diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,17514 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 2496,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.001201923076923077,
+      "grad_norm": 52.36468756203271,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 11.2231,
+      "step": 1
+    },
+    {
+      "epoch": 0.002403846153846154,
+      "grad_norm": 53.590174010245626,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 11.2492,
+      "step": 2
+    },
+    {
+      "epoch": 0.003605769230769231,
+      "grad_norm": 51.25869942793319,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 11.3683,
+      "step": 3
+    },
+    {
+      "epoch": 0.004807692307692308,
+      "grad_norm": 50.65458069751499,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 11.2018,
+      "step": 4
+    },
+    {
+      "epoch": 0.006009615384615385,
+      "grad_norm": 51.34402858179148,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 11.2402,
+      "step": 5
+    },
+    {
+      "epoch": 0.007211538461538462,
+      "grad_norm": 50.88547342117953,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 11.2026,
+      "step": 6
+    },
+    {
+      "epoch": 0.008413461538461538,
+      "grad_norm": 53.156180793942994,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 11.1428,
+      "step": 7
+    },
+    {
+      "epoch": 0.009615384615384616,
+      "grad_norm": 59.716031164436714,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 10.8538,
+      "step": 8
+    },
+    {
+      "epoch": 0.010817307692307692,
+      "grad_norm": 58.28117284763272,
+      "learning_rate": 1.8e-06,
+      "loss": 10.8037,
+      "step": 9
+    },
+    {
+      "epoch": 0.01201923076923077,
+      "grad_norm": 59.17099435945327,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 10.7573,
+      "step": 10
+    },
+    {
+      "epoch": 0.013221153846153846,
+      "grad_norm": 79.10334265983677,
+      "learning_rate": 2.2e-06,
+      "loss": 9.804,
+      "step": 11
+    },
+    {
+      "epoch": 0.014423076923076924,
+      "grad_norm": 80.60366209324407,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 9.4999,
+      "step": 12
+    },
+    {
+      "epoch": 0.015625,
+      "grad_norm": 91.66656257344529,
+      "learning_rate": 2.6e-06,
+      "loss": 9.2711,
+      "step": 13
+    },
+    {
+      "epoch": 0.016826923076923076,
+      "grad_norm": 80.66939622322255,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 4.9548,
+      "step": 14
+    },
+    {
+      "epoch": 0.018028846153846152,
+      "grad_norm": 63.33331025718187,
+      "learning_rate": 3e-06,
+      "loss": 3.9141,
+      "step": 15
+    },
+    {
+      "epoch": 0.019230769230769232,
+      "grad_norm": 58.59729252605625,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 3.7729,
+      "step": 16
+    },
+    {
+      "epoch": 0.020432692307692308,
+      "grad_norm": 41.42893527970717,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 3.0216,
+      "step": 17
+    },
+    {
+      "epoch": 0.021634615384615384,
+      "grad_norm": 32.75816252828612,
+      "learning_rate": 3.6e-06,
+      "loss": 2.7047,
+      "step": 18
+    },
+    {
+      "epoch": 0.02283653846153846,
+      "grad_norm": 12.383466278091234,
+      "learning_rate": 3.8e-06,
+      "loss": 1.798,
+      "step": 19
+    },
+    {
+      "epoch": 0.02403846153846154,
+      "grad_norm": 6.358882752031173,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.6083,
+      "step": 20
+    },
+    {
+      "epoch": 0.025240384615384616,
+      "grad_norm": 5.140029391921371,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.4684,
+      "step": 21
+    },
+    {
+      "epoch": 0.026442307692307692,
+      "grad_norm": 4.344569779624694,
+      "learning_rate": 4.4e-06,
+      "loss": 1.5334,
+      "step": 22
+    },
+    {
+      "epoch": 0.027644230769230768,
+      "grad_norm": 3.638170656052622,
+      "learning_rate": 4.6e-06,
+      "loss": 1.3926,
+      "step": 23
+    },
+    {
+      "epoch": 0.028846153846153848,
+      "grad_norm": 2.6571730084227463,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.364,
+      "step": 24
+    },
+    {
+      "epoch": 0.030048076923076924,
+      "grad_norm": 2.201458566586615,
+      "learning_rate": 5e-06,
+      "loss": 1.2675,
+      "step": 25
+    },
+    {
+      "epoch": 0.03125,
+      "grad_norm": 1.621829098307958,
+      "learning_rate": 5.2e-06,
+      "loss": 1.2041,
+      "step": 26
+    },
+    {
+      "epoch": 0.03245192307692308,
+      "grad_norm": 11.823405872978002,
+      "learning_rate": 5.4e-06,
+      "loss": 1.1429,
+      "step": 27
+    },
+    {
+      "epoch": 0.03365384615384615,
+      "grad_norm": 18.95129908053936,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.1561,
+      "step": 28
+    },
+    {
+      "epoch": 0.03485576923076923,
+      "grad_norm": 2.0541540917815544,
+      "learning_rate": 5.8e-06,
+      "loss": 1.1321,
+      "step": 29
+    },
+    {
+      "epoch": 0.036057692307692304,
+      "grad_norm": 1.315004624851504,
+      "learning_rate": 6e-06,
+      "loss": 1.0904,
