193837663830765,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.266640393427602e-05,
+      "loss": 1.1856,
+      "step": 3424
+    },
+    {
+      "epoch": 0.25201195677167165,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.2662305576164794e-05,
+      "loss": 1.0716,
+      "step": 3425
+    },
+    {
+      "epoch": 0.25208553690503566,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.265820627015099e-05,
+      "loss": 1.0572,
+      "step": 3426
+    },
+    {
+      "epoch": 0.2521591170383996,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.265410601645461e-05,
+      "loss": 0.7815,
+      "step": 3427
+    },
+    {
+      "epoch": 0.2522326971717636,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.26500048152957e-05,
+      "loss": 1.116,
+      "step": 3428
+    },
+    {
+      "epoch": 0.2523062773051276,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.2645902666894366e-05,
+      "loss": 1.1073,
+      "step": 3429
+    },
+    {
+      "epoch": 0.2523798574384916,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.264179957147077e-05,
+      "loss": 1.2016,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2524534375718556,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.2637695529245104e-05,
+      "loss": 0.8755,
+      "step": 3431
+    },
+    {
+      "epoch": 0.25252701770521957,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.263359054043763e-05,
+      "loss": 0.6902,
+      "step": 3432
+    },
+    {
+      "epoch": 0.2526005978385836,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.262948460526867e-05,
+      "loss": 0.8044,
+      "step": 3433
+    },
+    {
+      "epoch": 0.2526741779719476,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.262537772395856e-05,
+      "loss": 0.8182,
+      "step": 3434
+    },
+    {
+      "epoch": 0.2527477581053116,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.262126989672772e-05,
+      "loss": 0.6779,
+      "step": 3435
+    },
+    {
+      "epoch": 0.25282133823867553,
+      "grad_norm": 1.2578125,
+      "learning_rate": 4.26171611237966e-05,
+      "loss": 0.9907,
+      "step": 3436
+    },
+    {
+      "epoch": 0.25289491837203953,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.261305140538572e-05,
+      "loss": 1.2386,
+      "step": 3437
+    },
+    {
+      "epoch": 0.25296849850540354,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.260894074171564e-05,
+      "loss": 1.3068,
+      "step": 3438
+    },
+    {
+      "epoch": 0.25304207863876754,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.260482913300697e-05,
+      "loss": 0.7777,
+      "step": 3439
+    },
+    {
+      "epoch": 0.25311565877213155,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.260071657948036e-05,
+      "loss": 1.0844,
+      "step": 3440
+    },
+    {
+      "epoch": 0.2531892389054955,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.259660308135655e-05,
+      "loss": 0.8623,
+      "step": 3441
+    },
+    {
+      "epoch": 0.2532628190388595,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.2592488638856274e-05,
+      "loss": 1.1151,
+      "step": 3442
+    },
+    {
+      "epoch": 0.2533363991722235,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.2588373252200374e-05,
+      "loss": 0.8084,
+      "step": 3443
+    },
+    {
+      "epoch": 0.2534099793055875,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.258425692160969e-05,
+      "loss": 0.9513,
+      "step": 3444
+    },
+    {
+      "epoch": 0.25348355943895146,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.258013964730515e-05,
+      "loss": 0.8542,
+      "step": 3445
+    },
+    {
+      "epoch": 0.25355713957231546,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.257602142950773e-05,
+      "loss": 1.0133,
+      "step": 3446
+    },
+    {
+      "epoch": 0.25363071970567946,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.2571902268438426e-05,
+      "loss": 0.8895,
+      "step": 3447
+    },
+    {
+      "epoch": 0.25370429983904347,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.256778216431832e-05,
+      "loss": 1.0018,
+      "step": 3448
+    },
+    {
+      "epoch": 0.25377787997240747,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.2563661117368524e-05,
+      "loss": 0.938,
+      "step": 3449
+    },
+    {
+      "epoch": 0.2538514601057714,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.255953912781021e-05,
+      "loss": 1.0347,
+      "step": 3450
+    },
+    {
+      "epoch": 0.2539250402391354,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.255541619586459e-05,
+      "loss": 0.8427,
+      "step": 3451
+    },
+    {
+      "epoch": 0.25399862037249943,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.255129232175295e-05,
+      "loss": 1.2341,
+      "step": 3452
+    },
+    {
+      "epoch": 0.25407220050586343,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.25471675056966e-05,
+      "loss": 0.7688,
+      "step": 3453
+    },
+    {
+      "epoch": 0.2541457806392274,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.254304174791691e-05,
+      "loss": 1.219,
+      "step": 3454
+    },
+    {
+      "epoch": 0.2542193607725914,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.25389150486353e-05,
+      "loss": 0.6127,
+      "step": 3455
+    },
+    {
+      "epoch": 0.2542929409059554,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.253478740807325e-05,
+      "loss": 1.0402,
+      "step": 3456
+    },
+    {
+      "epoch": 0.2543665210393194,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.253065882645228e-05,
+      "loss": 0.9826,
+      "step": 3457
+    },
+    {
+      "epoch": 0.2544401011726834,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.252652930399395e-05,
+      "loss": 0.9286,
+      "step": 3458
+    },
+    {
+      "epoch": 0.25451368130604735,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.252239884091991e-05,
+      "loss": 0.6667,
+      "step": 3459
+    },
+    {
+      "epoch": 0.25458726143941135,
+      "grad_norm": 0.68359375,
+      "learning_rate": 4.251826743745181e-05,
+      "loss": 0.6704,
+      "step": 3460
+    },
+    {
+      "epoch": 0.25466084157277535,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.251413509381138e-05,
+      "loss": 1.0261,
+      "step": 3461
+    },
+    {
+      "epoch": 0.25473442170613936,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.2510001810220415e-05,
+      "loss": 0.7498,
+      "step": 3462
+    },
+    {
+      "epoch": 0.2548080018395033,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.2505867586900705e-05,
+      "loss": 0.8054,
+      "step": 3463
+    },
+    {
+      "epoch": 0.2548815819728673,
+      "grad_norm": 0.5234375,
+      "learning_rate": 4.250173242407416e-05,
+      "loss": 0.4933,
+      "step": 3464
+    },
+    {
+      "epoch": 0.2549551621062313,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.2497596321962685e-05,
+      "loss": 1.0394,
+      "step": 3465
+    },
+    {
+      "epoch": 0.2550287422395953,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.249345928078827e-05,
+      "loss": 0.8976,
+      "step": 3466
+    },
+    {
+      "epoch": 0.2551023223729593,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.2489321300772927e-05,
+      "loss": 1.02,
+      "step": 3467
+    },
+    {
+      "epoch": 0.25517590250632327,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.248518238213875e-05,
+      "loss": 1.0248,
+      "step": 3468
+    },
+    {
+      "epoch": 0.2552494826396873,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.2481042525107854e-05,
+      "loss": 1.3725,
+      "step": 3469
+    },
+    {
+      "epoch": 0.2553230627730513,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.247690172990242e-05,
+      "loss": 1.2737,
+      "step": 3470
+    },
+    {
+      "epoch": 0.2553966429064153,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.247275999674469e-05,
+      "loss": 0.8972,
+      "step": 3471
+    },
+    {
+      "epoch": 0.2554702230397793,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.2468617325856924e-05,
+      "loss": 0.8607,
+      "step": 3472
+    },
+    {
+      "epoch": 0.25554380317314324,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.246447371746147e-05,
+      "loss": 0.7903,
+      "step": 3473
+    },
+    {
+      "epoch": 0.25561738330650724,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.2460329171780684e-05,
+      "loss": 0.7285,
+      "step": 3474
+    },
+    {
+      "epoch": 0.25569096343987124,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.245618368903702e-05,
+      "loss": 1.1474,
+      "step": 3475
+    },
+    {
+      "epoch": 0.25576454357323525,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.245203726945294e-05,
+      "loss": 1.3079,
+      "step": 3476
+    },
+    {
+      "epoch": 0.2558381237065992,
+      "grad_norm": 1.0,
+      "learning_rate": 4.244788991325098e-05,
+      "loss": 1.0751,
+      "step": 3477
+    },
+    {
+      "epoch": 0.2559117038399632,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.244374162065373e-05,
+      "loss": 1.1544,
+      "step": 3478
+    },
+    {
+      "epoch": 0.2559852839733272,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.243959239188381e-05,
+      "loss": 1.2168,
+      "step": 3479
+    },
+    {
+      "epoch": 0.2560588641066912,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.243544222716391e-05,
+      "loss": 0.9321,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2561324442400552,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.243129112671674e-05,
+      "loss": 0.6829,
+      "step": 3481
+    },
+    {
+      "epoch": 0.25620602437341916,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.242713909076512e-05,
+      "loss": 0.7281,
+      "step": 3482
+    },
+    {
+      "epoch": 0.25627960450678317,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.242298611953185e-05,
+      "loss": 1.1156,
+      "step": 3483
+    },
+    {
+      "epoch": 0.25635318464014717,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.241883221323982e-05,
+      "loss": 0.9548,
+      "step": 3484
+    },
+    {
+      "epoch": 0.2564267647735112,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.241467737211197e-05,
+      "loss": 1.0641,
+      "step": 3485
+    },
+    {
+      "epoch": 0.2565003449068751,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.241052159637128e-05,
+      "loss": 0.8736,
+      "step": 3486
+    },
+    {
+      "epoch": 0.2565739250402391,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.240636488624077e-05,
+      "loss": 0.9449,
+      "step": 3487
+    },
+    {
+      "epoch": 0.25664750517360313,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.2402207241943534e-05,
+      "loss": 0.5731,
+      "step": 3488
+    },
+    {
+      "epoch": 0.25672108530696713,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.239804866370271e-05,
+      "loss": 0.7721,
+      "step": 3489
+    },
+    {
+      "epoch": 0.25679466544033114,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.239388915174147e-05,
+      "loss": 1.068,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2568682455736951,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.2389728706283056e-05,
+      "loss": 0.6954,
+      "step": 3491
+    },
+    {
+      "epoch": 0.2569418257070591,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.2385567327550743e-05,
+      "loss": 0.7014,
+      "step": 3492
+    },
+    {
+      "epoch": 0.2570154058404231,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.238140501576786e-05,
+      "loss": 0.9295,
+      "step": 3493
+    },
+    {
+      "epoch": 0.2570889859737871,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.237724177115781e-05,
+      "loss": 0.8264,
+      "step": 3494
+    },
+    {
+      "epoch": 0.25716256610715105,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.237307759394401e-05,
+      "loss": 1.0209,
+      "step": 3495
+    },
+    {
+      "epoch": 0.25723614624051505,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.236891248434995e-05,
+      "loss": 1.5645,
+      "step": 3496
+    },
+    {
+      "epoch": 0.25730972637387906,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.236474644259917e-05,
+      "loss": 1.1092,
+      "step": 3497
+    },
+    {
+      "epoch": 0.25738330650724306,
+      "grad_norm": 0.63671875,
+      "learning_rate": 4.236057946891524e-05,
+      "loss": 0.7523,
+      "step": 3498
+    },
+    {
+      "epoch": 0.25745688664060706,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.2356411563521794e-05,
+      "loss": 0.7251,
+      "step": 3499
+    },
+    {
+      "epoch": 0.257530466773971,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.235224272664253e-05,
+      "loss": 0.7612,
+      "step": 3500
+    },
+    {
+      "epoch": 0.257604046907335,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.234807295850117e-05,
+      "loss": 0.7898,
+      "step": 3501
+    },
+    {
+      "epoch": 0.257677627040699,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.234390225932149e-05,
+      "loss": 1.2333,
+      "step": 3502
+    },
+    {
+      "epoch": 0.257751207174063,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.2339730629327346e-05,
+      "loss": 0.9314,
+      "step": 3503
+    },
+    {
+      "epoch": 0.257824787307427,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.233555806874261e-05,
+      "loss": 1.0388,
+      "step": 3504
+    },
+    {
+      "epoch": 0.257898367440791,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.233138457779121e-05,
+      "loss": 0.9291,
+      "step": 3505
+    },
+    {
+      "epoch": 0.257971947574155,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.232721015669713e-05,
+      "loss": 0.8482,
+      "step": 3506
+    },
+    {
+      "epoch": 0.258045527707519,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.232303480568441e-05,
+      "loss": 0.7691,
+      "step": 3507
+    },
+    {
+      "epoch": 0.258119107840883,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.2318858524977134e-05,
+      "loss": 0.6183,
+      "step": 3508
+    },
+    {
+      "epoch": 0.25819268797424694,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.231468131479943e-05,
+      "loss": 0.8489,
+      "step": 3509
+    },
+    {
+      "epoch": 0.25826626810761094,
+      "grad_norm": 1.3203125,
+      "learning_rate": 4.2310503175375484e-05,
+      "loss": 1.2199,
+      "step": 3510
+    },
+    {
+      "epoch": 0.25833984824097495,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.230632410692952e-05,
+      "loss": 0.6541,
+      "step": 3511
+    },
+    {
+      "epoch": 0.25841342837433895,
+      "grad_norm": 2.375,
+      "learning_rate": 4.230214410968584e-05,
+      "loss": 1.0912,
+      "step": 3512
+    },
+    {
+      "epoch": 0.2584870085077029,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.229796318386875e-05,
+      "loss": 0.8423,
+      "step": 3513
+    },
+    {
+      "epoch": 0.2585605886410669,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.229378132970266e-05,
+      "loss": 1.0736,
+      "step": 3514
+    },
+    {
+      "epoch": 0.2586341687744309,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.228959854741198e-05,
+      "loss": 0.8272,
+      "step": 3515
+    },
+    {
+      "epoch": 0.2587077489077949,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.228541483722121e-05,
+      "loss": 0.9831,
+      "step": 3516
+    },
+    {
+      "epoch": 0.2587813290411589,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.228123019935487e-05,
+      "loss": 1.0159,
+      "step": 3517
+    },
+    {
+      "epoch": 0.25885490917452286,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.227704463403754e-05,
+      "loss": 1.1013,
+      "step": 3518
+    },
+    {
+      "epoch": 0.25892848930788687,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.2272858141493864e-05,
+      "loss": 0.8541,
+      "step": 3519
+    },
+    {
+      "epoch": 0.25900206944125087,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.226867072194851e-05,
+      "loss": 0.6593,
+      "step": 3520
+    },
+    {
+      "epoch": 0.2590756495746149,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.226448237562621e-05,
+      "loss": 0.8965,
+      "step": 3521
+    },
+    {
+      "epoch": 0.2591492297079788,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.226029310275176e-05,
+      "loss": 0.6424,
+      "step": 3522
+    },
+    {
+      "epoch": 0.2592228098413428,
+      "grad_norm": 1.125,
+      "learning_rate": 4.2256102903549976e-05,
+      "loss": 1.1124,
+      "step": 3523
+    },
+    {
+      "epoch": 0.25929638997470683,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.225191177824574e-05,
+      "loss": 1.2578,
+      "step": 3524
+    },
+    {
+      "epoch": 0.25936997010807084,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.224771972706398e-05,
+      "loss": 0.7997,
+      "step": 3525
+    },
+    {
+      "epoch": 0.25944355024143484,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.224352675022968e-05,
+      "loss": 0.7046,
+      "step": 3526
+    },
+    {
+      "epoch": 0.2595171303747988,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.223933284796787e-05,
+      "loss": 0.974,
+      "step": 3527
+    },
+    {
+      "epoch": 0.2595907105081628,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.223513802050363e-05,
+      "loss": 0.9775,
+      "step": 3528
+    },
+    {
+      "epoch": 0.2596642906415268,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.223094226806207e-05,
+      "loss": 1.0559,
+      "step": 3529
+    },
+    {
+      "epoch": 0.2597378707748908,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.22267455908684e-05,
+      "loss": 0.9653,
+      "step": 3530
+    },
+    {
+      "epoch": 0.25981145090825475,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.222254798914782e-05,
+      "loss": 1.349,
+      "step": 3531
+    },
+    {
+      "epoch": 0.25988503104161875,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.221834946312563e-05,
+      "loss": 0.8148,
+      "step": 3532
+    },
+    {
+      "epoch": 0.25995861117498276,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.221415001302713e-05,
+      "loss": 0.6854,
+      "step": 3533
+    },
+    {
+      "epoch": 0.26003219130834676,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.220994963907772e-05,
+      "loss": 0.6915,
+      "step": 3534
+    },
+    {
+      "epoch": 0.26010577144171076,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.220574834150281e-05,
+      "loss": 1.1557,
+      "step": 3535
+    },
+    {
+      "epoch": 0.2601793515750747,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.22015461205279e-05,
+      "loss": 0.8733,
+      "step": 3536
+    },
+    {
+      "epoch": 0.2602529317084387,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.219734297637849e-05,
+      "loss": 1.2008,
+      "step": 3537
+    },
+    {
+      "epoch": 0.2603265118418027,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.2193138909280163e-05,
+      "loss": 1.0873,
+      "step": 3538
+    },
+    {
+      "epoch": 0.2604000919751667,
+      "grad_norm": 1.921875,
+      "learning_rate": 4.218893391945854e-05,
+      "loss": 0.6881,
+      "step": 3539
+    },
+    {
+      "epoch": 0.2604736721085307,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.218472800713931e-05,
+      "loss": 0.7883,
+      "step": 3540
+    },
+    {
+      "epoch": 0.2605472522418947,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.218052117254817e-05,
+      "loss": 0.8724,
+      "step": 3541
+    },
+    {
+      "epoch": 0.2606208323752587,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.217631341591092e-05,
+      "loss": 0.7856,
+      "step": 3542
+    },
+    {
+      "epoch": 0.2606944125086227,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.2172104737453365e-05,
+      "loss": 0.8145,
+      "step": 3543
+    },
+    {
+      "epoch": 0.2607679926419867,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.216789513740139e-05,
+      "loss": 1.2071,
+      "step": 3544
+    },
+    {
+      "epoch": 0.26084157277535064,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.21636846159809e-05,
+      "loss": 1.4071,
+      "step": 3545
+    },
+    {
+      "epoch": 0.26091515290871464,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.2159473173417875e-05,
+      "loss": 0.6581,
+      "step": 3546
+    },
+    {
+      "epoch": 0.26098873304207865,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.2155260809938335e-05,
+      "loss": 0.8168,
+      "step": 3547
+    },
+    {
+      "epoch": 0.26106231317544265,
+      "grad_norm": 1.234375,
+      "learning_rate": 4.215104752576835e-05,
+      "loss": 0.8958,
+      "step": 3548
+    },
+    {
+      "epoch": 0.2611358933088066,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.214683332113404e-05,
+      "loss": 0.8052,
+      "step": 3549
+    },
+    {
+      "epoch": 0.2612094734421706,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.214261819626156e-05,
+      "loss": 0.6614,
+      "step": 3550
+    },
+    {
+      "epoch": 0.2612830535755346,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.2138402151377156e-05,
+      "loss": 0.9995,
+      "step": 3551
+    },
+    {
+      "epoch": 0.2613566337088986,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.213418518670707e-05,
+      "loss": 0.6394,
+      "step": 3552
+    },
+    {
+      "epoch": 0.2614302138422626,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.212996730247763e-05,
+      "loss": 0.6211,
+      "step": 3553
+    },
+    {
+      "epoch": 0.26150379397562656,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.212574849891519e-05,
+      "loss": 0.799,
+      "step": 3554
+    },
+    {
+      "epoch": 0.26157737410899057,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.212152877624619e-05,
+      "loss": 0.8535,
+      "step": 3555
+    },
+    {
+      "epoch": 0.26165095424235457,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.211730813469706e-05,
+      "loss": 1.134,
+      "step": 3556
+    },
+    {
+      "epoch": 0.2617245343757186,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.211308657449434e-05,
+      "loss": 1.0679,
+      "step": 3557
+    },
+    {
+      "epoch": 0.2617981145090825,
+      "grad_norm": 0.65234375,
+      "learning_rate": 4.2108864095864595e-05,
+      "loss": 0.6499,
+      "step": 3558
+    },
+    {
+      "epoch": 0.26187169464244653,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.210464069903442e-05,
+      "loss": 0.9486,
+      "step": 3559
+    },
+    {
+      "epoch": 0.26194527477581053,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.210041638423049e-05,
+      "loss": 0.9701,
+      "step": 3560
+    },
+    {
+      "epoch": 0.26201885490917454,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.20961911516795e-05,
+      "loss": 1.1911,
+      "step": 3561
+    },
+    {
+      "epoch": 0.26209243504253854,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.209196500160823e-05,
+      "loss": 0.8494,
+      "step": 3562
+    },
+    {
+      "epoch": 0.2621660151759025,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.2087737934243475e-05,
+      "loss": 0.8744,
+      "step": 3563
+    },
+    {
+      "epoch": 0.2622395953092665,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.2083509949812104e-05,
+      "loss": 0.8249,
+      "step": 3564
+    },
+    {
+      "epoch": 0.2623131754426305,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.2079281048541016e-05,
+      "loss": 0.9578,
+      "step": 3565
+    },
+    {
+      "epoch": 0.2623867555759945,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.207505123065717e-05,
+      "loss": 1.1456,
+      "step": 3566
+    },
+    {
+      "epoch": 0.26246033570935845,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.2070820496387584e-05,
+      "loss": 0.8844,
+      "step": 3567
+    },
+    {
+      "epoch": 0.26253391584272245,
+      "grad_norm": 0.8359375,
+      "learning_rate": 4.20665888459593e-05,
+      "loss": 1.0938,
+      "step": 3568
+    },
+    {
+      "epoch": 0.26260749597608646,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.206235627959942e-05,
+      "loss": 0.656,
+      "step": 3569
+    },
+    {
+      "epoch": 0.26268107610945046,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.2058122797535117e-05,
+      "loss": 0.7738,
+      "step": 3570
+    },
+    {
+      "epoch": 0.26275465624281447,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.205388839999358e-05,
+      "loss": 1.1333,
+      "step": 3571
+    },
+    {
+      "epoch": 0.2628282363761784,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.2049653087202045e-05,
+      "loss": 1.2255,
+      "step": 3572
+    },
+    {
+      "epoch": 0.2629018165095424,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.2045416859387844e-05,
+      "loss": 1.0472,
+      "step": 3573
+    },
+    {
+      "epoch": 0.2629753966429064,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.2041179716778324e-05,
+      "loss": 1.0815,
+      "step": 3574
+    },
+    {
+      "epoch": 0.2630489767762704,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.2036941659600856e-05,
+      "loss": 0.9453,
+      "step": 3575
+    },
+    {
+      "epoch": 0.2631225569096344,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.203270268808292e-05,
+      "loss": 0.9314,
+      "step": 3576
+    },
+    {
+      "epoch": 0.2631961370429984,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.202846280245199e-05,
+      "loss": 0.8123,
+      "step": 3577
+    },
+    {
+      "epoch": 0.2632697171763624,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.2024222002935644e-05,
+      "loss": 1.0372,
+      "step": 3578
+    },
+    {
+      "epoch": 0.2633432973097264,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.201998028976144e-05,
+      "loss": 1.0443,
+      "step": 3579
+    },
+    {
+      "epoch": 0.2634168774430904,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.201573766315704e-05,
+      "loss": 0.843,
+      "step": 3580
+    },
+    {
+      "epoch": 0.26349045757645434,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.201149412335015e-05,
+      "loss": 1.1496,
+      "step": 3581
+    },
+    {
+      "epoch": 0.26356403770981834,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.200724967056848e-05,
+      "loss": 0.6851,
+      "step": 3582
+    },
+    {
+      "epoch": 0.26363761784318235,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.200300430503986e-05,
+      "loss": 1.3379,
+      "step": 3583
+    },
+    {
+      "epoch": 0.26371119797654635,
+      "grad_norm": 1.1171875,
+      "learning_rate": 4.199875802699211e-05,
+      "loss": 1.044,
+      "step": 3584
+    },
+    {
+      "epoch": 0.2637847781099103,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.199451083665312e-05,
+      "loss": 1.3473,
+      "step": 3585
+    },
+    {
+      "epoch": 0.2638583582432743,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.199026273425083e-05,
+      "loss": 1.007,
+      "step": 3586
+    },
+    {
+      "epoch": 0.2639319383766383,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.198601372001323e-05,
+      "loss": 0.936,
+      "step": 3587
+    },
+    {
+      "epoch": 0.2640055185100023,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.198176379416836e-05,
+      "loss": 0.9573,
+      "step": 3588
+    },
+    {
+      "epoch": 0.2640790986433663,
+      "grad_norm": 0.6484375,
+      "learning_rate": 4.19775129569443e-05,
+      "loss": 0.6779,
+      "step": 3589
+    },
+    {
+      "epoch": 0.26415267877673027,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.197326120856918e-05,
+      "loss": 0.9555,
+      "step": 3590
+    },
+    {
+      "epoch": 0.26422625891009427,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.196900854927119e-05,
+      "loss": 0.6246,
+      "step": 3591
+    },
+    {
+      "epoch": 0.2642998390434583,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.1964754979278564e-05,
+      "loss": 0.7344,
+      "step": 3592
+    },
+    {
+      "epoch": 0.2643734191768223,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.1960500498819586e-05,
+      "loss": 0.6947,
+      "step": 3593
+    },
+    {
+      "epoch": 0.2644469993101862,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.195624510812257e-05,
+      "loss": 0.9936,
+      "step": 3594
+    },
+    {
+      "epoch": 0.26452057944355023,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.195198880741591e-05,
+      "loss": 0.8097,
+      "step": 3595
+    },
+    {
+      "epoch": 0.26459415957691423,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.194773159692803e-05,
+      "loss": 0.9053,
+      "step": 3596
+    },
+    {
+      "epoch": 0.26466773971027824,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.19434734768874e-05,
+      "loss": 1.2433,
+      "step": 3597
+    },
+    {
+      "epoch": 0.26474131984364224,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.193921444752255e-05,
+      "loss": 1.0103,
+      "step": 3598
+    },
+    {
+      "epoch": 0.2648148999770062,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.1934954509062056e-05,
+      "loss": 0.9681,
+      "step": 3599
+    },
+    {
+      "epoch": 0.2648884801103702,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.1930693661734545e-05,
+      "loss": 0.8749,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2649620602437342,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.1926431905768675e-05,
+      "loss": 1.3861,
+      "step": 3601
+    },
+    {
+      "epoch": 0.2650356403770982,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.1922169241393174e-05,
+      "loss": 1.0495,
+      "step": 3602
+    },
+    {
+      "epoch": 0.26510922051046215,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.191790566883681e-05,
+      "loss": 0.8725,
+      "step": 3603
+    },
+    {
+      "epoch": 0.26518280064382616,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.1913641188328415e-05,
+      "loss": 1.3962,
+      "step": 3604
+    },
+    {
+      "epoch": 0.26525638077719016,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.1909375800096826e-05,
+      "loss": 0.8629,
+      "step": 3605
+    },
+    {
+      "epoch": 0.26532996091055416,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.190510950437099e-05,
+      "loss": 0.7154,
+      "step": 3606
+    },
+    {
+      "epoch": 0.26540354104391817,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.190084230137984e-05,
+      "loss": 0.84,
+      "step": 3607
+    },
+    {
+      "epoch": 0.2654771211772821,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.1896574191352424e-05,
+      "loss": 1.0756,
+      "step": 3608
+    },
+    {
+      "epoch": 0.2655507013106461,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.189230517451777e-05,
+      "loss": 0.8495,
+      "step": 3609
+    },
+    {
+      "epoch": 0.2656242814440101,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.188803525110501e-05,
+      "loss": 1.0097,
+      "step": 3610
+    },
+    {
+      "epoch": 0.26569786157737413,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.188376442134329e-05,
+      "loss": 0.9431,
+      "step": 3611
+    },
+    {
+      "epoch": 0.2657714417107381,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.187949268546182e-05,
+      "loss": 0.7255,
+      "step": 3612
+    },
+    {
+      "epoch": 0.2658450218441021,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.187522004368986e-05,
+      "loss": 0.6103,
+      "step": 3613
+    },
+    {
+      "epoch": 0.2659186019774661,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.1870946496256714e-05,
+      "loss": 0.8542,
+      "step": 3614
+    },
+    {
+      "epoch": 0.2659921821108301,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.186667204339174e-05,
+      "loss": 1.4835,
+      "step": 3615
+    },
+    {
+      "epoch": 0.2660657622441941,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.1862396685324325e-05,
+      "loss": 0.961,
+      "step": 3616
+    },
+    {
+      "epoch": 0.26613934237755804,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.185812042228393e-05,
+      "loss": 0.8702,
+      "step": 3617
+    },
+    {
+      "epoch": 0.26621292251092205,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.185384325450006e-05,
+      "loss": 1.7418,
+      "step": 3618
+    },
+    {
+      "epoch": 0.26628650264428605,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.1849565182202246e-05,
+      "loss": 0.9595,
+      "step": 3619
+    },
+    {
+      "epoch": 0.26636008277765005,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.18452862056201e-05,
+      "loss": 0.7958,
+      "step": 3620
+    },
+    {
+      "epoch": 0.266433662911014,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.1841006324983256e-05,
+      "loss": 0.9107,
+      "step": 3621
+    },
+    {
+      "epoch": 0.266507243044378,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.183672554052142e-05,
+      "loss": 0.8926,
+      "step": 3622
+    },
+    {
+      "epoch": 0.266580823177742,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.1832443852464306e-05,
+      "loss": 1.1984,
+      "step": 3623
+    },
+    {
+      "epoch": 0.266654403311106,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.182816126104173e-05,
+      "loss": 1.0964,
+      "step": 3624
+    },
+    {
+      "epoch": 0.26672798344447,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.1823877766483524e-05,
+      "loss": 1.3117,
+      "step": 3625
+    },
+    {
+      "epoch": 0.26680156357783397,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.181959336901958e-05,
+      "loss": 1.7506,
+      "step": 3626
+    },
+    {
+      "epoch": 0.26687514371119797,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.181530806887982e-05,
+      "loss": 0.7347,
+      "step": 3627
+    },
+    {
+      "epoch": 0.266948723844562,
+      "grad_norm": 1.125,
+      "learning_rate": 4.181102186629424e-05,
+      "loss": 0.8751,
+      "step": 3628
+    },
+    {
+      "epoch": 0.267022303977926,
+      "grad_norm": 1.0,
+      "learning_rate": 4.1806734761492866e-05,
+      "loss": 1.3935,
+      "step": 3629
+    },
+    {
+      "epoch": 0.2670958841112899,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.1802446754705784e-05,
+      "loss": 1.4557,
+      "step": 3630
+    },
+    {
+      "epoch": 0.26716946424465393,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.179815784616311e-05,
+      "loss": 0.8388,
+      "step": 3631
+    },
+    {
+      "epoch": 0.26724304437801794,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.179386803609504e-05,
+      "loss": 0.806,
+      "step": 3632
+    },
+    {
+      "epoch": 0.26731662451138194,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.1789577324731786e-05,
+      "loss": 0.8701,
+      "step": 3633
+    },
+    {
+      "epoch": 0.26739020464474594,
+      "grad_norm": 1.1640625,
+      "learning_rate": 4.1785285712303636e-05,
+      "loss": 1.1736,
+      "step": 3634
+    },
+    {
+      "epoch": 0.2674637847781099,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.178099319904091e-05,
+      "loss": 0.6696,
+      "step": 3635
+    },
+    {
+      "epoch": 0.2675373649114739,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.1776699785173955e-05,
+      "loss": 1.1624,
+      "step": 3636
+    },
+    {
+      "epoch": 0.2676109450448379,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.177240547093322e-05,
+      "loss": 1.2793,
+      "step": 3637
+    },
+    {
+      "epoch": 0.2676845251782019,
+      "grad_norm": 1.0,
+      "learning_rate": 4.176811025654917e-05,
+      "loss": 1.1772,
+      "step": 3638
+    },
+    {
+      "epoch": 0.26775810531156585,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.176381414225229e-05,
+      "loss": 0.9055,
+      "step": 3639
+    },
+    {
+      "epoch": 0.26783168544492986,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.1759517128273184e-05,
+      "loss": 0.8579,
+      "step": 3640
+    },
+    {
+      "epoch": 0.26790526557829386,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.175521921484245e-05,
+      "loss": 0.6146,
+      "step": 3641
+    },
+    {
+      "epoch": 0.26797884571165786,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.175092040219074e-05,
+      "loss": 0.6374,
+      "step": 3642
+    },
+    {
+      "epoch": 0.26805242584502187,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.174662069054877e-05,
+      "loss": 1.7015,
+      "step": 3643
+    },
+    {
+      "epoch": 0.2681260059783858,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.1742320080147304e-05,
+      "loss": 0.7913,
+      "step": 3644
+    },
+    {
+      "epoch": 0.2681995861117498,
+      "grad_norm": 0.96875,
+      "learning_rate": 4.173801857121713e-05,
+      "loss": 0.8472,
+      "step": 3645
+    },
+    {
+      "epoch": 0.2682731662451138,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.173371616398912e-05,
+      "loss": 0.9246,
+      "step": 3646
+    },
+    {
+      "epoch": 0.26834674637847783,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.172941285869417e-05,
+      "loss": 1.0593,
+      "step": 3647
+    },
+    {
+      "epoch": 0.2684203265118418,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.172510865556323e-05,
+      "loss": 0.7386,
+      "step": 3648
+    },
+    {
+      "epoch": 0.2684939066452058,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.172080355482729e-05,
+      "loss": 1.0615,
+      "step": 3649
+    },
+    {
+      "epoch": 0.2685674867785698,
+      "grad_norm": 0.7265625,
+      "learning_rate": 4.171649755671741e-05,
+      "loss": 0.6225,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2686410669119338,
+      "grad_norm": 0.81640625,
+      "learning_rate": 4.171219066146468e-05,
+      "loss": 0.8728,
+      "step": 3651
+    },
+    {
+      "epoch": 0.2687146470452978,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.170788286930024e-05,
+      "loss": 0.8478,
+      "step": 3652
+    },
+    {
+      "epoch": 0.26878822717866174,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.170357418045529e-05,
+      "loss": 0.8094,
+      "step": 3653
+    },
+    {
+      "epoch": 0.26886180731202575,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.169926459516105e-05,
+      "loss": 0.8884,
+      "step": 3654
+    },
+    {
+      "epoch": 0.26893538744538975,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.1694954113648823e-05,
+      "loss": 0.6956,
+      "step": 3655
+    },
+    {
+      "epoch": 0.26900896757875375,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.169064273614995e-05,
+      "loss": 0.5339,
+      "step": 3656
+    },
+    {
+      "epoch": 0.2690825477121177,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.16863304628958e-05,
+      "loss": 1.36,
+      "step": 3657
+    },
+    {
+      "epoch": 0.2691561278454817,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.16820172941178e-05,
+      "loss": 1.2954,
+      "step": 3658
+    },
+    {
+      "epoch": 0.2692297079788457,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.167770323004746e-05,
+      "loss": 1.2144,
+      "step": 3659
+    },
+    {
+      "epoch": 0.2693032881122097,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.167338827091627e-05,
+      "loss": 1.0776,
+      "step": 3660
+    },
+    {
+      "epoch": 0.2693768682455737,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.166907241695583e-05,
+      "loss": 1.0608,
+      "step": 3661
+    },
+    {
+      "epoch": 0.26945044837893767,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.1664755668397756e-05,
+      "loss": 1.0712,
+      "step": 3662
+    },
+    {
+      "epoch": 0.26952402851230167,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.1660438025473725e-05,
+      "loss": 0.9114,
+      "step": 3663
+    },
+    {
+      "epoch": 0.2695976086456657,
+      "grad_norm": 0.9765625,
+      "learning_rate": 4.165611948841545e-05,
+      "loss": 0.8202,
+      "step": 3664
+    },
+    {
+      "epoch": 0.2696711887790297,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.165180005745469e-05,
+      "loss": 0.7816,
+      "step": 3665
+    },
+    {
+      "epoch": 0.26974476891239363,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.164747973282328e-05,
+      "loss": 1.0676,
+      "step": 3666
+    },
+    {
+      "epoch": 0.26981834904575763,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.1643158514753076e-05,
+      "loss": 1.0146,
+      "step": 3667
+    },
+    {
+      "epoch": 0.26989192917912164,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.163883640347599e-05,
+      "loss": 0.9597,
+      "step": 3668
+    },
+    {
+      "epoch": 0.26996550931248564,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.163451339922399e-05,
+      "loss": 1.0143,
+      "step": 3669
+    },
+    {
+      "epoch": 0.27003908944584964,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.163018950222905e-05,
+      "loss": 1.0431,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2701126695792136,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.162586471272327e-05,
+      "loss": 0.7312,
+      "step": 3671
+    },
+    {
+      "epoch": 0.2701862497125776,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.162153903093873e-05,
+      "loss": 1.2639,
+      "step": 3672
+    },
+    {
+      "epoch": 0.2702598298459416,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.1617212457107576e-05,
+      "loss": 0.7579,
+      "step": 3673
+    },
+    {
+      "epoch": 0.2703334099793056,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.161288499146202e-05,
+      "loss": 0.9605,
+      "step": 3674
+    },
+    {
+      "epoch": 0.27040699011266955,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.1608556634234296e-05,
+      "loss": 0.8284,
+      "step": 3675
+    },
+    {
+      "epoch": 0.27048057024603356,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.1604227385656714e-05,
+      "loss": 0.9931,
+      "step": 3676
+    },
+    {
+      "epoch": 0.27055415037939756,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.15998972459616e-05,
+      "loss": 0.9074,
+      "step": 3677
+    },
+    {
+      "epoch": 0.27062773051276157,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.159556621538137e-05,
+      "loss": 0.9218,
+      "step": 3678
+    },
+    {
+      "epoch": 0.27070131064612557,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.159123429414844e-05,
+      "loss": 0.8277,
+      "step": 3679
+    },
+    {
+      "epoch": 0.2707748907794895,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.158690148249529e-05,
+      "loss": 0.5017,
+      "step": 3680
+    },
+    {
+      "epoch": 0.2708484709128535,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.158256778065448e-05,
+      "loss": 1.0813,
+      "step": 3681
+    },
+    {
+      "epoch": 0.2709220510462175,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.157823318885857e-05,
+      "loss": 1.0314,
+      "step": 3682
+    },
+    {
+      "epoch": 0.27099563117958153,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.15738977073402e-05,
+      "loss": 0.676,
+      "step": 3683
+    },
+    {
+      "epoch": 0.2710692113129455,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.156956133633204e-05,
+      "loss": 0.9024,
+      "step": 3684
+    },
+    {
+      "epoch": 0.2711427914463095,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.1565224076066824e-05,
+      "loss": 1.1291,
+      "step": 3685
+    },
+    {
+      "epoch": 0.2712163715796735,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.1560885926777314e-05,
+      "loss": 0.7727,
+      "step": 3686
+    },
+    {
+      "epoch": 0.2712899517130375,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.1556546888696336e-05,
+      "loss": 1.0155,
+      "step": 3687
+    },
+    {
+      "epoch": 0.2713635318464015,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.1552206962056764e-05,
+      "loss": 1.2204,
+      "step": 3688
+    },
+    {
+      "epoch": 0.27143711197976544,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.1547866147091496e-05,
+      "loss": 0.7764,
+      "step": 3689
+    },
+    {
+      "epoch": 0.27151069211312945,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.154352444403351e-05,
+      "loss": 0.7791,
+      "step": 3690
+    },
+    {
+      "epoch": 0.27158427224649345,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.1539181853115814e-05,
+      "loss": 0.8875,
+      "step": 3691
+    },
+    {
+      "epoch": 0.27165785237985746,
+      "grad_norm": 0.875,
+      "learning_rate": 4.153483837457147e-05,
+      "loss": 1.0873,
+      "step": 3692
+    },
+    {
+      "epoch": 0.2717314325132214,
+      "grad_norm": 1.0,
+      "learning_rate": 4.153049400863358e-05,
+      "loss": 1.1457,
+      "step": 3693
+    },
+    {
+      "epoch": 0.2718050126465854,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.1526148755535296e-05,
+      "loss": 0.8534,
+      "step": 3694
+    },
+    {
+      "epoch": 0.2718785927799494,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.152180261550982e-05,
+      "loss": 0.8841,
+      "step": 3695
+    },
+    {
+      "epoch": 0.2719521729133134,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.1517455588790397e-05,
+      "loss": 0.9285,
+      "step": 3696
+    },
+    {
+      "epoch": 0.2720257530466774,
+      "grad_norm": 0.875,
+      "learning_rate": 4.151310767561034e-05,
+      "loss": 0.7892,
+      "step": 3697
+    },
+    {
+      "epoch": 0.27209933318004137,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.150875887620298e-05,
+      "loss": 0.9136,
+      "step": 3698
+    },
+    {
+      "epoch": 0.2721729133134054,
+      "grad_norm": 1.1875,
+      "learning_rate": 4.150440919080172e-05,
+      "loss": 1.0957,
+      "step": 3699
+    },
+    {
+      "epoch": 0.2722464934467694,
+      "grad_norm": 0.75,
+      "learning_rate": 4.1500058619639984e-05,
+      "loss": 0.8967,
+      "step": 3700
+    },
+    {
+      "epoch": 0.2723200735801334,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.149570716295126e-05,
+      "loss": 1.0409,
+      "step": 3701
+    },
+    {
+      "epoch": 0.27239365371349733,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.1491354820969094e-05,
+      "loss": 1.4121,
+      "step": 3702
+    },
+    {
+      "epoch": 0.27246723384686133,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.1487001593927055e-05,
+      "loss": 0.7186,
+      "step": 3703
+    },
+    {
+      "epoch": 0.27254081398022534,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.1482647482058787e-05,
+      "loss": 1.3317,
+      "step": 3704
+    },
+    {
+      "epoch": 0.27261439411358934,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.147829248559796e-05,
+      "loss": 0.6391,
+      "step": 3705
+    },
+    {
+      "epoch": 0.27268797424695335,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.147393660477831e-05,
+      "loss": 1.6144,
+      "step": 3706
+    },
+    {
+      "epoch": 0.2727615543803173,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.1469579839833577e-05,
+      "loss": 1.0032,
+      "step": 3707
+    },
+    {
+      "epoch": 0.2728351345136813,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.146522219099761e-05,
+      "loss": 0.8316,
+      "step": 3708
+    },
+    {
+      "epoch": 0.2729087146470453,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.146086365850427e-05,
+      "loss": 1.172,
+      "step": 3709
+    },
+    {
+      "epoch": 0.2729822947804093,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.1456504242587454e-05,
+      "loss": 1.3697,
+      "step": 3710
+    },
+    {
+      "epoch": 0.27305587491377326,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.145214394348115e-05,
+      "loss": 1.021,
+      "step": 3711
+    },
+    {
+      "epoch": 0.27312945504713726,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.1447782761419354e-05,
+      "loss": 1.0258,
+      "step": 3712
+    },
+    {
+      "epoch": 0.27320303518050126,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.144342069663611e-05,
+      "loss": 0.861,
+      "step": 3713
+    },
+    {
+      "epoch": 0.27327661531386527,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.143905774936555e-05,
+      "loss": 0.8782,
+      "step": 3714
+    },
+    {
+      "epoch": 0.27335019544722927,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.1434693919841805e-05,
+      "loss": 1.148,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2734237755805932,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.143032920829907e-05,
+      "loss": 1.0507,
+      "step": 3716
+    },
+    {
+      "epoch": 0.2734973557139572,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.142596361497161e-05,
+      "loss": 0.9985,
+      "step": 3717
+    },
+    {
+      "epoch": 0.27357093584732123,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.14215971400937e-05,
+      "loss": 1.3125,
+      "step": 3718
+    },
+    {
+      "epoch": 0.27364451598068523,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.141722978389969e-05,
+      "loss": 1.0741,
+      "step": 3719
+    },
+    {
+      "epoch": 0.2737180961140492,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.141286154662397e-05,
+      "loss": 1.418,
+      "step": 3720
+    },
+    {
+      "epoch": 0.2737916762474132,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.140849242850096e-05,
+      "loss": 0.7892,
+      "step": 3721
+    },
+    {
+      "epoch": 0.2738652563807772,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.140412242976516e-05,
+      "loss": 1.0057,
+      "step": 3722
+    },
+    {
+      "epoch": 0.2739388365141412,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.1399751550651087e-05,
+      "loss": 0.7086,
+      "step": 3723
+    },
+    {
+      "epoch": 0.2740124166475052,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.139537979139333e-05,
+      "loss": 0.9284,
+      "step": 3724
+    },
+    {
+      "epoch": 0.27408599678086915,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.1391007152226494e-05,
+      "loss": 0.8726,
+      "step": 3725
+    },
+    {
+      "epoch": 0.27415957691423315,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.1386633633385276e-05,
+      "loss": 0.9388,
+      "step": 3726
+    },
+    {
+      "epoch": 0.27423315704759715,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.138225923510437e-05,
+      "loss": 1.3097,
+      "step": 3727
+    },
+    {
+      "epoch": 0.27430673718096116,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.1377883957618556e-05,
+      "loss": 0.6689,
+      "step": 3728
+    },
+    {
+      "epoch": 0.2743803173143251,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.137350780116265e-05,
+      "loss": 1.0086,
+      "step": 3729
+    },
+    {
+      "epoch": 0.2744538974476891,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.13691307659715e-05,
+      "loss": 1.1734,
+      "step": 3730
+    },
+    {
+      "epoch": 0.2745274775810531,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.136475285228002e-05,
+      "loss": 0.7806,
+      "step": 3731
+    },
+    {
+      "epoch": 0.2746010577144171,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.136037406032316e-05,
+      "loss": 0.6985,
+      "step": 3732
+    },
+    {
+      "epoch": 0.2746746378477811,
+      "grad_norm": 0.87109375,
+      "learning_rate": 4.135599439033593e-05,
+      "loss": 1.0879,
+      "step": 3733
+    },
+    {
+      "epoch": 0.27474821798114507,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.135161384255337e-05,
+      "loss": 0.9441,
+      "step": 3734
+    },
+    {
+      "epoch": 0.2748217981145091,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.1347232417210575e-05,
+      "loss": 1.4543,
+      "step": 3735
+    },
+    {
+      "epoch": 0.2748953782478731,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.13428501145427e-05,
+      "loss": 0.8849,
+      "step": 3736
+    },
+    {
+      "epoch": 0.2749689583812371,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.133846693478492e-05,
+      "loss": 0.9882,
+      "step": 3737
+    },
+    {
+      "epoch": 0.27504253851460103,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.133408287817248e-05,
+      "loss": 0.9064,
+      "step": 3738
+    },
+    {
+      "epoch": 0.27511611864796504,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.132969794494066e-05,
+      "loss": 0.6894,
+      "step": 3739
+    },
+    {
+      "epoch": 0.27518969878132904,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.13253121353248e-05,
+      "loss": 0.8499,
+      "step": 3740
+    },
+    {
+      "epoch": 0.27526327891469304,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.1320925449560265e-05,
+      "loss": 0.7634,
+      "step": 3741
+    },
+    {
+      "epoch": 0.27533685904805705,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.131653788788249e-05,
+      "loss": 1.2907,
+      "step": 3742
+    },
+    {
+      "epoch": 0.275410439181421,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.1312149450526946e-05,
+      "loss": 1.1168,
+      "step": 3743
+    },
+    {
+      "epoch": 0.275484019314785,
+      "grad_norm": 0.97265625,
+      "learning_rate": 4.130776013772916e-05,
+      "loss": 1.072,
+      "step": 3744
+    },
+    {
+      "epoch": 0.275557599448149,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.130336994972467e-05,
+      "loss": 1.2104,
+      "step": 3745
+    },
+    {
+      "epoch": 0.275631179581513,
+      "grad_norm": 0.9453125,
+      "learning_rate": 4.1298978886749115e-05,
+      "loss": 0.7803,
+      "step": 3746
+    },
+    {
+      "epoch": 0.27570475971487696,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.129458694903815e-05,
+      "loss": 1.5481,
+      "step": 3747
+    },
+    {
+      "epoch": 0.27577833984824096,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.1290194136827476e-05,
+      "loss": 0.9913,
+      "step": 3748
+    },
+    {
+      "epoch": 0.27585191998160496,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.128580045035285e-05,
+      "loss": 1.2736,
+      "step": 3749
+    },
+    {
+      "epoch": 0.27592550011496897,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.128140588985008e-05,
+      "loss": 1.0355,
+      "step": 3750
+    },
+    {
+      "epoch": 0.275999080248333,
+      "grad_norm": 1.421875,
+      "learning_rate": 4.1277010455555e-05,
+      "loss": 1.2461,
+      "step": 3751
+    },
+    {
+      "epoch": 0.2760726603816969,
+      "grad_norm": 0.9296875,
+      "learning_rate": 4.127261414770351e-05,
+      "loss": 0.7185,
+      "step": 3752
+    },
+    {
+      "epoch": 0.2761462405150609,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.126821696653156e-05,
+      "loss": 0.9304,
+      "step": 3753
+    },
+    {
+      "epoch": 0.27621982064842493,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.126381891227513e-05,
+      "loss": 0.9105,
+      "step": 3754
+    },
+    {
+      "epoch": 0.27629340078178893,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.125941998517026e-05,
+      "loss": 0.9678,
+      "step": 3755
+    },
+    {
+      "epoch": 0.2763669809151529,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.125502018545302e-05,
+      "loss": 1.1382,
+      "step": 3756
+    },
+    {
+      "epoch": 0.2764405610485169,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.125061951335956e-05,
+      "loss": 0.811,
+      "step": 3757
+    },
+    {
+      "epoch": 0.2765141411818809,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.1246217969126036e-05,
+      "loss": 1.2773,
+      "step": 3758
+    },
+    {
+      "epoch": 0.2765877213152449,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.1241815552988675e-05,
+      "loss": 1.1922,
+      "step": 3759
+    },
+    {
+      "epoch": 0.2766613014486089,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.123741226518375e-05,
+      "loss": 0.8484,
+      "step": 3760
+    },
+    {
+      "epoch": 0.27673488158197285,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.1233008105947584e-05,
+      "loss": 1.1286,
+      "step": 3761
+    },
+    {
+      "epoch": 0.27680846171533685,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.1228603075516526e-05,
+      "loss": 0.9195,
+      "step": 3762
+    },
+    {
+      "epoch": 0.27688204184870086,
+      "grad_norm": 1.2109375,
+      "learning_rate": 4.1224197174127e-05,
+      "loss": 0.9709,
+      "step": 3763
+    },
+    {
+      "epoch": 0.27695562198206486,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.1219790402015444e-05,
+      "loss": 1.0277,
+      "step": 3764
+    },
+    {
+      "epoch": 0.2770292021154288,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.121538275941839e-05,
+      "loss": 0.7648,
+      "step": 3765
+    },
+    {
+      "epoch": 0.2771027822487928,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.121097424657235e-05,
+      "loss": 0.6587,
+      "step": 3766
+    },
+    {
+      "epoch": 0.2771763623821568,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.120656486371395e-05,
+      "loss": 0.8048,
+      "step": 3767
+    },
+    {
+      "epoch": 0.2772499425155208,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.120215461107982e-05,
+      "loss": 0.8197,
+      "step": 3768
+    },
+    {
+      "epoch": 0.2773235226488848,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.1197743488906656e-05,
+      "loss": 0.5523,
+      "step": 3769
+    },
+    {
+      "epoch": 0.2773971027822488,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.11933314974312e-05,
+      "loss": 0.8875,
+      "step": 3770
+    },
+    {
+      "epoch": 0.2774706829156128,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.1188918636890216e-05,
+      "loss": 0.9472,
+      "step": 3771
+    },
+    {
+      "epoch": 0.2775442630489768,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.118450490752055e-05,
+      "loss": 0.6643,
+      "step": 3772
+    },
+    {
+      "epoch": 0.2776178431823408,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.1180090309559075e-05,
+      "loss": 1.3515,
+      "step": 3773
+    },
+    {
+      "epoch": 0.27769142331570473,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.117567484324271e-05,
+      "loss": 1.155,
+      "step": 3774
+    },
+    {
+      "epoch": 0.27776500344906874,
+      "grad_norm": 0.921875,
+      "learning_rate": 4.117125850880842e-05,
+      "loss": 0.8974,
+      "step": 3775
+    },
+    {
+      "epoch": 0.27783858358243274,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.116684130649324e-05,
+      "loss": 0.8932,
+      "step": 3776
+    },
+    {
+      "epoch": 0.27791216371579675,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.116242323653422e-05,
+      "loss": 1.1929,
+      "step": 3777
+    },
+    {
+      "epoch": 0.27798574384916075,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.1158004299168465e-05,
+      "loss": 0.9176,
+      "step": 3778
+    },
+    {
+      "epoch": 0.2780593239825247,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.1153584494633144e-05,
+      "loss": 0.672,
+      "step": 3779
+    },
+    {
+      "epoch": 0.2781329041158887,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.114916382316546e-05,
+      "loss": 1.2671,
+      "step": 3780
+    },
+    {
+      "epoch": 0.2782064842492527,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.114474228500264e-05,
+      "loss": 1.0137,
+      "step": 3781
+    },
+    {
+      "epoch": 0.2782800643826167,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.1140319880382e-05,
+      "loss": 0.913,
+      "step": 3782
+    },
+    {
+      "epoch": 0.27835364451598066,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.113589660954088e-05,
+      "loss": 0.8609,
+      "step": 3783
+    },
+    {
+      "epoch": 0.27842722464934466,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.113147247271667e-05,
+      "loss": 0.8475,
+      "step": 3784
+    },
+    {
+      "epoch": 0.27850080478270867,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.1127047470146786e-05,
+      "loss": 1.2032,
+      "step": 3785
+    },
+    {
+      "epoch": 0.27857438491607267,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.112262160206873e-05,
+      "loss": 1.2808,
+      "step": 3786
+    },
+    {
+      "epoch": 0.2786479650494367,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.1118194868720025e-05,
+      "loss": 0.7413,
+      "step": 3787
+    },
+    {
+      "epoch": 0.2787215451828006,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.111376727033825e-05,
+      "loss": 0.7744,
+      "step": 3788
+    },
+    {
+      "epoch": 0.2787951253161646,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.110933880716101e-05,
+      "loss": 1.0279,
+      "step": 3789
+    },
+    {
+      "epoch": 0.27886870544952863,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.110490947942599e-05,
+      "loss": 0.8174,
+      "step": 3790
+    },
+    {
+      "epoch": 0.27894228558289264,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.1100479287370896e-05,
+      "loss": 0.9783,
+      "step": 3791
+    },
+    {
+      "epoch": 0.2790158657162566,
+      "grad_norm": 0.91015625,
+      "learning_rate": 4.109604823123349e-05,
+      "loss": 1.1988,
+      "step": 3792
+    },
+    {
+      "epoch": 0.2790894458496206,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.109161631125157e-05,
+      "loss": 1.1926,
+      "step": 3793
+    },
+    {
+      "epoch": 0.2791630259829846,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.1087183527663e-05,
+      "loss": 0.7411,
+      "step": 3794
+    },
+    {
+      "epoch": 0.2792366061163486,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.108274988070567e-05,
+      "loss": 0.967,
+      "step": 3795
+    },
+    {
+      "epoch": 0.2793101862497126,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.107831537061753e-05,
+      "loss": 1.1322,
+      "step": 3796
+    },
+    {
+      "epoch": 0.27938376638307655,
+      "grad_norm": 1.078125,
+      "learning_rate": 4.107387999763658e-05,
+      "loss": 0.9204,
+      "step": 3797
+    },
+    {
+      "epoch": 0.27945734651644055,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.1069443762000835e-05,
+      "loss": 0.7403,
+      "step": 3798
+    },
+    {
+      "epoch": 0.27953092664980456,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.1065006663948406e-05,
+      "loss": 0.9051,
+      "step": 3799
+    },
+    {
+      "epoch": 0.27960450678316856,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.106056870371741e-05,
+      "loss": 1.509,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2796780869165325,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.1056129881546024e-05,
+      "loss": 0.878,
+      "step": 3801
+    },
+    {
+      "epoch": 0.2797516670498965,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.105169019767248e-05,
+      "loss": 0.8086,
+      "step": 3802
+    },
+    {
+      "epoch": 0.2798252471832605,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.104724965233503e-05,
+      "loss": 0.8789,
+      "step": 3803
+    },
+    {
+      "epoch": 0.2798988273166245,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.1042808245772005e-05,
+      "loss": 1.0014,
+      "step": 3804
+    },
+    {
+      "epoch": 0.2799724074499885,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.103836597822176e-05,
+      "loss": 0.8212,
+      "step": 3805
+    },
+    {
+      "epoch": 0.2800459875833525,
+      "grad_norm": 1.109375,
+      "learning_rate": 4.1033922849922706e-05,
+      "loss": 1.792,
+      "step": 3806
+    },
+    {
+      "epoch": 0.2801195677167165,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.1029478861113295e-05,
+      "loss": 0.8233,
+      "step": 3807
+    },
+    {
+      "epoch": 0.2801931478500805,
+      "grad_norm": 0.66015625,
+      "learning_rate": 4.102503401203203e-05,
+      "loss": 0.8307,
+      "step": 3808
+    },
+    {
+      "epoch": 0.2802667279834445,
+      "grad_norm": 0.62890625,
+      "learning_rate": 4.102058830291746e-05,
+      "loss": 0.7511,
+      "step": 3809
+    },
+    {
+      "epoch": 0.28034030811680843,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.1016141734008165e-05,
+      "loss": 1.0038,
+      "step": 3810
+    },
+    {
+      "epoch": 0.28041388825017244,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.1011694305542805e-05,
+      "loss": 0.8683,
+      "step": 3811
+    },
+    {
+      "epoch": 0.28048746838353644,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.1007246017760047e-05,
+      "loss": 1.0173,
+      "step": 3812
+    },
+    {
+      "epoch": 0.28056104851690045,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.100279687089863e-05,
+      "loss": 1.0265,
+      "step": 3813
+    },
+    {
+      "epoch": 0.28063462865026445,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.099834686519733e-05,
+      "loss": 0.9441,
+      "step": 3814
+    },
+    {
+      "epoch": 0.2807082087836284,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.099389600089497e-05,
+      "loss": 0.7323,
+      "step": 3815
+    },
+    {
+      "epoch": 0.2807817889169924,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.098944427823041e-05,
+      "loss": 1.2657,
+      "step": 3816
+    },
+    {
+      "epoch": 0.2808553690503564,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.0984991697442596e-05,
+      "loss": 0.7879,
+      "step": 3817
+    },
+    {
+      "epoch": 0.2809289491837204,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.098053825877046e-05,
+      "loss": 0.7809,
+      "step": 3818
+    },
+    {
+      "epoch": 0.28100252931708436,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.097608396245301e-05,
+      "loss": 0.8438,
+      "step": 3819
+    },
+    {
+      "epoch": 0.28107610945044836,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.097162880872932e-05,
+      "loss": 0.6625,
+      "step": 3820
+    },
+    {
+      "epoch": 0.28114968958381237,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.096717279783847e-05,
+      "loss": 0.8132,
+      "step": 3821
+    },
+    {
+      "epoch": 0.28122326971717637,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.096271593001962e-05,
+      "loss": 1.2908,
+      "step": 3822
+    },
+    {
+      "epoch": 0.2812968498505404,
+      "grad_norm": 1.3359375,
+      "learning_rate": 4.095825820551195e-05,
+      "loss": 1.0392,
+      "step": 3823
+    },
+    {
+      "epoch": 0.2813704299839043,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.09537996245547e-05,
+      "loss": 0.8572,
+      "step": 3824
+    },
+    {
+      "epoch": 0.28144401011726833,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.094934018738716e-05,
+      "loss": 0.856,
+      "step": 3825
+    },
+    {
+      "epoch": 0.28151759025063233,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.094487989424866e-05,
+      "loss": 0.7391,
+      "step": 3826
+    },
+    {
+      "epoch": 0.28159117038399634,
+      "grad_norm": 1.0,
+      "learning_rate": 4.094041874537857e-05,
+      "loss": 1.0149,
+      "step": 3827
+    },
+    {
+      "epoch": 0.2816647505173603,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.0935956741016313e-05,
+      "loss": 1.3908,
+      "step": 3828
+    },
+    {
+      "epoch": 0.2817383306507243,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.0931493881401364e-05,
+      "loss": 0.9702,
+      "step": 3829
+    },
+    {
+      "epoch": 0.2818119107840883,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.0927030166773217e-05,
+      "loss": 1.2533,
+      "step": 3830
+    },
+    {
+      "epoch": 0.2818854909174523,
+      "grad_norm": 1.0,
+      "learning_rate": 4.092256559737144e-05,
+      "loss": 0.9352,
+      "step": 3831
+    },
+    {
+      "epoch": 0.2819590710508163,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.091810017343565e-05,
+      "loss": 0.9035,
+      "step": 3832
+    },
+    {
+      "epoch": 0.28203265118418025,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.0913633895205484e-05,
+      "loss": 0.8236,
+      "step": 3833
+    },
+    {
+      "epoch": 0.28210623131754425,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.090916676292065e-05,
+      "loss": 0.9557,
+      "step": 3834
+    },
+    {
+      "epoch": 0.28217981145090826,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.0904698776820885e-05,
+      "loss": 0.9953,
+      "step": 3835
+    },
+    {
+      "epoch": 0.28225339158427226,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.090022993714596e-05,
+      "loss": 1.0776,
+      "step": 3836
+    },
+    {
+      "epoch": 0.28232697171763627,
+      "grad_norm": 1.0703125,
+      "learning_rate": 4.089576024413574e-05,
+      "loss": 1.0095,
+      "step": 3837
+    },
+    {
+      "epoch": 0.2824005518510002,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.089128969803009e-05,
+      "loss": 0.8806,
+      "step": 3838
+    },
+    {
+      "epoch": 0.2824741319843642,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.088681829906893e-05,
+      "loss": 0.944,
+      "step": 3839
+    },
+    {
+      "epoch": 0.2825477121177282,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.0882346047492235e-05,
+      "loss": 0.8527,
+      "step": 3840
+    },
+    {
+      "epoch": 0.2826212922510922,
+      "grad_norm": 0.875,
+      "learning_rate": 4.087787294354004e-05,
+      "loss": 1.0146,
+      "step": 3841
+    },
+    {
+      "epoch": 0.2826948723844562,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.0873398987452384e-05,
+      "loss": 1.078,
+      "step": 3842
+    },
+    {
+      "epoch": 0.2827684525178202,
+      "grad_norm": 0.7578125,
+      "learning_rate": 4.086892417946938e-05,
+      "loss": 0.8147,
+      "step": 3843
+    },
+    {
+      "epoch": 0.2828420326511842,
+      "grad_norm": 1.0,
+      "learning_rate": 4.086444851983119e-05,
+      "loss": 1.2531,
+      "step": 3844
+    },
+    {
+      "epoch": 0.2829156127845482,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.085997200877803e-05,
+      "loss": 1.0398,
+      "step": 3845
+    },
+    {
+      "epoch": 0.2829891929179122,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.0855494646550094e-05,
+      "loss": 0.8554,
+      "step": 3846
+    },
+    {
+      "epoch": 0.28306277305127614,
+      "grad_norm": 0.90625,
+      "learning_rate": 4.085101643338774e-05,
+      "loss": 0.8474,
+      "step": 3847
+    },
+    {
+      "epoch": 0.28313635318464014,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.084653736953125e-05,
+      "loss": 0.7307,
+      "step": 3848
+    },
+    {
+      "epoch": 0.28320993331800415,
+      "grad_norm": 1.1015625,
+      "learning_rate": 4.084205745522104e-05,
+      "loss": 1.3031,
+      "step": 3849
+    },
+    {
+      "epoch": 0.28328351345136815,
+      "grad_norm": 0.9140625,
+      "learning_rate": 4.0837576690697524e-05,
+      "loss": 1.1648,
+      "step": 3850
+    },
+    {
+      "epoch": 0.2833570935847321,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.083309507620118e-05,
+      "loss": 0.7618,
+      "step": 3851
+    },
+    {
+      "epoch": 0.2834306737180961,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.0828612611972526e-05,
+      "loss": 0.686,
+      "step": 3852
+    },
+    {
+      "epoch": 0.2835042538514601,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.0824129298252126e-05,
+      "loss": 0.6035,
+      "step": 3853
+    },
+    {
+      "epoch": 0.2835778339848241,
+      "grad_norm": 0.88671875,
+      "learning_rate": 4.08196451352806e-05,
+      "loss": 1.3134,
+      "step": 3854
+    },
+    {
+      "epoch": 0.2836514141181881,
+      "grad_norm": 0.6875,
+      "learning_rate": 4.08151601232986e-05,
+      "loss": 0.8152,
+      "step": 3855
+    },
+    {
+      "epoch": 0.28372499425155207,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.081067426254682e-05,
+      "loss": 1.1926,
+      "step": 3856
+    },
+    {
+      "epoch": 0.28379857438491607,
+      "grad_norm": 0.875,
+      "learning_rate": 4.0806187553266024e-05,
+      "loss": 1.1827,
+      "step": 3857
+    },
+    {
+      "epoch": 0.2838721545182801,
+      "grad_norm": 1.1796875,
+      "learning_rate": 4.080169999569699e-05,
+      "loss": 0.921,
+      "step": 3858
+    },
+    {
+      "epoch": 0.2839457346516441,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.079721159008056e-05,
+      "loss": 0.6629,
+      "step": 3859
+    },
+    {
+      "epoch": 0.284019314785008,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.079272233665763e-05,
+      "loss": 0.7299,
+      "step": 3860
+    },
+    {
+      "epoch": 0.28409289491837203,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.078823223566911e-05,
+      "loss": 0.9084,
+      "step": 3861
+    },
+    {
+      "epoch": 0.28416647505173603,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.0783741287355994e-05,
+      "loss": 1.5359,
+      "step": 3862
+    },
+    {
+      "epoch": 0.28424005518510004,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.077924949195929e-05,
+      "loss": 0.8932,
+      "step": 3863
+    },
+    {
+      "epoch": 0.28431363531846404,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.077475684972006e-05,
+      "loss": 0.8656,
+      "step": 3864
+    },
+    {
+      "epoch": 0.284387215451828,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.0770263360879435e-05,
+      "loss": 0.7003,
+      "step": 3865
+    },
+    {
+      "epoch": 0.284460795585192,
+      "grad_norm": 0.828125,
+      "learning_rate": 4.076576902567856e-05,
+      "loss": 1.0429,
+      "step": 3866
+    },
+    {
+      "epoch": 0.284534375718556,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.076127384435864e-05,
+      "loss": 1.1955,
+      "step": 3867
+    },
+    {
+      "epoch": 0.28460795585192,
+      "grad_norm": 0.75,
+      "learning_rate": 4.075677781716092e-05,
+      "loss": 0.8434,
+      "step": 3868
+    },
+    {
+      "epoch": 0.28468153598528395,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.0752280944326694e-05,
+      "loss": 0.8366,
+      "step": 3869
+    },
+    {
+      "epoch": 0.28475511611864796,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.07477832260973e-05,
+      "loss": 0.9159,
+      "step": 3870
+    },
+    {
+      "epoch": 0.28482869625201196,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.074328466271413e-05,
+      "loss": 1.3194,
+      "step": 3871
+    },
+    {
+      "epoch": 0.28490227638537596,
+      "grad_norm": 0.8828125,
+      "learning_rate": 4.07387852544186e-05,
+      "loss": 1.0448,
+      "step": 3872
+    },
+    {
+      "epoch": 0.28497585651873997,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.0734285001452194e-05,
+      "loss": 1.4761,
+      "step": 3873
+    },
+    {
+      "epoch": 0.2850494366521039,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.072978390405643e-05,
+      "loss": 0.9652,
+      "step": 3874
+    },
+    {
+      "epoch": 0.2851230167854679,
+      "grad_norm": 1.203125,
+      "learning_rate": 4.0725281962472875e-05,
+      "loss": 1.1459,
+      "step": 3875
+    },
+    {
+      "epoch": 0.2851965969188319,
+      "grad_norm": 0.90234375,
+      "learning_rate": 4.072077917694314e-05,
+      "loss": 0.8921,
+      "step": 3876
+    },
+    {
+      "epoch": 0.28527017705219593,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.071627554770887e-05,
+      "loss": 0.8015,
+      "step": 3877
+    },
+    {
+      "epoch": 0.2853437571855599,
+      "grad_norm": 0.69921875,
+      "learning_rate": 4.071177107501178e-05,
+      "loss": 0.803,
+      "step": 3878
+    },
+    {
+      "epoch": 0.2854173373189239,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.070726575909361e-05,
+      "loss": 1.129,
+      "step": 3879
+    },
+    {
+      "epoch": 0.2854909174522879,
+      "grad_norm": 0.64453125,
+      "learning_rate": 4.0702759600196156e-05,
+      "loss": 0.7708,
+      "step": 3880
+    },
+    {
+      "epoch": 0.2855644975856519,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.069825259856125e-05,
+      "loss": 0.656,
+      "step": 3881
+    },
+    {
+      "epoch": 0.2856380777190159,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.0693744754430786e-05,
+      "loss": 0.7339,
+      "step": 3882
+    },
+    {
+      "epoch": 0.28571165785237984,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.0689236068046665e-05,
+      "loss": 1.2382,
+      "step": 3883
+    },
+    {
+      "epoch": 0.28578523798574385,
+      "grad_norm": 0.98046875,
+      "learning_rate": 4.068472653965089e-05,
+      "loss": 0.8877,
+      "step": 3884
+    },
+    {
+      "epoch": 0.28585881811910785,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.0680216169485466e-05,
+      "loss": 1.1071,
+      "step": 3885
+    },
+    {
+      "epoch": 0.28593239825247185,
+      "grad_norm": 1.0234375,
+      "learning_rate": 4.067570495779245e-05,
+      "loss": 1.5672,
+      "step": 3886
+    },
+    {
+      "epoch": 0.2860059783858358,
+      "grad_norm": 0.73046875,
+      "learning_rate": 4.067119290481396e-05,
+      "loss": 0.9924,
+      "step": 3887
+    },
+    {
+      "epoch": 0.2860795585191998,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.066668001079214e-05,
+      "loss": 1.0488,
+      "step": 3888
+    },
+    {
+      "epoch": 0.2861531386525638,
+      "grad_norm": 0.76953125,
+      "learning_rate": 4.066216627596921e-05,
+      "loss": 0.7105,
+      "step": 3889
+    },
+    {
+      "epoch": 0.2862267187859278,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.0657651700587376e-05,
+      "loss": 1.1021,
+      "step": 3890
+    },
+    {
+      "epoch": 0.2863002989192918,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.065313628488896e-05,
+      "loss": 0.8837,
+      "step": 3891
+    },
+    {
+      "epoch": 0.28637387905265577,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.064862002911628e-05,
+      "loss": 1.0584,
+      "step": 3892
+    },
+    {
+      "epoch": 0.28644745918601977,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.064410293351172e-05,
+      "loss": 0.8034,
+      "step": 3893
+    },
+    {
+      "epoch": 0.2865210393193838,
+      "grad_norm": 1.0078125,
+      "learning_rate": 4.0639584998317706e-05,
+      "loss": 1.5604,
+      "step": 3894
+    },
+    {
+      "epoch": 0.2865946194527478,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.0635066223776706e-05,
+      "loss": 0.8334,
+      "step": 3895
+    },
+    {
+      "epoch": 0.2866681995861117,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.0630546610131224e-05,
+      "loss": 1.2412,
+      "step": 3896
+    },
+    {
+      "epoch": 0.28674177971947573,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.062602615762384e-05,
+      "loss": 0.7904,
+      "step": 3897
+    },
+    {
+      "epoch": 0.28681535985283974,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.0621504866497136e-05,
+      "loss": 0.9662,
+      "step": 3898
+    },
+    {
+      "epoch": 0.28688893998620374,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.061698273699377e-05,
+      "loss": 1.0871,
+      "step": 3899
+    },
+    {
+      "epoch": 0.28696252011956774,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.0612459769356434e-05,
+      "loss": 1.1765,
+      "step": 3900
+    },
+    {
+      "epoch": 0.2870361002529317,
+      "grad_norm": 0.93359375,
+      "learning_rate": 4.060793596382788e-05,
+      "loss": 0.6849,
+      "step": 3901
+    },
+    {
+      "epoch": 0.2871096803862957,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.060341132065088e-05,
+      "loss": 1.3514,
+      "step": 3902
+    },
+    {
+      "epoch": 0.2871832605196597,
+      "grad_norm": 1.03125,
+      "learning_rate": 4.0598885840068264e-05,
+      "loss": 1.0643,
+      "step": 3903
+    },
+    {
+      "epoch": 0.2872568406530237,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.05943595223229e-05,
+      "loss": 0.9268,
+      "step": 3904
+    },
+    {
+      "epoch": 0.28733042078638765,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.0589832367657724e-05,
+      "loss": 1.1582,
+      "step": 3905
+    },
+    {
+      "epoch": 0.28740400091975166,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.058530437631568e-05,
+      "loss": 0.8383,
+      "step": 3906
+    },
+    {
+      "epoch": 0.28747758105311566,
+      "grad_norm": 1.0,
+      "learning_rate": 4.05807755485398e-05,
+      "loss": 1.1832,
+      "step": 3907
+    },
+    {
+      "epoch": 0.28755116118647966,
+      "grad_norm": 0.99609375,
+      "learning_rate": 4.0576245884573105e-05,
+      "loss": 1.1852,
+      "step": 3908
+    },
+    {
+      "epoch": 0.28762474131984367,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.057171538465873e-05,
+      "loss": 0.9573,
+      "step": 3909
+    },
+    {
+      "epoch": 0.2876983214532076,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.0567184049039796e-05,
+      "loss": 0.8536,
+      "step": 3910
+    },
+    {
+      "epoch": 0.2877719015865716,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.056265187795949e-05,
+      "loss": 1.3186,
+      "step": 3911
+    },
+    {
+      "epoch": 0.2878454817199356,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.055811887166106e-05,
+      "loss": 1.015,
+      "step": 3912
+    },
+    {
+      "epoch": 0.28791906185329963,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.055358503038777e-05,
+      "loss": 0.789,
+      "step": 3913
+    },
+    {
+      "epoch": 0.2879926419866636,
+      "grad_norm": 0.8203125,
+      "learning_rate": 4.054905035438295e-05,
+      "loss": 0.8662,
+      "step": 3914
+    },
+    {
+      "epoch": 0.2880662221200276,
+      "grad_norm": 0.75,
+      "learning_rate": 4.054451484388996e-05,
+      "loss": 0.5611,
+      "step": 3915
+    },
+    {
+      "epoch": 0.2881398022533916,
+      "grad_norm": 0.984375,
+      "learning_rate": 4.0539978499152235e-05,
+      "loss": 1.1378,
+      "step": 3916
+    },
+    {
+      "epoch": 0.2882133823867556,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.0535441320413194e-05,
+      "loss": 0.9651,
+      "step": 3917
+    },
+    {
+      "epoch": 0.2882869625201196,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.053090330791637e-05,
+      "loss": 1.051,
+      "step": 3918
+    },
+    {
+      "epoch": 0.28836054265348354,
+      "grad_norm": 0.94921875,
+      "learning_rate": 4.0526364461905295e-05,
+      "loss": 1.1585,
+      "step": 3919
+    },
+    {
+      "epoch": 0.28843412278684755,
+      "grad_norm": 0.86328125,
+      "learning_rate": 4.052182478262357e-05,
+      "loss": 1.3894,
+      "step": 3920
+    },
+    {
+      "epoch": 0.28850770292021155,
+      "grad_norm": 0.703125,
+      "learning_rate": 4.0517284270314826e-05,
+      "loss": 0.6976,
+      "step": 3921
+    },
+    {
+      "epoch": 0.28858128305357555,
+      "grad_norm": 0.96484375,
+      "learning_rate": 4.051274292522273e-05,
+      "loss": 0.9478,
+      "step": 3922
+    },
+    {
+      "epoch": 0.2886548631869395,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.050820074759104e-05,
+      "loss": 1.0256,
+      "step": 3923
+    },
+    {
+      "epoch": 0.2887284433203035,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.050365773766349e-05,
+      "loss": 0.939,
+      "step": 3924
+    },
+    {
+      "epoch": 0.2888020234536675,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.0499113895683927e-05,
+      "loss": 0.9522,
+      "step": 3925
+    },
+    {
+      "epoch": 0.2888756035870315,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.049456922189618e-05,
+      "loss": 0.9204,
+      "step": 3926
+    },
+    {
+      "epoch": 0.2889491837203955,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.049002371654418e-05,
+      "loss": 0.7246,
+      "step": 3927
+    },
+    {
+      "epoch": 0.28902276385375947,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.048547737987185e-05,
+      "loss": 0.8078,
+      "step": 3928
+    },
+    {
+      "epoch": 0.28909634398712347,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.04809302121232e-05,
+      "loss": 0.992,
+      "step": 3929
+    },
+    {
+      "epoch": 0.2891699241204875,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.047638221354228e-05,
+      "loss": 0.9145,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2892435042538515,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.047183338437314e-05,
+      "loss": 0.995,
+      "step": 3931
+    },
+    {
+      "epoch": 0.28931708438721543,
+      "grad_norm": 1.0859375,
+      "learning_rate": 4.046728372485994e-05,
+      "loss": 1.2221,
+      "step": 3932
+    },
+    {
+      "epoch": 0.28939066452057943,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.046273323524682e-05,
+      "loss": 0.8771,
+      "step": 3933
+    },
+    {
+      "epoch": 0.28946424465394344,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.0458181915778026e-05,
+      "loss": 1.0416,
+      "step": 3934
+    },
+    {
+      "epoch": 0.28953782478730744,
+      "grad_norm": 0.84375,
+      "learning_rate": 4.0453629766697796e-05,
+      "loss": 0.9703,
+      "step": 3935
+    },
+    {
+      "epoch": 0.28961140492067144,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.0449076788250446e-05,
+      "loss": 0.7855,
+      "step": 3936
+    },
+    {
+      "epoch": 0.2896849850540354,
+      "grad_norm": 1.125,
+      "learning_rate": 4.044452298068033e-05,
+      "loss": 1.1892,
+      "step": 3937
+    },
+    {
+      "epoch": 0.2897585651873994,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.043996834423183e-05,
+      "loss": 1.0772,
+      "step": 3938
+    },
+    {
+      "epoch": 0.2898321453207634,
+      "grad_norm": 0.6796875,
+      "learning_rate": 4.043541287914939e-05,
+      "loss": 0.6697,
+      "step": 3939
+    },
+    {
+      "epoch": 0.2899057254541274,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.043085658567749e-05,
+      "loss": 0.8478,
+      "step": 3940
+    },
+    {
+      "epoch": 0.28997930558749135,
+      "grad_norm": 1.0390625,
+      "learning_rate": 4.042629946406067e-05,
+      "loss": 1.2608,
+      "step": 3941
+    },
+    {
+      "epoch": 0.29005288572085536,
+      "grad_norm": 0.80078125,
+      "learning_rate": 4.042174151454349e-05,
+      "loss": 0.935,
+      "step": 3942
+    },
+    {
+      "epoch": 0.29012646585421936,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.0417182737370574e-05,
+      "loss": 0.989,
+      "step": 3943
+    },
+    {
+      "epoch": 0.29020004598758337,
+      "grad_norm": 1.15625,
+      "learning_rate": 4.041262313278657e-05,
+      "loss": 1.1673,
+      "step": 3944
+    },
+    {
+      "epoch": 0.29027362612094737,
+      "grad_norm": 1.09375,
+      "learning_rate": 4.040806270103621e-05,
+      "loss": 1.3329,
+      "step": 3945
+    },
+    {
+      "epoch": 0.2903472062543113,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.0403501442364213e-05,
+      "loss": 0.9676,
+      "step": 3946
+    },
+    {
+      "epoch": 0.2904207863876753,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.039893935701539e-05,
+      "loss": 0.8828,
+      "step": 3947
+    },
+    {
+      "epoch": 0.2904943665210393,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.039437644523458e-05,
+      "loss": 0.9777,
+      "step": 3948
+    },
+    {
+      "epoch": 0.29056794665440333,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.038981270726666e-05,
+      "loss": 0.8939,
+      "step": 3949
+    },
+    {
+      "epoch": 0.2906415267877673,
+      "grad_norm": 0.79296875,
+      "learning_rate": 4.038524814335656e-05,
+      "loss": 0.8317,
+      "step": 3950
+    },
+    {
+      "epoch": 0.2907151069211313,
+      "grad_norm": 0.875,
+      "learning_rate": 4.0380682753749245e-05,
+      "loss": 1.1056,
+      "step": 3951
+    },
+    {
+      "epoch": 0.2907886870544953,
+      "grad_norm": 0.7109375,
+      "learning_rate": 4.037611653868974e-05,
+      "loss": 0.8329,
+      "step": 3952
+    },
+    {
+      "epoch": 0.2908622671878593,
+      "grad_norm": 0.84765625,
+      "learning_rate": 4.03715494984231e-05,
+      "loss": 1.1901,
+      "step": 3953
+    },
+    {
+      "epoch": 0.2909358473212233,
+      "grad_norm": 0.78125,
+      "learning_rate": 4.0366981633194434e-05,
+      "loss": 0.8132,
+      "step": 3954
+    },
+    {
+      "epoch": 0.29100942745458724,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.036241294324889e-05,
+      "loss": 1.1858,
+      "step": 3955
+    },
+    {
+      "epoch": 0.29108300758795125,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.035784342883165e-05,
+      "loss": 0.6956,
+      "step": 3956
+    },
+    {
+      "epoch": 0.29115658772131525,
+      "grad_norm": 0.73828125,
+      "learning_rate": 4.0353273090187974e-05,
+      "loss": 1.0991,
+      "step": 3957
+    },
+    {
+      "epoch": 0.29123016785467926,
+      "grad_norm": 0.890625,
+      "learning_rate": 4.034870192756311e-05,
+      "loss": 1.155,
+      "step": 3958
+    },
+    {
+      "epoch": 0.2913037479880432,
+      "grad_norm": 1.2421875,
+      "learning_rate": 4.034412994120242e-05,
+      "loss": 1.4814,
+      "step": 3959
+    },
+    {
+      "epoch": 0.2913773281214072,
+      "grad_norm": 0.77734375,
+      "learning_rate": 4.033955713135126e-05,
+      "loss": 0.8542,
+      "step": 3960
+    },
+    {
+      "epoch": 0.2914509082547712,
+      "grad_norm": 0.75,
+      "learning_rate": 4.033498349825502e-05,
+      "loss": 0.8671,
+      "step": 3961
+    },
+    {
+      "epoch": 0.2915244883881352,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.03304090421592e-05,
+      "loss": 0.8192,
+      "step": 3962
+    },
+    {
+      "epoch": 0.2915980685214992,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.032583376330927e-05,
+      "loss": 0.7636,
+      "step": 3963
+    },
+    {
+      "epoch": 0.29167164865486317,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.032125766195079e-05,
+      "loss": 0.7492,
+      "step": 3964
+    },
+    {
+      "epoch": 0.2917452287882272,
+      "grad_norm": 0.9921875,
+      "learning_rate": 4.031668073832935e-05,
+      "loss": 0.8939,
+      "step": 3965
+    },
+    {
+      "epoch": 0.2918188089215912,
+      "grad_norm": 0.8671875,
+      "learning_rate": 4.031210299269059e-05,
+      "loss": 1.0382,
+      "step": 3966
+    },
+    {
+      "epoch": 0.2918923890549552,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.030752442528017e-05,
+      "loss": 0.8662,
+      "step": 3967
+    },
+    {
+      "epoch": 0.29196596918831913,
+      "grad_norm": 0.8125,
+      "learning_rate": 4.030294503634384e-05,
+      "loss": 1.2106,
+      "step": 3968
+    },
+    {
+      "epoch": 0.29203954932168313,
+      "grad_norm": 0.765625,
+      "learning_rate": 4.029836482612734e-05,
+      "loss": 0.7495,
+      "step": 3969
+    },
+    {
+      "epoch": 0.29211312945504714,
+      "grad_norm": 0.71875,
+      "learning_rate": 4.0293783794876504e-05,
+      "loss": 0.4663,
+      "step": 3970
+    },
+    {
+      "epoch": 0.29218670958841114,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.0289201942837174e-05,
+      "loss": 0.9168,
+      "step": 3971
+    },
+    {
+      "epoch": 0.29226028972177515,
+      "grad_norm": 0.83203125,
+      "learning_rate": 4.028461927025525e-05,
+      "loss": 0.6534,
+      "step": 3972
+    },
+    {
+      "epoch": 0.2923338698551391,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.028003577737669e-05,
+      "loss": 0.7976,
+      "step": 3973
+    },
+    {
+      "epoch": 0.2924074499885031,
+      "grad_norm": 1.0546875,
+      "learning_rate": 4.0275451464447454e-05,
+      "loss": 0.9889,
+      "step": 3974
+    },
+    {
+      "epoch": 0.2924810301218671,
+      "grad_norm": 1.015625,
+      "learning_rate": 4.02708663317136e-05,
+      "loss": 1.248,
+      "step": 3975
+    },
+    {
+      "epoch": 0.2925546102552311,
+      "grad_norm": 1.0625,
+      "learning_rate": 4.0266280379421195e-05,
+      "loss": 0.9817,
+      "step": 3976
+    },
+    {
+      "epoch": 0.29262819038859506,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.0261693607816344e-05,
+      "loss": 0.7109,
+      "step": 3977
+    },
+    {
+      "epoch": 0.29270177052195906,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.025710601714523e-05,
+      "loss": 1.7039,
+      "step": 3978
+    },
+    {
+      "epoch": 0.29277535065532306,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.025251760765405e-05,
+      "loss": 0.9992,
+      "step": 3979
+    },
+    {
+      "epoch": 0.29284893078868707,
+      "grad_norm": 0.75390625,
+      "learning_rate": 4.024792837958906e-05,
+      "loss": 0.6932,
+      "step": 3980
+    },
+    {
+      "epoch": 0.29292251092205107,
+      "grad_norm": 0.87890625,
+      "learning_rate": 4.0243338333196556e-05,
+      "loss": 0.812,
+      "step": 3981
+    },
+    {
+      "epoch": 0.292996091055415,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.023874746872287e-05,
+      "loss": 0.5889,
+      "step": 3982
+    },
+    {
+      "epoch": 0.293069671188779,
+      "grad_norm": 1.8671875,
+      "learning_rate": 4.023415578641438e-05,
+      "loss": 1.2658,
+      "step": 3983
+    },
+    {
+      "epoch": 0.29314325132214303,
+      "grad_norm": 0.85546875,
+      "learning_rate": 4.022956328651754e-05,
+      "loss": 0.8574,
+      "step": 3984
+    },
+    {
+      "epoch": 0.29321683145550703,
+      "grad_norm": 0.78515625,
+      "learning_rate": 4.022496996927879e-05,
+      "loss": 0.7555,
+      "step": 3985
+    },
+    {
+      "epoch": 0.293290411588871,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.022037583494466e-05,
+      "loss": 0.921,
+      "step": 3986
+    },
+    {
+      "epoch": 0.293363991722235,
+      "grad_norm": 0.76171875,
+      "learning_rate": 4.021578088376171e-05,
+      "loss": 0.7318,
+      "step": 3987
+    },
+    {
+      "epoch": 0.293437571855599,
+      "grad_norm": 0.875,
+      "learning_rate": 4.021118511597654e-05,
+      "loss": 0.6764,
+      "step": 3988
+    },
+    {
+      "epoch": 0.293511151988963,
+      "grad_norm": 0.8984375,
+      "learning_rate": 4.0206588531835795e-05,
+      "loss": 1.1966,
+      "step": 3989
+    },
+    {
+      "epoch": 0.293584732122327,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.0201991131586156e-05,
+      "loss": 0.663,
+      "step": 3990
+    },
+    {
+      "epoch": 0.29365831225569095,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.0197392915474375e-05,
+      "loss": 0.6867,
+      "step": 3991
+    },
+    {
+      "epoch": 0.29373189238905495,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.019279388374722e-05,
+      "loss": 0.9117,
+      "step": 3992
+    },
+    {
+      "epoch": 0.29380547252241895,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.01881940366515e-05,
+      "loss": 1.0109,
+      "step": 3993
+    },
+    {
+      "epoch": 0.29387905265578296,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.0183593374434106e-05,
+      "loss": 0.8342,
+      "step": 3994
+    },
+    {
+      "epoch": 0.2939526327891469,
+      "grad_norm": 0.8515625,
+      "learning_rate": 4.0178991897341925e-05,