+      "step": 30
+    },
+    {
+      "epoch": 0.037259615384615384,
+      "grad_norm": 1.081030900225818,
+      "learning_rate": 6.2e-06,
+      "loss": 1.0327,
+      "step": 31
+    },
+    {
+      "epoch": 0.038461538461538464,
+      "grad_norm": 0.9938258474916815,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.0726,
+      "step": 32
+    },
+    {
+      "epoch": 0.039663461538461536,
+      "grad_norm": 0.9351030105526601,
+      "learning_rate": 6.6e-06,
+      "loss": 1.0424,
+      "step": 33
+    },
+    {
+      "epoch": 0.040865384615384616,
+      "grad_norm": 0.84583365518549,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.9714,
+      "step": 34
+    },
+    {
+      "epoch": 0.042067307692307696,
+      "grad_norm": 0.765906549831305,
+      "learning_rate": 7.000000000000001e-06,
+      "loss": 0.9538,
+      "step": 35
+    },
+    {
+      "epoch": 0.04326923076923077,
+      "grad_norm": 0.8498194142936001,
+      "learning_rate": 7.2e-06,
+      "loss": 0.9542,
+      "step": 36
+    },
+    {
+      "epoch": 0.04447115384615385,
+      "grad_norm": 0.7402954369945861,
+      "learning_rate": 7.4e-06,
+      "loss": 0.9666,
+      "step": 37
+    },
+    {
+      "epoch": 0.04567307692307692,
+      "grad_norm": 0.579803449943185,
+      "learning_rate": 7.6e-06,
+      "loss": 0.8736,
+      "step": 38
+    },
+    {
+      "epoch": 0.046875,
+      "grad_norm": 0.6792605098545642,
+      "learning_rate": 7.8e-06,
+      "loss": 0.9027,
+      "step": 39
+    },
+    {
+      "epoch": 0.04807692307692308,
+      "grad_norm": 0.7239360653350769,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.8908,
+      "step": 40
+    },
+    {
+      "epoch": 0.04927884615384615,
+      "grad_norm": 0.6878385887756712,
+      "learning_rate": 8.200000000000001e-06,
+      "loss": 0.8648,
+      "step": 41
+    },
+    {
+      "epoch": 0.05048076923076923,
+      "grad_norm": 0.5387224535232076,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.887,
+      "step": 42
+    },
+    {
+      "epoch": 0.051682692307692304,
+      "grad_norm": 0.5704682977712934,
+      "learning_rate": 8.599999999999999e-06,
+      "loss": 0.8504,
+      "step": 43
+    },
+    {
+      "epoch": 0.052884615384615384,
+      "grad_norm": 0.5287315776847448,
+      "learning_rate": 8.8e-06,
+      "loss": 0.8846,
+      "step": 44
+    },
+    {
+      "epoch": 0.054086538461538464,
+      "grad_norm": 0.5177845551134469,
+      "learning_rate": 9e-06,
+      "loss": 0.8835,
+      "step": 45
+    },
+    {
+      "epoch": 0.055288461538461536,
+      "grad_norm": 0.5145966885995084,
+      "learning_rate": 9.2e-06,
+      "loss": 0.8449,
+      "step": 46
+    },
+    {
+      "epoch": 0.056490384615384616,
+      "grad_norm": 0.5003317319013577,
+      "learning_rate": 9.4e-06,
+      "loss": 0.8205,
+      "step": 47
+    },
+    {
+      "epoch": 0.057692307692307696,
+      "grad_norm": 0.4671544418943671,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.8141,
+      "step": 48
+    },
+    {
+      "epoch": 0.05889423076923077,
+      "grad_norm": 0.406969877976986,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.788,
+      "step": 49
+    },
+    {
+      "epoch": 0.06009615384615385,
+      "grad_norm": 0.42765420323556286,
+      "learning_rate": 1e-05,
+      "loss": 0.8406,
+      "step": 50
+    },
+    {
+      "epoch": 0.06129807692307692,
+      "grad_norm": 0.45256912801368543,
+      "learning_rate": 1.02e-05,
+      "loss": 0.84,
+      "step": 51
+    },
+    {
+      "epoch": 0.0625,
+      "grad_norm": 0.46182003355102735,
+      "learning_rate": 1.04e-05,
+      "loss": 0.844,
+      "step": 52
+    },
+    {
+      "epoch": 0.06370192307692307,
+      "grad_norm": 0.3782657402634094,
+      "learning_rate": 1.06e-05,
+      "loss": 0.7578,
+      "step": 53
+    },
+    {
+      "epoch": 0.06490384615384616,
+      "grad_norm": 0.36310988662243066,
+      "learning_rate": 1.08e-05,
+      "loss": 0.8406,
+      "step": 54
+    },
+    {
+      "epoch": 0.06610576923076923,
+      "grad_norm": 0.393045306657235,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.7844,
+      "step": 55
+    },
+    {
+      "epoch": 0.0673076923076923,
+      "grad_norm": 0.397033066009843,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 0.754,
+      "step": 56
+    },
+    {
+      "epoch": 0.06850961538461539,
+      "grad_norm": 0.3764304083714938,
+      "learning_rate": 1.1400000000000001e-05,
+      "loss": 0.7859,
+      "step": 57
+    },
+    {
+      "epoch": 0.06971153846153846,
+      "grad_norm": 0.3366556611015854,
+      "learning_rate": 1.16e-05,
+      "loss": 0.7588,
+      "step": 58
+    },
+    {
+      "epoch": 0.07091346153846154,
+      "grad_norm": 0.3680329493268861,
+      "learning_rate": 1.18e-05,
+      "loss": 0.7886,
+      "step": 59
+    },
+    {
+      "epoch": 0.07211538461538461,
+      "grad_norm": 0.3637694909698947,
+      "learning_rate": 1.2e-05,
+      "loss": 0.7621,
+      "step": 60
+    },
+    {
+      "epoch": 0.0733173076923077,
+      "grad_norm": 0.30601830648688,