+      "loss": 1.0372,
+      "step": 3995
+    },
+    {
+      "epoch": 0.2940262129225109,
+      "grad_norm": 0.6953125,
+      "learning_rate": 4.017438960562192e-05,
+      "loss": 0.7008,
+      "step": 3996
+    },
+    {
+      "epoch": 0.2940997930558749,
+      "grad_norm": 0.71484375,
+      "learning_rate": 4.0169786499521083e-05,
+      "loss": 0.7199,
+      "step": 3997
+    },
+    {
+      "epoch": 0.2941733731892389,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.0165182579286467e-05,
+      "loss": 1.1108,
+      "step": 3998
+    },
+    {
+      "epoch": 0.2942469533226029,
+      "grad_norm": 0.6640625,
+      "learning_rate": 4.016057784516513e-05,
+      "loss": 0.6917,
+      "step": 3999
+    },
+    {
+      "epoch": 0.29432053345596687,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.015597229740422e-05,
+      "loss": 0.8002,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2943941135893309,
+      "grad_norm": 0.9609375,
+      "learning_rate": 4.015136593625091e-05,
+      "loss": 1.214,
+      "step": 4001
+    },
+    {
+      "epoch": 0.2944676937226949,
+      "grad_norm": 0.74609375,
+      "learning_rate": 4.0146758761952396e-05,
+      "loss": 0.6247,
+      "step": 4002
+    },
+    {
+      "epoch": 0.2945412738560589,
+      "grad_norm": 1.1484375,
+      "learning_rate": 4.014215077475596e-05,
+      "loss": 1.5746,
+      "step": 4003
+    },
+    {
+      "epoch": 0.29461485398942283,
+      "grad_norm": 1.171875,
+      "learning_rate": 4.013754197490888e-05,
+      "loss": 1.0741,
+      "step": 4004
+    },
+    {
+      "epoch": 0.29468843412278684,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.0132932362658516e-05,
+      "loss": 1.278,
+      "step": 4005
+    },
+    {
+      "epoch": 0.29476201425615084,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.0128321938252254e-05,
+      "loss": 1.0738,
+      "step": 4006
+    },
+    {
+      "epoch": 0.29483559438951484,
+      "grad_norm": 1.828125,
+      "learning_rate": 4.012371070193753e-05,
+      "loss": 0.8013,
+      "step": 4007
+    },
+    {
+      "epoch": 0.29490917452287885,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.011909865396181e-05,
+      "loss": 1.0525,
+      "step": 4008
+    },
+    {
+      "epoch": 0.2949827546562428,
+      "grad_norm": 0.7890625,
+      "learning_rate": 4.011448579457263e-05,
+      "loss": 0.8665,
+      "step": 4009
+    },
+    {
+      "epoch": 0.2950563347896068,
+      "grad_norm": 1.1328125,
+      "learning_rate": 4.010987212401754e-05,
+      "loss": 1.0984,
+      "step": 4010
+    },
+    {
+      "epoch": 0.2951299149229708,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.010525764254415e-05,
+      "loss": 0.7722,
+      "step": 4011
+    },
+    {
+      "epoch": 0.2952034950563348,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.010064235040012e-05,
+      "loss": 1.1803,
+      "step": 4012
+    },
+    {
+      "epoch": 0.29527707518969876,
+      "grad_norm": 1.125,
+      "learning_rate": 4.009602624783312e-05,
+      "loss": 1.4804,
+      "step": 4013
+    },
+    {
+      "epoch": 0.29535065532306276,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.009140933509092e-05,
+      "loss": 1.0147,
+      "step": 4014
+    },
+    {
+      "epoch": 0.29542423545642676,
+      "grad_norm": 0.80859375,
+      "learning_rate": 4.008679161242128e-05,
+      "loss": 0.826,
+      "step": 4015
+    },
+    {
+      "epoch": 0.29549781558979077,
+      "grad_norm": 0.67578125,
+      "learning_rate": 4.008217308007203e-05,
+      "loss": 0.8106,
+      "step": 4016
+    },
+    {
+      "epoch": 0.2955713957231548,
+      "grad_norm": 0.796875,
+      "learning_rate": 4.007755373829103e-05,
+      "loss": 1.3554,
+      "step": 4017
+    },
+    {
+      "epoch": 0.2956449758565187,
+      "grad_norm": 0.9375,
+      "learning_rate": 4.00729335873262e-05,
+      "loss": 1.0787,
+      "step": 4018
+    },
+    {
+      "epoch": 0.2957185559898827,
+      "grad_norm": 0.94140625,
+      "learning_rate": 4.00683126274255e-05,
+      "loss": 1.0574,
+      "step": 4019
+    },
+    {
+      "epoch": 0.29579213612324673,
+      "grad_norm": 1.1953125,
+      "learning_rate": 4.0063690858836914e-05,
+      "loss": 0.8312,
+      "step": 4020
+    },
+    {
+      "epoch": 0.29586571625661073,
+      "grad_norm": 0.859375,
+      "learning_rate": 4.0059068281808495e-05,
+      "loss": 0.9378,
+      "step": 4021
+    },
+    {
+      "epoch": 0.2959392963899747,
+      "grad_norm": 0.82421875,
+      "learning_rate": 4.0054444896588326e-05,
+      "loss": 0.832,
+      "step": 4022
+    },
+    {
+      "epoch": 0.2960128765233387,
+      "grad_norm": 0.953125,
+      "learning_rate": 4.0049820703424534e-05,
+      "loss": 1.4243,
+      "step": 4023
+    },
+    {
+      "epoch": 0.2960864566567027,
+      "grad_norm": 1.125,
+      "learning_rate": 4.0045195702565285e-05,
+      "loss": 0.8421,
+      "step": 4024
+    },
+    {
+      "epoch": 0.2961600367900667,
+      "grad_norm": 0.95703125,
+      "learning_rate": 4.0040569894258794e-05,
+      "loss": 1.042,
+      "step": 4025
+    },
+    {
+      "epoch": 0.2962336169234307,
+      "grad_norm": 0.83984375,
+      "learning_rate": 4.003594327875334e-05,
+      "loss": 0.8849,
+      "step": 4026
+    },
+    {
+      "epoch": 0.29630719705679465,
+      "grad_norm": 0.92578125,
+      "learning_rate": 4.003131585629719e-05,
+      "loss": 1.2332,
+      "step": 4027
+    },
+    {
+      "epoch": 0.29638077719015865,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.002668762713873e-05,
+      "loss": 0.7748,
+      "step": 4028
+    },
+    {
+      "epoch": 0.29645435732352265,
+      "grad_norm": 1.0,
+      "learning_rate": 4.002205859152631e-05,
+      "loss": 1.0201,
+      "step": 4029
+    },
+    {
+      "epoch": 0.29652793745688666,
+      "grad_norm": 0.7421875,
+      "learning_rate": 4.0017428749708385e-05,
+      "loss": 0.6225,
+      "step": 4030
+    },
+    {
+      "epoch": 0.2966015175902506,
+      "grad_norm": 0.98828125,
+      "learning_rate": 4.0012798101933414e-05,
+      "loss": 1.0481,
+      "step": 4031
+    },
+    {
+      "epoch": 0.2966750977236146,
+      "grad_norm": 0.875,
+      "learning_rate": 4.000816664844993e-05,
+      "loss": 0.7896,
+      "step": 4032
+    },
+    {
+      "epoch": 0.2967486778569786,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.000353438950649e-05,
+      "loss": 0.963,
+      "step": 4033
+    },
+    {
+      "epoch": 0.2968222579903426,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.9998901325351694e-05,
+      "loss": 1.1171,
+      "step": 4034
+    },
+    {
+      "epoch": 0.2968958381237066,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.99942674562342e-05,
+      "loss": 0.9836,
+      "step": 4035
+    },
+    {
+      "epoch": 0.29696941825707057,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.998963278240268e-05,
+      "loss": 1.0004,
+      "step": 4036
+    },
+    {
+      "epoch": 0.2970429983904346,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.9984997304105885e-05,
+      "loss": 1.484,
+      "step": 4037
+    },
+    {
+      "epoch": 0.2971165785237986,
+      "grad_norm": 1.1171875,
+      "learning_rate": 3.998036102159259e-05,
+      "loss": 0.897,
+      "step": 4038
+    },
+    {
+      "epoch": 0.2971901586571626,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.9975723935111614e-05,
+      "loss": 0.9784,
+      "step": 4039
+    },
+    {
+      "epoch": 0.29726373879052653,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.997108604491182e-05,
+      "loss": 1.2529,
+      "step": 4040
+    },
+    {
+      "epoch": 0.29733731892389054,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.9966447351242106e-05,
+      "loss": 0.8156,
+      "step": 4041
+    },
+    {
+      "epoch": 0.29741089905725454,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.996180785435144e-05,
+      "loss": 1.016,
+      "step": 4042
+    },
+    {
+      "epoch": 0.29748447919061854,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.9957167554488795e-05,
+      "loss": 1.1219,
+      "step": 4043
+    },
+    {
+      "epoch": 0.29755805932398255,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.995252645190323e-05,
+      "loss": 1.1444,
+      "step": 4044
+    },
+    {
+      "epoch": 0.2976316394573465,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.99478845468438e-05,
+      "loss": 0.906,
+      "step": 4045
+    },
+    {
+      "epoch": 0.2977052195907105,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.994324183955964e-05,
+      "loss": 0.8373,
+      "step": 4046
+    },
+    {
+      "epoch": 0.2977787997240745,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.993859833029993e-05,
+      "loss": 1.2315,
+      "step": 4047
+    },
+    {
+      "epoch": 0.2978523798574385,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.9933954019313844e-05,
+      "loss": 0.8823,
+      "step": 4048
+    },
+    {
+      "epoch": 0.29792595999080246,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.992930890685066e-05,
+      "loss": 0.5843,
+      "step": 4049
+    },
+    {
+      "epoch": 0.29799954012416646,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.992466299315965e-05,
+      "loss": 0.8507,
+      "step": 4050
+    },
+    {
+      "epoch": 0.29807312025753047,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.992001627849019e-05,
+      "loss": 0.9282,
+      "step": 4051
+    },
+    {
+      "epoch": 0.29814670039089447,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.991536876309162e-05,
+      "loss": 0.8898,
+      "step": 4052
+    },
+    {
+      "epoch": 0.2982202805242585,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.991072044721339e-05,
+      "loss": 0.8895,
+      "step": 4053
+    },
+    {
+      "epoch": 0.2982938606576224,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.990607133110495e-05,
+      "loss": 0.909,
+      "step": 4054
+    },
+    {
+      "epoch": 0.2983674407909864,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.9901421415015815e-05,
+      "loss": 1.1622,
+      "step": 4055
+    },
+    {
+      "epoch": 0.29844102092435043,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.989677069919554e-05,
+      "loss": 0.8779,
+      "step": 4056
+    },
+    {
+      "epoch": 0.29851460105771443,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.9892119183893715e-05,
+      "loss": 0.9261,
+      "step": 4057
+    },
+    {
+      "epoch": 0.2985881811910784,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.988746686935998e-05,
+      "loss": 0.8476,
+      "step": 4058
+    },
+    {
+      "epoch": 0.2986617613244424,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.9882813755844015e-05,
+      "loss": 1.0539,
+      "step": 4059
+    },
+    {
+      "epoch": 0.2987353414578064,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.9878159843595554e-05,
+      "loss": 1.2995,
+      "step": 4060
+    },
+    {
+      "epoch": 0.2988089215911704,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.987350513286435e-05,
+      "loss": 0.9445,
+      "step": 4061
+    },
+    {
+      "epoch": 0.2988825017245344,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.986884962390022e-05,
+      "loss": 1.2249,
+      "step": 4062
+    },
+    {
+      "epoch": 0.29895608185789835,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.986419331695301e-05,
+      "loss": 0.9489,
+      "step": 4063
+    },
+    {
+      "epoch": 0.29902966199126235,
+      "grad_norm": 1.3515625,
+      "learning_rate": 3.985953621227262e-05,
+      "loss": 0.9554,
+      "step": 4064
+    },
+    {
+      "epoch": 0.29910324212462636,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.985487831010899e-05,
+      "loss": 1.1644,
+      "step": 4065
+    },
+    {
+      "epoch": 0.29917682225799036,
+      "grad_norm": 0.875,
+      "learning_rate": 3.985021961071209e-05,
+      "loss": 0.7879,
+      "step": 4066
+    },
+    {
+      "epoch": 0.2992504023913543,
+      "grad_norm": 1.578125,
+      "learning_rate": 3.984556011433196e-05,
+      "loss": 1.0147,
+      "step": 4067
+    },
+    {
+      "epoch": 0.2993239825247183,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.984089982121865e-05,
+      "loss": 0.8522,
+      "step": 4068
+    },
+    {
+      "epoch": 0.2993975626580823,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.983623873162229e-05,
+      "loss": 0.751,
+      "step": 4069
+    },
+    {
+      "epoch": 0.2994711427914463,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.983157684579301e-05,
+      "loss": 0.8797,
+      "step": 4070
+    },
+    {
+      "epoch": 0.2995447229248103,
+      "grad_norm": 0.75,
+      "learning_rate": 3.982691416398101e-05,
+      "loss": 0.7336,
+      "step": 4071
+    },
+    {
+      "epoch": 0.2996183030581743,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.9822250686436534e-05,
+      "loss": 1.3268,
+      "step": 4072
+    },
+    {
+      "epoch": 0.2996918831915383,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.981758641340986e-05,
+      "loss": 0.9972,
+      "step": 4073
+    },
+    {
+      "epoch": 0.2997654633249023,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.981292134515131e-05,
+      "loss": 1.0735,
+      "step": 4074
+    },
+    {
+      "epoch": 0.2998390434582663,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.980825548191125e-05,
+      "loss": 1.1444,
+      "step": 4075
+    },
+    {
+      "epoch": 0.29991262359163023,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.9803588823940085e-05,
+      "loss": 0.849,
+      "step": 4076
+    },
+    {
+      "epoch": 0.29998620372499424,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.979892137148827e-05,
+      "loss": 0.8298,
+      "step": 4077
+    },
+    {
+      "epoch": 0.30005978385835824,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.97942531248063e-05,
+      "loss": 1.0243,
+      "step": 4078
+    },
+    {
+      "epoch": 0.30013336399172225,
+      "grad_norm": 1.0,
+      "learning_rate": 3.97895840841447e-05,
+      "loss": 1.2506,
+      "step": 4079
+    },
+    {
+      "epoch": 0.30020694412508625,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.978491424975406e-05,
+      "loss": 0.9184,
+      "step": 4080
+    },
+    {
+      "epoch": 0.3002805242584502,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.9780243621884997e-05,
+      "loss": 1.4987,
+      "step": 4081
+    },
+    {
+      "epoch": 0.3003541043918142,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.977557220078817e-05,
+      "loss": 1.0628,
+      "step": 4082
+    },
+    {
+      "epoch": 0.3004276845251782,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.977089998671429e-05,
+      "loss": 0.6733,
+      "step": 4083
+    },
+    {
+      "epoch": 0.3005012646585422,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.9766226979914104e-05,
+      "loss": 0.7021,
+      "step": 4084
+    },
+    {
+      "epoch": 0.30057484479190616,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.976155318063841e-05,
+      "loss": 0.6074,
+      "step": 4085
+    },
+    {
+      "epoch": 0.30064842492527016,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.9756878589138044e-05,
+      "loss": 0.7702,
+      "step": 4086
+    },
+    {
+      "epoch": 0.30072200505863417,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.9752203205663865e-05,
+      "loss": 0.6078,
+      "step": 4087
+    },
+    {
+      "epoch": 0.30079558519199817,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.9747527030466805e-05,
+      "loss": 0.6987,
+      "step": 4088
+    },
+    {
+      "epoch": 0.3008691653253622,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.974285006379783e-05,
+      "loss": 1.3702,
+      "step": 4089
+    },
+    {
+      "epoch": 0.3009427454587261,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.9738172305907936e-05,
+      "loss": 1.4945,
+      "step": 4090
+    },
+    {
+      "epoch": 0.30101632559209013,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.973349375704816e-05,
+      "loss": 1.2491,
+      "step": 4091
+    },
+    {
+      "epoch": 0.30108990572545413,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.972881441746962e-05,
+      "loss": 0.9268,
+      "step": 4092
+    },
+    {
+      "epoch": 0.30116348585881814,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.9724134287423406e-05,
+      "loss": 0.7378,
+      "step": 4093
+    },
+    {
+      "epoch": 0.3012370659921821,
+      "grad_norm": 1.1171875,
+      "learning_rate": 3.971945336716074e-05,
+      "loss": 0.9854,
+      "step": 4094
+    },
+    {
+      "epoch": 0.3013106461255461,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.971477165693279e-05,
+      "loss": 0.8025,
+      "step": 4095
+    },
+    {
+      "epoch": 0.3013842262589101,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.9710089156990856e-05,
+      "loss": 1.0144,
+      "step": 4096
+    },
+    {
+      "epoch": 0.3014578063922741,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.970540586758621e-05,
+      "loss": 0.9574,
+      "step": 4097
+    },
+    {
+      "epoch": 0.3015313865256381,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.970072178897021e-05,
+      "loss": 0.7508,
+      "step": 4098
+    },
+    {
+      "epoch": 0.30160496665900205,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.969603692139423e-05,
+      "loss": 1.0062,
+      "step": 4099
+    },
+    {
+      "epoch": 0.30167854679236605,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.969135126510971e-05,
+      "loss": 0.9119,
+      "step": 4100
+    },
+    {
+      "epoch": 0.30175212692573006,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.968666482036812e-05,
+      "loss": 1.28,
+      "step": 4101
+    },
+    {
+      "epoch": 0.30182570705909406,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.968197758742096e-05,
+      "loss": 1.1114,
+      "step": 4102
+    },
+    {
+      "epoch": 0.301899287192458,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.9677289566519796e-05,
+      "loss": 0.7357,
+      "step": 4103
+    },
+    {
+      "epoch": 0.301972867325822,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.967260075791622e-05,
+      "loss": 0.5691,
+      "step": 4104
+    },
+    {
+      "epoch": 0.302046447459186,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.966791116186188e-05,
+      "loss": 0.8774,
+      "step": 4105
+    },
+    {
+      "epoch": 0.30212002759255,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.966322077860846e-05,
+      "loss": 1.015,
+      "step": 4106
+    },
+    {
+      "epoch": 0.302193607725914,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.965852960840766e-05,
+      "loss": 0.8266,
+      "step": 4107
+    },
+    {
+      "epoch": 0.302267187859278,
+      "grad_norm": 0.875,
+      "learning_rate": 3.9653837651511266e-05,
+      "loss": 1.1217,
+      "step": 4108
+    },
+    {
+      "epoch": 0.302340767992642,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.964914490817108e-05,
+      "loss": 0.9724,
+      "step": 4109
+    },
+    {
+      "epoch": 0.302414348126006,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.9644451378638956e-05,
+      "loss": 0.853,
+      "step": 4110
+    },
+    {
+      "epoch": 0.30248792825937,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.963975706316679e-05,
+      "loss": 0.8187,
+      "step": 4111
+    },
+    {
+      "epoch": 0.30256150839273394,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.963506196200651e-05,
+      "loss": 0.979,
+      "step": 4112
+    },
+    {
+      "epoch": 0.30263508852609794,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.96303660754101e-05,
+      "loss": 1.1585,
+      "step": 4113
+    },
+    {
+      "epoch": 0.30270866865946194,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.9625669403629574e-05,
+      "loss": 0.923,
+      "step": 4114
+    },
+    {
+      "epoch": 0.30278224879282595,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.9620971946916996e-05,
+      "loss": 0.8744,
+      "step": 4115
+    },
+    {
+      "epoch": 0.30285582892618995,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.961627370552447e-05,
+      "loss": 1.3116,
+      "step": 4116
+    },
+    {
+      "epoch": 0.3029294090595539,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.961157467970413e-05,
+      "loss": 0.8754,
+      "step": 4117
+    },
+    {
+      "epoch": 0.3030029891929179,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.9606874869708186e-05,
+      "loss": 0.8331,
+      "step": 4118
+    },
+    {
+      "epoch": 0.3030765693262819,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.960217427578885e-05,
+      "loss": 1.3008,
+      "step": 4119
+    },
+    {
+      "epoch": 0.3031501494596459,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.9597472898198404e-05,
+      "loss": 0.9466,
+      "step": 4120
+    },
+    {
+      "epoch": 0.30322372959300986,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.9592770737189155e-05,
+      "loss": 0.8472,
+      "step": 4121
+    },
+    {
+      "epoch": 0.30329730972637386,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.9588067793013466e-05,
+      "loss": 0.9979,
+      "step": 4122
+    },
+    {
+      "epoch": 0.30337088985973787,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.9583364065923727e-05,
+      "loss": 1.0135,
+      "step": 4123
+    },
+    {
+      "epoch": 0.3034444699931019,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.9578659556172386e-05,
+      "loss": 1.0402,
+      "step": 4124
+    },
+    {
+      "epoch": 0.3035180501264659,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.957395426401192e-05,
+      "loss": 0.8235,
+      "step": 4125
+    },
+    {
+      "epoch": 0.3035916302598298,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.9569248189694863e-05,
+      "loss": 1.3151,
+      "step": 4126
+    },
+    {
+      "epoch": 0.30366521039319383,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.956454133347376e-05,
+      "loss": 1.1963,
+      "step": 4127
+    },
+    {
+      "epoch": 0.30373879052655783,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.955983369560124e-05,
+      "loss": 1.0767,
+      "step": 4128
+    },
+    {
+      "epoch": 0.30381237065992184,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.955512527632994e-05,
+      "loss": 0.7796,
+      "step": 4129
+    },
+    {
+      "epoch": 0.3038859507932858,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.955041607591256e-05,
+      "loss": 1.0109,
+      "step": 4130
+    },
+    {
+      "epoch": 0.3039595309266498,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.954570609460183e-05,
+      "loss": 1.0324,
+      "step": 4131
+    },
+    {
+      "epoch": 0.3040331110600138,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.954099533265053e-05,
+      "loss": 1.2995,
+      "step": 4132
+    },
+    {
+      "epoch": 0.3041066911933778,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.953628379031147e-05,
+      "loss": 0.7554,
+      "step": 4133
+    },
+    {
+      "epoch": 0.3041802713267418,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.953157146783751e-05,
+      "loss": 0.8913,
+      "step": 4134
+    },
+    {
+      "epoch": 0.30425385146010575,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.952685836548157e-05,
+      "loss": 0.813,
+      "step": 4135
+    },
+    {
+      "epoch": 0.30432743159346975,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.952214448349657e-05,
+      "loss": 1.051,
+      "step": 4136
+    },
+    {
+      "epoch": 0.30440101172683376,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.951742982213551e-05,
+      "loss": 0.946,
+      "step": 4137
+    },
+    {
+      "epoch": 0.30447459186019776,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.9512714381651406e-05,
+      "loss": 1.1763,
+      "step": 4138
+    },
+    {
+      "epoch": 0.3045481719935617,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.950799816229733e-05,
+      "loss": 0.7493,
+      "step": 4139
+    },
+    {
+      "epoch": 0.3046217521269257,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.95032811643264e-05,
+      "loss": 1.0026,
+      "step": 4140
+    },
+    {
+      "epoch": 0.3046953322602897,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.949856338799175e-05,
+      "loss": 1.0317,
+      "step": 4141
+    },
+    {
+      "epoch": 0.3047689123936537,
+      "grad_norm": 4.25,
+      "learning_rate": 3.94938448335466e-05,
+      "loss": 0.8519,
+      "step": 4142
+    },
+    {
+      "epoch": 0.3048424925270177,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.948912550124417e-05,
+      "loss": 1.1186,
+      "step": 4143
+    },
+    {
+      "epoch": 0.3049160726603817,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.9484405391337744e-05,
+      "loss": 0.6545,
+      "step": 4144
+    },
+    {
+      "epoch": 0.3049896527937457,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.947968450408063e-05,
+      "loss": 1.0666,
+      "step": 4145
+    },
+    {
+      "epoch": 0.3050632329271097,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.947496283972619e-05,
+      "loss": 0.9816,
+      "step": 4146
+    },
+    {
+      "epoch": 0.3051368130604737,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.9470240398527846e-05,
+      "loss": 1.1742,
+      "step": 4147
+    },
+    {
+      "epoch": 0.30521039319383764,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.946551718073903e-05,
+      "loss": 0.8515,
+      "step": 4148
+    },
+    {
+      "epoch": 0.30528397332720164,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.946079318661323e-05,
+      "loss": 1.1281,
+      "step": 4149
+    },
+    {
+      "epoch": 0.30535755346056564,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.945606841640397e-05,
+      "loss": 0.5859,
+      "step": 4150
+    },
+    {
+      "epoch": 0.30543113359392965,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.9451342870364816e-05,
+      "loss": 0.6277,
+      "step": 4151
+    },
+    {
+      "epoch": 0.30550471372729365,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.944661654874939e-05,
+      "loss": 1.4611,
+      "step": 4152
+    },
+    {
+      "epoch": 0.3055782938606576,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.944188945181134e-05,
+      "loss": 1.3086,
+      "step": 4153
+    },
+    {
+      "epoch": 0.3056518739940216,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.9437161579804355e-05,
+      "loss": 0.9333,
+      "step": 4154
+    },
+    {
+      "epoch": 0.3057254541273856,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.943243293298218e-05,
+      "loss": 0.8708,
+      "step": 4155
+    },
+    {
+      "epoch": 0.3057990342607496,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.942770351159859e-05,
+      "loss": 0.7147,
+      "step": 4156
+    },
+    {
+      "epoch": 0.30587261439411356,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.942297331590739e-05,
+      "loss": 0.7497,
+      "step": 4157
+    },
+    {
+      "epoch": 0.30594619452747757,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.9418242346162474e-05,
+      "loss": 1.3933,
+      "step": 4158
+    },
+    {
+      "epoch": 0.30601977466084157,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.9413510602617706e-05,
+      "loss": 1.102,
+      "step": 4159
+    },
+    {
+      "epoch": 0.3060933547942056,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.940877808552706e-05,
+      "loss": 1.2623,
+      "step": 4160
+    },
+    {
+      "epoch": 0.3061669349275696,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.94040447951445e-05,
+      "loss": 1.2528,
+      "step": 4161
+    },
+    {
+      "epoch": 0.3062405150609335,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.939931073172406e-05,
+      "loss": 0.9589,
+      "step": 4162
+    },
+    {
+      "epoch": 0.30631409519429753,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.939457589551982e-05,
+      "loss": 1.0826,
+      "step": 4163
+    },
+    {
+      "epoch": 0.30638767532766154,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.938984028678587e-05,
+      "loss": 0.7509,
+      "step": 4164
+    },
+    {
+      "epoch": 0.30646125546102554,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.9385103905776374e-05,
+      "loss": 0.8302,
+      "step": 4165
+    },
+    {
+      "epoch": 0.3065348355943895,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.938036675274552e-05,
+      "loss": 0.9071,
+      "step": 4166
+    },
+    {
+      "epoch": 0.3066084157277535,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.937562882794754e-05,
+      "loss": 0.6937,
+      "step": 4167
+    },
+    {
+      "epoch": 0.3066819958611175,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.937089013163672e-05,
+      "loss": 0.9938,
+      "step": 4168
+    },
+    {
+      "epoch": 0.3067555759944815,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.936615066406737e-05,
+      "loss": 0.6215,
+      "step": 4169
+    },
+    {
+      "epoch": 0.3068291561278455,
+      "grad_norm": 3.578125,
+      "learning_rate": 3.936141042549384e-05,
+      "loss": 0.8972,
+      "step": 4170
+    },
+    {
+      "epoch": 0.30690273626120945,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.935666941617054e-05,
+      "loss": 0.8855,
+      "step": 4171
+    },
+    {
+      "epoch": 0.30697631639457346,
+      "grad_norm": 1.59375,
+      "learning_rate": 3.9351927636351905e-05,
+      "loss": 1.0294,
+      "step": 4172
+    },
+    {
+      "epoch": 0.30704989652793746,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.9347185086292424e-05,
+      "loss": 1.1058,
+      "step": 4173
+    },
+    {
+      "epoch": 0.30712347666130146,
+      "grad_norm": 1.0,
+      "learning_rate": 3.934244176624662e-05,
+      "loss": 0.7408,
+      "step": 4174
+    },
+    {
+      "epoch": 0.3071970567946654,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.9337697676469046e-05,
+      "loss": 1.0399,
+      "step": 4175
+    },
+    {
+      "epoch": 0.3072706369280294,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.933295281721433e-05,
+      "loss": 0.7997,
+      "step": 4176
+    },
+    {
+      "epoch": 0.3073442170613934,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.93282071887371e-05,
+      "loss": 0.8517,
+      "step": 4177
+    },
+    {
+      "epoch": 0.3074177971947574,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.9323460791292055e-05,
+      "loss": 1.04,
+      "step": 4178
+    },
+    {
+      "epoch": 0.30749137732812143,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.9318713625133926e-05,
+      "loss": 1.0115,
+      "step": 4179
+    },
+    {
+      "epoch": 0.3075649574614854,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.9313965690517475e-05,
+      "loss": 0.7472,
+      "step": 4180
+    },
+    {
+      "epoch": 0.3076385375948494,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.930921698769752e-05,
+      "loss": 0.9818,
+      "step": 4181
+    },
+    {
+      "epoch": 0.3077121177282134,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.930446751692892e-05,
+      "loss": 1.0383,
+      "step": 4182
+    },
+    {
+      "epoch": 0.3077856978615774,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.9299717278466566e-05,
+      "loss": 1.16,
+      "step": 4183
+    },
+    {
+      "epoch": 0.30785927799494134,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.929496627256539e-05,
+      "loss": 0.9647,
+      "step": 4184
+    },
+    {
+      "epoch": 0.30793285812830534,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.929021449948037e-05,
+      "loss": 0.8962,
+      "step": 4185
+    },
+    {
+      "epoch": 0.30800643826166935,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.928546195946654e-05,
+      "loss": 1.3702,
+      "step": 4186
+    },
+    {
+      "epoch": 0.30808001839503335,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.928070865277894e-05,
+      "loss": 0.9198,
+      "step": 4187
+    },
+    {
+      "epoch": 0.30815359852839735,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.927595457967268e-05,
+      "loss": 0.8546,
+      "step": 4188
+    },
+    {
+      "epoch": 0.3082271786617613,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.92711997404029e-05,
+      "loss": 1.02,
+      "step": 4189
+    },
+    {
+      "epoch": 0.3083007587951253,
+      "grad_norm": 0.640625,
+      "learning_rate": 3.9266444135224786e-05,
+      "loss": 0.5583,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3083743389284893,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.926168776439356e-05,
+      "loss": 0.9121,
+      "step": 4191
+    },
+    {
+      "epoch": 0.3084479190618533,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.925693062816449e-05,
+      "loss": 0.7983,
+      "step": 4192
+    },
+    {
+      "epoch": 0.30852149919521726,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.925217272679288e-05,
+      "loss": 1.0658,
+      "step": 4193
+    },
+    {
+      "epoch": 0.30859507932858127,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.924741406053407e-05,
+      "loss": 1.1073,
+      "step": 4194
+    },
+    {
+      "epoch": 0.30866865946194527,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.924265462964347e-05,
+      "loss": 1.1864,
+      "step": 4195
+    },
+    {
+      "epoch": 0.3087422395953093,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.923789443437649e-05,
+      "loss": 0.9088,
+      "step": 4196
+    },
+    {
+      "epoch": 0.3088158197286733,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.923313347498861e-05,
+      "loss": 0.7344,
+      "step": 4197
+    },
+    {
+      "epoch": 0.30888939986203723,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.9228371751735336e-05,
+      "loss": 0.8802,
+      "step": 4198
+    },
+    {
+      "epoch": 0.30896297999540123,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.922360926487223e-05,
+      "loss": 0.6581,
+      "step": 4199
+    },
+    {
+      "epoch": 0.30903656012876524,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.921884601465487e-05,
+      "loss": 1.1206,
+      "step": 4200
+    },
+    {
+      "epoch": 0.30911014026212924,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.921408200133891e-05,
+      "loss": 1.3348,
+      "step": 4201
+    },
+    {
+      "epoch": 0.30918372039549324,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.9209317225180006e-05,
+      "loss": 0.9765,
+      "step": 4202
+    },
+    {
+      "epoch": 0.3092573005288572,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.920455168643389e-05,
+      "loss": 1.094,
+      "step": 4203
+    },
+    {
+      "epoch": 0.3093308806622212,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.9199785385356314e-05,
+      "loss": 0.9887,
+      "step": 4204
+    },
+    {
+      "epoch": 0.3094044607955852,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.919501832220307e-05,
+      "loss": 1.0445,
+      "step": 4205
+    },
+    {
+      "epoch": 0.3094780409289492,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.919025049723001e-05,
+      "loss": 0.7977,
+      "step": 4206
+    },
+    {
+      "epoch": 0.30955162106231315,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.9185481910693004e-05,
+      "loss": 1.2536,
+      "step": 4207
+    },
+    {
+      "epoch": 0.30962520119567716,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.9180712562847974e-05,
+      "loss": 1.1189,
+      "step": 4208
+    },
+    {
+      "epoch": 0.30969878132904116,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.917594245395089e-05,
+      "loss": 1.6528,
+      "step": 4209
+    },
+    {
+      "epoch": 0.30977236146240517,
+      "grad_norm": 1.0,
+      "learning_rate": 3.917117158425774e-05,
+      "loss": 1.353,
+      "step": 4210
+    },
+    {
+      "epoch": 0.30984594159576917,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.916639995402459e-05,
+      "loss": 0.9726,
+      "step": 4211
+    },
+    {
+      "epoch": 0.3099195217291331,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.9161627563507494e-05,
+      "loss": 0.8628,
+      "step": 4212
+    },
+    {
+      "epoch": 0.3099931018624971,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.915685441296261e-05,
+      "loss": 0.8632,
+      "step": 4213
+    },
+    {
+      "epoch": 0.3100666819958611,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.915208050264608e-05,
+      "loss": 0.9593,
+      "step": 4214
+    },
+    {
+      "epoch": 0.31014026212922513,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.914730583281412e-05,
+      "loss": 0.7832,
+      "step": 4215
+    },
+    {
+      "epoch": 0.3102138422625891,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.9142530403722976e-05,
+      "loss": 1.0504,
+      "step": 4216
+    },
+    {
+      "epoch": 0.3102874223959531,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.9137754215628944e-05,
+      "loss": 0.9269,
+      "step": 4217
+    },
+    {
+      "epoch": 0.3103610025293171,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.913297726878834e-05,
+      "loss": 0.8606,
+      "step": 4218
+    },
+    {
+      "epoch": 0.3104345826626811,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.912819956345754e-05,
+      "loss": 0.852,
+      "step": 4219
+    },
+    {
+      "epoch": 0.3105081627960451,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.912342109989296e-05,
+      "loss": 0.9616,
+      "step": 4220
+    },
+    {
+      "epoch": 0.31058174292940904,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.911864187835103e-05,
+      "loss": 1.1499,
+      "step": 4221
+    },
+    {
+      "epoch": 0.31065532306277305,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.911386189908826e-05,
+      "loss": 0.7484,
+      "step": 4222
+    },
+    {
+      "epoch": 0.31072890319613705,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.910908116236118e-05,
+      "loss": 0.7645,
+      "step": 4223
+    },
+    {
+      "epoch": 0.31080248332950106,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.9104299668426375e-05,
+      "loss": 0.8799,
+      "step": 4224
+    },
+    {
+      "epoch": 0.310876063462865,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.909951741754043e-05,
+      "loss": 1.3231,
+      "step": 4225
+    },
+    {
+      "epoch": 0.310949643596229,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.9094734409960026e-05,
+      "loss": 0.8432,
+      "step": 4226
+    },
+    {
+      "epoch": 0.311023223729593,
+      "grad_norm": 1.3828125,
+      "learning_rate": 3.908995064594185e-05,
+      "loss": 1.1038,
+      "step": 4227
+    },
+    {
+      "epoch": 0.311096803862957,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.908516612574262e-05,
+      "loss": 0.8895,
+      "step": 4228
+    },
+    {
+      "epoch": 0.311170383996321,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.908038084961914e-05,
+      "loss": 0.944,
+      "step": 4229
+    },
+    {
+      "epoch": 0.31124396412968497,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.907559481782821e-05,
+      "loss": 0.7459,
+      "step": 4230
+    },
+    {
+      "epoch": 0.311317544263049,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.907080803062669e-05,
+      "loss": 0.8815,
+      "step": 4231
+    },
+    {
+      "epoch": 0.311391124396413,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.906602048827148e-05,
+      "loss": 1.0102,
+      "step": 4232
+    },
+    {
+      "epoch": 0.311464704529777,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.906123219101952e-05,
+      "loss": 0.9443,
+      "step": 4233
+    },
+    {
+      "epoch": 0.31153828466314093,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.905644313912778e-05,
+      "loss": 0.8356,
+      "step": 4234
+    },
+    {
+      "epoch": 0.31161186479650493,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.905165333285329e-05,
+      "loss": 1.1552,
+      "step": 4235
+    },
+    {
+      "epoch": 0.31168544492986894,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.904686277245311e-05,
+      "loss": 1.1431,
+      "step": 4236
+    },
+    {
+      "epoch": 0.31175902506323294,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.9042071458184323e-05,
+      "loss": 0.6391,
+      "step": 4237
+    },
+    {
+      "epoch": 0.31183260519659695,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.903727939030409e-05,
+      "loss": 0.7324,
+      "step": 4238
+    },
+    {
+      "epoch": 0.3119061853299609,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.903248656906958e-05,
+      "loss": 0.7205,
+      "step": 4239
+    },
+    {
+      "epoch": 0.3119797654633249,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.902769299473803e-05,
+      "loss": 0.7757,
+      "step": 4240
+    },
+    {
+      "epoch": 0.3120533455966889,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.9022898667566686e-05,
+      "loss": 0.7457,
+      "step": 4241
+    },
+    {
+      "epoch": 0.3121269257300529,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.901810358781286e-05,
+      "loss": 1.1793,
+      "step": 4242
+    },
+    {
+      "epoch": 0.31220050586341686,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.901330775573389e-05,
+      "loss": 0.9182,
+      "step": 4243
+    },
+    {
+      "epoch": 0.31227408599678086,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.9008511171587145e-05,
+      "loss": 1.2152,
+      "step": 4244
+    },
+    {
+      "epoch": 0.31234766613014486,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.900371383563008e-05,
+      "loss": 0.9771,
+      "step": 4245
+    },
+    {
+      "epoch": 0.31242124626350887,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.899891574812014e-05,
+      "loss": 0.7187,
+      "step": 4246
+    },
+    {
+      "epoch": 0.31249482639687287,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.899411690931482e-05,
+      "loss": 1.0377,
+      "step": 4247
+    },
+    {
+      "epoch": 0.3125684065302368,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.89893173194717e-05,
+      "loss": 1.0612,
+      "step": 4248
+    },
+    {
+      "epoch": 0.3126419866636008,
+      "grad_norm": 1.78125,
+      "learning_rate": 3.8984516978848326e-05,
+      "loss": 1.1835,
+      "step": 4249
+    },
+    {
+      "epoch": 0.31271556679696483,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.8979715887702336e-05,
+      "loss": 1.229,
+      "step": 4250
+    },
+    {
+      "epoch": 0.31278914693032883,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.89749140462914e-05,
+      "loss": 0.6719,
+      "step": 4251
+    },
+    {
+      "epoch": 0.3128627270636928,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.8970111454873225e-05,
+      "loss": 0.9952,
+      "step": 4252
+    },
+    {
+      "epoch": 0.3129363071970568,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.8965308113705553e-05,
+      "loss": 0.7742,
+      "step": 4253
+    },
+    {
+      "epoch": 0.3130098873304208,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.896050402304618e-05,
+      "loss": 0.9227,
+      "step": 4254
+    },
+    {
+      "epoch": 0.3130834674637848,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.895569918315292e-05,
+      "loss": 1.1841,
+      "step": 4255
+    },
+    {
+      "epoch": 0.3131570475971488,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.8950893594283636e-05,
+      "loss": 1.2208,
+      "step": 4256
+    },
+    {
+      "epoch": 0.31323062773051275,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.894608725669624e-05,
+      "loss": 0.7798,
+      "step": 4257
+    },
+    {
+      "epoch": 0.31330420786387675,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.894128017064869e-05,
+      "loss": 0.8113,
+      "step": 4258
+    },
+    {
+      "epoch": 0.31337778799724075,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.8936472336398965e-05,
+      "loss": 1.0558,
+      "step": 4259
+    },
+    {
+      "epoch": 0.31345136813060476,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.8931663754205086e-05,
+      "loss": 0.8551,
+      "step": 4260
+    },
+    {
+      "epoch": 0.3135249482639687,
+      "grad_norm": 1.4375,
+      "learning_rate": 3.8926854424325135e-05,
+      "loss": 0.9985,
+      "step": 4261
+    },
+    {
+      "epoch": 0.3135985283973327,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.89220443470172e-05,
+      "loss": 0.6664,
+      "step": 4262
+    },
+    {
+      "epoch": 0.3136721085306967,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.891723352253944e-05,
+      "loss": 0.9043,
+      "step": 4263
+    },
+    {
+      "epoch": 0.3137456886640607,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.8912421951150055e-05,
+      "loss": 1.2252,
+      "step": 4264
+    },
+    {
+      "epoch": 0.3138192687974247,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.890760963310725e-05,
+      "loss": 0.7947,
+      "step": 4265
+    },
+    {
+      "epoch": 0.31389284893078867,
+      "grad_norm": 1.0,
+      "learning_rate": 3.890279656866931e-05,
+      "loss": 0.916,
+      "step": 4266
+    },
+    {
+      "epoch": 0.3139664290641527,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.889798275809453e-05,
+      "loss": 0.6326,
+      "step": 4267
+    },
+    {
+      "epoch": 0.3140400091975167,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.889316820164127e-05,
+      "loss": 0.8687,
+      "step": 4268
+    },
+    {
+      "epoch": 0.3141135893308807,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.888835289956792e-05,
+      "loss": 0.9091,
+      "step": 4269
+    },
+    {
+      "epoch": 0.31418716946424463,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.888353685213289e-05,
+      "loss": 1.1087,
+      "step": 4270
+    },
+    {
+      "epoch": 0.31426074959760864,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.887872005959466e-05,
+      "loss": 1.0186,
+      "step": 4271
+    },
+    {
+      "epoch": 0.31433432973097264,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.887390252221174e-05,
+      "loss": 0.7486,
+      "step": 4272
+    },
+    {
+      "epoch": 0.31440790986433664,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.886908424024268e-05,
+      "loss": 1.0472,
+      "step": 4273
+    },
+    {
+      "epoch": 0.31448148999770065,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.886426521394606e-05,
+      "loss": 0.5371,
+      "step": 4274
+    },
+    {
+      "epoch": 0.3145550701310646,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.885944544358051e-05,
+      "loss": 0.7701,
+      "step": 4275
+    },
+    {
+      "epoch": 0.3146286502644286,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.8854624929404704e-05,
+      "loss": 1.0232,
+      "step": 4276
+    },
+    {
+      "epoch": 0.3147022303977926,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.8849803671677344e-05,
+      "loss": 1.0513,
+      "step": 4277
+    },
+    {
+      "epoch": 0.3147758105311566,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.8844981670657174e-05,
+      "loss": 0.8778,
+      "step": 4278
+    },
+    {
+      "epoch": 0.31484939066452056,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.8840158926603e-05,
+      "loss": 0.7227,
+      "step": 4279
+    },
+    {
+      "epoch": 0.31492297079788456,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.8835335439773624e-05,
+      "loss": 1.1575,
+      "step": 4280
+    },
+    {
+      "epoch": 0.31499655093124856,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.883051121042793e-05,
+      "loss": 1.0151,
+      "step": 4281
+    },
+    {
+      "epoch": 0.31507013106461257,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.882568623882482e-05,
+      "loss": 0.4754,
+      "step": 4282
+    },
+    {
+      "epoch": 0.3151437111979766,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.882086052522325e-05,
+      "loss": 1.3109,
+      "step": 4283
+    },
+    {
+      "epoch": 0.3152172913313405,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.881603406988219e-05,
+      "loss": 0.7013,
+      "step": 4284
+    },
+    {
+      "epoch": 0.3152908714647045,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.881120687306068e-05,
+      "loss": 1.0863,
+      "step": 4285
+    },
+    {
+      "epoch": 0.31536445159806853,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.8806378935017785e-05,
+      "loss": 0.7815,
+      "step": 4286
+    },
+    {
+      "epoch": 0.31543803173143253,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.8801550256012606e-05,
+      "loss": 0.9918,
+      "step": 4287
+    },
+    {
+      "epoch": 0.3155116118647965,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.87967208363043e-05,
+      "loss": 0.7851,
+      "step": 4288
+    },
+    {
+      "epoch": 0.3155851919981605,
+      "grad_norm": 1.8359375,
+      "learning_rate": 3.8791890676152036e-05,
+      "loss": 1.0747,
+      "step": 4289
+    },
+    {
+      "epoch": 0.3156587721315245,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.8787059775815055e-05,
+      "loss": 1.0735,
+      "step": 4290
+    },
+    {
+      "epoch": 0.3157323522648885,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.878222813555261e-05,
+      "loss": 0.8061,
+      "step": 4291
+    },
+    {
+      "epoch": 0.3158059323982525,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.877739575562401e-05,
+      "loss": 1.1215,
+      "step": 4292
+    },
+    {
+      "epoch": 0.31587951253161645,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.8772562636288614e-05,
+      "loss": 0.868,
+      "step": 4293
+    },
+    {
+      "epoch": 0.31595309266498045,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.876772877780578e-05,
+      "loss": 1.1162,
+      "step": 4294
+    },
+    {
+      "epoch": 0.31602667279834445,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.8762894180434956e-05,
+      "loss": 0.6902,
+      "step": 4295
+    },
+    {
+      "epoch": 0.31610025293170846,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.875805884443559e-05,
+      "loss": 0.8879,
+      "step": 4296
+    },
+    {
+      "epoch": 0.3161738330650724,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.875322277006719e-05,
+      "loss": 0.8738,
+      "step": 4297
+    },
+    {
+      "epoch": 0.3162474131984364,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.874838595758931e-05,
+      "loss": 0.9271,
+      "step": 4298
+    },
+    {
+      "epoch": 0.3163209933318004,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.874354840726151e-05,
+      "loss": 0.8894,
+      "step": 4299
+    },
+    {
+      "epoch": 0.3163945734651644,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.8738710119343435e-05,
+      "loss": 0.9597,
+      "step": 4300
+    },
+    {
+      "epoch": 0.3164681535985284,
+      "grad_norm": 1.125,
+      "learning_rate": 3.873387109409473e-05,
+      "loss": 0.971,
+      "step": 4301
+    },
+    {
+      "epoch": 0.31654173373189237,
+      "grad_norm": 0.875,
+      "learning_rate": 3.872903133177511e-05,
+      "loss": 0.867,
+      "step": 4302
+    },
+    {
+      "epoch": 0.3166153138652564,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.8724190832644294e-05,
+      "loss": 1.3698,
+      "step": 4303
+    },
+    {
+      "epoch": 0.3166888939986204,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.87193495969621e-05,
+      "loss": 1.3375,
+      "step": 4304
+    },
+    {
+      "epoch": 0.3167624741319844,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.8714507624988304e-05,
+      "loss": 1.3526,
+      "step": 4305
+    },
+    {
+      "epoch": 0.31683605426534833,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.87096649169828e-05,
+      "loss": 1.6838,
+      "step": 4306
+    },
+    {
+      "epoch": 0.31690963439871234,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.8704821473205466e-05,
+      "loss": 0.5086,
+      "step": 4307
+    },
+    {
+      "epoch": 0.31698321453207634,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.869997729391625e-05,
+      "loss": 1.0475,
+      "step": 4308
+    },
+    {
+      "epoch": 0.31705679466544034,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.869513237937513e-05,
+      "loss": 0.8904,
+      "step": 4309
+    },
+    {
+      "epoch": 0.31713037479880435,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.869028672984212e-05,
+      "loss": 0.7615,
+      "step": 4310
+    },
+    {
+      "epoch": 0.3172039549321683,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.868544034557728e-05,
+      "loss": 1.0894,
+      "step": 4311
+    },
+    {
+      "epoch": 0.3172775350655323,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.86805932268407e-05,
+      "loss": 0.9805,
+      "step": 4312
+    },
+    {
+      "epoch": 0.3173511151988963,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.867574537389253e-05,
+      "loss": 0.9955,
+      "step": 4313
+    },
+    {
+      "epoch": 0.3174246953322603,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.8670896786992926e-05,
+      "loss": 1.1447,
+      "step": 4314
+    },
+    {
+      "epoch": 0.31749827546562426,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.8666047466402125e-05,
+      "loss": 1.008,
+      "step": 4315
+    },
+    {
+      "epoch": 0.31757185559898826,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.8661197412380356e-05,
+      "loss": 1.0325,
+      "step": 4316
+    },
+    {
+      "epoch": 0.31764543573235227,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.865634662518792e-05,
+      "loss": 0.8053,
+      "step": 4317
+    },
+    {
+      "epoch": 0.31771901586571627,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.8651495105085164e-05,
+      "loss": 1.0136,
+      "step": 4318
+    },
+    {
+      "epoch": 0.3177925959990803,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.864664285233245e-05,
+      "loss": 0.8909,
+      "step": 4319
+    },
+    {
+      "epoch": 0.3178661761324442,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.864178986719019e-05,
+      "loss": 0.923,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3179397562658082,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.863693614991883e-05,
+      "loss": 1.127,
+      "step": 4321
+    },
+    {
+      "epoch": 0.31801333639917223,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.8632081700778865e-05,
+      "loss": 0.8364,
+      "step": 4322
+    },
+    {
+      "epoch": 0.31808691653253623,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.8627226520030816e-05,
+      "loss": 1.1214,
+      "step": 4323
+    },
+    {
+      "epoch": 0.3181604966659002,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.862237060793526e-05,
+      "loss": 0.765,
+      "step": 4324
+    },
+    {
+      "epoch": 0.3182340767992642,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.861751396475281e-05,
+      "loss": 0.951,
+      "step": 4325
+    },
+    {
+      "epoch": 0.3183076569326282,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.86126565907441e-05,
+      "loss": 0.9109,
+      "step": 4326
+    },
+    {
+      "epoch": 0.3183812370659922,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.860779848616982e-05,
+      "loss": 0.7895,
+      "step": 4327
+    },
+    {
+      "epoch": 0.3184548171993562,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.86029396512907e-05,
+      "loss": 0.8631,
+      "step": 4328
+    },
+    {
+      "epoch": 0.31852839733272015,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.85980800863675e-05,
+      "loss": 0.6359,
+      "step": 4329
+    },
+    {
+      "epoch": 0.31860197746608415,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.859321979166102e-05,
+      "loss": 1.3715,
+      "step": 4330
+    },
+    {
+      "epoch": 0.31867555759944816,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.858835876743211e-05,
+      "loss": 0.9113,
+      "step": 4331
+    },
+    {
+      "epoch": 0.31874913773281216,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.858349701394166e-05,
+      "loss": 1.0455,
+      "step": 4332
+    },
+    {
+      "epoch": 0.3188227178661761,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.857863453145057e-05,
+      "loss": 0.9149,
+      "step": 4333
+    },
+    {
+      "epoch": 0.3188962979995401,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.857377132021982e-05,
+      "loss": 0.84,
+      "step": 4334
+    },
+    {
+      "epoch": 0.3189698781329041,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.856890738051039e-05,
+      "loss": 0.6142,
+      "step": 4335
+    },
+    {
+      "epoch": 0.3190434582662681,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.856404271258334e-05,
+      "loss": 0.9173,
+      "step": 4336
+    },
+    {
+      "epoch": 0.3191170383996321,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.855917731669972e-05,
+      "loss": 0.9181,
+      "step": 4337
+    },
+    {
+      "epoch": 0.3191906185329961,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.8554311193120674e-05,
+      "loss": 0.9264,
+      "step": 4338
+    },
+    {
+      "epoch": 0.3192641986663601,
+      "grad_norm": 1.0,
+      "learning_rate": 3.8549444342107345e-05,
+      "loss": 1.0953,
+      "step": 4339
+    },
+    {
+      "epoch": 0.3193377787997241,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.8544576763920934e-05,
+      "loss": 0.9389,
+      "step": 4340
+    },
+    {
+      "epoch": 0.3194113589330881,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.8539708458822665e-05,
+      "loss": 0.9894,
+      "step": 4341
+    },
+    {
+      "epoch": 0.31948493906645203,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.853483942707382e-05,
+      "loss": 0.7656,
+      "step": 4342
+    },
+    {
+      "epoch": 0.31955851919981604,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.85299696689357e-05,
+      "loss": 0.6863,
+      "step": 4343
+    },
+    {
+      "epoch": 0.31963209933318004,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.852509918466967e-05,
+      "loss": 1.1532,
+      "step": 4344
+    },
+    {
+      "epoch": 0.31970567946654405,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.8520227974537106e-05,
+      "loss": 0.8935,
+      "step": 4345
+    },
+    {
+      "epoch": 0.31977925959990805,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.851535603879944e-05,
+      "loss": 0.7089,
+      "step": 4346
+    },
+    {
+      "epoch": 0.319852839733272,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.851048337771815e-05,
+      "loss": 0.975,
+      "step": 4347
+    },
+    {
+      "epoch": 0.319926419866636,
+      "grad_norm": 1.5546875,
+      "learning_rate": 3.850560999155473e-05,
+      "loss": 1.3401,
+      "step": 4348
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.8500735880570734e-05,
+      "loss": 1.1325,
+      "step": 4349
+    },
+    {
+      "epoch": 0.320073580133364,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.849586104502774e-05,
+      "loss": 0.852,
+      "step": 4350
+    },
+    {
+      "epoch": 0.32014716026672796,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.849098548518738e-05,
+      "loss": 0.7993,
+      "step": 4351
+    },
+    {
+      "epoch": 0.32022074040009196,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.84861092013113e-05,
+      "loss": 0.8838,
+      "step": 4352
+    },
+    {
+      "epoch": 0.32029432053345597,
+      "grad_norm": 1.0,
+      "learning_rate": 3.848123219366121e-05,
+      "loss": 0.8858,
+      "step": 4353
+    },
+    {
+      "epoch": 0.32036790066681997,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.847635446249886e-05,
+      "loss": 0.8729,
+      "step": 4354
+    },
+    {
+      "epoch": 0.320441480800184,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.8471476008086014e-05,
+      "loss": 0.689,
+      "step": 4355
+    },
+    {
+      "epoch": 0.3205150609335479,
+      "grad_norm": 1.0,
+      "learning_rate": 3.84665968306845e-05,
+      "loss": 1.2473,
+      "step": 4356
+    },
+    {
+      "epoch": 0.32058864106691193,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.8461716930556164e-05,
+      "loss": 0.8909,
+      "step": 4357
+    },
+    {
+      "epoch": 0.32066222120027593,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.845683630796291e-05,
+      "loss": 1.0895,
+      "step": 4358
+    },
+    {
+      "epoch": 0.32073580133363994,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.845195496316666e-05,
+      "loss": 1.2606,
+      "step": 4359
+    },
+    {
+      "epoch": 0.3208093814670039,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.844707289642939e-05,
+      "loss": 0.7466,
+      "step": 4360
+    },
+    {
+      "epoch": 0.3208829616003679,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.8442190108013135e-05,
+      "loss": 1.5815,
+      "step": 4361
+    },
+    {
+      "epoch": 0.3209565417337319,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.843730659817991e-05,
+      "loss": 0.8769,
+      "step": 4362
+    },
+    {
+      "epoch": 0.3210301218670959,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.843242236719182e-05,
+      "loss": 0.7549,
+      "step": 4363
+    },
+    {
+      "epoch": 0.3211037020004599,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.8427537415311e-05,
+      "loss": 0.7548,
+      "step": 4364
+    },
+    {
+      "epoch": 0.32117728213382385,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.84226517427996e-05,
+      "loss": 0.6908,
+      "step": 4365
+    },
+    {
+      "epoch": 0.32125086226718785,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.8417765349919835e-05,
+      "loss": 0.8573,
+      "step": 4366
+    },
+    {
+      "epoch": 0.32132444240055186,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.8412878236933946e-05,
+      "loss": 0.8099,
+      "step": 4367
+    },
+    {
+      "epoch": 0.32139802253391586,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.840799040410422e-05,
+      "loss": 0.7745,
+      "step": 4368
+    },
+    {
+      "epoch": 0.3214716026672798,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.840310185169297e-05,
+      "loss": 0.8537,
+      "step": 4369
+    },
+    {
+      "epoch": 0.3215451828006438,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.839821257996256e-05,
+      "loss": 0.6836,
+      "step": 4370
+    },
+    {
+      "epoch": 0.3216187629340078,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.839332258917539e-05,
+      "loss": 0.7052,
+      "step": 4371
+    },
+    {
+      "epoch": 0.3216923430673718,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.838843187959388e-05,
+      "loss": 1.4513,
+      "step": 4372
+    },
+    {
+      "epoch": 0.3217659232007358,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.8383540451480527e-05,
+      "loss": 0.8022,
+      "step": 4373
+    },
+    {
+      "epoch": 0.3218395033340998,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.837864830509783e-05,
+      "loss": 1.1341,
+      "step": 4374
+    },
+    {
+      "epoch": 0.3219130834674638,
+      "grad_norm": 0.875,
+      "learning_rate": 3.837375544070836e-05,
+      "loss": 0.7469,
+      "step": 4375
+    },
+    {
+      "epoch": 0.3219866636008278,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.8368861858574684e-05,
+      "loss": 0.8379,
+      "step": 4376
+    },
+    {
+      "epoch": 0.3220602437341918,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.8363967558959444e-05,
+      "loss": 1.1046,
+      "step": 4377
+    },
+    {
+      "epoch": 0.32213382386755574,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.8359072542125305e-05,
+      "loss": 0.7402,
+      "step": 4378
+    },
+    {
+      "epoch": 0.32220740400091974,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.835417680833499e-05,
+      "loss": 1.3135,
+      "step": 4379
+    },
+    {
+      "epoch": 0.32228098413428374,
+      "grad_norm": 1.140625,
+      "learning_rate": 3.83492803578512e-05,
+      "loss": 1.6153,
+      "step": 4380
+    },
+    {
+      "epoch": 0.32235456426764775,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.834438319093676e-05,
+      "loss": 0.6938,
+      "step": 4381
+    },
+    {
+      "epoch": 0.32242814440101175,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.8339485307854484e-05,
+      "loss": 1.081,
+      "step": 4382
+    },
+    {
+      "epoch": 0.3225017245343757,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.833458670886723e-05,
+      "loss": 0.8331,
+      "step": 4383
+    },
+    {
+      "epoch": 0.3225753046677397,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.8329687394237876e-05,
+      "loss": 0.7388,
+      "step": 4384
+    },
+    {
+      "epoch": 0.3226488848011037,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.8324787364229386e-05,
+      "loss": 1.7329,
+      "step": 4385
+    },
+    {
+      "epoch": 0.3227224649344677,
+      "grad_norm": 1.1484375,
+      "learning_rate": 3.831988661910473e-05,
+      "loss": 1.2951,
+      "step": 4386
+    },
+    {
+      "epoch": 0.32279604506783166,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.83149851591269e-05,
+      "loss": 0.8811,
+      "step": 4387
+    },
+    {
+      "epoch": 0.32286962520119566,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.8310082984558973e-05,
+      "loss": 0.7405,
+      "step": 4388
+    },
+    {
+      "epoch": 0.32294320533455967,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.830518009566404e-05,
+      "loss": 0.7229,
+      "step": 4389
+    },
+    {
+      "epoch": 0.3230167854679237,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.830027649270521e-05,
+      "loss": 1.0959,
+      "step": 4390
+    },
+    {
+      "epoch": 0.3230903656012877,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.829537217594567e-05,
+      "loss": 1.6317,
+      "step": 4391
+    },
+    {
+      "epoch": 0.3231639457346516,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.8290467145648615e-05,
+      "loss": 0.97,
+      "step": 4392
+    },
+    {
+      "epoch": 0.32323752586801563,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.828556140207729e-05,
+      "loss": 1.2009,
+      "step": 4393
+    },
+    {
+      "epoch": 0.32331110600137963,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.828065494549497e-05,
+      "loss": 1.0788,
+      "step": 4394
+    },
+    {
+      "epoch": 0.32338468613474364,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.827574777616499e-05,
+      "loss": 0.824,
+      "step": 4395
+    },
+    {
+      "epoch": 0.3234582662681076,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.8270839894350694e-05,
+      "loss": 0.7157,
+      "step": 4396
+    },
+    {
+      "epoch": 0.3235318464014716,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.826593130031549e-05,
+      "loss": 0.9199,
+      "step": 4397
+    },
+    {
+      "epoch": 0.3236054265348356,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.826102199432282e-05,
+      "loss": 1.3482,
+      "step": 4398
+    },
+    {
+      "epoch": 0.3236790066681996,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.8256111976636135e-05,
+      "loss": 0.672,
+      "step": 4399
+    },
+    {
+      "epoch": 0.3237525868015636,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.825120124751895e-05,
+      "loss": 1.0889,
+      "step": 4400
+    },
+    {
+      "epoch": 0.32382616693492755,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.8246289807234837e-05,
+      "loss": 0.9362,
+      "step": 4401
+    },
+    {
+      "epoch": 0.32389974706829155,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.824137765604736e-05,
+      "loss": 0.558,
+      "step": 4402
+    },
+    {
+      "epoch": 0.32397332720165556,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.823646479422016e-05,
+      "loss": 0.6618,
+      "step": 4403
+    },
+    {
+      "epoch": 0.32404690733501956,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.8231551222016884e-05,
+      "loss": 0.7082,
+      "step": 4404
+    },
+    {
+      "epoch": 0.3241204874683835,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.822663693970126e-05,
+      "loss": 0.8813,
+      "step": 4405
+    },
+    {
+      "epoch": 0.3241940676017475,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.8221721947537e-05,
+      "loss": 0.7983,
+      "step": 4406
+    },
+    {
+      "epoch": 0.3242676477351115,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.82168062457879e-05,
+      "loss": 0.8671,
+      "step": 4407
+    },
+    {
+      "epoch": 0.3243412278684755,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.8211889834717775e-05,
+      "loss": 0.64,
+      "step": 4408
+    },
+    {
+      "epoch": 0.3244148080018395,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.8206972714590474e-05,
+      "loss": 0.9115,
+      "step": 4409
+    },
+    {
+      "epoch": 0.3244883881352035,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.820205488566989e-05,
+      "loss": 1.4206,
+      "step": 4410
+    },
+    {
+      "epoch": 0.3245619682685675,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.819713634821995e-05,
+      "loss": 0.6944,
+      "step": 4411
+    },
+    {
+      "epoch": 0.3246355484019315,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.819221710250464e-05,
+      "loss": 0.614,
+      "step": 4412
+    },
+    {
+      "epoch": 0.3247091285352955,
+      "grad_norm": 1.3984375,
+      "learning_rate": 3.818729714878795e-05,
+      "loss": 0.764,
+      "step": 4413
+    },
+    {
+      "epoch": 0.32478270866865944,
+      "grad_norm": 1.2109375,
+      "learning_rate": 3.8182376487333925e-05,
+      "loss": 1.1189,
+      "step": 4414
+    },
+    {
+      "epoch": 0.32485628880202344,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.817745511840665e-05,
+      "loss": 0.6934,
+      "step": 4415
+    },
+    {
+      "epoch": 0.32492986893538744,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.8172533042270255e-05,
+      "loss": 1.1779,
+      "step": 4416
+    },
+    {
+      "epoch": 0.32500344906875145,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.816761025918889e-05,
+      "loss": 1.0069,
+      "step": 4417
+    },
+    {
+      "epoch": 0.32507702920211545,
+      "grad_norm": 1.140625,
+      "learning_rate": 3.816268676942676e-05,
+      "loss": 1.2744,
+      "step": 4418
+    },
+    {
+      "epoch": 0.3251506093354794,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.815776257324808e-05,
+      "loss": 0.8949,
+      "step": 4419
+    },
+    {
+      "epoch": 0.3252241894688434,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.815283767091713e-05,
+      "loss": 0.8184,
+      "step": 4420
+    },
+    {
+      "epoch": 0.3252977696022074,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.814791206269823e-05,
+      "loss": 0.9131,
+      "step": 4421
+    },
+    {
+      "epoch": 0.3253713497355714,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.8142985748855726e-05,
+      "loss": 1.1168,
+      "step": 4422
+    },
+    {
+      "epoch": 0.32544492986893536,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.8138058729654e-05,
+      "loss": 0.8302,
+      "step": 4423
+    },
+    {
+      "epoch": 0.32551851000229937,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.813313100535747e-05,
+      "loss": 0.8193,
+      "step": 4424
+    },
+    {
+      "epoch": 0.32559209013566337,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.8128202576230616e-05,
+      "loss": 1.2543,
+      "step": 4425
+    },
+    {
+      "epoch": 0.3256656702690274,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.8123273442537914e-05,
+      "loss": 1.0079,
+      "step": 4426
+    },
+    {
+      "epoch": 0.3257392504023914,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.8118343604543916e-05,
+      "loss": 0.7694,
+      "step": 4427
+    },
+    {
+      "epoch": 0.3258128305357553,
+      "grad_norm": 1.1875,
+      "learning_rate": 3.8113413062513195e-05,
+      "loss": 0.9604,
+      "step": 4428
+    },
+    {
+      "epoch": 0.32588641066911933,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.8108481816710365e-05,
+      "loss": 1.205,
+      "step": 4429
+    },
+    {
+      "epoch": 0.32595999080248333,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.810354986740007e-05,
+      "loss": 1.4123,
+      "step": 4430
+    },
+    {
+      "epoch": 0.32603357093584734,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.8098617214847e-05,
+      "loss": 1.187,
+      "step": 4431
+    },
+    {
+      "epoch": 0.3261071510692113,
+      "grad_norm": 0.75,
+      "learning_rate": 3.809368385931589e-05,
+      "loss": 0.8348,
+      "step": 4432
+    },
+    {
+      "epoch": 0.3261807312025753,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.80887498010715e-05,
+      "loss": 1.1723,
+      "step": 4433
+    },
+    {
+      "epoch": 0.3262543113359393,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.808381504037862e-05,
+      "loss": 0.6599,
+      "step": 4434
+    },
+    {
+      "epoch": 0.3263278914693033,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.8078879577502105e-05,
+      "loss": 0.7088,
+      "step": 4435
+    },
+    {
+      "epoch": 0.3264014716026673,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.807394341270682e-05,
+      "loss": 0.7271,
+      "step": 4436
+    },
+    {
+      "epoch": 0.32647505173603125,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.806900654625769e-05,
+      "loss": 0.7788,
+      "step": 4437
+    },
+    {
+      "epoch": 0.32654863186939526,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.806406897841966e-05,
+      "loss": 1.0432,
+      "step": 4438
+    },
+    {
+      "epoch": 0.32662221200275926,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.805913070945773e-05,
+      "loss": 1.1296,
+      "step": 4439
+    },
+    {
+      "epoch": 0.32669579213612326,
+      "grad_norm": 0.75,
+      "learning_rate": 3.8054191739636916e-05,
+      "loss": 0.849,
+      "step": 4440
+    },
+    {
+      "epoch": 0.3267693722694872,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.804925206922229e-05,
+      "loss": 0.8604,
+      "step": 4441
+    },
+    {
+      "epoch": 0.3268429524028512,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.804431169847894e-05,
+      "loss": 0.853,
+      "step": 4442
+    },
+    {
+      "epoch": 0.3269165325362152,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.8039370627672026e-05,
+      "loss": 0.747,
+      "step": 4443
+    },
+    {
+      "epoch": 0.3269901126695792,
+      "grad_norm": 1.71875,
+      "learning_rate": 3.8034428857066716e-05,
+      "loss": 0.8982,
+      "step": 4444
+    },
+    {
+      "epoch": 0.32706369280294323,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.802948638692823e-05,
+      "loss": 0.8946,
+      "step": 4445
+    },
+    {
+      "epoch": 0.3271372729363072,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.802454321752182e-05,
+      "loss": 1.32,
+      "step": 4446
+    },
+    {
+      "epoch": 0.3272108530696712,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.801959934911277e-05,
+      "loss": 0.8762,
+      "step": 4447
+    },
+    {
+      "epoch": 0.3272844332030352,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.801465478196642e-05,
+      "loss": 1.0046,
+      "step": 4448
+    },
+    {
+      "epoch": 0.3273580133363992,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.800970951634812e-05,
+      "loss": 0.6777,
+      "step": 4449
+    },
+    {
+      "epoch": 0.32743159346976314,
+      "grad_norm": 0.75,
+      "learning_rate": 3.8004763552523296e-05,
+      "loss": 0.9015,
+      "step": 4450
+    },
+    {
+      "epoch": 0.32750517360312714,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.799981689075737e-05,
+      "loss": 0.7905,
+      "step": 4451
+    },
+    {
+      "epoch": 0.32757875373649115,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.799486953131582e-05,
+      "loss": 1.1968,
+      "step": 4452
+    },
+    {
+      "epoch": 0.32765233386985515,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.798992147446417e-05,
+      "loss": 0.8914,
+      "step": 4453
+    },
+    {
+      "epoch": 0.32772591400321915,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.798497272046796e-05,
+      "loss": 0.8838,
+      "step": 4454
+    },
+    {
+      "epoch": 0.3277994941365831,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.79800232695928e-05,
+      "loss": 0.7554,
+      "step": 4455
+    },
+    {
+      "epoch": 0.3278730742699471,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.79750731221043e-05,
+      "loss": 0.6986,
+      "step": 4456
+    },
+    {
+      "epoch": 0.3279466544033111,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.7970122278268136e-05,
+      "loss": 1.3797,
+      "step": 4457
+    },
+    {
+      "epoch": 0.3280202345366751,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.7965170738350006e-05,
+      "loss": 0.8397,
+      "step": 4458
+    },
+    {
+      "epoch": 0.32809381467003906,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.7960218502615655e-05,
+      "loss": 0.8418,
+      "step": 4459
+    },
+    {
+      "epoch": 0.32816739480340307,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.795526557133085e-05,
+      "loss": 0.946,
+      "step": 4460
+    },
+    {
+      "epoch": 0.32824097493676707,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.7950311944761405e-05,
+      "loss": 1.3836,
+      "step": 4461
+    },
+    {
+      "epoch": 0.3283145550701311,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.7945357623173185e-05,
+      "loss": 0.8702,
+      "step": 4462
+    },
+    {
+      "epoch": 0.3283881352034951,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.794040260683207e-05,
+      "loss": 0.7486,
+      "step": 4463
+    },
+    {
+      "epoch": 0.32846171533685903,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.793544689600399e-05,
+      "loss": 0.8161,
+      "step": 4464
+    },
+    {
+      "epoch": 0.32853529547022303,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.7930490490954904e-05,
+      "loss": 1.0262,
+      "step": 4465
+    },
+    {
+      "epoch": 0.32860887560358704,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.792553339195082e-05,
+      "loss": 0.7622,
+      "step": 4466
+    },
+    {
+      "epoch": 0.32868245573695104,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.7920575599257754e-05,
+      "loss": 0.7319,
+      "step": 4467
+    },
+    {
+      "epoch": 0.328756035870315,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.791561711314182e-05,
+      "loss": 0.994,
+      "step": 4468
+    },
+    {
+      "epoch": 0.328829616003679,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.79106579338691e-05,
+      "loss": 0.9103,
+      "step": 4469
+    },
+    {
+      "epoch": 0.328903196137043,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.790569806170576e-05,
+      "loss": 0.8773,
+      "step": 4470
+    },
+    {
+      "epoch": 0.328976776270407,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.790073749691798e-05,
+      "loss": 1.2341,
+      "step": 4471
+    },
+    {
+      "epoch": 0.329050356403771,
+      "grad_norm": 0.75,
+      "learning_rate": 3.789577623977198e-05,
+      "loss": 0.6619,
+      "step": 4472
+    },
+    {
+      "epoch": 0.32912393653713495,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.789081429053403e-05,
+      "loss": 0.7464,
+      "step": 4473
+    },
+    {
+      "epoch": 0.32919751667049896,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.7885851649470415e-05,
+      "loss": 0.8021,
+      "step": 4474
+    },
+    {
+      "epoch": 0.32927109680386296,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.788088831684749e-05,
+      "loss": 0.8931,
+      "step": 4475
+    },
+    {
+      "epoch": 0.32934467693722697,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.787592429293161e-05,
+      "loss": 0.9185,
+      "step": 4476
+    },
+    {
+      "epoch": 0.3294182570705909,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.787095957798919e-05,
+      "loss": 0.9234,
+      "step": 4477
+    },
+    {
+      "epoch": 0.3294918372039549,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.78659941722867e-05,
+      "loss": 0.9418,
+      "step": 4478
+    },
+    {
+      "epoch": 0.3295654173373189,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.786102807609059e-05,
+      "loss": 1.0555,
+      "step": 4479
+    },
+    {
+      "epoch": 0.3296389974706829,
+      "grad_norm": 0.875,
+      "learning_rate": 3.785606128966739e-05,
+      "loss": 0.8578,
+      "step": 4480
+    },
+    {
+      "epoch": 0.32971257760404693,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.785109381328365e-05,
+      "loss": 0.7296,
+      "step": 4481
+    },
+    {
+      "epoch": 0.3297861577374109,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.7846125647206e-05,
+      "loss": 0.9887,
+      "step": 4482
+    },
+    {
+      "epoch": 0.3298597378707749,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.784115679170105e-05,
+      "loss": 1.4146,
+      "step": 4483
+    },
+    {
+      "epoch": 0.3299333180041389,
+      "grad_norm": 1.453125,
+      "learning_rate": 3.783618724703546e-05,
+      "loss": 0.9495,
+      "step": 4484
+    },
+    {
+      "epoch": 0.3300068981375029,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.783121701347594e-05,
+      "loss": 1.0103,
+      "step": 4485
+    },
+    {
+      "epoch": 0.33008047827086684,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.782624609128924e-05,
+      "loss": 1.0129,
+      "step": 4486
+    },
+    {
+      "epoch": 0.33015405840423084,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.782127448074214e-05,
+      "loss": 0.658,
+      "step": 4487
+    },
+    {
+      "epoch": 0.33022763853759485,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.7816302182101456e-05,
+      "loss": 1.1676,
+      "step": 4488
+    },
+    {
+      "epoch": 0.33030121867095885,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.7811329195634044e-05,
+      "loss": 0.7587,
+      "step": 4489
+    },
+    {
+      "epoch": 0.33037479880432286,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.7806355521606787e-05,
+      "loss": 1.4104,
+      "step": 4490
+    },
+    {
+      "epoch": 0.3304483789376868,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.780138116028662e-05,
+      "loss": 1.2279,
+      "step": 4491
+    },
+    {
+      "epoch": 0.3305219590710508,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.77964061119405e-05,
+      "loss": 1.0594,
+      "step": 4492
+    },
+    {
+      "epoch": 0.3305955392044148,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.7791430376835425e-05,
+      "loss": 0.8484,
+      "step": 4493
+    },
+    {
+      "epoch": 0.3306691193377788,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.778645395523845e-05,
+      "loss": 1.0097,
+      "step": 4494
+    },
+    {
+      "epoch": 0.33074269947114276,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.7781476847416645e-05,
+      "loss": 0.8819,
+      "step": 4495
+    },
+    {
+      "epoch": 0.33081627960450677,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.777649905363712e-05,
+      "loss": 1.1707,
+      "step": 4496
+    },
+    {
+      "epoch": 0.3308898597378708,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.7771520574167006e-05,
+      "loss": 0.679,
+      "step": 4497
+    },
+    {
+      "epoch": 0.3309634398712348,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.7766541409273524e-05,
+      "loss": 0.758,
+      "step": 4498
+    },
+    {
+      "epoch": 0.3310370200045988,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.7761561559223865e-05,
+      "loss": 0.8395,
+      "step": 4499
+    },
+    {
+      "epoch": 0.33111060013796273,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.77565810242853e-05,
+      "loss": 0.7264,
+      "step": 4500
+    },
+    {
+      "epoch": 0.33118418027132673,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.775159980472513e-05,
+      "loss": 1.4052,
+      "step": 4501
+    },
+    {
+      "epoch": 0.33125776040469074,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.7746617900810685e-05,
+      "loss": 0.8163,
+      "step": 4502
+    },
+    {
+      "epoch": 0.33133134053805474,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.774163531280933e-05,
+      "loss": 0.9579,
+      "step": 4503
+    },
+    {
+      "epoch": 0.3314049206714187,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.7736652040988476e-05,
+      "loss": 0.6861,
+      "step": 4504
+    },
+    {
+      "epoch": 0.3314785008047827,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.773166808561556e-05,
+      "loss": 0.6254,
+      "step": 4505
+    },
+    {
+      "epoch": 0.3315520809381467,
+      "grad_norm": 0.75,
+      "learning_rate": 3.772668344695806e-05,
+      "loss": 0.9836,
+      "step": 4506
+    },
+    {
+      "epoch": 0.3316256610715107,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.77216981252835e-05,
+      "loss": 0.708,
+      "step": 4507
+    },
+    {
+      "epoch": 0.3316992412048747,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.7716712120859435e-05,
+      "loss": 0.9957,
+      "step": 4508
+    },
+    {
+      "epoch": 0.33177282133823865,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.771172543395344e-05,
+      "loss": 0.7541,
+      "step": 4509
+    },
+    {
+      "epoch": 0.33184640147160266,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.7706738064833155e-05,
+      "loss": 1.4388,
+      "step": 4510
+    },
+    {
+      "epoch": 0.33191998160496666,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.770175001376623e-05,
+      "loss": 0.6819,
+      "step": 4511
+    },
+    {
+      "epoch": 0.33199356173833067,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.7696761281020387e-05,
+      "loss": 1.098,
+      "step": 4512
+    },
+    {
+      "epoch": 0.3320671418716946,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.769177186686334e-05,
+      "loss": 0.9928,
+      "step": 4513
+    },
+    {
+      "epoch": 0.3321407220050586,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.768678177156286e-05,
+      "loss": 0.759,
+      "step": 4514
+    },
+    {
+      "epoch": 0.3322143021384226,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.768179099538678e-05,
+      "loss": 0.8186,
+      "step": 4515
+    },
+    {
+      "epoch": 0.3322878822717866,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.767679953860292e-05,
+      "loss": 0.7773,
+      "step": 4516
+    },
+    {
+      "epoch": 0.33236146240515063,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.767180740147918e-05,
+      "loss": 0.9745,
+      "step": 4517
+    },
+    {
+      "epoch": 0.3324350425385146,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.766681458428346e-05,
+      "loss": 0.7881,
+      "step": 4518
+    },
+    {
+      "epoch": 0.3325086226718786,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.7661821087283726e-05,
+      "loss": 0.8318,
+      "step": 4519
+    },
+    {
+      "epoch": 0.3325822028052426,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.765682691074797e-05,
+      "loss": 0.9919,
+      "step": 4520
+    },
+    {
+      "epoch": 0.3326557829386066,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.765183205494422e-05,
+      "loss": 0.9731,
+      "step": 4521
+    },
+    {
+      "epoch": 0.33272936307197054,
+      "grad_norm": 1.125,
+      "learning_rate": 3.7646836520140536e-05,
+      "loss": 1.3843,
+      "step": 4522
+    },
+    {
+      "epoch": 0.33280294320533454,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.7641840306605034e-05,
+      "loss": 0.79,
+      "step": 4523
+    },
+    {
+      "epoch": 0.33287652333869855,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.763684341460583e-05,
+      "loss": 0.6936,
+      "step": 4524
+    },
+    {
+      "epoch": 0.33295010347206255,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.763184584441111e-05,
+      "loss": 0.9758,
+      "step": 4525
+    },
+    {
+      "epoch": 0.33302368360542656,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.762684759628908e-05,
+      "loss": 0.7863,
+      "step": 4526
+    },
+    {
+      "epoch": 0.3330972637387905,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.7621848670508e-05,
+      "loss": 0.6345,
+      "step": 4527
+    },
+    {
+      "epoch": 0.3331708438721545,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.761684906733613e-05,
+      "loss": 1.0841,
+      "step": 4528
+    },
+    {
+      "epoch": 0.3332444240055185,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.7611848787041794e-05,
+      "loss": 1.062,
+      "step": 4529
+    },
+    {
+      "epoch": 0.3333180041388825,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.760684782989337e-05,
+      "loss": 1.0346,
+      "step": 4530
+    },
+    {
+      "epoch": 0.33339158427224647,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.760184619615922e-05,
+      "loss": 1.297,
+      "step": 4531
+    },
+    {
+      "epoch": 0.33346516440561047,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.759684388610779e-05,
+      "loss": 0.8569,
+      "step": 4532
+    },
+    {
+      "epoch": 0.3335387445389745,
+      "grad_norm": 1.1328125,
+      "learning_rate": 3.7591840900007544e-05,
+      "loss": 1.1531,
+      "step": 4533
+    },
+    {
+      "epoch": 0.3336123246723385,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.758683723812698e-05,
+      "loss": 0.9587,
+      "step": 4534
+    },
+    {
+      "epoch": 0.3336859048057025,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.758183290073463e-05,
+      "loss": 1.0634,
+      "step": 4535
+    },
+    {
+      "epoch": 0.33375948493906643,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.757682788809907e-05,
+      "loss": 0.9081,
+      "step": 4536
+    },
+    {
+      "epoch": 0.33383306507243043,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.757182220048893e-05,
+      "loss": 0.7661,
+      "step": 4537
+    },
+    {
+      "epoch": 0.33390664520579444,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.7566815838172815e-05,
+      "loss": 0.9866,
+      "step": 4538
+    },
+    {
+      "epoch": 0.33398022533915844,
+      "grad_norm": 0.6015625,
+      "learning_rate": 3.756180880141944e-05,
+      "loss": 0.526,
+      "step": 4539
+    },
+    {
+      "epoch": 0.3340538054725224,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.7556801090497504e-05,
+      "loss": 1.2827,
+      "step": 4540
+    },
+    {
+      "epoch": 0.3341273856058864,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.755179270567578e-05,
+      "loss": 0.9763,
+      "step": 4541
+    },
+    {
+      "epoch": 0.3342009657392504,
+      "grad_norm": 1.15625,
+      "learning_rate": 3.754678364722304e-05,
+      "loss": 1.3209,
+      "step": 4542
+    },
+    {
+      "epoch": 0.3342745458726144,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.754177391540812e-05,
+      "loss": 1.079,
+      "step": 4543
+    },
+    {
+      "epoch": 0.3343481260059784,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.7536763510499895e-05,
+      "loss": 1.1404,
+      "step": 4544
+    },
+    {
+      "epoch": 0.33442170613934236,
+      "grad_norm": 1.0,
+      "learning_rate": 3.753175243276724e-05,
+      "loss": 1.2987,
+      "step": 4545
+    },
+    {
+      "epoch": 0.33449528627270636,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.752674068247911e-05,
+      "loss": 1.0372,
+      "step": 4546
+    },
+    {
+      "epoch": 0.33456886640607036,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.752172825990446e-05,
+      "loss": 0.9713,
+      "step": 4547
+    },
+    {
+      "epoch": 0.33464244653943437,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.75167151653123e-05,
+      "loss": 0.6626,
+      "step": 4548
+    },
+    {
+      "epoch": 0.3347160266727983,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.751170139897168e-05,
+      "loss": 0.8907,
+      "step": 4549
+    },
+    {
+      "epoch": 0.3347896068061623,
+      "grad_norm": 1.15625,
+      "learning_rate": 3.750668696115167e-05,
+      "loss": 0.7901,
+      "step": 4550
+    },
+    {
+      "epoch": 0.3348631869395263,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.7501671852121414e-05,
+      "loss": 1.1131,
+      "step": 4551
+    },
+    {
+      "epoch": 0.33493676707289033,
+      "grad_norm": 1.3828125,
+      "learning_rate": 3.7496656072150025e-05,
+      "loss": 0.8705,
+      "step": 4552
+    },
+    {
+      "epoch": 0.33501034720625433,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.749163962150671e-05,
+      "loss": 1.1594,
+      "step": 4553
+    },
+    {
+      "epoch": 0.3350839273396183,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.74866225004607e-05,
+      "loss": 0.7878,
+      "step": 4554
+    },
+    {
+      "epoch": 0.3351575074729823,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.7481604709281234e-05,
+      "loss": 0.806,
+      "step": 4555
+    },
+    {
+      "epoch": 0.3352310876063463,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.7476586248237616e-05,
+      "loss": 1.0637,
+      "step": 4556
+    },
+    {
+      "epoch": 0.3353046677397103,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.747156711759918e-05,
+      "loss": 1.2394,
+      "step": 4557
+    },
+    {
+      "epoch": 0.33537824787307424,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.74665473176353e-05,
+      "loss": 0.7652,
+      "step": 4558
+    },
+    {
+      "epoch": 0.33545182800643825,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.746152684861537e-05,
+      "loss": 1.0927,
+      "step": 4559
+    },
+    {
+      "epoch": 0.33552540813980225,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.745650571080882e-05,
+      "loss": 1.2981,
+      "step": 4560
+    },
+    {
+      "epoch": 0.33559898827316625,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.745148390448515e-05,
+      "loss": 0.8137,
+      "step": 4561
+    },
+    {
+      "epoch": 0.33567256840653026,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.744646142991385e-05,
+      "loss": 1.33,
+      "step": 4562
+    },
+    {
+      "epoch": 0.3357461485398942,
+      "grad_norm": 1.0,
+      "learning_rate": 3.7441438287364466e-05,
+      "loss": 1.2443,
+      "step": 4563
+    },
+    {
+      "epoch": 0.3358197286732582,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.74364144771066e-05,
+      "loss": 0.9518,
+      "step": 4564
+    },
+    {
+      "epoch": 0.3358933088066222,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.743138999940985e-05,
+      "loss": 0.7083,
+      "step": 4565
+    },
+    {
+      "epoch": 0.3359668889399862,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.742636485454388e-05,
+      "loss": 0.8364,
+      "step": 4566
+    },
+    {
+      "epoch": 0.3360404690733502,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.742133904277838e-05,
+      "loss": 1.3439,
+      "step": 4567
+    },
+    {
+      "epoch": 0.33611404920671417,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.741631256438307e-05,
+      "loss": 0.9982,
+      "step": 4568
+    },
+    {
+      "epoch": 0.3361876293400782,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.7411285419627716e-05,
+      "loss": 0.6144,
+      "step": 4569
+    },
+    {
+      "epoch": 0.3362612094734422,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.740625760878212e-05,
+      "loss": 0.8794,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3363347896068062,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.740122913211611e-05,
+      "loss": 1.2987,
+      "step": 4571
+    },
+    {
+      "epoch": 0.33640836974017013,
+      "grad_norm": 1.1171875,
+      "learning_rate": 3.739619998989955e-05,
+      "loss": 1.0957,
+      "step": 4572
+    },
+    {
+      "epoch": 0.33648194987353414,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.739117018240235e-05,
+      "loss": 0.7344,
+      "step": 4573
+    },
+    {
+      "epoch": 0.33655553000689814,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.7386139709894455e-05,
+      "loss": 1.4638,
+      "step": 4574
+    },
+    {
+      "epoch": 0.33662911014026214,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.7381108572645835e-05,
+      "loss": 0.7897,
+      "step": 4575
+    },
+    {
+      "epoch": 0.33670269027362615,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.73760767709265e-05,
+      "loss": 1.1372,
+      "step": 4576
+    },
+    {
+      "epoch": 0.3367762704069901,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.73710443050065e-05,
+      "loss": 0.927,
+      "step": 4577
+    },
+    {
+      "epoch": 0.3368498505403541,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.736601117515592e-05,
+      "loss": 1.03,
+      "step": 4578
+    },
+    {
+      "epoch": 0.3369234306737181,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.736097738164487e-05,
+      "loss": 0.7169,
+      "step": 4579
+    },
+    {
+      "epoch": 0.3369970108070821,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.7355942924743525e-05,
+      "loss": 0.9262,
+      "step": 4580
+    },
+    {
+      "epoch": 0.33707059094044606,
+      "grad_norm": 1.1328125,
+      "learning_rate": 3.7350907804722047e-05,
+      "loss": 0.9447,
+      "step": 4581
+    },
+    {
+      "epoch": 0.33714417107381006,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.734587202185067e-05,
+      "loss": 1.1006,
+      "step": 4582
+    },
+    {
+      "epoch": 0.33721775120717407,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.7340835576399675e-05,
+      "loss": 1.0276,
+      "step": 4583
+    },
+    {
+      "epoch": 0.33729133134053807,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.733579846863933e-05,
+      "loss": 0.8182,
+      "step": 4584
+    },
+    {
+      "epoch": 0.3373649114739021,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.7330760698839995e-05,
+      "loss": 0.806,
+      "step": 4585
+    },
+    {
+      "epoch": 0.337438491607266,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.732572226727201e-05,
+      "loss": 0.7503,
+      "step": 4586
+    },
+    {
+      "epoch": 0.33751207174063,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.7320683174205794e-05,
+      "loss": 1.0204,
+      "step": 4587
+    },
+    {
+      "epoch": 0.33758565187399403,
+      "grad_norm": 1.0,
+      "learning_rate": 3.731564341991178e-05,
+      "loss": 0.897,
+      "step": 4588
+    },
+    {
+      "epoch": 0.33765923200735803,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.7310603004660446e-05,
+      "loss": 0.8043,
+      "step": 4589
+    },
+    {
+      "epoch": 0.337732812140722,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.7305561928722306e-05,
+      "loss": 0.848,
+      "step": 4590
+    },
+    {
+      "epoch": 0.337806392274086,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.730052019236789e-05,
+      "loss": 0.8005,
+      "step": 4591
+    },
+    {
+      "epoch": 0.33787997240745,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.72954777958678e-05,
+      "loss": 0.8389,
+      "step": 4592
+    },
+    {
+      "epoch": 0.337953552540814,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.729043473949263e-05,
+      "loss": 1.0695,
+      "step": 4593
+    },
+    {
+      "epoch": 0.338027132674178,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.728539102351305e-05,
+      "loss": 1.1062,
+      "step": 4594
+    },
+    {
+      "epoch": 0.33810071280754195,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.7280346648199724e-05,
+      "loss": 0.7073,
+      "step": 4595
+    },
+    {
+      "epoch": 0.33817429294090595,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.72753016138234e-05,
+      "loss": 1.2053,
+      "step": 4596
+    },
+    {
+      "epoch": 0.33824787307426996,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.727025592065481e-05,
+      "loss": 0.8578,
+      "step": 4597
+    },
+    {
+      "epoch": 0.33832145320763396,
+      "grad_norm": 0.75,
+      "learning_rate": 3.726520956896477e-05,
+      "loss": 0.8075,
+      "step": 4598
+    },
+    {
+      "epoch": 0.3383950333409979,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.7260162559024095e-05,
+      "loss": 0.9552,
+      "step": 4599
+    },
+    {
+      "epoch": 0.3384686134743619,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.725511489110365e-05,
+      "loss": 0.6871,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3385421936077259,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.725006656547435e-05,
+      "loss": 1.2754,
+      "step": 4601
+    },
+    {
+      "epoch": 0.3386157737410899,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.7245017582407095e-05,
+      "loss": 0.7249,
+      "step": 4602
+    },
+    {
+      "epoch": 0.3386893538744539,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.723996794217288e-05,
+      "loss": 0.9319,
+      "step": 4603
+    },
+    {
+      "epoch": 0.3387629340078179,
+      "grad_norm": 1.0,
+      "learning_rate": 3.7234917645042706e-05,
+      "loss": 0.8907,
+      "step": 4604
+    },
+    {
+      "epoch": 0.3388365141411819,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.722986669128761e-05,
+      "loss": 0.9709,
+      "step": 4605
+    },
+    {
+      "epoch": 0.3389100942745459,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.7224815081178666e-05,
+      "loss": 1.0093,
+      "step": 4606
+    },
+    {
+      "epoch": 0.3389836744079099,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.721976281498699e-05,
+      "loss": 1.0445,
+      "step": 4607
+    },
+    {
+      "epoch": 0.33905725454127383,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.7214709892983725e-05,
+      "loss": 0.9872,
+      "step": 4608
+    },
+    {
+      "epoch": 0.33913083467463784,
+      "grad_norm": 1.1171875,
+      "learning_rate": 3.720965631544004e-05,
+      "loss": 1.0319,
+      "step": 4609
+    },
+    {
+      "epoch": 0.33920441480800184,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.720460208262717e-05,
+      "loss": 0.9197,
+      "step": 4610
+    },
+    {
+      "epoch": 0.33927799494136585,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.719954719481635e-05,
+      "loss": 0.7862,
+      "step": 4611
+    },
+    {
+      "epoch": 0.33935157507472985,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.7194491652278876e-05,
+      "loss": 0.9574,
+      "step": 4612
+    },
+    {
+      "epoch": 0.3394251552080938,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.718943545528607e-05,
+      "loss": 0.8347,
+      "step": 4613
+    },
+    {
+      "epoch": 0.3394987353414578,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.718437860410929e-05,
+      "loss": 0.9657,
+      "step": 4614
+    },
+    {
+      "epoch": 0.3395723154748218,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.7179321099019916e-05,
+      "loss": 0.8613,
+      "step": 4615
+    },
+    {
+      "epoch": 0.3396458956081858,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.717426294028938e-05,
+      "loss": 0.8471,
+      "step": 4616
+    },
+    {
+      "epoch": 0.33971947574154976,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.7169204128189154e-05,
+      "loss": 0.7595,
+      "step": 4617
+    },
+    {
+      "epoch": 0.33979305587491376,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.716414466299072e-05,
+      "loss": 0.6496,
+      "step": 4618
+    },
+    {
+      "epoch": 0.33986663600827777,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.715908454496563e-05,
+      "loss": 1.1779,
+      "step": 4619
+    },
+    {
+      "epoch": 0.33994021614164177,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.715402377438542e-05,
+      "loss": 1.117,
+      "step": 4620
+    },
+    {
+      "epoch": 0.3400137962750058,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.714896235152172e-05,
+      "loss": 1.1262,
+      "step": 4621
+    },
+    {
+      "epoch": 0.3400873764083697,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.714390027664615e-05,
+      "loss": 0.8273,
+      "step": 4622
+    },
+    {
+      "epoch": 0.3401609565417337,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.713883755003039e-05,
+      "loss": 0.925,
+      "step": 4623
+    },
+    {
+      "epoch": 0.34023453667509773,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.713377417194616e-05,
+      "loss": 0.9008,
+      "step": 4624
+    },
+    {
+      "epoch": 0.34030811680846174,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.7128710142665166e-05,
+      "loss": 0.7594,
+      "step": 4625
+    },
+    {
+      "epoch": 0.3403816969418257,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.712364546245922e-05,
+      "loss": 0.8971,
+      "step": 4626
+    },
+    {
+      "epoch": 0.3404552770751897,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.711858013160012e-05,
+      "loss": 1.0136,
+      "step": 4627
+    },
+    {
+      "epoch": 0.3405288572085537,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.711351415035971e-05,
+      "loss": 1.2684,
+      "step": 4628
+    },
+    {
+      "epoch": 0.3406024373419177,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.710844751900988e-05,
+      "loss": 0.8905,
+      "step": 4629
+    },
+    {
+      "epoch": 0.3406760174752817,
+      "grad_norm": 1.28125,
+      "learning_rate": 3.7103380237822525e-05,
+      "loss": 0.9414,
+      "step": 4630
+    },
+    {
+      "epoch": 0.34074959760864565,
+      "grad_norm": 1.140625,
+      "learning_rate": 3.7098312307069626e-05,
+      "loss": 1.092,
+      "step": 4631
+    },
+    {
+      "epoch": 0.34082317774200965,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.7093243727023154e-05,
+      "loss": 0.9318,
+      "step": 4632
+    },
+    {
+      "epoch": 0.34089675787537366,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.7088174497955136e-05,
+      "loss": 0.6058,
+      "step": 4633
+    },
+    {
+      "epoch": 0.34097033800873766,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.7083104620137624e-05,
+      "loss": 0.9116,
+      "step": 4634
+    },
+    {
+      "epoch": 0.3410439181421016,
+      "grad_norm": 1.0,
+      "learning_rate": 3.70780340938427e-05,
+      "loss": 1.1056,
+      "step": 4635
+    },
+    {
+      "epoch": 0.3411174982754656,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.70729629193425e-05,
+      "loss": 1.045,
+      "step": 4636
+    },
+    {
+      "epoch": 0.3411910784088296,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.706789109690919e-05,
+      "loss": 1.3281,
+      "step": 4637
+    },
+    {
+      "epoch": 0.3412646585421936,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.706281862681495e-05,
+      "loss": 0.6027,
+      "step": 4638
+    },
+    {
+      "epoch": 0.3413382386755576,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.705774550933202e-05,
+      "loss": 1.3672,
+      "step": 4639
+    },
+    {
+      "epoch": 0.3414118188089216,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.705267174473267e-05,
+      "loss": 1.2316,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3414853989422856,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.704759733328918e-05,
+      "loss": 1.1305,
+      "step": 4641
+    },
+    {
+      "epoch": 0.3415589790756496,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.704252227527391e-05,
+      "loss": 1.2627,
+      "step": 4642
+    },
+    {
+      "epoch": 0.3416325592090136,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.703744657095919e-05,
+      "loss": 0.7576,
+      "step": 4643
+    },
+    {
+      "epoch": 0.34170613934237754,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.7032370220617476e-05,
+      "loss": 1.2629,
+      "step": 4644
+    },
+    {
+      "epoch": 0.34177971947574154,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.702729322452116e-05,
+      "loss": 0.898,
+      "step": 4645
+    },
+    {
+      "epoch": 0.34185329960910554,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.702221558294274e-05,
+      "loss": 1.3024,
+      "step": 4646
+    },
+    {
+      "epoch": 0.34192687974246955,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.701713729615471e-05,
+      "loss": 1.4048,
+      "step": 4647
+    },
+    {
+      "epoch": 0.34200045987583355,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.701205836442963e-05,
+      "loss": 0.8842,
+      "step": 4648
+    },
+    {
+      "epoch": 0.3420740400091975,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.700697878804006e-05,
+      "loss": 0.9754,
+      "step": 4649
+    },
+    {
+      "epoch": 0.3421476201425615,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.7001898567258605e-05,
+      "loss": 1.0291,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3422212002759255,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.699681770235794e-05,
+      "loss": 0.9822,
+      "step": 4651
+    },
+    {
+      "epoch": 0.3422947804092895,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.6991736193610724e-05,
+      "loss": 1.2103,
+      "step": 4652
+    },
+    {
+      "epoch": 0.34236836054265346,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.698665404128967e-05,
+      "loss": 1.2218,
+      "step": 4653
+    },
+    {
+      "epoch": 0.34244194067601746,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.698157124566753e-05,
+      "loss": 1.1661,
+      "step": 4654
+    },
+    {
+      "epoch": 0.34251552080938147,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.6976487807017104e-05,
+      "loss": 0.911,
+      "step": 4655
+    },
+    {
+      "epoch": 0.3425891009427455,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.6971403725611186e-05,
+      "loss": 0.7984,
+      "step": 4656
+    },
+    {
+      "epoch": 0.3426626810761095,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.6966319001722646e-05,
+      "loss": 0.6459,
+      "step": 4657
+    },
+    {
+      "epoch": 0.3427362612094734,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.6961233635624364e-05,
+      "loss": 1.0434,
+      "step": 4658
+    },
+    {
+      "epoch": 0.34280984134283743,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.695614762758927e-05,
+      "loss": 1.3347,
+      "step": 4659
+    },
+    {
+      "epoch": 0.34288342147620143,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.6951060977890305e-05,
+      "loss": 0.828,
+      "step": 4660
+    },
+    {
+      "epoch": 0.34295700160956544,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.694597368680048e-05,
+      "loss": 0.766,
+      "step": 4661
+    },
+    {
+      "epoch": 0.3430305817429294,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.69408857545928e-05,
+      "loss": 1.3484,
+      "step": 4662
+    },
+    {
+      "epoch": 0.3431041618762934,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.693579718154034e-05,
+      "loss": 0.6742,
+      "step": 4663
+    },
+    {
+      "epoch": 0.3431777420096574,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.693070796791619e-05,
+      "loss": 0.9636,
+      "step": 4664
+    },
+    {
+      "epoch": 0.3432513221430214,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.692561811399348e-05,
+      "loss": 0.7635,
+      "step": 4665
+    },
+    {
+      "epoch": 0.3433249022763854,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.692052762004536e-05,
+      "loss": 1.2835,
+      "step": 4666
+    },
+    {
+      "epoch": 0.34339848240974935,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.691543648634505e-05,
+      "loss": 0.7873,
+      "step": 4667
+    },
+    {
+      "epoch": 0.34347206254311335,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.691034471316576e-05,
+      "loss": 1.097,
+      "step": 4668
+    },
+    {
+      "epoch": 0.34354564267647736,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.6905252300780765e-05,
+      "loss": 0.7454,
+      "step": 4669
+    },
+    {
+      "epoch": 0.34361922280984136,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.6900159249463364e-05,
+      "loss": 0.9617,
+      "step": 4670
+    },
+    {
+      "epoch": 0.3436928029432053,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.6895065559486894e-05,
+      "loss": 0.8249,
+      "step": 4671
+    },
+    {
+      "epoch": 0.3437663830765693,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.6889971231124725e-05,
+      "loss": 1.4166,
+      "step": 4672
+    },
+    {
+      "epoch": 0.3438399632099333,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.688487626465025e-05,
+      "loss": 0.6122,
+      "step": 4673
+    },
+    {
+      "epoch": 0.3439135433432973,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.687978066033693e-05,
+      "loss": 0.8879,
+      "step": 4674
+    },
+    {
+      "epoch": 0.3439871234766613,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.68746844184582e-05,
+      "loss": 0.8693,
+      "step": 4675
+    },
+    {
+      "epoch": 0.3440607036100253,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.686958753928759e-05,
+      "loss": 1.2325,
+      "step": 4676
+    },
+    {
+      "epoch": 0.3441342837433893,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.686449002309864e-05,
+      "loss": 0.9057,
+      "step": 4677
+    },
+    {
+      "epoch": 0.3442078638767533,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.685939187016492e-05,
+      "loss": 1.0158,
+      "step": 4678
+    },
+    {
+      "epoch": 0.3442814440101173,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.6854293080760036e-05,
+      "loss": 0.5932,
+      "step": 4679
+    },
+    {
+      "epoch": 0.34435502414348124,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.684919365515762e-05,
+      "loss": 0.7456,
+      "step": 4680
+    },
+    {
+      "epoch": 0.34442860427684524,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.684409359363138e-05,
+      "loss": 0.5749,
+      "step": 4681
+    },
+    {
+      "epoch": 0.34450218441020924,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.6838992896455e-05,
+      "loss": 0.7502,
+      "step": 4682
+    },
+    {
+      "epoch": 0.34457576454357325,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.6833891563902225e-05,
+      "loss": 0.8746,
+      "step": 4683
+    },
+    {
+      "epoch": 0.34464934467693725,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.6828789596246846e-05,
+      "loss": 0.9033,
+      "step": 4684
+    },
+    {
+      "epoch": 0.3447229248103012,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.682368699376268e-05,
+      "loss": 1.0363,
+      "step": 4685
+    },
+    {
+      "epoch": 0.3447965049436652,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.681858375672355e-05,
+      "loss": 1.1239,
+      "step": 4686
+    },
+    {
+      "epoch": 0.3448700850770292,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.6813479885403355e-05,
+      "loss": 0.8014,
+      "step": 4687
+    },
+    {
+      "epoch": 0.3449436652103932,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.680837538007601e-05,
+      "loss": 1.2095,
+      "step": 4688
+    },
+    {
+      "epoch": 0.34501724534375716,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.6803270241015465e-05,
+      "loss": 0.6501,
+      "step": 4689
+    },
+    {
+      "epoch": 0.34509082547712117,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.6798164468495696e-05,
+      "loss": 1.0622,
+      "step": 4690
+    },
+    {
+      "epoch": 0.34516440561048517,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.679305806279072e-05,
+      "loss": 0.8261,
+      "step": 4691
+    },
+    {
+      "epoch": 0.3452379857438492,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.67879510241746e-05,
+      "loss": 0.7577,
+      "step": 4692
+    },
+    {
+      "epoch": 0.3453115658772132,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.67828433529214e-05,
+      "loss": 1.22,
+      "step": 4693
+    },
+    {
+      "epoch": 0.3453851460105771,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.677773504930526e-05,
+      "loss": 0.9988,
+      "step": 4694
+    },
+    {
+      "epoch": 0.34545872614394113,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.677262611360033e-05,
+      "loss": 1.5305,
+      "step": 4695
+    },
+    {
+      "epoch": 0.34553230627730513,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.6767516546080786e-05,
+      "loss": 0.6164,
+      "step": 4696
+    },
+    {
+      "epoch": 0.34560588641066914,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.6762406347020856e-05,
+      "loss": 1.3206,
+      "step": 4697
+    },
+    {
+      "epoch": 0.3456794665440331,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.675729551669479e-05,
+      "loss": 0.7833,
+      "step": 4698
+    },
+    {
+      "epoch": 0.3457530466773971,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.6752184055376886e-05,
+      "loss": 0.6527,
+      "step": 4699
+    },
+    {
+      "epoch": 0.3458266268107611,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.6747071963341454e-05,
+      "loss": 1.1474,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3459002069441251,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.674195924086287e-05,
+      "loss": 0.599,
+      "step": 4701
+    },
+    {
+      "epoch": 0.3459737870774891,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.673684588821549e-05,
+      "loss": 0.8061,
+      "step": 4702
+    },
+    {
+      "epoch": 0.34604736721085305,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.6731731905673774e-05,
+      "loss": 1.018,
+      "step": 4703
+    },
+    {
+      "epoch": 0.34612094734421706,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.672661729351216e-05,
+      "loss": 0.7041,
+      "step": 4704
+    },
+    {
+      "epoch": 0.34619452747758106,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.672150205200514e-05,
+      "loss": 0.9247,
+      "step": 4705
+    },
+    {
+      "epoch": 0.34626810761094506,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.671638618142725e-05,
+      "loss": 0.936,
+      "step": 4706
+    },
+    {
+      "epoch": 0.346341687744309,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.671126968205304e-05,
+      "loss": 0.629,
+      "step": 4707
+    },
+    {
+      "epoch": 0.346415267877673,
+      "grad_norm": 1.1640625,
+      "learning_rate": 3.670615255415711e-05,
+      "loss": 1.2083,
+      "step": 4708
+    },
+    {
+      "epoch": 0.346488848011037,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.670103479801408e-05,
+      "loss": 0.7944,
+      "step": 4709
+    },
+    {
+      "epoch": 0.346562428144401,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.6695916413898603e-05,
+      "loss": 0.7856,
+      "step": 4710
+    },
+    {
+      "epoch": 0.34663600827776503,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.669079740208539e-05,
+      "loss": 0.9476,
+      "step": 4711
+    },
+    {
+      "epoch": 0.346709588411129,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.6685677762849165e-05,
+      "loss": 0.6409,
+      "step": 4712
+    },
+    {
+      "epoch": 0.346783168544493,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.6680557496464684e-05,
+      "loss": 1.1041,
+      "step": 4713
+    },
+    {
+      "epoch": 0.346856748677857,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.667543660320674e-05,
+      "loss": 0.7868,
+      "step": 4714
+    },
+    {
+      "epoch": 0.346930328811221,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.667031508335017e-05,
+      "loss": 1.2149,
+      "step": 4715
+    },
+    {
+      "epoch": 0.34700390894458494,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.666519293716983e-05,
+      "loss": 0.9748,
+      "step": 4716
+    },
+    {
+      "epoch": 0.34707748907794894,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.6660070164940615e-05,
+      "loss": 0.925,
+      "step": 4717
+    },
+    {
+      "epoch": 0.34715106921131295,
+      "grad_norm": 0.75,
+      "learning_rate": 3.665494676693745e-05,
+      "loss": 0.7965,
+      "step": 4718
+    },
+    {
+      "epoch": 0.34722464934467695,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.664982274343531e-05,
+      "loss": 0.7948,
+      "step": 4719
+    },
+    {
+      "epoch": 0.34729822947804095,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.66446980947092e-05,
+      "loss": 0.8549,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3473718096114049,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.663957282103412e-05,
+      "loss": 0.9215,
+      "step": 4721
+    },
+    {
+      "epoch": 0.3474453897447689,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.663444692268517e-05,
+      "loss": 0.7448,
+      "step": 4722
+    },
+    {
+      "epoch": 0.3475189698781329,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.6629320399937414e-05,
+      "loss": 0.9834,
+      "step": 4723
+    },
+    {
+      "epoch": 0.3475925500114969,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.6624193253065996e-05,
+      "loss": 1.0663,
+      "step": 4724
+    },
+    {
+      "epoch": 0.34766613014486086,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.6619065482346084e-05,
+      "loss": 0.5999,
+      "step": 4725
+    },
+    {
+      "epoch": 0.34773971027822487,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.6613937088052876e-05,
+      "loss": 0.9717,
+      "step": 4726
+    },
+    {
+      "epoch": 0.34781329041158887,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.660880807046159e-05,
+      "loss": 0.9176,
+      "step": 4727
+    },
+    {
+      "epoch": 0.3478868705449529,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.660367842984751e-05,
+      "loss": 1.0512,
+      "step": 4728
+    },
+    {
+      "epoch": 0.3479604506783169,
+      "grad_norm": 3.578125,
+      "learning_rate": 3.659854816648592e-05,
+      "loss": 1.1632,
+      "step": 4729
+    },
+    {
+      "epoch": 0.34803403081168083,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.6593417280652164e-05,
+      "loss": 0.7006,
+      "step": 4730
+    },
+    {
+      "epoch": 0.34810761094504483,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.6588285772621586e-05,
+      "loss": 0.9917,
+      "step": 4731
+    },
+    {
+      "epoch": 0.34818119107840884,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.658315364266961e-05,
+      "loss": 1.096,
+      "step": 4732
+    },
+    {
+      "epoch": 0.34825477121177284,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.657802089107165e-05,
+      "loss": 0.9812,
+      "step": 4733
+    },
+    {
+      "epoch": 0.3483283513451368,
+      "grad_norm": 0.62109375,
+      "learning_rate": 3.657288751810318e-05,
+      "loss": 0.6361,
+      "step": 4734
+    },
+    {
+      "epoch": 0.3484019314785008,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.656775352403969e-05,
+      "loss": 0.6014,
+      "step": 4735
+    },
+    {
+      "epoch": 0.3484755116118648,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.6562618909156726e-05,
+      "loss": 0.8307,
+      "step": 4736
+    },
+    {
+      "epoch": 0.3485490917452288,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.6557483673729834e-05,
+      "loss": 1.0536,
+      "step": 4737
+    },
+    {
+      "epoch": 0.3486226718785928,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.6552347818034626e-05,
+      "loss": 1.0115,
+      "step": 4738
+    },
+    {
+      "epoch": 0.34869625201195675,
+      "grad_norm": 1.0,
+      "learning_rate": 3.654721134234673e-05,
+      "loss": 1.311,
+      "step": 4739
+    },
+    {
+      "epoch": 0.34876983214532076,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.6542074246941816e-05,
+      "loss": 0.9436,
+      "step": 4740
+    },
+    {
+      "epoch": 0.34884341227868476,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.6536936532095565e-05,
+      "loss": 0.9542,
+      "step": 4741
+    },
+    {
+      "epoch": 0.34891699241204877,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.653179819808373e-05,
+      "loss": 1.1838,
+      "step": 4742
+    },
+    {
+      "epoch": 0.3489905725454127,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.652665924518206e-05,
+      "loss": 0.938,
+      "step": 4743
+    },
+    {
+      "epoch": 0.3490641526787767,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.652151967366637e-05,
+      "loss": 0.9013,
+      "step": 4744
+    },
+    {
+      "epoch": 0.3491377328121407,
+      "grad_norm": 5.28125,
+      "learning_rate": 3.651637948381247e-05,
+      "loss": 1.3578,
+      "step": 4745
+    },
+    {
+      "epoch": 0.3492113129455047,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.651123867589623e-05,
+      "loss": 0.8332,
+      "step": 4746
+    },
+    {
+      "epoch": 0.34928489307886873,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.650609725019356e-05,
+      "loss": 0.8459,
+      "step": 4747
+    },
+    {
+      "epoch": 0.3493584732122327,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.650095520698038e-05,
+      "loss": 0.8967,
+      "step": 4748
+    },
+    {
+      "epoch": 0.3494320533455967,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.6495812546532664e-05,
+      "loss": 0.8182,
+      "step": 4749
+    },
+    {
+      "epoch": 0.3495056334789607,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.649066926912639e-05,
+      "loss": 0.9408,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3495792136123247,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.64855253750376e-05,
+      "loss": 0.6891,
+      "step": 4751
+    },
+    {
+      "epoch": 0.34965279374568864,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.648038086454236e-05,
+      "loss": 0.6999,
+      "step": 4752
+    },
+    {
+      "epoch": 0.34972637387905264,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.647523573791677e-05,
+      "loss": 0.8032,
+      "step": 4753
+    },
+    {
+      "epoch": 0.34979995401241665,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.647008999543694e-05,
+      "loss": 1.1803,
+      "step": 4754
+    },
+    {
+      "epoch": 0.34987353414578065,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.6464943637379053e-05,
+      "loss": 1.5012,
+      "step": 4755
+    },
+    {
+      "epoch": 0.34994711427914466,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.645979666401929e-05,
+      "loss": 1.2367,
+      "step": 4756
+    },
+    {
+      "epoch": 0.3500206944125086,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.6454649075633885e-05,
+      "loss": 1.372,
+      "step": 4757
+    },
+    {
+      "epoch": 0.3500942745458726,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.64495008724991e-05,
+      "loss": 0.8756,
+      "step": 4758
+    },
+    {
+      "epoch": 0.3501678546792366,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.644435205489122e-05,
+      "loss": 1.0225,
+      "step": 4759
+    },
+    {
+      "epoch": 0.3502414348126006,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.643920262308659e-05,
+      "loss": 0.677,
+      "step": 4760
+    },
+    {
+      "epoch": 0.35031501494596456,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.6434052577361556e-05,
+      "loss": 1.448,
+      "step": 4761
+    },
+    {
+      "epoch": 0.35038859507932857,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.642890191799252e-05,
+      "loss": 0.6937,
+      "step": 4762
+    },
+    {
+      "epoch": 0.3504621752126926,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.642375064525591e-05,
+      "loss": 0.6267,
+      "step": 4763
+    },
+    {
+      "epoch": 0.3505357553460566,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.641859875942816e-05,
+      "loss": 1.0058,
+      "step": 4764
+    },
+    {
+      "epoch": 0.3506093354794206,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.6413446260785797e-05,
+      "loss": 1.1573,
+      "step": 4765
+    },
+    {
+      "epoch": 0.35068291561278453,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.640829314960532e-05,
+      "loss": 0.7338,
+      "step": 4766
+    },
+    {
+      "epoch": 0.35075649574614853,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.64031394261633e-05,
+      "loss": 1.1617,
+      "step": 4767
+    },
+    {
+      "epoch": 0.35083007587951254,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.6397985090736336e-05,
+      "loss": 0.8588,
+      "step": 4768
+    },
+    {
+      "epoch": 0.35090365601287654,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.639283014360103e-05,
+      "loss": 0.822,
+      "step": 4769
+    },
+    {
+      "epoch": 0.3509772361462405,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.638767458503405e-05,
+      "loss": 0.8289,
+      "step": 4770
+    },
+    {
+      "epoch": 0.3510508162796045,
+      "grad_norm": 4.53125,
+      "learning_rate": 3.638251841531208e-05,
+      "loss": 0.7602,
+      "step": 4771
+    },
+    {
+      "epoch": 0.3511243964129685,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.637736163471185e-05,
+      "loss": 1.2078,
+      "step": 4772
+    },
+    {
+      "epoch": 0.3511979765463325,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.63722042435101e-05,
+      "loss": 0.9631,
+      "step": 4773
+    },
+    {
+      "epoch": 0.3512715566796965,
+      "grad_norm": 0.6328125,
+      "learning_rate": 3.6367046241983644e-05,
+      "loss": 0.6661,
+      "step": 4774
+    },
+    {
+      "epoch": 0.35134513681306045,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.636188763040928e-05,
+      "loss": 0.9251,
+      "step": 4775
+    },
+    {
+      "epoch": 0.35141871694642446,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.6356728409063856e-05,
+      "loss": 1.3103,
+      "step": 4776
+    },
+    {
+      "epoch": 0.35149229707978846,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.6351568578224275e-05,
+      "loss": 1.4042,
+      "step": 4777
+    },
+    {
+      "epoch": 0.35156587721315247,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.6346408138167455e-05,
+      "loss": 0.6177,
+      "step": 4778
+    },
+    {
+      "epoch": 0.3516394573465164,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.634124708917033e-05,
+      "loss": 0.868,
+      "step": 4779
+    },
+    {
+      "epoch": 0.3517130374798804,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.6336085431509895e-05,
+      "loss": 0.7677,
+      "step": 4780
+    },
+    {
+      "epoch": 0.3517866176132444,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.6330923165463174e-05,
+      "loss": 0.8498,
+      "step": 4781
+    },
+    {
+      "epoch": 0.3518601977466084,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.63257602913072e-05,
+      "loss": 0.8257,
+      "step": 4782
+    },
+    {
+      "epoch": 0.35193377787997243,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.6320596809319064e-05,
+      "loss": 0.7374,
+      "step": 4783
+    },
+    {
+      "epoch": 0.3520073580133364,
+      "grad_norm": 1.1640625,
+      "learning_rate": 3.6315432719775885e-05,
+      "loss": 1.3143,
+      "step": 4784
+    },
+    {
+      "epoch": 0.3520809381467004,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.63102680229548e-05,
+      "loss": 1.0398,
+      "step": 4785
+    },
+    {
+      "epoch": 0.3521545182800644,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.630510271913298e-05,
+      "loss": 0.9623,
+      "step": 4786
+    },
+    {
+      "epoch": 0.3522280984134284,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.6299936808587666e-05,
+      "loss": 0.877,
+      "step": 4787
+    },
+    {
+      "epoch": 0.35230167854679234,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.629477029159608e-05,
+      "loss": 0.7758,
+      "step": 4788
+    },
+    {
+      "epoch": 0.35237525868015634,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.6289603168435504e-05,
+      "loss": 1.0819,
+      "step": 4789
+    },
+    {
+      "epoch": 0.35244883881352035,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.6284435439383254e-05,
+      "loss": 0.7601,
+      "step": 4790
+    },
+    {
+      "epoch": 0.35252241894688435,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.6279267104716664e-05,
+      "loss": 1.6886,
+      "step": 4791
+    },
+    {
+      "epoch": 0.35259599908024836,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.627409816471311e-05,
+      "loss": 0.8755,
+      "step": 4792
+    },
+    {
+      "epoch": 0.3526695792136123,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.626892861965e-05,
+      "loss": 1.0923,
+      "step": 4793
+    },
+    {
+      "epoch": 0.3527431593469763,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.626375846980477e-05,
+      "loss": 0.8383,
+      "step": 4794
+    },
+    {
+      "epoch": 0.3528167394803403,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.625858771545492e-05,
+      "loss": 0.8668,
+      "step": 4795
+    },
+    {
+      "epoch": 0.3528903196137043,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.625341635687791e-05,
+      "loss": 1.0451,
+      "step": 4796
+    },
+    {
+      "epoch": 0.35296389974706827,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.624824439435132e-05,
+      "loss": 0.8534,
+      "step": 4797
+    },
+    {
+      "epoch": 0.35303747988043227,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.624307182815268e-05,
+      "loss": 0.8749,
+      "step": 4798
+    },
+    {
+      "epoch": 0.3531110600137963,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.623789865855961e-05,
+      "loss": 0.9576,
+      "step": 4799
+    },
+    {
+      "epoch": 0.3531846401471603,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.6232724885849745e-05,
+      "loss": 0.9915,
+      "step": 4800
+    },
+    {
+      "epoch": 0.3532582202805243,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.6227550510300754e-05,
+      "loss": 1.1721,
+      "step": 4801
+    },
+    {
+      "epoch": 0.35333180041388823,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.622237553219034e-05,
+      "loss": 0.8471,
+      "step": 4802
+    },
+    {
+      "epoch": 0.35340538054725223,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.621719995179622e-05,
+      "loss": 1.0099,
+      "step": 4803
+    },
+    {
+      "epoch": 0.35347896068061624,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.6212023769396165e-05,
+      "loss": 0.7313,
+      "step": 4804
+    },
+    {
+      "epoch": 0.35355254081398024,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.620684698526797e-05,
+      "loss": 0.7327,
+      "step": 4805
+    },
+    {
+      "epoch": 0.3536261209473442,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.6201669599689465e-05,
+      "loss": 0.7027,
+      "step": 4806
+    },
+    {
+      "epoch": 0.3536997010807082,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.619649161293851e-05,
+      "loss": 1.1363,
+      "step": 4807
+    },
+    {
+      "epoch": 0.3537732812140722,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.6191313025292996e-05,
+      "loss": 0.7394,
+      "step": 4808
+    },
+    {
+      "epoch": 0.3538468613474362,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.6186133837030856e-05,
+      "loss": 0.8749,
+      "step": 4809
+    },
+    {
+      "epoch": 0.3539204414808002,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.618095404843003e-05,
+      "loss": 0.9417,
+      "step": 4810
+    },
+    {
+      "epoch": 0.35399402161416416,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.617577365976853e-05,
+      "loss": 0.8377,
+      "step": 4811
+    },
+    {
+      "epoch": 0.35406760174752816,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.617059267132435e-05,
+      "loss": 0.9462,
+      "step": 4812
+    },
+    {
+      "epoch": 0.35414118188089216,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.6165411083375575e-05,
+      "loss": 0.8838,
+      "step": 4813
+    },
+    {
+      "epoch": 0.35421476201425617,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.6160228896200265e-05,
+      "loss": 1.2216,
+      "step": 4814
+    },
+    {
+      "epoch": 0.3542883421476201,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.6155046110076554e-05,
+      "loss": 0.9618,
+      "step": 4815
+    },
+    {
+      "epoch": 0.3543619222809841,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.6149862725282587e-05,
+      "loss": 0.8811,
+      "step": 4816
+    },
+    {
+      "epoch": 0.3544355024143481,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.614467874209654e-05,
+      "loss": 1.0577,
+      "step": 4817
+    },
+    {
+      "epoch": 0.35450908254771213,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.613949416079665e-05,
+      "loss": 0.5258,
+      "step": 4818
+    },
+    {
+      "epoch": 0.35458266268107613,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.613430898166113e-05,
+      "loss": 1.047,
+      "step": 4819
+    },
+    {
+      "epoch": 0.3546562428144401,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.612912320496829e-05,
+      "loss": 0.9518,
+      "step": 4820
+    },
+    {
+      "epoch": 0.3547298229478041,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.612393683099642e-05,
+      "loss": 1.0906,
+      "step": 4821
+    },
+    {
+      "epoch": 0.3548034030811681,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.611874986002387e-05,
+      "loss": 0.935,
+      "step": 4822
+    },
+    {
+      "epoch": 0.3548769832145321,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.6113562292329016e-05,
+      "loss": 0.6102,
+      "step": 4823
+    },
+    {
+      "epoch": 0.35495056334789604,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.610837412819027e-05,
+      "loss": 1.1,
+      "step": 4824
+    },
+    {
+      "epoch": 0.35502414348126005,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.610318536788606e-05,
+      "loss": 1.2297,
+      "step": 4825
+    },
+    {
+      "epoch": 0.35509772361462405,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.609799601169486e-05,
+      "loss": 0.7288,
+      "step": 4826
+    },
+    {
+      "epoch": 0.35517130374798805,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.609280605989518e-05,
+      "loss": 0.8733,
+      "step": 4827
+    },
+    {
+      "epoch": 0.35524488388135206,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.608761551276555e-05,
+      "loss": 1.189,
+      "step": 4828
+    },
+    {
+      "epoch": 0.355318464014716,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.608242437058454e-05,
+      "loss": 0.6602,
+      "step": 4829
+    },
+    {
+      "epoch": 0.35539204414808,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.607723263363074e-05,
+      "loss": 0.7282,
+      "step": 4830
+    },
+    {
+      "epoch": 0.355465624281444,
+      "grad_norm": 0.75,
+      "learning_rate": 3.6072040302182784e-05,
+      "loss": 0.9348,
+      "step": 4831
+    },
+    {
+      "epoch": 0.355539204414808,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.6066847376519345e-05,
+      "loss": 1.2901,
+      "step": 4832
+    },
+    {
+      "epoch": 0.35561278454817197,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.60616538569191e-05,
+      "loss": 1.0461,
+      "step": 4833
+    },
+    {
+      "epoch": 0.35568636468153597,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.60564597436608e-05,
+      "loss": 0.6251,
+      "step": 4834
+    },
+    {
+      "epoch": 0.3557599448149,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.6051265037023176e-05,
+      "loss": 0.9186,
+      "step": 4835
+    },
+    {
+      "epoch": 0.355833524948264,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.604606973728504e-05,
+      "loss": 0.984,
+      "step": 4836
+    },
+    {
+      "epoch": 0.355907105081628,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.604087384472519e-05,
+      "loss": 1.1697,
+      "step": 4837
+    },
+    {
+      "epoch": 0.35598068521499193,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.603567735962251e-05,
+      "loss": 0.9941,
+      "step": 4838
+    },
+    {
+      "epoch": 0.35605426534835594,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.603048028225585e-05,
+      "loss": 1.1957,
+      "step": 4839
+    },
+    {
+      "epoch": 0.35612784548171994,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.6025282612904157e-05,
+      "loss": 1.4154,
+      "step": 4840
+    },
+    {
+      "epoch": 0.35620142561508394,
+      "grad_norm": 2.328125,
+      "learning_rate": 3.602008435184638e-05,
+      "loss": 1.1748,
+      "step": 4841
+    },
+    {
+      "epoch": 0.3562750057484479,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.6014885499361476e-05,
+      "loss": 1.0554,
+      "step": 4842
+    },
+    {
+      "epoch": 0.3563485858818119,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.600968605572848e-05,
+      "loss": 0.9385,
+      "step": 4843
+    },
+    {
+      "epoch": 0.3564221660151759,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.600448602122643e-05,
+      "loss": 0.729,
+      "step": 4844
+    },
+    {
+      "epoch": 0.3564957461485399,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.599928539613439e-05,
+      "loss": 0.7915,
+      "step": 4845
+    },
+    {
+      "epoch": 0.3565693262819039,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.599408418073147e-05,
+      "loss": 0.8664,
+      "step": 4846
+    },
+    {
+      "epoch": 0.35664290641526786,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.598888237529684e-05,
+      "loss": 0.7553,
+      "step": 4847
+    },
+    {
+      "epoch": 0.35671648654863186,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.598367998010963e-05,
+      "loss": 0.9498,
+      "step": 4848
+    },
+    {
+      "epoch": 0.35679006668199587,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.5978476995449066e-05,
+      "loss": 0.9862,
+      "step": 4849
+    },
+    {
+      "epoch": 0.35686364681535987,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.597327342159438e-05,
+      "loss": 0.981,
+      "step": 4850
+    },
+    {
+      "epoch": 0.3569372269487238,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.596806925882483e-05,
+      "loss": 0.9306,
+      "step": 4851
+    },
+    {
+      "epoch": 0.3570108070820878,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.596286450741973e-05,
+      "loss": 0.6954,
+      "step": 4852
+    },
+    {
+      "epoch": 0.3570843872154518,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.595765916765838e-05,
+      "loss": 0.77,
+      "step": 4853
+    },
+    {
+      "epoch": 0.35715796734881583,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.595245323982017e-05,
+      "loss": 0.8674,
+      "step": 4854
+    },
+    {
+      "epoch": 0.35723154748217983,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.594724672418448e-05,
+      "loss": 1.17,
+      "step": 4855
+    },
+    {
+      "epoch": 0.3573051276155438,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.594203962103073e-05,
+      "loss": 0.8265,
+      "step": 4856
+    },
+    {
+      "epoch": 0.3573787077489078,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.5936831930638395e-05,
+      "loss": 0.9776,
+      "step": 4857
+    },
+    {
+      "epoch": 0.3574522878822718,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.593162365328693e-05,
+      "loss": 1.1234,
+      "step": 4858
+    },
+    {
+      "epoch": 0.3575258680156358,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.5926414789255875e-05,
+      "loss": 0.9733,
+      "step": 4859
+    },
+    {
+      "epoch": 0.35759944814899974,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.592120533882477e-05,
+      "loss": 1.1785,
+      "step": 4860
+    },
+    {
+      "epoch": 0.35767302828236375,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.5915995302273214e-05,
+      "loss": 1.1252,
+      "step": 4861
+    },
+    {
+      "epoch": 0.35774660841572775,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.5910784679880805e-05,
+      "loss": 1.2581,
+      "step": 4862
+    },
+    {
+      "epoch": 0.35782018854909176,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.590557347192719e-05,
+      "loss": 1.158,
+      "step": 4863
+    },
+    {
+      "epoch": 0.35789376868245576,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.590036167869203e-05,
+      "loss": 0.921,
+      "step": 4864
+    },
+    {
+      "epoch": 0.3579673488158197,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.5895149300455063e-05,
+      "loss": 0.9467,
+      "step": 4865
+    },
+    {
+      "epoch": 0.3580409289491837,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.5889936337496e-05,
+      "loss": 1.5529,
+      "step": 4866
+    },
+    {
+      "epoch": 0.3581145090825477,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.5884722790094626e-05,
+      "loss": 1.187,
+      "step": 4867
+    },
+    {
+      "epoch": 0.3581880892159117,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.587950865853075e-05,
+      "loss": 1.0382,
+      "step": 4868
+    },
+    {
+      "epoch": 0.35826166934927567,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.587429394308418e-05,
+      "loss": 1.2916,
+      "step": 4869
+    },
+    {
+      "epoch": 0.3583352494826397,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.586907864403479e-05,
+      "loss": 0.7197,
+      "step": 4870
+    },
+    {
+      "epoch": 0.3584088296160037,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.5863862761662485e-05,
+      "loss": 0.9155,
+      "step": 4871
+    },
+    {
+      "epoch": 0.3584824097493677,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.5858646296247186e-05,
+      "loss": 1.0277,
+      "step": 4872
+    },
+    {
+      "epoch": 0.3585559898827317,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.585342924806884e-05,
+      "loss": 1.0115,
+      "step": 4873
+    },
+    {
+      "epoch": 0.35862957001609563,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.584821161740745e-05,
+      "loss": 0.9582,
+      "step": 4874
+    },
+    {
+      "epoch": 0.35870315014945964,
+      "grad_norm": 0.875,
+      "learning_rate": 3.5842993404543034e-05,
+      "loss": 1.0438,
+      "step": 4875
+    },
+    {
+      "epoch": 0.35877673028282364,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.583777460975564e-05,
+      "loss": 0.9466,
+      "step": 4876
+    },
+    {
+      "epoch": 0.35885031041618765,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.583255523332536e-05,
+      "loss": 1.8376,
+      "step": 4877
+    },
+    {
+      "epoch": 0.3589238905495516,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.5827335275532293e-05,
+      "loss": 1.1748,
+      "step": 4878
+    },
+    {
+      "epoch": 0.3589974706829156,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.582211473665659e-05,
+      "loss": 0.6493,
+      "step": 4879
+    },
+    {
+      "epoch": 0.3590710508162796,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.581689361697843e-05,
+      "loss": 1.1134,
+      "step": 4880
+    },
+    {
+      "epoch": 0.3591446309496436,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.581167191677802e-05,
+      "loss": 0.9733,
+      "step": 4881
+    },
+    {
+      "epoch": 0.3592182110830076,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.5806449636335606e-05,
+      "loss": 0.8222,
+      "step": 4882
+    },
+    {
+      "epoch": 0.35929179121637156,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.580122677593144e-05,
+      "loss": 1.1154,
+      "step": 4883
+    },
+    {
+      "epoch": 0.35936537134973556,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.579600333584584e-05,
+      "loss": 0.85,
+      "step": 4884
+    },
+    {
+      "epoch": 0.35943895148309957,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.579077931635913e-05,
+      "loss": 0.74,
+      "step": 4885
+    },
+    {
+      "epoch": 0.35951253161646357,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.578555471775167e-05,
+      "loss": 0.8637,
+      "step": 4886
+    },
+    {
+      "epoch": 0.3595861117498275,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.5780329540303865e-05,
+      "loss": 0.859,
+      "step": 4887
+    },
+    {
+      "epoch": 0.3596596918831915,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.5775103784296135e-05,
+      "loss": 0.8033,
+      "step": 4888
+    },
+    {
+      "epoch": 0.3597332720165555,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.5769877450008935e-05,
+      "loss": 0.6638,
+      "step": 4889
+    },
+    {
+      "epoch": 0.35980685214991953,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.576465053772275e-05,
+      "loss": 1.4388,
+      "step": 4890
+    },
+    {
+      "epoch": 0.35988043228328354,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.575942304771811e-05,
+      "loss": 1.4897,
+      "step": 4891
+    },
+    {
+      "epoch": 0.3599540124166475,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.5754194980275554e-05,
+      "loss": 1.3071,
+      "step": 4892
+    },
+    {
+      "epoch": 0.3600275925500115,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.574896633567566e-05,
+      "loss": 0.7576,
+      "step": 4893
+    },
+    {
+      "epoch": 0.3601011726833755,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.5743737114199045e-05,
+      "loss": 1.513,
+      "step": 4894
+    },
+    {
+      "epoch": 0.3601747528167395,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.573850731612636e-05,
+      "loss": 0.8965,
+      "step": 4895
+    },
+    {
+      "epoch": 0.36024833295010344,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.573327694173826e-05,
+      "loss": 0.7815,
+      "step": 4896
+    },
+    {
+      "epoch": 0.36032191308346745,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.572804599131546e-05,
+      "loss": 0.9349,
+      "step": 4897
+    },
+    {
+      "epoch": 0.36039549321683145,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.57228144651387e-05,
+      "loss": 0.8704,
+      "step": 4898
+    },
+    {
+      "epoch": 0.36046907335019546,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.5717582363488735e-05,
+      "loss": 1.1795,
+      "step": 4899
+    },
+    {
+      "epoch": 0.36054265348355946,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.5712349686646365e-05,
+      "loss": 0.9342,
+      "step": 4900
+    },
+    {
+      "epoch": 0.3606162336169234,
+      "grad_norm": 1.0,
+      "learning_rate": 3.5707116434892424e-05,
+      "loss": 1.1258,
+      "step": 4901
+    },
+    {
+      "epoch": 0.3606898137502874,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.570188260850777e-05,
+      "loss": 0.8858,
+      "step": 4902
+    },
+    {
+      "epoch": 0.3607633938836514,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.569664820777329e-05,
+      "loss": 0.8702,
+      "step": 4903
+    },
+    {
+      "epoch": 0.3608369740170154,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.56914132329699e-05,
+      "loss": 1.3725,
+      "step": 4904
+    },
+    {
+      "epoch": 0.36091055415037937,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.568617768437855e-05,
+      "loss": 0.6883,
+      "step": 4905
+    },
+    {
+      "epoch": 0.3609841342837434,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.5680941562280235e-05,
+      "loss": 0.8077,
+      "step": 4906
+    },
+    {
+      "epoch": 0.3610577144171074,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.5675704866955954e-05,
+      "loss": 1.11,
+      "step": 4907
+    },
+    {
+      "epoch": 0.3611312945504714,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.5670467598686756e-05,
+      "loss": 1.0383,
+      "step": 4908
+    },
+    {
+      "epoch": 0.3612048746838354,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.566522975775373e-05,
+      "loss": 0.7719,
+      "step": 4909
+    },
+    {
+      "epoch": 0.36127845481719933,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.565999134443795e-05,
+      "loss": 1.0769,
+      "step": 4910
+    },
+    {
+      "epoch": 0.36135203495056334,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.565475235902058e-05,
+      "loss": 1.0033,
+      "step": 4911
+    },
+    {
+      "epoch": 0.36142561508392734,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.5649512801782764e-05,
+      "loss": 1.0606,
+      "step": 4912
+    },
+    {
+      "epoch": 0.36149919521729135,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.5644272673005715e-05,
+      "loss": 0.9322,
+      "step": 4913
+    },
+    {
+      "epoch": 0.3615727753506553,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.5639031972970646e-05,
+      "loss": 0.9531,
+      "step": 4914
+    },
+    {
+      "epoch": 0.3616463554840193,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.563379070195883e-05,
+      "loss": 0.8194,
+      "step": 4915
+    },
+    {
+      "epoch": 0.3617199356173833,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.5628548860251565e-05,
+      "loss": 0.8207,
+      "step": 4916
+    },
+    {
+      "epoch": 0.3617935157507473,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.562330644813015e-05,
+      "loss": 1.0512,
+      "step": 4917
+    },
+    {
+      "epoch": 0.3618670958841113,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.561806346587594e-05,
+      "loss": 0.8682,
+      "step": 4918
+    },
+    {
+      "epoch": 0.36194067601747526,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.5612819913770325e-05,
+      "loss": 0.864,
+      "step": 4919
+    },
+    {
+      "epoch": 0.36201425615083926,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.56075757920947e-05,
+      "loss": 0.788,
+      "step": 4920
+    },
+    {
+      "epoch": 0.36208783628420327,
+      "grad_norm": 1.4765625,
+      "learning_rate": 3.560233110113052e-05,
+      "loss": 1.4926,
+      "step": 4921
+    },
+    {
+      "epoch": 0.3621614164175673,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.559708584115925e-05,
+      "loss": 0.6587,
+      "step": 4922
+    },
+    {
+      "epoch": 0.3622349965509312,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.559184001246241e-05,
+      "loss": 0.9058,
+      "step": 4923
+    },
+    {
+      "epoch": 0.3623085766842952,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.558659361532152e-05,
+      "loss": 0.9397,
+      "step": 4924
+    },
+    {
+      "epoch": 0.36238215681765923,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.558134665001814e-05,
+      "loss": 1.0313,
+      "step": 4925
+    },
+    {
+      "epoch": 0.36245573695102323,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.5576099116833874e-05,
+      "loss": 0.9287,
+      "step": 4926
+    },
+    {
+      "epoch": 0.36252931708438724,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.557085101605034e-05,
+      "loss": 0.8493,
+      "step": 4927
+    },
+    {
+      "epoch": 0.3626028972177512,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.55656023479492e-05,
+      "loss": 0.6525,
+      "step": 4928
+    },
+    {
+      "epoch": 0.3626764773511152,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.556035311281213e-05,
+      "loss": 1.2842,
+      "step": 4929
+    },
+    {
+      "epoch": 0.3627500574844792,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.555510331092087e-05,
+      "loss": 1.1034,
+      "step": 4930
+    },
+    {
+      "epoch": 0.3628236376178432,
+      "grad_norm": 1.6171875,
+      "learning_rate": 3.554985294255714e-05,
+      "loss": 1.0617,
+      "step": 4931
+    },
+    {
+      "epoch": 0.3628972177512072,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.554460200800273e-05,
+      "loss": 0.876,
+      "step": 4932
+    },
+    {
+      "epoch": 0.36297079788457115,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.553935050753945e-05,
+      "loss": 0.7704,
+      "step": 4933
+    },
+    {
+      "epoch": 0.36304437801793515,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.553409844144913e-05,
+      "loss": 0.6025,
+      "step": 4934
+    },
+    {
+      "epoch": 0.36311795815129916,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.552884581001364e-05,
+      "loss": 0.7969,
+      "step": 4935
+    },
+    {
+      "epoch": 0.36319153828466316,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.5523592613514886e-05,
+      "loss": 0.7456,
+      "step": 4936
+    },
+    {
+      "epoch": 0.3632651184180271,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.551833885223479e-05,
+      "loss": 1.182,
+      "step": 4937
+    },
+    {
+      "epoch": 0.3633386985513911,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.551308452645532e-05,
+      "loss": 1.0166,
+      "step": 4938
+    },
+    {
+      "epoch": 0.3634122786847551,
+      "grad_norm": 1.0,
+      "learning_rate": 3.5507829636458454e-05,
+      "loss": 1.0311,
+      "step": 4939
+    },
+    {
+      "epoch": 0.3634858588181191,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.550257418252622e-05,
+      "loss": 0.6928,
+      "step": 4940
+    },
+    {
+      "epoch": 0.3635594389514831,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.5497318164940665e-05,
+      "loss": 1.1217,
+      "step": 4941
+    },
+    {
+      "epoch": 0.3636330190848471,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.549206158398387e-05,
+      "loss": 1.3708,
+      "step": 4942
+    },
+    {
+      "epoch": 0.3637065992182111,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.5486804439937954e-05,
+      "loss": 0.8204,
+      "step": 4943
+    },
+    {
+      "epoch": 0.3637801793515751,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.548154673308504e-05,
+      "loss": 0.6268,
+      "step": 4944
+    },
+    {
+      "epoch": 0.3638537594849391,
+      "grad_norm": 1.21875,
+      "learning_rate": 3.547628846370731e-05,
+      "loss": 1.4384,
+      "step": 4945
+    },
+    {
+      "epoch": 0.36392733961830304,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.547102963208698e-05,
+      "loss": 0.9204,
+      "step": 4946
+    },
+    {
+      "epoch": 0.36400091975166704,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.546577023850625e-05,
+      "loss": 0.8621,
+      "step": 4947
+    },
+    {
+      "epoch": 0.36407449988503104,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.546051028324741e-05,
+      "loss": 0.9103,
+      "step": 4948
+    },
+    {
+      "epoch": 0.36414808001839505,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.545524976659273e-05,
+      "loss": 0.9211,
+      "step": 4949
+    },
+    {
+      "epoch": 0.36422166015175905,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.544998868882455e-05,
+      "loss": 1.1029,
+      "step": 4950
+    },
+    {
+      "epoch": 0.364295240285123,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.544472705022521e-05,
+      "loss": 0.831,
+      "step": 4951
+    },
+    {
+      "epoch": 0.364368820418487,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.5439464851077096e-05,
+      "loss": 0.7651,
+      "step": 4952
+    },
+    {
+      "epoch": 0.364442400551851,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.5434202091662625e-05,
+      "loss": 1.0167,
+      "step": 4953
+    },
+    {
+      "epoch": 0.364515980685215,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.542893877226423e-05,
+      "loss": 1.185,
+      "step": 4954
+    },
+    {
+      "epoch": 0.36458956081857896,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.542367489316439e-05,
+      "loss": 0.7352,
+      "step": 4955
+    },
+    {
+      "epoch": 0.36466314095194297,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.5418410454645606e-05,
+      "loss": 0.7912,
+      "step": 4956
+    },
+    {
+      "epoch": 0.36473672108530697,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.5413145456990414e-05,
+      "loss": 0.9649,
+      "step": 4957
+    },
+    {
+      "epoch": 0.364810301218671,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.540787990048137e-05,
+      "loss": 0.7054,
+      "step": 4958
+    },
+    {
+      "epoch": 0.364883881352035,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.5402613785401076e-05,
+      "loss": 0.8387,
+      "step": 4959
+    },
+    {
+      "epoch": 0.3649574614853989,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.539734711203214e-05,
+      "loss": 0.9519,
+      "step": 4960
+    },
+    {
+      "epoch": 0.36503104161876293,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.539207988065722e-05,
+      "loss": 0.8418,
+      "step": 4961
+    },
+    {
+      "epoch": 0.36510462175212693,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.5386812091559e-05,
+      "loss": 0.8378,
+      "step": 4962
+    },
+    {
+      "epoch": 0.36517820188549094,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.53815437450202e-05,
+      "loss": 1.0,
+      "step": 4963
+    },
+    {
+      "epoch": 0.3652517820188549,
+      "grad_norm": 1.125,
+      "learning_rate": 3.5376274841323557e-05,
+      "loss": 1.3167,
+      "step": 4964
+    },
+    {
+      "epoch": 0.3653253621522189,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.537100538075183e-05,
+      "loss": 1.0696,
+      "step": 4965
+    },
+    {
+      "epoch": 0.3653989422855829,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.536573536358785e-05,
+      "loss": 1.2404,
+      "step": 4966
+    },
+    {
+      "epoch": 0.3654725224189469,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.5360464790114416e-05,
+      "loss": 0.9548,
+      "step": 4967
+    },
+    {
+      "epoch": 0.3655461025523109,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.535519366061441e-05,
+      "loss": 1.042,
+      "step": 4968
+    },
+    {
+      "epoch": 0.36561968268567485,
+      "grad_norm": 1.0,
+      "learning_rate": 3.534992197537071e-05,
+      "loss": 1.0249,
+      "step": 4969
+    },
+    {
+      "epoch": 0.36569326281903886,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.5344649734666256e-05,
+      "loss": 0.7701,
+      "step": 4970
+    },
+    {
+      "epoch": 0.36576684295240286,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.533937693878399e-05,
+      "loss": 0.9024,
+      "step": 4971
+    },
+    {
+      "epoch": 0.36584042308576686,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.533410358800689e-05,
+      "loss": 1.2592,
+      "step": 4972
+    },
+    {
+      "epoch": 0.3659140032191308,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.532882968261797e-05,
+      "loss": 0.8673,
+      "step": 4973
+    },
+    {
+      "epoch": 0.3659875833524948,
+      "grad_norm": 1.2578125,
+      "learning_rate": 3.532355522290026e-05,
+      "loss": 1.1793,
+      "step": 4974
+    },
+    {
+      "epoch": 0.3660611634858588,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.531828020913685e-05,
+      "loss": 0.9704,
+      "step": 4975
+    },
+    {
+      "epoch": 0.3661347436192228,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.5313004641610825e-05,
+      "loss": 0.7921,
+      "step": 4976
+    },
+    {
+      "epoch": 0.36620832375258683,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.530772852060532e-05,
+      "loss": 0.616,
+      "step": 4977
+    },
+    {
+      "epoch": 0.3662819038859508,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.5302451846403496e-05,
+      "loss": 1.0702,
+      "step": 4978
+    },
+    {
+      "epoch": 0.3663554840193148,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.5297174619288536e-05,
+      "loss": 0.9164,
+      "step": 4979
+    },
+    {
+      "epoch": 0.3664290641526788,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.529189683954367e-05,
+      "loss": 0.931,
+      "step": 4980
+    },
+    {
+      "epoch": 0.3665026442860428,
+      "grad_norm": 1.1171875,
+      "learning_rate": 3.528661850745213e-05,
+      "loss": 0.9794,
+      "step": 4981
+    },
+    {
+      "epoch": 0.36657622441940674,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.528133962329721e-05,
+      "loss": 0.8532,
+      "step": 4982
+    },
+    {
+      "epoch": 0.36664980455277074,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.52760601873622e-05,
+      "loss": 1.0545,
+      "step": 4983
+    },
+    {
+      "epoch": 0.36672338468613475,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.5270780199930465e-05,
+      "loss": 1.0776,
+      "step": 4984
+    },
+    {
+      "epoch": 0.36679696481949875,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.526549966128535e-05,
+      "loss": 1.0756,
+      "step": 4985
+    },
+    {
+      "epoch": 0.36687054495286275,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.526021857171025e-05,
+      "loss": 1.2265,
+      "step": 4986
+    },
+    {
+      "epoch": 0.3669441250862267,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.525493693148861e-05,
+      "loss": 1.0376,
+      "step": 4987
+    },
+    {
+      "epoch": 0.3670177052195907,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.524965474090386e-05,
+      "loss": 0.8881,
+      "step": 4988
+    },
+    {
+      "epoch": 0.3670912853529547,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.524437200023951e-05,
+      "loss": 0.7254,
+      "step": 4989
+    },
+    {
+      "epoch": 0.3671648654863187,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.523908870977906e-05,
+      "loss": 0.8913,
+      "step": 4990
+    },
+    {
+      "epoch": 0.36723844561968266,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.523380486980605e-05,
+      "loss": 0.9651,
+      "step": 4991
+    },
+    {
+      "epoch": 0.36731202575304667,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.522852048060407e-05,
+      "loss": 0.7655,
+      "step": 4992
+    },
+    {
+      "epoch": 0.36738560588641067,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.522323554245671e-05,
+      "loss": 1.1908,
+      "step": 4993
+    },
+    {
+      "epoch": 0.3674591860197747,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.5217950055647616e-05,
+      "loss": 1.0507,
+      "step": 4994
+    },
+    {
+      "epoch": 0.3675327661531387,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.521266402046044e-05,
+      "loss": 0.756,
+      "step": 4995
+    },
+    {
+      "epoch": 0.3676063462865026,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.520737743717886e-05,
+      "loss": 0.7442,
+      "step": 4996
+    },
+    {
+      "epoch": 0.36767992641986663,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.520209030608662e-05,
+      "loss": 0.8751,
+      "step": 4997
+    },
+    {
+      "epoch": 0.36775350655323064,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.519680262746747e-05,
+      "loss": 1.0225,
+      "step": 4998
+    },
+    {
+      "epoch": 0.36782708668659464,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.5191514401605176e-05,
+      "loss": 0.9934,
+      "step": 4999
+    },
+    {
+      "epoch": 0.3679006668199586,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.518622562878356e-05,
+      "loss": 1.2016,
+      "step": 5000
+    },
+    {
+      "epoch": 0.3679742469533226,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.518093630928644e-05,
+      "loss": 0.695,
+      "step": 5001
+    },
+    {
+      "epoch": 0.3680478270866866,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.51756464433977e-05,
+      "loss": 1.1051,
+      "step": 5002
+    },
+    {
+      "epoch": 0.3681214072200506,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.517035603140125e-05,
+      "loss": 0.8496,
+      "step": 5003
+    },
+    {
+      "epoch": 0.3681949873534146,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.516506507358099e-05,
+      "loss": 0.8501,
+      "step": 5004
+    },
+    {
+      "epoch": 0.36826856748677855,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.515977357022089e-05,
+      "loss": 0.8302,
+      "step": 5005
+    },
+    {
+      "epoch": 0.36834214762014256,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.515448152160492e-05,
+      "loss": 0.8431,
+      "step": 5006
+    },
+    {
+      "epoch": 0.36841572775350656,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.514918892801712e-05,
+      "loss": 1.1764,
+      "step": 5007
+    },
+    {
+      "epoch": 0.36848930788687057,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.5143895789741516e-05,
+      "loss": 1.0617,
+      "step": 5008
+    },
+    {
+      "epoch": 0.3685628880202345,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.5138602107062174e-05,
+      "loss": 0.7635,
+      "step": 5009
+    },
+    {
+      "epoch": 0.3686364681535985,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.513330788026322e-05,
+      "loss": 0.8319,
+      "step": 5010
+    },
+    {
+      "epoch": 0.3687100482869625,
+      "grad_norm": 1.1484375,
+      "learning_rate": 3.512801310962876e-05,
+      "loss": 0.9136,
+      "step": 5011
+    },
+    {
+      "epoch": 0.3687836284203265,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.512271779544297e-05,
+      "loss": 0.8254,
+      "step": 5012
+    },
+    {
+      "epoch": 0.36885720855369053,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.5117421937990036e-05,
+      "loss": 1.3523,
+      "step": 5013
+    },
+    {
+      "epoch": 0.3689307886870545,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.511212553755418e-05,
+      "loss": 1.2158,
+      "step": 5014
+    },
+    {
+      "epoch": 0.3690043688204185,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.510682859441964e-05,
+      "loss": 1.1216,
+      "step": 5015
+    },
+    {
+      "epoch": 0.3690779489537825,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.51015311088707e-05,
+      "loss": 1.0272,
+      "step": 5016
+    },
+    {
+      "epoch": 0.3691515290871465,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.509623308119166e-05,
+      "loss": 1.136,
+      "step": 5017
+    },
+    {
+      "epoch": 0.36922510922051044,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.509093451166686e-05,
+      "loss": 1.0315,
+      "step": 5018
+    },
+    {
+      "epoch": 0.36929868935387444,
+      "grad_norm": 1.1640625,
+      "learning_rate": 3.5085635400580675e-05,
+      "loss": 1.2436,
+      "step": 5019
+    },
+    {
+      "epoch": 0.36937226948723845,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.508033574821747e-05,
+      "loss": 1.3433,
+      "step": 5020
+    },
+    {
+      "epoch": 0.36944584962060245,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.50750355548617e-05,
+      "loss": 0.6279,
+      "step": 5021
+    },
+    {
+      "epoch": 0.36951942975396646,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.506973482079779e-05,
+      "loss": 0.6971,
+      "step": 5022
+    },
+    {
+      "epoch": 0.3695930098873304,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.5064433546310236e-05,
+      "loss": 0.6801,
+      "step": 5023
+    },
+    {
+      "epoch": 0.3696665900206944,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.505913173168354e-05,
+      "loss": 0.8268,
+      "step": 5024
+    },
+    {
+      "epoch": 0.3697401701540584,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.505382937720225e-05,
+      "loss": 0.9737,
+      "step": 5025
+    },
+    {
+      "epoch": 0.3698137502874224,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.504852648315092e-05,
+      "loss": 0.8916,
+      "step": 5026
+    },
+    {
+      "epoch": 0.36988733042078636,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.5043223049814156e-05,
+      "loss": 0.9102,
+      "step": 5027
+    },
+    {
+      "epoch": 0.36996091055415037,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.5037919077476576e-05,
+      "loss": 1.4119,
+      "step": 5028
+    },
+    {
+      "epoch": 0.3700344906875144,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.503261456642284e-05,
+      "loss": 0.8671,
+      "step": 5029
+    },
+    {
+      "epoch": 0.3701080708208784,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.502730951693763e-05,
+      "loss": 1.2578,
+      "step": 5030
+    },
+    {
+      "epoch": 0.3701816509542424,
+      "grad_norm": 0.875,
+      "learning_rate": 3.502200392930565e-05,
+      "loss": 0.7428,
+      "step": 5031
+    },
+    {
+      "epoch": 0.37025523108760633,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.501669780381166e-05,
+      "loss": 0.8569,
+      "step": 5032
+    },
+    {
+      "epoch": 0.37032881122097033,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.501139114074041e-05,
+      "loss": 0.7413,
+      "step": 5033
+    },
+    {
+      "epoch": 0.37040239135433434,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.5006083940376714e-05,
+      "loss": 0.9027,
+      "step": 5034
+    },
+    {
+      "epoch": 0.37047597148769834,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.5000776203005384e-05,
+      "loss": 1.1972,
+      "step": 5035
+    },
+    {
+      "epoch": 0.3705495516210623,
+      "grad_norm": 0.66015625,
+      "learning_rate": 3.499546792891128e-05,
+      "loss": 0.7057,
+      "step": 5036
+    },
+    {
+      "epoch": 0.3706231317544263,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.49901591183793e-05,
+      "loss": 0.7118,
+      "step": 5037
+    },
+    {
+      "epoch": 0.3706967118877903,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.4984849771694345e-05,
+      "loss": 0.8855,
+      "step": 5038
+    },
+    {
+      "epoch": 0.3707702920211543,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.497953988914137e-05,
+      "loss": 0.9269,
+      "step": 5039
+    },
+    {
+      "epoch": 0.3708438721545183,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.497422947100533e-05,
+      "loss": 1.2386,
+      "step": 5040
+    },
+    {
+      "epoch": 0.37091745228788225,
+      "grad_norm": 0.66796875,
+      "learning_rate": 3.496891851757123e-05,
+      "loss": 0.5296,
+      "step": 5041
+    },
+    {
+      "epoch": 0.37099103242124626,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.4963607029124094e-05,
+      "loss": 0.9789,
+      "step": 5042
+    },
+    {
+      "epoch": 0.37106461255461026,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.4958295005949e-05,
+      "loss": 0.7716,
+      "step": 5043
+    },
+    {
+      "epoch": 0.37113819268797427,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.495298244833102e-05,
+      "loss": 0.9389,
+      "step": 5044
+    },
+    {
+      "epoch": 0.3712117728213382,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.494766935655527e-05,
+      "loss": 0.5788,
+      "step": 5045
+    },
+    {
+      "epoch": 0.3712853529547022,
+      "grad_norm": 1.1328125,
+      "learning_rate": 3.49423557309069e-05,
+      "loss": 0.6275,
+      "step": 5046
+    },
+    {
+      "epoch": 0.3713589330880662,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.493704157167107e-05,
+      "loss": 0.8667,
+      "step": 5047
+    },
+    {
+      "epoch": 0.3714325132214302,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.493172687913299e-05,
+      "loss": 0.875,
+      "step": 5048
+    },
+    {
+      "epoch": 0.37150609335479423,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.492641165357788e-05,
+      "loss": 0.8389,
+      "step": 5049
+    },
+    {
+      "epoch": 0.3715796734881582,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.492109589529101e-05,
+      "loss": 1.293,
+      "step": 5050
+    },
+    {
+      "epoch": 0.3716532536215222,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.491577960455766e-05,
+      "loss": 0.9775,
+      "step": 5051
+    },
+    {
+      "epoch": 0.3717268337548862,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.4910462781663146e-05,
+      "loss": 0.8613,
+      "step": 5052
+    },
+    {
+      "epoch": 0.3718004138882502,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.4905145426892824e-05,
+      "loss": 1.4802,
+      "step": 5053
+    },
+    {
+      "epoch": 0.37187399402161414,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.489982754053204e-05,
+      "loss": 0.7911,
+      "step": 5054
+    },
+    {
+      "epoch": 0.37194757415497814,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.4894509122866216e-05,
+      "loss": 0.7443,
+      "step": 5055
+    },
+    {
+      "epoch": 0.37202115428834215,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.4889190174180776e-05,
+      "loss": 0.6426,
+      "step": 5056
+    },
+    {
+      "epoch": 0.37209473442170615,
+      "grad_norm": 1.0,
+      "learning_rate": 3.4883870694761175e-05,
+      "loss": 1.1715,
+      "step": 5057
+    },
+    {
+      "epoch": 0.37216831455507016,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.48785506848929e-05,
+      "loss": 0.7959,
+      "step": 5058
+    },
+    {
+      "epoch": 0.3722418946884341,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.487323014486147e-05,
+      "loss": 0.7008,
+      "step": 5059
+    },
+    {
+      "epoch": 0.3723154748217981,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.486790907495243e-05,
+      "loss": 0.5899,
+      "step": 5060
+    },
+    {
+      "epoch": 0.3723890549551621,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.486258747545135e-05,
+      "loss": 0.7655,
+      "step": 5061
+    },
+    {
+      "epoch": 0.3724626350885261,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.485726534664382e-05,
+      "loss": 0.7658,
+      "step": 5062
+    },
+    {
+      "epoch": 0.37253621522189007,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.485194268881547e-05,
+      "loss": 1.0227,
+      "step": 5063
+    },
+    {
+      "epoch": 0.37260979535525407,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.484661950225198e-05,
+      "loss": 1.016,
+      "step": 5064
+    },
+    {
+      "epoch": 0.3726833754886181,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.4841295787239015e-05,
+      "loss": 1.0784,
+      "step": 5065
+    },
+    {
+      "epoch": 0.3727569556219821,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.483597154406228e-05,
+      "loss": 0.9107,
+      "step": 5066
+    },
+    {
+      "epoch": 0.3728305357553461,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.4830646773007544e-05,
+      "loss": 1.178,
+      "step": 5067
+    },
+    {
+      "epoch": 0.37290411588871003,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.482532147436056e-05,
+      "loss": 1.0666,
+      "step": 5068
+    },
+    {
+      "epoch": 0.37297769602207403,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.4819995648407125e-05,
+      "loss": 1.0449,
+      "step": 5069
+    },
+    {
+      "epoch": 0.37305127615543804,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.481466929543308e-05,
+      "loss": 0.8709,
+      "step": 5070
+    },
+    {
+      "epoch": 0.37312485628880204,
+      "grad_norm": 0.875,
+      "learning_rate": 3.4809342415724266e-05,
+      "loss": 0.8263,
+      "step": 5071
+    },
+    {
+      "epoch": 0.373198436422166,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.480401500956657e-05,
+      "loss": 0.7905,
+      "step": 5072
+    },
+    {
+      "epoch": 0.37327201655553,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.479868707724591e-05,
+      "loss": 1.0048,
+      "step": 5073
+    },
+    {
+      "epoch": 0.373345596688894,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.479335861904822e-05,
+      "loss": 0.8132,
+      "step": 5074
+    },
+    {
+      "epoch": 0.373419176822258,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.478802963525947e-05,
+      "loss": 0.8332,
+      "step": 5075
+    },
+    {
+      "epoch": 0.373492756955622,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.478270012616565e-05,
+      "loss": 0.8352,
+      "step": 5076
+    },
+    {
+      "epoch": 0.37356633708898596,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.477737009205279e-05,
+      "loss": 0.9812,
+      "step": 5077
+    },
+    {
+      "epoch": 0.37363991722234996,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.4772039533206954e-05,
+      "loss": 0.9344,
+      "step": 5078
+    },
+    {
+      "epoch": 0.37371349735571396,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.476670844991421e-05,
+      "loss": 0.9901,
+      "step": 5079
+    },
+    {
+      "epoch": 0.37378707748907797,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.476137684246067e-05,
+      "loss": 1.0108,
+      "step": 5080
+    },
+    {
+      "epoch": 0.3738606576224419,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.475604471113247e-05,
+      "loss": 1.5467,
+      "step": 5081
+    },
+    {
+      "epoch": 0.3739342377558059,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.475071205621578e-05,
+      "loss": 1.1166,
+      "step": 5082
+    },
+    {
+      "epoch": 0.3740078178891699,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.474537887799678e-05,
+      "loss": 0.8274,
+      "step": 5083
+    },
+    {
+      "epoch": 0.37408139802253393,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.474004517676171e-05,
+      "loss": 0.6075,
+      "step": 5084
+    },
+    {
+      "epoch": 0.37415497815589793,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.473471095279682e-05,
+      "loss": 0.9428,
+      "step": 5085
+    },
+    {
+      "epoch": 0.3742285582892619,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.472937620638837e-05,
+      "loss": 0.9936,
+      "step": 5086
+    },
+    {
+      "epoch": 0.3743021384226259,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.472404093782268e-05,
+      "loss": 0.9604,
+      "step": 5087
+    },
+    {
+      "epoch": 0.3743757185559899,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.4718705147386075e-05,
+      "loss": 0.8169,
+      "step": 5088
+    },
+    {
+      "epoch": 0.3744492986893539,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.471336883536491e-05,
+      "loss": 0.7736,
+      "step": 5089
+    },
+    {
+      "epoch": 0.37452287882271784,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.47080320020456e-05,
+      "loss": 0.9711,
+      "step": 5090
+    },
+    {
+      "epoch": 0.37459645895608185,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.470269464771454e-05,
+      "loss": 0.9966,
+      "step": 5091
+    },
+    {
+      "epoch": 0.37467003908944585,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.469735677265819e-05,
+      "loss": 0.7787,
+      "step": 5092
+    },
+    {
+      "epoch": 0.37474361922280985,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.4692018377163024e-05,
+      "loss": 1.1409,
+      "step": 5093
+    },
+    {
+      "epoch": 0.37481719935617386,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.4686679461515534e-05,
+      "loss": 1.6977,
+      "step": 5094
+    },
+    {
+      "epoch": 0.3748907794895378,
+      "grad_norm": 4.625,
+      "learning_rate": 3.4681340026002243e-05,
+      "loss": 0.7545,
+      "step": 5095
+    },
+    {
+      "epoch": 0.3749643596229018,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.467600007090972e-05,
+      "loss": 0.9121,
+      "step": 5096
+    },
+    {
+      "epoch": 0.3750379397562658,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.467065959652456e-05,
+      "loss": 1.0055,
+      "step": 5097
+    },
+    {
+      "epoch": 0.3751115198896298,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.466531860313335e-05,
+      "loss": 0.9226,
+      "step": 5098
+    },
+    {
+      "epoch": 0.37518510002299377,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.465997709102276e-05,
+      "loss": 1.2116,
+      "step": 5099
+    },
+    {
+      "epoch": 0.37525868015635777,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.4654635060479434e-05,
+      "loss": 0.8605,
+      "step": 5100
+    },
+    {
+      "epoch": 0.3753322602897218,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.4649292511790085e-05,
+      "loss": 0.8016,
+      "step": 5101
+    },
+    {
+      "epoch": 0.3754058404230858,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.464394944524143e-05,
+      "loss": 0.6412,
+      "step": 5102
+    },
+    {
+      "epoch": 0.3754794205564498,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.463860586112022e-05,
+      "loss": 1.0554,
+      "step": 5103
+    },
+    {
+      "epoch": 0.37555300068981373,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.463326175971324e-05,
+      "loss": 0.8955,
+      "step": 5104
+    },
+    {
+      "epoch": 0.37562658082317774,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.4627917141307295e-05,
+      "loss": 0.8602,
+      "step": 5105
+    },
+    {
+      "epoch": 0.37570016095654174,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.462257200618923e-05,
+      "loss": 0.9276,
+      "step": 5106
+    },
+    {
+      "epoch": 0.37577374108990574,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.4617226354645894e-05,
+      "loss": 0.6035,
+      "step": 5107
+    },
+    {
+      "epoch": 0.3758473212232697,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.4611880186964185e-05,
+      "loss": 1.0858,
+      "step": 5108
+    },
+    {
+      "epoch": 0.3759209013566337,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.4606533503431015e-05,
+      "loss": 0.9035,
+      "step": 5109
+    },
+    {
+      "epoch": 0.3759944814899977,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.460118630433334e-05,
+      "loss": 0.7965,
+      "step": 5110
+    },
+    {
+      "epoch": 0.3760680616233617,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.459583858995813e-05,
+      "loss": 0.9757,
+      "step": 5111
+    },
+    {
+      "epoch": 0.3761416417567257,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.459049036059239e-05,
+      "loss": 0.8417,
+      "step": 5112
+    },
+    {
+      "epoch": 0.37621522189008966,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.458514161652314e-05,
+      "loss": 1.2755,
+      "step": 5113
+    },
+    {
+      "epoch": 0.37628880202345366,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.457979235803744e-05,
+      "loss": 1.1431,
+      "step": 5114
+    },
+    {
+      "epoch": 0.37636238215681767,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.45744425854224e-05,
+      "loss": 1.1099,
+      "step": 5115
+    },
+    {
+      "epoch": 0.37643596229018167,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.456909229896509e-05,
+      "loss": 0.772,
+      "step": 5116
+    },
+    {
+      "epoch": 0.3765095424235456,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.4563741498952674e-05,
+      "loss": 0.8679,
+      "step": 5117
+    },
+    {
+      "epoch": 0.3765831225569096,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.455839018567231e-05,
+      "loss": 0.796,
+      "step": 5118
+    },
+    {
+      "epoch": 0.3766567026902736,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.455303835941121e-05,
+      "loss": 0.7404,
+      "step": 5119
+    },
+    {
+      "epoch": 0.37673028282363763,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.4547686020456574e-05,
+      "loss": 0.604,
+      "step": 5120
+    },
+    {
+      "epoch": 0.37680386295700163,
+      "grad_norm": 0.625,
+      "learning_rate": 3.454233316909567e-05,
+      "loss": 0.6039,
+      "step": 5121
+    },
+    {
+      "epoch": 0.3768774430903656,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.453697980561576e-05,
+      "loss": 0.8094,
+      "step": 5122
+    },
+    {
+      "epoch": 0.3769510232237296,
+      "grad_norm": 0.875,
+      "learning_rate": 3.4531625930304155e-05,
+      "loss": 0.9612,
+      "step": 5123
+    },
+    {
+      "epoch": 0.3770246033570936,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.4526271543448196e-05,
+      "loss": 0.601,
+      "step": 5124
+    },
+    {
+      "epoch": 0.3770981834904576,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.452091664533523e-05,
+      "loss": 0.9024,
+      "step": 5125
+    },
+    {
+      "epoch": 0.37717176362382154,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.451556123625266e-05,
+      "loss": 0.7551,
+      "step": 5126
+    },
+    {
+      "epoch": 0.37724534375718555,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.4510205316487885e-05,
+      "loss": 0.9634,
+      "step": 5127
+    },
+    {
+      "epoch": 0.37731892389054955,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.450484888632836e-05,
+      "loss": 1.0097,
+      "step": 5128
+    },
+    {
+      "epoch": 0.37739250402391356,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.4499491946061534e-05,
+      "loss": 0.6972,
+      "step": 5129
+    },
+    {
+      "epoch": 0.37746608415727756,
+      "grad_norm": 0.875,
+      "learning_rate": 3.449413449597492e-05,
+      "loss": 0.7327,
+      "step": 5130
+    },
+    {
+      "epoch": 0.3775396642906415,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.448877653635604e-05,
+      "loss": 1.0995,
+      "step": 5131
+    },
+    {
+      "epoch": 0.3776132444240055,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.448341806749245e-05,
+      "loss": 1.0709,
+      "step": 5132
+    },
+    {
+      "epoch": 0.3776868245573695,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.4478059089671725e-05,
+      "loss": 0.7534,
+      "step": 5133
+    },
+    {
+      "epoch": 0.3777604046907335,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.447269960318147e-05,
+      "loss": 1.3757,
+      "step": 5134
+    },
+    {
+      "epoch": 0.37783398482409747,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.4467339608309316e-05,
+      "loss": 0.6253,
+      "step": 5135
+    },
+    {
+      "epoch": 0.3779075649574615,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.4461979105342925e-05,
+      "loss": 0.7474,
+      "step": 5136
+    },
+    {
+      "epoch": 0.3779811450908255,
+      "grad_norm": 1.140625,
+      "learning_rate": 3.445661809456999e-05,
+      "loss": 1.2046,
+      "step": 5137
+    },
+    {
+      "epoch": 0.3780547252241895,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.4451256576278215e-05,
+      "loss": 0.9753,
+      "step": 5138
+    },
+    {
+      "epoch": 0.3781283053575535,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.4445894550755356e-05,
+      "loss": 0.7771,
+      "step": 5139
+    },
+    {
+      "epoch": 0.37820188549091743,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.444053201828918e-05,
+      "loss": 1.2612,
+      "step": 5140
+    },
+    {
+      "epoch": 0.37827546562428144,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.4435168979167476e-05,
+      "loss": 1.3281,
+      "step": 5141