+      "learning_rate": 1.22e-05,
+      "loss": 0.694,
+      "step": 61
+    },
+    {
+      "epoch": 0.07451923076923077,
+      "grad_norm": 0.30807961381167387,
+      "learning_rate": 1.24e-05,
+      "loss": 0.7126,
+      "step": 62
+    },
+    {
+      "epoch": 0.07572115384615384,
+      "grad_norm": 0.3525421202354727,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 0.8087,
+      "step": 63
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 0.3723165692826317,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 0.7565,
+      "step": 64
+    },
+    {
+      "epoch": 0.078125,
+      "grad_norm": 0.32421628267991015,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.7059,
+      "step": 65
+    },
+    {
+      "epoch": 0.07932692307692307,
+      "grad_norm": 0.2988538495265667,
+      "learning_rate": 1.32e-05,
+      "loss": 0.7024,
+      "step": 66
+    },
+    {
+      "epoch": 0.08052884615384616,
+      "grad_norm": 0.30749589684554157,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 0.7362,
+      "step": 67
+    },
+    {
+      "epoch": 0.08173076923076923,
+      "grad_norm": 0.3082910288176201,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 0.7192,
+      "step": 68
+    },
+    {
+      "epoch": 0.0829326923076923,
+      "grad_norm": 0.2920355067312445,
+      "learning_rate": 1.3800000000000002e-05,
+      "loss": 0.7165,
+      "step": 69
+    },
+    {
+      "epoch": 0.08413461538461539,
+      "grad_norm": 0.3044256519645847,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 0.7655,
+      "step": 70
+    },
+    {
+      "epoch": 0.08533653846153846,
+      "grad_norm": 0.2906848149440174,
+      "learning_rate": 1.42e-05,
+      "loss": 0.7189,
+      "step": 71
+    },
+    {
+      "epoch": 0.08653846153846154,
+      "grad_norm": 0.286243263719988,
+      "learning_rate": 1.44e-05,
+      "loss": 0.7203,
+      "step": 72
+    },
+    {
+      "epoch": 0.08774038461538461,
+      "grad_norm": 0.25904081070641494,
+      "learning_rate": 1.4599999999999999e-05,
+      "loss": 0.6947,
+      "step": 73
+    },
+    {
+      "epoch": 0.0889423076923077,
+      "grad_norm": 0.28086332851466217,
+      "learning_rate": 1.48e-05,
+      "loss": 0.7426,
+      "step": 74
+    },
+    {
+      "epoch": 0.09014423076923077,
+      "grad_norm": 0.28559062950935016,
+      "learning_rate": 1.5e-05,
+      "loss": 0.7195,
+      "step": 75
+    },
+    {
+      "epoch": 0.09134615384615384,
+      "grad_norm": 0.25697432391790126,
+      "learning_rate": 1.52e-05,
+      "loss": 0.6963,
+      "step": 76
+    },
+    {
+      "epoch": 0.09254807692307693,
+      "grad_norm": 0.2766603552414007,
+      "learning_rate": 1.54e-05,
+      "loss": 0.6813,
+      "step": 77
+    },
+    {
+      "epoch": 0.09375,
+      "grad_norm": 0.2776118850058663,
+      "learning_rate": 1.56e-05,
+      "loss": 0.7338,
+      "step": 78
+    },
+    {
+      "epoch": 0.09495192307692307,
+      "grad_norm": 0.253478275392958,
+      "learning_rate": 1.58e-05,
+      "loss": 0.6974,
+      "step": 79
+    },
+    {
+      "epoch": 0.09615384615384616,
+      "grad_norm": 0.2817610591890845,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.6943,
+      "step": 80
+    },
+    {
+      "epoch": 0.09735576923076923,
+      "grad_norm": 0.2824300141567493,
+      "learning_rate": 1.62e-05,
+      "loss": 0.6776,
+      "step": 81
+    },
+    {
+      "epoch": 0.0985576923076923,
+      "grad_norm": 0.29185153482222376,
+      "learning_rate": 1.6400000000000002e-05,
+      "loss": 0.6878,
+      "step": 82
+    },
+    {
+      "epoch": 0.09975961538461539,
+      "grad_norm": 0.2883797972667889,
+      "learning_rate": 1.66e-05,
+      "loss": 0.7224,
+      "step": 83
+    },
+    {
+      "epoch": 0.10096153846153846,
+      "grad_norm": 0.25668626445836606,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 0.6759,
+      "step": 84
+    },
+    {
+      "epoch": 0.10216346153846154,
+      "grad_norm": 0.2698972295305341,
+      "learning_rate": 1.7000000000000003e-05,
+      "loss": 0.6834,
+      "step": 85
+    },
+    {
+      "epoch": 0.10336538461538461,
+      "grad_norm": 0.27500229878959487,
+      "learning_rate": 1.7199999999999998e-05,
+      "loss": 0.7405,
+      "step": 86
+    },
+    {
+      "epoch": 0.1045673076923077,
+      "grad_norm": 0.3017580728427446,
+      "learning_rate": 1.74e-05,
+      "loss": 0.7095,
+      "step": 87
+    },
+    {
+      "epoch": 0.10576923076923077,
+      "grad_norm": 0.27128821493881566,
+      "learning_rate": 1.76e-05,
+      "loss": 0.6598,
+      "step": 88
+    },
+    {
+      "epoch": 0.10697115384615384,
+      "grad_norm": 0.2655857571663556,
+      "learning_rate": 1.78e-05,
+      "loss": 0.673,
+      "step": 89
+    },
+    {
+      "epoch": 0.10817307692307693,
+      "grad_norm": 0.3399254924245528,
+      "learning_rate": 1.8e-05,
+      "loss": 0.7097,
+      "step": 90
+    },
+    {
+      "epoch": 0.109375,
+      "grad_norm": 0.274303620191625,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 0.6639,
+      "step": 91
+    },
+    {