+    },
+    {
+      "epoch": 0.37834904575764544,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.442980543367808e-05,
+      "loss": 1.2024,
+      "step": 5142
+    },
+    {
+      "epoch": 0.37842262589100945,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.442444138210883e-05,
+      "loss": 1.2625,
+      "step": 5143
+    },
+    {
+      "epoch": 0.3784962060243734,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.441907682474762e-05,
+      "loss": 0.888,
+      "step": 5144
+    },
+    {
+      "epoch": 0.3785697861577374,
+      "grad_norm": 0.75,
+      "learning_rate": 3.441371176188233e-05,
+      "loss": 0.782,
+      "step": 5145
+    },
+    {
+      "epoch": 0.3786433662911014,
+      "grad_norm": 1.0,
+      "learning_rate": 3.440834619380092e-05,
+      "loss": 0.8574,
+      "step": 5146
+    },
+    {
+      "epoch": 0.3787169464244654,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.4402980120791345e-05,
+      "loss": 1.3946,
+      "step": 5147
+    },
+    {
+      "epoch": 0.3787905265578294,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.439761354314158e-05,
+      "loss": 1.2674,
+      "step": 5148
+    },
+    {
+      "epoch": 0.37886410669119336,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.4392246461139656e-05,
+      "loss": 0.7175,
+      "step": 5149
+    },
+    {
+      "epoch": 0.37893768682455736,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.4386878875073594e-05,
+      "loss": 0.9561,
+      "step": 5150
+    },
+    {
+      "epoch": 0.37901126695792137,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.438151078523147e-05,
+      "loss": 0.6987,
+      "step": 5151
+    },
+    {
+      "epoch": 0.37908484709128537,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.4376142191901385e-05,
+      "loss": 0.8381,
+      "step": 5152
+    },
+    {
+      "epoch": 0.3791584272246493,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.437077309537146e-05,
+      "loss": 1.326,
+      "step": 5153
+    },
+    {
+      "epoch": 0.3792320073580133,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.436540349592984e-05,
+      "loss": 0.779,
+      "step": 5154
+    },
+    {
+      "epoch": 0.3793055874913773,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.4360033393864696e-05,
+      "loss": 0.8868,
+      "step": 5155
+    },
+    {
+      "epoch": 0.37937916762474133,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.4354662789464246e-05,
+      "loss": 0.8471,
+      "step": 5156
+    },
+    {
+      "epoch": 0.37945274775810534,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.4349291683016715e-05,
+      "loss": 0.8737,
+      "step": 5157
+    },
+    {
+      "epoch": 0.3795263278914693,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.434392007481035e-05,
+      "loss": 0.7175,
+      "step": 5158
+    },
+    {
+      "epoch": 0.3795999080248333,
+      "grad_norm": 1.0,
+      "learning_rate": 3.433854796513344e-05,
+      "loss": 0.786,
+      "step": 5159
+    },
+    {
+      "epoch": 0.3796734881581973,
+      "grad_norm": 1.0,
+      "learning_rate": 3.433317535427432e-05,
+      "loss": 0.9675,
+      "step": 5160
+    },
+    {
+      "epoch": 0.3797470682915613,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.432780224252129e-05,
+      "loss": 0.8816,
+      "step": 5161
+    },
+    {
+      "epoch": 0.37982064842492524,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.432242863016273e-05,
+      "loss": 1.2071,
+      "step": 5162
+    },
+    {
+      "epoch": 0.37989422855828925,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.431705451748703e-05,
+      "loss": 1.0243,
+      "step": 5163
+    },
+    {
+      "epoch": 0.37996780869165325,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.4311679904782625e-05,
+      "loss": 1.092,
+      "step": 5164
+    },
+    {
+      "epoch": 0.38004138882501726,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.4306304792337934e-05,
+      "loss": 0.7073,
+      "step": 5165
+    },
+    {
+      "epoch": 0.38011496895838126,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.430092918044145e-05,
+      "loss": 0.805,
+      "step": 5166
+    },
+    {
+      "epoch": 0.3801885490917452,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.4295553069381664e-05,
+      "loss": 0.7357,
+      "step": 5167
+    },
+    {
+      "epoch": 0.3802621292251092,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.4290176459447096e-05,
+      "loss": 0.7566,
+      "step": 5168
+    },
+    {
+      "epoch": 0.3803357093584732,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.428479935092631e-05,
+      "loss": 0.8424,
+      "step": 5169
+    },
+    {
+      "epoch": 0.3804092894918372,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.427942174410787e-05,
+      "loss": 0.8054,
+      "step": 5170
+    },
+    {
+      "epoch": 0.38048286962520117,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.42740436392804e-05,
+      "loss": 1.303,
+      "step": 5171
+    },
+    {
+      "epoch": 0.3805564497585652,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.426866503673252e-05,
+      "loss": 0.9888,
+      "step": 5172
+    },
+    {
+      "epoch": 0.3806300298919292,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.42632859367529e-05,
+      "loss": 1.2661,
+      "step": 5173
+    },
+    {
+      "epoch": 0.3807036100252932,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.425790633963021e-05,
+      "loss": 0.8517,
+      "step": 5174
+    },
+    {
+      "epoch": 0.3807771901586572,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.4252526245653184e-05,
+      "loss": 0.8681,
+      "step": 5175
+    },
+    {
+      "epoch": 0.38085077029202113,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.4247145655110544e-05,
+      "loss": 1.0429,
+      "step": 5176
+    },
+    {
+      "epoch": 0.38092435042538514,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.4241764568291057e-05,
+      "loss": 1.257,
+      "step": 5177
+    },
+    {
+      "epoch": 0.38099793055874914,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.423638298548352e-05,
+      "loss": 0.7095,
+      "step": 5178
+    },
+    {
+      "epoch": 0.38107151069211315,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.423100090697676e-05,
+      "loss": 1.2795,
+      "step": 5179
+    },
+    {
+      "epoch": 0.3811450908254771,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.422561833305962e-05,
+      "loss": 1.0184,
+      "step": 5180
+    },
+    {
+      "epoch": 0.3812186709588411,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.422023526402096e-05,
+      "loss": 1.1854,
+      "step": 5181
+    },
+    {
+      "epoch": 0.3812922510922051,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.421485170014969e-05,
+      "loss": 1.0371,
+      "step": 5182
+    },
+    {
+      "epoch": 0.3813658312255691,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.420946764173474e-05,
+      "loss": 0.8685,
+      "step": 5183
+    },
+    {
+      "epoch": 0.3814394113589331,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.4204083089065054e-05,
+      "loss": 1.1983,
+      "step": 5184
+    },
+    {
+      "epoch": 0.38151299149229706,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.4198698042429604e-05,
+      "loss": 1.0159,
+      "step": 5185
+    },
+    {
+      "epoch": 0.38158657162566106,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.419331250211741e-05,
+      "loss": 0.9349,
+      "step": 5186
+    },
+    {
+      "epoch": 0.38166015175902507,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.418792646841749e-05,
+      "loss": 0.7611,
+      "step": 5187
+    },
+    {
+      "epoch": 0.38173373189238907,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.418253994161892e-05,
+      "loss": 1.2321,
+      "step": 5188
+    },
+    {
+      "epoch": 0.381807312025753,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.4177152922010776e-05,
+      "loss": 0.8143,
+      "step": 5189
+    },
+    {
+      "epoch": 0.381880892159117,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.4171765409882165e-05,
+      "loss": 0.8956,
+      "step": 5190
+    },
+    {
+      "epoch": 0.38195447229248103,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.416637740552222e-05,
+      "loss": 0.9225,
+      "step": 5191
+    },
+    {
+      "epoch": 0.38202805242584503,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.416098890922012e-05,
+      "loss": 1.4415,
+      "step": 5192
+    },
+    {
+      "epoch": 0.38210163255920904,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.4155599921265044e-05,
+      "loss": 0.7804,
+      "step": 5193
+    },
+    {
+      "epoch": 0.382175212692573,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.415021044194622e-05,
+      "loss": 0.8919,
+      "step": 5194
+    },
+    {
+      "epoch": 0.382248792825937,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.414482047155288e-05,
+      "loss": 0.8574,
+      "step": 5195
+    },
+    {
+      "epoch": 0.382322372959301,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.413943001037429e-05,
+      "loss": 0.9972,
+      "step": 5196
+    },
+    {
+      "epoch": 0.382395953092665,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.4134039058699765e-05,
+      "loss": 1.4238,
+      "step": 5197
+    },
+    {
+      "epoch": 0.38246953322602895,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.412864761681861e-05,
+      "loss": 1.0294,
+      "step": 5198
+    },
+    {
+      "epoch": 0.38254311335939295,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.412325568502018e-05,
+      "loss": 0.786,
+      "step": 5199
+    },
+    {
+      "epoch": 0.38261669349275695,
+      "grad_norm": 1.1328125,
+      "learning_rate": 3.411786326359384e-05,
+      "loss": 1.2168,
+      "step": 5200
+    },
+    {
+      "epoch": 0.38269027362612096,
+      "grad_norm": 0.75,
+      "learning_rate": 3.411247035282902e-05,
+      "loss": 0.588,
+      "step": 5201
+    },
+    {
+      "epoch": 0.38276385375948496,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.410707695301511e-05,
+      "loss": 0.7805,
+      "step": 5202
+    },
+    {
+      "epoch": 0.3828374338928489,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.410168306444158e-05,
+      "loss": 0.7293,
+      "step": 5203
+    },
+    {
+      "epoch": 0.3829110140262129,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.409628868739793e-05,
+      "loss": 0.9363,
+      "step": 5204
+    },
+    {
+      "epoch": 0.3829845941595769,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.409089382217362e-05,
+      "loss": 0.8522,
+      "step": 5205
+    },
+    {
+      "epoch": 0.3830581742929409,
+      "grad_norm": 1.171875,
+      "learning_rate": 3.408549846905822e-05,
+      "loss": 0.8536,
+      "step": 5206
+    },
+    {
+      "epoch": 0.38313175442630487,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.408010262834128e-05,
+      "loss": 0.9996,
+      "step": 5207
+    },
+    {
+      "epoch": 0.3832053345596689,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.4074706300312385e-05,
+      "loss": 0.6353,
+      "step": 5208
+    },
+    {
+      "epoch": 0.3832789146930329,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.406930948526114e-05,
+      "loss": 0.91,
+      "step": 5209
+    },
+    {
+      "epoch": 0.3833524948263969,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.4063912183477186e-05,
+      "loss": 1.0982,
+      "step": 5210
+    },
+    {
+      "epoch": 0.3834260749597609,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.405851439525018e-05,
+      "loss": 1.0828,
+      "step": 5211
+    },
+    {
+      "epoch": 0.38349965509312484,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.405311612086981e-05,
+      "loss": 1.0176,
+      "step": 5212
+    },
+    {
+      "epoch": 0.38357323522648884,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.4047717360625804e-05,
+      "loss": 0.9045,
+      "step": 5213
+    },
+    {
+      "epoch": 0.38364681535985284,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.404231811480789e-05,
+      "loss": 0.507,
+      "step": 5214
+    },
+    {
+      "epoch": 0.38372039549321685,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.403691838370585e-05,
+      "loss": 0.8231,
+      "step": 5215
+    },
+    {
+      "epoch": 0.3837939756265808,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.403151816760947e-05,
+      "loss": 1.22,
+      "step": 5216
+    },
+    {
+      "epoch": 0.3838675557599448,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.402611746680857e-05,
+      "loss": 0.5894,
+      "step": 5217
+    },
+    {
+      "epoch": 0.3839411358933088,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.4020716281592974e-05,
+      "loss": 0.8356,
+      "step": 5218
+    },
+    {
+      "epoch": 0.3840147160266728,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.401531461225258e-05,
+      "loss": 0.8682,
+      "step": 5219
+    },
+    {
+      "epoch": 0.3840882961600368,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.400991245907729e-05,
+      "loss": 1.2747,
+      "step": 5220
+    },
+    {
+      "epoch": 0.38416187629340076,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.400450982235701e-05,
+      "loss": 1.047,
+      "step": 5221
+    },
+    {
+      "epoch": 0.38423545642676477,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.39991067023817e-05,
+      "loss": 1.0059,
+      "step": 5222
+    },
+    {
+      "epoch": 0.38430903656012877,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.3993703099441323e-05,
+      "loss": 0.7447,
+      "step": 5223
+    },
+    {
+      "epoch": 0.3843826166934928,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.3988299013825894e-05,
+      "loss": 1.1768,
+      "step": 5224
+    },
+    {
+      "epoch": 0.3844561968268567,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.398289444582542e-05,
+      "loss": 0.8489,
+      "step": 5225
+    },
+    {
+      "epoch": 0.3845297769602207,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.397748939572998e-05,
+      "loss": 0.9217,
+      "step": 5226
+    },
+    {
+      "epoch": 0.38460335709358473,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.397208386382963e-05,
+      "loss": 1.3784,
+      "step": 5227
+    },
+    {
+      "epoch": 0.38467693722694873,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.396667785041449e-05,
+      "loss": 0.9415,
+      "step": 5228
+    },
+    {
+      "epoch": 0.38475051736031274,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.396127135577469e-05,
+      "loss": 0.7007,
+      "step": 5229
+    },
+    {
+      "epoch": 0.3848240974936767,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.3955864380200374e-05,
+      "loss": 1.1998,
+      "step": 5230
+    },
+    {
+      "epoch": 0.3848976776270407,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.3950456923981736e-05,
+      "loss": 1.1884,
+      "step": 5231
+    },
+    {
+      "epoch": 0.3849712577604047,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.394504898740898e-05,
+      "loss": 0.924,
+      "step": 5232
+    },
+    {
+      "epoch": 0.3850448378937687,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.3939640570772344e-05,
+      "loss": 1.4184,
+      "step": 5233
+    },
+    {
+      "epoch": 0.38511841802713265,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.393423167436208e-05,
+      "loss": 0.9991,
+      "step": 5234
+    },
+    {
+      "epoch": 0.38519199816049665,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.392882229846847e-05,
+      "loss": 0.7082,
+      "step": 5235
+    },
+    {
+      "epoch": 0.38526557829386066,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.392341244338184e-05,
+      "loss": 0.6199,
+      "step": 5236
+    },
+    {
+      "epoch": 0.38533915842722466,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.391800210939251e-05,
+      "loss": 1.1754,
+      "step": 5237
+    },
+    {
+      "epoch": 0.38541273856058866,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.391259129679086e-05,
+      "loss": 1.2509,
+      "step": 5238
+    },
+    {
+      "epoch": 0.3854863186939526,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.390718000586725e-05,
+      "loss": 0.9425,
+      "step": 5239
+    },
+    {
+      "epoch": 0.3855598988273166,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.390176823691214e-05,
+      "loss": 0.893,
+      "step": 5240
+    },
+    {
+      "epoch": 0.3856334789606806,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.3896355990215925e-05,
+      "loss": 0.9134,
+      "step": 5241
+    },
+    {
+      "epoch": 0.3857070590940446,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.389094326606909e-05,
+      "loss": 0.7703,
+      "step": 5242
+    },
+    {
+      "epoch": 0.3857806392274086,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.388553006476212e-05,
+      "loss": 1.2572,
+      "step": 5243
+    },
+    {
+      "epoch": 0.3858542193607726,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.388011638658554e-05,
+      "loss": 0.9101,
+      "step": 5244
+    },
+    {
+      "epoch": 0.3859277994941366,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.3874702231829883e-05,
+      "loss": 0.7354,
+      "step": 5245
+    },
+    {
+      "epoch": 0.3860013796275006,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.386928760078571e-05,
+      "loss": 1.211,
+      "step": 5246
+    },
+    {
+      "epoch": 0.3860749597608646,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.386387249374364e-05,
+      "loss": 1.1345,
+      "step": 5247
+    },
+    {
+      "epoch": 0.38614853989422854,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.385845691099426e-05,
+      "loss": 0.6798,
+      "step": 5248
+    },
+    {
+      "epoch": 0.38622212002759254,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.385304085282824e-05,
+      "loss": 0.9198,
+      "step": 5249
+    },
+    {
+      "epoch": 0.38629570016095655,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.384762431953623e-05,
+      "loss": 0.7693,
+      "step": 5250
+    },
+    {
+      "epoch": 0.38636928029432055,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.384220731140894e-05,
+      "loss": 0.7477,
+      "step": 5251
+    },
+    {
+      "epoch": 0.3864428604276845,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.3836789828737074e-05,
+      "loss": 0.8239,
+      "step": 5252
+    },
+    {
+      "epoch": 0.3865164405610485,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.383137187181139e-05,
+      "loss": 0.8524,
+      "step": 5253
+    },
+    {
+      "epoch": 0.3865900206944125,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.382595344092267e-05,
+      "loss": 1.1263,
+      "step": 5254
+    },
+    {
+      "epoch": 0.3866636008277765,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.382053453636169e-05,
+      "loss": 1.233,
+      "step": 5255
+    },
+    {
+      "epoch": 0.3867371809611405,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.3815115158419287e-05,
+      "loss": 1.3229,
+      "step": 5256
+    },
+    {
+      "epoch": 0.38681076109450446,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.3809695307386294e-05,
+      "loss": 0.9049,
+      "step": 5257
+    },
+    {
+      "epoch": 0.38688434122786847,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.38042749835536e-05,
+      "loss": 0.7716,
+      "step": 5258
+    },
+    {
+      "epoch": 0.38695792136123247,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.379885418721209e-05,
+      "loss": 1.0446,
+      "step": 5259
+    },
+    {
+      "epoch": 0.3870315014945965,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.3793432918652695e-05,
+      "loss": 1.0852,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3871050816279604,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.378801117816637e-05,
+      "loss": 0.8312,
+      "step": 5261
+    },
+    {
+      "epoch": 0.3871786617613244,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.378258896604408e-05,
+      "loss": 0.795,
+      "step": 5262
+    },
+    {
+      "epoch": 0.38725224189468843,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.377716628257683e-05,
+      "loss": 1.0622,
+      "step": 5263
+    },
+    {
+      "epoch": 0.38732582202805244,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.3771743128055645e-05,
+      "loss": 0.953,
+      "step": 5264
+    },
+    {
+      "epoch": 0.38739940216141644,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.3766319502771565e-05,
+      "loss": 0.901,
+      "step": 5265
+    },
+    {
+      "epoch": 0.3874729822947804,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.376089540701568e-05,
+      "loss": 0.9593,
+      "step": 5266
+    },
+    {
+      "epoch": 0.3875465624281444,
+      "grad_norm": 0.625,
+      "learning_rate": 3.375547084107908e-05,
+      "loss": 0.7164,
+      "step": 5267
+    },
+    {
+      "epoch": 0.3876201425615084,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.375004580525291e-05,
+      "loss": 1.3671,
+      "step": 5268
+    },
+    {
+      "epoch": 0.3876937226948724,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.37446202998283e-05,
+      "loss": 1.0811,
+      "step": 5269
+    },
+    {
+      "epoch": 0.38776730282823635,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.3739194325096436e-05,
+      "loss": 1.0696,
+      "step": 5270
+    },
+    {
+      "epoch": 0.38784088296160035,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.373376788134852e-05,
+      "loss": 1.0669,
+      "step": 5271
+    },
+    {
+      "epoch": 0.38791446309496436,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.3728340968875773e-05,
+      "loss": 0.7129,
+      "step": 5272
+    },
+    {
+      "epoch": 0.38798804322832836,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.372291358796945e-05,
+      "loss": 0.8409,
+      "step": 5273
+    },
+    {
+      "epoch": 0.38806162336169236,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.371748573892084e-05,
+      "loss": 1.2108,
+      "step": 5274
+    },
+    {
+      "epoch": 0.3881352034950563,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.3712057422021224e-05,
+      "loss": 0.7328,
+      "step": 5275
+    },
+    {
+      "epoch": 0.3882087836284203,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.370662863756194e-05,
+      "loss": 0.97,
+      "step": 5276
+    },
+    {
+      "epoch": 0.3882823637617843,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.370119938583436e-05,
+      "loss": 0.8496,
+      "step": 5277
+    },
+    {
+      "epoch": 0.3883559438951483,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.369576966712982e-05,
+      "loss": 0.8405,
+      "step": 5278
+    },
+    {
+      "epoch": 0.3884295240285123,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.369033948173976e-05,
+      "loss": 0.9884,
+      "step": 5279
+    },
+    {
+      "epoch": 0.3885031041618763,
+      "grad_norm": 0.875,
+      "learning_rate": 3.3684908829955586e-05,
+      "loss": 0.7494,
+      "step": 5280
+    },
+    {
+      "epoch": 0.3885766842952403,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.367947771206877e-05,
+      "loss": 0.7816,
+      "step": 5281
+    },
+    {
+      "epoch": 0.3886502644286043,
+      "grad_norm": 0.875,
+      "learning_rate": 3.3674046128370766e-05,
+      "loss": 0.6771,
+      "step": 5282
+    },
+    {
+      "epoch": 0.3887238445619683,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.366861407915309e-05,
+      "loss": 0.9264,
+      "step": 5283
+    },
+    {
+      "epoch": 0.38879742469533224,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.3663181564707286e-05,
+      "loss": 0.8988,
+      "step": 5284
+    },
+    {
+      "epoch": 0.38887100482869624,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.365774858532487e-05,
+      "loss": 1.3125,
+      "step": 5285
+    },
+    {
+      "epoch": 0.38894458496206025,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.365231514129745e-05,
+      "loss": 1.2386,
+      "step": 5286
+    },
+    {
+      "epoch": 0.38901816509542425,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.364688123291662e-05,
+      "loss": 0.934,
+      "step": 5287
+    },
+    {
+      "epoch": 0.3890917452287882,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.3641446860474e-05,
+      "loss": 0.8637,
+      "step": 5288
+    },
+    {
+      "epoch": 0.3891653253621522,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.363601202426124e-05,
+      "loss": 0.9003,
+      "step": 5289
+    },
+    {
+      "epoch": 0.3892389054955162,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.3630576724570046e-05,
+      "loss": 0.9079,
+      "step": 5290
+    },
+    {
+      "epoch": 0.3893124856288802,
+      "grad_norm": 0.59765625,
+      "learning_rate": 3.3625140961692084e-05,
+      "loss": 0.6688,
+      "step": 5291
+    },
+    {
+      "epoch": 0.3893860657622442,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.361970473591911e-05,
+      "loss": 0.7819,
+      "step": 5292
+    },
+    {
+      "epoch": 0.38945964589560816,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.361426804754285e-05,
+      "loss": 0.6088,
+      "step": 5293
+    },
+    {
+      "epoch": 0.38953322602897217,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.36088308968551e-05,
+      "loss": 1.0231,
+      "step": 5294
+    },
+    {
+      "epoch": 0.38960680616233617,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.3603393284147656e-05,
+      "loss": 1.3235,
+      "step": 5295
+    },
+    {
+      "epoch": 0.3896803862957002,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.3597955209712337e-05,
+      "loss": 0.8804,
+      "step": 5296
+    },
+    {
+      "epoch": 0.3897539664290642,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.359251667384101e-05,
+      "loss": 1.1223,
+      "step": 5297
+    },
+    {
+      "epoch": 0.38982754656242813,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.358707767682554e-05,
+      "loss": 0.8199,
+      "step": 5298
+    },
+    {
+      "epoch": 0.38990112669579213,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.358163821895783e-05,
+      "loss": 0.661,
+      "step": 5299
+    },
+    {
+      "epoch": 0.38997470682915614,
+      "grad_norm": 1.296875,
+      "learning_rate": 3.3576198300529804e-05,
+      "loss": 1.1993,
+      "step": 5300
+    },
+    {
+      "epoch": 0.39004828696252014,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.357075792183341e-05,
+      "loss": 1.0676,
+      "step": 5301
+    },
+    {
+      "epoch": 0.3901218670958841,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.356531708316063e-05,
+      "loss": 1.09,
+      "step": 5302
+    },
+    {
+      "epoch": 0.3901954472292481,
+      "grad_norm": 0.7109375,
+      "learning_rate": 3.3559875784803465e-05,
+      "loss": 0.92,
+      "step": 5303
+    },
+    {
+      "epoch": 0.3902690273626121,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.3554434027053926e-05,
+      "loss": 1.0002,
+      "step": 5304
+    },
+    {
+      "epoch": 0.3903426074959761,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.354899181020407e-05,
+      "loss": 1.2513,
+      "step": 5305
+    },
+    {
+      "epoch": 0.3904161876293401,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.3543549134545975e-05,
+      "loss": 0.6741,
+      "step": 5306
+    },
+    {
+      "epoch": 0.39048976776270405,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.353810600037173e-05,
+      "loss": 1.0504,
+      "step": 5307
+    },
+    {
+      "epoch": 0.39056334789606806,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.353266240797346e-05,
+      "loss": 0.5393,
+      "step": 5308
+    },
+    {
+      "epoch": 0.39063692802943206,
+      "grad_norm": 0.75,
+      "learning_rate": 3.352721835764333e-05,
+      "loss": 1.1147,
+      "step": 5309
+    },
+    {
+      "epoch": 0.39071050816279607,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.3521773849673475e-05,
+      "loss": 0.9932,
+      "step": 5310
+    },
+    {
+      "epoch": 0.39078408829616,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.351632888435613e-05,
+      "loss": 0.7278,
+      "step": 5311
+    },
+    {
+      "epoch": 0.390857668429524,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.351088346198349e-05,
+      "loss": 0.7127,
+      "step": 5312
+    },
+    {
+      "epoch": 0.390931248562888,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.350543758284781e-05,
+      "loss": 0.9187,
+      "step": 5313
+    },
+    {
+      "epoch": 0.391004828696252,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.349999124724136e-05,
+      "loss": 0.9182,
+      "step": 5314
+    },
+    {
+      "epoch": 0.39107840882961603,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.349454445545644e-05,
+      "loss": 0.7731,
+      "step": 5315
+    },
+    {
+      "epoch": 0.39115198896298,
+      "grad_norm": 1.640625,
+      "learning_rate": 3.348909720778535e-05,
+      "loss": 0.8143,
+      "step": 5316
+    },
+    {
+      "epoch": 0.391225569096344,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.348364950452046e-05,
+      "loss": 1.1468,
+      "step": 5317
+    },
+    {
+      "epoch": 0.391299149229708,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.347820134595412e-05,
+      "loss": 0.9937,
+      "step": 5318
+    },
+    {
+      "epoch": 0.391372729363072,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.347275273237872e-05,
+      "loss": 1.2644,
+      "step": 5319
+    },
+    {
+      "epoch": 0.39144630949643594,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.346730366408669e-05,
+      "loss": 0.9775,
+      "step": 5320
+    },
+    {
+      "epoch": 0.39151988962979994,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.346185414137046e-05,
+      "loss": 0.8551,
+      "step": 5321
+    },
+    {
+      "epoch": 0.39159346976316395,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.345640416452251e-05,
+      "loss": 1.1624,
+      "step": 5322
+    },
+    {
+      "epoch": 0.39166704989652795,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.345095373383531e-05,
+      "loss": 0.8431,
+      "step": 5323
+    },
+    {
+      "epoch": 0.39174063002989196,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.344550284960139e-05,
+      "loss": 0.5652,
+      "step": 5324
+    },
+    {
+      "epoch": 0.3918142101632559,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.3440051512113285e-05,
+      "loss": 0.7254,
+      "step": 5325
+    },
+    {
+      "epoch": 0.3918877902966199,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.343459972166355e-05,
+      "loss": 0.899,
+      "step": 5326
+    },
+    {
+      "epoch": 0.3919613704299839,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.342914747854478e-05,
+      "loss": 1.1617,
+      "step": 5327
+    },
+    {
+      "epoch": 0.3920349505633479,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.342369478304958e-05,
+      "loss": 1.0367,
+      "step": 5328
+    },
+    {
+      "epoch": 0.39210853069671187,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.341824163547061e-05,
+      "loss": 1.0766,
+      "step": 5329
+    },
+    {
+      "epoch": 0.39218211083007587,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.3412788036100504e-05,
+      "loss": 0.8493,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3922556909634399,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.340733398523195e-05,
+      "loss": 1.1773,
+      "step": 5331
+    },
+    {
+      "epoch": 0.3923292710968039,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.3401879483157655e-05,
+      "loss": 0.8706,
+      "step": 5332
+    },
+    {
+      "epoch": 0.3924028512301679,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.339642453017036e-05,
+      "loss": 1.0878,
+      "step": 5333
+    },
+    {
+      "epoch": 0.39247643136353183,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.3390969126562824e-05,
+      "loss": 0.688,
+      "step": 5334
+    },
+    {
+      "epoch": 0.39255001149689583,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.3385513272627816e-05,
+      "loss": 0.8408,
+      "step": 5335
+    },
+    {
+      "epoch": 0.39262359163025984,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.338005696865816e-05,
+      "loss": 1.3848,
+      "step": 5336
+    },
+    {
+      "epoch": 0.39269717176362384,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.337460021494666e-05,
+      "loss": 0.8371,
+      "step": 5337
+    },
+    {
+      "epoch": 0.3927707518969878,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.3369143011786195e-05,
+      "loss": 0.8894,
+      "step": 5338
+    },
+    {
+      "epoch": 0.3928443320303518,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.336368535946963e-05,
+      "loss": 0.9933,
+      "step": 5339
+    },
+    {
+      "epoch": 0.3929179121637158,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.3358227258289865e-05,
+      "loss": 0.8759,
+      "step": 5340
+    },
+    {
+      "epoch": 0.3929914922970798,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.335276870853983e-05,
+      "loss": 1.1022,
+      "step": 5341
+    },
+    {
+      "epoch": 0.3930650724304438,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.334730971051247e-05,
+      "loss": 0.9794,
+      "step": 5342
+    },
+    {
+      "epoch": 0.39313865256380776,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.334185026450077e-05,
+      "loss": 0.8649,
+      "step": 5343
+    },
+    {
+      "epoch": 0.39321223269717176,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.333639037079772e-05,
+      "loss": 0.7994,
+      "step": 5344
+    },
+    {
+      "epoch": 0.39328581283053576,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.333093002969635e-05,
+      "loss": 0.9591,
+      "step": 5345
+    },
+    {
+      "epoch": 0.39335939296389977,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.3325469241489696e-05,
+      "loss": 0.8415,
+      "step": 5346
+    },
+    {
+      "epoch": 0.3934329730972637,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.3320008006470825e-05,
+      "loss": 1.1434,
+      "step": 5347
+    },
+    {
+      "epoch": 0.3935065532306277,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.331454632493284e-05,
+      "loss": 1.3751,
+      "step": 5348
+    },
+    {
+      "epoch": 0.3935801333639917,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.330908419716886e-05,
+      "loss": 0.9702,
+      "step": 5349
+    },
+    {
+      "epoch": 0.39365371349735573,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.330362162347204e-05,
+      "loss": 1.0524,
+      "step": 5350
+    },
+    {
+      "epoch": 0.39372729363071973,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.329815860413551e-05,
+      "loss": 0.8381,
+      "step": 5351
+    },
+    {
+      "epoch": 0.3938008737640837,
+      "grad_norm": 0.875,
+      "learning_rate": 3.329269513945249e-05,
+      "loss": 0.9965,
+      "step": 5352
+    },
+    {
+      "epoch": 0.3938744538974477,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.328723122971619e-05,
+      "loss": 1.1901,
+      "step": 5353
+    },
+    {
+      "epoch": 0.3939480340308117,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.328176687521983e-05,
+      "loss": 0.9843,
+      "step": 5354
+    },
+    {
+      "epoch": 0.3940216141641757,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.327630207625668e-05,
+      "loss": 0.827,
+      "step": 5355
+    },
+    {
+      "epoch": 0.39409519429753964,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.327083683312004e-05,
+      "loss": 0.8483,
+      "step": 5356
+    },
+    {
+      "epoch": 0.39416877443090365,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.326537114610321e-05,
+      "loss": 1.094,
+      "step": 5357
+    },
+    {
+      "epoch": 0.39424235456426765,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.325990501549952e-05,
+      "loss": 0.8335,
+      "step": 5358
+    },
+    {
+      "epoch": 0.39431593469763165,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.325443844160233e-05,
+      "loss": 0.8375,
+      "step": 5359
+    },
+    {
+      "epoch": 0.39438951483099566,
+      "grad_norm": 0.875,
+      "learning_rate": 3.324897142470502e-05,
+      "loss": 0.8763,
+      "step": 5360
+    },
+    {
+      "epoch": 0.3944630949643596,
+      "grad_norm": 1.390625,
+      "learning_rate": 3.324350396510099e-05,
+      "loss": 1.296,
+      "step": 5361
+    },
+    {
+      "epoch": 0.3945366750977236,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.323803606308367e-05,
+      "loss": 1.3584,
+      "step": 5362
+    },
+    {
+      "epoch": 0.3946102552310876,
+      "grad_norm": 0.75,
+      "learning_rate": 3.3232567718946516e-05,
+      "loss": 0.7868,
+      "step": 5363
+    },
+    {
+      "epoch": 0.3946838353644516,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.3227098932983e-05,
+      "loss": 0.9081,
+      "step": 5364
+    },
+    {
+      "epoch": 0.39475741549781557,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.3221629705486627e-05,
+      "loss": 0.8263,
+      "step": 5365
+    },
+    {
+      "epoch": 0.39483099563117957,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.321616003675091e-05,
+      "loss": 1.1382,
+      "step": 5366
+    },
+    {
+      "epoch": 0.3949045757645436,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.32106899270694e-05,
+      "loss": 1.0956,
+      "step": 5367
+    },
+    {
+      "epoch": 0.3949781558979076,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.320521937673568e-05,
+      "loss": 0.9621,
+      "step": 5368
+    },
+    {
+      "epoch": 0.3950517360312716,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.319974838604333e-05,
+      "loss": 0.6868,
+      "step": 5369
+    },
+    {
+      "epoch": 0.39512531616463553,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.3194276955285976e-05,
+      "loss": 1.1196,
+      "step": 5370
+    },
+    {
+      "epoch": 0.39519889629799954,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.3188805084757244e-05,
+      "loss": 0.6754,
+      "step": 5371
+    },
+    {
+      "epoch": 0.39527247643136354,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.318333277475081e-05,
+      "loss": 0.7185,
+      "step": 5372
+    },
+    {
+      "epoch": 0.39534605656472754,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.317786002556037e-05,
+      "loss": 0.7248,
+      "step": 5373
+    },
+    {
+      "epoch": 0.3954196366980915,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.3172386837479625e-05,
+      "loss": 1.2999,
+      "step": 5374
+    },
+    {
+      "epoch": 0.3954932168314555,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.316691321080231e-05,
+      "loss": 0.7879,
+      "step": 5375
+    },
+    {
+      "epoch": 0.3955667969648195,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.3161439145822194e-05,
+      "loss": 0.7453,
+      "step": 5376
+    },
+    {
+      "epoch": 0.3956403770981835,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.315596464283306e-05,
+      "loss": 0.8616,
+      "step": 5377
+    },
+    {
+      "epoch": 0.3957139572315475,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.315048970212869e-05,
+      "loss": 0.9999,
+      "step": 5378
+    },
+    {
+      "epoch": 0.39578753736491146,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.3145014324002944e-05,
+      "loss": 1.459,
+      "step": 5379
+    },
+    {
+      "epoch": 0.39586111749827546,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.313953850874966e-05,
+      "loss": 0.9809,
+      "step": 5380
+    },
+    {
+      "epoch": 0.39593469763163947,
+      "grad_norm": 1.1171875,
+      "learning_rate": 3.3134062256662714e-05,
+      "loss": 0.9415,
+      "step": 5381
+    },
+    {
+      "epoch": 0.39600827776500347,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.3128585568036014e-05,
+      "loss": 1.1217,
+      "step": 5382
+    },
+    {
+      "epoch": 0.3960818578983674,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.3123108443163474e-05,
+      "loss": 0.6739,
+      "step": 5383
+    },
+    {
+      "epoch": 0.3961554380317314,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.3117630882339054e-05,
+      "loss": 1.3538,
+      "step": 5384
+    },
+    {
+      "epoch": 0.3962290181650954,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.3112152885856716e-05,
+      "loss": 0.6795,
+      "step": 5385
+    },
+    {
+      "epoch": 0.39630259829845943,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.3106674454010454e-05,
+      "loss": 1.1909,
+      "step": 5386
+    },
+    {
+      "epoch": 0.39637617843182343,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.310119558709428e-05,
+      "loss": 0.9258,
+      "step": 5387
+    },
+    {
+      "epoch": 0.3964497585651874,
+      "grad_norm": 1.3515625,
+      "learning_rate": 3.309571628540224e-05,
+      "loss": 1.1506,
+      "step": 5388
+    },
+    {
+      "epoch": 0.3965233386985514,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.30902365492284e-05,
+      "loss": 1.0797,
+      "step": 5389
+    },
+    {
+      "epoch": 0.3965969188319154,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.3084756378866845e-05,
+      "loss": 0.7262,
+      "step": 5390
+    },
+    {
+      "epoch": 0.3966704989652794,
+      "grad_norm": 0.65234375,
+      "learning_rate": 3.307927577461169e-05,
+      "loss": 0.5526,
+      "step": 5391
+    },
+    {
+      "epoch": 0.39674407909864334,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.3073794736757056e-05,
+      "loss": 0.8195,
+      "step": 5392
+    },
+    {
+      "epoch": 0.39681765923200735,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.306831326559712e-05,
+      "loss": 0.8256,
+      "step": 5393
+    },
+    {
+      "epoch": 0.39689123936537135,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.306283136142603e-05,
+      "loss": 1.4114,
+      "step": 5394
+    },
+    {
+      "epoch": 0.39696481949873536,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.305734902453802e-05,
+      "loss": 0.7751,
+      "step": 5395
+    },
+    {
+      "epoch": 0.39703839963209936,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.305186625522731e-05,
+      "loss": 0.7799,
+      "step": 5396
+    },
+    {
+      "epoch": 0.3971119797654633,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.304638305378814e-05,
+      "loss": 0.937,
+      "step": 5397
+    },
+    {
+      "epoch": 0.3971855598988273,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.3040899420514793e-05,
+      "loss": 0.6851,
+      "step": 5398
+    },
+    {
+      "epoch": 0.3972591400321913,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.303541535570156e-05,
+      "loss": 1.0003,
+      "step": 5399
+    },
+    {
+      "epoch": 0.3973327201655553,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.302993085964276e-05,
+      "loss": 1.0351,
+      "step": 5400
+    },
+    {
+      "epoch": 0.39740630029891927,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.302444593263273e-05,
+      "loss": 0.8263,
+      "step": 5401
+    },
+    {
+      "epoch": 0.3974798804322833,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.3018960574965864e-05,
+      "loss": 0.7955,
+      "step": 5402
+    },
+    {
+      "epoch": 0.3975534605656473,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.301347478693651e-05,
+      "loss": 0.8923,
+      "step": 5403
+    },
+    {
+      "epoch": 0.3976270406990113,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.3007988568839104e-05,
+      "loss": 0.7069,
+      "step": 5404
+    },
+    {
+      "epoch": 0.3977006208323753,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.300250192096808e-05,
+      "loss": 0.979,
+      "step": 5405
+    },
+    {
+      "epoch": 0.39777420096573923,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.299701484361789e-05,
+      "loss": 0.9844,
+      "step": 5406
+    },
+    {
+      "epoch": 0.39784778109910324,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.299152733708303e-05,
+      "loss": 0.7956,
+      "step": 5407
+    },
+    {
+      "epoch": 0.39792136123246724,
+      "grad_norm": 1.140625,
+      "learning_rate": 3.298603940165797e-05,
+      "loss": 1.2359,
+      "step": 5408
+    },
+    {
+      "epoch": 0.39799494136583125,
+      "grad_norm": 1.140625,
+      "learning_rate": 3.298055103763727e-05,
+      "loss": 1.1349,
+      "step": 5409
+    },
+    {
+      "epoch": 0.3980685214991952,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.297506224531547e-05,
+      "loss": 0.852,
+      "step": 5410
+    },
+    {
+      "epoch": 0.3981421016325592,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.2969573024987136e-05,
+      "loss": 0.706,
+      "step": 5411
+    },
+    {
+      "epoch": 0.3982156817659232,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.296408337694687e-05,
+      "loss": 0.7747,
+      "step": 5412
+    },
+    {
+      "epoch": 0.3982892618992872,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.2958593301489296e-05,
+      "loss": 0.7966,
+      "step": 5413
+    },
+    {
+      "epoch": 0.3983628420326512,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.2953102798909055e-05,
+      "loss": 1.4578,
+      "step": 5414
+    },
+    {
+      "epoch": 0.39843642216601516,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.29476118695008e-05,
+      "loss": 0.6992,
+      "step": 5415
+    },
+    {
+      "epoch": 0.39851000229937916,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.294212051355923e-05,
+      "loss": 1.0596,
+      "step": 5416
+    },
+    {
+      "epoch": 0.39858358243274317,
+      "grad_norm": 1.1484375,
+      "learning_rate": 3.293662873137906e-05,
+      "loss": 1.3676,
+      "step": 5417
+    },
+    {
+      "epoch": 0.39865716256610717,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.293113652325501e-05,
+      "loss": 0.8509,
+      "step": 5418
+    },
+    {
+      "epoch": 0.3987307426994711,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.292564388948184e-05,
+      "loss": 0.7004,
+      "step": 5419
+    },
+    {
+      "epoch": 0.3988043228328351,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.292015083035433e-05,
+      "loss": 0.8331,
+      "step": 5420
+    },
+    {
+      "epoch": 0.3988779029661991,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.291465734616729e-05,
+      "loss": 0.736,
+      "step": 5421
+    },
+    {
+      "epoch": 0.39895148309956313,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.290916343721553e-05,
+      "loss": 1.1495,
+      "step": 5422
+    },
+    {
+      "epoch": 0.39902506323292714,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.2903669103793916e-05,
+      "loss": 0.5683,
+      "step": 5423
+    },
+    {
+      "epoch": 0.3990986433662911,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.2898174346197306e-05,
+      "loss": 1.2228,
+      "step": 5424
+    },
+    {
+      "epoch": 0.3991722234996551,
+      "grad_norm": 1.3359375,
+      "learning_rate": 3.289267916472059e-05,
+      "loss": 1.1784,
+      "step": 5425
+    },
+    {
+      "epoch": 0.3992458036330191,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.288718355965871e-05,
+      "loss": 0.9894,
+      "step": 5426
+    },
+    {
+      "epoch": 0.3993193837663831,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.288168753130657e-05,
+      "loss": 0.9713,
+      "step": 5427
+    },
+    {
+      "epoch": 0.39939296389974704,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.287619107995914e-05,
+      "loss": 0.9314,
+      "step": 5428
+    },
+    {
+      "epoch": 0.39946654403311105,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.287069420591142e-05,
+      "loss": 1.343,
+      "step": 5429
+    },
+    {
+      "epoch": 0.39954012416647505,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.286519690945841e-05,
+      "loss": 0.7085,
+      "step": 5430
+    },
+    {
+      "epoch": 0.39961370429983906,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.2859699190895135e-05,
+      "loss": 0.9537,
+      "step": 5431
+    },
+    {
+      "epoch": 0.39968728443320306,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.285420105051665e-05,
+      "loss": 0.7191,
+      "step": 5432
+    },
+    {
+      "epoch": 0.399760864566567,
+      "grad_norm": 0.625,
+      "learning_rate": 3.2848702488618025e-05,
+      "loss": 0.4963,
+      "step": 5433
+    },
+    {
+      "epoch": 0.399834444699931,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.284320350549436e-05,
+      "loss": 0.707,
+      "step": 5434
+    },
+    {
+      "epoch": 0.399908024833295,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.283770410144078e-05,
+      "loss": 1.1303,
+      "step": 5435
+    },
+    {
+      "epoch": 0.399981604966659,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.2832204276752435e-05,
+      "loss": 1.1261,
+      "step": 5436
+    },
+    {
+      "epoch": 0.40005518510002297,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.282670403172447e-05,
+      "loss": 1.2074,
+      "step": 5437
+    },
+    {
+      "epoch": 0.400128765233387,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.282120336665207e-05,
+      "loss": 0.7685,
+      "step": 5438
+    },
+    {
+      "epoch": 0.400202345366751,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.281570228183047e-05,
+      "loss": 1.1608,
+      "step": 5439
+    },
+    {
+      "epoch": 0.400275925500115,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.2810200777554887e-05,
+      "loss": 0.7845,
+      "step": 5440
+    },
+    {
+      "epoch": 0.400349505633479,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.280469885412058e-05,
+      "loss": 0.9856,
+      "step": 5441
+    },
+    {
+      "epoch": 0.40042308576684293,
+      "grad_norm": 0.875,
+      "learning_rate": 3.2799196511822835e-05,
+      "loss": 0.8395,
+      "step": 5442
+    },
+    {
+      "epoch": 0.40049666590020694,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.279369375095694e-05,
+      "loss": 0.8938,
+      "step": 5443
+    },
+    {
+      "epoch": 0.40057024603357094,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.278819057181821e-05,
+      "loss": 1.1752,
+      "step": 5444
+    },
+    {
+      "epoch": 0.40064382616693495,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.278268697470202e-05,
+      "loss": 1.0623,
+      "step": 5445
+    },
+    {
+      "epoch": 0.4007174063002989,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.2777182959903705e-05,
+      "loss": 0.7103,
+      "step": 5446
+    },
+    {
+      "epoch": 0.4007909864336629,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.277167852771867e-05,
+      "loss": 0.6163,
+      "step": 5447
+    },
+    {
+      "epoch": 0.4008645665670269,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.276617367844233e-05,
+      "loss": 0.9038,
+      "step": 5448
+    },
+    {
+      "epoch": 0.4009381467003909,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.276066841237012e-05,
+      "loss": 0.7005,
+      "step": 5449
+    },
+    {
+      "epoch": 0.4010117268337549,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.275516272979749e-05,
+      "loss": 1.1644,
+      "step": 5450
+    },
+    {
+      "epoch": 0.40108530696711886,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.274965663101993e-05,
+      "loss": 1.0017,
+      "step": 5451
+    },
+    {
+      "epoch": 0.40115888710048286,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.274415011633294e-05,
+      "loss": 1.1346,
+      "step": 5452
+    },
+    {
+      "epoch": 0.40123246723384687,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.2738643186032026e-05,
+      "loss": 0.9945,
+      "step": 5453
+    },
+    {
+      "epoch": 0.40130604736721087,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.273313584041276e-05,
+      "loss": 0.8145,
+      "step": 5454
+    },
+    {
+      "epoch": 0.4013796275005748,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.272762807977068e-05,
+      "loss": 1.1186,
+      "step": 5455
+    },
+    {
+      "epoch": 0.4014532076339388,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.272211990440141e-05,
+      "loss": 1.1701,
+      "step": 5456
+    },
+    {
+      "epoch": 0.40152678776730283,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.2716611314600554e-05,
+      "loss": 0.8724,
+      "step": 5457
+    },
+    {
+      "epoch": 0.40160036790066683,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.271110231066374e-05,
+      "loss": 0.9826,
+      "step": 5458
+    },
+    {
+      "epoch": 0.40167394803403084,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.270559289288663e-05,
+      "loss": 0.9174,
+      "step": 5459
+    },
+    {
+      "epoch": 0.4017475281673948,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.27000830615649e-05,
+      "loss": 1.2029,
+      "step": 5460
+    },
+    {
+      "epoch": 0.4018211083007588,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.269457281699426e-05,
+      "loss": 0.7479,
+      "step": 5461
+    },
+    {
+      "epoch": 0.4018946884341228,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.268906215947042e-05,
+      "loss": 0.7568,
+      "step": 5462
+    },
+    {
+      "epoch": 0.4019682685674868,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.2683551089289144e-05,
+      "loss": 0.8404,
+      "step": 5463
+    },
+    {
+      "epoch": 0.40204184870085075,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.267803960674619e-05,
+      "loss": 1.1092,
+      "step": 5464
+    },
+    {
+      "epoch": 0.40211542883421475,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.2672527712137356e-05,
+      "loss": 1.0652,
+      "step": 5465
+    },
+    {
+      "epoch": 0.40218900896757875,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.2667015405758453e-05,
+      "loss": 1.0759,
+      "step": 5466
+    },
+    {
+      "epoch": 0.40226258910094276,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.266150268790531e-05,
+      "loss": 1.0637,
+      "step": 5467
+    },
+    {
+      "epoch": 0.40233616923430676,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.265598955887379e-05,
+      "loss": 1.2914,
+      "step": 5468
+    },
+    {
+      "epoch": 0.4024097493676707,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.2650476018959765e-05,
+      "loss": 0.7047,
+      "step": 5469
+    },
+    {
+      "epoch": 0.4024833295010347,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.264496206845915e-05,
+      "loss": 1.0054,
+      "step": 5470
+    },
+    {
+      "epoch": 0.4025569096343987,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.263944770766785e-05,
+      "loss": 1.0365,
+      "step": 5471
+    },
+    {
+      "epoch": 0.4026304897677627,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.2633932936881825e-05,
+      "loss": 1.5661,
+      "step": 5472
+    },
+    {
+      "epoch": 0.40270406990112667,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.262841775639705e-05,
+      "loss": 1.0951,
+      "step": 5473
+    },
+    {
+      "epoch": 0.4027776500344907,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.262290216650948e-05,
+      "loss": 0.7258,
+      "step": 5474
+    },
+    {
+      "epoch": 0.4028512301678547,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.2617386167515164e-05,
+      "loss": 0.8595,
+      "step": 5475
+    },
+    {
+      "epoch": 0.4029248103012187,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.261186975971011e-05,
+      "loss": 0.6583,
+      "step": 5476
+    },
+    {
+      "epoch": 0.4029983904345827,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.260635294339039e-05,
+      "loss": 0.9518,
+      "step": 5477
+    },
+    {
+      "epoch": 0.40307197056794664,
+      "grad_norm": 1.0,
+      "learning_rate": 3.260083571885207e-05,
+      "loss": 1.1576,
+      "step": 5478
+    },
+    {
+      "epoch": 0.40314555070131064,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.259531808639126e-05,
+      "loss": 0.9954,
+      "step": 5479
+    },
+    {
+      "epoch": 0.40321913083467464,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.258980004630407e-05,
+      "loss": 1.269,
+      "step": 5480
+    },
+    {
+      "epoch": 0.40329271096803865,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.258428159888664e-05,
+      "loss": 1.5506,
+      "step": 5481
+    },
+    {
+      "epoch": 0.4033662911014026,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.257876274443515e-05,
+      "loss": 1.0325,
+      "step": 5482
+    },
+    {
+      "epoch": 0.4034398712347666,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.2573243483245774e-05,
+      "loss": 0.778,
+      "step": 5483
+    },
+    {
+      "epoch": 0.4035134513681306,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.2567723815614726e-05,
+      "loss": 0.7232,
+      "step": 5484
+    },
+    {
+      "epoch": 0.4035870315014946,
+      "grad_norm": 1.0,
+      "learning_rate": 3.256220374183823e-05,
+      "loss": 1.0646,
+      "step": 5485
+    },
+    {
+      "epoch": 0.4036606116348586,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.255668326221255e-05,
+      "loss": 1.1485,
+      "step": 5486
+    },
+    {
+      "epoch": 0.40373419176822256,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.2551162377033945e-05,
+      "loss": 1.0045,
+      "step": 5487
+    },
+    {
+      "epoch": 0.40380777190158657,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.254564108659872e-05,
+      "loss": 0.6102,
+      "step": 5488
+    },
+    {
+      "epoch": 0.40388135203495057,
+      "grad_norm": 0.9609375,
+      "learning_rate": 3.2540119391203186e-05,
+      "loss": 1.3953,
+      "step": 5489
+    },
+    {
+      "epoch": 0.4039549321683146,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.253459729114368e-05,
+      "loss": 0.9125,
+      "step": 5490
+    },
+    {
+      "epoch": 0.4040285123016785,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.252907478671659e-05,
+      "loss": 1.0289,
+      "step": 5491
+    },
+    {
+      "epoch": 0.4041020924350425,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.252355187821825e-05,
+      "loss": 0.8511,
+      "step": 5492
+    },
+    {
+      "epoch": 0.40417567256840653,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.2518028565945104e-05,
+      "loss": 1.0297,
+      "step": 5493
+    },
+    {
+      "epoch": 0.40424925270177053,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.251250485019357e-05,
+      "loss": 0.7107,
+      "step": 5494
+    },
+    {
+      "epoch": 0.40432283283513454,
+      "grad_norm": 0.72265625,
+      "learning_rate": 3.250698073126008e-05,
+      "loss": 0.6151,
+      "step": 5495
+    },
+    {
+      "epoch": 0.4043964129684985,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.2501456209441114e-05,
+      "loss": 0.717,
+      "step": 5496
+    },
+    {
+      "epoch": 0.4044699931018625,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.2495931285033166e-05,
+      "loss": 0.842,
+      "step": 5497
+    },
+    {
+      "epoch": 0.4045435732352265,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.249040595833274e-05,
+      "loss": 0.8624,
+      "step": 5498
+    },
+    {
+      "epoch": 0.4046171533685905,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.2484880229636375e-05,
+      "loss": 0.7128,
+      "step": 5499
+    },
+    {
+      "epoch": 0.40469073350195445,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.247935409924063e-05,
+      "loss": 0.6546,
+      "step": 5500
+    },
+    {
+      "epoch": 0.40476431363531845,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.247382756744207e-05,
+      "loss": 0.7038,
+      "step": 5501
+    },
+    {
+      "epoch": 0.40483789376868246,
+      "grad_norm": 1.1171875,
+      "learning_rate": 3.24683006345373e-05,
+      "loss": 1.0853,
+      "step": 5502
+    },
+    {
+      "epoch": 0.40491147390204646,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.246277330082295e-05,
+      "loss": 0.7791,
+      "step": 5503
+    },
+    {
+      "epoch": 0.40498505403541046,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.245724556659564e-05,
+      "loss": 0.7281,
+      "step": 5504
+    },
+    {
+      "epoch": 0.4050586341687744,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.245171743215205e-05,
+      "loss": 0.8079,
+      "step": 5505
+    },
+    {
+      "epoch": 0.4051322143021384,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.244618889778886e-05,
+      "loss": 0.9249,
+      "step": 5506
+    },
+    {
+      "epoch": 0.4052057944355024,
+      "grad_norm": 0.625,
+      "learning_rate": 3.2440659963802785e-05,
+      "loss": 0.7506,
+      "step": 5507
+    },
+    {
+      "epoch": 0.4052793745688664,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.243513063049053e-05,
+      "loss": 0.8894,
+      "step": 5508
+    },
+    {
+      "epoch": 0.4053529547022304,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.242960089814886e-05,
+      "loss": 1.2063,
+      "step": 5509
+    },
+    {
+      "epoch": 0.4054265348355944,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.2424070767074546e-05,
+      "loss": 1.0831,
+      "step": 5510
+    },
+    {
+      "epoch": 0.4055001149689584,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.241854023756437e-05,
+      "loss": 0.9957,
+      "step": 5511
+    },
+    {
+      "epoch": 0.4055736951023224,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.241300930991515e-05,
+      "loss": 0.7813,
+      "step": 5512
+    },
+    {
+      "epoch": 0.4056472752356864,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.2407477984423726e-05,
+      "loss": 0.9488,
+      "step": 5513
+    },
+    {
+      "epoch": 0.40572085536905034,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.240194626138696e-05,
+      "loss": 1.2827,
+      "step": 5514
+    },
+    {
+      "epoch": 0.40579443550241434,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.239641414110171e-05,
+      "loss": 1.1106,
+      "step": 5515
+    },
+    {
+      "epoch": 0.40586801563577835,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.239088162386487e-05,
+      "loss": 1.4845,
+      "step": 5516
+    },
+    {
+      "epoch": 0.40594159576914235,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.238534870997338e-05,
+      "loss": 0.9536,
+      "step": 5517
+    },
+    {
+      "epoch": 0.4060151759025063,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.237981539972418e-05,
+      "loss": 1.1186,
+      "step": 5518
+    },
+    {
+      "epoch": 0.4060887560358703,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.2374281693414226e-05,
+      "loss": 0.8825,
+      "step": 5519
+    },
+    {
+      "epoch": 0.4061623361692343,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.2368747591340496e-05,
+      "loss": 0.6731,
+      "step": 5520
+    },
+    {
+      "epoch": 0.4062359163025983,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.2363213093800006e-05,
+      "loss": 1.0177,
+      "step": 5521
+    },
+    {
+      "epoch": 0.4063094964359623,
+      "grad_norm": 1.0625,
+      "learning_rate": 3.235767820108977e-05,
+      "loss": 0.8926,
+      "step": 5522
+    },
+    {
+      "epoch": 0.40638307656932626,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.2352142913506846e-05,
+      "loss": 0.6149,
+      "step": 5523
+    },
+    {
+      "epoch": 0.40645665670269027,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.23466072313483e-05,
+      "loss": 1.2331,
+      "step": 5524
+    },
+    {
+      "epoch": 0.40653023683605427,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.2341071154911215e-05,
+      "loss": 1.1672,
+      "step": 5525
+    },
+    {
+      "epoch": 0.4066038169694183,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.233553468449271e-05,
+      "loss": 0.7292,
+      "step": 5526
+    },
+    {
+      "epoch": 0.4066773971027822,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.232999782038991e-05,
+      "loss": 0.6549,
+      "step": 5527
+    },
+    {
+      "epoch": 0.4067509772361462,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.232446056289997e-05,
+      "loss": 1.1656,
+      "step": 5528
+    },
+    {
+      "epoch": 0.40682455736951023,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.231892291232007e-05,
+      "loss": 1.3971,
+      "step": 5529
+    },
+    {
+      "epoch": 0.40689813750287424,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.23133848689474e-05,
+      "loss": 1.2544,
+      "step": 5530
+    },
+    {
+      "epoch": 0.40697171763623824,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.230784643307917e-05,
+      "loss": 0.8279,
+      "step": 5531
+    },
+    {
+      "epoch": 0.4070452977696022,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.230230760501264e-05,
+      "loss": 0.7329,
+      "step": 5532
+    },
+    {
+      "epoch": 0.4071188779029662,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.229676838504504e-05,
+      "loss": 1.2462,
+      "step": 5533
+    },
+    {
+      "epoch": 0.4071924580363302,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.229122877347366e-05,
+      "loss": 1.0062,
+      "step": 5534
+    },
+    {
+      "epoch": 0.4072660381696942,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.2285688770595815e-05,
+      "loss": 0.9326,
+      "step": 5535
+    },
+    {
+      "epoch": 0.40733961830305815,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.228014837670881e-05,
+      "loss": 0.763,
+      "step": 5536
+    },
+    {
+      "epoch": 0.40741319843642215,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.227460759210999e-05,
+      "loss": 0.7049,
+      "step": 5537
+    },
+    {
+      "epoch": 0.40748677856978616,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.2269066417096715e-05,
+      "loss": 0.78,
+      "step": 5538
+    },
+    {
+      "epoch": 0.40756035870315016,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.2263524851966385e-05,
+      "loss": 0.9355,
+      "step": 5539
+    },
+    {
+      "epoch": 0.40763393883651416,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.2257982897016395e-05,
+      "loss": 0.9373,
+      "step": 5540
+    },
+    {
+      "epoch": 0.4077075189698781,
+      "grad_norm": 0.71484375,
+      "learning_rate": 3.225244055254416e-05,
+      "loss": 0.722,
+      "step": 5541
+    },
+    {
+      "epoch": 0.4077810991032421,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.224689781884715e-05,
+      "loss": 0.9293,
+      "step": 5542
+    },
+    {
+      "epoch": 0.4078546792366061,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.224135469622282e-05,
+      "loss": 1.1588,
+      "step": 5543
+    },
+    {
+      "epoch": 0.4079282593699701,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.223581118496865e-05,
+      "loss": 0.8876,
+      "step": 5544
+    },
+    {
+      "epoch": 0.4080018395033341,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.223026728538216e-05,
+      "loss": 1.0028,
+      "step": 5545
+    },
+    {
+      "epoch": 0.4080754196366981,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.22247229977609e-05,
+      "loss": 1.0211,
+      "step": 5546
+    },
+    {
+      "epoch": 0.4081489997700621,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.221917832240239e-05,
+      "loss": 0.8526,
+      "step": 5547
+    },
+    {
+      "epoch": 0.4082225799034261,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.221363325960422e-05,
+      "loss": 0.8572,
+      "step": 5548
+    },
+    {
+      "epoch": 0.4082961600367901,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.2208087809663966e-05,
+      "loss": 0.5536,
+      "step": 5549
+    },
+    {
+      "epoch": 0.40836974017015404,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.2202541972879256e-05,
+      "loss": 0.7928,
+      "step": 5550
+    },
+    {
+      "epoch": 0.40844332030351804,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.2196995749547725e-05,
+      "loss": 0.8852,
+      "step": 5551
+    },
+    {
+      "epoch": 0.40851690043688205,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.219144913996702e-05,
+      "loss": 0.8989,
+      "step": 5552
+    },
+    {
+      "epoch": 0.40859048057024605,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.218590214443484e-05,
+      "loss": 0.8728,
+      "step": 5553
+    },
+    {
+      "epoch": 0.40866406070361,
+      "grad_norm": 1.0,
+      "learning_rate": 3.218035476324885e-05,
+      "loss": 1.4778,
+      "step": 5554
+    },
+    {
+      "epoch": 0.408737640836974,
+      "grad_norm": 0.91796875,
+      "learning_rate": 3.2174806996706785e-05,
+      "loss": 1.132,
+      "step": 5555
+    },
+    {
+      "epoch": 0.408811220970338,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.216925884510638e-05,
+      "loss": 0.6525,
+      "step": 5556
+    },
+    {
+      "epoch": 0.408884801103702,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.2163710308745385e-05,
+      "loss": 1.2691,
+      "step": 5557
+    },
+    {
+      "epoch": 0.408958381237066,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.21581613879216e-05,
+      "loss": 1.1924,
+      "step": 5558
+    },
+    {
+      "epoch": 0.40903196137042996,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.215261208293281e-05,
+      "loss": 1.179,
+      "step": 5559
+    },
+    {
+      "epoch": 0.40910554150379397,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.214706239407684e-05,
+      "loss": 0.7966,
+      "step": 5560
+    },
+    {
+      "epoch": 0.40917912163715797,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.214151232165152e-05,
+      "loss": 0.7218,
+      "step": 5561
+    },
+    {
+      "epoch": 0.409252701770522,
+      "grad_norm": 1.15625,
+      "learning_rate": 3.213596186595474e-05,
+      "loss": 1.6786,
+      "step": 5562
+    },
+    {
+      "epoch": 0.4093262819038859,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.213041102728435e-05,
+      "loss": 0.7659,
+      "step": 5563
+    },
+    {
+      "epoch": 0.40939986203724993,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.212485980593827e-05,
+      "loss": 0.9232,
+      "step": 5564
+    },
+    {
+      "epoch": 0.40947344217061393,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.211930820221443e-05,
+      "loss": 1.1072,
+      "step": 5565
+    },
+    {
+      "epoch": 0.40954702230397794,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.211375621641075e-05,
+      "loss": 0.8378,
+      "step": 5566
+    },
+    {
+      "epoch": 0.40962060243734194,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.210820384882522e-05,
+      "loss": 0.8205,
+      "step": 5567
+    },
+    {
+      "epoch": 0.4096941825707059,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.2102651099755815e-05,
+      "loss": 1.1259,
+      "step": 5568
+    },
+    {
+      "epoch": 0.4097677627040699,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.209709796950054e-05,
+      "loss": 1.0587,
+      "step": 5569
+    },
+    {
+      "epoch": 0.4098413428374339,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.209154445835742e-05,
+      "loss": 0.7716,
+      "step": 5570
+    },
+    {
+      "epoch": 0.4099149229707979,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.20859905666245e-05,
+      "loss": 1.0481,
+      "step": 5571
+    },
+    {
+      "epoch": 0.40998850310416185,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.208043629459986e-05,
+      "loss": 0.5851,
+      "step": 5572
+    },
+    {
+      "epoch": 0.41006208323752585,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.2074881642581564e-05,
+      "loss": 1.0272,
+      "step": 5573
+    },
+    {
+      "epoch": 0.41013566337088986,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.206932661086775e-05,
+      "loss": 0.9631,
+      "step": 5574
+    },
+    {
+      "epoch": 0.41020924350425386,
+      "grad_norm": 0.71875,
+      "learning_rate": 3.206377119975651e-05,
+      "loss": 0.612,
+      "step": 5575
+    },
+    {
+      "epoch": 0.41028282363761787,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.205821540954602e-05,
+      "loss": 0.769,
+      "step": 5576
+    },
+    {
+      "epoch": 0.4103564037709818,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.205265924053443e-05,
+      "loss": 0.9553,
+      "step": 5577
+    },
+    {
+      "epoch": 0.4104299839043458,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.204710269301996e-05,
+      "loss": 0.867,
+      "step": 5578
+    },
+    {
+      "epoch": 0.4105035640377098,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.2041545767300785e-05,
+      "loss": 0.8351,
+      "step": 5579
+    },
+    {
+      "epoch": 0.4105771441710738,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.203598846367515e-05,
+      "loss": 0.8639,
+      "step": 5580
+    },
+    {
+      "epoch": 0.4106507243044378,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.20304307824413e-05,
+      "loss": 0.6448,
+      "step": 5581
+    },
+    {
+      "epoch": 0.4107243044378018,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.2024872723897514e-05,
+      "loss": 0.5816,
+      "step": 5582
+    },
+    {
+      "epoch": 0.4107978845711658,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.2019314288342075e-05,
+      "loss": 0.8489,
+      "step": 5583
+    },
+    {
+      "epoch": 0.4108714647045298,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.2013755476073294e-05,
+      "loss": 0.9564,
+      "step": 5584
+    },
+    {
+      "epoch": 0.4109450448378938,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.200819628738951e-05,
+      "loss": 1.0322,
+      "step": 5585
+    },
+    {
+      "epoch": 0.41101862497125774,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.200263672258906e-05,
+      "loss": 0.9438,
+      "step": 5586
+    },
+    {
+      "epoch": 0.41109220510462174,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.199707678197033e-05,
+      "loss": 0.9212,
+      "step": 5587
+    },
+    {
+      "epoch": 0.41116578523798575,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.1991516465831696e-05,
+      "loss": 1.0347,
+      "step": 5588
+    },
+    {
+      "epoch": 0.41123936537134975,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.198595577447159e-05,
+      "loss": 1.0681,
+      "step": 5589
+    },
+    {
+      "epoch": 0.4113129455047137,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.1980394708188425e-05,
+      "loss": 0.8626,
+      "step": 5590
+    },
+    {
+      "epoch": 0.4113865256380777,
+      "grad_norm": 4.09375,
+      "learning_rate": 3.197483326728065e-05,
+      "loss": 0.9472,
+      "step": 5591
+    },
+    {
+      "epoch": 0.4114601057714417,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.196927145204676e-05,
+      "loss": 0.9151,
+      "step": 5592
+    },
+    {
+      "epoch": 0.4115336859048057,
+      "grad_norm": 1.125,
+      "learning_rate": 3.1963709262785234e-05,
+      "loss": 1.0061,
+      "step": 5593
+    },
+    {
+      "epoch": 0.4116072660381697,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.1958146699794586e-05,
+      "loss": 0.7624,
+      "step": 5594
+    },
+    {
+      "epoch": 0.41168084617153367,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.195258376337334e-05,
+      "loss": 1.3725,
+      "step": 5595
+    },
+    {
+      "epoch": 0.41175442630489767,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.1947020453820066e-05,
+      "loss": 0.8691,
+      "step": 5596
+    },
+    {
+      "epoch": 0.4118280064382617,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.194145677143331e-05,
+      "loss": 1.0105,
+      "step": 5597
+    },
+    {
+      "epoch": 0.4119015865716257,
+      "grad_norm": 0.75390625,
+      "learning_rate": 3.193589271651169e-05,
+      "loss": 0.9358,
+      "step": 5598
+    },
+    {
+      "epoch": 0.4119751667049896,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.1930328289353805e-05,
+      "loss": 1.1158,
+      "step": 5599
+    },
+    {
+      "epoch": 0.41204874683835363,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.192476349025829e-05,
+      "loss": 0.7623,
+      "step": 5600
+    },
+    {
+      "epoch": 0.41212232697171763,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.1919198319523805e-05,
+      "loss": 0.918,
+      "step": 5601
+    },
+    {
+      "epoch": 0.41219590710508164,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.1913632777449e-05,
+      "loss": 0.7094,
+      "step": 5602
+    },
+    {
+      "epoch": 0.41226948723844564,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.190806686433259e-05,
+      "loss": 0.9955,
+      "step": 5603
+    },
+    {
+      "epoch": 0.4123430673718096,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.190250058047328e-05,
+      "loss": 1.1954,
+      "step": 5604
+    },
+    {
+      "epoch": 0.4124166475051736,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.18969339261698e-05,
+      "loss": 1.2955,
+      "step": 5605
+    },
+    {
+      "epoch": 0.4124902276385376,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.1891366901720896e-05,
+      "loss": 1.2679,
+      "step": 5606
+    },
+    {
+      "epoch": 0.4125638077719016,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.188579950742535e-05,
+      "loss": 1.3088,
+      "step": 5607
+    },
+    {
+      "epoch": 0.41263738790526555,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.188023174358196e-05,
+      "loss": 1.1992,
+      "step": 5608
+    },
+    {
+      "epoch": 0.41271096803862956,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.187466361048952e-05,
+      "loss": 1.1695,
+      "step": 5609
+    },
+    {
+      "epoch": 0.41278454817199356,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.1869095108446864e-05,
+      "loss": 1.1898,
+      "step": 5610
+    },
+    {
+      "epoch": 0.41285812830535756,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.186352623775285e-05,
+      "loss": 0.981,
+      "step": 5611
+    },
+    {
+      "epoch": 0.41293170843872157,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.185795699870635e-05,
+      "loss": 0.73,
+      "step": 5612
+    },
+    {
+      "epoch": 0.4130052885720855,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.185238739160624e-05,
+      "loss": 0.8553,
+      "step": 5613
+    },
+    {
+      "epoch": 0.4130788687054495,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.184681741675145e-05,
+      "loss": 0.7428,
+      "step": 5614
+    },
+    {
+      "epoch": 0.4131524488388135,
+      "grad_norm": 1.1171875,
+      "learning_rate": 3.1841247074440906e-05,
+      "loss": 0.9858,
+      "step": 5615
+    },
+    {
+      "epoch": 0.41322602897217753,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.1835676364973546e-05,
+      "loss": 0.9968,
+      "step": 5616
+    },
+    {
+      "epoch": 0.4132996091055415,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.183010528864835e-05,
+      "loss": 0.9381,
+      "step": 5617
+    },
+    {
+      "epoch": 0.4133731892389055,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.182453384576429e-05,
+      "loss": 0.7072,
+      "step": 5618
+    },
+    {
+      "epoch": 0.4134467693722695,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.18189620366204e-05,
+      "loss": 0.8605,
+      "step": 5619
+    },
+    {
+      "epoch": 0.4135203495056335,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.1813389861515706e-05,
+      "loss": 0.9305,
+      "step": 5620
+    },
+    {
+      "epoch": 0.4135939296389975,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.1807817320749236e-05,
+      "loss": 0.6096,
+      "step": 5621
+    },
+    {
+      "epoch": 0.41366750977236144,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.1802244414620065e-05,
+      "loss": 1.0141,
+      "step": 5622
+    },
+    {
+      "epoch": 0.41374108990572545,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.179667114342729e-05,
+      "loss": 0.7475,
+      "step": 5623
+    },
+    {
+      "epoch": 0.41381467003908945,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.179109750747e-05,
+      "loss": 1.0894,
+      "step": 5624
+    },
+    {
+      "epoch": 0.41388825017245345,
+      "grad_norm": 0.94921875,
+      "learning_rate": 3.1785523507047345e-05,
+      "loss": 1.0363,
+      "step": 5625
+    },
+    {
+      "epoch": 0.4139618303058174,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.177994914245846e-05,
+      "loss": 0.9267,
+      "step": 5626
+    },
+    {
+      "epoch": 0.4140354104391814,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.1774374414002505e-05,
+      "loss": 0.8263,
+      "step": 5627
+    },
+    {
+      "epoch": 0.4141089905725454,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.1768799321978674e-05,
+      "loss": 0.5761,
+      "step": 5628
+    },
+    {
+      "epoch": 0.4141825707059094,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.176322386668617e-05,
+      "loss": 0.7389,
+      "step": 5629
+    },
+    {
+      "epoch": 0.4142561508392734,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.17576480484242e-05,
+      "loss": 0.6863,
+      "step": 5630
+    },
+    {
+      "epoch": 0.41432973097263737,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.1752071867492037e-05,
+      "loss": 0.8422,
+      "step": 5631
+    },
+    {
+      "epoch": 0.41440331110600137,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.174649532418893e-05,
+      "loss": 1.3692,
+      "step": 5632
+    },
+    {
+      "epoch": 0.4144768912393654,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.1740918418814156e-05,
+      "loss": 1.0136,
+      "step": 5633
+    },
+    {
+      "epoch": 0.4145504713727294,
+      "grad_norm": 1.1640625,
+      "learning_rate": 3.173534115166702e-05,
+      "loss": 1.1615,
+      "step": 5634
+    },
+    {
+      "epoch": 0.4146240515060933,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.1729763523046864e-05,
+      "loss": 0.5981,
+      "step": 5635
+    },
+    {
+      "epoch": 0.41469763163945733,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.172418553325299e-05,
+      "loss": 1.1503,
+      "step": 5636
+    },
+    {
+      "epoch": 0.41477121177282134,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.171860718258478e-05,
+      "loss": 0.7839,
+      "step": 5637
+    },
+    {
+      "epoch": 0.41484479190618534,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.1713028471341624e-05,
+      "loss": 1.2209,
+      "step": 5638
+    },
+    {
+      "epoch": 0.41491837203954934,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.17074493998229e-05,
+      "loss": 0.5049,
+      "step": 5639
+    },
+    {
+      "epoch": 0.4149919521729133,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.1701869968328036e-05,
+      "loss": 0.8211,
+      "step": 5640
+    },
+    {
+      "epoch": 0.4150655323062773,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.169629017715647e-05,
+      "loss": 0.8303,
+      "step": 5641
+    },
+    {
+      "epoch": 0.4151391124396413,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.169071002660766e-05,
+      "loss": 0.6896,
+      "step": 5642
+    },
+    {
+      "epoch": 0.4152126925730053,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.1685129516981076e-05,
+      "loss": 1.0469,
+      "step": 5643
+    },
+    {
+      "epoch": 0.41528627270636925,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.1679548648576216e-05,
+      "loss": 0.9086,
+      "step": 5644
+    },
+    {
+      "epoch": 0.41535985283973326,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.167396742169259e-05,
+      "loss": 1.2287,
+      "step": 5645
+    },
+    {
+      "epoch": 0.41543343297309726,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.166838583662975e-05,
+      "loss": 0.5987,
+      "step": 5646
+    },
+    {
+      "epoch": 0.41550701310646126,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.1662803893687244e-05,
+      "loss": 1.229,
+      "step": 5647
+    },
+    {
+      "epoch": 0.41558059323982527,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.165722159316462e-05,
+      "loss": 0.9456,
+      "step": 5648
+    },
+    {
+      "epoch": 0.4156541733731892,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.16516389353615e-05,
+      "loss": 0.9693,
+      "step": 5649
+    },
+    {
+      "epoch": 0.4157277535065532,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.164605592057747e-05,
+      "loss": 0.6719,
+      "step": 5650
+    },
+    {
+      "epoch": 0.4158013336399172,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.1640472549112174e-05,
+      "loss": 0.6725,
+      "step": 5651
+    },
+    {
+      "epoch": 0.41587491377328123,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.163488882126526e-05,
+      "loss": 0.6966,
+      "step": 5652
+    },
+    {
+      "epoch": 0.4159484939066452,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.1629304737336404e-05,
+      "loss": 1.3396,
+      "step": 5653
+    },
+    {
+      "epoch": 0.4160220740400092,
+      "grad_norm": 0.69921875,
+      "learning_rate": 3.162372029762527e-05,
+      "loss": 0.647,
+      "step": 5654
+    },
+    {
+      "epoch": 0.4160956541733732,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.161813550243158e-05,
+      "loss": 0.958,
+      "step": 5655
+    },
+    {
+      "epoch": 0.4161692343067372,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.161255035205507e-05,
+      "loss": 1.1894,
+      "step": 5656
+    },
+    {
+      "epoch": 0.4162428144401012,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.160696484679546e-05,
+      "loss": 0.9956,
+      "step": 5657
+    },
+    {
+      "epoch": 0.41631639457346514,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.160137898695252e-05,
+      "loss": 0.8505,
+      "step": 5658
+    },
+    {
+      "epoch": 0.41638997470682915,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.159579277282605e-05,
+      "loss": 0.9437,
+      "step": 5659
+    },
+    {
+      "epoch": 0.41646355484019315,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.159020620471584e-05,
+      "loss": 0.9145,
+      "step": 5660
+    },
+    {
+      "epoch": 0.41653713497355715,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.1584619282921704e-05,
+      "loss": 1.3915,
+      "step": 5661
+    },
+    {
+      "epoch": 0.41661071510692116,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.1579032007743485e-05,
+      "loss": 0.8081,
+      "step": 5662
+    },
+    {
+      "epoch": 0.4166842952402851,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.157344437948105e-05,
+      "loss": 0.8382,
+      "step": 5663
+    },
+    {
+      "epoch": 0.4167578753736491,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.156785639843427e-05,
+      "loss": 0.824,
+      "step": 5664
+    },
+    {
+      "epoch": 0.4168314555070131,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.156226806490304e-05,
+      "loss": 1.2242,
+      "step": 5665
+    },
+    {
+      "epoch": 0.4169050356403771,
+      "grad_norm": 0.75,
+      "learning_rate": 3.1556679379187274e-05,
+      "loss": 1.0485,
+      "step": 5666
+    },
+    {
+      "epoch": 0.41697861577374107,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.1551090341586917e-05,
+      "loss": 1.1262,
+      "step": 5667
+    },
+    {
+      "epoch": 0.41705219590710507,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.154550095240191e-05,
+      "loss": 0.8812,
+      "step": 5668
+    },
+    {
+      "epoch": 0.4171257760404691,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.153991121193224e-05,
+      "loss": 0.686,
+      "step": 5669
+    },
+    {
+      "epoch": 0.4171993561738331,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.153432112047787e-05,
+      "loss": 0.8766,
+      "step": 5670
+    },
+    {
+      "epoch": 0.4172729363071971,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.152873067833884e-05,
+      "loss": 0.9772,
+      "step": 5671
+    },
+    {
+      "epoch": 0.41734651644056103,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.1523139885815154e-05,
+      "loss": 1.5448,
+      "step": 5672
+    },
+    {
+      "epoch": 0.41742009657392504,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.151754874320688e-05,
+      "loss": 1.0638,
+      "step": 5673
+    },
+    {
+      "epoch": 0.41749367670728904,
+      "grad_norm": 1.0234375,
+      "learning_rate": 3.151195725081408e-05,
+      "loss": 1.1067,
+      "step": 5674
+    },
+    {
+      "epoch": 0.41756725684065304,
+      "grad_norm": 0.97265625,
+      "learning_rate": 3.150636540893682e-05,
+      "loss": 1.1005,
+      "step": 5675
+    },
+    {
+      "epoch": 0.417640836974017,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.150077321787523e-05,
+      "loss": 0.9598,
+      "step": 5676
+    },
+    {
+      "epoch": 0.417714417107381,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.149518067792941e-05,
+      "loss": 1.0019,
+      "step": 5677
+    },
+    {
+      "epoch": 0.417787997240745,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.148958778939951e-05,
+      "loss": 0.9,
+      "step": 5678
+    },
+    {
+      "epoch": 0.417861577374109,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.1483994552585696e-05,
+      "loss": 1.0215,
+      "step": 5679
+    },
+    {
+      "epoch": 0.417935157507473,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.147840096778813e-05,
+      "loss": 0.9514,
+      "step": 5680
+    },
+    {
+      "epoch": 0.41800873764083696,
+      "grad_norm": 0.99609375,
+      "learning_rate": 3.147280703530704e-05,
+      "loss": 0.9191,
+      "step": 5681
+    },
+    {
+      "epoch": 0.41808231777420096,
+      "grad_norm": 0.7734375,
+      "learning_rate": 3.146721275544261e-05,
+      "loss": 0.6177,
+      "step": 5682
+    },
+    {
+      "epoch": 0.41815589790756497,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.146161812849509e-05,
+      "loss": 0.9278,
+      "step": 5683
+    },
+    {
+      "epoch": 0.41822947804092897,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.1456023154764725e-05,
+      "loss": 1.3063,
+      "step": 5684
+    },
+    {
+      "epoch": 0.4183030581742929,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.1450427834551795e-05,
+      "loss": 0.8765,
+      "step": 5685
+    },
+    {
+      "epoch": 0.4183766383076569,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.144483216815658e-05,
+      "loss": 1.439,
+      "step": 5686
+    },
+    {
+      "epoch": 0.4184502184410209,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.14392361558794e-05,
+      "loss": 1.2136,
+      "step": 5687
+    },
+    {
+      "epoch": 0.41852379857438493,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.1433639798020575e-05,
+      "loss": 0.8733,
+      "step": 5688
+    },
+    {
+      "epoch": 0.41859737870774893,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.142804309488045e-05,
+      "loss": 0.8334,
+      "step": 5689
+    },
+    {
+      "epoch": 0.4186709588411129,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.142244604675941e-05,
+      "loss": 0.7069,
+      "step": 5690
+    },
+    {
+      "epoch": 0.4187445389744769,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.14168486539578e-05,
+      "loss": 0.8814,
+      "step": 5691
+    },
+    {
+      "epoch": 0.4188181191078409,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.141125091677605e-05,
+      "loss": 0.6196,
+      "step": 5692
+    },
+    {
+      "epoch": 0.4188916992412049,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.140565283551458e-05,
+      "loss": 0.9069,
+      "step": 5693
+    },
+    {
+      "epoch": 0.41896527937456884,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.140005441047381e-05,
+      "loss": 1.2138,
+      "step": 5694
+    },
+    {
+      "epoch": 0.41903885950793285,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.139445564195421e-05,
+      "loss": 1.1058,
+      "step": 5695
+    },
+    {
+      "epoch": 0.41911243964129685,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.138885653025626e-05,
+      "loss": 0.955,
+      "step": 5696
+    },
+    {
+      "epoch": 0.41918601977466086,
+      "grad_norm": 0.8046875,
+      "learning_rate": 3.1383257075680446e-05,
+      "loss": 0.722,
+      "step": 5697
+    },
+    {
+      "epoch": 0.41925959990802486,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.1377657278527274e-05,
+      "loss": 1.0323,
+      "step": 5698
+    },
+    {
+      "epoch": 0.4193331800413888,
+      "grad_norm": 0.7578125,
+      "learning_rate": 3.1372057139097284e-05,
+      "loss": 0.8831,
+      "step": 5699
+    },
+    {
+      "epoch": 0.4194067601747528,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.136645665769102e-05,
+      "loss": 1.2314,
+      "step": 5700
+    },
+    {
+      "epoch": 0.4194803403081168,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.136085583460906e-05,
+      "loss": 1.176,
+      "step": 5701
+    },
+    {
+      "epoch": 0.4195539204414808,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.135525467015197e-05,
+      "loss": 0.8508,
+      "step": 5702
+    },
+    {
+      "epoch": 0.41962750057484477,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.134965316462037e-05,
+      "loss": 0.9366,
+      "step": 5703
+    },
+    {
+      "epoch": 0.4197010807082088,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.134405131831489e-05,
+      "loss": 0.623,
+      "step": 5704
+    },
+    {
+      "epoch": 0.4197746608415728,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.133844913153614e-05,
+      "loss": 0.7714,
+      "step": 5705
+    },
+    {
+      "epoch": 0.4198482409749368,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.133284660458481e-05,
+      "loss": 1.3795,
+      "step": 5706
+    },
+    {
+      "epoch": 0.4199218211083008,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.1327243737761555e-05,
+      "loss": 0.888,
+      "step": 5707
+    },
+    {
+      "epoch": 0.41999540124166473,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.132164053136708e-05,
+      "loss": 0.8543,
+      "step": 5708
+    },
+    {
+      "epoch": 0.42006898137502874,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.13160369857021e-05,
+      "loss": 0.7844,
+      "step": 5709
+    },
+    {
+      "epoch": 0.42014256150839274,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.131043310106735e-05,
+      "loss": 1.0513,
+      "step": 5710
+    },
+    {
+      "epoch": 0.42021614164175675,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.130482887776356e-05,
+      "loss": 1.0161,
+      "step": 5711
+    },
+    {
+      "epoch": 0.4202897217751207,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.129922431609152e-05,
+      "loss": 0.8555,
+      "step": 5712
+    },
+    {
+      "epoch": 0.4203633019084847,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.1293619416352005e-05,
+      "loss": 0.7085,
+      "step": 5713
+    },
+    {
+      "epoch": 0.4204368820418487,
+      "grad_norm": 0.875,
+      "learning_rate": 3.1288014178845824e-05,