+      "epoch": 0.11057692307692307,
+      "grad_norm": 0.25505551557522693,
+      "learning_rate": 1.84e-05,
+      "loss": 0.6586,
+      "step": 92
+    },
+    {
+      "epoch": 0.11177884615384616,
+      "grad_norm": 0.27236975973438904,
+      "learning_rate": 1.86e-05,
+      "loss": 0.6754,
+      "step": 93
+    },
+    {
+      "epoch": 0.11298076923076923,
+      "grad_norm": 0.2929075580433976,
+      "learning_rate": 1.88e-05,
+      "loss": 0.6504,
+      "step": 94
+    },
+    {
+      "epoch": 0.1141826923076923,
+      "grad_norm": 0.3095089446557706,
+      "learning_rate": 1.9e-05,
+      "loss": 0.6749,
+      "step": 95
+    },
+    {
+      "epoch": 0.11538461538461539,
+      "grad_norm": 0.3168175872359653,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.7331,
+      "step": 96
+    },
+    {
+      "epoch": 0.11658653846153846,
+      "grad_norm": 0.2946714222388024,
+      "learning_rate": 1.94e-05,
+      "loss": 0.6929,
+      "step": 97
+    },
+    {
+      "epoch": 0.11778846153846154,
+      "grad_norm": 0.27207487079870024,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 0.6764,
+      "step": 98
+    },
+    {
+      "epoch": 0.11899038461538461,
+      "grad_norm": 0.2747169757398028,
+      "learning_rate": 1.9800000000000004e-05,
+      "loss": 0.7196,
+      "step": 99
+    },
+    {
+      "epoch": 0.1201923076923077,
+      "grad_norm": 0.3083390327442034,
+      "learning_rate": 2e-05,
+      "loss": 0.6884,
+      "step": 100
+    },
+    {
+      "epoch": 0.12139423076923077,
+      "grad_norm": 0.3383233058430867,
+      "learning_rate": 2.0200000000000003e-05,
+      "loss": 0.7028,
+      "step": 101
+    },
+    {
+      "epoch": 0.12259615384615384,
+      "grad_norm": 0.29368392949335553,
+      "learning_rate": 2.04e-05,
+      "loss": 0.6738,
+      "step": 102
+    },
+    {
+      "epoch": 0.12379807692307693,
+      "grad_norm": 0.36674648922392455,
+      "learning_rate": 2.06e-05,
+      "loss": 0.6333,
+      "step": 103
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 0.27908519802738857,
+      "learning_rate": 2.08e-05,
+      "loss": 0.6609,
+      "step": 104
+    },
+    {
+      "epoch": 0.12620192307692307,
+      "grad_norm": 0.30471863878294875,
+      "learning_rate": 2.1e-05,
+      "loss": 0.629,
+      "step": 105
+    },
+    {
+      "epoch": 0.12740384615384615,
+      "grad_norm": 0.3366610939687902,
+      "learning_rate": 2.12e-05,
+      "loss": 0.6748,
+      "step": 106
+    },
+    {
+      "epoch": 0.12860576923076922,
+      "grad_norm": 0.31160080827702136,
+      "learning_rate": 2.1400000000000002e-05,
+      "loss": 0.7079,
+      "step": 107
+    },
+    {
+      "epoch": 0.12980769230769232,
+      "grad_norm": 0.31702094739936687,
+      "learning_rate": 2.16e-05,
+      "loss": 0.6238,
+      "step": 108
+    },
+    {
+      "epoch": 0.1310096153846154,
+      "grad_norm": 0.3216221758526942,
+      "learning_rate": 2.18e-05,
+      "loss": 0.7107,
+      "step": 109
+    },
+    {
+      "epoch": 0.13221153846153846,
+      "grad_norm": 0.3610438454291373,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 0.6532,
+      "step": 110
+    },
+    {
+      "epoch": 0.13341346153846154,
+      "grad_norm": 0.32165622853937376,
+      "learning_rate": 2.22e-05,
+      "loss": 0.6754,
+      "step": 111
+    },
+    {
+      "epoch": 0.1346153846153846,
+      "grad_norm": 0.3555262212245758,
+      "learning_rate": 2.2400000000000002e-05,
+      "loss": 0.7061,
+      "step": 112
+    },
+    {
+      "epoch": 0.13581730769230768,
+      "grad_norm": 0.34571015677338013,
+      "learning_rate": 2.26e-05,
+      "loss": 0.6421,
+      "step": 113
+    },
+    {
+      "epoch": 0.13701923076923078,
+      "grad_norm": 0.37483504827913616,
+      "learning_rate": 2.2800000000000002e-05,
+      "loss": 0.649,
+      "step": 114
+    },
+    {
+      "epoch": 0.13822115384615385,
+      "grad_norm": 0.2880165576646606,
+      "learning_rate": 2.3000000000000003e-05,
+      "loss": 0.6503,
+      "step": 115
+    },
+    {
+      "epoch": 0.13942307692307693,
+      "grad_norm": 0.35631809643557855,
+      "learning_rate": 2.32e-05,
+      "loss": 0.65,
+      "step": 116
+    },
+    {
+      "epoch": 0.140625,
+      "grad_norm": 0.33792366938040236,
+      "learning_rate": 2.3400000000000003e-05,
+      "loss": 0.6705,
+      "step": 117
+    },
+    {
+      "epoch": 0.14182692307692307,
+      "grad_norm": 0.2898312675693551,
+      "learning_rate": 2.36e-05,
+      "loss": 0.6216,
+      "step": 118
+    },
+    {
+      "epoch": 0.14302884615384615,
+      "grad_norm": 0.3413317556521742,
+      "learning_rate": 2.38e-05,
+      "loss": 0.6564,
+      "step": 119
+    },
+    {
+      "epoch": 0.14423076923076922,
+      "grad_norm": 0.31709397790424876,
+      "learning_rate": 2.4e-05,
+      "loss": 0.6775,
+      "step": 120
+    },
+    {
+      "epoch": 0.14543269230769232,
+      "grad_norm": 0.34543125295154054,
+      "learning_rate": 2.4200000000000002e-05,
+      "loss": 0.6338,
+      "step": 121
+    },
+    {
+      "epoch": 0.1466346153846154,