+      "loss": 0.7576,
+      "step": 5714
+    },
+    {
+      "epoch": 0.4205104621752127,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.128240860387381e-05,
+      "loss": 0.6418,
+      "step": 5715
+    },
+    {
+      "epoch": 0.4205840423085767,
+      "grad_norm": 0.69140625,
+      "learning_rate": 3.127680269173678e-05,
+      "loss": 0.6519,
+      "step": 5716
+    },
+    {
+      "epoch": 0.42065762244194066,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.1271196442735606e-05,
+      "loss": 1.6772,
+      "step": 5717
+    },
+    {
+      "epoch": 0.42073120257530466,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.126558985717117e-05,
+      "loss": 0.9215,
+      "step": 5718
+    },
+    {
+      "epoch": 0.42080478270866867,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.125998293534435e-05,
+      "loss": 0.8935,
+      "step": 5719
+    },
+    {
+      "epoch": 0.42087836284203267,
+      "grad_norm": 1.1328125,
+      "learning_rate": 3.1254375677556076e-05,
+      "loss": 1.3714,
+      "step": 5720
+    },
+    {
+      "epoch": 0.4209519429753966,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.1248768084107265e-05,
+      "loss": 0.8136,
+      "step": 5721
+    },
+    {
+      "epoch": 0.4210255231087606,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.1243160155298886e-05,
+      "loss": 1.1025,
+      "step": 5722
+    },
+    {
+      "epoch": 0.42109910324212463,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.123755189143188e-05,
+      "loss": 1.3903,
+      "step": 5723
+    },
+    {
+      "epoch": 0.42117268337548863,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.123194329280726e-05,
+      "loss": 1.1325,
+      "step": 5724
+    },
+    {
+      "epoch": 0.42124626350885264,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.1226334359726e-05,
+      "loss": 1.3851,
+      "step": 5725
+    },
+    {
+      "epoch": 0.4213198436422166,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.122072509248914e-05,
+      "loss": 0.7743,
+      "step": 5726
+    },
+    {
+      "epoch": 0.4213934237755806,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.1215115491397714e-05,
+      "loss": 1.2776,
+      "step": 5727
+    },
+    {
+      "epoch": 0.4214670039089446,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.120950555675277e-05,
+      "loss": 1.1721,
+      "step": 5728
+    },
+    {
+      "epoch": 0.4215405840423086,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.120389528885541e-05,
+      "loss": 1.3343,
+      "step": 5729
+    },
+    {
+      "epoch": 0.42161416417567255,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.119828468800669e-05,
+      "loss": 0.9707,
+      "step": 5730
+    },
+    {
+      "epoch": 0.42168774430903655,
+      "grad_norm": 1.109375,
+      "learning_rate": 3.1192673754507754e-05,
+      "loss": 1.2327,
+      "step": 5731
+    },
+    {
+      "epoch": 0.42176132444240055,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.11870624886597e-05,
+      "loss": 0.8943,
+      "step": 5732
+    },
+    {
+      "epoch": 0.42183490457576456,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.118145089076369e-05,
+      "loss": 1.0463,
+      "step": 5733
+    },
+    {
+      "epoch": 0.42190848470912856,
+      "grad_norm": 1.0,
+      "learning_rate": 3.117583896112088e-05,
+      "loss": 0.8832,
+      "step": 5734
+    },
+    {
+      "epoch": 0.4219820648424925,
+      "grad_norm": 1.1015625,
+      "learning_rate": 3.117022670003246e-05,
+      "loss": 1.1644,
+      "step": 5735
+    },
+    {
+      "epoch": 0.4220556449758565,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.116461410779964e-05,
+      "loss": 0.916,
+      "step": 5736
+    },
+    {
+      "epoch": 0.4221292251092205,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.1159001184723615e-05,
+      "loss": 1.1364,
+      "step": 5737
+    },
+    {
+      "epoch": 0.4222028052425845,
+      "grad_norm": 1.1171875,
+      "learning_rate": 3.115338793110563e-05,
+      "loss": 1.3442,
+      "step": 5738
+    },
+    {
+      "epoch": 0.42227638537594847,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.114777434724694e-05,
+      "loss": 0.7471,
+      "step": 5739
+    },
+    {
+      "epoch": 0.4223499655093125,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.11421604334488e-05,
+      "loss": 1.0077,
+      "step": 5740
+    },
+    {
+      "epoch": 0.4224235456426765,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.1136546190012524e-05,
+      "loss": 0.7534,
+      "step": 5741
+    },
+    {
+      "epoch": 0.4224971257760405,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.11309316172394e-05,
+      "loss": 0.9952,
+      "step": 5742
+    },
+    {
+      "epoch": 0.4225707059094045,
+      "grad_norm": 1.234375,
+      "learning_rate": 3.1125316715430755e-05,
+      "loss": 1.0639,
+      "step": 5743
+    },
+    {
+      "epoch": 0.42264428604276844,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.111970148488793e-05,
+      "loss": 0.8631,
+      "step": 5744
+    },
+    {
+      "epoch": 0.42271786617613244,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.111408592591229e-05,
+      "loss": 1.0575,
+      "step": 5745
+    },
+    {
+      "epoch": 0.42279144630949644,
+      "grad_norm": 0.76953125,
+      "learning_rate": 3.11084700388052e-05,
+      "loss": 0.7257,
+      "step": 5746
+    },
+    {
+      "epoch": 0.42286502644286045,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.110285382386806e-05,
+      "loss": 0.6819,
+      "step": 5747
+    },
+    {
+      "epoch": 0.4229386065762244,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.109723728140229e-05,
+      "loss": 0.7694,
+      "step": 5748
+    },
+    {
+      "epoch": 0.4230121867095884,
+      "grad_norm": 0.83984375,
+      "learning_rate": 3.10916204117093e-05,
+      "loss": 1.0799,
+      "step": 5749
+    },
+    {
+      "epoch": 0.4230857668429524,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.108600321509056e-05,
+      "loss": 0.9085,
+      "step": 5750
+    },
+    {
+      "epoch": 0.4231593469763164,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.108038569184751e-05,
+      "loss": 1.1044,
+      "step": 5751
+    },
+    {
+      "epoch": 0.4232329271096804,
+      "grad_norm": 0.9921875,
+      "learning_rate": 3.1074767842281654e-05,
+      "loss": 1.2157,
+      "step": 5752
+    },
+    {
+      "epoch": 0.42330650724304436,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.1069149666694474e-05,
+      "loss": 1.0037,
+      "step": 5753
+    },
+    {
+      "epoch": 0.42338008737640836,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.10635311653875e-05,
+      "loss": 0.9024,
+      "step": 5754
+    },
+    {
+      "epoch": 0.42345366750977237,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.105791233866226e-05,
+      "loss": 1.5471,
+      "step": 5755
+    },
+    {
+      "epoch": 0.4235272476431364,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.10522931868203e-05,
+      "loss": 1.5793,
+      "step": 5756
+    },
+    {
+      "epoch": 0.4236008277765003,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.1046673710163206e-05,
+      "loss": 0.7626,
+      "step": 5757
+    },
+    {
+      "epoch": 0.4236744079098643,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.104105390899255e-05,
+      "loss": 0.947,
+      "step": 5758
+    },
+    {
+      "epoch": 0.42374798804322833,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.103543378360994e-05,
+      "loss": 0.9181,
+      "step": 5759
+    },
+    {
+      "epoch": 0.42382156817659233,
+      "grad_norm": 0.875,
+      "learning_rate": 3.1029813334316994e-05,
+      "loss": 0.784,
+      "step": 5760
+    },
+    {
+      "epoch": 0.42389514830995634,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.102419256141536e-05,
+      "loss": 0.951,
+      "step": 5761
+    },
+    {
+      "epoch": 0.4239687284433203,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.101857146520669e-05,
+      "loss": 0.9759,
+      "step": 5762
+    },
+    {
+      "epoch": 0.4240423085766843,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.101295004599266e-05,
+      "loss": 1.1313,
+      "step": 5763
+    },
+    {
+      "epoch": 0.4241158887100483,
+      "grad_norm": 1.5703125,
+      "learning_rate": 3.100732830407495e-05,
+      "loss": 1.0304,
+      "step": 5764
+    },
+    {
+      "epoch": 0.4241894688434123,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.1001706239755286e-05,
+      "loss": 0.9478,
+      "step": 5765
+    },
+    {
+      "epoch": 0.42426304897677625,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.0996083853335374e-05,
+      "loss": 0.8076,
+      "step": 5766
+    },
+    {
+      "epoch": 0.42433662911014025,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.099046114511696e-05,
+      "loss": 0.5262,
+      "step": 5767
+    },
+    {
+      "epoch": 0.42441020924350426,
+      "grad_norm": 1.1640625,
+      "learning_rate": 3.098483811540183e-05,
+      "loss": 1.2619,
+      "step": 5768
+    },
+    {
+      "epoch": 0.42448378937686826,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.097921476449173e-05,
+      "loss": 0.7225,
+      "step": 5769
+    },
+    {
+      "epoch": 0.42455736951023226,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.097359109268847e-05,
+      "loss": 1.2377,
+      "step": 5770
+    },
+    {
+      "epoch": 0.4246309496435962,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.096796710029386e-05,
+      "loss": 1.4803,
+      "step": 5771
+    },
+    {
+      "epoch": 0.4247045297769602,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.0962342787609725e-05,
+      "loss": 1.1074,
+      "step": 5772
+    },
+    {
+      "epoch": 0.4247781099103242,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.0956718154937915e-05,
+      "loss": 1.1368,
+      "step": 5773
+    },
+    {
+      "epoch": 0.4248516900436882,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.095109320258029e-05,
+      "loss": 1.1014,
+      "step": 5774
+    },
+    {
+      "epoch": 0.4249252701770522,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.094546793083873e-05,
+      "loss": 1.9882,
+      "step": 5775
+    },
+    {
+      "epoch": 0.4249988503104162,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.0939842340015145e-05,
+      "loss": 1.282,
+      "step": 5776
+    },
+    {
+      "epoch": 0.4250724304437802,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.093421643041144e-05,
+      "loss": 0.8593,
+      "step": 5777
+    },
+    {
+      "epoch": 0.4251460105771442,
+      "grad_norm": 1.0,
+      "learning_rate": 3.092859020232954e-05,
+      "loss": 0.9462,
+      "step": 5778
+    },
+    {
+      "epoch": 0.4252195907105082,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.092296365607141e-05,
+      "loss": 0.8246,
+      "step": 5779
+    },
+    {
+      "epoch": 0.42529317084387214,
+      "grad_norm": 0.94140625,
+      "learning_rate": 3.091733679193899e-05,
+      "loss": 0.9693,
+      "step": 5780
+    },
+    {
+      "epoch": 0.42536675097723614,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.09117096102343e-05,
+      "loss": 0.7689,
+      "step": 5781
+    },
+    {
+      "epoch": 0.42544033111060015,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.090608211125931e-05,
+      "loss": 0.9306,
+      "step": 5782
+    },
+    {
+      "epoch": 0.42551391124396415,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.090045429531605e-05,
+      "loss": 1.2965,
+      "step": 5783
+    },
+    {
+      "epoch": 0.4255874913773281,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.089482616270656e-05,
+      "loss": 0.8974,
+      "step": 5784
+    },
+    {
+      "epoch": 0.4256610715106921,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.088919771373287e-05,
+      "loss": 1.1757,
+      "step": 5785
+    },
+    {
+      "epoch": 0.4257346516440561,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.0883568948697077e-05,
+      "loss": 1.1472,
+      "step": 5786
+    },
+    {
+      "epoch": 0.4258082317774201,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.0877939867901235e-05,
+      "loss": 0.823,
+      "step": 5787
+    },
+    {
+      "epoch": 0.4258818119107841,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.0872310471647474e-05,
+      "loss": 1.0921,
+      "step": 5788
+    },
+    {
+      "epoch": 0.42595539204414806,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.08666807602379e-05,
+      "loss": 0.9897,
+      "step": 5789
+    },
+    {
+      "epoch": 0.42602897217751207,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.086105073397465e-05,
+      "loss": 0.7113,
+      "step": 5790
+    },
+    {
+      "epoch": 0.42610255231087607,
+      "grad_norm": 0.68359375,
+      "learning_rate": 3.0855420393159874e-05,
+      "loss": 0.6338,
+      "step": 5791
+    },
+    {
+      "epoch": 0.4261761324442401,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.084978973809574e-05,
+      "loss": 1.2566,
+      "step": 5792
+    },
+    {
+      "epoch": 0.426249712577604,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.084415876908444e-05,
+      "loss": 0.7628,
+      "step": 5793
+    },
+    {
+      "epoch": 0.426323292710968,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.083852748642818e-05,
+      "loss": 0.7678,
+      "step": 5794
+    },
+    {
+      "epoch": 0.42639687284433203,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.083289589042918e-05,
+      "loss": 0.9107,
+      "step": 5795
+    },
+    {
+      "epoch": 0.42647045297769604,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.082726398138968e-05,
+      "loss": 0.8132,
+      "step": 5796
+    },
+    {
+      "epoch": 0.42654403311106004,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.0821631759611925e-05,
+      "loss": 0.616,
+      "step": 5797
+    },
+    {
+      "epoch": 0.426617613244424,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.0815999225398186e-05,
+      "loss": 1.4227,
+      "step": 5798
+    },
+    {
+      "epoch": 0.426691193377788,
+      "grad_norm": 1.0546875,
+      "learning_rate": 3.081036637905075e-05,
+      "loss": 1.1151,
+      "step": 5799
+    },
+    {
+      "epoch": 0.426764773511152,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.080473322087193e-05,
+      "loss": 1.221,
+      "step": 5800
+    },
+    {
+      "epoch": 0.426838353644516,
+      "grad_norm": 0.75,
+      "learning_rate": 3.0799099751164036e-05,
+      "loss": 0.8428,
+      "step": 5801
+    },
+    {
+      "epoch": 0.42691193377787995,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.079346597022942e-05,
+      "loss": 0.8761,
+      "step": 5802
+    },
+    {
+      "epoch": 0.42698551391124395,
+      "grad_norm": 0.88671875,
+      "learning_rate": 3.078783187837042e-05,
+      "loss": 1.2068,
+      "step": 5803
+    },
+    {
+      "epoch": 0.42705909404460796,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.078219747588942e-05,
+      "loss": 0.7164,
+      "step": 5804
+    },
+    {
+      "epoch": 0.42713267417797196,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.077656276308879e-05,
+      "loss": 0.8161,
+      "step": 5805
+    },
+    {
+      "epoch": 0.42720625431133596,
+      "grad_norm": 0.9140625,
+      "learning_rate": 3.077092774027096e-05,
+      "loss": 1.4555,
+      "step": 5806
+    },
+    {
+      "epoch": 0.4272798344446999,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.076529240773833e-05,
+      "loss": 0.9777,
+      "step": 5807
+    },
+    {
+      "epoch": 0.4273534145780639,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.0759656765793356e-05,
+      "loss": 1.1901,
+      "step": 5808
+    },
+    {
+      "epoch": 0.4274269947114279,
+      "grad_norm": 0.9765625,
+      "learning_rate": 3.075402081473848e-05,
+      "loss": 1.0205,
+      "step": 5809
+    },
+    {
+      "epoch": 0.4275005748447919,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.074838455487616e-05,
+      "loss": 1.1187,
+      "step": 5810
+    },
+    {
+      "epoch": 0.4275741549781559,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.074274798650892e-05,
+      "loss": 0.909,
+      "step": 5811
+    },
+    {
+      "epoch": 0.4276477351115199,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.073711110993923e-05,
+      "loss": 0.9265,
+      "step": 5812
+    },
+    {
+      "epoch": 0.4277213152448839,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.0731473925469616e-05,
+      "loss": 0.868,
+      "step": 5813
+    },
+    {
+      "epoch": 0.4277948953782479,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.072583643340263e-05,
+      "loss": 1.1088,
+      "step": 5814
+    },
+    {
+      "epoch": 0.4278684755116119,
+      "grad_norm": 1.1484375,
+      "learning_rate": 3.072019863404082e-05,
+      "loss": 1.3737,
+      "step": 5815
+    },
+    {
+      "epoch": 0.42794205564497584,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.071456052768675e-05,
+      "loss": 1.4669,
+      "step": 5816
+    },
+    {
+      "epoch": 0.42801563577833984,
+      "grad_norm": 1.203125,
+      "learning_rate": 3.0708922114643e-05,
+      "loss": 1.3207,
+      "step": 5817
+    },
+    {
+      "epoch": 0.42808921591170385,
+      "grad_norm": 1.3046875,
+      "learning_rate": 3.07032833952122e-05,
+      "loss": 1.0223,
+      "step": 5818
+    },
+    {
+      "epoch": 0.42816279604506785,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.0697644369696945e-05,
+      "loss": 0.7622,
+      "step": 5819
+    },
+    {
+      "epoch": 0.4282363761784318,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.069200503839988e-05,
+      "loss": 0.7641,
+      "step": 5820
+    },
+    {
+      "epoch": 0.4283099563117958,
+      "grad_norm": 0.80078125,
+      "learning_rate": 3.068636540162364e-05,
+      "loss": 0.8504,
+      "step": 5821
+    },
+    {
+      "epoch": 0.4283835364451598,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.068072545967093e-05,
+      "loss": 1.165,
+      "step": 5822
+    },
+    {
+      "epoch": 0.4284571165785238,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.0675085212844403e-05,
+      "loss": 1.3198,
+      "step": 5823
+    },
+    {
+      "epoch": 0.4285306967118878,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.066944466144677e-05,
+      "loss": 0.7301,
+      "step": 5824
+    },
+    {
+      "epoch": 0.42860427684525176,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.066380380578075e-05,
+      "loss": 0.8298,
+      "step": 5825
+    },
+    {
+      "epoch": 0.42867785697861577,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.065816264614909e-05,
+      "loss": 0.9894,
+      "step": 5826
+    },
+    {
+      "epoch": 0.42875143711197977,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.065252118285451e-05,
+      "loss": 0.8816,
+      "step": 5827
+    },
+    {
+      "epoch": 0.4288250172453438,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.0646879416199805e-05,
+      "loss": 1.0246,
+      "step": 5828
+    },
+    {
+      "epoch": 0.4288985973787077,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.0641237346487747e-05,
+      "loss": 0.8419,
+      "step": 5829
+    },
+    {
+      "epoch": 0.42897217751207173,
+      "grad_norm": 0.9375,
+      "learning_rate": 3.063559497402113e-05,
+      "loss": 0.8907,
+      "step": 5830
+    },
+    {
+      "epoch": 0.42904575764543573,
+      "grad_norm": 1.0703125,
+      "learning_rate": 3.062995229910278e-05,
+      "loss": 0.8183,
+      "step": 5831
+    },
+    {
+      "epoch": 0.42911933777879974,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.062430932203552e-05,
+      "loss": 0.9873,
+      "step": 5832
+    },
+    {
+      "epoch": 0.42919291791216374,
+      "grad_norm": 0.89453125,
+      "learning_rate": 3.06186660431222e-05,
+      "loss": 1.1113,
+      "step": 5833
+    },
+    {
+      "epoch": 0.4292664980455277,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.061302246266569e-05,
+      "loss": 1.0394,
+      "step": 5834
+    },
+    {
+      "epoch": 0.4293400781788917,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.060737858096886e-05,
+      "loss": 0.9237,
+      "step": 5835
+    },
+    {
+      "epoch": 0.4294136583122557,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.0601734398334626e-05,
+      "loss": 0.8831,
+      "step": 5836
+    },
+    {
+      "epoch": 0.4294872384456197,
+      "grad_norm": 0.6953125,
+      "learning_rate": 3.0596089915065875e-05,
+      "loss": 0.5404,
+      "step": 5837
+    },
+    {
+      "epoch": 0.42956081857898365,
+      "grad_norm": 0.75,
+      "learning_rate": 3.059044513146555e-05,
+      "loss": 0.7114,
+      "step": 5838
+    },
+    {
+      "epoch": 0.42963439871234765,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.058480004783659e-05,
+      "loss": 0.7721,
+      "step": 5839
+    },
+    {
+      "epoch": 0.42970797884571166,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.0579154664481966e-05,
+      "loss": 0.8697,
+      "step": 5840
+    },
+    {
+      "epoch": 0.42978155897907566,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.057350898170464e-05,
+      "loss": 0.8042,
+      "step": 5841
+    },
+    {
+      "epoch": 0.42985513911243967,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.056786299980763e-05,
+      "loss": 1.1312,
+      "step": 5842
+    },
+    {
+      "epoch": 0.4299287192458036,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.0562216719093913e-05,
+      "loss": 1.0338,
+      "step": 5843
+    },
+    {
+      "epoch": 0.4300022993791676,
+      "grad_norm": 0.77734375,
+      "learning_rate": 3.055657013986654e-05,
+      "loss": 1.1038,
+      "step": 5844
+    },
+    {
+      "epoch": 0.4300758795125316,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.055092326242854e-05,
+      "loss": 0.9665,
+      "step": 5845
+    },
+    {
+      "epoch": 0.4301494596458956,
+      "grad_norm": 1.7578125,
+      "learning_rate": 3.054527608708298e-05,
+      "loss": 1.4164,
+      "step": 5846
+    },
+    {
+      "epoch": 0.4302230397792596,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.053962861413291e-05,
+      "loss": 1.3889,
+      "step": 5847
+    },
+    {
+      "epoch": 0.4302966199126236,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.053398084388144e-05,
+      "loss": 1.0581,
+      "step": 5848
+    },
+    {
+      "epoch": 0.4303702000459876,
+      "grad_norm": 0.82421875,
+      "learning_rate": 3.052833277663167e-05,
+      "loss": 0.7859,
+      "step": 5849
+    },
+    {
+      "epoch": 0.4304437801793516,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.052268441268673e-05,
+      "loss": 0.7511,
+      "step": 5850
+    },
+    {
+      "epoch": 0.4305173603127156,
+      "grad_norm": 1.1796875,
+      "learning_rate": 3.0517035752349743e-05,
+      "loss": 1.367,
+      "step": 5851
+    },
+    {
+      "epoch": 0.43059094044607954,
+      "grad_norm": 1.09375,
+      "learning_rate": 3.051138679592387e-05,
+      "loss": 0.9445,
+      "step": 5852
+    },
+    {
+      "epoch": 0.43066452057944354,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.0505737543712275e-05,
+      "loss": 1.543,
+      "step": 5853
+    },
+    {
+      "epoch": 0.43073810071280755,
+      "grad_norm": 0.6796875,
+      "learning_rate": 3.050008799601814e-05,
+      "loss": 0.5615,
+      "step": 5854
+    },
+    {
+      "epoch": 0.43081168084617155,
+      "grad_norm": 1.0,
+      "learning_rate": 3.0494438153144676e-05,
+      "loss": 0.6474,
+      "step": 5855
+    },
+    {
+      "epoch": 0.4308852609795355,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.0488788015395087e-05,
+      "loss": 0.8403,
+      "step": 5856
+    },
+    {
+      "epoch": 0.4309588411128995,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.0483137583072623e-05,
+      "loss": 0.8642,
+      "step": 5857
+    },
+    {
+      "epoch": 0.4310324212462635,
+      "grad_norm": 0.91015625,
+      "learning_rate": 3.0477486856480515e-05,
+      "loss": 0.9637,
+      "step": 5858
+    },
+    {
+      "epoch": 0.4311060013796275,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.0471835835922034e-05,
+      "loss": 0.6809,
+      "step": 5859
+    },
+    {
+      "epoch": 0.4311795815129915,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.046618452170046e-05,
+      "loss": 0.9921,
+      "step": 5860
+    },
+    {
+      "epoch": 0.43125316164635547,
+      "grad_norm": 0.9453125,
+      "learning_rate": 3.0460532914119083e-05,
+      "loss": 1.0586,
+      "step": 5861
+    },
+    {
+      "epoch": 0.43132674177971947,
+      "grad_norm": 1.078125,
+      "learning_rate": 3.0454881013481217e-05,
+      "loss": 1.4062,
+      "step": 5862
+    },
+    {
+      "epoch": 0.4314003219130835,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.0449228820090198e-05,
+      "loss": 1.2889,
+      "step": 5863
+    },
+    {
+      "epoch": 0.4314739020464475,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.0443576334249357e-05,
+      "loss": 1.1536,
+      "step": 5864
+    },
+    {
+      "epoch": 0.4315474821798114,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.043792355626205e-05,
+      "loss": 0.5844,
+      "step": 5865
+    },
+    {
+      "epoch": 0.43162106231317543,
+      "grad_norm": 0.98828125,
+      "learning_rate": 3.0432270486431663e-05,
+      "loss": 0.7336,
+      "step": 5866
+    },
+    {
+      "epoch": 0.43169464244653943,
+      "grad_norm": 1.2421875,
+      "learning_rate": 3.042661712506158e-05,
+      "loss": 1.3385,
+      "step": 5867
+    },
+    {
+      "epoch": 0.43176822257990344,
+      "grad_norm": 0.90625,
+      "learning_rate": 3.04209634724552e-05,
+      "loss": 0.7551,
+      "step": 5868
+    },
+    {
+      "epoch": 0.43184180271326744,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.0415309528915946e-05,
+      "loss": 0.8322,
+      "step": 5869
+    },
+    {
+      "epoch": 0.4319153828466314,
+      "grad_norm": 0.875,
+      "learning_rate": 3.040965529474727e-05,
+      "loss": 0.7568,
+      "step": 5870
+    },
+    {
+      "epoch": 0.4319889629799954,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.04040007702526e-05,
+      "loss": 0.9458,
+      "step": 5871
+    },
+    {
+      "epoch": 0.4320625431133594,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.039834595573542e-05,
+      "loss": 0.9895,
+      "step": 5872
+    },
+    {
+      "epoch": 0.4321361232467234,
+      "grad_norm": 0.953125,
+      "learning_rate": 3.039269085149921e-05,
+      "loss": 1.1591,
+      "step": 5873
+    },
+    {
+      "epoch": 0.43220970338008735,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.038703545784747e-05,
+      "loss": 0.8226,
+      "step": 5874
+    },
+    {
+      "epoch": 0.43228328351345136,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.03813797750837e-05,
+      "loss": 0.9298,
+      "step": 5875
+    },
+    {
+      "epoch": 0.43235686364681536,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.0375723803511447e-05,
+      "loss": 0.5958,
+      "step": 5876
+    },
+    {
+      "epoch": 0.43243044378017936,
+      "grad_norm": 0.765625,
+      "learning_rate": 3.0370067543434254e-05,
+      "loss": 0.8827,
+      "step": 5877
+    },
+    {
+      "epoch": 0.43250402391354337,
+      "grad_norm": 1.03125,
+      "learning_rate": 3.036441099515568e-05,
+      "loss": 1.133,
+      "step": 5878
+    },
+    {
+      "epoch": 0.4325776040469073,
+      "grad_norm": 0.859375,
+      "learning_rate": 3.0358754158979298e-05,
+      "loss": 0.796,
+      "step": 5879
+    },
+    {
+      "epoch": 0.4326511841802713,
+      "grad_norm": 0.8359375,
+      "learning_rate": 3.03530970352087e-05,
+      "loss": 0.7298,
+      "step": 5880
+    },
+    {
+      "epoch": 0.4327247643136353,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.0347439624147493e-05,
+      "loss": 0.6449,
+      "step": 5881
+    },
+    {
+      "epoch": 0.43279834444699933,
+      "grad_norm": 0.8125,
+      "learning_rate": 3.0341781926099305e-05,
+      "loss": 0.9,
+      "step": 5882
+    },
+    {
+      "epoch": 0.4328719245803633,
+      "grad_norm": 0.8828125,
+      "learning_rate": 3.0336123941367768e-05,
+      "loss": 1.0585,
+      "step": 5883
+    },
+    {
+      "epoch": 0.4329455047137273,
+      "grad_norm": 0.78125,
+      "learning_rate": 3.0330465670256542e-05,
+      "loss": 0.7246,
+      "step": 5884
+    },
+    {
+      "epoch": 0.4330190848470913,
+      "grad_norm": 0.87109375,
+      "learning_rate": 3.032480711306928e-05,
+      "loss": 1.0527,
+      "step": 5885
+    },
+    {
+      "epoch": 0.4330926649804553,
+      "grad_norm": 1.015625,
+      "learning_rate": 3.0319148270109693e-05,
+      "loss": 1.1721,
+      "step": 5886
+    },
+    {
+      "epoch": 0.4331662451138193,
+      "grad_norm": 1.1953125,
+      "learning_rate": 3.0313489141681456e-05,
+      "loss": 1.5106,
+      "step": 5887
+    },
+    {
+      "epoch": 0.43323982524718324,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.030782972808829e-05,
+      "loss": 0.7848,
+      "step": 5888
+    },
+    {
+      "epoch": 0.43331340538054725,
+      "grad_norm": 0.90234375,
+      "learning_rate": 3.030217002963393e-05,
+      "loss": 1.0025,
+      "step": 5889
+    },
+    {
+      "epoch": 0.43338698551391125,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.0296510046622118e-05,
+      "loss": 0.6883,
+      "step": 5890
+    },
+    {
+      "epoch": 0.43346056564727525,
+      "grad_norm": 0.8984375,
+      "learning_rate": 3.0290849779356623e-05,
+      "loss": 1.1494,
+      "step": 5891
+    },
+    {
+      "epoch": 0.4335341457806392,
+      "grad_norm": 0.73828125,
+      "learning_rate": 3.0285189228141202e-05,
+      "loss": 0.8172,
+      "step": 5892
+    },
+    {
+      "epoch": 0.4336077259140032,
+      "grad_norm": 0.734375,
+      "learning_rate": 3.0279528393279664e-05,
+      "loss": 0.7093,
+      "step": 5893
+    },
+    {
+      "epoch": 0.4336813060473672,
+      "grad_norm": 0.6171875,
+      "learning_rate": 3.0273867275075808e-05,
+      "loss": 0.5568,
+      "step": 5894
+    },
+    {
+      "epoch": 0.4337548861807312,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.0268205873833456e-05,
+      "loss": 0.8622,
+      "step": 5895
+    },
+    {
+      "epoch": 0.4338284663140952,
+      "grad_norm": 0.8515625,
+      "learning_rate": 3.0262544189856446e-05,
+      "loss": 1.0844,
+      "step": 5896
+    },
+    {
+      "epoch": 0.43390204644745917,
+      "grad_norm": 1.046875,
+      "learning_rate": 3.0256882223448625e-05,
+      "loss": 1.2497,
+      "step": 5897
+    },
+    {
+      "epoch": 0.43397562658082317,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.025121997491387e-05,
+      "loss": 0.7019,
+      "step": 5898
+    },
+    {
+      "epoch": 0.4340492067141872,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.024555744455605e-05,
+      "loss": 1.0185,
+      "step": 5899
+    },
+    {
+      "epoch": 0.4341227868475512,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.0239894632679078e-05,
+      "loss": 0.9628,
+      "step": 5900
+    },
+    {
+      "epoch": 0.4341963669809151,
+      "grad_norm": 0.80859375,
+      "learning_rate": 3.023423153958685e-05,
+      "loss": 0.612,
+      "step": 5901
+    },
+    {
+      "epoch": 0.43426994711427913,
+      "grad_norm": 0.81640625,
+      "learning_rate": 3.0228568165583303e-05,
+      "loss": 0.999,
+      "step": 5902
+    },
+    {
+      "epoch": 0.43434352724764314,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.0222904510972377e-05,
+      "loss": 1.065,
+      "step": 5903
+    },
+    {
+      "epoch": 0.43441710738100714,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.0217240576058033e-05,
+      "loss": 0.9642,
+      "step": 5904
+    },
+    {
+      "epoch": 0.43449068751437114,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.0211576361144245e-05,
+      "loss": 1.1554,
+      "step": 5905
+    },
+    {
+      "epoch": 0.4345642676477351,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.020591186653499e-05,
+      "loss": 1.2682,
+      "step": 5906
+    },
+    {
+      "epoch": 0.4346378477810991,
+      "grad_norm": 0.921875,
+      "learning_rate": 3.0200247092534285e-05,
+      "loss": 0.7066,
+      "step": 5907
+    },
+    {
+      "epoch": 0.4347114279144631,
+      "grad_norm": 0.671875,
+      "learning_rate": 3.0194582039446136e-05,
+      "loss": 0.7009,
+      "step": 5908
+    },
+    {
+      "epoch": 0.4347850080478271,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.018891670757458e-05,
+      "loss": 0.9667,
+      "step": 5909
+    },
+    {
+      "epoch": 0.43485858818119105,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.0183251097223664e-05,
+      "loss": 0.8592,
+      "step": 5910
+    },
+    {
+      "epoch": 0.43493216831455506,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.0177585208697456e-05,
+      "loss": 0.9382,
+      "step": 5911
+    },
+    {
+      "epoch": 0.43500574844791906,
+      "grad_norm": 0.796875,
+      "learning_rate": 3.0171919042300023e-05,
+      "loss": 1.0079,
+      "step": 5912
+    },
+    {
+      "epoch": 0.43507932858128306,
+      "grad_norm": 0.7265625,
+      "learning_rate": 3.0166252598335466e-05,
+      "loss": 0.8211,
+      "step": 5913
+    },
+    {
+      "epoch": 0.43515290871464707,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.016058587710789e-05,
+      "loss": 1.1583,
+      "step": 5914
+    },
+    {
+      "epoch": 0.435226488848011,
+      "grad_norm": 0.984375,
+      "learning_rate": 3.015491887892141e-05,
+      "loss": 1.0118,
+      "step": 5915
+    },
+    {
+      "epoch": 0.435300068981375,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.014925160408018e-05,
+      "loss": 1.1678,
+      "step": 5916
+    },
+    {
+      "epoch": 0.435373649114739,
+      "grad_norm": 0.98046875,
+      "learning_rate": 3.0143584052888334e-05,
+      "loss": 0.9329,
+      "step": 5917
+    },
+    {
+      "epoch": 0.43544722924810303,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.0137916225650054e-05,
+      "loss": 0.5739,
+      "step": 5918
+    },
+    {
+      "epoch": 0.435520809381467,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.0132248122669514e-05,
+      "loss": 0.8449,
+      "step": 5919
+    },
+    {
+      "epoch": 0.435594389514831,
+      "grad_norm": 1.0,
+      "learning_rate": 3.012657974425091e-05,
+      "loss": 0.9105,
+      "step": 5920
+    },
+    {
+      "epoch": 0.435667969648195,
+      "grad_norm": 0.875,
+      "learning_rate": 3.012091109069845e-05,
+      "loss": 0.821,
+      "step": 5921
+    },
+    {
+      "epoch": 0.435741549781559,
+      "grad_norm": 1.2890625,
+      "learning_rate": 3.0115242162316365e-05,
+      "loss": 1.7447,
+      "step": 5922
+    },
+    {
+      "epoch": 0.435815129914923,
+      "grad_norm": 0.92578125,
+      "learning_rate": 3.0109572959408894e-05,
+      "loss": 1.1237,
+      "step": 5923
+    },
+    {
+      "epoch": 0.43588871004828694,
+      "grad_norm": 0.8203125,
+      "learning_rate": 3.0103903482280292e-05,
+      "loss": 1.156,
+      "step": 5924
+    },
+    {
+      "epoch": 0.43596229018165095,
+      "grad_norm": 0.875,
+      "learning_rate": 3.009823373123484e-05,
+      "loss": 0.682,
+      "step": 5925
+    },
+    {
+      "epoch": 0.43603587031501495,
+      "grad_norm": 0.85546875,
+      "learning_rate": 3.0092563706576798e-05,
+      "loss": 0.8481,
+      "step": 5926
+    },
+    {
+      "epoch": 0.43610945044837895,
+      "grad_norm": 0.828125,
+      "learning_rate": 3.0086893408610495e-05,
+      "loss": 0.7081,
+      "step": 5927
+    },
+    {
+      "epoch": 0.4361830305817429,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.0081222837640215e-05,
+      "loss": 1.0804,
+      "step": 5928
+    },
+    {
+      "epoch": 0.4362566107151069,
+      "grad_norm": 0.890625,
+      "learning_rate": 3.0075551993970314e-05,
+      "loss": 0.9245,
+      "step": 5929
+    },
+    {
+      "epoch": 0.4363301908484709,
+      "grad_norm": 0.75,
+      "learning_rate": 3.0069880877905116e-05,
+      "loss": 0.5928,
+      "step": 5930
+    },
+    {
+      "epoch": 0.4364037709818349,
+      "grad_norm": 1.0859375,
+      "learning_rate": 3.006420948974899e-05,
+      "loss": 0.9989,
+      "step": 5931
+    },
+    {
+      "epoch": 0.4364773511151989,
+      "grad_norm": 1.1640625,
+      "learning_rate": 3.005853782980631e-05,
+      "loss": 0.9612,
+      "step": 5932
+    },
+    {
+      "epoch": 0.43655093124856287,
+      "grad_norm": 0.87890625,
+      "learning_rate": 3.0052865898381456e-05,
+      "loss": 1.058,
+      "step": 5933
+    },
+    {
+      "epoch": 0.43662451138192687,
+      "grad_norm": 0.9296875,
+      "learning_rate": 3.0047193695778836e-05,
+      "loss": 1.1295,
+      "step": 5934
+    },
+    {
+      "epoch": 0.4366980915152909,
+      "grad_norm": 0.86328125,
+      "learning_rate": 3.0041521222302853e-05,
+      "loss": 0.966,
+      "step": 5935
+    },
+    {
+      "epoch": 0.4367716716486549,
+      "grad_norm": 0.83203125,
+      "learning_rate": 3.0035848478257956e-05,
+      "loss": 0.882,
+      "step": 5936
+    },
+    {
+      "epoch": 0.43684525178201883,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.003017546394858e-05,
+      "loss": 0.7737,
+      "step": 5937
+    },
+    {
+      "epoch": 0.43691883191538283,
+      "grad_norm": 1.0390625,
+      "learning_rate": 3.0024502179679187e-05,
+      "loss": 1.3886,
+      "step": 5938
+    },
+    {
+      "epoch": 0.43699241204874684,
+      "grad_norm": 0.93359375,
+      "learning_rate": 3.0018828625754258e-05,
+      "loss": 1.2207,
+      "step": 5939
+    },
+    {
+      "epoch": 0.43706599218211084,
+      "grad_norm": 0.875,
+      "learning_rate": 3.0013154802478266e-05,
+      "loss": 0.8199,
+      "step": 5940
+    },
+    {
+      "epoch": 0.43713957231547484,
+      "grad_norm": 0.78515625,
+      "learning_rate": 3.0007480710155732e-05,
+      "loss": 0.7809,
+      "step": 5941
+    },
+    {
+      "epoch": 0.4372131524488388,
+      "grad_norm": 0.84765625,
+      "learning_rate": 3.0001806349091166e-05,
+      "loss": 0.9857,
+      "step": 5942
+    },
+    {
+      "epoch": 0.4372867325822028,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.9996131719589092e-05,
+      "loss": 1.1154,
+      "step": 5943
+    },
+    {
+      "epoch": 0.4373603127155668,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.999045682195407e-05,
+      "loss": 1.4869,
+      "step": 5944
+    },
+    {
+      "epoch": 0.4374338928489308,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.9984781656490656e-05,
+      "loss": 0.6608,
+      "step": 5945
+    },
+    {
+      "epoch": 0.43750747298229475,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.9979106223503435e-05,
+      "loss": 0.9823,
+      "step": 5946
+    },
+    {
+      "epoch": 0.43758105311565876,
+      "grad_norm": 1.25,
+      "learning_rate": 2.9973430523296974e-05,
+      "loss": 1.1349,
+      "step": 5947
+    },
+    {
+      "epoch": 0.43765463324902276,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.9967754556175896e-05,
+      "loss": 1.3051,
+      "step": 5948
+    },
+    {
+      "epoch": 0.43772821338238677,
+      "grad_norm": 1.0,
+      "learning_rate": 2.9962078322444815e-05,
+      "loss": 1.2846,
+      "step": 5949
+    },
+    {
+      "epoch": 0.43780179351575077,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.9956401822408357e-05,
+      "loss": 0.8564,
+      "step": 5950
+    },
+    {
+      "epoch": 0.4378753736491147,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.9950725056371177e-05,
+      "loss": 1.1169,
+      "step": 5951
+    },
+    {
+      "epoch": 0.4379489537824787,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.9945048024637935e-05,
+      "loss": 0.9506,
+      "step": 5952
+    },
+    {
+      "epoch": 0.4380225339158427,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.9939370727513306e-05,
+      "loss": 0.671,
+      "step": 5953
+    },
+    {
+      "epoch": 0.43809611404920673,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.9933693165301975e-05,
+      "loss": 1.1273,
+      "step": 5954
+    },
+    {
+      "epoch": 0.4381696941825707,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.9928015338308655e-05,
+      "loss": 0.9897,
+      "step": 5955
+    },
+    {
+      "epoch": 0.4382432743159347,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.992233724683805e-05,
+      "loss": 0.7975,
+      "step": 5956
+    },
+    {
+      "epoch": 0.4383168544492987,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.9916658891194917e-05,
+      "loss": 0.6962,
+      "step": 5957
+    },
+    {
+      "epoch": 0.4383904345826627,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.9910980271683975e-05,
+      "loss": 0.7968,
+      "step": 5958
+    },
+    {
+      "epoch": 0.4384640147160267,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.9905301388610008e-05,
+      "loss": 1.1579,
+      "step": 5959
+    },
+    {
+      "epoch": 0.43853759484939064,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.9899622242277773e-05,
+      "loss": 0.7132,
+      "step": 5960
+    },
+    {
+      "epoch": 0.43861117498275465,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.989394283299207e-05,
+      "loss": 0.9599,
+      "step": 5961
+    },
+    {
+      "epoch": 0.43868475511611865,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.9888263161057696e-05,
+      "loss": 0.7488,
+      "step": 5962
+    },
+    {
+      "epoch": 0.43875833524948266,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.9882583226779466e-05,
+      "loss": 0.7828,
+      "step": 5963
+    },
+    {
+      "epoch": 0.4388319153828466,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.9876903030462222e-05,
+      "loss": 0.9261,
+      "step": 5964
+    },
+    {
+      "epoch": 0.4389054955162106,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.9871222572410805e-05,
+      "loss": 0.8842,
+      "step": 5965
+    },
+    {
+      "epoch": 0.4389790756495746,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.9865541852930075e-05,
+      "loss": 1.0741,
+      "step": 5966
+    },
+    {
+      "epoch": 0.4390526557829386,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.9859860872324896e-05,
+      "loss": 0.8826,
+      "step": 5967
+    },
+    {
+      "epoch": 0.4391262359163026,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.985417963090017e-05,
+      "loss": 0.7303,
+      "step": 5968
+    },
+    {
+      "epoch": 0.43919981604966657,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.9848498128960785e-05,
+      "loss": 0.7794,
+      "step": 5969
+    },
+    {
+      "epoch": 0.4392733961830306,
+      "grad_norm": 1.2109375,
+      "learning_rate": 2.984281636681166e-05,
+      "loss": 1.0307,
+      "step": 5970
+    },
+    {
+      "epoch": 0.4393469763163946,
+      "grad_norm": 1.140625,
+      "learning_rate": 2.983713434475774e-05,
+      "loss": 1.1013,
+      "step": 5971
+    },
+    {
+      "epoch": 0.4394205564497586,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.9831452063103944e-05,
+      "loss": 1.058,
+      "step": 5972
+    },
+    {
+      "epoch": 0.43949413658312253,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.982576952215525e-05,
+      "loss": 0.942,
+      "step": 5973
+    },
+    {
+      "epoch": 0.43956771671648653,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.9820086722216616e-05,
+      "loss": 0.873,
+      "step": 5974
+    },
+    {
+      "epoch": 0.43964129684985054,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.9814403663593036e-05,
+      "loss": 0.8835,
+      "step": 5975
+    },
+    {
+      "epoch": 0.43971487698321454,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.9808720346589493e-05,
+      "loss": 0.5536,
+      "step": 5976
+    },
+    {
+      "epoch": 0.43978845711657855,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.9803036771511016e-05,
+      "loss": 0.9971,
+      "step": 5977
+    },
+    {
+      "epoch": 0.4398620372499425,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.9797352938662633e-05,
+      "loss": 0.7169,
+      "step": 5978
+    },
+    {
+      "epoch": 0.4399356173833065,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.9791668848349376e-05,
+      "loss": 0.7012,
+      "step": 5979
+    },
+    {
+      "epoch": 0.4400091975166705,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.9785984500876303e-05,
+      "loss": 0.7088,
+      "step": 5980
+    },
+    {
+      "epoch": 0.4400827776500345,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.978029989654848e-05,
+      "loss": 0.962,
+      "step": 5981
+    },
+    {
+      "epoch": 0.44015635778339846,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.9774615035670996e-05,
+      "loss": 1.1626,
+      "step": 5982
+    },
+    {
+      "epoch": 0.44022993791676246,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.9768929918548926e-05,
+      "loss": 0.7757,
+      "step": 5983
+    },
+    {
+      "epoch": 0.44030351805012646,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.9763244545487406e-05,
+      "loss": 0.5121,
+      "step": 5984
+    },
+    {
+      "epoch": 0.44037709818349047,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.9757558916791545e-05,
+      "loss": 0.9457,
+      "step": 5985
+    },
+    {
+      "epoch": 0.44045067831685447,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.9751873032766482e-05,
+      "loss": 0.961,
+      "step": 5986
+    },
+    {
+      "epoch": 0.4405242584502184,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.974618689371737e-05,
+      "loss": 0.9199,
+      "step": 5987
+    },
+    {
+      "epoch": 0.4405978385835824,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.974050049994937e-05,
+      "loss": 0.6986,
+      "step": 5988
+    },
+    {
+      "epoch": 0.44067141871694643,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.9734813851767662e-05,
+      "loss": 0.6249,
+      "step": 5989
+    },
+    {
+      "epoch": 0.44074499885031043,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.9729126949477436e-05,
+      "loss": 1.158,
+      "step": 5990
+    },
+    {
+      "epoch": 0.4408185789836744,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.9723439793383907e-05,
+      "loss": 0.6793,
+      "step": 5991
+    },
+    {
+      "epoch": 0.4408921591170384,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.9717752383792276e-05,
+      "loss": 0.9961,
+      "step": 5992
+    },
+    {
+      "epoch": 0.4409657392504024,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.971206472100779e-05,
+      "loss": 0.7426,
+      "step": 5993
+    },
+    {
+      "epoch": 0.4410393193837664,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.970637680533569e-05,
+      "loss": 0.8417,
+      "step": 5994
+    },
+    {
+      "epoch": 0.4411128995171304,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.970068863708123e-05,
+      "loss": 1.532,
+      "step": 5995
+    },
+    {
+      "epoch": 0.44118647965049435,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.9695000216549695e-05,
+      "loss": 1.146,
+      "step": 5996
+    },
+    {
+      "epoch": 0.44126005978385835,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.9689311544046362e-05,
+      "loss": 0.5832,
+      "step": 5997
+    },
+    {
+      "epoch": 0.44133363991722235,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.9683622619876544e-05,
+      "loss": 0.8079,
+      "step": 5998
+    },
+    {
+      "epoch": 0.44140722005058636,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.9677933444345535e-05,
+      "loss": 0.5841,
+      "step": 5999
+    },
+    {
+      "epoch": 0.4414808001839503,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.9672244017758683e-05,
+      "loss": 0.9227,
+      "step": 6000
+    },
+    {
+      "epoch": 0.4415543803173143,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.9666554340421314e-05,
+      "loss": 0.7703,
+      "step": 6001
+    },
+    {
+      "epoch": 0.4416279604506783,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.966086441263879e-05,
+      "loss": 0.6359,
+      "step": 6002
+    },
+    {
+      "epoch": 0.4417015405840423,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.965517423471647e-05,
+      "loss": 0.5752,
+      "step": 6003
+    },
+    {
+      "epoch": 0.4417751207174063,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.9649483806959742e-05,
+      "loss": 0.8808,
+      "step": 6004
+    },
+    {
+      "epoch": 0.44184870085077027,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.964379312967401e-05,
+      "loss": 0.8262,
+      "step": 6005
+    },
+    {
+      "epoch": 0.4419222809841343,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.9638102203164663e-05,
+      "loss": 1.0234,
+      "step": 6006
+    },
+    {
+      "epoch": 0.4419958611174983,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.9632411027737135e-05,
+      "loss": 1.6336,
+      "step": 6007
+    },
+    {
+      "epoch": 0.4420694412508623,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.9626719603696852e-05,
+      "loss": 1.2063,
+      "step": 6008
+    },
+    {
+      "epoch": 0.44214302138422623,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.9621027931349267e-05,
+      "loss": 1.0634,
+      "step": 6009
+    },
+    {
+      "epoch": 0.44221660151759024,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.9615336010999844e-05,
+      "loss": 1.0263,
+      "step": 6010
+    },
+    {
+      "epoch": 0.44229018165095424,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.960964384295405e-05,
+      "loss": 1.041,
+      "step": 6011
+    },
+    {
+      "epoch": 0.44236376178431824,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.9603951427517385e-05,
+      "loss": 0.983,
+      "step": 6012
+    },
+    {
+      "epoch": 0.44243734191768225,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.9598258764995336e-05,
+      "loss": 0.6677,
+      "step": 6013
+    },
+    {
+      "epoch": 0.4425109220510462,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.959256585569343e-05,
+      "loss": 1.0042,
+      "step": 6014
+    },
+    {
+      "epoch": 0.4425845021844102,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.9586872699917178e-05,
+      "loss": 0.783,
+      "step": 6015
+    },
+    {
+      "epoch": 0.4426580823177742,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.958117929797214e-05,
+      "loss": 0.7919,
+      "step": 6016
+    },
+    {
+      "epoch": 0.4427316624511382,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.957548565016386e-05,
+      "loss": 1.0166,
+      "step": 6017
+    },
+    {
+      "epoch": 0.44280524258450216,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.9569791756797905e-05,
+      "loss": 0.9252,
+      "step": 6018
+    },
+    {
+      "epoch": 0.44287882271786616,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.956409761817986e-05,
+      "loss": 1.3885,
+      "step": 6019
+    },
+    {
+      "epoch": 0.44295240285123016,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.955840323461532e-05,
+      "loss": 1.5966,
+      "step": 6020
+    },
+    {
+      "epoch": 0.44302598298459417,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.9552708606409885e-05,
+      "loss": 1.6081,
+      "step": 6021
+    },
+    {
+      "epoch": 0.4430995631179582,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.9547013733869172e-05,
+      "loss": 0.6657,
+      "step": 6022
+    },
+    {
+      "epoch": 0.4431731432513221,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.9541318617298825e-05,
+      "loss": 0.9565,
+      "step": 6023
+    },
+    {
+      "epoch": 0.4432467233846861,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.953562325700448e-05,
+      "loss": 1.0535,
+      "step": 6024
+    },
+    {
+      "epoch": 0.44332030351805013,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.9529927653291806e-05,
+      "loss": 0.9568,
+      "step": 6025
+    },
+    {
+      "epoch": 0.44339388365141413,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.9524231806466475e-05,
+      "loss": 1.438,
+      "step": 6026
+    },
+    {
+      "epoch": 0.44346746378477814,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.9518535716834162e-05,
+      "loss": 0.9918,
+      "step": 6027
+    },
+    {
+      "epoch": 0.4435410439181421,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.9512839384700575e-05,
+      "loss": 0.8983,
+      "step": 6028
+    },
+    {
+      "epoch": 0.4436146240515061,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.950714281037142e-05,
+      "loss": 1.0855,
+      "step": 6029
+    },
+    {
+      "epoch": 0.4436882041848701,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.950144599415242e-05,
+      "loss": 1.0439,
+      "step": 6030
+    },
+    {
+      "epoch": 0.4437617843182341,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.9495748936349316e-05,
+      "loss": 1.113,
+      "step": 6031
+    },
+    {
+      "epoch": 0.44383536445159805,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.9490051637267864e-05,
+      "loss": 1.3181,
+      "step": 6032
+    },
+    {
+      "epoch": 0.44390894458496205,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.9484354097213818e-05,
+      "loss": 0.8374,
+      "step": 6033
+    },
+    {
+      "epoch": 0.44398252471832605,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.9478656316492952e-05,
+      "loss": 0.9774,
+      "step": 6034
+    },
+    {
+      "epoch": 0.44405610485169006,
+      "grad_norm": 0.75,
+      "learning_rate": 2.9472958295411067e-05,
+      "loss": 0.9186,
+      "step": 6035
+    },
+    {
+      "epoch": 0.44412968498505406,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.9467260034273957e-05,
+      "loss": 0.7875,
+      "step": 6036
+    },
+    {
+      "epoch": 0.444203265118418,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.9461561533387437e-05,
+      "loss": 1.146,
+      "step": 6037
+    },
+    {
+      "epoch": 0.444276845251782,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.9455862793057338e-05,
+      "loss": 1.0432,
+      "step": 6038
+    },
+    {
+      "epoch": 0.444350425385146,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.9450163813589498e-05,
+      "loss": 0.966,
+      "step": 6039
+    },
+    {
+      "epoch": 0.44442400551851,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.9444464595289772e-05,
+      "loss": 0.8573,
+      "step": 6040
+    },
+    {
+      "epoch": 0.44449758565187397,
+      "grad_norm": 1.1328125,
+      "learning_rate": 2.943876513846403e-05,
+      "loss": 1.5034,
+      "step": 6041
+    },
+    {
+      "epoch": 0.444571165785238,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.9433065443418138e-05,
+      "loss": 0.7055,
+      "step": 6042
+    },
+    {
+      "epoch": 0.444644745918602,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.9427365510458e-05,
+      "loss": 0.6595,
+      "step": 6043
+    },
+    {
+      "epoch": 0.444718326051966,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.9421665339889515e-05,
+      "loss": 0.8481,
+      "step": 6044
+    },
+    {
+      "epoch": 0.44479190618533,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.94159649320186e-05,
+      "loss": 0.5326,
+      "step": 6045
+    },
+    {
+      "epoch": 0.44486548631869394,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.94102642871512e-05,
+      "loss": 0.8251,
+      "step": 6046
+    },
+    {
+      "epoch": 0.44493906645205794,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.9404563405593235e-05,
+      "loss": 0.6753,
+      "step": 6047
+    },
+    {
+      "epoch": 0.44501264658542194,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.939886228765068e-05,
+      "loss": 0.8832,
+      "step": 6048
+    },
+    {
+      "epoch": 0.44508622671878595,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.939316093362948e-05,
+      "loss": 0.9699,
+      "step": 6049
+    },
+    {
+      "epoch": 0.4451598068521499,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.9387459343835644e-05,
+      "loss": 0.6979,
+      "step": 6050
+    },
+    {
+      "epoch": 0.4452333869855139,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.9381757518575142e-05,
+      "loss": 0.8925,
+      "step": 6051
+    },
+    {
+      "epoch": 0.4453069671188779,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.9376055458153994e-05,
+      "loss": 0.821,
+      "step": 6052
+    },
+    {
+      "epoch": 0.4453805472522419,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.9370353162878216e-05,
+      "loss": 0.8584,
+      "step": 6053
+    },
+    {
+      "epoch": 0.4454541273856059,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.9364650633053838e-05,
+      "loss": 1.0352,
+      "step": 6054
+    },
+    {
+      "epoch": 0.44552770751896986,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.935894786898691e-05,
+      "loss": 0.9812,
+      "step": 6055
+    },
+    {
+      "epoch": 0.44560128765233387,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.935324487098347e-05,
+      "loss": 0.9025,
+      "step": 6056
+    },
+    {
+      "epoch": 0.44567486778569787,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.934754163934961e-05,
+      "loss": 1.0327,
+      "step": 6057
+    },
+    {
+      "epoch": 0.4457484479190619,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.934183817439139e-05,
+      "loss": 0.8906,
+      "step": 6058
+    },
+    {
+      "epoch": 0.4458220280524258,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.9336134476414923e-05,
+      "loss": 0.7753,
+      "step": 6059
+    },
+    {
+      "epoch": 0.4458956081857898,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.933043054572631e-05,
+      "loss": 0.884,
+      "step": 6060
+    },
+    {
+      "epoch": 0.44596918831915383,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.932472638263167e-05,
+      "loss": 0.8998,
+      "step": 6061
+    },
+    {
+      "epoch": 0.44604276845251783,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.9319021987437134e-05,
+      "loss": 1.1624,
+      "step": 6062
+    },
+    {
+      "epoch": 0.44611634858588184,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.931331736044884e-05,
+      "loss": 1.1614,
+      "step": 6063
+    },
+    {
+      "epoch": 0.4461899287192458,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.9307612501972948e-05,
+      "loss": 0.7683,
+      "step": 6064
+    },
+    {
+      "epoch": 0.4462635088526098,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.9301907412315632e-05,
+      "loss": 1.1088,
+      "step": 6065
+    },
+    {
+      "epoch": 0.4463370889859738,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.9296202091783072e-05,
+      "loss": 0.9251,
+      "step": 6066
+    },
+    {
+      "epoch": 0.4464106691193378,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.929049654068146e-05,
+      "loss": 0.7168,
+      "step": 6067
+    },
+    {
+      "epoch": 0.44648424925270175,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.9284790759317004e-05,
+      "loss": 1.1251,
+      "step": 6068
+    },
+    {
+      "epoch": 0.44655782938606575,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.9279084747995922e-05,
+      "loss": 1.3019,
+      "step": 6069
+    },
+    {
+      "epoch": 0.44663140951942976,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.9273378507024435e-05,
+      "loss": 0.6461,
+      "step": 6070
+    },
+    {
+      "epoch": 0.44670498965279376,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.92676720367088e-05,
+      "loss": 1.0662,
+      "step": 6071
+    },
+    {
+      "epoch": 0.44677856978615776,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.926196533735527e-05,
+      "loss": 1.0884,
+      "step": 6072
+    },
+    {
+      "epoch": 0.4468521499195217,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.925625840927011e-05,
+      "loss": 1.2211,
+      "step": 6073
+    },
+    {
+      "epoch": 0.4469257300528857,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.9250551252759595e-05,
+      "loss": 1.1065,
+      "step": 6074
+    },
+    {
+      "epoch": 0.4469993101862497,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.9244843868130023e-05,
+      "loss": 0.94,
+      "step": 6075
+    },
+    {
+      "epoch": 0.4470728903196137,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.9239136255687698e-05,
+      "loss": 0.7705,
+      "step": 6076
+    },
+    {
+      "epoch": 0.4471464704529777,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.923342841573894e-05,
+      "loss": 1.1884,
+      "step": 6077
+    },
+    {
+      "epoch": 0.4472200505863417,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.922772034859007e-05,
+      "loss": 0.9155,
+      "step": 6078
+    },
+    {
+      "epoch": 0.4472936307197057,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.9222012054547436e-05,
+      "loss": 0.9214,
+      "step": 6079
+    },
+    {
+      "epoch": 0.4473672108530697,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.9216303533917394e-05,
+      "loss": 1.0133,
+      "step": 6080
+    },
+    {
+      "epoch": 0.4474407909864337,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.9210594787006302e-05,
+      "loss": 0.9676,
+      "step": 6081
+    },
+    {
+      "epoch": 0.44751437111979764,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.9204885814120535e-05,
+      "loss": 1.2342,
+      "step": 6082
+    },
+    {
+      "epoch": 0.44758795125316164,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.9199176615566498e-05,
+      "loss": 1.4221,
+      "step": 6083
+    },
+    {
+      "epoch": 0.44766153138652565,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.9193467191650577e-05,
+      "loss": 0.6799,
+      "step": 6084
+    },
+    {
+      "epoch": 0.44773511151988965,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.918775754267919e-05,
+      "loss": 0.8441,
+      "step": 6085
+    },
+    {
+      "epoch": 0.4478086916532536,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.9182047668958766e-05,
+      "loss": 0.7616,
+      "step": 6086
+    },
+    {
+      "epoch": 0.4478822717866176,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.9176337570795752e-05,
+      "loss": 0.8768,
+      "step": 6087
+    },
+    {
+      "epoch": 0.4479558519199816,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.917062724849658e-05,
+      "loss": 0.6916,
+      "step": 6088
+    },
+    {
+      "epoch": 0.4480294320533456,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.9164916702367722e-05,
+      "loss": 0.6812,
+      "step": 6089
+    },
+    {
+      "epoch": 0.4481030121867096,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.915920593271565e-05,
+      "loss": 1.0242,
+      "step": 6090
+    },
+    {
+      "epoch": 0.44817659232007356,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.9153494939846855e-05,
+      "loss": 0.8894,
+      "step": 6091
+    },
+    {
+      "epoch": 0.44825017245343757,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.914778372406782e-05,
+      "loss": 1.1701,
+      "step": 6092
+    },
+    {
+      "epoch": 0.44832375258680157,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.914207228568508e-05,
+      "loss": 1.1768,
+      "step": 6093
+    },
+    {
+      "epoch": 0.4483973327201656,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.9136360625005142e-05,
+      "loss": 0.9539,
+      "step": 6094
+    },
+    {
+      "epoch": 0.4484709128535295,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.913064874233454e-05,
+      "loss": 1.1154,
+      "step": 6095
+    },
+    {
+      "epoch": 0.44854449298689353,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.912493663797982e-05,
+      "loss": 1.4624,
+      "step": 6096
+    },
+    {
+      "epoch": 0.44861807312025753,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.911922431224754e-05,
+      "loss": 1.0469,
+      "step": 6097
+    },
+    {
+      "epoch": 0.44869165325362154,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.9113511765444272e-05,
+      "loss": 0.5635,
+      "step": 6098
+    },
+    {
+      "epoch": 0.44876523338698554,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.91077989978766e-05,
+      "loss": 1.336,
+      "step": 6099
+    },
+    {
+      "epoch": 0.4488388135203495,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.910208600985111e-05,
+      "loss": 1.1571,
+      "step": 6100
+    },
+    {
+      "epoch": 0.4489123936537135,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.9096372801674416e-05,
+      "loss": 0.7791,
+      "step": 6101
+    },
+    {
+      "epoch": 0.4489859737870775,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.909065937365313e-05,
+      "loss": 0.8365,
+      "step": 6102
+    },
+    {
+      "epoch": 0.4490595539204415,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.908494572609388e-05,
+      "loss": 0.9837,
+      "step": 6103
+    },
+    {
+      "epoch": 0.44913313405380545,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.9079231859303306e-05,
+      "loss": 1.0402,
+      "step": 6104
+    },
+    {
+      "epoch": 0.44920671418716945,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.9073517773588067e-05,
+      "loss": 1.1519,
+      "step": 6105
+    },
+    {
+      "epoch": 0.44928029432053346,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.9067803469254818e-05,
+      "loss": 1.0381,
+      "step": 6106
+    },
+    {
+      "epoch": 0.44935387445389746,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.9062088946610244e-05,
+      "loss": 0.8927,
+      "step": 6107
+    },
+    {
+      "epoch": 0.44942745458726147,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.905637420596103e-05,
+      "loss": 0.5416,
+      "step": 6108
+    },
+    {
+      "epoch": 0.4495010347206254,
+      "grad_norm": 1.1640625,
+      "learning_rate": 2.905065924761387e-05,
+      "loss": 1.0362,
+      "step": 6109
+    },
+    {
+      "epoch": 0.4495746148539894,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.904494407187548e-05,
+      "loss": 0.749,
+      "step": 6110
+    },
+    {
+      "epoch": 0.4496481949873534,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.9039228679052583e-05,
+      "loss": 0.9613,
+      "step": 6111
+    },
+    {
+      "epoch": 0.4497217751207174,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.9033513069451905e-05,
+      "loss": 1.0949,
+      "step": 6112
+    },
+    {
+      "epoch": 0.4497953552540814,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.9027797243380205e-05,
+      "loss": 0.7375,
+      "step": 6113
+    },
+    {
+      "epoch": 0.4498689353874454,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.902208120114423e-05,
+      "loss": 0.6392,
+      "step": 6114
+    },
+    {
+      "epoch": 0.4499425155208094,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.9016364943050763e-05,
+      "loss": 1.0716,
+      "step": 6115
+    },
+    {
+      "epoch": 0.4500160956541734,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.9010648469406565e-05,
+      "loss": 1.2349,
+      "step": 6116
+    },
+    {
+      "epoch": 0.4500896757875374,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.900493178051845e-05,
+      "loss": 0.6422,
+      "step": 6117
+    },
+    {
+      "epoch": 0.45016325592090134,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.8999214876693205e-05,
+      "loss": 1.2383,
+      "step": 6118
+    },
+    {
+      "epoch": 0.45023683605426534,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.899349775823764e-05,
+      "loss": 0.9533,
+      "step": 6119
+    },
+    {
+      "epoch": 0.45031041618762935,
+      "grad_norm": 1.2265625,
+      "learning_rate": 2.8987780425458604e-05,
+      "loss": 1.0481,
+      "step": 6120
+    },
+    {
+      "epoch": 0.45038399632099335,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.898206287866293e-05,
+      "loss": 1.1189,
+      "step": 6121
+    },
+    {
+      "epoch": 0.4504575764543573,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.897634511815745e-05,
+      "loss": 0.6942,
+      "step": 6122
+    },
+    {
+      "epoch": 0.4505311565877213,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.897062714424904e-05,
+      "loss": 1.0282,
+      "step": 6123
+    },
+    {
+      "epoch": 0.4506047367210853,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.896490895724458e-05,
+      "loss": 1.0714,
+      "step": 6124
+    },
+    {
+      "epoch": 0.4506783168544493,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.8959190557450937e-05,
+      "loss": 0.9996,
+      "step": 6125
+    },
+    {
+      "epoch": 0.4507518969878133,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.8953471945175013e-05,
+      "loss": 1.0714,
+      "step": 6126
+    },
+    {
+      "epoch": 0.45082547712117726,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.894775312072372e-05,
+      "loss": 1.0969,
+      "step": 6127
+    },
+    {
+      "epoch": 0.45089905725454127,
+      "grad_norm": 1.2890625,
+      "learning_rate": 2.894203408440398e-05,
+      "loss": 1.3571,
+      "step": 6128
+    },
+    {
+      "epoch": 0.4509726373879053,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.8936314836522706e-05,
+      "loss": 0.7812,
+      "step": 6129
+    },
+    {
+      "epoch": 0.4510462175212693,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.893059537738685e-05,
+      "loss": 0.8675,
+      "step": 6130
+    },
+    {
+      "epoch": 0.4511197976546332,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.8924875707303366e-05,
+      "loss": 1.1733,
+      "step": 6131
+    },
+    {
+      "epoch": 0.45119337778799723,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.8919155826579215e-05,
+      "loss": 1.1205,
+      "step": 6132
+    },
+    {
+      "epoch": 0.45126695792136123,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.8913435735521377e-05,
+      "loss": 0.9608,
+      "step": 6133
+    },
+    {
+      "epoch": 0.45134053805472524,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.8907715434436834e-05,
+      "loss": 0.8731,
+      "step": 6134
+    },
+    {
+      "epoch": 0.45141411818808924,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.8901994923632582e-05,
+      "loss": 1.1364,
+      "step": 6135
+    },
+    {
+      "epoch": 0.4514876983214532,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.889627420341563e-05,
+      "loss": 0.8074,
+      "step": 6136
+    },
+    {
+      "epoch": 0.4515612784548172,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.8890553274093006e-05,
+      "loss": 0.909,
+      "step": 6137
+    },
+    {
+      "epoch": 0.4516348585881812,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.8884832135971735e-05,
+      "loss": 0.6877,
+      "step": 6138
+    },
+    {
+      "epoch": 0.4517084387215452,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.887911078935885e-05,
+      "loss": 0.9929,
+      "step": 6139
+    },
+    {
+      "epoch": 0.45178201885490915,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.8873389234561428e-05,
+      "loss": 0.6979,
+      "step": 6140
+    },
+    {
+      "epoch": 0.45185559898827315,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.8867667471886518e-05,
+      "loss": 1.1729,
+      "step": 6141
+    },
+    {
+      "epoch": 0.45192917912163716,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.88619455016412e-05,
+      "loss": 0.9276,
+      "step": 6142
+    },
+    {
+      "epoch": 0.45200275925500116,
+      "grad_norm": 1.0,
+      "learning_rate": 2.885622332413256e-05,
+      "loss": 1.0578,
+      "step": 6143
+    },
+    {
+      "epoch": 0.45207633938836517,
+      "grad_norm": 1.0,
+      "learning_rate": 2.88505009396677e-05,
+      "loss": 1.0555,
+      "step": 6144
+    },
+    {
+      "epoch": 0.4521499195217291,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.8844778348553724e-05,
+      "loss": 1.0048,
+      "step": 6145
+    },
+    {
+      "epoch": 0.4522234996550931,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.8839055551097755e-05,
+      "loss": 0.837,
+      "step": 6146
+    },
+    {
+      "epoch": 0.4522970797884571,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.8833332547606927e-05,
+      "loss": 0.6533,
+      "step": 6147
+    },
+    {
+      "epoch": 0.4523706599218211,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.8827609338388385e-05,
+      "loss": 1.1388,
+      "step": 6148
+    },
+    {
+      "epoch": 0.4524442400551851,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.8821885923749275e-05,
+      "loss": 0.8837,
+      "step": 6149
+    },
+    {
+      "epoch": 0.4525178201885491,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.881616230399677e-05,
+      "loss": 0.9821,
+      "step": 6150
+    },
+    {
+      "epoch": 0.4525914003219131,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.881043847943804e-05,
+      "loss": 1.0957,
+      "step": 6151
+    },
+    {
+      "epoch": 0.4526649804552771,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.8804714450380266e-05,
+      "loss": 0.7325,
+      "step": 6152
+    },
+    {
+      "epoch": 0.4527385605886411,
+      "grad_norm": 1.0,
+      "learning_rate": 2.879899021713066e-05,
+      "loss": 0.6788,
+      "step": 6153
+    },
+    {
+      "epoch": 0.45281214072200504,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.8793265779996427e-05,
+      "loss": 0.6262,
+      "step": 6154
+    },
+    {
+      "epoch": 0.45288572085536904,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.878754113928478e-05,
+      "loss": 0.8315,
+      "step": 6155
+    },
+    {
+      "epoch": 0.45295930098873305,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.878181629530296e-05,
+      "loss": 0.8807,
+      "step": 6156
+    },
+    {
+      "epoch": 0.45303288112209705,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.8776091248358194e-05,
+      "loss": 0.78,
+      "step": 6157
+    },
+    {
+      "epoch": 0.453106461255461,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.8770365998757754e-05,
+      "loss": 0.7171,
+      "step": 6158
+    },
+    {
+      "epoch": 0.453180041388825,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.8764640546808873e-05,
+      "loss": 0.9987,
+      "step": 6159
+    },
+    {
+      "epoch": 0.453253621522189,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.875891489281886e-05,
+      "loss": 0.6616,
+      "step": 6160
+    },
+    {
+      "epoch": 0.453327201655553,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.875318903709498e-05,
+      "loss": 0.9825,
+      "step": 6161
+    },
+    {
+      "epoch": 0.453400781788917,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.8747462979944534e-05,
+      "loss": 0.7909,
+      "step": 6162
+    },
+    {
+      "epoch": 0.45347436192228097,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.8741736721674827e-05,
+      "loss": 0.7882,
+      "step": 6163
+    },
+    {
+      "epoch": 0.45354794205564497,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.8736010262593178e-05,
+      "loss": 0.8349,
+      "step": 6164
+    },
+    {
+      "epoch": 0.453621522189009,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.8730283603006907e-05,
+      "loss": 0.9046,
+      "step": 6165
+    },
+    {
+      "epoch": 0.453695102322373,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.872455674322337e-05,
+      "loss": 1.1353,
+      "step": 6166
+    },
+    {
+      "epoch": 0.4537686824557369,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.871882968354991e-05,
+      "loss": 0.9633,
+      "step": 6167
+    },
+    {
+      "epoch": 0.45384226258910093,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.8713102424293874e-05,
+      "loss": 0.6544,
+      "step": 6168
+    },
+    {
+      "epoch": 0.45391584272246494,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.870737496576265e-05,
+      "loss": 1.5257,
+      "step": 6169
+    },
+    {
+      "epoch": 0.45398942285582894,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.8701647308263613e-05,
+      "loss": 0.6725,
+      "step": 6170
+    },
+    {
+      "epoch": 0.45406300298919294,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.8695919452104157e-05,
+      "loss": 1.0626,
+      "step": 6171
+    },
+    {
+      "epoch": 0.4541365831225569,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.8690191397591682e-05,
+      "loss": 0.9497,
+      "step": 6172
+    },
+    {
+      "epoch": 0.4542101632559209,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.86844631450336e-05,
+      "loss": 0.7885,
+      "step": 6173
+    },
+    {
+      "epoch": 0.4542837433892849,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.8678734694737352e-05,
+      "loss": 0.8002,
+      "step": 6174
+    },
+    {
+      "epoch": 0.4543573235226489,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.8673006047010353e-05,
+      "loss": 0.8922,
+      "step": 6175
+    },
+    {
+      "epoch": 0.45443090365601285,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.8667277202160066e-05,
+      "loss": 1.1074,
+      "step": 6176
+    },
+    {
+      "epoch": 0.45450448378937686,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.8661548160493927e-05,
+      "loss": 0.7483,
+      "step": 6177
+    },
+    {
+      "epoch": 0.45457806392274086,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.865581892231942e-05,
+      "loss": 0.8306,
+      "step": 6178
+    },
+    {
+      "epoch": 0.45465164405610486,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.865008948794401e-05,
+      "loss": 1.1476,
+      "step": 6179
+    },
+    {
+      "epoch": 0.45472522418946887,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.8644359857675195e-05,
+      "loss": 0.7028,
+      "step": 6180
+    },
+    {
+      "epoch": 0.4547988043228328,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.8638630031820472e-05,
+      "loss": 1.1273,
+      "step": 6181
+    },
+    {
+      "epoch": 0.4548723844561968,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.863290001068735e-05,
+      "loss": 1.0302,
+      "step": 6182
+    },
+    {
+      "epoch": 0.4549459645895608,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.862716979458334e-05,
+      "loss": 1.0111,
+      "step": 6183
+    },
+    {
+      "epoch": 0.45501954472292483,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.8621439383815985e-05,
+      "loss": 0.8821,
+      "step": 6184
+    },
+    {
+      "epoch": 0.4550931248562888,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.8615708778692807e-05,
+      "loss": 0.9798,
+      "step": 6185
+    },
+    {
+      "epoch": 0.4551667049896528,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.8609977979521375e-05,
+      "loss": 1.1728,
+      "step": 6186
+    },
+    {
+      "epoch": 0.4552402851230168,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.860424698660924e-05,
+      "loss": 0.7231,
+      "step": 6187
+    },
+    {
+      "epoch": 0.4553138652563808,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.859851580026398e-05,
+      "loss": 0.9579,
+      "step": 6188
+    },
+    {
+      "epoch": 0.4553874453897448,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.8592784420793173e-05,
+      "loss": 0.9078,
+      "step": 6189
+    },
+    {
+      "epoch": 0.45546102552310874,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.8587052848504414e-05,
+      "loss": 1.1242,
+      "step": 6190
+    },
+    {
+      "epoch": 0.45553460565647275,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.8581321083705298e-05,
+      "loss": 1.0109,
+      "step": 6191
+    },
+    {
+      "epoch": 0.45560818578983675,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.8575589126703452e-05,
+      "loss": 0.6212,
+      "step": 6192
+    },
+    {
+      "epoch": 0.45568176592320075,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.856985697780648e-05,
+      "loss": 0.9691,
+      "step": 6193
+    },
+    {
+      "epoch": 0.4557553460565647,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.856412463732204e-05,
+      "loss": 0.9252,
+      "step": 6194
+    },
+    {
+      "epoch": 0.4558289261899287,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.8558392105557746e-05,
+      "loss": 0.978,
+      "step": 6195
+    },
+    {
+      "epoch": 0.4559025063232927,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.8552659382821277e-05,
+      "loss": 0.6545,
+      "step": 6196
+    },
+    {
+      "epoch": 0.4559760864566567,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.8546926469420293e-05,
+      "loss": 1.3891,
+      "step": 6197
+    },
+    {
+      "epoch": 0.4560496665900207,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.854119336566246e-05,
+      "loss": 0.9865,
+      "step": 6198
+    },
+    {
+      "epoch": 0.45612324672338467,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.8535460071855464e-05,
+      "loss": 1.1086,
+      "step": 6199
+    },
+    {
+      "epoch": 0.45619682685674867,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.8529726588307005e-05,
+      "loss": 1.3484,
+      "step": 6200
+    },
+    {
+      "epoch": 0.4562704069901127,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.8523992915324794e-05,
+      "loss": 0.8123,
+      "step": 6201
+    },
+    {
+      "epoch": 0.4563439871234767,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.8518259053216534e-05,
+      "loss": 0.9677,
+      "step": 6202
+    },
+    {
+      "epoch": 0.45641756725684063,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.8512525002289954e-05,
+      "loss": 1.1213,
+      "step": 6203
+    },
+    {
+      "epoch": 0.45649114739020463,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.8506790762852796e-05,
+      "loss": 1.123,
+      "step": 6204
+    },
+    {
+      "epoch": 0.45656472752356864,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.8501056335212794e-05,
+      "loss": 1.2317,
+      "step": 6205
+    },
+    {
+      "epoch": 0.45663830765693264,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.8495321719677714e-05,
+      "loss": 1.1782,
+      "step": 6206
+    },
+    {
+      "epoch": 0.45671188779029664,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.8489586916555322e-05,
+      "loss": 0.6275,
+      "step": 6207
+    },
+    {
+      "epoch": 0.4567854679236606,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.8483851926153393e-05,
+      "loss": 0.6487,
+      "step": 6208
+    },
+    {
+      "epoch": 0.4568590480570246,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.8478116748779705e-05,
+      "loss": 0.946,
+      "step": 6209
+    },
+    {
+      "epoch": 0.4569326281903886,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.847238138474207e-05,
+      "loss": 1.0903,
+      "step": 6210
+    },
+    {
+      "epoch": 0.4570062083237526,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.8466645834348272e-05,
+      "loss": 0.7167,
+      "step": 6211
+    },
+    {
+      "epoch": 0.45707978845711655,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.8460910097906148e-05,
+      "loss": 1.0082,
+      "step": 6212
+    },
+    {
+      "epoch": 0.45715336859048056,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.8455174175723508e-05,
+      "loss": 0.9808,
+      "step": 6213
+    },
+    {
+      "epoch": 0.45722694872384456,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.8449438068108198e-05,
+      "loss": 0.6252,
+      "step": 6214
+    },
+    {
+      "epoch": 0.45730052885720857,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.844370177536807e-05,
+      "loss": 0.8817,
+      "step": 6215
+    },
+    {
+      "epoch": 0.45737410899057257,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.8437965297810966e-05,
+      "loss": 1.0954,
+      "step": 6216
+    },
+    {
+      "epoch": 0.4574476891239365,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.843222863574476e-05,
+      "loss": 1.0867,
+      "step": 6217
+    },
+    {
+      "epoch": 0.4575212692573005,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.842649178947732e-05,
+      "loss": 1.1105,
+      "step": 6218
+    },
+    {
+      "epoch": 0.4575948493906645,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.842075475931654e-05,
+      "loss": 0.9596,
+      "step": 6219
+    },
+    {
+      "epoch": 0.45766842952402853,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.8415017545570318e-05,
+      "loss": 1.0159,
+      "step": 6220
+    },
+    {
+      "epoch": 0.4577420096573925,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.8409280148546548e-05,
+      "loss": 1.4714,
+      "step": 6221
+    },
+    {
+      "epoch": 0.4578155897907565,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.840354256855316e-05,
+      "loss": 0.8535,
+      "step": 6222
+    },
+    {
+      "epoch": 0.4578891699241205,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.8397804805898058e-05,
+      "loss": 1.2393,
+      "step": 6223
+    },
+    {
+      "epoch": 0.4579627500574845,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.8392066860889206e-05,
+      "loss": 0.7248,
+      "step": 6224
+    },
+    {
+      "epoch": 0.4580363301908485,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.8386328733834522e-05,
+      "loss": 0.9474,
+      "step": 6225
+    },
+    {
+      "epoch": 0.45810991032421244,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.838059042504197e-05,
+      "loss": 0.9051,
+      "step": 6226
+    },
+    {
+      "epoch": 0.45818349045757645,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.8374851934819514e-05,
+      "loss": 1.2669,
+      "step": 6227
+    },
+    {
+      "epoch": 0.45825707059094045,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.836911326347513e-05,
+      "loss": 0.6519,
+      "step": 6228
+    },
+    {
+      "epoch": 0.45833065072430446,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.8363374411316806e-05,
+      "loss": 1.4039,
+      "step": 6229
+    },
+    {
+      "epoch": 0.4584042308576684,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.8357635378652527e-05,
+      "loss": 0.9145,
+      "step": 6230
+    },
+    {
+      "epoch": 0.4584778109910324,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.83518961657903e-05,
+      "loss": 1.0376,
+      "step": 6231
+    },
+    {
+      "epoch": 0.4585513911243964,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.834615677303813e-05,
+      "loss": 0.8081,
+      "step": 6232
+    },
+    {
+      "epoch": 0.4586249712577604,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.834041720070405e-05,
+      "loss": 0.8228,
+      "step": 6233
+    },
+    {
+      "epoch": 0.4586985513911244,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.8334677449096088e-05,
+      "loss": 0.5794,
+      "step": 6234
+    },
+    {
+      "epoch": 0.45877213152448837,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.832893751852228e-05,
+      "loss": 0.7654,
+      "step": 6235
+    },
+    {
+      "epoch": 0.4588457116578524,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.8323197409290696e-05,
+      "loss": 1.2892,
+      "step": 6236
+    },
+    {
+      "epoch": 0.4589192917912164,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.831745712170937e-05,
+      "loss": 1.0228,
+      "step": 6237
+    },
+    {
+      "epoch": 0.4589928719245804,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.83117166560864e-05,
+      "loss": 0.7601,
+      "step": 6238
+    },
+    {
+      "epoch": 0.45906645205794433,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.8305976012729845e-05,
+      "loss": 0.65,
+      "step": 6239
+    },
+    {
+      "epoch": 0.45914003219130833,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.83002351919478e-05,
+      "loss": 1.0663,
+      "step": 6240
+    },
+    {
+      "epoch": 0.45921361232467234,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.8294494194048365e-05,
+      "loss": 1.2407,
+      "step": 6241
+    },
+    {
+      "epoch": 0.45928719245803634,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.8288753019339653e-05,
+      "loss": 1.0828,
+      "step": 6242
+    },
+    {
+      "epoch": 0.45936077259140035,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.8283011668129777e-05,
+      "loss": 0.9263,
+      "step": 6243
+    },
+    {
+      "epoch": 0.4594343527247643,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.8277270140726863e-05,
+      "loss": 0.6277,
+      "step": 6244
+    },
+    {
+      "epoch": 0.4595079328581283,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.8271528437439054e-05,
+      "loss": 0.8775,
+      "step": 6245
+    },
+    {
+      "epoch": 0.4595815129914923,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.8265786558574486e-05,
+      "loss": 1.0949,
+      "step": 6246
+    },
+    {
+      "epoch": 0.4596550931248563,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.826004450444133e-05,
+      "loss": 1.1995,
+      "step": 6247
+    },
+    {
+      "epoch": 0.45972867325822026,
+      "grad_norm": 1.15625,
+      "learning_rate": 2.825430227534773e-05,
+      "loss": 1.6563,
+      "step": 6248
+    },
+    {
+      "epoch": 0.45980225339158426,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.8248559871601887e-05,
+      "loss": 0.9782,
+      "step": 6249
+    },
+    {
+      "epoch": 0.45987583352494826,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.824281729351197e-05,
+      "loss": 0.7899,
+      "step": 6250
+    },
+    {
+      "epoch": 0.45994941365831227,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.8237074541386167e-05,
+      "loss": 0.9182,
+      "step": 6251
+    },
+    {
+      "epoch": 0.46002299379167627,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.8231331615532697e-05,
+      "loss": 0.6859,
+      "step": 6252
+    },
+    {
+      "epoch": 0.4600965739250402,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.8225588516259748e-05,
+      "loss": 1.0853,
+      "step": 6253
+    },
+    {
+      "epoch": 0.4601701540584042,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.8219845243875565e-05,
+      "loss": 1.3617,
+      "step": 6254
+    },
+    {
+      "epoch": 0.46024373419176823,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.8214101798688363e-05,
+      "loss": 1.1836,
+      "step": 6255
+    },
+    {
+      "epoch": 0.46031731432513223,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.82083581810064e-05,
+      "loss": 0.8941,
+      "step": 6256
+    },
+    {
+      "epoch": 0.4603908944584962,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.8202614391137906e-05,
+      "loss": 0.7217,
+      "step": 6257
+    },
+    {
+      "epoch": 0.4604644745918602,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.8196870429391147e-05,
+      "loss": 0.6514,
+      "step": 6258
+    },
+    {
+      "epoch": 0.4605380547252242,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.819112629607439e-05,
+      "loss": 0.9014,
+      "step": 6259
+    },
+    {
+      "epoch": 0.4606116348585882,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.8185381991495908e-05,
+      "loss": 0.708,
+      "step": 6260
+    },
+    {
+      "epoch": 0.4606852149919522,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.8179637515963996e-05,
+      "loss": 1.1131,
+      "step": 6261
+    },
+    {
+      "epoch": 0.46075879512531615,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.817389286978694e-05,
+      "loss": 1.6942,
+      "step": 6262
+    },
+    {
+      "epoch": 0.46083237525868015,
+      "grad_norm": 1.25,
+      "learning_rate": 2.8168148053273053e-05,
+      "loss": 1.1356,
+      "step": 6263
+    },
+    {
+      "epoch": 0.46090595539204415,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.8162403066730643e-05,
+      "loss": 1.2119,
+      "step": 6264
+    },
+    {
+      "epoch": 0.46097953552540816,
+      "grad_norm": 0.875,
+      "learning_rate": 2.815665791046803e-05,
+      "loss": 1.3804,
+      "step": 6265
+    },
+    {
+      "epoch": 0.4610531156587721,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.8150912584793554e-05,
+      "loss": 1.095,
+      "step": 6266
+    },
+    {
+      "epoch": 0.4611266957921361,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.8145167090015546e-05,
+      "loss": 1.2664,
+      "step": 6267
+    },
+    {
+      "epoch": 0.4612002759255001,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.8139421426442357e-05,
+      "loss": 0.993,
+      "step": 6268
+    },
+    {
+      "epoch": 0.4612738560588641,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.813367559438235e-05,
+      "loss": 0.9281,
+      "step": 6269
+    },
+    {
+      "epoch": 0.4613474361922281,
+      "grad_norm": 1.1328125,
+      "learning_rate": 2.8127929594143903e-05,
+      "loss": 1.0383,
+      "step": 6270
+    },
+    {
+      "epoch": 0.46142101632559207,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.8122183426035377e-05,
+      "loss": 0.9486,
+      "step": 6271
+    },
+    {
+      "epoch": 0.4614945964589561,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.8116437090365166e-05,
+      "loss": 1.5287,
+      "step": 6272
+    },
+    {
+      "epoch": 0.4615681765923201,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.8110690587441656e-05,
+      "loss": 1.0122,
+      "step": 6273
+    },
+    {
+      "epoch": 0.4616417567256841,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.8104943917573262e-05,
+      "loss": 1.0152,
+      "step": 6274
+    },
+    {
+      "epoch": 0.46171533685904803,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.809919708106839e-05,
+      "loss": 1.0841,
+      "step": 6275
+    },
+    {
+      "epoch": 0.46178891699241204,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.809345007823546e-05,
+      "loss": 1.2076,
+      "step": 6276
+    },
+    {
+      "epoch": 0.46186249712577604,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.8087702909382918e-05,
+      "loss": 1.175,
+      "step": 6277
+    },
+    {
+      "epoch": 0.46193607725914004,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.8081955574819184e-05,
+      "loss": 0.932,
+      "step": 6278
+    },
+    {
+      "epoch": 0.46200965739250405,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.8076208074852728e-05,
+      "loss": 0.9306,
+      "step": 6279
+    },
+    {
+      "epoch": 0.462083237525868,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.807046040979198e-05,
+      "loss": 1.2228,
+      "step": 6280
+    },
+    {
+      "epoch": 0.462156817659232,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.806471257994543e-05,
+      "loss": 0.7424,
+      "step": 6281
+    },
+    {
+      "epoch": 0.462230397792596,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.805896458562154e-05,
+      "loss": 0.6086,
+      "step": 6282
+    },
+    {
+      "epoch": 0.46230397792596,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.8053216427128796e-05,
+      "loss": 0.768,
+      "step": 6283
+    },
+    {
+      "epoch": 0.46237755805932396,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.8047468104775697e-05,
+      "loss": 0.8747,
+      "step": 6284
+    },
+    {
+      "epoch": 0.46245113819268796,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.8041719618870737e-05,
+      "loss": 0.739,
+      "step": 6285
+    },
+    {
+      "epoch": 0.46252471832605196,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.8035970969722436e-05,
+      "loss": 0.9525,
+      "step": 6286
+    },
+    {
+      "epoch": 0.46259829845941597,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.8030222157639308e-05,
+      "loss": 1.1422,
+      "step": 6287
+    },
+    {
+      "epoch": 0.46267187859278,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.8024473182929872e-05,
+      "loss": 0.64,
+      "step": 6288
+    },
+    {
+      "epoch": 0.4627454587261439,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.8018724045902673e-05,
+      "loss": 0.8409,
+      "step": 6289
+    },
+    {
+      "epoch": 0.4628190388595079,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.8012974746866266e-05,
+      "loss": 0.823,
+      "step": 6290
+    },
+    {
+      "epoch": 0.46289261899287193,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.8007225286129187e-05,
+      "loss": 0.7293,
+      "step": 6291
+    },
+    {
+      "epoch": 0.46296619912623593,
+      "grad_norm": 0.875,
+      "learning_rate": 2.8001475664000004e-05,
+      "loss": 0.8793,
+      "step": 6292
+    },
+    {
+      "epoch": 0.4630397792595999,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.79957258807873e-05,
+      "loss": 1.0189,
+      "step": 6293
+    },
+    {
+      "epoch": 0.4631133593929639,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.798997593679964e-05,
+      "loss": 0.839,
+      "step": 6294
+    },
+    {
+      "epoch": 0.4631869395263279,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.7984225832345624e-05,
+      "loss": 0.7649,
+      "step": 6295
+    },
+    {
+      "epoch": 0.4632605196596919,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.7978475567733836e-05,
+      "loss": 1.1753,
+      "step": 6296
+    },
+    {
+      "epoch": 0.4633340997930559,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.7972725143272898e-05,
+      "loss": 0.9838,
+      "step": 6297
+    },
+    {
+      "epoch": 0.46340767992641985,
+      "grad_norm": 1.0,
+      "learning_rate": 2.7966974559271407e-05,
+      "loss": 1.024,
+      "step": 6298
+    },
+    {
+      "epoch": 0.46348126005978385,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.7961223816038008e-05,
+      "loss": 0.7331,
+      "step": 6299
+    },
+    {
+      "epoch": 0.46355484019314785,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.795547291388131e-05,
+      "loss": 0.9146,
+      "step": 6300
+    },
+    {
+      "epoch": 0.46362842032651186,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.794972185310996e-05,
+      "loss": 0.5683,
+      "step": 6301
+    },
+    {
+      "epoch": 0.4637020004598758,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.7943970634032618e-05,
+      "loss": 1.18,
+      "step": 6302
+    },
+    {
+      "epoch": 0.4637755805932398,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.7938219256957926e-05,
+      "loss": 0.6956,
+      "step": 6303
+    },
+    {
+      "epoch": 0.4638491607266038,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.7932467722194562e-05,
+      "loss": 0.8688,
+      "step": 6304
+    },
+    {
+      "epoch": 0.4639227408599678,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.7926716030051193e-05,
+      "loss": 1.1108,
+      "step": 6305
+    },
+    {
+      "epoch": 0.4639963209933318,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.79209641808365e-05,
+      "loss": 0.4767,
+      "step": 6306
+    },
+    {
+      "epoch": 0.46406990112669577,
+      "grad_norm": 0.75,
+      "learning_rate": 2.7915212174859177e-05,
+      "loss": 0.7869,
+      "step": 6307
+    },
+    {
+      "epoch": 0.4641434812600598,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.7909460012427923e-05,
+      "loss": 1.1705,
+      "step": 6308
+    },
+    {
+      "epoch": 0.4642170613934238,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.7903707693851444e-05,
+      "loss": 0.7204,
+      "step": 6309
+    },
+    {
+      "epoch": 0.4642906415267878,
+      "grad_norm": 0.625,
+      "learning_rate": 2.7897955219438454e-05,
+      "loss": 0.7186,
+      "step": 6310
+    },
+    {
+      "epoch": 0.46436422166015173,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.789220258949769e-05,
+      "loss": 0.8165,
+      "step": 6311
+    },
+    {
+      "epoch": 0.46443780179351574,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.7886449804337867e-05,
+      "loss": 0.8139,
+      "step": 6312
+    },
+    {
+      "epoch": 0.46451138192687974,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.788069686426774e-05,
+      "loss": 1.2423,
+      "step": 6313
+    },
+    {
+      "epoch": 0.46458496206024374,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.787494376959605e-05,
+      "loss": 0.674,
+      "step": 6314
+    },
+    {
+      "epoch": 0.46465854219360775,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.786919052063156e-05,
+      "loss": 1.146,
+      "step": 6315
+    },
+    {
+      "epoch": 0.4647321223269717,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.7863437117683034e-05,
+      "loss": 1.2663,
+      "step": 6316
+    },
+    {
+      "epoch": 0.4648057024603357,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.7857683561059245e-05,
+      "loss": 0.5894,
+      "step": 6317
+    },
+    {
+      "epoch": 0.4648792825936997,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.785192985106898e-05,
+      "loss": 1.1378,
+      "step": 6318
+    },
+    {
+      "epoch": 0.4649528627270637,
+      "grad_norm": 1.0,
+      "learning_rate": 2.784617598802102e-05,
+      "loss": 0.7876,
+      "step": 6319
+    },
+    {
+      "epoch": 0.46502644286042766,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.784042197222418e-05,
+      "loss": 0.8514,
+      "step": 6320
+    },
+    {
+      "epoch": 0.46510002299379166,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.7834667803987247e-05,
+      "loss": 1.0082,
+      "step": 6321
+    },
+    {
+      "epoch": 0.46517360312715567,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.782891348361905e-05,
+      "loss": 1.0139,
+      "step": 6322
+    },
+    {
+      "epoch": 0.46524718326051967,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.7823159011428412e-05,
+      "loss": 0.634,
+      "step": 6323
+    },
+    {
+      "epoch": 0.4653207633938837,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.7817404387724154e-05,
+      "loss": 0.6427,
+      "step": 6324
+    },
+    {
+      "epoch": 0.4653943435272476,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.7811649612815138e-05,
+      "loss": 0.7685,
+      "step": 6325
+    },
+    {
+      "epoch": 0.4654679236606116,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.7805894687010186e-05,
+      "loss": 0.8788,
+      "step": 6326
+    },
+    {
+      "epoch": 0.46554150379397563,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.7800139610618176e-05,
+      "loss": 0.6293,
+      "step": 6327
+    },
+    {
+      "epoch": 0.46561508392733963,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.7794384383947945e-05,
+      "loss": 0.7029,
+      "step": 6328
+    },
+    {
+      "epoch": 0.4656886640607036,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.7788629007308396e-05,
+      "loss": 1.1446,
+      "step": 6329
+    },
+    {
+      "epoch": 0.4657622441940676,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.77828734810084e-05,
+      "loss": 0.9256,
+      "step": 6330
+    },
+    {
+      "epoch": 0.4658358243274316,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.7777117805356834e-05,
+      "loss": 1.0913,
+      "step": 6331
+    },
+    {
+      "epoch": 0.4659094044607956,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.7771361980662596e-05,
+      "loss": 1.1589,
+      "step": 6332
+    },
+    {
+      "epoch": 0.4659829845941596,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.7765606007234597e-05,
+      "loss": 0.7919,
+      "step": 6333
+    },
+    {
+      "epoch": 0.46605656472752355,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.775984988538175e-05,
+      "loss": 0.9905,
+      "step": 6334
+    },
+    {
+      "epoch": 0.46613014486088755,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.7754093615412963e-05,
+      "loss": 0.6927,
+      "step": 6335
+    },
+    {
+      "epoch": 0.46620372499425156,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.7748337197637186e-05,
+      "loss": 0.8133,
+      "step": 6336
+    },
+    {
+      "epoch": 0.46627730512761556,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.7742580632363336e-05,
+      "loss": 0.9428,
+      "step": 6337
+    },
+    {
+      "epoch": 0.4663508852609795,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.7736823919900367e-05,
+      "loss": 0.9724,
+      "step": 6338
+    },
+    {
+      "epoch": 0.4664244653943435,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.7731067060557224e-05,
+      "loss": 0.9285,
+      "step": 6339
+    },
+    {
+      "epoch": 0.4664980455277075,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.7725310054642866e-05,
+      "loss": 1.1349,
+      "step": 6340
+    },
+    {
+      "epoch": 0.4665716256610715,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.7719552902466273e-05,
+      "loss": 0.9095,
+      "step": 6341
+    },
+    {
+      "epoch": 0.4666452057944355,
+      "grad_norm": 1.0,
+      "learning_rate": 2.7713795604336407e-05,
+      "loss": 1.5698,
+      "step": 6342
+    },
+    {
+      "epoch": 0.4667187859277995,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.770803816056226e-05,
+      "loss": 0.9831,
+      "step": 6343
+    },
+    {
+      "epoch": 0.4667923660611635,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.7702280571452817e-05,
+      "loss": 0.7587,
+      "step": 6344
+    },
+    {
+      "epoch": 0.4668659461945275,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.7696522837317085e-05,
+      "loss": 0.9243,
+      "step": 6345
+    },
+    {
+      "epoch": 0.4669395263278915,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.769076495846406e-05,
+      "loss": 0.9187,
+      "step": 6346
+    },
+    {
+      "epoch": 0.46701310646125543,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.7685006935202768e-05,
+      "loss": 0.8834,
+      "step": 6347
+    },
+    {
+      "epoch": 0.46708668659461944,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.767924876784221e-05,
+      "loss": 0.8859,
+      "step": 6348
+    },
+    {
+      "epoch": 0.46716026672798344,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.7673490456691442e-05,
+      "loss": 1.0394,
+      "step": 6349
+    },
+    {
+      "epoch": 0.46723384686134745,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.7667732002059493e-05,
+      "loss": 0.8982,
+      "step": 6350
+    },
+    {
+      "epoch": 0.46730742699471145,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.76619734042554e-05,
+      "loss": 0.9029,
+      "step": 6351
+    },
+    {
+      "epoch": 0.4673810071280754,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.7656214663588226e-05,
+      "loss": 0.7857,
+      "step": 6352
+    },
+    {
+      "epoch": 0.4674545872614394,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.765045578036703e-05,
+      "loss": 1.067,
+      "step": 6353
+    },
+    {
+      "epoch": 0.4675281673948034,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.7644696754900878e-05,
+      "loss": 0.9241,
+      "step": 6354
+    },
+    {
+      "epoch": 0.4676017475281674,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.7638937587498835e-05,
+      "loss": 0.723,
+      "step": 6355
+    },
+    {
+      "epoch": 0.46767532766153136,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.763317827847e-05,
+      "loss": 1.4426,
+      "step": 6356
+    },
+    {
+      "epoch": 0.46774890779489536,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.762741882812347e-05,
+      "loss": 1.2153,
+      "step": 6357
+    },
+    {
+      "epoch": 0.46782248792825937,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.7621659236768326e-05,
+      "loss": 0.7467,
+      "step": 6358
+    },
+    {
+      "epoch": 0.46789606806162337,
+      "grad_norm": 1.0,
+      "learning_rate": 2.7615899504713687e-05,
+      "loss": 1.0205,
+      "step": 6359
+    },
+    {
+      "epoch": 0.4679696481949874,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.761013963226866e-05,
+      "loss": 0.689,
+      "step": 6360
+    },
+    {
+      "epoch": 0.4680432283283513,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.7604379619742378e-05,
+      "loss": 0.7143,
+      "step": 6361
+    },
+    {
+      "epoch": 0.46811680846171533,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.7598619467443943e-05,
+      "loss": 0.7523,
+      "step": 6362
+    },
+    {
+      "epoch": 0.46819038859507933,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.7592859175682517e-05,
+      "loss": 0.6483,
+      "step": 6363
+    },
+    {
+      "epoch": 0.46826396872844334,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.7587098744767238e-05,
+      "loss": 0.57,
+      "step": 6364
+    },
+    {
+      "epoch": 0.4683375488618073,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.7581338175007253e-05,
+      "loss": 0.8446,
+      "step": 6365
+    },
+    {
+      "epoch": 0.4684111289951713,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.7575577466711733e-05,
+      "loss": 1.0267,
+      "step": 6366
+    },
+    {
+      "epoch": 0.4684847091285353,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.7569816620189825e-05,
+      "loss": 0.8569,
+      "step": 6367
+    },
+    {
+      "epoch": 0.4685582892618993,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.7564055635750712e-05,
+      "loss": 0.9914,
+      "step": 6368
+    },
+    {
+      "epoch": 0.4686318693952633,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.7558294513703575e-05,
+      "loss": 0.8706,
+      "step": 6369
+    },
+    {
+      "epoch": 0.46870544952862725,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.7552533254357615e-05,
+      "loss": 1.0821,
+      "step": 6370
+    },
+    {
+      "epoch": 0.46877902966199125,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.7546771858022006e-05,
+      "loss": 0.7812,
+      "step": 6371
+    },
+    {
+      "epoch": 0.46885260979535526,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.7541010325005963e-05,
+      "loss": 0.8116,
+      "step": 6372
+    },
+    {
+      "epoch": 0.46892618992871926,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.7535248655618702e-05,
+      "loss": 0.9588,
+      "step": 6373
+    },
+    {
+      "epoch": 0.4689997700620832,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.7529486850169422e-05,
+      "loss": 0.974,
+      "step": 6374
+    },
+    {
+      "epoch": 0.4690733501954472,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.7523724908967367e-05,
+      "loss": 1.1405,
+      "step": 6375
+    },
+    {
+      "epoch": 0.4691469303288112,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.7517962832321763e-05,
+      "loss": 1.1151,
+      "step": 6376
+    },
+    {
+      "epoch": 0.4692205104621752,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.751220062054185e-05,
+      "loss": 0.965,
+      "step": 6377
+    },
+    {
+      "epoch": 0.4692940905955392,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.7506438273936874e-05,
+      "loss": 0.6209,
+      "step": 6378
+    },
+    {
+      "epoch": 0.4693676707289032,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.7500675792816094e-05,
+      "loss": 1.389,
+      "step": 6379
+    },
+    {
+      "epoch": 0.4694412508622672,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.7494913177488756e-05,
+      "loss": 1.2503,
+      "step": 6380
+    },
+    {
+      "epoch": 0.4695148309956312,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.748915042826415e-05,
+      "loss": 1.0505,
+      "step": 6381
+    },
+    {
+      "epoch": 0.4695884111289952,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.7483387545451535e-05,
+      "loss": 0.9022,
+      "step": 6382
+    },
+    {
+      "epoch": 0.46966199126235914,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.74776245293602e-05,
+      "loss": 0.6629,
+      "step": 6383
+    },
+    {
+      "epoch": 0.46973557139572314,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.7471861380299447e-05,
+      "loss": 0.7861,
+      "step": 6384
+    },
+    {
+      "epoch": 0.46980915152908714,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.7466098098578557e-05,
+      "loss": 1.0548,
+      "step": 6385
+    },
+    {
+      "epoch": 0.46988273166245115,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.746033468450684e-05,
+      "loss": 0.7586,
+      "step": 6386
+    },
+    {
+      "epoch": 0.46995631179581515,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.7454571138393603e-05,
+      "loss": 1.3519,
+      "step": 6387
+    },
+    {
+      "epoch": 0.4700298919291791,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.7448807460548174e-05,
+      "loss": 0.9955,
+      "step": 6388
+    },
+    {
+      "epoch": 0.4701034720625431,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.744304365127987e-05,
+      "loss": 1.0774,
+      "step": 6389
+    },
+    {
+      "epoch": 0.4701770521959071,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.7437279710898027e-05,
+      "loss": 0.6787,
+      "step": 6390
+    },
+    {
+      "epoch": 0.4702506323292711,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.7431515639711992e-05,
+      "loss": 1.358,
+      "step": 6391
+    },
+    {
+      "epoch": 0.4703242124626351,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.7425751438031098e-05,
+      "loss": 1.093,
+      "step": 6392
+    },
+    {
+      "epoch": 0.47039779259599906,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.7419987106164714e-05,
+      "loss": 0.5847,
+      "step": 6393
+    },
+    {
+      "epoch": 0.47047137272936307,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.741422264442218e-05,
+      "loss": 0.9338,
+      "step": 6394
+    },
+    {
+      "epoch": 0.4705449528627271,
+      "grad_norm": 1.0,
+      "learning_rate": 2.740845805311289e-05,
+      "loss": 1.2731,
+      "step": 6395
+    },
+    {
+      "epoch": 0.4706185329960911,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.7402693332546198e-05,
+      "loss": 1.114,
+      "step": 6396
+    },
+    {
+      "epoch": 0.470692113129455,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.739692848303149e-05,
+      "loss": 1.4891,
+      "step": 6397
+    },
+    {
+      "epoch": 0.47076569326281903,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.7391163504878164e-05,
+      "loss": 1.2744,
+      "step": 6398
+    },
+    {
+      "epoch": 0.47083927339618303,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.7385398398395606e-05,
+      "loss": 0.7885,
+      "step": 6399
+    },
+    {
+      "epoch": 0.47091285352954704,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.7379633163893226e-05,
+      "loss": 0.6159,
+      "step": 6400
+    },
+    {
+      "epoch": 0.47098643366291104,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.737386780168042e-05,
+      "loss": 1.0149,
+      "step": 6401
+    },
+    {
+      "epoch": 0.471060013796275,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.7368102312066618e-05,
+      "loss": 0.9604,
+      "step": 6402
+    },
+    {
+      "epoch": 0.471133593929639,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.7362336695361235e-05,
+      "loss": 0.6421,
+      "step": 6403
+    },
+    {
+      "epoch": 0.471207174063003,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.735657095187371e-05,
+      "loss": 1.1529,
+      "step": 6404
+    },
+    {
+      "epoch": 0.471280754196367,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.7350805081913466e-05,
+      "loss": 0.7504,
+      "step": 6405
+    },
+    {
+      "epoch": 0.47135433432973095,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.7345039085789953e-05,
+      "loss": 0.8606,
+      "step": 6406
+    },
+    {
+      "epoch": 0.47142791446309495,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.733927296381263e-05,
+      "loss": 0.7918,
+      "step": 6407
+    },
+    {
+      "epoch": 0.47150149459645896,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.7333506716290937e-05,
+      "loss": 0.9462,
+      "step": 6408
+    },
+    {
+      "epoch": 0.47157507472982296,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.7327740343534346e-05,
+      "loss": 1.1555,
+      "step": 6409
+    },
+    {
+      "epoch": 0.47164865486318697,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.732197384585233e-05,
+      "loss": 0.964,
+      "step": 6410
+    },
+    {
+      "epoch": 0.4717222349965509,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.7316207223554364e-05,
+      "loss": 0.6545,
+      "step": 6411
+    },
+    {
+      "epoch": 0.4717958151299149,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.731044047694993e-05,
+      "loss": 1.1812,
+      "step": 6412
+    },
+    {
+      "epoch": 0.4718693952632789,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.7304673606348514e-05,
+      "loss": 1.0098,
+      "step": 6413
+    },
+    {
+      "epoch": 0.4719429753966429,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.7298906612059628e-05,
+      "loss": 0.5711,
+      "step": 6414
+    },
+    {
+      "epoch": 0.4720165555300069,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.729313949439276e-05,
+      "loss": 0.9497,
+      "step": 6415
+    },
+    {
+      "epoch": 0.4720901356633709,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.7287372253657422e-05,
+      "loss": 0.9081,
+      "step": 6416
+    },
+    {
+      "epoch": 0.4721637157967349,
+      "grad_norm": 0.875,
+      "learning_rate": 2.7281604890163144e-05,
+      "loss": 0.9944,
+      "step": 6417
+    },
+    {
+      "epoch": 0.4722372959300989,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.727583740421944e-05,
+      "loss": 1.0096,
+      "step": 6418
+    },
+    {
+      "epoch": 0.4723108760634629,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.727006979613584e-05,
+      "loss": 1.1038,
+      "step": 6419
+    },
+    {
+      "epoch": 0.47238445619682684,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.726430206622188e-05,
+      "loss": 0.9111,
+      "step": 6420
+    },
+    {
+      "epoch": 0.47245803633019084,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.7258534214787108e-05,
+      "loss": 0.6363,
+      "step": 6421
+    },
+    {
+      "epoch": 0.47253161646355485,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.7252766242141066e-05,
+      "loss": 0.745,
+      "step": 6422
+    },
+    {
+      "epoch": 0.47260519659691885,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.7246998148593313e-05,
+      "loss": 0.906,
+      "step": 6423
+    },
+    {
+      "epoch": 0.4726787767302828,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.7241229934453417e-05,
+      "loss": 0.6845,
+      "step": 6424
+    },
+    {
+      "epoch": 0.4727523568636468,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.7235461600030947e-05,
+      "loss": 0.7243,
+      "step": 6425
+    },
+    {
+      "epoch": 0.4728259369970108,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.7229693145635472e-05,
+      "loss": 0.8047,
+      "step": 6426
+    },
+    {
+      "epoch": 0.4728995171303748,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.7223924571576577e-05,
+      "loss": 1.1103,
+      "step": 6427
+    },
+    {
+      "epoch": 0.4729730972637388,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.7218155878163847e-05,
+      "loss": 0.7126,
+      "step": 6428
+    },
+    {
+      "epoch": 0.47304667739710277,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.7212387065706885e-05,
+      "loss": 0.7595,
+      "step": 6429
+    },
+    {
+      "epoch": 0.47312025753046677,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.7206618134515284e-05,
+      "loss": 1.0391,
+      "step": 6430
+    },
+    {
+      "epoch": 0.4731938376638308,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.7200849084898654e-05,
+      "loss": 1.1466,
+      "step": 6431
+    },
+    {
+      "epoch": 0.4732674177971948,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.719507991716662e-05,
+      "loss": 0.9107,
+      "step": 6432
+    },
+    {
+      "epoch": 0.4733409979305587,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.718931063162879e-05,
+      "loss": 0.9272,
+      "step": 6433
+    },
+    {
+      "epoch": 0.47341457806392273,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.7183541228594796e-05,
+      "loss": 0.8903,
+      "step": 6434
+    },
+    {
+      "epoch": 0.47348815819728673,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.7177771708374257e-05,
+      "loss": 1.2076,
+      "step": 6435
+    },
+    {
+      "epoch": 0.47356173833065074,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.7172002071276832e-05,
+      "loss": 0.684,
+      "step": 6436
+    },
+    {
+      "epoch": 0.47363531846401474,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.716623231761215e-05,
+      "loss": 1.1711,
+      "step": 6437
+    },
+    {
+      "epoch": 0.4737088985973787,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.7160462447689877e-05,
+      "loss": 0.6916,
+      "step": 6438
+    },
+    {
+      "epoch": 0.4737824787307427,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.715469246181967e-05,
+      "loss": 0.7315,
+      "step": 6439
+    },
+    {
+      "epoch": 0.4738560588641067,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.714892236031118e-05,
+      "loss": 1.0714,
+      "step": 6440
+    },
+    {
+      "epoch": 0.4739296389974707,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.7143152143474093e-05,
+      "loss": 0.6242,
+      "step": 6441
+    },
+    {
+      "epoch": 0.47400321913083465,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.7137381811618073e-05,
+      "loss": 1.0698,
+      "step": 6442
+    },
+    {
+      "epoch": 0.47407679926419866,
+      "grad_norm": 0.75,
+      "learning_rate": 2.7131611365052806e-05,
+      "loss": 0.6787,
+      "step": 6443
+    },
+    {
+      "epoch": 0.47415037939756266,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.7125840804087987e-05,
+      "loss": 0.756,
+      "step": 6444
+    },
+    {
+      "epoch": 0.47422395953092666,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.7120070129033302e-05,
+      "loss": 0.7813,
+      "step": 6445
+    },
+    {
+      "epoch": 0.47429753966429067,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.7114299340198467e-05,
+      "loss": 1.0591,
+      "step": 6446
+    },
+    {
+      "epoch": 0.4743711197976546,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.710852843789317e-05,
+      "loss": 0.7675,
+      "step": 6447
+    },
+    {
+      "epoch": 0.4744446999310186,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.7102757422427145e-05,
+      "loss": 0.7435,
+      "step": 6448
+    },
+    {
+      "epoch": 0.4745182800643826,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.7096986294110094e-05,
+      "loss": 0.8249,
+      "step": 6449
+    },
+    {
+      "epoch": 0.47459186019774663,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.7091215053251745e-05,
+      "loss": 0.8246,
+      "step": 6450
+    },
+    {
+      "epoch": 0.4746654403311106,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.708544370016184e-05,
+      "loss": 0.9574,
+      "step": 6451
+    },
+    {
+      "epoch": 0.4747390204644746,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.707967223515011e-05,
+      "loss": 0.989,
+      "step": 6452
+    },
+    {
+      "epoch": 0.4748126005978386,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.7073900658526295e-05,
+      "loss": 1.0891,
+      "step": 6453
+    },
+    {
+      "epoch": 0.4748861807312026,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.7068128970600152e-05,
+      "loss": 1.2637,
+      "step": 6454
+    },
+    {
+      "epoch": 0.4749597608645666,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.7062357171681435e-05,
+      "loss": 0.8857,
+      "step": 6455
+    },
+    {
+      "epoch": 0.47503334099793054,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.7056585262079902e-05,
+      "loss": 1.1733,
+      "step": 6456
+    },
+    {
+      "epoch": 0.47510692113129455,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.705081324210532e-05,
+      "loss": 1.2251,
+      "step": 6457
+    },
+    {
+      "epoch": 0.47518050126465855,
+      "grad_norm": 1.1328125,
+      "learning_rate": 2.7045041112067465e-05,
+      "loss": 1.0351,
+      "step": 6458
+    },
+    {
+      "epoch": 0.47525408139802255,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.7039268872276124e-05,
+      "loss": 0.6846,
+      "step": 6459
+    },
+    {
+      "epoch": 0.4753276615313865,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.703349652304107e-05,
+      "loss": 0.6974,
+      "step": 6460
+    },
+    {
+      "epoch": 0.4754012416647505,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.7027724064672088e-05,
+      "loss": 0.8048,
+      "step": 6461
+    },
+    {
+      "epoch": 0.4754748217981145,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.7021951497479003e-05,
+      "loss": 0.6476,
+      "step": 6462
+    },
+    {
+      "epoch": 0.4755484019314785,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.7016178821771586e-05,
+      "loss": 1.1329,
+      "step": 6463
+    },
+    {
+      "epoch": 0.4756219820648425,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.7010406037859663e-05,
+      "loss": 0.6575,
+      "step": 6464
+    },
+    {
+      "epoch": 0.47569556219820647,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.7004633146053043e-05,
+      "loss": 0.8041,
+      "step": 6465
+    },
+    {
+      "epoch": 0.47576914233157047,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.699886014666156e-05,
+      "loss": 1.2123,
+      "step": 6466
+    },
+    {
+      "epoch": 0.4758427224649345,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.6993087039995017e-05,
+      "loss": 0.9338,
+      "step": 6467
+    },
+    {
+      "epoch": 0.4759163025982985,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.6987313826363264e-05,
+      "loss": 1.7897,
+      "step": 6468
+    },
+    {
+      "epoch": 0.47598988273166243,
+      "grad_norm": 1.0,
+      "learning_rate": 2.6981540506076126e-05,
+      "loss": 1.1717,
+      "step": 6469
+    },
+    {
+      "epoch": 0.47606346286502643,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.6975767079443454e-05,
+      "loss": 1.0878,
+      "step": 6470
+    },
+    {
+      "epoch": 0.47613704299839044,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.696999354677509e-05,
+      "loss": 0.8289,
+      "step": 6471
+    },
+    {
+      "epoch": 0.47621062313175444,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.6964219908380895e-05,
+      "loss": 0.9862,
+      "step": 6472
+    },
+    {
+      "epoch": 0.47628420326511844,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.6958446164570734e-05,
+      "loss": 0.7615,
+      "step": 6473
+    },
+    {
+      "epoch": 0.4763577833984824,
+      "grad_norm": 1.2421875,
+      "learning_rate": 2.6952672315654458e-05,
+      "loss": 1.434,
+      "step": 6474
+    },
+    {
+      "epoch": 0.4764313635318464,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.694689836194195e-05,
+      "loss": 1.4117,
+      "step": 6475
+    },
+    {
+      "epoch": 0.4765049436652104,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.694112430374309e-05,
+      "loss": 0.7745,
+      "step": 6476
+    },
+    {
+      "epoch": 0.4765785237985744,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.6935350141367742e-05,
+      "loss": 0.9303,
+      "step": 6477
+    },
+    {
+      "epoch": 0.47665210393193835,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.6929575875125817e-05,
+      "loss": 0.9707,
+      "step": 6478
+    },
+    {
+      "epoch": 0.47672568406530236,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.6923801505327196e-05,
+      "loss": 0.8115,
+      "step": 6479
+    },
+    {
+      "epoch": 0.47679926419866636,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.6918027032281784e-05,
+      "loss": 0.6961,
+      "step": 6480
+    },
+    {
+      "epoch": 0.47687284433203037,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.691225245629948e-05,
+      "loss": 0.7492,
+      "step": 6481
+    },
+    {
+      "epoch": 0.47694642446539437,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.690647777769021e-05,
+      "loss": 1.1451,
+      "step": 6482
+    },
+    {
+      "epoch": 0.4770200045987583,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.6900702996763866e-05,
+      "loss": 0.9732,
+      "step": 6483
+    },
+    {
+      "epoch": 0.4770935847321223,
+      "grad_norm": 1.125,
+      "learning_rate": 2.689492811383038e-05,
+      "loss": 1.2119,
+      "step": 6484
+    },
+    {
+      "epoch": 0.4771671648654863,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.688915312919969e-05,
+      "loss": 0.5543,
+      "step": 6485
+    },
+    {
+      "epoch": 0.47724074499885033,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.6883378043181713e-05,
+      "loss": 1.4937,
+      "step": 6486
+    },
+    {
+      "epoch": 0.4773143251322143,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.68776028560864e-05,
+      "loss": 1.0516,
+      "step": 6487
+    },
+    {
+      "epoch": 0.4773879052655783,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.6871827568223684e-05,
+      "loss": 1.3265,
+      "step": 6488
+    },
+    {
+      "epoch": 0.4774614853989423,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.6866052179903522e-05,
+      "loss": 1.0175,
+      "step": 6489
+    },
+    {
+      "epoch": 0.4775350655323063,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.686027669143586e-05,
+      "loss": 1.1962,
+      "step": 6490
+    },
+    {
+      "epoch": 0.4776086456656703,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.6854501103130657e-05,
+      "loss": 0.9129,
+      "step": 6491
+    },
+    {
+      "epoch": 0.47768222579903424,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.6848725415297887e-05,
+      "loss": 0.8448,
+      "step": 6492
+    },
+    {
+      "epoch": 0.47775580593239825,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.6842949628247516e-05,
+      "loss": 1.237,
+      "step": 6493
+    },
+    {
+      "epoch": 0.47782938606576225,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.6837173742289524e-05,
+      "loss": 0.9473,
+      "step": 6494
+    },
+    {
+      "epoch": 0.47790296619912626,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.683139775773388e-05,
+      "loss": 0.8712,
+      "step": 6495
+    },
+    {
+      "epoch": 0.4779765463324902,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.6825621674890584e-05,
+      "loss": 0.7254,
+      "step": 6496
+    },
+    {
+      "epoch": 0.4780501264658542,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.681984549406962e-05,
+      "loss": 1.2051,
+      "step": 6497
+    },
+    {
+      "epoch": 0.4781237065992182,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.6814069215580978e-05,
+      "loss": 0.8268,
+      "step": 6498
+    },
+    {
+      "epoch": 0.4781972867325822,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.680829283973467e-05,
+      "loss": 0.9537,
+      "step": 6499
+    },
+    {
+      "epoch": 0.4782708668659462,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.6802516366840706e-05,
+      "loss": 0.7822,
+      "step": 6500
+    },
+    {
+      "epoch": 0.47834444699931017,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.6796739797209087e-05,
+      "loss": 0.791,
+      "step": 6501
+    },
+    {
+      "epoch": 0.4784180271326742,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.679096313114984e-05,
+      "loss": 1.1303,
+      "step": 6502
+    },
+    {
+      "epoch": 0.4784916072660382,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.6785186368972987e-05,
+      "loss": 1.2051,
+      "step": 6503
+    },
+    {
+      "epoch": 0.4785651873994022,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.677940951098855e-05,
+      "loss": 1.0447,
+      "step": 6504
+    },
+    {
+      "epoch": 0.47863876753276613,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.677363255750656e-05,
+      "loss": 0.9259,
+      "step": 6505
+    },
+    {
+      "epoch": 0.47871234766613013,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.676785550883707e-05,
+      "loss": 0.8799,
+      "step": 6506
+    },
+    {
+      "epoch": 0.47878592779949414,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.676207836529011e-05,
+      "loss": 1.212,
+      "step": 6507
+    },
+    {
+      "epoch": 0.47885950793285814,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.6756301127175732e-05,
+      "loss": 0.8355,
+      "step": 6508
+    },
+    {
+      "epoch": 0.47893308806622215,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.6750523794803988e-05,
+      "loss": 0.687,
+      "step": 6509
+    },
+    {
+      "epoch": 0.4790066681995861,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.674474636848494e-05,
+      "loss": 0.7383,
+      "step": 6510
+    },
+    {
+      "epoch": 0.4790802483329501,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.6738968848528647e-05,
+      "loss": 0.742,
+      "step": 6511
+    },
+    {
+      "epoch": 0.4791538284663141,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.6733191235245185e-05,
+      "loss": 0.6763,
+      "step": 6512
+    },
+    {
+      "epoch": 0.4792274085996781,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.672741352894462e-05,
+      "loss": 0.8088,
+      "step": 6513
+    },
+    {
+      "epoch": 0.47930098873304205,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.672163572993704e-05,
+      "loss": 1.2608,
+      "step": 6514
+    },
+    {
+      "epoch": 0.47937456886640606,
+      "grad_norm": 1.21875,
+      "learning_rate": 2.6715857838532516e-05,
+      "loss": 1.5293,
+      "step": 6515
+    },
+    {
+      "epoch": 0.47944814899977006,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.6710079855041142e-05,
+      "loss": 0.7698,
+      "step": 6516
+    },
+    {
+      "epoch": 0.47952172913313407,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.6704301779773016e-05,
+      "loss": 0.8382,
+      "step": 6517
+    },
+    {
+      "epoch": 0.47959530926649807,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.6698523613038223e-05,
+      "loss": 1.1378,
+      "step": 6518
+    },
+    {
+      "epoch": 0.479668889399862,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.6692745355146887e-05,
+      "loss": 1.0447,
+      "step": 6519
+    },
+    {
+      "epoch": 0.479742469533226,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.66869670064091e-05,
+      "loss": 0.9409,
+      "step": 6520
+    },
+    {
+      "epoch": 0.47981604966659,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.668118856713498e-05,
+      "loss": 0.7465,
+      "step": 6521
+    },
+    {
+      "epoch": 0.47988962979995403,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.6675410037634646e-05,
+      "loss": 0.6655,
+      "step": 6522
+    },
+    {
+      "epoch": 0.479963209933318,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.6669631418218223e-05,
+      "loss": 0.7218,
+      "step": 6523
+    },
+    {
+      "epoch": 0.480036790066682,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.666385270919583e-05,
+      "loss": 0.81,
+      "step": 6524
+    },
+    {
+      "epoch": 0.480110370200046,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.6658073910877603e-05,
+      "loss": 0.8634,
+      "step": 6525
+    },
+    {
+      "epoch": 0.48018395033341,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.665229502357368e-05,
+      "loss": 0.8237,
+      "step": 6526
+    },
+    {
+      "epoch": 0.480257530466774,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.6646516047594206e-05,
+      "loss": 0.9938,
+      "step": 6527
+    },
+    {
+      "epoch": 0.48033111060013794,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.6640736983249332e-05,
+      "loss": 0.7124,
+      "step": 6528
+    },
+    {
+      "epoch": 0.48040469073350195,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.66349578308492e-05,
+      "loss": 0.9326,
+      "step": 6529
+    },
+    {
+      "epoch": 0.48047827086686595,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.6629178590703968e-05,
+      "loss": 1.2848,
+      "step": 6530
+    },
+    {
+      "epoch": 0.48055185100022996,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.6623399263123792e-05,
+      "loss": 1.0676,
+      "step": 6531
+    },
+    {
+      "epoch": 0.4806254311335939,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.6617619848418852e-05,
+      "loss": 0.6388,
+      "step": 6532
+    },
+    {
+      "epoch": 0.4806990112669579,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.661184034689931e-05,
+      "loss": 0.8049,
+      "step": 6533
+    },
+    {
+      "epoch": 0.4807725914003219,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.6606060758875333e-05,
+      "loss": 1.0493,
+      "step": 6534
+    },
+    {
+      "epoch": 0.4808461715336859,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.660028108465712e-05,
+      "loss": 0.8408,
+      "step": 6535
+    },
+    {
+      "epoch": 0.4809197516670499,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.6594501324554833e-05,
+      "loss": 1.2574,
+      "step": 6536
+    },
+    {
+      "epoch": 0.48099333180041387,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.6588721478878682e-05,
+      "loss": 1.0273,
+      "step": 6537
+    },
+    {
+      "epoch": 0.4810669119337779,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.6582941547938832e-05,
+      "loss": 0.9924,
+      "step": 6538
+    },
+    {
+      "epoch": 0.4811404920671419,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.6577161532045515e-05,
+      "loss": 0.9943,
+      "step": 6539
+    },
+    {
+      "epoch": 0.4812140722005059,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.6571381431508913e-05,
+      "loss": 1.3333,
+      "step": 6540
+    },
+    {
+      "epoch": 0.48128765233386983,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.6565601246639245e-05,
+      "loss": 0.9713,
+      "step": 6541
+    },
+    {
+      "epoch": 0.48136123246723383,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.6559820977746703e-05,
+      "loss": 0.7901,
+      "step": 6542
+    },
+    {
+      "epoch": 0.48143481260059784,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.655404062514152e-05,
+      "loss": 0.9806,
+      "step": 6543
+    },
+    {
+      "epoch": 0.48150839273396184,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.6548260189133904e-05,
+      "loss": 0.6986,
+      "step": 6544
+    },
+    {
+      "epoch": 0.48158197286732585,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.6542479670034098e-05,
+      "loss": 1.0739,
+      "step": 6545
+    },
+    {
+      "epoch": 0.4816555530006898,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.6536699068152322e-05,
+      "loss": 0.6641,
+      "step": 6546
+    },
+    {
+      "epoch": 0.4817291331340538,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.6530918383798804e-05,
+      "loss": 1.1925,
+      "step": 6547
+    },
+    {
+      "epoch": 0.4818027132674178,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.6525137617283797e-05,
+      "loss": 0.7034,
+      "step": 6548
+    },
+    {
+      "epoch": 0.4818762934007818,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.6519356768917524e-05,
+      "loss": 0.6368,
+      "step": 6549
+    },
+    {
+      "epoch": 0.48194987353414576,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.6513575839010246e-05,
+      "loss": 0.8171,
+      "step": 6550
+    },
+    {
+      "epoch": 0.48202345366750976,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.6507794827872212e-05,
+      "loss": 0.9612,
+      "step": 6551
+    },
+    {
+      "epoch": 0.48209703380087376,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.6502013735813676e-05,
+      "loss": 0.8156,
+      "step": 6552
+    },
+    {
+      "epoch": 0.48217061393423777,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.64962325631449e-05,
+      "loss": 0.9645,
+      "step": 6553
+    },
+    {
+      "epoch": 0.4822441940676018,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.649045131017615e-05,
+      "loss": 1.0374,
+      "step": 6554
+    },
+    {
+      "epoch": 0.4823177742009657,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.6484669977217696e-05,
+      "loss": 0.697,
+      "step": 6555
+    },
+    {
+      "epoch": 0.4823913543343297,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.6478888564579808e-05,
+      "loss": 0.7014,
+      "step": 6556
+    },
+    {
+      "epoch": 0.48246493446769373,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.647310707257276e-05,
+      "loss": 1.2796,
+      "step": 6557
+    },
+    {
+      "epoch": 0.48253851460105773,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.6467325501506834e-05,
+      "loss": 0.9344,
+      "step": 6558
+    },
+    {
+      "epoch": 0.4826120947344217,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.646154385169232e-05,
+      "loss": 0.6295,
+      "step": 6559
+    },
+    {
+      "epoch": 0.4826856748677857,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.6455762123439522e-05,
+      "loss": 0.8593,
+      "step": 6560
+    },
+    {
+      "epoch": 0.4827592550011497,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.6449980317058708e-05,
+      "loss": 0.743,
+      "step": 6561
+    },
+    {
+      "epoch": 0.4828328351345137,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.6444198432860197e-05,
+      "loss": 1.0652,
+      "step": 6562
+    },
+    {
+      "epoch": 0.4829064152678777,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.6438416471154275e-05,
+      "loss": 0.6952,
+      "step": 6563
+    },
+    {
+      "epoch": 0.48297999540124165,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.643263443225126e-05,
+      "loss": 1.2082,
+      "step": 6564
+    },
+    {
+      "epoch": 0.48305357553460565,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.6426852316461465e-05,
+      "loss": 0.8351,
+      "step": 6565
+    },
+    {
+      "epoch": 0.48312715566796965,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.6421070124095194e-05,
+      "loss": 0.7706,
+      "step": 6566
+    },
+    {
+      "epoch": 0.48320073580133366,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.6415287855462784e-05,
+      "loss": 0.6269,
+      "step": 6567
+    },
+    {
+      "epoch": 0.4832743159346976,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.6409505510874538e-05,
+      "loss": 1.1887,
+      "step": 6568
+    },
+    {
+      "epoch": 0.4833478960680616,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.64037230906408e-05,
+      "loss": 0.6661,
+      "step": 6569
+    },
+    {
+      "epoch": 0.4834214762014256,
+      "grad_norm": 0.875,
+      "learning_rate": 2.639794059507189e-05,
+      "loss": 0.8258,
+      "step": 6570
+    },
+    {
+      "epoch": 0.4834950563347896,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.6392158024478154e-05,
+      "loss": 1.0068,
+      "step": 6571
+    },
+    {
+      "epoch": 0.4835686364681536,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.6386375379169918e-05,
+      "loss": 1.1187,
+      "step": 6572
+    },
+    {
+      "epoch": 0.48364221660151757,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.6380592659457543e-05,
+      "loss": 0.9549,
+      "step": 6573
+    },
+    {
+      "epoch": 0.4837157967348816,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.6374809865651363e-05,
+      "loss": 1.1353,
+      "step": 6574
+    },
+    {
+      "epoch": 0.4837893768682456,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.6369026998061736e-05,
+      "loss": 1.303,
+      "step": 6575
+    },
+    {
+      "epoch": 0.4838629570016096,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.6363244056999014e-05,
+      "loss": 1.2945,
+      "step": 6576
+    },
+    {
+      "epoch": 0.48393653713497353,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.635746104277356e-05,
+      "loss": 1.0926,
+      "step": 6577
+    },
+    {
+      "epoch": 0.48401011726833754,
+      "grad_norm": 0.875,
+      "learning_rate": 2.6351677955695732e-05,
+      "loss": 0.8453,
+      "step": 6578
+    },
+    {
+      "epoch": 0.48408369740170154,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.6345894796075904e-05,
+      "loss": 0.7784,
+      "step": 6579
+    },
+    {
+      "epoch": 0.48415727753506554,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.6340111564224447e-05,
+      "loss": 1.159,
+      "step": 6580
+    },
+    {
+      "epoch": 0.48423085766842955,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.6334328260451728e-05,
+      "loss": 0.9067,
+      "step": 6581
+    },
+    {
+      "epoch": 0.4843044378017935,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.6328544885068136e-05,
+      "loss": 0.8354,
+      "step": 6582
+    },
+    {
+      "epoch": 0.4843780179351575,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.6322761438384048e-05,
+      "loss": 0.6787,
+      "step": 6583
+    },
+    {
+      "epoch": 0.4844515980685215,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.631697792070985e-05,
+      "loss": 1.0695,
+      "step": 6584
+    },
+    {
+      "epoch": 0.4845251782018855,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.631119433235593e-05,
+      "loss": 0.9824,
+      "step": 6585
+    },
+    {
+      "epoch": 0.48459875833524946,
+      "grad_norm": 1.0,
+      "learning_rate": 2.630541067363269e-05,
+      "loss": 1.0727,
+      "step": 6586
+    },
+    {
+      "epoch": 0.48467233846861346,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.6299626944850537e-05,
+      "loss": 0.7621,
+      "step": 6587
+    },
+    {
+      "epoch": 0.48474591860197747,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.6293843146319846e-05,
+      "loss": 0.7323,
+      "step": 6588
+    },
+    {
+      "epoch": 0.48481949873534147,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.6288059278351047e-05,
+      "loss": 0.9427,
+      "step": 6589
+    },
+    {
+      "epoch": 0.4848930788687055,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.628227534125453e-05,
+      "loss": 1.0747,
+      "step": 6590
+    },
+    {
+      "epoch": 0.4849666590020694,
+      "grad_norm": 0.64453125,
+      "learning_rate": 2.627649133534072e-05,
+      "loss": 0.6269,
+      "step": 6591
+    },
+    {
+      "epoch": 0.4850402391354334,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.627070726092003e-05,
+      "loss": 1.057,
+      "step": 6592
+    },
+    {
+      "epoch": 0.48511381926879743,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.6264923118302882e-05,
+      "loss": 0.838,
+      "step": 6593
+    },
+    {
+      "epoch": 0.48518739940216143,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.625913890779971e-05,
+      "loss": 0.9063,
+      "step": 6594
+    },
+    {
+      "epoch": 0.4852609795355254,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.6253354629720918e-05,
+      "loss": 0.9571,
+      "step": 6595
+    },
+    {
+      "epoch": 0.4853345596688894,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.6247570284376954e-05,
+      "loss": 0.859,
+      "step": 6596
+    },
+    {
+      "epoch": 0.4854081398022534,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.624178587207825e-05,
+      "loss": 1.0043,
+      "step": 6597
+    },
+    {
+      "epoch": 0.4854817199356174,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.623600139313524e-05,
+      "loss": 0.5801,
+      "step": 6598
+    },
+    {
+      "epoch": 0.4855553000689814,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.623021684785837e-05,
+      "loss": 0.8346,
+      "step": 6599
+    },
+    {
+      "epoch": 0.48562888020234535,
+      "grad_norm": 1.5234375,
+      "learning_rate": 2.6224432236558088e-05,
+      "loss": 0.663,
+      "step": 6600
+    },
+    {
+      "epoch": 0.48570246033570935,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.6218647559544844e-05,
+      "loss": 1.0242,
+      "step": 6601
+    },
+    {
+      "epoch": 0.48577604046907336,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.621286281712908e-05,
+      "loss": 0.8787,
+      "step": 6602
+    },
+    {
+      "epoch": 0.48584962060243736,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.6207078009621266e-05,
+      "loss": 0.6553,
+      "step": 6603
+    },
+    {
+      "epoch": 0.4859232007358013,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.620129313733185e-05,
+      "loss": 0.7066,
+      "step": 6604
+    },
+    {
+      "epoch": 0.4859967808691653,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.6195508200571305e-05,
+      "loss": 0.9585,
+      "step": 6605
+    },
+    {
+      "epoch": 0.4860703610025293,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.6189723199650084e-05,
+      "loss": 1.4974,
+      "step": 6606
+    },
+    {
+      "epoch": 0.4861439411358933,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.6183938134878673e-05,
+      "loss": 1.0687,
+      "step": 6607
+    },
+    {
+      "epoch": 0.4862175212692573,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.6178153006567542e-05,
+      "loss": 0.5521,
+      "step": 6608
+    },
+    {
+      "epoch": 0.4862911014026213,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.617236781502716e-05,
+      "loss": 1.0084,
+      "step": 6609
+    },
+    {
+      "epoch": 0.4863646815359853,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.6166582560568016e-05,
+      "loss": 0.8275,
+      "step": 6610
+    },
+    {
+      "epoch": 0.4864382616693493,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.616079724350058e-05,
+      "loss": 0.702,
+      "step": 6611
+    },
+    {
+      "epoch": 0.4865118418027133,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.6155011864135355e-05,
+      "loss": 0.8587,
+      "step": 6612
+    },
+    {
+      "epoch": 0.48658542193607723,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.614922642278282e-05,
+      "loss": 0.635,
+      "step": 6613
+    },
+    {
+      "epoch": 0.48665900206944124,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.614344091975347e-05,
+      "loss": 1.03,
+      "step": 6614
+    },
+    {
+      "epoch": 0.48673258220280524,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.6137655355357822e-05,
+      "loss": 1.0933,
+      "step": 6615
+    },
+    {
+      "epoch": 0.48680616233616925,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.6131869729906344e-05,
+      "loss": 0.9497,
+      "step": 6616
+    },
+    {
+      "epoch": 0.48687974246953325,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.6126084043709563e-05,
+      "loss": 0.6989,
+      "step": 6617
+    },
+    {
+      "epoch": 0.4869533226028972,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.6120298297077978e-05,
+      "loss": 0.7913,
+      "step": 6618
+    },
+    {
+      "epoch": 0.4870269027362612,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.6114512490322096e-05,
+      "loss": 0.6741,
+      "step": 6619
+    },
+    {
+      "epoch": 0.4871004828696252,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.6108726623752434e-05,
+      "loss": 0.7712,
+      "step": 6620
+    },
+    {
+      "epoch": 0.4871740630029892,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.6102940697679514e-05,
+      "loss": 0.8339,
+      "step": 6621
+    },
+    {
+      "epoch": 0.48724764313635316,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.609715471241384e-05,
+      "loss": 0.8066,
+      "step": 6622
+    },
+    {
+      "epoch": 0.48732122326971716,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.6091368668265954e-05,
+      "loss": 1.4414,
+      "step": 6623
+    },
+    {
+      "epoch": 0.48739480340308117,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.608558256554637e-05,
+      "loss": 1.1509,
+      "step": 6624
+    },
+    {
+      "epoch": 0.48746838353644517,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.6079796404565622e-05,
+      "loss": 1.0638,
+      "step": 6625
+    },
+    {
+      "epoch": 0.4875419636698092,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.6074010185634235e-05,
+      "loss": 1.0043,
+      "step": 6626
+    },
+    {
+      "epoch": 0.4876155438031731,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.606822390906275e-05,
+      "loss": 1.0775,
+      "step": 6627
+    },
+    {
+      "epoch": 0.48768912393653713,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.6062437575161708e-05,
+      "loss": 1.2867,
+      "step": 6628
+    },
+    {
+      "epoch": 0.48776270406990113,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.605665118424165e-05,
+      "loss": 1.2731,
+      "step": 6629
+    },
+    {
+      "epoch": 0.48783628420326514,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.6050864736613112e-05,
+      "loss": 0.72,
+      "step": 6630
+    },
+    {
+      "epoch": 0.4879098643366291,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.6045078232586657e-05,
+      "loss": 0.9182,
+      "step": 6631
+    },
+    {
+      "epoch": 0.4879834444699931,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.603929167247282e-05,
+      "loss": 1.1477,
+      "step": 6632
+    },
+    {
+      "epoch": 0.4880570246033571,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.6033505056582164e-05,
+      "loss": 1.0072,
+      "step": 6633
+    },
+    {
+      "epoch": 0.4881306047367211,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.6027718385225245e-05,
+      "loss": 0.8941,
+      "step": 6634
+    },
+    {
+      "epoch": 0.4882041848700851,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.6021931658712624e-05,
+      "loss": 0.7858,
+      "step": 6635
+    },
+    {
+      "epoch": 0.48827776500344905,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.601614487735486e-05,
+      "loss": 0.5726,
+      "step": 6636
+    },
+    {
+      "epoch": 0.48835134513681305,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.601035804146252e-05,
+      "loss": 0.9406,
+      "step": 6637
+    },
+    {
+      "epoch": 0.48842492527017706,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.600457115134617e-05,
+      "loss": 1.0871,
+      "step": 6638
+    },
+    {
+      "epoch": 0.48849850540354106,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.5998784207316378e-05,
+      "loss": 0.9156,
+      "step": 6639
+    },
+    {
+      "epoch": 0.488572085536905,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.599299720968373e-05,
+      "loss": 0.835,
+      "step": 6640
+    },
+    {
+      "epoch": 0.488645665670269,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.5987210158758797e-05,
+      "loss": 0.8283,
+      "step": 6641
+    },
+    {
+      "epoch": 0.488719245803633,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.598142305485216e-05,
+      "loss": 0.8809,
+      "step": 6642
+    },
+    {
+      "epoch": 0.488792825936997,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.5975635898274397e-05,
+      "loss": 0.689,
+      "step": 6643
+    },
+    {
+      "epoch": 0.488866406070361,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.596984868933611e-05,
+      "loss": 0.8381,
+      "step": 6644
+    },
+    {
+      "epoch": 0.488939986203725,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.5964061428347862e-05,
+      "loss": 0.8655,
+      "step": 6645
+    },
+    {
+      "epoch": 0.489013566337089,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.5958274115620262e-05,
+      "loss": 1.0838,
+      "step": 6646
+    },
+    {
+      "epoch": 0.489087146470453,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.59524867514639e-05,
+      "loss": 1.3738,
+      "step": 6647
+    },
+    {
+      "epoch": 0.489160726603817,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.5946699336189373e-05,
+      "loss": 1.0237,
+      "step": 6648
+    },
+    {
+      "epoch": 0.48923430673718094,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.5940911870107288e-05,
+      "loss": 0.9828,
+      "step": 6649
+    },
+    {
+      "epoch": 0.48930788687054494,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.593512435352823e-05,
+      "loss": 1.157,
+      "step": 6650
+    },
+    {
+      "epoch": 0.48938146700390894,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.592933678676282e-05,
+      "loss": 0.7981,
+      "step": 6651
+    },
+    {
+      "epoch": 0.48945504713727295,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.592354917012166e-05,
+      "loss": 1.1765,
+      "step": 6652
+    },
+    {
+      "epoch": 0.48952862727063695,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.5917761503915355e-05,
+      "loss": 1.0321,
+      "step": 6653
+    },
+    {
+      "epoch": 0.4896022074040009,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.5911973788454525e-05,
+      "loss": 0.7883,
+      "step": 6654
+    },
+    {
+      "epoch": 0.4896757875373649,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.5906186024049784e-05,
+      "loss": 0.9707,
+      "step": 6655
+    },
+    {
+      "epoch": 0.4897493676707289,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.5900398211011755e-05,
+      "loss": 0.7421,
+      "step": 6656
+    },
+    {
+      "epoch": 0.4898229478040929,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.5894610349651055e-05,
+      "loss": 0.833,
+      "step": 6657
+    },
+    {
+      "epoch": 0.48989652793745686,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.588882244027831e-05,
+      "loss": 0.6903,
+      "step": 6658
+    },
+    {
+      "epoch": 0.48997010807082086,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.5883034483204133e-05,
+      "loss": 0.9731,
+      "step": 6659
+    },
+    {
+      "epoch": 0.49004368820418487,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.5877246478739165e-05,
+      "loss": 1.0695,
+      "step": 6660
+    },
+    {
+      "epoch": 0.4901172683375489,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.587145842719404e-05,
+      "loss": 0.5633,
+      "step": 6661
+    },
+    {
+      "epoch": 0.4901908484709129,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.5865670328879387e-05,
+      "loss": 0.8886,
+      "step": 6662
+    },
+    {
+      "epoch": 0.4902644286042768,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.5859882184105844e-05,
+      "loss": 1.2564,
+      "step": 6663
+    },
+    {
+      "epoch": 0.49033800873764083,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.585409399318404e-05,
+      "loss": 0.7831,
+      "step": 6664
+    },
+    {
+      "epoch": 0.49041158887100483,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.5848305756424635e-05,
+      "loss": 1.1321,
+      "step": 6665
+    },
+    {
+      "epoch": 0.49048516900436884,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.5842517474138254e-05,
+      "loss": 1.2362,
+      "step": 6666
+    },
+    {
+      "epoch": 0.4905587491377328,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.583672914663555e-05,
+      "loss": 0.9461,
+      "step": 6667
+    },
+    {
+      "epoch": 0.4906323292710968,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.583094077422718e-05,
+      "loss": 1.2423,
+      "step": 6668
+    },
+    {
+      "epoch": 0.4907059094044608,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.5825152357223792e-05,
+      "loss": 0.6962,
+      "step": 6669
+    },
+    {
+      "epoch": 0.4907794895378248,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.5819363895936026e-05,
+      "loss": 0.946,
+      "step": 6670
+    },
+    {
+      "epoch": 0.4908530696711888,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.5813575390674548e-05,
+      "loss": 0.9372,
+      "step": 6671
+    },
+    {
+      "epoch": 0.49092664980455275,
+      "grad_norm": 0.875,
+      "learning_rate": 2.5807786841750025e-05,
+      "loss": 0.9871,
+      "step": 6672
+    },
+    {
+      "epoch": 0.49100022993791675,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.5801998249473096e-05,
+      "loss": 0.7371,
+      "step": 6673
+    },
+    {
+      "epoch": 0.49107381007128076,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.5796209614154442e-05,
+      "loss": 1.0865,
+      "step": 6674
+    },
+    {
+      "epoch": 0.49114739020464476,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.5790420936104714e-05,
+      "loss": 0.8629,
+      "step": 6675
+    },
+    {
+      "epoch": 0.4912209703380087,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.5784632215634603e-05,
+      "loss": 0.7611,
+      "step": 6676
+    },
+    {
+      "epoch": 0.4912945504713727,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.577884345305475e-05,
+      "loss": 1.1233,
+      "step": 6677
+    },
+    {
+      "epoch": 0.4913681306047367,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.577305464867585e-05,
+      "loss": 0.9199,
+      "step": 6678
+    },
+    {
+      "epoch": 0.4914417107381007,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.5767265802808555e-05,
+      "loss": 0.9293,
+      "step": 6679
+    },
+    {
+      "epoch": 0.4915152908714647,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.576147691576356e-05,
+      "loss": 1.1749,
+      "step": 6680
+    },
+    {
+      "epoch": 0.4915888710048287,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.5755687987851534e-05,
+      "loss": 0.9712,
+      "step": 6681
+    },
+    {
+      "epoch": 0.4916624511381927,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.5749899019383163e-05,
+      "loss": 1.1325,
+      "step": 6682
+    },
+    {
+      "epoch": 0.4917360312715567,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.5744110010669133e-05,
+      "loss": 1.0317,
+      "step": 6683
+    },
+    {
+      "epoch": 0.4918096114049207,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.573832096202011e-05,
+      "loss": 0.7013,
+      "step": 6684
+    },
+    {
+      "epoch": 0.49188319153828464,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.573253187374681e-05,
+      "loss": 0.9334,
+      "step": 6685
+    },
+    {
+      "epoch": 0.49195677167164864,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.5726742746159906e-05,
+      "loss": 1.036,
+      "step": 6686
+    },
+    {
+      "epoch": 0.49203035180501264,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.5720953579570078e-05,
+      "loss": 0.9906,
+      "step": 6687
+    },
+    {
+      "epoch": 0.49210393193837665,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.571516437428805e-05,
+      "loss": 0.7805,
+      "step": 6688
+    },
+    {
+      "epoch": 0.49217751207174065,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.570937513062449e-05,
+      "loss": 0.7876,
+      "step": 6689
+    },
+    {
+      "epoch": 0.4922510922051046,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.5703585848890116e-05,
+      "loss": 0.5749,
+      "step": 6690
+    },
+    {
+      "epoch": 0.4923246723384686,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.569779652939561e-05,
+      "loss": 0.7909,
+      "step": 6691
+    },
+    {
+      "epoch": 0.4923982524718326,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.5692007172451693e-05,
+      "loss": 0.8743,
+      "step": 6692
+    },
+    {
+      "epoch": 0.4924718326051966,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.568621777836905e-05,
+      "loss": 0.7506,
+      "step": 6693
+    },
+    {
+      "epoch": 0.49254541273856056,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.568042834745839e-05,
+      "loss": 0.89,
+      "step": 6694
+    },
+    {
+      "epoch": 0.49261899287192457,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.5674638880030427e-05,
+      "loss": 0.841,
+      "step": 6695
+    },
+    {
+      "epoch": 0.49269257300528857,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.5668849376395876e-05,
+      "loss": 1.1024,
+      "step": 6696
+    },
+    {
+      "epoch": 0.4927661531386526,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.5663059836865444e-05,
+      "loss": 0.8662,
+      "step": 6697
+    },
+    {
+      "epoch": 0.4928397332720166,
+      "grad_norm": 0.875,
+      "learning_rate": 2.5657270261749834e-05,
+      "loss": 0.7646,
+      "step": 6698
+    },
+    {
+      "epoch": 0.4929133134053805,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.5651480651359777e-05,
+      "loss": 1.05,
+      "step": 6699
+    },
+    {
+      "epoch": 0.49298689353874453,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.5645691006005985e-05,
+      "loss": 1.0519,
+      "step": 6700
+    },
+    {
+      "epoch": 0.49306047367210853,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.563990132599917e-05,
+      "loss": 1.1284,
+      "step": 6701
+    },
+    {
+      "epoch": 0.49313405380547254,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.5634111611650063e-05,
+      "loss": 0.6174,
+      "step": 6702
+    },
+    {
+      "epoch": 0.4932076339388365,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.5628321863269377e-05,
+      "loss": 1.2426,
+      "step": 6703
+    },
+    {
+      "epoch": 0.4932812140722005,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.5622532081167855e-05,
+      "loss": 1.1306,
+      "step": 6704
+    },
+    {
+      "epoch": 0.4933547942055645,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.5616742265656207e-05,
+      "loss": 0.9979,
+      "step": 6705
+    },
+    {
+      "epoch": 0.4934283743389285,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.561095241704517e-05,
+      "loss": 0.8668,
+      "step": 6706
+    },
+    {
+      "epoch": 0.4935019544722925,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.560516253564546e-05,
+      "loss": 0.6973,
+      "step": 6707
+    },
+    {
+      "epoch": 0.49357553460565645,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.5599372621767833e-05,
+      "loss": 1.0704,
+      "step": 6708
+    },
+    {
+      "epoch": 0.49364911473902046,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.5593582675723e-05,
+      "loss": 0.9875,
+      "step": 6709
+    },
+    {
+      "epoch": 0.49372269487238446,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.558779269782171e-05,
+      "loss": 0.9234,
+      "step": 6710
+    },
+    {
+      "epoch": 0.49379627500574846,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.558200268837469e-05,
+      "loss": 0.888,
+      "step": 6711
+    },
+    {
+      "epoch": 0.4938698551391124,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.5576212647692688e-05,
+      "loss": 0.6125,
+      "step": 6712
+    },
+    {
+      "epoch": 0.4939434352724764,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.557042257608645e-05,
+      "loss": 0.7742,
+      "step": 6713
+    },
+    {
+      "epoch": 0.4940170154058404,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.5564632473866694e-05,
+      "loss": 1.027,
+      "step": 6714
+    },
+    {
+      "epoch": 0.4940905955392044,
+      "grad_norm": 0.875,
+      "learning_rate": 2.5558842341344193e-05,
+      "loss": 1.3594,
+      "step": 6715
+    },
+    {
+      "epoch": 0.49416417567256843,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.555305217882967e-05,
+      "loss": 0.7544,
+      "step": 6716
+    },
+    {
+      "epoch": 0.4942377558059324,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.5547261986633892e-05,
+      "loss": 0.7817,
+      "step": 6717
+    },
+    {
+      "epoch": 0.4943113359392964,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.554147176506759e-05,
+      "loss": 1.3649,
+      "step": 6718
+    },
+    {
+      "epoch": 0.4943849160726604,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.553568151444152e-05,
+      "loss": 0.8064,
+      "step": 6719
+    },
+    {
+      "epoch": 0.4944584962060244,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.552989123506644e-05,
+      "loss": 1.2102,
+      "step": 6720
+    },
+    {
+      "epoch": 0.49453207633938834,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.5524100927253085e-05,
+      "loss": 1.3482,
+      "step": 6721
+    },
+    {
+      "epoch": 0.49460565647275234,
+      "grad_norm": 0.75,
+      "learning_rate": 2.551831059131224e-05,
+      "loss": 0.8116,
+      "step": 6722
+    },
+    {
+      "epoch": 0.49467923660611635,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.5512520227554637e-05,
+      "loss": 0.573,
+      "step": 6723
+    },
+    {
+      "epoch": 0.49475281673948035,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.5506729836291045e-05,
+      "loss": 0.8246,
+      "step": 6724
+    },
+    {
+      "epoch": 0.49482639687284435,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.5500939417832213e-05,
+      "loss": 0.8664,
+      "step": 6725
+    },
+    {
+      "epoch": 0.4948999770062083,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.549514897248892e-05,
+      "loss": 0.8978,
+      "step": 6726
+    },
+    {
+      "epoch": 0.4949735571395723,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.5489358500571896e-05,
+      "loss": 0.9654,
+      "step": 6727
+    },
+    {
+      "epoch": 0.4950471372729363,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.548356800239194e-05,
+      "loss": 1.2929,
+      "step": 6728
+    },
+    {
+      "epoch": 0.4951207174063003,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.5477777478259805e-05,
+      "loss": 0.8073,
+      "step": 6729
+    },
+    {
+      "epoch": 0.49519429753966426,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.5471986928486245e-05,
+      "loss": 0.9441,
+      "step": 6730
+    },
+    {
+      "epoch": 0.49526787767302827,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.5466196353382053e-05,
+      "loss": 1.1509,
+      "step": 6731
+    },
+    {
+      "epoch": 0.49534145780639227,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.5460405753257972e-05,
+      "loss": 1.2072,
+      "step": 6732
+    },
+    {
+      "epoch": 0.4954150379397563,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.545461512842479e-05,
+      "loss": 0.8782,
+      "step": 6733
+    },
+    {
+      "epoch": 0.4954886180731203,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.5448824479193262e-05,
+      "loss": 1.0232,
+      "step": 6734
+    },
+    {
+      "epoch": 0.49556219820648423,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.5443033805874177e-05,
+      "loss": 1.0602,
+      "step": 6735
+    },
+    {
+      "epoch": 0.49563577833984823,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.5437243108778307e-05,
+      "loss": 0.7239,
+      "step": 6736
+    },
+    {
+      "epoch": 0.49570935847321224,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.5431452388216427e-05,
+      "loss": 0.6611,
+      "step": 6737
+    },
+    {
+      "epoch": 0.49578293860657624,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.5425661644499315e-05,
+      "loss": 0.9489,
+      "step": 6738
+    },
+    {
+      "epoch": 0.4958565187399402,
+      "grad_norm": 0.75,
+      "learning_rate": 2.541987087793774e-05,
+      "loss": 0.5325,
+      "step": 6739
+    },
+    {
+      "epoch": 0.4959300988733042,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.5414080088842484e-05,
+      "loss": 1.1207,
+      "step": 6740
+    },
+    {
+      "epoch": 0.4960036790066682,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.540828927752434e-05,
+      "loss": 1.1418,
+      "step": 6741
+    },
+    {
+      "epoch": 0.4960772591400322,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.540249844429408e-05,
+      "loss": 0.8199,
+      "step": 6742
+    },
+    {
+      "epoch": 0.4961508392733962,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.5396707589462486e-05,
+      "loss": 0.7246,
+      "step": 6743
+    },
+    {
+      "epoch": 0.49622441940676015,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.5390916713340345e-05,
+      "loss": 0.943,
+      "step": 6744
+    },
+    {
+      "epoch": 0.49629799954012416,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.5385125816238448e-05,
+      "loss": 1.1701,
+      "step": 6745
+    },
+    {
+      "epoch": 0.49637157967348816,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.5379334898467565e-05,
+      "loss": 0.8514,
+      "step": 6746
+    },
+    {
+      "epoch": 0.49644515980685217,
+      "grad_norm": 1.1875,
+      "learning_rate": 2.5373543960338503e-05,
+      "loss": 1.1307,
+      "step": 6747
+    },
+    {
+      "epoch": 0.4965187399402161,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.5367753002162037e-05,
+      "loss": 0.6164,
+      "step": 6748
+    },
+    {
+      "epoch": 0.4965923200735801,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.536196202424897e-05,
+      "loss": 0.7547,
+      "step": 6749
+    },
+    {
+      "epoch": 0.4966659002069441,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.5356171026910074e-05,
+      "loss": 0.7357,
+      "step": 6750
+    },
+    {
+      "epoch": 0.4967394803403081,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.535038001045616e-05,
+      "loss": 1.8553,
+      "step": 6751
+    },
+    {
+      "epoch": 0.49681306047367213,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.5344588975198012e-05,
+      "loss": 0.8724,
+      "step": 6752
+    },
+    {
+      "epoch": 0.4968866406070361,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.5338797921446422e-05,
+      "loss": 1.2423,
+      "step": 6753
+    },
+    {
+      "epoch": 0.4969602207404001,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.5333006849512185e-05,
+      "loss": 1.4631,
+      "step": 6754
+    },
+    {
+      "epoch": 0.4970338008737641,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.53272157597061e-05,
+      "loss": 1.0206,
+      "step": 6755
+    },
+    {
+      "epoch": 0.4971073810071281,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.5321424652338972e-05,
+      "loss": 0.6425,
+      "step": 6756
+    },
+    {
+      "epoch": 0.4971809611404921,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.5315633527721583e-05,
+      "loss": 0.9942,
+      "step": 6757
+    },
+    {
+      "epoch": 0.49725454127385604,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.5309842386164744e-05,
+      "loss": 0.7313,
+      "step": 6758
+    },
+    {
+      "epoch": 0.49732812140722005,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.5304051227979243e-05,
+      "loss": 1.2216,
+      "step": 6759
+    },
+    {
+      "epoch": 0.49740170154058405,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.5298260053475885e-05,
+      "loss": 0.9072,
+      "step": 6760
+    },
+    {
+      "epoch": 0.49747528167394806,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.5292468862965472e-05,
+      "loss": 1.1626,
+      "step": 6761
+    },
+    {
+      "epoch": 0.497548861807312,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.5286677656758817e-05,
+      "loss": 0.6964,
+      "step": 6762
+    },
+    {
+      "epoch": 0.497622441940676,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.5280886435166713e-05,
+      "loss": 0.8919,
+      "step": 6763
+    },
+    {
+      "epoch": 0.49769602207404,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.5275095198499953e-05,
+      "loss": 1.1505,
+      "step": 6764
+    },
+    {
+      "epoch": 0.497769602207404,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.5269303947069368e-05,
+      "loss": 1.2343,
+      "step": 6765
+    },
+    {
+      "epoch": 0.497843182340768,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.5263512681185742e-05,
+      "loss": 1.6502,
+      "step": 6766
+    },
+    {
+      "epoch": 0.49791676247413197,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.5257721401159884e-05,
+      "loss": 1.0583,
+      "step": 6767
+    },
+    {
+      "epoch": 0.497990342607496,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.5251930107302608e-05,
+      "loss": 0.9987,
+      "step": 6768
+    },
+    {
+      "epoch": 0.49806392274086,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.5246138799924718e-05,
+      "loss": 0.9529,
+      "step": 6769
+    },
+    {
+      "epoch": 0.498137502874224,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.5240347479337027e-05,
+      "loss": 0.8805,
+      "step": 6770
+    },
+    {
+      "epoch": 0.49821108300758793,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.5234556145850343e-05,
+      "loss": 0.9518,
+      "step": 6771
+    },
+    {
+      "epoch": 0.49828466314095193,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.5228764799775468e-05,
+      "loss": 1.0451,
+      "step": 6772
+    },
+    {
+      "epoch": 0.49835824327431594,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.522297344142322e-05,
+      "loss": 0.9713,
+      "step": 6773
+    },
+    {
+      "epoch": 0.49843182340767994,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.5217182071104405e-05,
+      "loss": 0.8702,
+      "step": 6774
+    },
+    {
+      "epoch": 0.49850540354104395,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.5211390689129844e-05,
+      "loss": 1.041,
+      "step": 6775
+    },
+    {
+      "epoch": 0.4985789836744079,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.5205599295810338e-05,
+      "loss": 0.898,
+      "step": 6776
+    },
+    {
+      "epoch": 0.4986525638077719,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.5199807891456716e-05,
+      "loss": 0.6358,
+      "step": 6777
+    },
+    {
+      "epoch": 0.4987261439411359,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.5194016476379773e-05,
+      "loss": 0.7588,
+      "step": 6778
+    },
+    {
+      "epoch": 0.4987997240744999,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.518822505089034e-05,
+      "loss": 0.5863,
+      "step": 6779
+    },
+    {
+      "epoch": 0.49887330420786385,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.5182433615299215e-05,
+      "loss": 1.0457,
+      "step": 6780
+    },
+    {
+      "epoch": 0.49894688434122786,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.5176642169917225e-05,
+      "loss": 0.8337,
+      "step": 6781
+    },
+    {
+      "epoch": 0.49902046447459186,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.517085071505518e-05,
+      "loss": 1.1624,
+      "step": 6782
+    },
+    {
+      "epoch": 0.49909404460795587,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.5165059251023915e-05,
+      "loss": 0.7193,
+      "step": 6783
+    },
+    {
+      "epoch": 0.49916762474131987,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.515926777813422e-05,
+      "loss": 1.4729,
+      "step": 6784
+    },
+    {
+      "epoch": 0.4992412048746838,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.5153476296696922e-05,
+      "loss": 0.6558,
+      "step": 6785
+    },
+    {
+      "epoch": 0.4993147850080478,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.5147684807022847e-05,
+      "loss": 0.6489,
+      "step": 6786
+    },
+    {
+      "epoch": 0.4993883651414118,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.5141893309422804e-05,
+      "loss": 1.0101,
+      "step": 6787
+    },
+    {
+      "epoch": 0.49946194527477583,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.513610180420762e-05,
+      "loss": 1.3912,
+      "step": 6788
+    },
+    {
+      "epoch": 0.4995355254081398,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.5130310291688097e-05,
+      "loss": 0.9714,
+      "step": 6789
+    },
+    {
+      "epoch": 0.4996091055415038,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.512451877217508e-05,
+      "loss": 0.7537,
+      "step": 6790
+    },
+    {
+      "epoch": 0.4996826856748678,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.5118727245979372e-05,
+      "loss": 0.8053,
+      "step": 6791
+    },
+    {
+      "epoch": 0.4997562658082318,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.5112935713411796e-05,
+      "loss": 1.1749,
+      "step": 6792
+    },
+    {
+      "epoch": 0.4998298459415958,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.5107144174783174e-05,
+      "loss": 0.5942,
+      "step": 6793
+    },
+    {
+      "epoch": 0.49990342607495974,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.5101352630404324e-05,
+      "loss": 0.7524,
+      "step": 6794
+    },
+    {
+      "epoch": 0.49997700620832375,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.509556108058607e-05,
+      "loss": 0.8138,
+      "step": 6795
+    },
+    {
+      "epoch": 0.5000505863416878,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.5089769525639235e-05,
+      "loss": 0.6325,
+      "step": 6796
+    },
+    {
+      "epoch": 0.5001241664750518,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.5083977965874637e-05,
+      "loss": 0.8426,
+      "step": 6797
+    },
+    {
+      "epoch": 0.5001977466084158,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.5078186401603103e-05,
+      "loss": 0.6433,
+      "step": 6798
+    },
+    {
+      "epoch": 0.5002713267417798,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.507239483313546e-05,
+      "loss": 0.7694,
+      "step": 6799
+    },
+    {
+      "epoch": 0.5003449068751437,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.506660326078251e-05,
+      "loss": 0.6423,
+      "step": 6800
+    },
+    {
+      "epoch": 0.5004184870085077,
+      "grad_norm": 0.75,
+      "learning_rate": 2.506081168485509e-05,
+      "loss": 0.7475,
+      "step": 6801
+    },
+    {
+      "epoch": 0.5004920671418717,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.5055020105664024e-05,
+      "loss": 0.7871,
+      "step": 6802
+    },
+    {
+      "epoch": 0.5005656472752357,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.5049228523520135e-05,
+      "loss": 1.1883,
+      "step": 6803
+    },
+    {
+      "epoch": 0.5006392274085997,
+      "grad_norm": 0.875,
+      "learning_rate": 2.5043436938734245e-05,
+      "loss": 1.0403,
+      "step": 6804
+    },
+    {
+      "epoch": 0.5007128075419637,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.503764535161717e-05,
+      "loss": 1.0646,
+      "step": 6805
+    },
+    {
+      "epoch": 0.5007863876753277,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.503185376247975e-05,
+      "loss": 0.8878,
+      "step": 6806
+    },
+    {
+      "epoch": 0.5008599678086917,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.5026062171632796e-05,
+      "loss": 0.9242,
+      "step": 6807
+    },
+    {
+      "epoch": 0.5009335479420557,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.5020270579387127e-05,
+      "loss": 0.9098,
+      "step": 6808
+    },
+    {
+      "epoch": 0.5010071280754196,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.501447898605358e-05,
+      "loss": 0.9524,
+      "step": 6809
+    },
+    {
+      "epoch": 0.5010807082087836,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.500868739194297e-05,
+      "loss": 0.8059,
+      "step": 6810
+    },
+    {
+      "epoch": 0.5011542883421476,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.5002895797366134e-05,
+      "loss": 0.758,
+      "step": 6811
+    },
+    {
+      "epoch": 0.5012278684755116,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.4997104202633875e-05,
+      "loss": 0.7122,
+      "step": 6812
+    },
+    {
+      "epoch": 0.5013014486088756,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.499131260805703e-05,
+      "loss": 0.7517,
+      "step": 6813
+    },
+    {
+      "epoch": 0.5013750287422396,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.4985521013946425e-05,
+      "loss": 1.0842,
+      "step": 6814
+    },
+    {
+      "epoch": 0.5014486088756036,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.4979729420612875e-05,
+      "loss": 1.4459,
+      "step": 6815
+    },
+    {
+      "epoch": 0.5015221890089676,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.497393782836721e-05,
+      "loss": 1.1544,
+      "step": 6816
+    },
+    {
+      "epoch": 0.5015957691423316,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.496814623752025e-05,
+      "loss": 1.3312,
+      "step": 6817
+    },
+    {
+      "epoch": 0.5016693492756955,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.4962354648382827e-05,
+      "loss": 1.2026,
+      "step": 6818
+    },
+    {
+      "epoch": 0.5017429294090595,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.4956563061265764e-05,
+      "loss": 0.5985,
+      "step": 6819
+    },
+    {
+      "epoch": 0.5018165095424235,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.495077147647987e-05,
+      "loss": 0.8615,
+      "step": 6820
+    },
+    {
+      "epoch": 0.5018900896757875,
+      "grad_norm": 1.140625,
+      "learning_rate": 2.494497989433598e-05,
+      "loss": 1.1456,
+      "step": 6821
+    },
+    {
+      "epoch": 0.5019636698091515,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.4939188315144914e-05,
+      "loss": 0.897,
+      "step": 6822
+    },
+    {
+      "epoch": 0.5020372499425155,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.49333967392175e-05,
+      "loss": 1.0049,
+      "step": 6823
+    },
+    {
+      "epoch": 0.5021108300758795,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.4927605166864554e-05,
+      "loss": 0.7346,
+      "step": 6824
+    },
+    {
+      "epoch": 0.5021844102092435,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.49218135983969e-05,
+      "loss": 1.0402,
+      "step": 6825
+    },
+    {
+      "epoch": 0.5022579903426075,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.4916022034125365e-05,
+      "loss": 1.0197,
+      "step": 6826
+    },
+    {
+      "epoch": 0.5023315704759714,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.4910230474360764e-05,
+      "loss": 0.7358,
+      "step": 6827
+    },
+    {
+      "epoch": 0.5024051506093354,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.4904438919413938e-05,
+      "loss": 1.1908,
+      "step": 6828
+    },
+    {
+      "epoch": 0.5024787307426994,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.4898647369595682e-05,
+      "loss": 0.6124,
+      "step": 6829
+    },
+    {
+      "epoch": 0.5025523108760634,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.489285582521683e-05,
+      "loss": 0.9041,
+      "step": 6830
+    },
+    {
+      "epoch": 0.5026258910094275,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.4887064286588206e-05,
+      "loss": 0.8017,
+      "step": 6831
+    },
+    {
+      "epoch": 0.5026994711427915,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.488127275402063e-05,
+      "loss": 0.9575,
+      "step": 6832
+    },
+    {
+      "epoch": 0.5027730512761555,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.4875481227824928e-05,
+      "loss": 0.889,
+      "step": 6833
+    },
+    {
+      "epoch": 0.5028466314095195,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.4869689708311905e-05,
+      "loss": 0.9494,
+      "step": 6834
+    },
+    {
+      "epoch": 0.5029202115428835,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.486389819579239e-05,
+      "loss": 0.8909,
+      "step": 6835
+    },
+    {
+      "epoch": 0.5029937916762474,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.4858106690577198e-05,
+      "loss": 1.0806,
+      "step": 6836
+    },
+    {
+      "epoch": 0.5030673718096114,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.4852315192977162e-05,
+      "loss": 1.1335,
+      "step": 6837
+    },
+    {
+      "epoch": 0.5031409519429754,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.484652370330308e-05,
+      "loss": 1.1262,
+      "step": 6838
+    },
+    {
+      "epoch": 0.5032145320763394,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.4840732221865788e-05,
+      "loss": 0.7607,
+      "step": 6839
+    },
+    {
+      "epoch": 0.5032881122097034,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.4834940748976095e-05,
+      "loss": 0.8848,
+      "step": 6840
+    },
+    {
+      "epoch": 0.5033616923430674,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.4829149284944814e-05,
+      "loss": 1.009,
+      "step": 6841
+    },
+    {
+      "epoch": 0.5034352724764314,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.482335783008278e-05,
+      "loss": 0.9775,
+      "step": 6842
+    },
+    {
+      "epoch": 0.5035088526097954,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.4817566384700787e-05,
+      "loss": 0.6439,
+      "step": 6843
+    },
+    {
+      "epoch": 0.5035824327431594,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.4811774949109668e-05,
+      "loss": 1.2899,
+      "step": 6844
+    },
+    {
+      "epoch": 0.5036560128765233,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.4805983523620226e-05,
+      "loss": 1.7501,
+      "step": 6845
+    },
+    {
+      "epoch": 0.5037295930098873,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.4800192108543296e-05,
+      "loss": 0.9806,
+      "step": 6846
+    },
+    {
+      "epoch": 0.5038031731432513,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.4794400704189664e-05,
+      "loss": 0.9335,
+      "step": 6847
+    },
+    {
+      "epoch": 0.5038767532766153,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.478860931087016e-05,
+      "loss": 0.9179,
+      "step": 6848
+    },
+    {
+      "epoch": 0.5039503334099793,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.4782817928895598e-05,
+      "loss": 0.6442,
+      "step": 6849
+    },
+    {
+      "epoch": 0.5040239135433433,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.477702655857678e-05,
+      "loss": 0.9782,
+      "step": 6850
+    },
+    {
+      "epoch": 0.5040974936767073,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.477123520022454e-05,
+      "loss": 0.677,
+      "step": 6851
+    },
+    {
+      "epoch": 0.5041710738100713,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.4765443854149667e-05,
+      "loss": 1.0652,
+      "step": 6852
+    },
+    {
+      "epoch": 0.5042446539434353,
+      "grad_norm": 0.75,
+      "learning_rate": 2.4759652520662975e-05,
+      "loss": 0.7857,
+      "step": 6853
+    },
+    {
+      "epoch": 0.5043182340767992,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.475386120007528e-05,
+      "loss": 0.6471,
+      "step": 6854
+    },
+    {
+      "epoch": 0.5043918142101632,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.4748069892697398e-05,
+      "loss": 0.799,
+      "step": 6855
+    },
+    {
+      "epoch": 0.5044653943435272,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.4742278598840122e-05,
+      "loss": 1.0061,
+      "step": 6856
+    },
+    {
+      "epoch": 0.5045389744768912,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.4736487318814267e-05,