+      "grad_norm": 0.33774565190864353,
+      "learning_rate": 2.44e-05,
+      "loss": 0.5984,
+      "step": 122
+    },
+    {
+      "epoch": 0.14783653846153846,
+      "grad_norm": 0.40355034744937746,
+      "learning_rate": 2.46e-05,
+      "loss": 0.6335,
+      "step": 123
+    },
+    {
+      "epoch": 0.14903846153846154,
+      "grad_norm": 0.34058962974358786,
+      "learning_rate": 2.48e-05,
+      "loss": 0.6324,
+      "step": 124
+    },
+    {
+      "epoch": 0.1502403846153846,
+      "grad_norm": 0.4142584688772935,
+      "learning_rate": 2.5e-05,
+      "loss": 0.6598,
+      "step": 125
+    },
+    {
+      "epoch": 0.15144230769230768,
+      "grad_norm": 0.31493636407469944,
+      "learning_rate": 2.5200000000000003e-05,
+      "loss": 0.6513,
+      "step": 126
+    },
+    {
+      "epoch": 0.15264423076923078,
+      "grad_norm": 0.3291571169192573,
+      "learning_rate": 2.54e-05,
+      "loss": 0.6522,
+      "step": 127
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.34214662982302946,
+      "learning_rate": 2.5600000000000002e-05,
+      "loss": 0.6247,
+      "step": 128
+    },
+    {
+      "epoch": 0.15504807692307693,
+      "grad_norm": 0.31384317462079325,
+      "learning_rate": 2.58e-05,
+      "loss": 0.6279,
+      "step": 129
+    },
+    {
+      "epoch": 0.15625,
+      "grad_norm": 0.3323987167816166,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 0.628,
+      "step": 130
+    },
+    {
+      "epoch": 0.15745192307692307,
+      "grad_norm": 0.3371568899149115,
+      "learning_rate": 2.6200000000000003e-05,
+      "loss": 0.6664,
+      "step": 131
+    },
+    {
+      "epoch": 0.15865384615384615,
+      "grad_norm": 0.3090195706185463,
+      "learning_rate": 2.64e-05,
+      "loss": 0.6282,
+      "step": 132
+    },
+    {
+      "epoch": 0.15985576923076922,
+      "grad_norm": 0.3383842472514099,
+      "learning_rate": 2.6600000000000003e-05,
+      "loss": 0.67,
+      "step": 133
+    },
+    {
+      "epoch": 0.16105769230769232,
+      "grad_norm": 0.3307785807691344,
+      "learning_rate": 2.6800000000000004e-05,
+      "loss": 0.6556,
+      "step": 134
+    },
+    {
+      "epoch": 0.1622596153846154,
+      "grad_norm": 0.34176980983444116,
+      "learning_rate": 2.7000000000000002e-05,
+      "loss": 0.6428,
+      "step": 135
+    },
+    {
+      "epoch": 0.16346153846153846,
+      "grad_norm": 0.36526768732051157,
+      "learning_rate": 2.7200000000000004e-05,
+      "loss": 0.6639,
+      "step": 136
+    },
+    {
+      "epoch": 0.16466346153846154,
+      "grad_norm": 0.31601991384641115,
+      "learning_rate": 2.7400000000000002e-05,
+      "loss": 0.6448,
+      "step": 137
+    },
+    {
+      "epoch": 0.1658653846153846,
+      "grad_norm": 0.3880457445531805,
+      "learning_rate": 2.7600000000000003e-05,
+      "loss": 0.6597,
+      "step": 138
+    },
+    {
+      "epoch": 0.16706730769230768,
+      "grad_norm": 0.32192086702010203,
+      "learning_rate": 2.7800000000000005e-05,
+      "loss": 0.6473,
+      "step": 139
+    },
+    {
+      "epoch": 0.16826923076923078,
+      "grad_norm": 0.4112353975307156,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 0.6745,
+      "step": 140
+    },
+    {
+      "epoch": 0.16947115384615385,
+      "grad_norm": 0.3675560320807144,
+      "learning_rate": 2.8199999999999998e-05,
+      "loss": 0.6627,
+      "step": 141
+    },
+    {
+      "epoch": 0.17067307692307693,
+      "grad_norm": 0.39606348350135545,
+      "learning_rate": 2.84e-05,
+      "loss": 0.652,
+      "step": 142
+    },
+    {
+      "epoch": 0.171875,
+      "grad_norm": 0.35369801220648794,
+      "learning_rate": 2.86e-05,
+      "loss": 0.6163,
+      "step": 143
+    },
+    {
+      "epoch": 0.17307692307692307,
+      "grad_norm": 0.34209773393120896,
+      "learning_rate": 2.88e-05,
+      "loss": 0.6235,
+      "step": 144
+    },
+    {
+      "epoch": 0.17427884615384615,
+      "grad_norm": 0.37888895522094534,
+      "learning_rate": 2.9e-05,
+      "loss": 0.5981,
+      "step": 145
+    },
+    {
+      "epoch": 0.17548076923076922,
+      "grad_norm": 0.32214975880781815,
+      "learning_rate": 2.9199999999999998e-05,
+      "loss": 0.6422,
+      "step": 146
+    },
+    {
+      "epoch": 0.17668269230769232,
+      "grad_norm": 0.4516931008230314,
+      "learning_rate": 2.94e-05,
+      "loss": 0.635,
+      "step": 147
+    },
+    {
+      "epoch": 0.1778846153846154,
+      "grad_norm": 0.3517447481476473,
+      "learning_rate": 2.96e-05,
+      "loss": 0.6854,
+      "step": 148
+    },
+    {
+      "epoch": 0.17908653846153846,
+      "grad_norm": 0.47485956730660245,
+      "learning_rate": 2.98e-05,
+      "loss": 0.637,
+      "step": 149
+    },
+    {
+      "epoch": 0.18028846153846154,
+      "grad_norm": 0.3021103532618611,
+      "learning_rate": 3e-05,
+      "loss": 0.5714,
+      "step": 150
+    },
+    {
+      "epoch": 0.1814903846153846,
+      "grad_norm": 0.3917400669641918,
+      "learning_rate": 3.02e-05,
+      "loss": 0.6196,
+      "step": 151
+    },
+    {
+      "epoch": 0.18269230769230768,
+      "grad_norm": 0.3066085227718499,
+      "learning_rate": 3.04e-05,
+      "loss": 0.6018,
+      "step": 152
+    },
+    {
+      "epoch": 0.18389423076923078,
+      "grad_norm": 0.35395373711803363,
+      "learning_rate": 3.06e-05,
+      "loss": 0.5963,
+      "step": 153
+    },
+    {
+      "epoch": 0.18509615384615385,
+      "grad_norm": 0.42245892846077576,
+      "learning_rate": 3.08e-05,
+      "loss": 0.6732,
+      "step": 154
+    },
+    {
+      "epoch": 0.18629807692307693,
+      "grad_norm": 0.3790104692214946,
+      "learning_rate": 3.1e-05,
+      "loss": 0.6016,
+      "step": 155
+    },
+    {
+      "epoch": 0.1875,
+      "grad_norm": 0.44025949672513465,
+      "learning_rate": 3.12e-05,
+      "loss": 0.6389,
+      "step": 156
+    },
+    {
+      "epoch": 0.18870192307692307,
+      "grad_norm": 0.49643071548556644,
+      "learning_rate": 3.1400000000000004e-05,
+      "loss": 0.623,
+      "step": 157
+    },
+    {
+      "epoch": 0.18990384615384615,
+      "grad_norm": 0.3950961651328371,
+      "learning_rate": 3.16e-05,
+      "loss": 0.5977,
+      "step": 158
+    },
+    {
+      "epoch": 0.19110576923076922,
+      "grad_norm": 0.42055216247796384,
+      "learning_rate": 3.18e-05,
+      "loss": 0.6135,
+      "step": 159
+    },
+    {
+      "epoch": 0.19230769230769232,
+      "grad_norm": 0.4669620544584079,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 0.6199,
+      "step": 160
+    },
+    {
+      "epoch": 0.1935096153846154,
+      "grad_norm": 0.41760010797309416,
+      "learning_rate": 3.2200000000000003e-05,
+      "loss": 0.5751,
+      "step": 161
+    },
+    {
+      "epoch": 0.19471153846153846,
+      "grad_norm": 0.3631476064800621,
+      "learning_rate": 3.24e-05,
+      "loss": 0.6049,
+      "step": 162
+    },
+    {
+      "epoch": 0.19591346153846154,
+      "grad_norm": 0.4793293681447984,
+      "learning_rate": 3.26e-05,
+      "loss": 0.6398,
+      "step": 163
+    },
+    {
+      "epoch": 0.1971153846153846,
+      "grad_norm": 0.4952890098061175,
+      "learning_rate": 3.2800000000000004e-05,
+      "loss": 0.6241,
+      "step": 164
+    },
+    {
+      "epoch": 0.19831730769230768,
+      "grad_norm": 0.42537048045980447,
+      "learning_rate": 3.3e-05,
+      "loss": 0.632,
+      "step": 165
+    },
+    {
+      "epoch": 0.19951923076923078,
+      "grad_norm": 0.4805113583166397,
+      "learning_rate": 3.32e-05,
+      "loss": 0.6595,
+      "step": 166
+    },
+    {
+      "epoch": 0.20072115384615385,
+      "grad_norm": 0.5091118792433852,
+      "learning_rate": 3.3400000000000005e-05,
+      "loss": 0.6421,
+      "step": 167
+    },
+    {
+      "epoch": 0.20192307692307693,
+      "grad_norm": 0.33918876630013806,
+      "learning_rate": 3.3600000000000004e-05,
+      "loss": 0.6194,
+      "step": 168
+    },
+    {
+      "epoch": 0.203125,
+      "grad_norm": 0.5241976329441247,
+      "learning_rate": 3.38e-05,
+      "loss": 0.587,
+      "step": 169
+    },
+    {
+      "epoch": 0.20432692307692307,
+      "grad_norm": 0.47990580369672126,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 0.6311,
+      "step": 170
+    },
+    {
+      "epoch": 0.20552884615384615,
+      "grad_norm": 0.5218878373179289,
+      "learning_rate": 3.4200000000000005e-05,
+      "loss": 0.6152,
+      "step": 171
+    },
+    {
+      "epoch": 0.20673076923076922,
+      "grad_norm": 0.5450020464558445,
+      "learning_rate": 3.4399999999999996e-05,
+      "loss": 0.6282,
+      "step": 172
+    },
+    {
+      "epoch": 0.20793269230769232,
+      "grad_norm": 0.46056605137555434,
+      "learning_rate": 3.46e-05,
+      "loss": 0.622,
+      "step": 173
+    },
+    {
+      "epoch": 0.2091346153846154,
+      "grad_norm": 0.42896071167517313,
+      "learning_rate": 3.48e-05,
+      "loss": 0.6075,
+      "step": 174
+    },
+    {
+      "epoch": 0.21033653846153846,
+      "grad_norm": 0.5749945001024139,
+      "learning_rate": 3.5e-05,
+      "loss": 0.6297,
+      "step": 175
+    },
+    {
+      "epoch": 0.21153846153846154,
+      "grad_norm": 0.4552613982828779,
+      "learning_rate": 3.52e-05,
+      "loss": 0.5955,
+      "step": 176
+    },
+    {
+      "epoch": 0.2127403846153846,
+      "grad_norm": 0.4675224851354325,
+      "learning_rate": 3.54e-05,
+      "loss": 0.5973,
+      "step": 177
+    },
+    {
+      "epoch": 0.21394230769230768,
+      "grad_norm": 0.5418779038250601,
+      "learning_rate": 3.56e-05,
+      "loss": 0.6532,
+      "step": 178
+    },
+    {
+      "epoch": 0.21514423076923078,
+      "grad_norm": 0.4351289693256165,
+      "learning_rate": 3.58e-05,
+      "loss": 0.6244,
+      "step": 179
+    },
+    {
+      "epoch": 0.21634615384615385,
+      "grad_norm": 0.4601756992799404,
+      "learning_rate": 3.6e-05,
+      "loss": 0.6027,
+      "step": 180
+    },
+    {
+      "epoch": 0.21754807692307693,
+      "grad_norm": 0.4600347110843093,
+      "learning_rate": 3.62e-05,
+      "loss": 0.607,
+      "step": 181
+    },
+    {
+      "epoch": 0.21875,