+      "loss": 0.7325,
+      "step": 6857
+    },
+    {
+      "epoch": 0.5046125546102552,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.4730696052930638e-05,
+      "loss": 0.8355,
+      "step": 6858
+    },
+    {
+      "epoch": 0.5046861347436192,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.4724904801500043e-05,
+      "loss": 0.8914,
+      "step": 6859
+    },
+    {
+      "epoch": 0.5047597148769832,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.4719113564833303e-05,
+      "loss": 1.6786,
+      "step": 6860
+    },
+    {
+      "epoch": 0.5048332950103472,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.4713322343241192e-05,
+      "loss": 1.0828,
+      "step": 6861
+    },
+    {
+      "epoch": 0.5049068751437112,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.470753113703453e-05,
+      "loss": 0.7747,
+      "step": 6862
+    },
+    {
+      "epoch": 0.5049804552770751,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.470173994652412e-05,
+      "loss": 1.0917,
+      "step": 6863
+    },
+    {
+      "epoch": 0.5050540354104391,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.4695948772020756e-05,
+      "loss": 1.1608,
+      "step": 6864
+    },
+    {
+      "epoch": 0.5051276155438031,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.469015761383527e-05,
+      "loss": 0.7306,
+      "step": 6865
+    },
+    {
+      "epoch": 0.5052011956771671,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.4684366472278423e-05,
+      "loss": 0.706,
+      "step": 6866
+    },
+    {
+      "epoch": 0.5052747758105312,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.4678575347661034e-05,
+      "loss": 0.7526,
+      "step": 6867
+    },
+    {
+      "epoch": 0.5053483559438952,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.4672784240293897e-05,
+      "loss": 0.9766,
+      "step": 6868
+    },
+    {
+      "epoch": 0.5054219360772592,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.4666993150487818e-05,
+      "loss": 1.0136,
+      "step": 6869
+    },
+    {
+      "epoch": 0.5054955162106232,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.4661202078553584e-05,
+      "loss": 1.1905,
+      "step": 6870
+    },
+    {
+      "epoch": 0.5055690963439872,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.465541102480199e-05,
+      "loss": 0.7406,
+      "step": 6871
+    },
+    {
+      "epoch": 0.5056426764773511,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.464961998954384e-05,
+      "loss": 0.6446,
+      "step": 6872
+    },
+    {
+      "epoch": 0.5057162566107151,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.464382897308992e-05,
+      "loss": 0.8695,
+      "step": 6873
+    },
+    {
+      "epoch": 0.5057898367440791,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.463803797575104e-05,
+      "loss": 1.2651,
+      "step": 6874
+    },
+    {
+      "epoch": 0.5058634168774431,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.463224699783797e-05,
+      "loss": 0.7668,
+      "step": 6875
+    },
+    {
+      "epoch": 0.5059369970108071,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.4626456039661503e-05,
+      "loss": 0.8345,
+      "step": 6876
+    },
+    {
+      "epoch": 0.5060105771441711,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.462066510153243e-05,
+      "loss": 1.0072,
+      "step": 6877
+    },
+    {
+      "epoch": 0.5060841572775351,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.4614874183761564e-05,
+      "loss": 1.4951,
+      "step": 6878
+    },
+    {
+      "epoch": 0.5061577374108991,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.460908328665966e-05,
+      "loss": 1.0984,
+      "step": 6879
+    },
+    {
+      "epoch": 0.5062313175442631,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.460329241053752e-05,
+      "loss": 1.1923,
+      "step": 6880
+    },
+    {
+      "epoch": 0.506304897677627,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.4597501555705925e-05,
+      "loss": 0.8413,
+      "step": 6881
+    },
+    {
+      "epoch": 0.506378477810991,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.4591710722475662e-05,
+      "loss": 1.3315,
+      "step": 6882
+    },
+    {
+      "epoch": 0.506452057944355,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.4585919911157522e-05,
+      "loss": 0.7327,
+      "step": 6883
+    },
+    {
+      "epoch": 0.506525638077719,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.4580129122062267e-05,
+      "loss": 0.7946,
+      "step": 6884
+    },
+    {
+      "epoch": 0.506599218211083,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.4574338355500694e-05,
+      "loss": 0.6864,
+      "step": 6885
+    },
+    {
+      "epoch": 0.506672798344447,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.4568547611783575e-05,
+      "loss": 0.7982,
+      "step": 6886
+    },
+    {
+      "epoch": 0.506746378477811,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.4562756891221698e-05,
+      "loss": 1.0393,
+      "step": 6887
+    },
+    {
+      "epoch": 0.506819958611175,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.4556966194125826e-05,
+      "loss": 1.0597,
+      "step": 6888
+    },
+    {
+      "epoch": 0.506893538744539,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.4551175520806744e-05,
+      "loss": 1.2202,
+      "step": 6889
+    },
+    {
+      "epoch": 0.5069671188779029,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.4545384871575215e-05,
+      "loss": 0.5673,
+      "step": 6890
+    },
+    {
+      "epoch": 0.5070406990112669,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.4539594246742027e-05,
+      "loss": 0.6118,
+      "step": 6891
+    },
+    {
+      "epoch": 0.5071142791446309,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.453380364661796e-05,
+      "loss": 0.9154,
+      "step": 6892
+    },
+    {
+      "epoch": 0.5071878592779949,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.4528013071513757e-05,
+      "loss": 0.7346,
+      "step": 6893
+    },
+    {
+      "epoch": 0.5072614394113589,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.45222225217402e-05,
+      "loss": 0.7266,
+      "step": 6894
+    },
+    {
+      "epoch": 0.5073350195447229,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.4516431997608062e-05,
+      "loss": 0.6065,
+      "step": 6895
+    },
+    {
+      "epoch": 0.5074085996780869,
+      "grad_norm": 0.6328125,
+      "learning_rate": 2.451064149942811e-05,
+      "loss": 0.6672,
+      "step": 6896
+    },
+    {
+      "epoch": 0.5074821798114509,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.4504851027511094e-05,
+      "loss": 0.7494,
+      "step": 6897
+    },
+    {
+      "epoch": 0.5075557599448149,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.449906058216779e-05,
+      "loss": 1.0767,
+      "step": 6898
+    },
+    {
+      "epoch": 0.5076293400781788,
+      "grad_norm": 1.125,
+      "learning_rate": 2.449327016370896e-05,
+      "loss": 1.2438,
+      "step": 6899
+    },
+    {
+      "epoch": 0.5077029202115428,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.4487479772445365e-05,
+      "loss": 0.752,
+      "step": 6900
+    },
+    {
+      "epoch": 0.5077765003449068,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.4481689408687772e-05,
+      "loss": 0.9818,
+      "step": 6901
+    },
+    {
+      "epoch": 0.5078500804782708,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.4475899072746918e-05,
+      "loss": 0.7712,
+      "step": 6902
+    },
+    {
+      "epoch": 0.5079236606116349,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.4470108764933564e-05,
+      "loss": 0.8489,
+      "step": 6903
+    },
+    {
+      "epoch": 0.5079972407449989,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.446431848555848e-05,
+      "loss": 0.9714,
+      "step": 6904
+    },
+    {
+      "epoch": 0.5080708208783629,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.445852823493241e-05,
+      "loss": 0.9575,
+      "step": 6905
+    },
+    {
+      "epoch": 0.5081444010117269,
+      "grad_norm": 1.125,
+      "learning_rate": 2.4452738013366117e-05,
+      "loss": 1.476,
+      "step": 6906
+    },
+    {
+      "epoch": 0.5082179811450909,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.444694782117033e-05,
+      "loss": 0.7326,
+      "step": 6907
+    },
+    {
+      "epoch": 0.5082915612784548,
+      "grad_norm": 1.0,
+      "learning_rate": 2.444115765865581e-05,
+      "loss": 0.8245,
+      "step": 6908
+    },
+    {
+      "epoch": 0.5083651414118188,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.4435367526133305e-05,
+      "loss": 0.7986,
+      "step": 6909
+    },
+    {
+      "epoch": 0.5084387215451828,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.442957742391356e-05,
+      "loss": 0.6466,
+      "step": 6910
+    },
+    {
+      "epoch": 0.5085123016785468,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.4423787352307314e-05,
+      "loss": 1.061,
+      "step": 6911
+    },
+    {
+      "epoch": 0.5085858818119108,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.4417997311625313e-05,
+      "loss": 0.6697,
+      "step": 6912
+    },
+    {
+      "epoch": 0.5086594619452748,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.4412207302178298e-05,
+      "loss": 0.9173,
+      "step": 6913
+    },
+    {
+      "epoch": 0.5087330420786388,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.4406417324277003e-05,
+      "loss": 1.1273,
+      "step": 6914
+    },
+    {
+      "epoch": 0.5088066222120028,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.4400627378232183e-05,
+      "loss": 1.1021,
+      "step": 6915
+    },
+    {
+      "epoch": 0.5088802023453668,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.4394837464354548e-05,
+      "loss": 0.8225,
+      "step": 6916
+    },
+    {
+      "epoch": 0.5089537824787307,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.438904758295484e-05,
+      "loss": 1.2826,
+      "step": 6917
+    },
+    {
+      "epoch": 0.5090273626120947,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.4383257734343796e-05,
+      "loss": 0.6253,
+      "step": 6918
+    },
+    {
+      "epoch": 0.5091009427454587,
+      "grad_norm": 1.25,
+      "learning_rate": 2.4377467918832157e-05,
+      "loss": 1.2076,
+      "step": 6919
+    },
+    {
+      "epoch": 0.5091745228788227,
+      "grad_norm": 1.0,
+      "learning_rate": 2.437167813673063e-05,
+      "loss": 1.6609,
+      "step": 6920
+    },
+    {
+      "epoch": 0.5092481030121867,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.4365888388349943e-05,
+      "loss": 1.054,
+      "step": 6921
+    },
+    {
+      "epoch": 0.5093216831455507,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.4360098674000835e-05,
+      "loss": 1.1305,
+      "step": 6922
+    },
+    {
+      "epoch": 0.5093952632789147,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.435430899399402e-05,
+      "loss": 0.6593,
+      "step": 6923
+    },
+    {
+      "epoch": 0.5094688434122787,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.434851934864023e-05,
+      "loss": 0.8338,
+      "step": 6924
+    },
+    {
+      "epoch": 0.5095424235456427,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.434272973825017e-05,
+      "loss": 1.2494,
+      "step": 6925
+    },
+    {
+      "epoch": 0.5096160036790066,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.4336940163134565e-05,
+      "loss": 1.015,
+      "step": 6926
+    },
+    {
+      "epoch": 0.5096895838123706,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.4331150623604126e-05,
+      "loss": 1.0062,
+      "step": 6927
+    },
+    {
+      "epoch": 0.5097631639457346,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.4325361119969582e-05,
+      "loss": 0.9374,
+      "step": 6928
+    },
+    {
+      "epoch": 0.5098367440790986,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.431957165254162e-05,
+      "loss": 0.6762,
+      "step": 6929
+    },
+    {
+      "epoch": 0.5099103242124626,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.431378222163096e-05,
+      "loss": 0.8244,
+      "step": 6930
+    },
+    {
+      "epoch": 0.5099839043458266,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.4307992827548316e-05,
+      "loss": 0.8764,
+      "step": 6931
+    },
+    {
+      "epoch": 0.5100574844791906,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.4302203470604388e-05,
+      "loss": 1.1631,
+      "step": 6932
+    },
+    {
+      "epoch": 0.5101310646125546,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.4296414151109893e-05,
+      "loss": 0.732,
+      "step": 6933
+    },
+    {
+      "epoch": 0.5102046447459186,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.4290624869375515e-05,
+      "loss": 0.9829,
+      "step": 6934
+    },
+    {
+      "epoch": 0.5102782248792826,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.4284835625711957e-05,
+      "loss": 0.95,
+      "step": 6935
+    },
+    {
+      "epoch": 0.5103518050126465,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.427904642042992e-05,
+      "loss": 1.4681,
+      "step": 6936
+    },
+    {
+      "epoch": 0.5104253851460105,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.42732572538401e-05,
+      "loss": 0.8922,
+      "step": 6937
+    },
+    {
+      "epoch": 0.5104989652793746,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.42674681262532e-05,
+      "loss": 0.966,
+      "step": 6938
+    },
+    {
+      "epoch": 0.5105725454127386,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.426167903797989e-05,
+      "loss": 1.0419,
+      "step": 6939
+    },
+    {
+      "epoch": 0.5106461255461026,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.4255889989330877e-05,
+      "loss": 0.7437,
+      "step": 6940
+    },
+    {
+      "epoch": 0.5107197056794666,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.425010098061684e-05,
+      "loss": 0.7891,
+      "step": 6941
+    },
+    {
+      "epoch": 0.5107932858128306,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.424431201214848e-05,
+      "loss": 0.8571,
+      "step": 6942
+    },
+    {
+      "epoch": 0.5108668659461946,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.423852308423645e-05,
+      "loss": 0.398,
+      "step": 6943
+    },
+    {
+      "epoch": 0.5109404460795586,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.4232734197191448e-05,
+      "loss": 0.8547,
+      "step": 6944
+    },
+    {
+      "epoch": 0.5110140262129225,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.422694535132416e-05,
+      "loss": 1.027,
+      "step": 6945
+    },
+    {
+      "epoch": 0.5110876063462865,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.422115654694525e-05,
+      "loss": 0.8691,
+      "step": 6946
+    },
+    {
+      "epoch": 0.5111611864796505,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.4215367784365413e-05,
+      "loss": 0.9444,
+      "step": 6947
+    },
+    {
+      "epoch": 0.5112347666130145,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.420957906389529e-05,
+      "loss": 0.8475,
+      "step": 6948
+    },
+    {
+      "epoch": 0.5113083467463785,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.4203790385845564e-05,
+      "loss": 0.78,
+      "step": 6949
+    },
+    {
+      "epoch": 0.5113819268797425,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.419800175052691e-05,
+      "loss": 1.2925,
+      "step": 6950
+    },
+    {
+      "epoch": 0.5114555070131065,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.4192213158249987e-05,
+      "loss": 0.9363,
+      "step": 6951
+    },
+    {
+      "epoch": 0.5115290871464705,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.4186424609325455e-05,
+      "loss": 0.8718,
+      "step": 6952
+    },
+    {
+      "epoch": 0.5116026672798345,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.418063610406398e-05,
+      "loss": 0.7089,
+      "step": 6953
+    },
+    {
+      "epoch": 0.5116762474131984,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.4174847642776217e-05,
+      "loss": 0.8138,
+      "step": 6954
+    },
+    {
+      "epoch": 0.5117498275465624,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.416905922577282e-05,
+      "loss": 0.8168,
+      "step": 6955
+    },
+    {
+      "epoch": 0.5118234076799264,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.416327085336445e-05,
+      "loss": 0.9951,
+      "step": 6956
+    },
+    {
+      "epoch": 0.5118969878132904,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.4157482525861748e-05,
+      "loss": 0.9895,
+      "step": 6957
+    },
+    {
+      "epoch": 0.5119705679466544,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.415169424357537e-05,
+      "loss": 0.7844,
+      "step": 6958
+    },
+    {
+      "epoch": 0.5120441480800184,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.414590600681596e-05,
+      "loss": 0.6938,
+      "step": 6959
+    },
+    {
+      "epoch": 0.5121177282133824,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.414011781589417e-05,
+      "loss": 1.1322,
+      "step": 6960
+    },
+    {
+      "epoch": 0.5121913083467464,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.4134329671120623e-05,
+      "loss": 1.162,
+      "step": 6961
+    },
+    {
+      "epoch": 0.5122648884801104,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.4128541572805966e-05,
+      "loss": 0.6704,
+      "step": 6962
+    },
+    {
+      "epoch": 0.5123384686134743,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.4122753521260837e-05,
+      "loss": 0.6643,
+      "step": 6963
+    },
+    {
+      "epoch": 0.5124120487468383,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.4116965516795873e-05,
+      "loss": 1.1199,
+      "step": 6964
+    },
+    {
+      "epoch": 0.5124856288802023,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.4111177559721703e-05,
+      "loss": 0.664,
+      "step": 6965
+    },
+    {
+      "epoch": 0.5125592090135663,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.4105389650348954e-05,
+      "loss": 0.8739,
+      "step": 6966
+    },
+    {
+      "epoch": 0.5126327891469303,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.4099601788988247e-05,
+      "loss": 1.1971,
+      "step": 6967
+    },
+    {
+      "epoch": 0.5127063692802943,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.4093813975950215e-05,
+      "loss": 0.9507,
+      "step": 6968
+    },
+    {
+      "epoch": 0.5127799494136583,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.4088026211545474e-05,
+      "loss": 0.8032,
+      "step": 6969
+    },
+    {
+      "epoch": 0.5128535295470223,
+      "grad_norm": 0.75,
+      "learning_rate": 2.408223849608465e-05,
+      "loss": 1.1861,
+      "step": 6970
+    },
+    {
+      "epoch": 0.5129271096803864,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.4076450829878347e-05,
+      "loss": 1.0445,
+      "step": 6971
+    },
+    {
+      "epoch": 0.5130006898137502,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.4070663213237185e-05,
+      "loss": 1.5042,
+      "step": 6972
+    },
+    {
+      "epoch": 0.5130742699471142,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.406487564647177e-05,
+      "loss": 0.7042,
+      "step": 6973
+    },
+    {
+      "epoch": 0.5131478500804783,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.4059088129892725e-05,
+      "loss": 1.1904,
+      "step": 6974
+    },
+    {
+      "epoch": 0.5132214302138423,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.4053300663810633e-05,
+      "loss": 0.8076,
+      "step": 6975
+    },
+    {
+      "epoch": 0.5132950103472063,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.4047513248536104e-05,
+      "loss": 0.7586,
+      "step": 6976
+    },
+    {
+      "epoch": 0.5133685904805703,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.4041725884379744e-05,
+      "loss": 0.5449,
+      "step": 6977
+    },
+    {
+      "epoch": 0.5134421706139343,
+      "grad_norm": 0.75,
+      "learning_rate": 2.4035938571652143e-05,
+      "loss": 0.7126,
+      "step": 6978
+    },
+    {
+      "epoch": 0.5135157507472983,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.4030151310663904e-05,
+      "loss": 0.8298,
+      "step": 6979
+    },
+    {
+      "epoch": 0.5135893308806623,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.402436410172561e-05,
+      "loss": 0.5451,
+      "step": 6980
+    },
+    {
+      "epoch": 0.5136629110140262,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.401857694514785e-05,
+      "loss": 0.7817,
+      "step": 6981
+    },
+    {
+      "epoch": 0.5137364911473902,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.4012789841241205e-05,
+      "loss": 0.6687,
+      "step": 6982
+    },
+    {
+      "epoch": 0.5138100712807542,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.4007002790316276e-05,
+      "loss": 1.3465,
+      "step": 6983
+    },
+    {
+      "epoch": 0.5138836514141182,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.4001215792683625e-05,
+      "loss": 1.1978,
+      "step": 6984
+    },
+    {
+      "epoch": 0.5139572315474822,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.399542884865384e-05,
+      "loss": 1.0106,
+      "step": 6985
+    },
+    {
+      "epoch": 0.5140308116808462,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.3989641958537486e-05,
+      "loss": 0.8311,
+      "step": 6986
+    },
+    {
+      "epoch": 0.5141043918142102,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.3983855122645142e-05,
+      "loss": 0.6216,
+      "step": 6987
+    },
+    {
+      "epoch": 0.5141779719475742,
+      "grad_norm": 0.875,
+      "learning_rate": 2.3978068341287385e-05,
+      "loss": 1.3475,
+      "step": 6988
+    },
+    {
+      "epoch": 0.5142515520809382,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.397228161477476e-05,
+      "loss": 0.9304,
+      "step": 6989
+    },
+    {
+      "epoch": 0.5143251322143021,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.3966494943417838e-05,
+      "loss": 0.944,
+      "step": 6990
+    },
+    {
+      "epoch": 0.5143987123476661,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.3960708327527184e-05,
+      "loss": 0.7263,
+      "step": 6991
+    },
+    {
+      "epoch": 0.5144722924810301,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.3954921767413352e-05,
+      "loss": 0.9915,
+      "step": 6992
+    },
+    {
+      "epoch": 0.5145458726143941,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.3949135263386894e-05,
+      "loss": 0.864,
+      "step": 6993
+    },
+    {
+      "epoch": 0.5146194527477581,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.3943348815758356e-05,
+      "loss": 1.1575,
+      "step": 6994
+    },
+    {
+      "epoch": 0.5146930328811221,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.3937562424838294e-05,
+      "loss": 0.9029,
+      "step": 6995
+    },
+    {
+      "epoch": 0.5147666130144861,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.3931776090937252e-05,
+      "loss": 0.859,
+      "step": 6996
+    },
+    {
+      "epoch": 0.5148401931478501,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.3925989814365774e-05,
+      "loss": 0.8189,
+      "step": 6997
+    },
+    {
+      "epoch": 0.5149137732812141,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.3920203595434387e-05,
+      "loss": 0.7732,
+      "step": 6998
+    },
+    {
+      "epoch": 0.514987353414578,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.3914417434453633e-05,
+      "loss": 0.9874,
+      "step": 6999
+    },
+    {
+      "epoch": 0.515060933547942,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.390863133173405e-05,
+      "loss": 1.3188,
+      "step": 7000
+    },
+    {
+      "epoch": 0.515134513681306,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.390284528758617e-05,
+      "loss": 1.1229,
+      "step": 7001
+    },
+    {
+      "epoch": 0.51520809381467,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.38970593023205e-05,
+      "loss": 1.0909,
+      "step": 7002
+    },
+    {
+      "epoch": 0.515281673948034,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.3891273376247572e-05,
+      "loss": 1.0579,
+      "step": 7003
+    },
+    {
+      "epoch": 0.515355254081398,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.388548750967791e-05,
+      "loss": 0.8992,
+      "step": 7004
+    },
+    {
+      "epoch": 0.515428834214762,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.3879701702922028e-05,
+      "loss": 0.8122,
+      "step": 7005
+    },
+    {
+      "epoch": 0.515502414348126,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.3873915956290446e-05,
+      "loss": 0.649,
+      "step": 7006
+    },
+    {
+      "epoch": 0.51557599448149,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.386813027009366e-05,
+      "loss": 0.831,
+      "step": 7007
+    },
+    {
+      "epoch": 0.515649574614854,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.3862344644642187e-05,
+      "loss": 0.6527,
+      "step": 7008
+    },
+    {
+      "epoch": 0.515723154748218,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.385655908024653e-05,
+      "loss": 0.7749,
+      "step": 7009
+    },
+    {
+      "epoch": 0.515796734881582,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.385077357721718e-05,
+      "loss": 0.7606,
+      "step": 7010
+    },
+    {
+      "epoch": 0.515870315014946,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.3844988135864654e-05,
+      "loss": 0.8744,
+      "step": 7011
+    },
+    {
+      "epoch": 0.51594389514831,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.3839202756499426e-05,
+      "loss": 0.6514,
+      "step": 7012
+    },
+    {
+      "epoch": 0.516017475281674,
+      "grad_norm": 1.125,
+      "learning_rate": 2.3833417439431993e-05,
+      "loss": 1.3973,
+      "step": 7013
+    },
+    {
+      "epoch": 0.516091055415038,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.3827632184972842e-05,
+      "loss": 0.8511,
+      "step": 7014
+    },
+    {
+      "epoch": 0.516164635548402,
+      "grad_norm": 0.9375,
+      "learning_rate": 2.382184699343247e-05,
+      "loss": 0.8448,
+      "step": 7015
+    },
+    {
+      "epoch": 0.516238215681766,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.3816061865121333e-05,
+      "loss": 0.9146,
+      "step": 7016
+    },
+    {
+      "epoch": 0.5163117958151299,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.3810276800349918e-05,
+      "loss": 0.6113,
+      "step": 7017
+    },
+    {
+      "epoch": 0.5163853759484939,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.38044917994287e-05,
+      "loss": 0.7426,
+      "step": 7018
+    },
+    {
+      "epoch": 0.5164589560818579,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.379870686266815e-05,
+      "loss": 0.8616,
+      "step": 7019
+    },
+    {
+      "epoch": 0.5165325362152219,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.379292199037874e-05,
+      "loss": 1.7812,
+      "step": 7020
+    },
+    {
+      "epoch": 0.5166061163485859,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.3787137182870926e-05,
+      "loss": 0.8274,
+      "step": 7021
+    },
+    {
+      "epoch": 0.5166796964819499,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.3781352440455162e-05,
+      "loss": 1.1996,
+      "step": 7022
+    },
+    {
+      "epoch": 0.5167532766153139,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.3775567763441915e-05,
+      "loss": 0.9538,
+      "step": 7023
+    },
+    {
+      "epoch": 0.5168268567486779,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.3769783152141633e-05,
+      "loss": 0.7558,
+      "step": 7024
+    },
+    {
+      "epoch": 0.5169004368820419,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.3763998606864764e-05,
+      "loss": 1.3663,
+      "step": 7025
+    },
+    {
+      "epoch": 0.5169740170154058,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.375821412792176e-05,
+      "loss": 1.1262,
+      "step": 7026
+    },
+    {
+      "epoch": 0.5170475971487698,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.3752429715623052e-05,
+      "loss": 1.0163,
+      "step": 7027
+    },
+    {
+      "epoch": 0.5171211772821338,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.3746645370279084e-05,
+      "loss": 1.0332,
+      "step": 7028
+    },
+    {
+      "epoch": 0.5171947574154978,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.3740861092200307e-05,
+      "loss": 1.1006,
+      "step": 7029
+    },
+    {
+      "epoch": 0.5172683375488618,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.3735076881697124e-05,
+      "loss": 1.0731,
+      "step": 7030
+    },
+    {
+      "epoch": 0.5173419176822258,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.3729292739079975e-05,
+      "loss": 0.7434,
+      "step": 7031
+    },
+    {
+      "epoch": 0.5174154978155898,
+      "grad_norm": 0.609375,
+      "learning_rate": 2.3723508664659286e-05,
+      "loss": 0.6844,
+      "step": 7032
+    },
+    {
+      "epoch": 0.5174890779489538,
+      "grad_norm": 1.125,
+      "learning_rate": 2.371772465874548e-05,
+      "loss": 1.2202,
+      "step": 7033
+    },
+    {
+      "epoch": 0.5175626580823178,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.3711940721648966e-05,
+      "loss": 0.7353,
+      "step": 7034
+    },
+    {
+      "epoch": 0.5176362382156817,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.370615685368016e-05,
+      "loss": 0.8891,
+      "step": 7035
+    },
+    {
+      "epoch": 0.5177098183490457,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.3700373055149472e-05,
+      "loss": 1.232,
+      "step": 7036
+    },
+    {
+      "epoch": 0.5177833984824097,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.3694589326367305e-05,
+      "loss": 0.8903,
+      "step": 7037
+    },
+    {
+      "epoch": 0.5178569786157737,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.3688805667644074e-05,
+      "loss": 1.089,
+      "step": 7038
+    },
+    {
+      "epoch": 0.5179305587491377,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.3683022079290158e-05,
+      "loss": 0.9619,
+      "step": 7039
+    },
+    {
+      "epoch": 0.5180041388825017,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.3677238561615958e-05,
+      "loss": 1.0581,
+      "step": 7040
+    },
+    {
+      "epoch": 0.5180777190158657,
+      "grad_norm": 1.0,
+      "learning_rate": 2.3671455114931867e-05,
+      "loss": 1.1741,
+      "step": 7041
+    },
+    {
+      "epoch": 0.5181512991492297,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.366567173954827e-05,
+      "loss": 0.871,
+      "step": 7042
+    },
+    {
+      "epoch": 0.5182248792825938,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.3659888435775565e-05,
+      "loss": 1.5283,
+      "step": 7043
+    },
+    {
+      "epoch": 0.5182984594159576,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.3654105203924105e-05,
+      "loss": 0.7814,
+      "step": 7044
+    },
+    {
+      "epoch": 0.5183720395493217,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.364832204430427e-05,
+      "loss": 0.8017,
+      "step": 7045
+    },
+    {
+      "epoch": 0.5184456196826857,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.364253895722644e-05,
+      "loss": 1.2968,
+      "step": 7046
+    },
+    {
+      "epoch": 0.5185191998160497,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.3636755943000995e-05,
+      "loss": 0.6544,
+      "step": 7047
+    },
+    {
+      "epoch": 0.5185927799494137,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.3630973001938273e-05,
+      "loss": 1.4115,
+      "step": 7048
+    },
+    {
+      "epoch": 0.5186663600827777,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.362519013434864e-05,
+      "loss": 0.6817,
+      "step": 7049
+    },
+    {
+      "epoch": 0.5187399402161417,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.361940734054246e-05,
+      "loss": 1.2427,
+      "step": 7050
+    },
+    {
+      "epoch": 0.5188135203495057,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.361362462083008e-05,
+      "loss": 0.8087,
+      "step": 7051
+    },
+    {
+      "epoch": 0.5188871004828697,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.3607841975521852e-05,
+      "loss": 1.005,
+      "step": 7052
+    },
+    {
+      "epoch": 0.5189606806162336,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.3602059404928112e-05,
+      "loss": 0.801,
+      "step": 7053
+    },
+    {
+      "epoch": 0.5190342607495976,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.35962769093592e-05,
+      "loss": 1.3282,
+      "step": 7054
+    },
+    {
+      "epoch": 0.5191078408829616,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.359049448912546e-05,
+      "loss": 0.7246,
+      "step": 7055
+    },
+    {
+      "epoch": 0.5191814210163256,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.358471214453723e-05,
+      "loss": 1.1222,
+      "step": 7056
+    },
+    {
+      "epoch": 0.5192550011496896,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.3578929875904812e-05,
+      "loss": 0.9401,
+      "step": 7057
+    },
+    {
+      "epoch": 0.5193285812830536,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.3573147683538544e-05,
+      "loss": 0.7846,
+      "step": 7058
+    },
+    {
+      "epoch": 0.5194021614164176,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.3567365567748744e-05,
+      "loss": 0.6962,
+      "step": 7059
+    },
+    {
+      "epoch": 0.5194757415497816,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.3561583528845724e-05,
+      "loss": 0.8994,
+      "step": 7060
+    },
+    {
+      "epoch": 0.5195493216831456,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.3555801567139816e-05,
+      "loss": 0.9372,
+      "step": 7061
+    },
+    {
+      "epoch": 0.5196229018165095,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.3550019682941295e-05,
+      "loss": 1.2208,
+      "step": 7062
+    },
+    {
+      "epoch": 0.5196964819498735,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.3544237876560484e-05,
+      "loss": 0.938,
+      "step": 7063
+    },
+    {
+      "epoch": 0.5197700620832375,
+      "grad_norm": 1.34375,
+      "learning_rate": 2.3538456148307674e-05,
+      "loss": 1.3258,
+      "step": 7064
+    },
+    {
+      "epoch": 0.5198436422166015,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.3532674498493172e-05,
+      "loss": 1.2045,
+      "step": 7065
+    },
+    {
+      "epoch": 0.5199172223499655,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.3526892927427248e-05,
+      "loss": 0.8768,
+      "step": 7066
+    },
+    {
+      "epoch": 0.5199908024833295,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.3521111435420198e-05,
+      "loss": 0.959,
+      "step": 7067
+    },
+    {
+      "epoch": 0.5200643826166935,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.3515330022782307e-05,
+      "loss": 0.958,
+      "step": 7068
+    },
+    {
+      "epoch": 0.5201379627500575,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.350954868982385e-05,
+      "loss": 0.7206,
+      "step": 7069
+    },
+    {
+      "epoch": 0.5202115428834215,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.3503767436855108e-05,
+      "loss": 0.8993,
+      "step": 7070
+    },
+    {
+      "epoch": 0.5202851230167854,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.349798626418633e-05,
+      "loss": 1.1485,
+      "step": 7071
+    },
+    {
+      "epoch": 0.5203587031501494,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.3492205172127797e-05,
+      "loss": 0.8782,
+      "step": 7072
+    },
+    {
+      "epoch": 0.5204322832835134,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.3486424160989753e-05,
+      "loss": 0.9922,
+      "step": 7073
+    },
+    {
+      "epoch": 0.5205058634168774,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.3480643231082475e-05,
+      "loss": 0.6067,
+      "step": 7074
+    },
+    {
+      "epoch": 0.5205794435502414,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.347486238271622e-05,
+      "loss": 0.9247,
+      "step": 7075
+    },
+    {
+      "epoch": 0.5206530236836054,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.34690816162012e-05,
+      "loss": 0.7634,
+      "step": 7076
+    },
+    {
+      "epoch": 0.5207266038169694,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.3463300931847684e-05,
+      "loss": 0.9113,
+      "step": 7077
+    },
+    {
+      "epoch": 0.5208001839503335,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.34575203299659e-05,
+      "loss": 0.9231,
+      "step": 7078
+    },
+    {
+      "epoch": 0.5208737640836975,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.34517398108661e-05,
+      "loss": 0.4933,
+      "step": 7079
+    },
+    {
+      "epoch": 0.5209473442170613,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.344595937485849e-05,
+      "loss": 0.8535,
+      "step": 7080
+    },
+    {
+      "epoch": 0.5210209243504254,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.3440179022253303e-05,
+      "loss": 0.8081,
+      "step": 7081
+    },
+    {
+      "epoch": 0.5210945044837894,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.3434398753360765e-05,
+      "loss": 1.1482,
+      "step": 7082
+    },
+    {
+      "epoch": 0.5211680846171534,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.3428618568491086e-05,
+      "loss": 1.1784,
+      "step": 7083
+    },
+    {
+      "epoch": 0.5212416647505174,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.3422838467954495e-05,
+      "loss": 0.8263,
+      "step": 7084
+    },
+    {
+      "epoch": 0.5213152448838814,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.341705845206117e-05,
+      "loss": 0.8077,
+      "step": 7085
+    },
+    {
+      "epoch": 0.5213888250172454,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.3411278521121327e-05,
+      "loss": 1.7572,
+      "step": 7086
+    },
+    {
+      "epoch": 0.5214624051506094,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.3405498675445166e-05,
+      "loss": 0.8529,
+      "step": 7087
+    },
+    {
+      "epoch": 0.5215359852839734,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.3399718915342893e-05,
+      "loss": 1.0166,
+      "step": 7088
+    },
+    {
+      "epoch": 0.5216095654173373,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.3393939241124672e-05,
+      "loss": 0.741,
+      "step": 7089
+    },
+    {
+      "epoch": 0.5216831455507013,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.33881596531007e-05,
+      "loss": 0.9169,
+      "step": 7090
+    },
+    {
+      "epoch": 0.5217567256840653,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.3382380151581154e-05,
+      "loss": 0.9867,
+      "step": 7091
+    },
+    {
+      "epoch": 0.5218303058174293,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.337660073687621e-05,
+      "loss": 1.0336,
+      "step": 7092
+    },
+    {
+      "epoch": 0.5219038859507933,
+      "grad_norm": 0.734375,
+      "learning_rate": 2.337082140929604e-05,
+      "loss": 1.063,
+      "step": 7093
+    },
+    {
+      "epoch": 0.5219774660841573,
+      "grad_norm": 1.125,
+      "learning_rate": 2.3365042169150808e-05,
+      "loss": 1.2778,
+      "step": 7094
+    },
+    {
+      "epoch": 0.5220510462175213,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.3359263016750673e-05,
+      "loss": 0.8718,
+      "step": 7095
+    },
+    {
+      "epoch": 0.5221246263508853,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.335348395240579e-05,
+      "loss": 0.4805,
+      "step": 7096
+    },
+    {
+      "epoch": 0.5221982064842493,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.3347704976426328e-05,
+      "loss": 0.8076,
+      "step": 7097
+    },
+    {
+      "epoch": 0.5222717866176132,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.334192608912241e-05,
+      "loss": 0.7714,
+      "step": 7098
+    },
+    {
+      "epoch": 0.5223453667509772,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.333614729080418e-05,
+      "loss": 0.9829,
+      "step": 7099
+    },
+    {
+      "epoch": 0.5224189468843412,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.3330368581781783e-05,
+      "loss": 1.0025,
+      "step": 7100
+    },
+    {
+      "epoch": 0.5224925270177052,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.3324589962365357e-05,
+      "loss": 0.8973,
+      "step": 7101
+    },
+    {
+      "epoch": 0.5225661071510692,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.3318811432865032e-05,
+      "loss": 1.0828,
+      "step": 7102
+    },
+    {
+      "epoch": 0.5226396872844332,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.3313032993590907e-05,
+      "loss": 0.9739,
+      "step": 7103
+    },
+    {
+      "epoch": 0.5227132674177972,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.3307254644853122e-05,
+      "loss": 0.9812,
+      "step": 7104
+    },
+    {
+      "epoch": 0.5227868475511612,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.330147638696178e-05,
+      "loss": 0.8089,
+      "step": 7105
+    },
+    {
+      "epoch": 0.5228604276845252,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.329569822022699e-05,
+      "loss": 0.848,
+      "step": 7106
+    },
+    {
+      "epoch": 0.5229340078178891,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.3289920144958864e-05,
+      "loss": 1.4171,
+      "step": 7107
+    },
+    {
+      "epoch": 0.5230075879512531,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.3284142161467493e-05,
+      "loss": 0.9272,
+      "step": 7108
+    },
+    {
+      "epoch": 0.5230811680846171,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.3278364270062966e-05,
+      "loss": 1.0096,
+      "step": 7109
+    },
+    {
+      "epoch": 0.5231547482179811,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.327258647105538e-05,
+      "loss": 0.9844,
+      "step": 7110
+    },
+    {
+      "epoch": 0.5232283283513451,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.3266808764754824e-05,
+      "loss": 1.5004,
+      "step": 7111
+    },
+    {
+      "epoch": 0.5233019084847091,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.326103115147136e-05,
+      "loss": 0.9452,
+      "step": 7112
+    },
+    {
+      "epoch": 0.5233754886180731,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.3255253631515062e-05,
+      "loss": 1.2341,
+      "step": 7113
+    },
+    {
+      "epoch": 0.5234490687514372,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.3249476205196014e-05,
+      "loss": 0.9148,
+      "step": 7114
+    },
+    {
+      "epoch": 0.5235226488848012,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.3243698872824267e-05,
+      "loss": 1.3095,
+      "step": 7115
+    },
+    {
+      "epoch": 0.523596229018165,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.32379216347099e-05,
+      "loss": 0.7405,
+      "step": 7116
+    },
+    {
+      "epoch": 0.523669809151529,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.3232144491162938e-05,
+      "loss": 1.0501,
+      "step": 7117
+    },
+    {
+      "epoch": 0.5237433892848931,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.3226367442493442e-05,
+      "loss": 0.9641,
+      "step": 7118
+    },
+    {
+      "epoch": 0.5238169694182571,
+      "grad_norm": 0.75,
+      "learning_rate": 2.3220590489011455e-05,
+      "loss": 0.7268,
+      "step": 7119
+    },
+    {
+      "epoch": 0.5238905495516211,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.3214813631027025e-05,
+      "loss": 1.2105,
+      "step": 7120
+    },
+    {
+      "epoch": 0.5239641296849851,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.3209036868850164e-05,
+      "loss": 0.9473,
+      "step": 7121
+    },
+    {
+      "epoch": 0.5240377098183491,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.3203260202790915e-05,
+      "loss": 1.2614,
+      "step": 7122
+    },
+    {
+      "epoch": 0.5241112899517131,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.31974836331593e-05,
+      "loss": 1.2449,
+      "step": 7123
+    },
+    {
+      "epoch": 0.5241848700850771,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.319170716026533e-05,
+      "loss": 1.1326,
+      "step": 7124
+    },
+    {
+      "epoch": 0.524258450218441,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.318593078441903e-05,
+      "loss": 1.1696,
+      "step": 7125
+    },
+    {
+      "epoch": 0.524332030351805,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.318015450593039e-05,
+      "loss": 1.0247,
+      "step": 7126
+    },
+    {
+      "epoch": 0.524405610485169,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.317437832510942e-05,
+      "loss": 0.9742,
+      "step": 7127
+    },
+    {
+      "epoch": 0.524479190618533,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.316860224226612e-05,
+      "loss": 0.8823,
+      "step": 7128
+    },
+    {
+      "epoch": 0.524552770751897,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.3162826257710488e-05,
+      "loss": 0.9043,
+      "step": 7129
+    },
+    {
+      "epoch": 0.524626350885261,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.315705037175249e-05,
+      "loss": 0.7274,
+      "step": 7130
+    },
+    {
+      "epoch": 0.524699931018625,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.315127458470212e-05,
+      "loss": 0.5809,
+      "step": 7131
+    },
+    {
+      "epoch": 0.524773511151989,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.3145498896869345e-05,
+      "loss": 1.0719,
+      "step": 7132
+    },
+    {
+      "epoch": 0.524847091285353,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.3139723308564146e-05,
+      "loss": 1.2744,
+      "step": 7133
+    },
+    {
+      "epoch": 0.5249206714187169,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.3133947820096487e-05,
+      "loss": 0.7277,
+      "step": 7134
+    },
+    {
+      "epoch": 0.5249942515520809,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.3128172431776322e-05,
+      "loss": 0.8872,
+      "step": 7135
+    },
+    {
+      "epoch": 0.5250678316854449,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.3122397143913604e-05,
+      "loss": 1.0752,
+      "step": 7136
+    },
+    {
+      "epoch": 0.5251414118188089,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.311662195681829e-05,
+      "loss": 0.7656,
+      "step": 7137
+    },
+    {
+      "epoch": 0.5252149919521729,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.3110846870800324e-05,
+      "loss": 1.03,
+      "step": 7138
+    },
+    {
+      "epoch": 0.5252885720855369,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.310507188616962e-05,
+      "loss": 0.8522,
+      "step": 7139
+    },
+    {
+      "epoch": 0.5253621522189009,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.309929700323614e-05,
+      "loss": 0.817,
+      "step": 7140
+    },
+    {
+      "epoch": 0.5254357323522649,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.30935222223098e-05,
+      "loss": 1.248,
+      "step": 7141
+    },
+    {
+      "epoch": 0.5255093124856289,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.3087747543700516e-05,
+      "loss": 0.8816,
+      "step": 7142
+    },
+    {
+      "epoch": 0.5255828926189928,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.3081972967718226e-05,
+      "loss": 0.814,
+      "step": 7143
+    },
+    {
+      "epoch": 0.5256564727523568,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.307619849467281e-05,
+      "loss": 0.8109,
+      "step": 7144
+    },
+    {
+      "epoch": 0.5257300528857208,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.307042412487419e-05,
+      "loss": 0.9709,
+      "step": 7145
+    },
+    {
+      "epoch": 0.5258036330190848,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.306464985863226e-05,
+      "loss": 0.5595,
+      "step": 7146
+    },
+    {
+      "epoch": 0.5258772131524488,
+      "grad_norm": 0.75,
+      "learning_rate": 2.3058875696256917e-05,
+      "loss": 0.7491,
+      "step": 7147
+    },
+    {
+      "epoch": 0.5259507932858128,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.3053101638058055e-05,
+      "loss": 1.3381,
+      "step": 7148
+    },
+    {
+      "epoch": 0.5260243734191768,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.3047327684345548e-05,
+      "loss": 0.8883,
+      "step": 7149
+    },
+    {
+      "epoch": 0.5260979535525409,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.304155383542927e-05,
+      "loss": 1.0357,
+      "step": 7150
+    },
+    {
+      "epoch": 0.5261715336859049,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.3035780091619104e-05,
+      "loss": 1.2783,
+      "step": 7151
+    },
+    {
+      "epoch": 0.5262451138192688,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.303000645322492e-05,
+      "loss": 0.7235,
+      "step": 7152
+    },
+    {
+      "epoch": 0.5263186939526328,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.3024232920556555e-05,
+      "loss": 1.0723,
+      "step": 7153
+    },
+    {
+      "epoch": 0.5263922740859968,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.301845949392388e-05,
+      "loss": 0.7592,
+      "step": 7154
+    },
+    {
+      "epoch": 0.5264658542193608,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.3012686173636742e-05,
+      "loss": 0.8355,
+      "step": 7155
+    },
+    {
+      "epoch": 0.5265394343527248,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.3006912960004985e-05,
+      "loss": 1.0347,
+      "step": 7156
+    },
+    {
+      "epoch": 0.5266130144860888,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.3001139853338453e-05,
+      "loss": 1.2346,
+      "step": 7157
+    },
+    {
+      "epoch": 0.5266865946194528,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.299536685394696e-05,
+      "loss": 0.8342,
+      "step": 7158
+    },
+    {
+      "epoch": 0.5267601747528168,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.298959396214034e-05,
+      "loss": 1.0139,
+      "step": 7159
+    },
+    {
+      "epoch": 0.5268337548861808,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.2983821178228416e-05,
+      "loss": 0.7412,
+      "step": 7160
+    },
+    {
+      "epoch": 0.5269073350195447,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.297804850252101e-05,
+      "loss": 1.248,
+      "step": 7161
+    },
+    {
+      "epoch": 0.5269809151529087,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.2972275935327914e-05,
+      "loss": 0.7006,
+      "step": 7162
+    },
+    {
+      "epoch": 0.5270544952862727,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.296650347695894e-05,
+      "loss": 0.9661,
+      "step": 7163
+    },
+    {
+      "epoch": 0.5271280754196367,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.2960731127723885e-05,
+      "loss": 0.8884,
+      "step": 7164
+    },
+    {
+      "epoch": 0.5272016555530007,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.2954958887932534e-05,
+      "loss": 1.1262,
+      "step": 7165
+    },
+    {
+      "epoch": 0.5272752356863647,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.2949186757894685e-05,
+      "loss": 0.7143,
+      "step": 7166
+    },
+    {
+      "epoch": 0.5273488158197287,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.29434147379201e-05,
+      "loss": 0.6592,
+      "step": 7167
+    },
+    {
+      "epoch": 0.5274223959530927,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.2937642828318568e-05,
+      "loss": 1.0319,
+      "step": 7168
+    },
+    {
+      "epoch": 0.5274959760864567,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.293187102939985e-05,
+      "loss": 1.0607,
+      "step": 7169
+    },
+    {
+      "epoch": 0.5275695562198206,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.2926099341473714e-05,
+      "loss": 1.3403,
+      "step": 7170
+    },
+    {
+      "epoch": 0.5276431363531846,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.29203277648499e-05,
+      "loss": 0.7048,
+      "step": 7171
+    },
+    {
+      "epoch": 0.5277167164865486,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.2914556299838166e-05,
+      "loss": 1.1061,
+      "step": 7172
+    },
+    {
+      "epoch": 0.5277902966199126,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.290878494674826e-05,
+      "loss": 0.6776,
+      "step": 7173
+    },
+    {
+      "epoch": 0.5278638767532766,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.2903013705889916e-05,
+      "loss": 0.7855,
+      "step": 7174
+    },
+    {
+      "epoch": 0.5279374568866406,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.2897242577572868e-05,
+      "loss": 0.9804,
+      "step": 7175
+    },
+    {
+      "epoch": 0.5280110370200046,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.2891471562106832e-05,
+      "loss": 1.1,
+      "step": 7176
+    },
+    {
+      "epoch": 0.5280846171533686,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.288570065980154e-05,
+      "loss": 1.1266,
+      "step": 7177
+    },
+    {
+      "epoch": 0.5281581972867326,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.28799298709667e-05,
+      "loss": 0.9623,
+      "step": 7178
+    },
+    {
+      "epoch": 0.5282317774200965,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.2874159195912015e-05,
+      "loss": 0.8287,
+      "step": 7179
+    },
+    {
+      "epoch": 0.5283053575534605,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.28683886349472e-05,
+      "loss": 0.9611,
+      "step": 7180
+    },
+    {
+      "epoch": 0.5283789376868245,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.2862618188381936e-05,
+      "loss": 0.9517,
+      "step": 7181
+    },
+    {
+      "epoch": 0.5284525178201885,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.2856847856525916e-05,
+      "loss": 0.797,
+      "step": 7182
+    },
+    {
+      "epoch": 0.5285260979535525,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.285107763968882e-05,
+      "loss": 1.0026,
+      "step": 7183
+    },
+    {
+      "epoch": 0.5285996780869165,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.2845307538180342e-05,
+      "loss": 0.999,
+      "step": 7184
+    },
+    {
+      "epoch": 0.5286732582202806,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.283953755231013e-05,
+      "loss": 1.0142,
+      "step": 7185
+    },
+    {
+      "epoch": 0.5287468383536446,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.2833767682387856e-05,
+      "loss": 0.991,
+      "step": 7186
+    },
+    {
+      "epoch": 0.5288204184870086,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.2827997928723177e-05,
+      "loss": 0.8853,
+      "step": 7187
+    },
+    {
+      "epoch": 0.5288939986203725,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.2822228291625746e-05,
+      "loss": 1.3182,
+      "step": 7188
+    },
+    {
+      "epoch": 0.5289675787537365,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.281645877140522e-05,
+      "loss": 0.722,
+      "step": 7189
+    },
+    {
+      "epoch": 0.5290411588871005,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.281068936837122e-05,
+      "loss": 0.8684,
+      "step": 7190
+    },
+    {
+      "epoch": 0.5291147390204645,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.2804920082833385e-05,
+      "loss": 0.7389,
+      "step": 7191
+    },
+    {
+      "epoch": 0.5291883191538285,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.279915091510134e-05,
+      "loss": 1.354,
+      "step": 7192
+    },
+    {
+      "epoch": 0.5292618992871925,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.279338186548472e-05,
+      "loss": 0.796,
+      "step": 7193
+    },
+    {
+      "epoch": 0.5293354794205565,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.278761293429312e-05,
+      "loss": 0.7321,
+      "step": 7194
+    },
+    {
+      "epoch": 0.5294090595539205,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.2781844121836155e-05,
+      "loss": 0.8823,
+      "step": 7195
+    },
+    {
+      "epoch": 0.5294826396872845,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.2776075428423426e-05,
+      "loss": 0.7476,
+      "step": 7196
+    },
+    {
+      "epoch": 0.5295562198206484,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.277030685436453e-05,
+      "loss": 0.8428,
+      "step": 7197
+    },
+    {
+      "epoch": 0.5296297999540124,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.2764538399969065e-05,
+      "loss": 0.8999,
+      "step": 7198
+    },
+    {
+      "epoch": 0.5297033800873764,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.275877006554659e-05,
+      "loss": 0.8448,
+      "step": 7199
+    },
+    {
+      "epoch": 0.5297769602207404,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.275300185140669e-05,
+      "loss": 0.9239,
+      "step": 7200
+    },
+    {
+      "epoch": 0.5298505403541044,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.274723375785894e-05,
+      "loss": 0.9951,
+      "step": 7201
+    },
+    {
+      "epoch": 0.5299241204874684,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.2741465785212905e-05,
+      "loss": 0.832,
+      "step": 7202
+    },
+    {
+      "epoch": 0.5299977006208324,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.273569793377813e-05,
+      "loss": 0.6314,
+      "step": 7203
+    },
+    {
+      "epoch": 0.5300712807541964,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.2729930203864167e-05,
+      "loss": 0.9671,
+      "step": 7204
+    },
+    {
+      "epoch": 0.5301448608875604,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.2724162595780564e-05,
+      "loss": 0.9649,
+      "step": 7205
+    },
+    {
+      "epoch": 0.5302184410209243,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.271839510983686e-05,
+      "loss": 0.7653,
+      "step": 7206
+    },
+    {
+      "epoch": 0.5302920211542883,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.271262774634258e-05,
+      "loss": 0.6205,
+      "step": 7207
+    },
+    {
+      "epoch": 0.5303656012876523,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.2706860505607246e-05,
+      "loss": 1.0973,
+      "step": 7208
+    },
+    {
+      "epoch": 0.5304391814210163,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.2701093387940378e-05,
+      "loss": 0.7639,
+      "step": 7209
+    },
+    {
+      "epoch": 0.5305127615543803,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.2695326393651485e-05,
+      "loss": 0.7974,
+      "step": 7210
+    },
+    {
+      "epoch": 0.5305863416877443,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.2689559523050073e-05,
+      "loss": 1.028,
+      "step": 7211
+    },
+    {
+      "epoch": 0.5306599218211083,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.268379277644565e-05,
+      "loss": 0.7399,
+      "step": 7212
+    },
+    {
+      "epoch": 0.5307335019544723,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.267802615414768e-05,
+      "loss": 1.0821,
+      "step": 7213
+    },
+    {
+      "epoch": 0.5308070820878363,
+      "grad_norm": 1.1953125,
+      "learning_rate": 2.267225965646566e-05,
+      "loss": 1.3199,
+      "step": 7214
+    },
+    {
+      "epoch": 0.5308806622212002,
+      "grad_norm": 1.296875,
+      "learning_rate": 2.266649328370907e-05,
+      "loss": 1.1325,
+      "step": 7215
+    },
+    {
+      "epoch": 0.5309542423545642,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.2660727036187384e-05,
+      "loss": 1.1914,
+      "step": 7216
+    },
+    {
+      "epoch": 0.5310278224879282,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.2654960914210053e-05,
+      "loss": 1.0181,
+      "step": 7217
+    },
+    {
+      "epoch": 0.5311014026212922,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.264919491808654e-05,
+      "loss": 0.9841,
+      "step": 7218
+    },
+    {
+      "epoch": 0.5311749827546562,
+      "grad_norm": 0.6171875,
+      "learning_rate": 2.2643429048126298e-05,
+      "loss": 0.6514,
+      "step": 7219
+    },
+    {
+      "epoch": 0.5312485628880202,
+      "grad_norm": 1.2578125,
+      "learning_rate": 2.2637663304638764e-05,
+      "loss": 1.4098,
+      "step": 7220
+    },
+    {
+      "epoch": 0.5313221430213843,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.2631897687933388e-05,
+      "loss": 1.1333,
+      "step": 7221
+    },
+    {
+      "epoch": 0.5313957231547483,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.2626132198319582e-05,
+      "loss": 1.2393,
+      "step": 7222
+    },
+    {
+      "epoch": 0.5314693032881123,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.262036683610678e-05,
+      "loss": 0.8314,
+      "step": 7223
+    },
+    {
+      "epoch": 0.5315428834214762,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.2614601601604393e-05,
+      "loss": 0.611,
+      "step": 7224
+    },
+    {
+      "epoch": 0.5316164635548402,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.2608836495121845e-05,
+      "loss": 0.7074,
+      "step": 7225
+    },
+    {
+      "epoch": 0.5316900436882042,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.2603071516968515e-05,
+      "loss": 1.0695,
+      "step": 7226
+    },
+    {
+      "epoch": 0.5317636238215682,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.259730666745381e-05,
+      "loss": 0.7008,
+      "step": 7227
+    },
+    {
+      "epoch": 0.5318372039549322,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.2591541946887118e-05,
+      "loss": 0.6894,
+      "step": 7228
+    },
+    {
+      "epoch": 0.5319107840882962,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.2585777355577814e-05,
+      "loss": 0.7755,
+      "step": 7229
+    },
+    {
+      "epoch": 0.5319843642216602,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.25800128938353e-05,
+      "loss": 0.986,
+      "step": 7230
+    },
+    {
+      "epoch": 0.5320579443550242,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.2574248561968904e-05,
+      "loss": 0.8011,
+      "step": 7231
+    },
+    {
+      "epoch": 0.5321315244883882,
+      "grad_norm": 0.875,
+      "learning_rate": 2.2568484360288014e-05,
+      "loss": 1.0675,
+      "step": 7232
+    },
+    {
+      "epoch": 0.5322051046217521,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.2562720289101975e-05,
+      "loss": 0.7311,
+      "step": 7233
+    },
+    {
+      "epoch": 0.5322786847551161,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.2556956348720138e-05,
+      "loss": 0.6406,
+      "step": 7234
+    },
+    {
+      "epoch": 0.5323522648884801,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.255119253945183e-05,
+      "loss": 0.6657,
+      "step": 7235
+    },
+    {
+      "epoch": 0.5324258450218441,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.25454288616064e-05,
+      "loss": 0.8719,
+      "step": 7236
+    },
+    {
+      "epoch": 0.5324994251552081,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.2539665315493164e-05,
+      "loss": 0.777,
+      "step": 7237
+    },
+    {
+      "epoch": 0.5325730052885721,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.2533901901421445e-05,
+      "loss": 1.1499,
+      "step": 7238
+    },
+    {
+      "epoch": 0.5326465854219361,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.2528138619700562e-05,
+      "loss": 0.8541,
+      "step": 7239
+    },
+    {
+      "epoch": 0.5327201655553001,
+      "grad_norm": 1.125,
+      "learning_rate": 2.25223754706398e-05,
+      "loss": 0.9575,
+      "step": 7240
+    },
+    {
+      "epoch": 0.5327937456886641,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.2516612454548468e-05,
+      "loss": 0.8939,
+      "step": 7241
+    },
+    {
+      "epoch": 0.532867325822028,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.2510849571735857e-05,
+      "loss": 0.7688,
+      "step": 7242
+    },
+    {
+      "epoch": 0.532940905955392,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.250508682251125e-05,
+      "loss": 0.7608,
+      "step": 7243
+    },
+    {
+      "epoch": 0.533014486088756,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.249932420718392e-05,
+      "loss": 1.3086,
+      "step": 7244
+    },
+    {
+      "epoch": 0.53308806622212,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.2493561726063135e-05,
+      "loss": 0.7568,
+      "step": 7245
+    },
+    {
+      "epoch": 0.533161646355484,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.2487799379458156e-05,
+      "loss": 0.693,
+      "step": 7246
+    },
+    {
+      "epoch": 0.533235226488848,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.248203716767824e-05,
+      "loss": 0.8562,
+      "step": 7247
+    },
+    {
+      "epoch": 0.533308806622212,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.247627509103264e-05,
+      "loss": 0.9862,
+      "step": 7248
+    },
+    {
+      "epoch": 0.533382386755576,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.247051314983058e-05,
+      "loss": 0.9593,
+      "step": 7249
+    },
+    {
+      "epoch": 0.53345596688894,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.2464751344381307e-05,
+      "loss": 1.0049,
+      "step": 7250
+    },
+    {
+      "epoch": 0.5335295470223039,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.245898967499404e-05,
+      "loss": 1.0317,
+      "step": 7251
+    },
+    {
+      "epoch": 0.5336031271556679,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.2453228141977996e-05,
+      "loss": 0.8445,
+      "step": 7252
+    },
+    {
+      "epoch": 0.5336767072890319,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.2447466745642397e-05,
+      "loss": 1.3378,
+      "step": 7253
+    },
+    {
+      "epoch": 0.5337502874223959,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.2441705486296427e-05,
+      "loss": 0.7162,
+      "step": 7254
+    },
+    {
+      "epoch": 0.53382386755576,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.2435944364249294e-05,
+      "loss": 1.0402,
+      "step": 7255
+    },
+    {
+      "epoch": 0.533897447689124,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.2430183379810178e-05,
+      "loss": 1.0057,
+      "step": 7256
+    },
+    {
+      "epoch": 0.533971027822488,
+      "grad_norm": 0.875,
+      "learning_rate": 2.242442253328828e-05,
+      "loss": 1.0033,
+      "step": 7257
+    },
+    {
+      "epoch": 0.534044607955852,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.241866182499275e-05,
+      "loss": 0.6219,
+      "step": 7258
+    },
+    {
+      "epoch": 0.534118188089216,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.2412901255232765e-05,
+      "loss": 1.0824,
+      "step": 7259
+    },
+    {
+      "epoch": 0.5341917682225799,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.2407140824317486e-05,
+      "loss": 0.7379,
+      "step": 7260
+    },
+    {
+      "epoch": 0.5342653483559439,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.240138053255606e-05,
+      "loss": 0.7896,
+      "step": 7261
+    },
+    {
+      "epoch": 0.5343389284893079,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.2395620380257638e-05,
+      "loss": 1.1455,
+      "step": 7262
+    },
+    {
+      "epoch": 0.5344125086226719,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.2389860367731346e-05,
+      "loss": 1.0093,
+      "step": 7263
+    },
+    {
+      "epoch": 0.5344860887560359,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.2384100495286315e-05,
+      "loss": 1.1271,
+      "step": 7264
+    },
+    {
+      "epoch": 0.5345596688893999,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.2378340763231673e-05,
+      "loss": 1.311,
+      "step": 7265
+    },
+    {
+      "epoch": 0.5346332490227639,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.237258117187654e-05,
+      "loss": 1.0288,
+      "step": 7266
+    },
+    {
+      "epoch": 0.5347068291561279,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.2366821721530005e-05,
+      "loss": 1.013,
+      "step": 7267
+    },
+    {
+      "epoch": 0.5347804092894919,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.236106241250117e-05,
+      "loss": 1.0508,
+      "step": 7268
+    },
+    {
+      "epoch": 0.5348539894228558,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.2355303245099128e-05,
+      "loss": 1.031,
+      "step": 7269
+    },
+    {
+      "epoch": 0.5349275695562198,
+      "grad_norm": 1.1796875,
+      "learning_rate": 2.2349544219632974e-05,
+      "loss": 1.4136,
+      "step": 7270
+    },
+    {
+      "epoch": 0.5350011496895838,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.2343785336411783e-05,
+      "loss": 0.6587,
+      "step": 7271
+    },
+    {
+      "epoch": 0.5350747298229478,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.2338026595744604e-05,
+      "loss": 0.6274,
+      "step": 7272
+    },
+    {
+      "epoch": 0.5351483099563118,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.2332267997940516e-05,
+      "loss": 1.0917,
+      "step": 7273
+    },
+    {
+      "epoch": 0.5352218900896758,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.232650954330856e-05,
+      "loss": 1.4517,
+      "step": 7274
+    },
+    {
+      "epoch": 0.5352954702230398,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.2320751232157793e-05,
+      "loss": 0.8355,
+      "step": 7275
+    },
+    {
+      "epoch": 0.5353690503564038,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.2314993064797244e-05,
+      "loss": 1.3266,
+      "step": 7276
+    },
+    {
+      "epoch": 0.5354426304897678,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.2309235041535947e-05,
+      "loss": 1.5163,
+      "step": 7277
+    },
+    {
+      "epoch": 0.5355162106231317,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.230347716268292e-05,
+      "loss": 1.0089,
+      "step": 7278
+    },
+    {
+      "epoch": 0.5355897907564957,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.2297719428547186e-05,
+      "loss": 1.3617,
+      "step": 7279
+    },
+    {
+      "epoch": 0.5356633708898597,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.229196183943775e-05,
+      "loss": 1.0673,
+      "step": 7280
+    },
+    {
+      "epoch": 0.5357369510232237,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.2286204395663602e-05,
+      "loss": 0.8204,
+      "step": 7281
+    },
+    {
+      "epoch": 0.5358105311565877,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.228044709753373e-05,
+      "loss": 1.0077,
+      "step": 7282
+    },
+    {
+      "epoch": 0.5358841112899517,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.2274689945357133e-05,
+      "loss": 0.9992,
+      "step": 7283
+    },
+    {
+      "epoch": 0.5359576914233157,
+      "grad_norm": 1.1171875,
+      "learning_rate": 2.226893293944278e-05,
+      "loss": 1.1013,
+      "step": 7284
+    },
+    {
+      "epoch": 0.5360312715566797,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.2263176080099645e-05,
+      "loss": 1.0745,
+      "step": 7285
+    },
+    {
+      "epoch": 0.5361048516900437,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.2257419367636673e-05,
+      "loss": 0.9617,
+      "step": 7286
+    },
+    {
+      "epoch": 0.5361784318234076,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.225166280236282e-05,
+      "loss": 0.7743,
+      "step": 7287
+    },
+    {
+      "epoch": 0.5362520119567716,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.2245906384587036e-05,
+      "loss": 1.2785,
+      "step": 7288
+    },
+    {
+      "epoch": 0.5363255920901356,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.224015011461826e-05,
+      "loss": 0.8162,
+      "step": 7289
+    },
+    {
+      "epoch": 0.5363991722234996,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.223439399276541e-05,
+      "loss": 0.7667,
+      "step": 7290
+    },
+    {
+      "epoch": 0.5364727523568636,
+      "grad_norm": 1.15625,
+      "learning_rate": 2.222863801933741e-05,
+      "loss": 1.3415,
+      "step": 7291
+    },
+    {
+      "epoch": 0.5365463324902277,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.2222882194643175e-05,
+      "loss": 0.9342,
+      "step": 7292
+    },
+    {
+      "epoch": 0.5366199126235917,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.2217126518991604e-05,
+      "loss": 0.8758,
+      "step": 7293
+    },
+    {
+      "epoch": 0.5366934927569557,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.221137099269161e-05,
+      "loss": 0.9747,
+      "step": 7294
+    },
+    {
+      "epoch": 0.5367670728903197,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.2205615616052057e-05,
+      "loss": 0.8359,
+      "step": 7295
+    },
+    {
+      "epoch": 0.5368406530236836,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.219986038938183e-05,
+      "loss": 1.2206,
+      "step": 7296
+    },
+    {
+      "epoch": 0.5369142331570476,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.2194105312989813e-05,
+      "loss": 0.6821,
+      "step": 7297
+    },
+    {
+      "epoch": 0.5369878132904116,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.218835038718487e-05,
+      "loss": 0.8949,
+      "step": 7298
+    },
+    {
+      "epoch": 0.5370613934237756,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.2182595612275848e-05,
+      "loss": 0.7176,
+      "step": 7299
+    },
+    {
+      "epoch": 0.5371349735571396,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.2176840988571594e-05,
+      "loss": 1.0669,
+      "step": 7300
+    },
+    {
+      "epoch": 0.5372085536905036,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.2171086516380955e-05,
+      "loss": 0.9216,
+      "step": 7301
+    },
+    {
+      "epoch": 0.5372821338238676,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.216533219601276e-05,
+      "loss": 0.6807,
+      "step": 7302
+    },
+    {
+      "epoch": 0.5373557139572316,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.2159578027775833e-05,
+      "loss": 0.9525,
+      "step": 7303
+    },
+    {
+      "epoch": 0.5374292940905956,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.2153824011978984e-05,
+      "loss": 0.631,
+      "step": 7304
+    },
+    {
+      "epoch": 0.5375028742239595,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.214807014893103e-05,
+      "loss": 1.2324,
+      "step": 7305
+    },
+    {
+      "epoch": 0.5375764543573235,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.2142316438940757e-05,
+      "loss": 1.3342,
+      "step": 7306
+    },
+    {
+      "epoch": 0.5376500344906875,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.213656288231698e-05,
+      "loss": 1.0739,
+      "step": 7307
+    },
+    {
+      "epoch": 0.5377236146240515,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.213080947936845e-05,
+      "loss": 0.8572,
+      "step": 7308
+    },
+    {
+      "epoch": 0.5377971947574155,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.212505623040395e-05,
+      "loss": 0.905,
+      "step": 7309
+    },
+    {
+      "epoch": 0.5378707748907795,
+      "grad_norm": 0.75,
+      "learning_rate": 2.211930313573226e-05,
+      "loss": 0.7273,
+      "step": 7310
+    },
+    {
+      "epoch": 0.5379443550241435,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.2113550195662132e-05,
+      "loss": 0.9139,
+      "step": 7311
+    },
+    {
+      "epoch": 0.5380179351575075,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.2107797410502324e-05,
+      "loss": 1.1933,
+      "step": 7312
+    },
+    {
+      "epoch": 0.5380915152908715,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.2102044780561552e-05,
+      "loss": 0.8918,
+      "step": 7313
+    },
+    {
+      "epoch": 0.5381650954242354,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.2096292306148565e-05,
+      "loss": 0.9323,
+      "step": 7314
+    },
+    {
+      "epoch": 0.5382386755575994,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.2090539987572083e-05,
+      "loss": 0.8173,
+      "step": 7315
+    },
+    {
+      "epoch": 0.5383122556909634,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.208478782514083e-05,
+      "loss": 1.0322,
+      "step": 7316
+    },
+    {
+      "epoch": 0.5383858358243274,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.207903581916351e-05,
+      "loss": 0.5873,
+      "step": 7317
+    },
+    {
+      "epoch": 0.5384594159576914,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.2073283969948816e-05,
+      "loss": 1.1847,
+      "step": 7318
+    },
+    {
+      "epoch": 0.5385329960910554,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.206753227780544e-05,
+      "loss": 1.084,
+      "step": 7319
+    },
+    {
+      "epoch": 0.5386065762244194,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.2061780743042073e-05,
+      "loss": 1.1231,
+      "step": 7320
+    },
+    {
+      "epoch": 0.5386801563577834,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.205602936596739e-05,
+      "loss": 0.9485,
+      "step": 7321
+    },
+    {
+      "epoch": 0.5387537364911474,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.2050278146890042e-05,
+      "loss": 0.7016,
+      "step": 7322
+    },
+    {
+      "epoch": 0.5388273166245113,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.2044527086118693e-05,
+      "loss": 1.0751,
+      "step": 7323
+    },
+    {
+      "epoch": 0.5389008967578753,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.2038776183961998e-05,
+      "loss": 0.8132,
+      "step": 7324
+    },
+    {
+      "epoch": 0.5389744768912393,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.203302544072859e-05,
+      "loss": 0.8207,
+      "step": 7325
+    },
+    {
+      "epoch": 0.5390480570246033,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.2027274856727115e-05,
+      "loss": 1.4968,
+      "step": 7326
+    },
+    {
+      "epoch": 0.5391216371579673,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.2021524432266173e-05,
+      "loss": 1.3769,
+      "step": 7327
+    },
+    {
+      "epoch": 0.5391952172913314,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.2015774167654386e-05,
+      "loss": 1.0462,
+      "step": 7328
+    },
+    {
+      "epoch": 0.5392687974246954,
+      "grad_norm": 0.61328125,
+      "learning_rate": 2.2010024063200364e-05,
+      "loss": 0.5404,
+      "step": 7329
+    },
+    {
+      "epoch": 0.5393423775580594,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.200427411921271e-05,
+      "loss": 1.2398,
+      "step": 7330
+    },
+    {
+      "epoch": 0.5394159576914234,
+      "grad_norm": 1.1484375,
+      "learning_rate": 2.1998524335999998e-05,
+      "loss": 1.2768,
+      "step": 7331
+    },
+    {
+      "epoch": 0.5394895378247873,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.199277471387082e-05,
+      "loss": 0.8722,
+      "step": 7332
+    },
+    {
+      "epoch": 0.5395631179581513,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.1987025253133743e-05,
+      "loss": 1.1583,
+      "step": 7333
+    },
+    {
+      "epoch": 0.5396366980915153,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.1981275954097323e-05,
+      "loss": 0.8982,
+      "step": 7334
+    },
+    {
+      "epoch": 0.5397102782248793,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.1975526817070137e-05,
+      "loss": 1.0524,
+      "step": 7335
+    },
+    {
+      "epoch": 0.5397838583582433,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.19697778423607e-05,
+      "loss": 0.9947,
+      "step": 7336
+    },
+    {
+      "epoch": 0.5398574384916073,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.1964029030277567e-05,
+      "loss": 1.2866,
+      "step": 7337
+    },
+    {
+      "epoch": 0.5399310186249713,
+      "grad_norm": 0.68359375,
+      "learning_rate": 2.1958280381129263e-05,
+      "loss": 0.707,
+      "step": 7338
+    },
+    {
+      "epoch": 0.5400045987583353,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.1952531895224313e-05,
+      "loss": 0.5779,
+      "step": 7339
+    },
+    {
+      "epoch": 0.5400781788916993,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.194678357287121e-05,
+      "loss": 0.7908,
+      "step": 7340
+    },
+    {
+      "epoch": 0.5401517590250632,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.194103541437847e-05,
+      "loss": 1.3806,
+      "step": 7341
+    },
+    {
+      "epoch": 0.5402253391584272,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.193528742005458e-05,
+      "loss": 0.8333,
+      "step": 7342
+    },
+    {
+      "epoch": 0.5402989192917912,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.192953959020802e-05,
+      "loss": 1.1764,
+      "step": 7343
+    },
+    {
+      "epoch": 0.5403724994251552,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.1923791925147285e-05,
+      "loss": 0.8591,
+      "step": 7344
+    },
+    {
+      "epoch": 0.5404460795585192,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.191804442518082e-05,
+      "loss": 0.6764,
+      "step": 7345
+    },
+    {
+      "epoch": 0.5405196596918832,
+      "grad_norm": 0.6484375,
+      "learning_rate": 2.1912297090617084e-05,
+      "loss": 0.5337,
+      "step": 7346
+    },
+    {
+      "epoch": 0.5405932398252472,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.1906549921764535e-05,
+      "loss": 0.9451,
+      "step": 7347
+    },
+    {
+      "epoch": 0.5406668199586112,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.190080291893161e-05,
+      "loss": 0.6899,
+      "step": 7348
+    },
+    {
+      "epoch": 0.5407404000919752,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.1895056082426744e-05,
+      "loss": 0.7317,
+      "step": 7349
+    },
+    {
+      "epoch": 0.5408139802253391,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.1889309412558346e-05,
+      "loss": 0.7372,
+      "step": 7350
+    },
+    {
+      "epoch": 0.5408875603587031,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.188356290963484e-05,
+      "loss": 1.036,
+      "step": 7351
+    },
+    {
+      "epoch": 0.5409611404920671,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.1877816573964626e-05,
+      "loss": 0.9117,
+      "step": 7352
+    },
+    {
+      "epoch": 0.5410347206254311,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.1872070405856106e-05,
+      "loss": 0.9015,
+      "step": 7353
+    },
+    {
+      "epoch": 0.5411083007587951,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.186632440561765e-05,
+      "loss": 0.8015,
+      "step": 7354
+    },
+    {
+      "epoch": 0.5411818808921591,
+      "grad_norm": 1.125,
+      "learning_rate": 2.186057857355765e-05,
+      "loss": 1.2927,
+      "step": 7355
+    },
+    {
+      "epoch": 0.5412554610255231,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.1854832909984463e-05,
+      "loss": 0.5719,
+      "step": 7356
+    },
+    {
+      "epoch": 0.5413290411588871,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.1849087415206455e-05,
+      "loss": 1.0546,
+      "step": 7357
+    },
+    {
+      "epoch": 0.5414026212922511,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.184334208953198e-05,
+      "loss": 0.7692,
+      "step": 7358
+    },
+    {
+      "epoch": 0.541476201425615,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.1837596933269366e-05,
+      "loss": 1.1095,
+      "step": 7359
+    },
+    {
+      "epoch": 0.541549781558979,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.1831851946726953e-05,
+      "loss": 0.5641,
+      "step": 7360
+    },
+    {
+      "epoch": 0.541623361692343,
+      "grad_norm": 0.875,
+      "learning_rate": 2.182610713021306e-05,
+      "loss": 1.0079,
+      "step": 7361
+    },
+    {
+      "epoch": 0.541696941825707,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.182036248403601e-05,
+      "loss": 0.9787,
+      "step": 7362
+    },
+    {
+      "epoch": 0.541770521959071,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.1814618008504094e-05,
+      "loss": 0.9466,
+      "step": 7363
+    },
+    {
+      "epoch": 0.541844102092435,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.1808873703925616e-05,
+      "loss": 0.6536,
+      "step": 7364
+    },
+    {
+      "epoch": 0.5419176822257991,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.1803129570608858e-05,
+      "loss": 0.8832,
+      "step": 7365
+    },
+    {
+      "epoch": 0.5419912623591631,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.1797385608862093e-05,
+      "loss": 0.6238,
+      "step": 7366
+    },
+    {
+      "epoch": 0.5420648424925271,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.179164181899361e-05,
+      "loss": 1.0085,
+      "step": 7367
+    },
+    {
+      "epoch": 0.542138422625891,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.178589820131164e-05,
+      "loss": 0.8215,
+      "step": 7368
+    },
+    {
+      "epoch": 0.542212002759255,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.178015475612444e-05,
+      "loss": 0.6883,
+      "step": 7369
+    },
+    {
+      "epoch": 0.542285582892619,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.1774411483740255e-05,
+      "loss": 0.9847,
+      "step": 7370
+    },
+    {
+      "epoch": 0.542359163025983,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.1768668384467316e-05,
+      "loss": 1.1023,
+      "step": 7371
+    },
+    {
+      "epoch": 0.542432743159347,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.176292545861384e-05,
+      "loss": 0.6608,
+      "step": 7372
+    },
+    {
+      "epoch": 0.542506323292711,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.175718270648804e-05,
+      "loss": 0.9281,
+      "step": 7373
+    },
+    {
+      "epoch": 0.542579903426075,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.1751440128398115e-05,
+      "loss": 0.8566,
+      "step": 7374
+    },
+    {
+      "epoch": 0.542653483559439,
+      "grad_norm": 0.859375,
+      "learning_rate": 2.1745697724652268e-05,
+      "loss": 1.1914,
+      "step": 7375
+    },
+    {
+      "epoch": 0.542727063692803,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.1739955495558678e-05,
+      "loss": 0.9672,
+      "step": 7376
+    },
+    {
+      "epoch": 0.5428006438261669,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.1734213441425516e-05,
+      "loss": 0.8312,
+      "step": 7377
+    },
+    {
+      "epoch": 0.5428742239595309,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.1728471562560955e-05,
+      "loss": 1.1319,
+      "step": 7378
+    },
+    {
+      "epoch": 0.5429478040928949,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.172272985927314e-05,
+      "loss": 0.6857,
+      "step": 7379
+    },
+    {
+      "epoch": 0.5430213842262589,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.1716988331870236e-05,
+      "loss": 0.8395,
+      "step": 7380
+    },
+    {
+      "epoch": 0.5430949643596229,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.171124698066036e-05,
+      "loss": 0.778,
+      "step": 7381
+    },
+    {
+      "epoch": 0.5431685444929869,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.1705505805951644e-05,
+      "loss": 0.9363,
+      "step": 7382
+    },
+    {
+      "epoch": 0.5432421246263509,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.1699764808052206e-05,
+      "loss": 1.06,
+      "step": 7383
+    },
+    {
+      "epoch": 0.5433157047597149,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.169402398727016e-05,
+      "loss": 1.117,
+      "step": 7384
+    },
+    {
+      "epoch": 0.5433892848930789,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.168828334391361e-05,
+      "loss": 0.7153,
+      "step": 7385
+    },
+    {
+      "epoch": 0.5434628650264428,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.168254287829063e-05,
+      "loss": 0.736,
+      "step": 7386
+    },
+    {
+      "epoch": 0.5435364451598068,
+      "grad_norm": 1.359375,
+      "learning_rate": 2.167680259070931e-05,
+      "loss": 0.6422,
+      "step": 7387
+    },
+    {
+      "epoch": 0.5436100252931708,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.1671062481477718e-05,
+      "loss": 0.7089,
+      "step": 7388
+    },
+    {
+      "epoch": 0.5436836054265348,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.1665322550903914e-05,
+      "loss": 0.9998,
+      "step": 7389
+    },
+    {
+      "epoch": 0.5437571855598988,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.1659582799295955e-05,
+      "loss": 0.9858,
+      "step": 7390
+    },
+    {
+      "epoch": 0.5438307656932628,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.1653843226961872e-05,
+      "loss": 1.0957,
+      "step": 7391
+    },
+    {
+      "epoch": 0.5439043458266268,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.1648103834209707e-05,
+      "loss": 0.9069,
+      "step": 7392
+    },
+    {
+      "epoch": 0.5439779259599908,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.1642364621347476e-05,
+      "loss": 1.0202,
+      "step": 7393
+    },
+    {
+      "epoch": 0.5440515060933548,
+      "grad_norm": 0.67578125,
+      "learning_rate": 2.1636625588683206e-05,
+      "loss": 0.7816,
+      "step": 7394
+    },
+    {
+      "epoch": 0.5441250862267187,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.1630886736524873e-05,
+      "loss": 0.9869,
+      "step": 7395
+    },
+    {
+      "epoch": 0.5441986663600827,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.1625148065180492e-05,
+      "loss": 1.2544,
+      "step": 7396
+    },
+    {
+      "epoch": 0.5442722464934467,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.161940957495804e-05,
+      "loss": 0.7264,
+      "step": 7397
+    },
+    {
+      "epoch": 0.5443458266268107,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.1613671266165487e-05,
+      "loss": 0.8085,
+      "step": 7398
+    },
+    {
+      "epoch": 0.5444194067601748,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.1607933139110807e-05,
+      "loss": 1.1389,
+      "step": 7399
+    },
+    {
+      "epoch": 0.5444929868935388,
+      "grad_norm": 1.390625,
+      "learning_rate": 2.1602195194101944e-05,
+      "loss": 0.8613,
+      "step": 7400
+    },
+    {
+      "epoch": 0.5445665670269028,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.1596457431446848e-05,
+      "loss": 0.7618,
+      "step": 7401
+    },
+    {
+      "epoch": 0.5446401471602668,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.1590719851453455e-05,
+      "loss": 0.6643,
+      "step": 7402
+    },
+    {
+      "epoch": 0.5447137272936308,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.1584982454429688e-05,
+      "loss": 0.8465,
+      "step": 7403
+    },
+    {
+      "epoch": 0.5447873074269947,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.157924524068346e-05,
+      "loss": 1.0482,
+      "step": 7404
+    },
+    {
+      "epoch": 0.5448608875603587,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.157350821052268e-05,
+      "loss": 0.7691,
+      "step": 7405
+    },
+    {
+      "epoch": 0.5449344676937227,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.1567771364255244e-05,
+      "loss": 0.7051,
+      "step": 7406
+    },
+    {
+      "epoch": 0.5450080478270867,
+      "grad_norm": 0.75,
+      "learning_rate": 2.1562034702189033e-05,
+      "loss": 0.696,
+      "step": 7407
+    },
+    {
+      "epoch": 0.5450816279604507,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.1556298224631942e-05,
+      "loss": 0.6731,
+      "step": 7408
+    },
+    {
+      "epoch": 0.5451552080938147,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.1550561931891805e-05,
+      "loss": 0.7866,
+      "step": 7409
+    },
+    {
+      "epoch": 0.5452287882271787,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.1544825824276498e-05,
+      "loss": 0.5782,
+      "step": 7410
+    },
+    {
+      "epoch": 0.5453023683605427,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.153908990209386e-05,
+      "loss": 0.9274,
+      "step": 7411
+    },
+    {
+      "epoch": 0.5453759484939067,
+      "grad_norm": 0.6953125,
+      "learning_rate": 2.1533354165651737e-05,
+      "loss": 0.7583,
+      "step": 7412
+    },
+    {
+      "epoch": 0.5454495286272706,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.1527618615257943e-05,
+      "loss": 1.0156,
+      "step": 7413
+    },
+    {
+      "epoch": 0.5455231087606346,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.1521883251220297e-05,
+      "loss": 1.203,
+      "step": 7414
+    },
+    {
+      "epoch": 0.5455966888939986,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.1516148073846613e-05,
+      "loss": 0.7928,
+      "step": 7415
+    },
+    {
+      "epoch": 0.5456702690273626,
+      "grad_norm": 0.91796875,
+      "learning_rate": 2.151041308344468e-05,
+      "loss": 0.9713,
+      "step": 7416
+    },
+    {
+      "epoch": 0.5457438491607266,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.150467828032229e-05,
+      "loss": 0.876,
+      "step": 7417
+    },
+    {
+      "epoch": 0.5458174292940906,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.1498943664787208e-05,
+      "loss": 1.079,
+      "step": 7418
+    },
+    {
+      "epoch": 0.5458910094274546,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.149320923714721e-05,
+      "loss": 1.0431,
+      "step": 7419
+    },
+    {
+      "epoch": 0.5459645895608186,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.148747499771005e-05,
+      "loss": 0.6573,
+      "step": 7420
+    },
+    {
+      "epoch": 0.5460381696941826,
+      "grad_norm": 1.140625,
+      "learning_rate": 2.1481740946783468e-05,
+      "loss": 1.3769,
+      "step": 7421
+    },
+    {
+      "epoch": 0.5461117498275465,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.147600708467522e-05,
+      "loss": 1.1254,
+      "step": 7422
+    },
+    {
+      "epoch": 0.5461853299609105,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.1470273411692998e-05,
+      "loss": 0.5865,
+      "step": 7423
+    },
+    {
+      "epoch": 0.5462589100942745,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.146453992814454e-05,
+      "loss": 1.0619,
+      "step": 7424
+    },
+    {
+      "epoch": 0.5463324902276385,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.145880663433754e-05,
+      "loss": 0.8634,
+      "step": 7425
+    },
+    {
+      "epoch": 0.5464060703610025,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.1453073530579716e-05,
+      "loss": 0.5685,
+      "step": 7426
+    },
+    {
+      "epoch": 0.5464796504943665,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.144734061717873e-05,
+      "loss": 0.991,
+      "step": 7427
+    },
+    {
+      "epoch": 0.5465532306277305,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.144160789444226e-05,
+      "loss": 0.9245,
+      "step": 7428
+    },
+    {
+      "epoch": 0.5466268107610945,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.143587536267797e-05,
+      "loss": 0.9547,
+      "step": 7429
+    },
+    {
+      "epoch": 0.5467003908944585,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.143014302219352e-05,
+      "loss": 1.04,
+      "step": 7430
+    },
+    {
+      "epoch": 0.5467739710278224,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.1424410873296557e-05,
+      "loss": 1.016,
+      "step": 7431
+    },
+    {
+      "epoch": 0.5468475511611864,
+      "grad_norm": 1.171875,
+      "learning_rate": 2.1418678916294705e-05,
+      "loss": 1.1255,
+      "step": 7432
+    },
+    {
+      "epoch": 0.5469211312945504,
+      "grad_norm": 1.046875,
+      "learning_rate": 2.141294715149559e-05,
+      "loss": 1.1142,
+      "step": 7433
+    },
+    {
+      "epoch": 0.5469947114279144,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.1407215579206826e-05,
+      "loss": 0.9451,
+      "step": 7434
+    },
+    {
+      "epoch": 0.5470682915612785,
+      "grad_norm": 0.75390625,
+      "learning_rate": 2.140148419973603e-05,
+      "loss": 0.8057,
+      "step": 7435
+    },
+    {
+      "epoch": 0.5471418716946425,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.1395753013390765e-05,
+      "loss": 0.7773,
+      "step": 7436
+    },
+    {
+      "epoch": 0.5472154518280065,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.139002202047863e-05,
+      "loss": 0.9999,
+      "step": 7437
+    },
+    {
+      "epoch": 0.5472890319613705,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.1384291221307195e-05,
+      "loss": 1.0387,
+      "step": 7438
+    },
+    {
+      "epoch": 0.5473626120947345,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.1378560616184017e-05,
+      "loss": 0.7955,
+      "step": 7439
+    },
+    {
+      "epoch": 0.5474361922280984,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.1372830205416666e-05,
+      "loss": 1.2673,
+      "step": 7440
+    },
+    {
+      "epoch": 0.5475097723614624,
+      "grad_norm": 0.90234375,
+      "learning_rate": 2.1367099989312657e-05,
+      "loss": 0.7981,
+      "step": 7441
+    },
+    {
+      "epoch": 0.5475833524948264,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.136136996817953e-05,
+      "loss": 1.0126,
+      "step": 7442
+    },
+    {
+      "epoch": 0.5476569326281904,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.1355640142324804e-05,
+      "loss": 0.8105,
+      "step": 7443
+    },
+    {
+      "epoch": 0.5477305127615544,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.1349910512055992e-05,
+      "loss": 0.7491,
+      "step": 7444
+    },
+    {
+      "epoch": 0.5478040928949184,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.1344181077680585e-05,
+      "loss": 0.8243,
+      "step": 7445
+    },
+    {
+      "epoch": 0.5478776730282824,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.1338451839506075e-05,
+      "loss": 0.9097,
+      "step": 7446
+    },
+    {
+      "epoch": 0.5479512531616464,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.1332722797839937e-05,
+      "loss": 0.8273,
+      "step": 7447
+    },
+    {
+      "epoch": 0.5480248332950104,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.1326993952989642e-05,
+      "loss": 1.2072,
+      "step": 7448
+    },
+    {
+      "epoch": 0.5480984134283743,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.1321265305262654e-05,
+      "loss": 0.629,
+      "step": 7449
+    },
+    {
+      "epoch": 0.5481719935617383,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.13155368549664e-05,
+      "loss": 1.0074,
+      "step": 7450
+    },
+    {
+      "epoch": 0.5482455736951023,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.1309808602408323e-05,
+      "loss": 0.7728,
+      "step": 7451
+    },
+    {
+      "epoch": 0.5483191538284663,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.1304080547895845e-05,
+      "loss": 1.2414,
+      "step": 7452
+    },
+    {
+      "epoch": 0.5483927339618303,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.1298352691736386e-05,
+      "loss": 0.8012,
+      "step": 7453
+    },
+    {
+      "epoch": 0.5484663140951943,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.1292625034237358e-05,
+      "loss": 0.4832,
+      "step": 7454
+    },
+    {
+      "epoch": 0.5485398942285583,
+      "grad_norm": 1.2734375,
+      "learning_rate": 2.1286897575706132e-05,
+      "loss": 0.9493,
+      "step": 7455
+    },
+    {
+      "epoch": 0.5486134743619223,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.12811703164501e-05,
+      "loss": 1.1735,
+      "step": 7456
+    },
+    {
+      "epoch": 0.5486870544952863,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.1275443256776632e-05,
+      "loss": 1.3607,
+      "step": 7457
+    },
+    {
+      "epoch": 0.5487606346286502,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.1269716396993095e-05,
+      "loss": 0.9992,
+      "step": 7458
+    },
+    {
+      "epoch": 0.5488342147620142,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.126398973740683e-05,
+      "loss": 0.6699,
+      "step": 7459
+    },
+    {
+      "epoch": 0.5489077948953782,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.125826327832518e-05,
+      "loss": 0.9599,
+      "step": 7460
+    },
+    {
+      "epoch": 0.5489813750287422,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.1252537020055468e-05,
+      "loss": 0.617,
+      "step": 7461
+    },
+    {
+      "epoch": 0.5490549551621062,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.1246810962905024e-05,
+      "loss": 1.0797,
+      "step": 7462
+    },
+    {
+      "epoch": 0.5491285352954702,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.1241085107181148e-05,
+      "loss": 0.6466,
+      "step": 7463
+    },
+    {
+      "epoch": 0.5492021154288342,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.1235359453191132e-05,
+      "loss": 1.1459,
+      "step": 7464
+    },
+    {
+      "epoch": 0.5492756955621982,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.1229634001242256e-05,
+      "loss": 1.2396,
+      "step": 7465
+    },
+    {
+      "epoch": 0.5493492756955622,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.1223908751641805e-05,
+      "loss": 0.8712,
+      "step": 7466
+    },
+    {
+      "epoch": 0.5494228558289261,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.1218183704697053e-05,
+      "loss": 0.7721,
+      "step": 7467
+    },
+    {
+      "epoch": 0.5494964359622901,
+      "grad_norm": 0.96484375,
+      "learning_rate": 2.1212458860715225e-05,
+      "loss": 1.315,
+      "step": 7468
+    },
+    {
+      "epoch": 0.5495700160956541,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.1206734220003575e-05,
+      "loss": 0.9055,
+      "step": 7469
+    },
+    {
+      "epoch": 0.5496435962290181,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.120100978286934e-05,
+      "loss": 1.1254,
+      "step": 7470
+    },
+    {
+      "epoch": 0.5497171763623822,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.1195285549619733e-05,
+      "loss": 1.2472,
+      "step": 7471
+    },
+    {
+      "epoch": 0.5497907564957462,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.118956152056197e-05,
+      "loss": 0.7265,
+      "step": 7472
+    },
+    {
+      "epoch": 0.5498643366291102,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.118383769600324e-05,
+      "loss": 1.5356,
+      "step": 7473
+    },
+    {
+      "epoch": 0.5499379167624742,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.1178114076250727e-05,
+      "loss": 0.8452,
+      "step": 7474
+    },
+    {
+      "epoch": 0.5500114968958382,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.1172390661611618e-05,
+      "loss": 0.8157,
+      "step": 7475
+    },
+    {
+      "epoch": 0.5500850770292021,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.1166667452393082e-05,
+      "loss": 0.8626,
+      "step": 7476
+    },
+    {
+      "epoch": 0.5501586571625661,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.1160944448902254e-05,
+      "loss": 1.1232,
+      "step": 7477
+    },
+    {
+      "epoch": 0.5502322372959301,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.1155221651446278e-05,
+      "loss": 1.1201,
+      "step": 7478
+    },
+    {
+      "epoch": 0.5503058174292941,
+      "grad_norm": 0.66015625,
+      "learning_rate": 2.1149499060332302e-05,
+      "loss": 0.8872,
+      "step": 7479
+    },
+    {
+      "epoch": 0.5503793975626581,
+      "grad_norm": 0.87109375,
+      "learning_rate": 2.114377667586744e-05,
+      "loss": 1.3235,
+      "step": 7480
+    },
+    {
+      "epoch": 0.5504529776960221,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.1138054498358808e-05,
+      "loss": 0.6922,
+      "step": 7481
+    },
+    {
+      "epoch": 0.5505265578293861,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.113233252811349e-05,
+      "loss": 0.5839,
+      "step": 7482
+    },
+    {
+      "epoch": 0.5506001379627501,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.1126610765438574e-05,
+      "loss": 0.7273,
+      "step": 7483
+    },
+    {
+      "epoch": 0.5506737180961141,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.1120889210641152e-05,
+      "loss": 0.626,
+      "step": 7484
+    },
+    {
+      "epoch": 0.550747298229478,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.1115167864028278e-05,
+      "loss": 0.7516,
+      "step": 7485
+    },
+    {
+      "epoch": 0.550820878362842,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.1109446725907003e-05,
+      "loss": 1.0033,
+      "step": 7486
+    },
+    {
+      "epoch": 0.550894458496206,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.1103725796584374e-05,
+      "loss": 1.243,
+      "step": 7487
+    },
+    {
+      "epoch": 0.55096803862957,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.109800507636742e-05,
+      "loss": 1.1234,
+      "step": 7488
+    },
+    {
+      "epoch": 0.551041618762934,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.1092284565563168e-05,
+      "loss": 0.795,
+      "step": 7489
+    },
+    {
+      "epoch": 0.551115198896298,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.1086564264478635e-05,
+      "loss": 0.94,
+      "step": 7490
+    },
+    {
+      "epoch": 0.551188779029662,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.108084417342079e-05,
+      "loss": 1.2097,
+      "step": 7491
+    },
+    {
+      "epoch": 0.551262359163026,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.1075124292696636e-05,
+      "loss": 0.8736,
+      "step": 7492
+    },
+    {
+      "epoch": 0.55133593929639,
+      "grad_norm": 1.3125,
+      "learning_rate": 2.106940462261315e-05,
+      "loss": 1.0572,
+      "step": 7493
+    },
+    {
+      "epoch": 0.5514095194297539,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.1063685163477296e-05,
+      "loss": 0.6779,
+      "step": 7494
+    },
+    {
+      "epoch": 0.5514830995631179,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.1057965915596034e-05,
+      "loss": 0.703,
+      "step": 7495
+    },
+    {
+      "epoch": 0.5515566796964819,
+      "grad_norm": 0.69140625,
+      "learning_rate": 2.1052246879276287e-05,
+      "loss": 0.6112,
+      "step": 7496
+    },
+    {
+      "epoch": 0.5516302598298459,
+      "grad_norm": 0.69921875,
+      "learning_rate": 2.104652805482499e-05,
+      "loss": 0.9929,
+      "step": 7497
+    },
+    {
+      "epoch": 0.5517038399632099,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.1040809442549068e-05,
+      "loss": 0.7051,
+      "step": 7498
+    },
+    {
+      "epoch": 0.5517774200965739,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.103509104275543e-05,
+      "loss": 0.8016,
+      "step": 7499
+    },
+    {
+      "epoch": 0.5518510002299379,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.1029372855750962e-05,
+      "loss": 0.8743,
+      "step": 7500
+    },
+    {
+      "epoch": 0.5519245803633019,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.1023654881842553e-05,
+      "loss": 1.0809,
+      "step": 7501
+    },
+    {
+      "epoch": 0.551998160496666,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.101793712133708e-05,
+      "loss": 1.3755,
+      "step": 7502
+    },
+    {
+      "epoch": 0.5520717406300298,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.1012219574541395e-05,
+      "loss": 0.5926,
+      "step": 7503
+    },
+    {
+      "epoch": 0.5521453207633938,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.1006502241762365e-05,
+      "loss": 0.9977,
+      "step": 7504
+    },
+    {
+      "epoch": 0.5522189008967578,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.1000785123306804e-05,
+      "loss": 0.7903,
+      "step": 7505
+    },
+    {
+      "epoch": 0.5522924810301219,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.0995068219481555e-05,
+      "loss": 0.812,
+      "step": 7506
+    },
+    {
+      "epoch": 0.5523660611634859,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.098935153059343e-05,
+      "loss": 0.8522,
+      "step": 7507
+    },
+    {
+      "epoch": 0.5524396412968499,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.0983635056949246e-05,
+      "loss": 0.7858,
+      "step": 7508
+    },
+    {
+      "epoch": 0.5525132214302139,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.0977918798855774e-05,
+      "loss": 0.7539,
+      "step": 7509
+    },
+    {
+      "epoch": 0.5525868015635779,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.0972202756619798e-05,
+      "loss": 1.0636,
+      "step": 7510
+    },
+    {
+      "epoch": 0.5526603816969419,
+      "grad_norm": 0.8203125,
+      "learning_rate": 2.0966486930548097e-05,
+      "loss": 0.6219,
+      "step": 7511
+    },
+    {
+      "epoch": 0.5527339618303058,
+      "grad_norm": 0.82421875,
+      "learning_rate": 2.0960771320947423e-05,
+      "loss": 0.9379,
+      "step": 7512
+    },
+    {
+      "epoch": 0.5528075419636698,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.095505592812453e-05,
+      "loss": 0.8834,
+      "step": 7513
+    },
+    {
+      "epoch": 0.5528811220970338,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.094934075238614e-05,
+      "loss": 0.8552,
+      "step": 7514
+    },
+    {
+      "epoch": 0.5529547022303978,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.0943625794038977e-05,
+      "loss": 0.9106,
+      "step": 7515
+    },
+    {
+      "epoch": 0.5530282823637618,
+      "grad_norm": 0.8828125,
+      "learning_rate": 2.093791105338976e-05,
+      "loss": 0.6977,
+      "step": 7516
+    },
+    {
+      "epoch": 0.5531018624971258,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.093219653074519e-05,
+      "loss": 1.0213,
+      "step": 7517
+    },
+    {
+      "epoch": 0.5531754426304898,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.092648222641194e-05,
+      "loss": 1.0649,
+      "step": 7518
+    },
+    {
+      "epoch": 0.5532490227638538,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.0920768140696696e-05,
+      "loss": 0.9968,
+      "step": 7519
+    },
+    {
+      "epoch": 0.5533226028972178,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.0915054273906125e-05,
+      "loss": 0.6715,
+      "step": 7520
+    },
+    {
+      "epoch": 0.5533961830305817,
+      "grad_norm": 0.72265625,
+      "learning_rate": 2.0909340626346874e-05,
+      "loss": 0.9704,
+      "step": 7521
+    },
+    {
+      "epoch": 0.5534697631639457,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.0903627198325593e-05,
+      "loss": 0.6492,
+      "step": 7522
+    },
+    {
+      "epoch": 0.5535433432973097,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.0897913990148895e-05,
+      "loss": 0.8551,
+      "step": 7523
+    },
+    {
+      "epoch": 0.5536169234306737,
+      "grad_norm": 0.95703125,
+      "learning_rate": 2.0892201002123406e-05,
+      "loss": 0.9271,
+      "step": 7524
+    },
+    {
+      "epoch": 0.5536905035640377,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.088648823455573e-05,
+      "loss": 0.826,
+      "step": 7525
+    },
+    {
+      "epoch": 0.5537640836974017,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.0880775687752464e-05,
+      "loss": 0.797,
+      "step": 7526
+    },
+    {
+      "epoch": 0.5538376638307657,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.087506336202019e-05,
+      "loss": 0.8945,
+      "step": 7527
+    },
+    {
+      "epoch": 0.5539112439641297,
+      "grad_norm": 0.625,
+      "learning_rate": 2.0869351257665467e-05,
+      "loss": 0.7587,
+      "step": 7528
+    },
+    {
+      "epoch": 0.5539848240974937,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.0863639374994863e-05,
+      "loss": 0.7524,
+      "step": 7529
+    },
+    {
+      "epoch": 0.5540584042308576,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.0857927714314922e-05,
+      "loss": 0.7819,
+      "step": 7530
+    },
+    {
+      "epoch": 0.5541319843642216,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.085221627593218e-05,
+      "loss": 1.0946,
+      "step": 7531
+    },
+    {
+      "epoch": 0.5542055644975856,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.084650506015315e-05,
+      "loss": 0.83,
+      "step": 7532
+    },
+    {
+      "epoch": 0.5542791446309496,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.0840794067284353e-05,
+      "loss": 0.9074,
+      "step": 7533
+    },
+    {
+      "epoch": 0.5543527247643136,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.083508329763228e-05,
+      "loss": 0.7968,
+      "step": 7534
+    },
+    {
+      "epoch": 0.5544263048976776,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.082937275150342e-05,
+      "loss": 0.8298,
+      "step": 7535
+    },
+    {
+      "epoch": 0.5544998850310416,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.082366242920426e-05,
+      "loss": 1.2276,
+      "step": 7536
+    },
+    {
+      "epoch": 0.5545734651644056,
+      "grad_norm": 0.77734375,
+      "learning_rate": 2.0817952331041236e-05,
+      "loss": 0.7956,
+      "step": 7537
+    },
+    {
+      "epoch": 0.5546470452977696,
+      "grad_norm": 1.125,
+      "learning_rate": 2.0812242457320812e-05,
+      "loss": 1.2686,
+      "step": 7538
+    },
+    {
+      "epoch": 0.5547206254311335,
+      "grad_norm": 0.98828125,
+      "learning_rate": 2.0806532808349426e-05,
+      "loss": 0.809,
+      "step": 7539
+    },
+    {
+      "epoch": 0.5547942055644975,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.080082338443351e-05,
+      "loss": 0.732,
+      "step": 7540
+    },
+    {
+      "epoch": 0.5548677856978615,
+      "grad_norm": 0.70703125,
+      "learning_rate": 2.0795114185879468e-05,
+      "loss": 0.8356,
+      "step": 7541
+    },
+    {
+      "epoch": 0.5549413658312256,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.0789405212993704e-05,
+      "loss": 0.9492,
+      "step": 7542
+    },
+    {
+      "epoch": 0.5550149459645896,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.078369646608261e-05,
+      "loss": 0.8767,
+      "step": 7543
+    },
+    {
+      "epoch": 0.5550885260979536,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.0777987945452563e-05,
+      "loss": 0.8245,
+      "step": 7544
+    },
+    {
+      "epoch": 0.5551621062313176,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.0772279651409933e-05,
+      "loss": 0.8415,
+      "step": 7545
+    },
+    {
+      "epoch": 0.5552356863646816,
+      "grad_norm": 0.80078125,
+      "learning_rate": 2.0766571584261066e-05,
+      "loss": 0.7504,
+      "step": 7546
+    },
+    {
+      "epoch": 0.5553092664980456,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.0760863744312305e-05,
+      "loss": 0.8587,
+      "step": 7547
+    },
+    {
+      "epoch": 0.5553828466314095,
+      "grad_norm": 0.75,
+      "learning_rate": 2.075515613186998e-05,
+      "loss": 0.5417,
+      "step": 7548
+    },
+    {
+      "epoch": 0.5554564267647735,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.0749448747240417e-05,
+      "loss": 0.8205,
+      "step": 7549
+    },
+    {
+      "epoch": 0.5555300068981375,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.0743741590729903e-05,
+      "loss": 0.7322,
+      "step": 7550
+    },
+    {
+      "epoch": 0.5556035870315015,
+      "grad_norm": 0.7421875,
+      "learning_rate": 2.073803466264474e-05,
+      "loss": 0.7334,
+      "step": 7551
+    },
+    {
+      "epoch": 0.5556771671648655,
+      "grad_norm": 0.875,
+      "learning_rate": 2.0732327963291203e-05,
+      "loss": 0.9157,
+      "step": 7552
+    },
+    {
+      "epoch": 0.5557507472982295,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.0726621492975567e-05,
+      "loss": 0.8135,
+      "step": 7553
+    },
+    {
+      "epoch": 0.5558243274315935,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.072091525200409e-05,
+      "loss": 0.6913,
+      "step": 7554
+    },
+    {
+      "epoch": 0.5558979075649575,
+      "grad_norm": 0.76171875,
+      "learning_rate": 2.0715209240683005e-05,
+      "loss": 0.5635,
+      "step": 7555
+    },
+    {
+      "epoch": 0.5559714876983215,
+      "grad_norm": 1.109375,
+      "learning_rate": 2.0709503459318544e-05,
+      "loss": 1.2376,
+      "step": 7556
+    },
+    {
+      "epoch": 0.5560450678316854,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.070379790821693e-05,
+      "loss": 0.9284,
+      "step": 7557
+    },
+    {
+      "epoch": 0.5561186479650494,
+      "grad_norm": 0.6796875,
+      "learning_rate": 2.0698092587684367e-05,
+      "loss": 1.0626,
+      "step": 7558
+    },
+    {
+      "epoch": 0.5561922280984134,
+      "grad_norm": 0.84375,
+      "learning_rate": 2.0692387498027055e-05,
+      "loss": 0.8049,
+      "step": 7559
+    },
+    {
+      "epoch": 0.5562658082317774,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.0686682639551167e-05,
+      "loss": 1.0139,
+      "step": 7560
+    },
+    {
+      "epoch": 0.5563393883651414,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.0680978012562875e-05,
+      "loss": 0.9984,
+      "step": 7561
+    },
+    {
+      "epoch": 0.5564129684985054,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.0675273617368334e-05,
+      "loss": 0.9501,
+      "step": 7562
+    },
+    {
+      "epoch": 0.5564865486318694,
+      "grad_norm": 0.65625,
+      "learning_rate": 2.0669569454273698e-05,
+      "loss": 0.7705,
+      "step": 7563
+    },
+    {
+      "epoch": 0.5565601287652334,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.0663865523585083e-05,
+      "loss": 0.9158,
+      "step": 7564
+    },
+    {
+      "epoch": 0.5566337088985974,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.0658161825608612e-05,
+      "loss": 0.8112,
+      "step": 7565
+    },
+    {
+      "epoch": 0.5567072890319613,
+      "grad_norm": 0.75,
+      "learning_rate": 2.06524583606504e-05,
+      "loss": 0.8144,
+      "step": 7566
+    },
+    {
+      "epoch": 0.5567808691653253,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.0646755129016535e-05,
+      "loss": 1.0407,
+      "step": 7567
+    },
+    {
+      "epoch": 0.5568544492986893,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.0641052131013107e-05,
+      "loss": 0.9335,
+      "step": 7568
+    },
+    {
+      "epoch": 0.5569280294320533,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.0635349366946168e-05,
+      "loss": 0.7903,
+      "step": 7569
+    },
+    {
+      "epoch": 0.5570016095654173,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.0629646837121787e-05,
+      "loss": 1.2399,
+      "step": 7570
+    },
+    {
+      "epoch": 0.5570751896987813,
+      "grad_norm": 0.875,
+      "learning_rate": 2.062394454184601e-05,
+      "loss": 1.1816,
+      "step": 7571
+    },
+    {
+      "epoch": 0.5571487698321453,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.0618242481424864e-05,
+      "loss": 0.9437,
+      "step": 7572
+    },
+    {
+      "epoch": 0.5572223499655093,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.0612540656164365e-05,
+      "loss": 0.9093,
+      "step": 7573
+    },
+    {
+      "epoch": 0.5572959300988733,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.060683906637052e-05,
+      "loss": 0.9084,
+      "step": 7574
+    },
+    {
+      "epoch": 0.5573695102322372,
+      "grad_norm": 0.765625,
+      "learning_rate": 2.0601137712349328e-05,
+      "loss": 1.0847,
+      "step": 7575
+    },
+    {
+      "epoch": 0.5574430903656012,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.0595436594406764e-05,
+      "loss": 0.9812,
+      "step": 7576
+    },
+    {
+      "epoch": 0.5575166704989652,
+      "grad_norm": 0.921875,
+      "learning_rate": 2.0589735712848813e-05,
+      "loss": 0.9761,
+      "step": 7577
+    },
+    {
+      "epoch": 0.5575902506323293,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.05840350679814e-05,
+      "loss": 1.1286,
+      "step": 7578
+    },
+    {
+      "epoch": 0.5576638307656933,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.057833466011049e-05,
+      "loss": 0.7901,
+      "step": 7579
+    },
+    {
+      "epoch": 0.5577374108990573,
+      "grad_norm": 0.79296875,
+      "learning_rate": 2.0572634489542007e-05,
+      "loss": 0.9269,
+      "step": 7580
+    },
+    {
+      "epoch": 0.5578109910324213,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.0566934556581874e-05,
+      "loss": 0.588,
+      "step": 7581
+    },
+    {
+      "epoch": 0.5578845711657853,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.0561234861535983e-05,
+      "loss": 1.3045,
+      "step": 7582
+    },
+    {
+      "epoch": 0.5579581512991493,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.0555535404710237e-05,
+      "loss": 0.7659,
+      "step": 7583
+    },
+    {
+      "epoch": 0.5580317314325132,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.0549836186410508e-05,
+      "loss": 0.8927,
+      "step": 7584
+    },
+    {
+      "epoch": 0.5581053115658772,
+      "grad_norm": 1.1015625,
+      "learning_rate": 2.0544137206942665e-05,
+      "loss": 1.1027,
+      "step": 7585
+    },
+    {
+      "epoch": 0.5581788916992412,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.053843846661257e-05,
+      "loss": 1.3398,
+      "step": 7586
+    },
+    {
+      "epoch": 0.5582524718326052,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.053273996572605e-05,
+      "loss": 0.9436,
+      "step": 7587
+    },
+    {
+      "epoch": 0.5583260519659692,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.052704170458894e-05,
+      "loss": 1.2582,
+      "step": 7588
+    },
+    {
+      "epoch": 0.5583996320993332,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.0521343683507047e-05,
+      "loss": 0.6782,
+      "step": 7589
+    },
+    {
+      "epoch": 0.5584732122326972,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.0515645902786184e-05,
+      "loss": 0.8031,
+      "step": 7590
+    },
+    {
+      "epoch": 0.5585467923660612,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.050994836273215e-05,
+      "loss": 0.6921,
+      "step": 7591
+    },
+    {
+      "epoch": 0.5586203724994252,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.050425106365069e-05,
+      "loss": 0.9654,
+      "step": 7592
+    },
+    {
+      "epoch": 0.5586939526327891,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.0498554005847588e-05,
+      "loss": 0.7598,
+      "step": 7593
+    },
+    {
+      "epoch": 0.5587675327661531,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.049285718962859e-05,
+      "loss": 0.8186,
+      "step": 7594
+    },
+    {
+      "epoch": 0.5588411128995171,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.0487160615299437e-05,
+      "loss": 0.9807,
+      "step": 7595
+    },
+    {
+      "epoch": 0.5589146930328811,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.0481464283165847e-05,
+      "loss": 1.0518,
+      "step": 7596
+    },
+    {
+      "epoch": 0.5589882731662451,
+      "grad_norm": 1.0703125,
+      "learning_rate": 2.047576819353353e-05,
+      "loss": 1.0099,
+      "step": 7597
+    },
+    {
+      "epoch": 0.5590618532996091,
+      "grad_norm": 0.875,
+      "learning_rate": 2.0470072346708196e-05,
+      "loss": 1.2192,
+      "step": 7598
+    },
+    {
+      "epoch": 0.5591354334329731,
+      "grad_norm": 0.9921875,
+      "learning_rate": 2.046437674299552e-05,
+      "loss": 1.0461,
+      "step": 7599
+    },
+    {
+      "epoch": 0.5592090135663371,
+      "grad_norm": 0.85546875,
+      "learning_rate": 2.0458681382701184e-05,
+      "loss": 1.4193,
+      "step": 7600
+    },
+    {
+      "epoch": 0.5592825936997011,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.0452986266130834e-05,
+      "loss": 1.0968,
+      "step": 7601
+    },
+    {
+      "epoch": 0.559356173833065,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.0447291393590124e-05,
+      "loss": 1.3195,
+      "step": 7602
+    },
+    {
+      "epoch": 0.559429753966429,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.0441596765384684e-05,
+      "loss": 0.8513,
+      "step": 7603
+    },
+    {
+      "epoch": 0.559503334099793,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.043590238182015e-05,
+      "loss": 1.076,
+      "step": 7604
+    },
+    {
+      "epoch": 0.559576914233157,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.0430208243202104e-05,
+      "loss": 0.9032,
+      "step": 7605
+    },
+    {
+      "epoch": 0.559650494366521,
+      "grad_norm": 0.9296875,
+      "learning_rate": 2.042451434983615e-05,
+      "loss": 0.9369,
+      "step": 7606
+    },
+    {
+      "epoch": 0.559724074499885,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.0418820702027866e-05,
+      "loss": 1.3684,
+      "step": 7607
+    },
+    {
+      "epoch": 0.559797654633249,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.0413127300082818e-05,
+      "loss": 0.9994,
+      "step": 7608
+    },
+    {
+      "epoch": 0.559871234766613,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.040743414430658e-05,
+      "loss": 0.7899,
+      "step": 7609
+    },
+    {
+      "epoch": 0.559944814899977,
+      "grad_norm": 1.0859375,
+      "learning_rate": 2.040174123500467e-05,
+      "loss": 1.2626,
+      "step": 7610
+    },
+    {
+      "epoch": 0.5600183950333409,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.039604857248262e-05,
+      "loss": 0.8315,
+      "step": 7611
+    },
+    {
+      "epoch": 0.560091975166705,
+      "grad_norm": 0.86328125,
+      "learning_rate": 2.039035615704595e-05,
+      "loss": 0.7149,
+      "step": 7612
+    },
+    {
+      "epoch": 0.560165555300069,
+      "grad_norm": 1.0234375,
+      "learning_rate": 2.0384663989000162e-05,
+      "loss": 1.2585,
+      "step": 7613
+    },
+    {
+      "epoch": 0.560239135433433,
+      "grad_norm": 0.7734375,
+      "learning_rate": 2.0378972068650736e-05,
+      "loss": 0.9356,
+      "step": 7614
+    },
+    {
+      "epoch": 0.560312715566797,
+      "grad_norm": 0.9609375,
+      "learning_rate": 2.037328039630315e-05,
+      "loss": 1.1026,
+      "step": 7615
+    },
+    {
+      "epoch": 0.560386295700161,
+      "grad_norm": 0.98046875,
+      "learning_rate": 2.0367588972262867e-05,
+      "loss": 0.9621,
+      "step": 7616
+    },
+    {
+      "epoch": 0.560459875833525,
+      "grad_norm": 1.0078125,
+      "learning_rate": 2.036189779683534e-05,
+      "loss": 1.2336,
+      "step": 7617
+    },
+    {
+      "epoch": 0.560533455966889,
+      "grad_norm": 0.89453125,
+      "learning_rate": 2.0356206870326004e-05,
+      "loss": 0.8108,
+      "step": 7618
+    },
+    {
+      "epoch": 0.560607036100253,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.035051619304026e-05,
+      "loss": 0.8924,
+      "step": 7619
+    },
+    {
+      "epoch": 0.5606806162336169,
+      "grad_norm": 0.92578125,
+      "learning_rate": 2.0344825765283535e-05,
+      "loss": 0.7975,
+      "step": 7620
+    },
+    {
+      "epoch": 0.5607541963669809,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.0339135587361213e-05,
+      "loss": 1.0401,
+      "step": 7621
+    },
+    {
+      "epoch": 0.5608277765003449,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.0333445659578695e-05,
+      "loss": 0.5405,
+      "step": 7622
+    },
+    {
+      "epoch": 0.5609013566337089,
+      "grad_norm": 0.94921875,
+      "learning_rate": 2.0327755982241326e-05,
+      "loss": 0.8101,
+      "step": 7623
+    },
+    {
+      "epoch": 0.5609749367670729,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.0322066555654467e-05,
+      "loss": 0.6919,
+      "step": 7624
+    },
+    {
+      "epoch": 0.5610485169004369,
+      "grad_norm": 0.99609375,
+      "learning_rate": 2.0316377380123465e-05,
+      "loss": 1.1004,
+      "step": 7625
+    },
+    {
+      "epoch": 0.5611220970338009,
+      "grad_norm": 0.62890625,
+      "learning_rate": 2.0310688455953637e-05,
+      "loss": 0.5486,
+      "step": 7626
+    },
+    {
+      "epoch": 0.5611956771671649,
+      "grad_norm": 0.93359375,
+      "learning_rate": 2.030499978345031e-05,
+      "loss": 1.0957,
+      "step": 7627
+    },
+    {
+      "epoch": 0.5612692573005289,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.0299311362918775e-05,
+      "loss": 1.3608,
+      "step": 7628
+    },
+    {
+      "epoch": 0.5613428374338928,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.0293623194664317e-05,
+      "loss": 0.8691,
+      "step": 7629
+    },
+    {
+      "epoch": 0.5614164175672568,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.0287935278992214e-05,
+      "loss": 0.9023,
+      "step": 7630
+    },
+    {
+      "epoch": 0.5614899977006208,
+      "grad_norm": 0.88671875,
+      "learning_rate": 2.0282247616207727e-05,
+      "loss": 0.7869,
+      "step": 7631
+    },
+    {
+      "epoch": 0.5615635778339848,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.027656020661611e-05,
+      "loss": 1.2842,
+      "step": 7632
+    },
+    {
+      "epoch": 0.5616371579673488,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.0270873050522566e-05,
+      "loss": 0.7527,
+      "step": 7633
+    },
+    {
+      "epoch": 0.5617107381007128,
+      "grad_norm": 0.96875,
+      "learning_rate": 2.0265186148232343e-05,
+      "loss": 1.1703,
+      "step": 7634
+    },
+    {
+      "epoch": 0.5617843182340768,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.025949950005063e-05,
+      "loss": 1.009,
+      "step": 7635
+    },
+    {
+      "epoch": 0.5618578983674408,
+      "grad_norm": 0.8046875,
+      "learning_rate": 2.025381310628264e-05,
+      "loss": 1.2368,
+      "step": 7636
+    },
+    {
+      "epoch": 0.5619314785008048,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.0248126967233524e-05,
+      "loss": 1.1802,
+      "step": 7637
+    },
+    {
+      "epoch": 0.5620050586341687,
+      "grad_norm": 0.65234375,
+      "learning_rate": 2.024244108320846e-05,
+      "loss": 0.6024,
+      "step": 7638
+    },
+    {
+      "epoch": 0.5620786387675327,
+      "grad_norm": 0.83203125,
+      "learning_rate": 2.02367554545126e-05,
+      "loss": 0.7543,
+      "step": 7639
+    },
+    {
+      "epoch": 0.5621522189008967,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.0231070081451076e-05,
+      "loss": 0.6373,
+      "step": 7640
+    },
+    {
+      "epoch": 0.5622257990342607,
+      "grad_norm": 0.74609375,
+      "learning_rate": 2.022538496432902e-05,
+      "loss": 0.7917,
+      "step": 7641
+    },
+    {
+      "epoch": 0.5622993791676247,
+      "grad_norm": 0.953125,
+      "learning_rate": 2.0219700103451528e-05,
+      "loss": 0.7983,
+      "step": 7642
+    },
+    {
+      "epoch": 0.5623729593009887,
+      "grad_norm": 1.0546875,
+      "learning_rate": 2.02140154991237e-05,
+      "loss": 1.0568,
+      "step": 7643
+    },
+    {
+      "epoch": 0.5624465394343527,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.0208331151650626e-05,
+      "loss": 0.6334,
+      "step": 7644
+    },
+    {
+      "epoch": 0.5625201195677167,
+      "grad_norm": 0.890625,
+      "learning_rate": 2.0202647061337376e-05,
+      "loss": 0.8103,
+      "step": 7645
+    },
+    {
+      "epoch": 0.5625936997010808,
+      "grad_norm": 0.703125,
+      "learning_rate": 2.019696322848899e-05,
+      "loss": 0.8565,
+      "step": 7646
+    },
+    {
+      "epoch": 0.5626672798344446,
+      "grad_norm": 0.84765625,
+      "learning_rate": 2.0191279653410513e-05,
+      "loss": 1.0922,
+      "step": 7647
+    },
+    {
+      "epoch": 0.5627408599678086,
+      "grad_norm": 0.984375,
+      "learning_rate": 2.018559633640697e-05,
+      "loss": 0.8309,
+      "step": 7648
+    },
+    {
+      "epoch": 0.5628144401011727,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.0179913277783386e-05,
+      "loss": 0.7959,
+      "step": 7649
+    },
+    {
+      "epoch": 0.5628880202345367,
+      "grad_norm": 0.8515625,
+      "learning_rate": 2.017423047784476e-05,
+      "loss": 1.374,
+      "step": 7650
+    },
+    {
+      "epoch": 0.5629616003679007,
+      "grad_norm": 0.828125,
+      "learning_rate": 2.016854793689606e-05,
+      "loss": 0.7029,
+      "step": 7651
+    },
+    {
+      "epoch": 0.5630351805012647,
+      "grad_norm": 0.73828125,
+      "learning_rate": 2.016286565524227e-05,
+      "loss": 0.7069,
+      "step": 7652
+    },
+    {
+      "epoch": 0.5631087606346287,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.015718363318834e-05,
+      "loss": 0.4827,
+      "step": 7653
+    },
+    {
+      "epoch": 0.5631823407679927,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.0151501871039224e-05,
+      "loss": 0.6133,
+      "step": 7654
+    },
+    {
+      "epoch": 0.5632559209013567,
+      "grad_norm": 0.8984375,
+      "learning_rate": 2.014582036909984e-05,
+      "loss": 0.8943,
+      "step": 7655
+    },
+    {
+      "epoch": 0.5633295010347206,
+      "grad_norm": 0.80859375,
+      "learning_rate": 2.014013912767511e-05,
+      "loss": 0.678,
+      "step": 7656
+    },
+    {
+      "epoch": 0.5634030811680846,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.013445814706993e-05,
+      "loss": 0.5697,
+      "step": 7657
+    },
+    {
+      "epoch": 0.5634766613014486,
+      "grad_norm": 0.71484375,
+      "learning_rate": 2.0128777427589198e-05,
+      "loss": 0.6967,
+      "step": 7658
+    },
+    {
+      "epoch": 0.5635502414348126,
+      "grad_norm": 0.8125,
+      "learning_rate": 2.0123096969537787e-05,
+      "loss": 1.3641,
+      "step": 7659
+    },
+    {
+      "epoch": 0.5636238215681766,
+      "grad_norm": 0.640625,
+      "learning_rate": 2.011741677322054e-05,
+      "loss": 0.666,
+      "step": 7660
+    },
+    {
+      "epoch": 0.5636974017015406,
+      "grad_norm": 0.6875,
+      "learning_rate": 2.0111736838942306e-05,
+      "loss": 0.5978,
+      "step": 7661
+    },
+    {
+      "epoch": 0.5637709818349046,
+      "grad_norm": 0.87890625,
+      "learning_rate": 2.0106057167007934e-05,
+      "loss": 0.6678,
+      "step": 7662
+    },
+    {
+      "epoch": 0.5638445619682686,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.010037775772223e-05,
+      "loss": 1.0527,
+      "step": 7663
+    },
+    {
+      "epoch": 0.5639181421016326,
+      "grad_norm": 0.9140625,
+      "learning_rate": 2.0094698611390005e-05,
+      "loss": 0.9668,
+      "step": 7664
+    },
+    {
+      "epoch": 0.5639917222349966,
+      "grad_norm": 0.91015625,
+      "learning_rate": 2.0089019728316027e-05,
+      "loss": 0.9403,
+      "step": 7665
+    },
+    {
+      "epoch": 0.5640653023683605,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.008334110880509e-05,
+      "loss": 1.039,
+      "step": 7666
+    },
+    {
+      "epoch": 0.5641388825017245,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.0077662753161946e-05,
+      "loss": 0.8635,
+      "step": 7667
+    },
+    {
+      "epoch": 0.5642124626350885,
+      "grad_norm": 0.94140625,
+      "learning_rate": 2.0071984661691354e-05,
+      "loss": 1.0429,
+      "step": 7668
+    },
+    {
+      "epoch": 0.5642860427684525,
+      "grad_norm": 0.8359375,
+      "learning_rate": 2.006630683469803e-05,
+      "loss": 0.7902,
+      "step": 7669
+    },
+    {
+      "epoch": 0.5643596229018165,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.00606292724867e-05,
+      "loss": 1.0544,
+      "step": 7670
+    },
+    {
+      "epoch": 0.5644332030351805,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.0054951975362067e-05,
+      "loss": 1.218,
+      "step": 7671
+    },
+    {
+      "epoch": 0.5645067831685445,
+      "grad_norm": 1.015625,
+      "learning_rate": 2.0049274943628822e-05,
+      "loss": 1.234,
+      "step": 7672
+    },
+    {
+      "epoch": 0.5645803633019085,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.0043598177591655e-05,
+      "loss": 0.7262,
+      "step": 7673
+    },
+    {
+      "epoch": 0.5646539434352725,
+      "grad_norm": 1.0625,
+      "learning_rate": 2.0037921677555194e-05,
+      "loss": 1.0449,
+      "step": 7674
+    },
+    {
+      "epoch": 0.5647275235686364,
+      "grad_norm": 0.7890625,
+      "learning_rate": 2.0032245443824106e-05,
+      "loss": 0.5502,
+      "step": 7675
+    },
+    {
+      "epoch": 0.5648011037020004,
+      "grad_norm": 1.0390625,
+      "learning_rate": 2.002656947670303e-05,
+      "loss": 1.3229,
+      "step": 7676
+    },
+    {
+      "epoch": 0.5648746838353644,
+      "grad_norm": 0.97265625,
+      "learning_rate": 2.002089377649658e-05,
+      "loss": 1.1109,
+      "step": 7677
+    },
+    {
+      "epoch": 0.5649482639687284,
+      "grad_norm": 0.9765625,
+      "learning_rate": 2.0015218343509347e-05,
+      "loss": 1.4111,
+      "step": 7678
+    },
+    {
+      "epoch": 0.5650218441020924,
+      "grad_norm": 0.81640625,
+      "learning_rate": 2.0009543178045932e-05,
+      "loss": 0.9055,
+      "step": 7679
+    },
+    {
+      "epoch": 0.5650954242354564,
+      "grad_norm": 0.78125,
+      "learning_rate": 2.000386828041091e-05,
+      "loss": 0.5753,
+      "step": 7680
+    },
+    {
+      "epoch": 0.5651690043688204,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9998193650908843e-05,
+      "loss": 1.1224,
+      "step": 7681
+    },
+    {
+      "epoch": 0.5652425845021845,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9992519289844274e-05,
+      "loss": 1.0986,
+      "step": 7682
+    },
+    {
+      "epoch": 0.5653161646355485,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9986845197521737e-05,
+      "loss": 0.6512,
+      "step": 7683
+    },
+    {
+      "epoch": 0.5653897447689124,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.9981171374245748e-05,
+      "loss": 1.4244,
+      "step": 7684
+    },
+    {
+      "epoch": 0.5654633249022764,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.9975497820320815e-05,
+      "loss": 0.7566,
+      "step": 7685
+    },
+    {
+      "epoch": 0.5655369050356404,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9969824536051433e-05,
+      "loss": 0.7247,
+      "step": 7686
+    },
+    {
+      "epoch": 0.5656104851690044,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9964151521742057e-05,
+      "loss": 0.6758,
+      "step": 7687
+    },
+    {
+      "epoch": 0.5656840653023684,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.995847877769715e-05,
+      "loss": 0.663,
+      "step": 7688
+    },
+    {
+      "epoch": 0.5657576454357324,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.9952806304221173e-05,
+      "loss": 0.6042,
+      "step": 7689
+    },
+    {
+      "epoch": 0.5658312255690964,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9947134101618547e-05,
+      "loss": 0.8123,
+      "step": 7690
+    },
+    {
+      "epoch": 0.5659048057024604,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.99414621701937e-05,
+      "loss": 1.0737,
+      "step": 7691
+    },
+    {
+      "epoch": 0.5659783858358244,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9935790510251013e-05,
+      "loss": 1.0011,
+      "step": 7692
+    },
+    {
+      "epoch": 0.5660519659691883,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.993011912209489e-05,
+      "loss": 0.7549,
+      "step": 7693
+    },
+    {
+      "epoch": 0.5661255461025523,
+      "grad_norm": 1.0,
+      "learning_rate": 1.9924448006029695e-05,
+      "loss": 1.2663,
+      "step": 7694
+    },
+    {
+      "epoch": 0.5661991262359163,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9918777162359787e-05,
+      "loss": 1.3989,
+      "step": 7695
+    },
+    {
+      "epoch": 0.5662727063692803,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9913106591389517e-05,
+      "loss": 0.7846,
+      "step": 7696
+    },
+    {
+      "epoch": 0.5663462865026443,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9907436293423208e-05,
+      "loss": 0.7626,
+      "step": 7697
+    },
+    {
+      "epoch": 0.5664198666360083,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.990176626876517e-05,
+      "loss": 1.2104,
+      "step": 7698
+    },
+    {
+      "epoch": 0.5664934467693723,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.989609651771971e-05,
+      "loss": 1.087,
+      "step": 7699
+    },
+    {
+      "epoch": 0.5665670269027363,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9890427040591116e-05,
+      "loss": 0.8319,
+      "step": 7700
+    },
+    {
+      "epoch": 0.5666406070361003,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.988475783768364e-05,
+      "loss": 0.7973,
+      "step": 7701
+    },
+    {
+      "epoch": 0.5667141871694642,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.9879088909301556e-05,
+      "loss": 1.1152,
+      "step": 7702
+    },
+    {
+      "epoch": 0.5667877673028282,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.9873420255749096e-05,
+      "loss": 1.093,
+      "step": 7703
+    },
+    {
+      "epoch": 0.5668613474361922,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.9867751877330488e-05,
+      "loss": 0.951,
+      "step": 7704
+    },
+    {
+      "epoch": 0.5669349275695562,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.9862083774349956e-05,
+      "loss": 1.2658,
+      "step": 7705
+    },
+    {
+      "epoch": 0.5670085077029202,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.985641594711167e-05,
+      "loss": 0.9686,
+      "step": 7706
+    },
+    {
+      "epoch": 0.5670820878362842,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.9850748395919826e-05,
+      "loss": 1.0193,
+      "step": 7707
+    },
+    {
+      "epoch": 0.5671556679696482,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.984508112107859e-05,
+      "loss": 0.7656,
+      "step": 7708
+    },
+    {
+      "epoch": 0.5672292481030122,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.983941412289212e-05,
+      "loss": 0.8086,
+      "step": 7709
+    },
+    {
+      "epoch": 0.5673028282363762,
+      "grad_norm": 0.875,
+      "learning_rate": 1.9833747401664543e-05,
+      "loss": 0.7632,
+      "step": 7710
+    },
+    {
+      "epoch": 0.5673764083697401,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9828080957699987e-05,
+      "loss": 0.4955,
+      "step": 7711
+    },
+    {
+      "epoch": 0.5674499885031041,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.982241479130255e-05,
+      "loss": 0.9354,
+      "step": 7712
+    },
+    {
+      "epoch": 0.5675235686364681,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.981674890277634e-05,
+      "loss": 1.072,
+      "step": 7713
+    },
+    {
+      "epoch": 0.5675971487698321,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.9811083292425427e-05,
+      "loss": 0.7631,
+      "step": 7714
+    },
+    {
+      "epoch": 0.5676707289031961,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.980541796055387e-05,
+      "loss": 0.9107,
+      "step": 7715
+    },
+    {
+      "epoch": 0.5677443090365601,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.9799752907465717e-05,
+      "loss": 0.9741,
+      "step": 7716
+    },
+    {
+      "epoch": 0.5678178891699242,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.9794088133465008e-05,
+      "loss": 0.9064,
+      "step": 7717
+    },
+    {
+      "epoch": 0.5678914693032882,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.9788423638855767e-05,
+      "loss": 0.9792,
+      "step": 7718
+    },
+    {
+      "epoch": 0.5679650494366522,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.978275942394197e-05,
+      "loss": 0.5625,
+      "step": 7719
+    },
+    {
+      "epoch": 0.568038629570016,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.977709548902763e-05,
+      "loss": 0.7593,
+      "step": 7720
+    },
+    {
+      "epoch": 0.5681122097033801,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.97714318344167e-05,
+      "loss": 0.6402,
+      "step": 7721
+    },
+    {
+      "epoch": 0.5681857898367441,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.9765768460413153e-05,
+      "loss": 1.0625,
+      "step": 7722
+    },
+    {
+      "epoch": 0.5682593699701081,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.9760105367320934e-05,
+      "loss": 1.1693,
+      "step": 7723
+    },
+    {
+      "epoch": 0.5683329501034721,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.9754442555443956e-05,
+      "loss": 1.3715,
+      "step": 7724
+    },
+    {
+      "epoch": 0.5684065302368361,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.9748780025086136e-05,
+      "loss": 0.9845,
+      "step": 7725
+    },
+    {
+      "epoch": 0.5684801103702001,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9743117776551377e-05,
+      "loss": 0.7554,
+      "step": 7726
+    },
+    {
+      "epoch": 0.5685536905035641,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9737455810143564e-05,
+      "loss": 0.791,
+      "step": 7727
+    },
+    {
+      "epoch": 0.5686272706369281,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.973179412616655e-05,
+      "loss": 1.1941,
+      "step": 7728
+    },
+    {
+      "epoch": 0.568700850770292,
+      "grad_norm": 0.875,
+      "learning_rate": 1.9726132724924195e-05,
+      "loss": 0.678,
+      "step": 7729
+    },
+    {
+      "epoch": 0.568774430903656,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.9720471606720338e-05,
+      "loss": 0.8929,
+      "step": 7730
+    },
+    {
+      "epoch": 0.56884801103702,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.9714810771858797e-05,
+      "loss": 0.554,
+      "step": 7731
+    },
+    {
+      "epoch": 0.568921591170384,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.970915022064339e-05,
+      "loss": 0.8933,
+      "step": 7732
+    },
+    {
+      "epoch": 0.568995171303748,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9703489953377888e-05,
+      "loss": 1.0544,
+      "step": 7733
+    },
+    {
+      "epoch": 0.569068751437112,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.9697829970366076e-05,
+      "loss": 0.9059,
+      "step": 7734
+    },
+    {
+      "epoch": 0.569142331570476,
+      "grad_norm": 1.375,
+      "learning_rate": 1.9692170271911717e-05,
+      "loss": 1.1534,
+      "step": 7735
+    },
+    {
+      "epoch": 0.56921591170384,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.9686510858318553e-05,
+      "loss": 1.1078,
+      "step": 7736
+    },
+    {
+      "epoch": 0.569289491837204,
+      "grad_norm": 1.46875,
+      "learning_rate": 1.968085172989032e-05,
+      "loss": 1.294,
+      "step": 7737
+    },
+    {
+      "epoch": 0.5693630719705679,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.9675192886930722e-05,
+      "loss": 0.9051,
+      "step": 7738
+    },
+    {
+      "epoch": 0.5694366521039319,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9669534329743467e-05,
+      "loss": 0.8163,
+      "step": 7739
+    },
+    {
+      "epoch": 0.5695102322372959,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.9663876058632235e-05,
+      "loss": 1.6457,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5695838123706599,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.9658218073900704e-05,
+      "loss": 0.618,
+      "step": 7741
+    },
+    {
+      "epoch": 0.5696573925040239,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.965256037585251e-05,
+      "loss": 0.9302,
+      "step": 7742
+    },
+    {
+      "epoch": 0.5697309726373879,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.9646902964791305e-05,
+      "loss": 1.4923,
+      "step": 7743
+    },
+    {
+      "epoch": 0.5698045527707519,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.9641245841020705e-05,
+      "loss": 0.9057,
+      "step": 7744
+    },
+    {
+      "epoch": 0.5698781329041159,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.9635589004844322e-05,
+      "loss": 1.0035,
+      "step": 7745
+    },
+    {
+      "epoch": 0.5699517130374799,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.9629932456565752e-05,
+      "loss": 0.849,
+      "step": 7746
+    },
+    {
+      "epoch": 0.5700252931708438,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.9624276196488556e-05,
+      "loss": 0.8605,
+      "step": 7747
+    },
+    {
+      "epoch": 0.5700988733042078,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.9618620224916304e-05,
+      "loss": 0.7747,
+      "step": 7748
+    },
+    {
+      "epoch": 0.5701724534375718,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.961296454215254e-05,
+      "loss": 1.1588,
+      "step": 7749
+    },
+    {
+      "epoch": 0.5702460335709358,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.96073091485008e-05,
+      "loss": 0.5469,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5703196137042998,
+      "grad_norm": 0.875,
+      "learning_rate": 1.9601654044264586e-05,
+      "loss": 0.9935,
+      "step": 7751
+    },
+    {
+      "epoch": 0.5703931938376638,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.9595999229747405e-05,
+      "loss": 0.9887,
+      "step": 7752
+    },
+    {
+      "epoch": 0.5704667739710279,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.959034470525274e-05,
+      "loss": 1.0697,
+      "step": 7753
+    },
+    {
+      "epoch": 0.5705403541043919,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.9584690471084053e-05,
+      "loss": 1.1769,
+      "step": 7754
+    },
+    {
+      "epoch": 0.5706139342377559,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.957903652754481e-05,
+      "loss": 1.0359,
+      "step": 7755
+    },
+    {
+      "epoch": 0.5706875143711198,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.957338287493843e-05,
+      "loss": 0.9603,
+      "step": 7756
+    },
+    {
+      "epoch": 0.5707610945044838,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.956772951356834e-05,
+      "loss": 0.9707,
+      "step": 7757
+    },
+    {
+      "epoch": 0.5708346746378478,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.9562076443737947e-05,
+      "loss": 0.5688,
+      "step": 7758
+    },
+    {
+      "epoch": 0.5709082547712118,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9556423665750655e-05,
+      "loss": 0.6643,
+      "step": 7759
+    },
+    {
+      "epoch": 0.5709818349045758,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.955077117990981e-05,
+      "loss": 0.8678,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5710554150379398,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9545118986518785e-05,
+      "loss": 0.7246,
+      "step": 7761
+    },
+    {
+      "epoch": 0.5711289951713038,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.953946708588092e-05,
+      "loss": 0.9106,
+      "step": 7762
+    },
+    {
+      "epoch": 0.5712025753046678,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.9533815478299543e-05,
+      "loss": 0.7549,
+      "step": 7763
+    },
+    {
+      "epoch": 0.5712761554380318,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.952816416407797e-05,
+      "loss": 1.1657,
+      "step": 7764
+    },
+    {
+      "epoch": 0.5713497355713957,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9522513143519488e-05,
+      "loss": 0.8916,
+      "step": 7765
+    },
+    {
+      "epoch": 0.5714233157047597,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.9516862416927383e-05,
+      "loss": 1.3224,
+      "step": 7766
+    },
+    {
+      "epoch": 0.5714968958381237,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.951121198460491e-05,
+      "loss": 0.8834,
+      "step": 7767
+    },
+    {
+      "epoch": 0.5715704759714877,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9505561846855326e-05,
+      "loss": 0.8804,
+      "step": 7768
+    },
+    {
+      "epoch": 0.5716440561048517,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.9499912003981864e-05,
+      "loss": 1.0037,
+      "step": 7769
+    },
+    {
+      "epoch": 0.5717176362382157,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.949426245628773e-05,
+      "loss": 1.2265,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5717912163715797,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9488613204076133e-05,
+      "loss": 0.714,
+      "step": 7771
+    },
+    {
+      "epoch": 0.5718647965049437,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.948296424765026e-05,
+      "loss": 0.9577,
+      "step": 7772
+    },
+    {
+      "epoch": 0.5719383766383077,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.947731558731328e-05,
+      "loss": 0.8312,
+      "step": 7773
+    },
+    {
+      "epoch": 0.5720119567716716,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.9471667223368333e-05,
+      "loss": 0.8835,
+      "step": 7774
+    },
+    {
+      "epoch": 0.5720855369050356,
+      "grad_norm": 0.81640625,
+      "learning_rate": 1.9466019156118565e-05,
+      "loss": 1.032,
+      "step": 7775
+    },
+    {
+      "epoch": 0.5721591170383996,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.9460371385867097e-05,
+      "loss": 0.89,
+      "step": 7776
+    },
+    {
+      "epoch": 0.5722326971717636,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.9454723912917035e-05,
+      "loss": 0.712,
+      "step": 7777
+    },
+    {
+      "epoch": 0.5723062773051276,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.9449076737571467e-05,
+      "loss": 1.0629,
+      "step": 7778
+    },
+    {
+      "epoch": 0.5723798574384916,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.9443429860133467e-05,
+      "loss": 1.1659,
+      "step": 7779
+    },
+    {
+      "epoch": 0.5724534375718556,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9437783280906086e-05,
+      "loss": 0.6964,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5725270177052196,
+      "grad_norm": 0.75390625,
+      "learning_rate": 1.9432137000192376e-05,
+      "loss": 0.9,
+      "step": 7781
+    },
+    {
+      "epoch": 0.5726005978385836,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.942649101829536e-05,
+      "loss": 1.2242,
+      "step": 7782
+    },
+    {
+      "epoch": 0.5726741779719475,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.9420845335518036e-05,
+      "loss": 0.907,
+      "step": 7783
+    },
+    {
+      "epoch": 0.5727477581053115,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.941519995216341e-05,
+      "loss": 0.5625,
+      "step": 7784
+    },
+    {
+      "epoch": 0.5728213382386755,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.940955486853445e-05,
+      "loss": 0.8627,
+      "step": 7785
+    },
+    {
+      "epoch": 0.5728949183720395,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.9403910084934128e-05,
+      "loss": 0.6583,
+      "step": 7786
+    },
+    {
+      "epoch": 0.5729684985054035,
+      "grad_norm": 0.66015625,
+      "learning_rate": 1.939826560166539e-05,
+      "loss": 0.5249,
+      "step": 7787
+    },
+    {
+      "epoch": 0.5730420786387675,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.939262141903114e-05,
+      "loss": 0.691,
+      "step": 7788
+    },
+    {
+      "epoch": 0.5731156587721316,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.9386977537334316e-05,
+      "loss": 1.3678,
+      "step": 7789
+    },
+    {
+      "epoch": 0.5731892389054956,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.93813339568778e-05,
+      "loss": 0.896,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5732628190388596,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.937569067796449e-05,
+      "loss": 0.9596,
+      "step": 7791
+    },
+    {
+      "epoch": 0.5733363991722235,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.9370047700897227e-05,
+      "loss": 0.849,
+      "step": 7792
+    },
+    {
+      "epoch": 0.5734099793055875,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.9364405025978877e-05,
+      "loss": 1.0411,
+      "step": 7793
+    },
+    {
+      "epoch": 0.5734835594389515,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.935876265351226e-05,
+      "loss": 0.7349,
+      "step": 7794
+    },
+    {
+      "epoch": 0.5735571395723155,
+      "grad_norm": 0.8359375,
+      "learning_rate": 1.9353120583800197e-05,
+      "loss": 0.716,
+      "step": 7795
+    },
+    {
+      "epoch": 0.5736307197056795,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.9347478817145492e-05,
+      "loss": 0.9886,
+      "step": 7796
+    },
+    {
+      "epoch": 0.5737042998390435,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.934183735385092e-05,
+      "loss": 0.8516,
+      "step": 7797
+    },
+    {
+      "epoch": 0.5737778799724075,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.933619619421925e-05,
+      "loss": 1.1078,
+      "step": 7798
+    },
+    {
+      "epoch": 0.5738514601057715,
+      "grad_norm": 0.875,
+      "learning_rate": 1.933055533855323e-05,
+      "loss": 1.1994,
+      "step": 7799
+    },
+    {
+      "epoch": 0.5739250402391355,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.93249147871556e-05,
+      "loss": 1.1573,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5739986203724994,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.9319274540329085e-05,
+      "loss": 0.9439,
+      "step": 7801
+    },
+    {
+      "epoch": 0.5740722005058634,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.9313634598376363e-05,
+      "loss": 0.8465,
+      "step": 7802
+    },
+    {
+      "epoch": 0.5741457806392274,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9307994961600136e-05,
+      "loss": 1.1433,
+      "step": 7803
+    },
+    {
+      "epoch": 0.5742193607725914,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.930235563030306e-05,
+      "loss": 1.1513,
+      "step": 7804
+    },
+    {
+      "epoch": 0.5742929409059554,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.929671660478781e-05,
+      "loss": 1.366,
+      "step": 7805
+    },
+    {
+      "epoch": 0.5743665210393194,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.9291077885357e-05,
+      "loss": 0.9652,
+      "step": 7806
+    },
+    {
+      "epoch": 0.5744401011726834,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.928543947231326e-05,
+      "loss": 0.8515,
+      "step": 7807
+    },
+    {
+      "epoch": 0.5745136813060474,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.927980136595919e-05,
+      "loss": 1.1187,
+      "step": 7808
+    },
+    {
+      "epoch": 0.5745872614394114,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9274163566597372e-05,
+      "loss": 1.021,
+      "step": 7809
+    },
+    {
+      "epoch": 0.5746608415727753,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9268526074530386e-05,
+      "loss": 0.9821,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5747344217061393,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.926288889006078e-05,
+      "loss": 0.8988,
+      "step": 7811
+    },
+    {
+      "epoch": 0.5748080018395033,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.9257252013491088e-05,
+      "loss": 0.7945,
+      "step": 7812
+    },
+    {
+      "epoch": 0.5748815819728673,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.9251615445123835e-05,
+      "loss": 0.8642,
+      "step": 7813
+    },
+    {
+      "epoch": 0.5749551621062313,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.9245979185261536e-05,
+      "loss": 0.7619,
+      "step": 7814
+    },
+    {
+      "epoch": 0.5750287422395953,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.9240343234206653e-05,
+      "loss": 0.549,
+      "step": 7815
+    },
+    {
+      "epoch": 0.5751023223729593,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.9234707592261672e-05,
+      "loss": 0.6836,
+      "step": 7816
+    },
+    {
+      "epoch": 0.5751759025063233,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9229072259729045e-05,
+      "loss": 0.6723,
+      "step": 7817
+    },
+    {
+      "epoch": 0.5752494826396873,
+      "grad_norm": 0.7421875,
+      "learning_rate": 1.9223437236911203e-05,
+      "loss": 0.5329,
+      "step": 7818
+    },
+    {
+      "epoch": 0.5753230627730512,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.9217802524110594e-05,
+      "loss": 1.3878,
+      "step": 7819
+    },
+    {
+      "epoch": 0.5753966429064152,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.921216812162959e-05,
+      "loss": 0.7419,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5754702230397792,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.920653402977059e-05,
+      "loss": 0.7552,
+      "step": 7821
+    },
+    {
+      "epoch": 0.5755438031731432,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.9200900248835967e-05,
+      "loss": 0.662,
+      "step": 7822
+    },
+    {
+      "epoch": 0.5756173833065072,
+      "grad_norm": 0.875,
+      "learning_rate": 1.919526677912808e-05,
+      "loss": 1.0735,
+      "step": 7823
+    },
+    {
+      "epoch": 0.5756909634398713,
+      "grad_norm": 0.9609375,
+      "learning_rate": 1.9189633620949256e-05,
+      "loss": 1.2137,
+      "step": 7824
+    },
+    {
+      "epoch": 0.5757645435732353,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.918400077460182e-05,
+      "loss": 0.5444,
+      "step": 7825
+    },
+    {
+      "epoch": 0.5758381237065993,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.917836824038808e-05,
+      "loss": 0.6851,
+      "step": 7826
+    },
+    {
+      "epoch": 0.5759117038399633,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9172736018610322e-05,
+      "loss": 0.9634,
+      "step": 7827
+    },
+    {
+      "epoch": 0.5759852839733272,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.9167104109570826e-05,
+      "loss": 0.8805,
+      "step": 7828
+    },
+    {
+      "epoch": 0.5760588641066912,
+      "grad_norm": 1.0625,
+      "learning_rate": 1.916147251357182e-05,
+      "loss": 0.9175,
+      "step": 7829
+    },
+    {
+      "epoch": 0.5761324442400552,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.915584123091556e-05,
+      "loss": 0.9094,
+      "step": 7830
+    },
+    {
+      "epoch": 0.5762060243734192,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.9150210261904257e-05,
+      "loss": 0.9626,
+      "step": 7831
+    },
+    {
+      "epoch": 0.5762796045067832,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.9144579606840142e-05,
+      "loss": 0.6984,
+      "step": 7832
+    },
+    {
+      "epoch": 0.5763531846401472,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.9138949266025362e-05,
+      "loss": 1.0027,
+      "step": 7833
+    },
+    {
+      "epoch": 0.5764267647735112,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.913331923976211e-05,
+      "loss": 0.8469,
+      "step": 7834
+    },
+    {
+      "epoch": 0.5765003449068752,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.9127689528352532e-05,
+      "loss": 0.7408,
+      "step": 7835
+    },
+    {
+      "epoch": 0.5765739250402392,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.9122060132098764e-05,
+      "loss": 0.8928,
+      "step": 7836
+    },
+    {
+      "epoch": 0.5766475051736031,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.9116431051302936e-05,
+      "loss": 0.9892,
+      "step": 7837
+    },
+    {
+      "epoch": 0.5767210853069671,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.9110802286267133e-05,
+      "loss": 0.6941,
+      "step": 7838
+    },
+    {
+      "epoch": 0.5767946654403311,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.9105173837293448e-05,
+      "loss": 0.8589,
+      "step": 7839
+    },
+    {
+      "epoch": 0.5768682455736951,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.909954570468395e-05,
+      "loss": 1.0987,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5769418257070591,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.909391788874069e-05,
+      "loss": 0.7726,
+      "step": 7841
+    },
+    {
+      "epoch": 0.5770154058404231,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.908829038976571e-05,
+      "loss": 0.7741,
+      "step": 7842
+    },
+    {
+      "epoch": 0.5770889859737871,
+      "grad_norm": 1.1328125,
+      "learning_rate": 1.9082663208061014e-05,
+      "loss": 0.9955,
+      "step": 7843
+    },
+    {
+      "epoch": 0.5771625661071511,
+      "grad_norm": 1.046875,
+      "learning_rate": 1.9077036343928596e-05,
+      "loss": 1.2145,
+      "step": 7844
+    },
+    {
+      "epoch": 0.5772361462405151,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.9071409797670462e-05,
+      "loss": 0.6294,
+      "step": 7845
+    },
+    {
+      "epoch": 0.577309726373879,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.9065783569588576e-05,
+      "loss": 0.8148,
+      "step": 7846
+    },
+    {
+      "epoch": 0.577383306507243,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.906015765998486e-05,
+      "loss": 0.9866,
+      "step": 7847
+    },
+    {
+      "epoch": 0.577456886640607,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.905453206916127e-05,
+      "loss": 0.8646,
+      "step": 7848
+    },
+    {
+      "epoch": 0.577530466773971,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.9048906797419713e-05,
+      "loss": 0.8968,
+      "step": 7849
+    },
+    {
+      "epoch": 0.577604046907335,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.9043281845062087e-05,
+      "loss": 1.5815,
+      "step": 7850
+    },
+    {
+      "epoch": 0.577677627040699,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.903765721239028e-05,
+      "loss": 0.7399,
+      "step": 7851
+    },
+    {
+      "epoch": 0.577751207174063,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.903203289970615e-05,
+      "loss": 0.694,
+      "step": 7852
+    },
+    {
+      "epoch": 0.577824787307427,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.9026408907311532e-05,
+      "loss": 0.7211,
+      "step": 7853
+    },
+    {
+      "epoch": 0.577898367440791,
+      "grad_norm": 0.92578125,
+      "learning_rate": 1.902078523550827e-05,
+      "loss": 0.979,
+      "step": 7854
+    },
+    {
+      "epoch": 0.5779719475741549,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.901516188459818e-05,
+      "loss": 1.4331,
+      "step": 7855
+    },
+    {
+      "epoch": 0.5780455277075189,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.900953885488304e-05,
+      "loss": 1.0824,
+      "step": 7856
+    },
+    {
+      "epoch": 0.5781191078408829,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.900391614666463e-05,
+      "loss": 0.7317,
+      "step": 7857
+    },
+    {
+      "epoch": 0.5781926879742469,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.899829376024472e-05,
+      "loss": 1.0449,
+      "step": 7858
+    },
+    {
+      "epoch": 0.578266268107611,
+      "grad_norm": 0.93359375,
+      "learning_rate": 1.899267169592505e-05,
+      "loss": 0.7986,
+      "step": 7859
+    },
+    {
+      "epoch": 0.578339848240975,
+      "grad_norm": 1.0,
+      "learning_rate": 1.898704995400735e-05,
+      "loss": 1.1066,
+      "step": 7860
+    },
+    {
+      "epoch": 0.578413428374339,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.8981428534793317e-05,
+      "loss": 0.6905,
+      "step": 7861
+    },
+    {
+      "epoch": 0.578487008507703,
+      "grad_norm": 0.640625,
+      "learning_rate": 1.8975807438584642e-05,
+      "loss": 0.7546,
+      "step": 7862
+    },
+    {
+      "epoch": 0.578560588641067,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.8970186665683005e-05,
+      "loss": 0.7572,
+      "step": 7863
+    },
+    {
+      "epoch": 0.5786341687744309,
+      "grad_norm": 1.1484375,
+      "learning_rate": 1.896456621639007e-05,
+      "loss": 1.2453,
+      "step": 7864
+    },
+    {
+      "epoch": 0.5787077489077949,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8958946091007458e-05,
+      "loss": 0.6999,
+      "step": 7865
+    },
+    {
+      "epoch": 0.5787813290411589,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.89533262898368e-05,
+      "loss": 0.8552,
+      "step": 7866
+    },
+    {
+      "epoch": 0.5788549091745229,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.89477068131797e-05,
+      "loss": 0.6997,
+      "step": 7867
+    },
+    {
+      "epoch": 0.5789284893078869,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.8942087661337742e-05,
+      "loss": 0.8844,
+      "step": 7868
+    },
+    {
+      "epoch": 0.5790020694412509,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.893646883461251e-05,
+      "loss": 0.9771,
+      "step": 7869
+    },
+    {
+      "epoch": 0.5790756495746149,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.8930850333305532e-05,
+      "loss": 1.0053,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5791492297079789,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.8925232157718352e-05,
+      "loss": 0.8334,
+      "step": 7871
+    },
+    {
+      "epoch": 0.5792228098413429,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.891961430815249e-05,
+      "loss": 0.9882,
+      "step": 7872
+    },
+    {
+      "epoch": 0.5792963899747068,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.8913996784909445e-05,
+      "loss": 0.9642,
+      "step": 7873
+    },
+    {
+      "epoch": 0.5793699701080708,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.8908379588290707e-05,
+      "loss": 0.8087,
+      "step": 7874
+    },
+    {
+      "epoch": 0.5794435502414348,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.890276271859772e-05,
+      "loss": 0.7891,
+      "step": 7875
+    },
+    {
+      "epoch": 0.5795171303747988,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.8897146176131945e-05,
+      "loss": 0.8852,
+      "step": 7876
+    },
+    {
+      "epoch": 0.5795907105081628,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.8891529961194804e-05,
+      "loss": 0.8946,
+      "step": 7877
+    },
+    {
+      "epoch": 0.5796642906415268,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.8885914074087722e-05,
+      "loss": 0.7218,
+      "step": 7878
+    },
+    {
+      "epoch": 0.5797378707748908,
+      "grad_norm": 1.0,
+      "learning_rate": 1.8880298515112073e-05,
+      "loss": 1.0425,
+      "step": 7879
+    },
+    {
+      "epoch": 0.5798114509082548,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.887468328456925e-05,
+      "loss": 0.5877,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5798850310416188,
+      "grad_norm": 0.953125,
+      "learning_rate": 1.8869068382760604e-05,
+      "loss": 0.9595,
+      "step": 7881
+    },
+    {
+      "epoch": 0.5799586111749827,
+      "grad_norm": 0.7890625,
+      "learning_rate": 1.8863453809987478e-05,
+      "loss": 0.7985,
+      "step": 7882
+    },
+    {
+      "epoch": 0.5800321913083467,
+      "grad_norm": 0.79296875,
+      "learning_rate": 1.8857839566551205e-05,
+      "loss": 0.734,
+      "step": 7883
+    },
+    {
+      "epoch": 0.5801057714417107,
+      "grad_norm": 0.69140625,
+      "learning_rate": 1.885222565275307e-05,
+      "loss": 0.6292,
+      "step": 7884
+    },
+    {
+      "epoch": 0.5801793515750747,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.8846612068894373e-05,
+      "loss": 0.7832,
+      "step": 7885
+    },
+    {
+      "epoch": 0.5802529317084387,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.8840998815276387e-05,
+      "loss": 1.0154,
+      "step": 7886
+    },
+    {
+      "epoch": 0.5803265118418027,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.883538589220037e-05,
+      "loss": 0.9185,
+      "step": 7887
+    },
+    {
+      "epoch": 0.5804000919751667,
+      "grad_norm": 0.84375,
+      "learning_rate": 1.882977329996754e-05,
+      "loss": 1.2152,
+      "step": 7888
+    },
+    {
+      "epoch": 0.5804736721085307,
+      "grad_norm": 0.828125,
+      "learning_rate": 1.8824161038879122e-05,
+      "loss": 0.8345,
+      "step": 7889
+    },
+    {
+      "epoch": 0.5805472522418947,
+      "grad_norm": 1.078125,
+      "learning_rate": 1.881854910923632e-05,
+      "loss": 0.6835,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5806208323752586,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.8812937511340307e-05,
+      "loss": 1.1624,
+      "step": 7891
+    },
+    {
+      "epoch": 0.5806944125086226,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.8807326245492262e-05,
+      "loss": 0.9678,
+      "step": 7892
+    },
+    {
+      "epoch": 0.5807679926419866,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8801715311993315e-05,
+      "loss": 1.1112,
+      "step": 7893
+    },
+    {
+      "epoch": 0.5808415727753506,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.87961047111446e-05,
+      "loss": 1.2048,
+      "step": 7894
+    },
+    {
+      "epoch": 0.5809151529087146,
+      "grad_norm": 0.78515625,
+      "learning_rate": 1.8790494443247225e-05,
+      "loss": 0.6695,
+      "step": 7895
+    },
+    {
+      "epoch": 0.5809887330420787,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.87848845086023e-05,
+      "loss": 0.6305,
+      "step": 7896
+    },
+    {
+      "epoch": 0.5810623131754427,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.8779274907510866e-05,
+      "loss": 0.7055,
+      "step": 7897
+    },
+    {
+      "epoch": 0.5811358933088067,
+      "grad_norm": 0.94140625,
+      "learning_rate": 1.8773665640274004e-05,
+      "loss": 0.9516,
+      "step": 7898
+    },
+    {
+      "epoch": 0.5812094734421707,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.8768056707192748e-05,
+      "loss": 0.7438,
+      "step": 7899
+    },
+    {
+      "epoch": 0.5812830535755346,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.876244810856812e-05,
+      "loss": 0.8965,
+      "step": 7900
+    },
+    {
+      "epoch": 0.5813566337088986,
+      "grad_norm": 0.69921875,
+      "learning_rate": 1.8756839844701126e-05,
+      "loss": 0.9464,
+      "step": 7901
+    },
+    {
+      "epoch": 0.5814302138422626,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.875123191589274e-05,
+      "loss": 0.8169,
+      "step": 7902
+    },
+    {
+      "epoch": 0.5815037939756266,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8745624322443933e-05,
+      "loss": 0.7822,
+      "step": 7903
+    },
+    {
+      "epoch": 0.5815773741089906,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.8740017064655655e-05,
+      "loss": 0.6579,
+      "step": 7904
+    },
+    {
+      "epoch": 0.5816509542423546,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.873441014282884e-05,
+      "loss": 0.6201,
+      "step": 7905
+    },
+    {
+      "epoch": 0.5817245343757186,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.8728803557264403e-05,
+      "loss": 0.8544,
+      "step": 7906
+    },
+    {
+      "epoch": 0.5817981145090826,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.8723197308263227e-05,
+      "loss": 0.6245,
+      "step": 7907
+    },
+    {
+      "epoch": 0.5818716946424466,
+      "grad_norm": 1.0859375,
+      "learning_rate": 1.87175913961262e-05,
+      "loss": 1.4219,
+      "step": 7908
+    },
+    {
+      "epoch": 0.5819452747758105,
+      "grad_norm": 0.85546875,
+      "learning_rate": 1.8711985821154172e-05,
+      "loss": 0.8163,
+      "step": 7909
+    },
+    {
+      "epoch": 0.5820188549091745,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.8706380583647998e-05,
+      "loss": 0.9422,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5820924350425385,
+      "grad_norm": 0.94921875,
+      "learning_rate": 1.8700775683908483e-05,
+      "loss": 0.9302,
+      "step": 7911
+    },
+    {
+      "epoch": 0.5821660151759025,
+      "grad_norm": 0.9375,
+      "learning_rate": 1.8695171122236444e-05,
+      "loss": 0.9279,
+      "step": 7912
+    },
+    {
+      "epoch": 0.5822395953092665,
+      "grad_norm": 0.99609375,
+      "learning_rate": 1.868956689893266e-05,
+      "loss": 1.2156,
+      "step": 7913
+    },
+    {
+      "epoch": 0.5823131754426305,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.86839630142979e-05,
+      "loss": 1.0219,
+      "step": 7914
+    },
+    {
+      "epoch": 0.5823867555759945,
+      "grad_norm": 0.8984375,
+      "learning_rate": 1.8678359468632926e-05,
+      "loss": 0.8994,
+      "step": 7915
+    },
+    {
+      "epoch": 0.5824603357093585,
+      "grad_norm": 0.9453125,
+      "learning_rate": 1.8672756262238454e-05,
+      "loss": 1.2641,
+      "step": 7916
+    },
+    {
+      "epoch": 0.5825339158427225,
+      "grad_norm": 0.88671875,
+      "learning_rate": 1.8667153395415198e-05,
+      "loss": 0.9052,
+      "step": 7917
+    },
+    {
+      "epoch": 0.5826074959760864,
+      "grad_norm": 0.875,
+      "learning_rate": 1.866155086846386e-05,
+      "loss": 0.8305,
+      "step": 7918
+    },
+    {
+      "epoch": 0.5826810761094504,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.8655948681685123e-05,
+      "loss": 0.7444,
+      "step": 7919
+    },
+    {
+      "epoch": 0.5827546562428144,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.865034683537963e-05,
+      "loss": 0.7477,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5828282363761784,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.8644745329848027e-05,
+      "loss": 0.6366,
+      "step": 7921
+    },
+    {
+      "epoch": 0.5829018165095424,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.8639144165390945e-05,
+      "loss": 0.8134,
+      "step": 7922
+    },
+    {
+      "epoch": 0.5829753966429064,
+      "grad_norm": 1.0546875,
+      "learning_rate": 1.863354334230898e-05,
+      "loss": 1.0882,
+      "step": 7923
+    },
+    {
+      "epoch": 0.5830489767762704,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.862794286090272e-05,
+      "loss": 0.7208,
+      "step": 7924
+    },
+    {
+      "epoch": 0.5831225569096344,
+      "grad_norm": 0.76953125,
+      "learning_rate": 1.8622342721472728e-05,
+      "loss": 0.5987,
+      "step": 7925
+    },
+    {
+      "epoch": 0.5831961370429984,
+      "grad_norm": 1.109375,
+      "learning_rate": 1.861674292431956e-05,
+      "loss": 1.1597,
+      "step": 7926
+    },
+    {
+      "epoch": 0.5832697171763623,
+      "grad_norm": 0.625,
+      "learning_rate": 1.861114346974374e-05,
+      "loss": 0.7863,
+      "step": 7927
+    },
+    {
+      "epoch": 0.5833432973097263,
+      "grad_norm": 1.03125,
+      "learning_rate": 1.8605544358045794e-05,
+      "loss": 0.9326,
+      "step": 7928
+    },
+    {
+      "epoch": 0.5834168774430903,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.8599945589526198e-05,
+      "loss": 0.6481,
+      "step": 7929
+    },
+    {
+      "epoch": 0.5834904575764543,
+      "grad_norm": 0.75,
+      "learning_rate": 1.8594347164485427e-05,
+      "loss": 0.803,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5835640377098184,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.858874908322395e-05,
+      "loss": 0.806,
+      "step": 7931
+    },
+    {
+      "epoch": 0.5836376178431824,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.8583151346042203e-05,
+      "loss": 0.9222,
+      "step": 7932
+    },
+    {
+      "epoch": 0.5837111979765464,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.8577553953240604e-05,
+      "loss": 0.9362,
+      "step": 7933
+    },
+    {
+      "epoch": 0.5837847781099104,
+      "grad_norm": 0.87109375,
+      "learning_rate": 1.857195690511955e-05,
+      "loss": 0.8713,
+      "step": 7934
+    },
+    {
+      "epoch": 0.5838583582432744,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.8566360201979427e-05,
+      "loss": 0.8874,
+      "step": 7935
+    },
+    {
+      "epoch": 0.5839319383766383,
+      "grad_norm": 1.015625,
+      "learning_rate": 1.8560763844120603e-05,
+      "loss": 0.9917,
+      "step": 7936
+    },
+    {
+      "epoch": 0.5840055185100023,
+      "grad_norm": 1.0234375,
+      "learning_rate": 1.8555167831843422e-05,
+      "loss": 1.3873,
+      "step": 7937
+    },
+    {
+      "epoch": 0.5840790986433663,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.8549572165448214e-05,
+      "loss": 0.8306,
+      "step": 7938
+    },
+    {
+      "epoch": 0.5841526787767303,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.8543976845235277e-05,
+      "loss": 1.0889,
+      "step": 7939
+    },
+    {
+      "epoch": 0.5842262589100943,
+      "grad_norm": 0.80078125,
+      "learning_rate": 1.8538381871504915e-05,
+      "loss": 0.8001,
+      "step": 7940
+    },
+    {
+      "epoch": 0.5842998390434583,
+      "grad_norm": 1.3515625,
+      "learning_rate": 1.8532787244557393e-05,
+      "loss": 1.2419,
+      "step": 7941
+    },
+    {
+      "epoch": 0.5843734191768223,
+      "grad_norm": 0.72265625,
+      "learning_rate": 1.852719296469297e-05,
+      "loss": 0.8252,
+      "step": 7942
+    },
+    {
+      "epoch": 0.5844469993101863,
+      "grad_norm": 0.765625,
+      "learning_rate": 1.8521599032211866e-05,
+      "loss": 0.9006,
+      "step": 7943
+    },
+    {
+      "epoch": 0.5845205794435503,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.851600544741431e-05,
+      "loss": 1.0111,
+      "step": 7944
+    },
+    {
+      "epoch": 0.5845941595769142,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.8510412210600493e-05,
+      "loss": 1.1289,
+      "step": 7945
+    },
+    {
+      "epoch": 0.5846677397102782,
+      "grad_norm": 0.8828125,
+      "learning_rate": 1.8504819322070595e-05,
+      "loss": 0.6984,
+      "step": 7946
+    },
+    {
+      "epoch": 0.5847413198436422,
+      "grad_norm": 0.7265625,
+      "learning_rate": 1.849922678212478e-05,
+      "loss": 0.8641,
+      "step": 7947
+    },
+    {
+      "epoch": 0.5848148999770062,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.8493634591063187e-05,
+      "loss": 0.5647,
+      "step": 7948
+    },
+    {
+      "epoch": 0.5848884801103702,
+      "grad_norm": 1.0,
+      "learning_rate": 1.848804274918593e-05,
+      "loss": 1.2141,
+      "step": 7949
+    },
+    {
+      "epoch": 0.5849620602437342,
+      "grad_norm": 0.86328125,
+      "learning_rate": 1.848245125679312e-05,
+      "loss": 0.7978,
+      "step": 7950
+    },
+    {
+      "epoch": 0.5850356403770982,
+      "grad_norm": 1.0078125,
+      "learning_rate": 1.8476860114184845e-05,
+      "loss": 0.869,
+      "step": 7951
+    },
+    {
+      "epoch": 0.5851092205104622,
+      "grad_norm": 0.71484375,
+      "learning_rate": 1.8471269321661167e-05,
+      "loss": 0.775,
+      "step": 7952
+    },
+    {
+      "epoch": 0.5851828006438262,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.846567887952213e-05,
+      "loss": 0.596,
+      "step": 7953
+    },
+    {
+      "epoch": 0.5852563807771901,
+      "grad_norm": 1.09375,
+      "learning_rate": 1.846008878806777e-05,
+      "loss": 1.1225,
+      "step": 7954
+    },
+    {
+      "epoch": 0.5853299609105541,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.845449904759809e-05,
+      "loss": 1.0799,
+      "step": 7955
+    },
+    {
+      "epoch": 0.5854035410439181,
+      "grad_norm": 0.8125,
+      "learning_rate": 1.8448909658413093e-05,
+      "loss": 1.0973,
+      "step": 7956
+    },
+    {
+      "epoch": 0.5854771211772821,
+      "grad_norm": 0.91796875,
+      "learning_rate": 1.844332062081273e-05,
+      "loss": 1.0958,
+      "step": 7957
+    },
+    {
+      "epoch": 0.5855507013106461,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.8437731935096967e-05,
+      "loss": 0.9853,
+      "step": 7958
+    },
+    {
+      "epoch": 0.5856242814440101,
+      "grad_norm": 0.984375,
+      "learning_rate": 1.8432143601565737e-05,
+      "loss": 0.988,
+      "step": 7959
+    },
+    {
+      "epoch": 0.5856978615773741,
+      "grad_norm": 0.7578125,
+      "learning_rate": 1.842655562051896e-05,
+      "loss": 0.6313,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5857714417107381,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.842096799225652e-05,
+      "loss": 0.8302,
+      "step": 7961
+    },
+    {
+      "epoch": 0.5858450218441021,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.8415380717078305e-05,
+      "loss": 0.7189,
+      "step": 7962
+    },
+    {
+      "epoch": 0.585918601977466,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.840979379528417e-05,
+      "loss": 0.6009,
+      "step": 7963
+    },
+    {
+      "epoch": 0.58599218211083,
+      "grad_norm": 0.921875,
+      "learning_rate": 1.8404207227173953e-05,
+      "loss": 0.8535,
+      "step": 7964
+    },
+    {
+      "epoch": 0.586065762244194,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.8398621013047483e-05,
+      "loss": 1.0447,
+      "step": 7965
+    },
+    {
+      "epoch": 0.586139342377558,
+      "grad_norm": 0.80859375,
+      "learning_rate": 1.8393035153204547e-05,
+      "loss": 1.0278,
+      "step": 7966
+    },
+    {
+      "epoch": 0.586212922510922,
+      "grad_norm": 0.90234375,
+      "learning_rate": 1.8387449647944938e-05,
+      "loss": 0.924,
+      "step": 7967
+    },
+    {
+      "epoch": 0.5862865026442861,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.838186449756842e-05,
+      "loss": 0.7968,
+      "step": 7968
+    },
+    {
+      "epoch": 0.5863600827776501,
+      "grad_norm": 0.87890625,
+      "learning_rate": 1.837627970237474e-05,
+      "loss": 1.2992,
+      "step": 7969
+    },
+    {
+      "epoch": 0.5864336629110141,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.837069526266361e-05,
+      "loss": 0.9364,
+      "step": 7970
+    },
+    {
+      "epoch": 0.5865072430443781,
+      "grad_norm": 0.91015625,
+      "learning_rate": 1.8365111178734745e-05,
+      "loss": 0.7273,
+      "step": 7971
+    },
+    {
+      "epoch": 0.586580823177742,
+      "grad_norm": 0.890625,
+      "learning_rate": 1.8359527450887828e-05,
+      "loss": 0.7213,
+      "step": 7972
+    },
+    {
+      "epoch": 0.586654403311106,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8353944079422533e-05,
+      "loss": 0.725,
+      "step": 7973
+    },
+    {
+      "epoch": 0.58672798344447,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.8348361064638513e-05,
+      "loss": 1.2185,
+      "step": 7974
+    },
+    {
+      "epoch": 0.586801563577834,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.8342778406835383e-05,
+      "loss": 0.766,
+      "step": 7975
+    },
+    {
+      "epoch": 0.586875143711198,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.8337196106312766e-05,
+      "loss": 1.5889,
+      "step": 7976
+    },
+    {
+      "epoch": 0.586948723844562,
+      "grad_norm": 0.9296875,
+      "learning_rate": 1.8331614163370247e-05,
+      "loss": 0.6474,
+      "step": 7977
+    },
+    {
+      "epoch": 0.587022303977926,
+      "grad_norm": 0.7734375,
+      "learning_rate": 1.83260325783074e-05,
+      "loss": 0.7061,
+      "step": 7978
+    },
+    {
+      "epoch": 0.58709588411129,
+      "grad_norm": 0.73828125,
+      "learning_rate": 1.832045135142379e-05,
+      "loss": 0.9112,
+      "step": 7979
+    },
+    {
+      "epoch": 0.587169464244654,
+      "grad_norm": 0.96875,
+      "learning_rate": 1.831487048301893e-05,
+      "loss": 0.6902,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5872430443780179,
+      "grad_norm": 0.9765625,
+      "learning_rate": 1.8309289973392347e-05,
+      "loss": 0.797,
+      "step": 7981
+    },
+    {
+      "epoch": 0.5873166245113819,
+      "grad_norm": 0.875,
+      "learning_rate": 1.8303709822843533e-05,
+      "loss": 0.8609,
+      "step": 7982
+    },
+    {
+      "epoch": 0.5873902046447459,
+      "grad_norm": 0.83984375,
+      "learning_rate": 1.8298130031671974e-05,
+      "loss": 1.0372,
+      "step": 7983
+    },
+    {
+      "epoch": 0.5874637847781099,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.8292550600177112e-05,
+      "loss": 1.3363,
+      "step": 7984
+    },
+    {
+      "epoch": 0.5875373649114739,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.8286971528658386e-05,
+      "loss": 0.8451,
+      "step": 7985
+    },
+    {
+      "epoch": 0.5876109450448379,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.8281392817415223e-05,
+      "loss": 1.6296,
+      "step": 7986
+    },
+    {
+      "epoch": 0.5876845251782019,
+      "grad_norm": 1.3046875,
+      "learning_rate": 1.827581446674701e-05,
+      "loss": 1.386,
+      "step": 7987
+    },
+    {
+      "epoch": 0.5877581053115659,
+      "grad_norm": 1.1796875,
+      "learning_rate": 1.827023647695315e-05,
+      "loss": 1.2393,
+      "step": 7988
+    },
+    {
+      "epoch": 0.5878316854449299,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.8264658848332977e-05,
+      "loss": 1.1769,
+      "step": 7989
+    },
+    {
+      "epoch": 0.5879052655782938,
+      "grad_norm": 0.82421875,
+      "learning_rate": 1.8259081581185843e-05,
+      "loss": 0.8126,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5879788457116578,
+      "grad_norm": 0.9921875,
+      "learning_rate": 1.8253504675811073e-05,
+      "loss": 0.9025,
+      "step": 7991
+    },
+    {
+      "epoch": 0.5880524258450218,
+      "grad_norm": 1.2109375,
+      "learning_rate": 1.8247928132507962e-05,
+      "loss": 1.0028,
+      "step": 7992
+    },
+    {
+      "epoch": 0.5881260059783858,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.82423519515758e-05,
+      "loss": 0.5055,
+      "step": 7993
+    },
+    {
+      "epoch": 0.5881995861117498,
+      "grad_norm": 0.90625,
+      "learning_rate": 1.8236776133313837e-05,
+      "loss": 0.7578,
+      "step": 7994
+    },
+    {
+      "epoch": 0.5882731662451138,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.8231200678021325e-05,
+      "loss": 0.7794,
+      "step": 7995
+    },
+    {
+      "epoch": 0.5883467463784778,
+      "grad_norm": 0.9140625,
+      "learning_rate": 1.8225625585997494e-05,
+      "loss": 1.2828,
+      "step": 7996
+    },
+    {
+      "epoch": 0.5884203265118418,
+      "grad_norm": 0.77734375,
+      "learning_rate": 1.8220050857541548e-05,
+      "loss": 0.993,
+      "step": 7997
+    },
+    {
+      "epoch": 0.5884939066452058,
+      "grad_norm": 0.98046875,
+      "learning_rate": 1.8214476492952658e-05,
+      "loss": 0.8535,
+      "step": 7998
+    },
+    {
+      "epoch": 0.5885674867785697,
+      "grad_norm": 0.8515625,
+      "learning_rate": 1.820890249253e-05,
+      "loss": 0.9465,
+      "step": 7999
+    },
+    {
+      "epoch": 0.5886410669119337,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.8203328856572716e-05,
+      "loss": 0.6409,
+      "step": 8000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 13591,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.0646749664124574e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}