+      "grad_norm": 0.4186249372381047,
+      "learning_rate": 3.6400000000000004e-05,
+      "loss": 0.5994,
+      "step": 182
+    },
+    {
+      "epoch": 0.21995192307692307,
+      "grad_norm": 0.4148035493337245,
+      "learning_rate": 3.66e-05,
+      "loss": 0.625,
+      "step": 183
+    },
+    {
+      "epoch": 0.22115384615384615,
+      "grad_norm": 0.5268873423132147,
+      "learning_rate": 3.68e-05,
+      "loss": 0.6273,
+      "step": 184
+    },
+    {
+      "epoch": 0.22235576923076922,
+      "grad_norm": 0.4428367628733136,
+      "learning_rate": 3.7e-05,
+      "loss": 0.6062,
+      "step": 185
+    },
+    {
+      "epoch": 0.22355769230769232,
+      "grad_norm": 0.44910428424072446,
+      "learning_rate": 3.72e-05,
+      "loss": 0.5635,
+      "step": 186
+    },
+    {
+      "epoch": 0.2247596153846154,
+      "grad_norm": 0.4416329276637058,
+      "learning_rate": 3.74e-05,
+      "loss": 0.605,
+      "step": 187
+    },
+    {
+      "epoch": 0.22596153846153846,
+      "grad_norm": 0.4589773251871935,
+      "learning_rate": 3.76e-05,
+      "loss": 0.6665,
+      "step": 188
+    },
+    {
+      "epoch": 0.22716346153846154,
+      "grad_norm": 0.44066257347746185,
+      "learning_rate": 3.7800000000000004e-05,
+      "loss": 0.6038,
+      "step": 189
+    },
+    {
+      "epoch": 0.2283653846153846,
+      "grad_norm": 0.42823496505503106,
+      "learning_rate": 3.8e-05,
+      "loss": 0.5908,
+      "step": 190
+    },
+    {
+      "epoch": 0.22956730769230768,
+      "grad_norm": 0.4580954488973302,
+      "learning_rate": 3.82e-05,
+      "loss": 0.6159,
+      "step": 191
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.4574710017326464,
+      "learning_rate": 3.8400000000000005e-05,
+      "loss": 0.6132,
+      "step": 192
+    },
+    {
+      "epoch": 0.23197115384615385,
+      "grad_norm": 0.43440014236043306,
+      "learning_rate": 3.86e-05,
+      "loss": 0.6165,
+      "step": 193
+    },
+    {
+      "epoch": 0.23317307692307693,
+      "grad_norm": 0.39653121352338333,
+      "learning_rate": 3.88e-05,
+      "loss": 0.6073,
+      "step": 194
+    },
+    {
+      "epoch": 0.234375,
+      "grad_norm": 0.3923699787537727,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 0.5966,
+      "step": 195
+    },
+    {
+      "epoch": 0.23557692307692307,
+      "grad_norm": 0.41479453394298105,
+      "learning_rate": 3.9200000000000004e-05,
+      "loss": 0.5671,
+      "step": 196
+    },
+    {
+      "epoch": 0.23677884615384615,
+      "grad_norm": 0.39359385390474927,
+      "learning_rate": 3.94e-05,
+      "loss": 0.5972,
+      "step": 197
+    },
+    {
+      "epoch": 0.23798076923076922,
+      "grad_norm": 0.40863815663566727,
+      "learning_rate": 3.960000000000001e-05,
+      "loss": 0.5835,
+      "step": 198
+    },
+    {
+      "epoch": 0.23918269230769232,
+      "grad_norm": 0.5726634196478491,
+      "learning_rate": 3.9800000000000005e-05,
+      "loss": 0.5701,
+      "step": 199
+    },
+    {
+      "epoch": 0.2403846153846154,
+      "grad_norm": 0.36444213307862267,
+      "learning_rate": 4e-05,
+      "loss": 0.5944,
+      "step": 200
+    },
+    {
+      "epoch": 0.24158653846153846,
+      "grad_norm": 0.49269780319884193,
+      "learning_rate": 4.02e-05,
+      "loss": 0.6319,
+      "step": 201
+    },
+    {
+      "epoch": 0.24278846153846154,
+      "grad_norm": 0.4744659127585973,
+      "learning_rate": 4.0400000000000006e-05,
+      "loss": 0.6518,
+      "step": 202
+    },
+    {
+      "epoch": 0.2439903846153846,
+      "grad_norm": 0.3799679873543414,
+      "learning_rate": 4.0600000000000004e-05,
+      "loss": 0.5995,
+      "step": 203
+    },
+    {
+      "epoch": 0.24519230769230768,
+      "grad_norm": 0.41607034212238364,
+      "learning_rate": 4.08e-05,
+      "loss": 0.5722,
+      "step": 204
+    },
+    {
+      "epoch": 0.24639423076923078,
+      "grad_norm": 0.4869148099640537,
+      "learning_rate": 4.1e-05,
+      "loss": 0.5858,
+      "step": 205
+    },
+    {
+      "epoch": 0.24759615384615385,
+      "grad_norm": 0.37133239984853006,
+      "learning_rate": 4.12e-05,
+      "loss": 0.607,
+      "step": 206
+    },
+    {
+      "epoch": 0.24879807692307693,
+      "grad_norm": 0.4844803913812173,
+      "learning_rate": 4.14e-05,
+      "loss": 0.5918,
+      "step": 207
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5364236786150066,
+      "learning_rate": 4.16e-05,
+      "loss": 0.584,
+      "step": 208
+    },
+    {
+      "epoch": 0.2512019230769231,
+      "grad_norm": 0.4505694273725652,
+      "learning_rate": 4.18e-05,
+      "loss": 0.6037,
+      "step": 209
+    },
+    {
+      "epoch": 0.25240384615384615,
+      "grad_norm": 0.39650825572996884,
+      "learning_rate": 4.2e-05,
+      "loss": 0.6107,
+      "step": 210
+    },
+    {
+      "epoch": 0.2536057692307692,
+      "grad_norm": 0.42971528570217493,
+      "learning_rate": 4.22e-05,
+      "loss": 0.5899,
+      "step": 211
+    },
+    {
+      "epoch": 0.2548076923076923,
+      "grad_norm": 0.4637706620078796,
+      "learning_rate": 4.24e-05,
+      "loss": 0.5888,
+      "step": 212
+    },
+    {
+      "epoch": 0.25600961538461536,
+      "grad_norm": 0.36617246873434806,
+      "learning_rate": 4.26e-05,
+      "loss": 0.6253,
+      "step": 213
+    },
+    {
+      "epoch": 0.25721153846153844,
+      "grad_norm": 0.4754471880434809,
+      "learning_rate": 4.2800000000000004e-05,
+      "loss": 0.6282,
+      "step": 214
+    },
+    {
+      "epoch": 0.25841346153846156,
+      "grad_norm": 0.57471092853946,
+      "learning_rate": 4.3e-05,
+      "loss": 0.6145,
+      "step": 215
+    },
+    {
+      "epoch": 0.25961538461538464,
+      "grad_norm": 0.40663450185983585,
+      "learning_rate": 4.32e-05,
+      "loss": 0.6002,
+      "step": 216
+    },
+    {
+      "epoch": 0.2608173076923077,
+      "grad_norm": 0.5037233099577721,
+      "learning_rate": 4.3400000000000005e-05,
+      "loss": 0.6187,
+      "step": 217
+    },
+    {
+      "epoch": 0.2620192307692308,
+      "grad_norm": 0.4606467272328667,
+      "learning_rate": 4.36e-05,
+      "loss": 0.5529,
+      "step": 218
+    },
+    {
+      "epoch": 0.26322115384615385,
+      "grad_norm": 0.3659759364237195,
+      "learning_rate": 4.38e-05,
+      "loss": 0.6067,
+      "step": 219
+    },
+    {
+      "epoch": 0.2644230769230769,
+      "grad_norm": 0.45942993542353716,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 0.5647,
+      "step": 220
+    },
+    {
+      "epoch": 0.265625,
+      "grad_norm": 0.5114236170384563,
+      "learning_rate": 4.4200000000000004e-05,
+      "loss": 0.6211,
+      "step": 221
+    },
+    {
+      "epoch": 0.2668269230769231,
+      "grad_norm": 0.4374624330421017,
+      "learning_rate": 4.44e-05,
+      "loss": 0.585,
+      "step": 222
+    },
+    {
+      "epoch": 0.26802884615384615,
+      "grad_norm": 0.4320595393198989,
+      "learning_rate": 4.46e-05,
+      "loss": 0.6104,
+      "step": 223
+    },
+    {
+      "epoch": 0.2692307692307692,
+      "grad_norm": 0.48796513656755397,
+      "learning_rate": 4.4800000000000005e-05,
+      "loss": 0.578,
+      "step": 224
+    },
+    {
+      "epoch": 0.2704326923076923,
+      "grad_norm": 0.40402602172588004,
+      "learning_rate": 4.5e-05,
+      "loss": 0.5999,
+      "step": 225
+    },
+    {
+      "epoch": 0.27163461538461536,
+      "grad_norm": 0.5344722956097983,
+      "learning_rate": 4.52e-05,
+      "loss": 0.6271,
+      "step": 226
+    },
+    {
+      "epoch": 0.27283653846153844,
+      "grad_norm": 0.4875801179577109,
+      "learning_rate": 4.5400000000000006e-05,
+      "loss": 0.5704,
+      "step": 227
+    },
+    {
+      "epoch": 0.27403846153846156,
+      "grad_norm": 0.4831259727978741,
+      "learning_rate": 4.5600000000000004e-05,
+      "loss": 0.5869,
+      "step": 228
+    },
+    {
+      "epoch": 0.27524038461538464,
+      "grad_norm": 0.4878005680099931,
+      "learning_rate": 4.58e-05,
+      "loss": 0.5923,
+      "step": 229
+    },
+    {
+      "epoch": 0.2764423076923077,
+      "grad_norm": 0.4377935977759504,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 0.6161,
+      "step": 230
+    },
+    {
+      "epoch": 0.2776442307692308,
+      "grad_norm": 0.4643597252038858,
+      "learning_rate": 4.6200000000000005e-05,
+      "loss": 0.6243,
+      "step": 231
+    },
+    {
+      "epoch": 0.27884615384615385,
+      "grad_norm": 0.4173513565911848,
+      "learning_rate": 4.64e-05,
+      "loss": 0.5896,
+      "step": 232
+    },
+    {
+      "epoch": 0.2800480769230769,
+      "grad_norm": 0.5668289910430891,
+      "learning_rate": 4.660000000000001e-05,
+      "loss": 0.61,
+      "step": 233
+    },
+    {
+      "epoch": 0.28125,
+      "grad_norm": 0.4572700730997457,
+      "learning_rate": 4.6800000000000006e-05,
+      "loss": 0.5767,
+      "step": 234
+    },
+    {
+      "epoch": 0.2824519230769231,
+      "grad_norm": 0.4351930512130122,
+      "learning_rate": 4.7e-05,
+      "loss": 0.5599,
+      "step": 235
+    },
+    {
+      "epoch": 0.28365384615384615,
+      "grad_norm": 0.4618445835253526,
+      "learning_rate": 4.72e-05,
+      "loss": 0.6092,
+      "step": 236
+    },
+    {
+      "epoch": 0.2848557692307692,
+      "grad_norm": 0.38073376242310975,
+      "learning_rate": 4.74e-05,
+      "loss": 0.6338,
+      "step": 237
+    },
+    {
+      "epoch": 0.2860576923076923,
+      "grad_norm": 0.5652961434513976,
+      "learning_rate": 4.76e-05,
+      "loss": 0.5603,
+      "step": 238
+    },
+    {
+      "epoch": 0.28725961538461536,
+      "grad_norm": 0.43115968557184137,
+      "learning_rate": 4.78e-05,
+      "loss": 0.5488,
+      "step": 239
+    },
+    {
+      "epoch": 0.28846153846153844,
+      "grad_norm": 0.4501975677071634,
+      "learning_rate": 4.8e-05,
+      "loss": 0.5827,
+      "step": 240
+    },
+    {
+      "epoch": 0.28966346153846156,
+      "grad_norm": 0.3786540996553594,
+      "learning_rate": 4.82e-05,
+      "loss": 0.6013,
+      "step": 241
+    },
+    {
+      "epoch": 0.29086538461538464,
+      "grad_norm": 0.4266395679642578,
+      "learning_rate": 4.8400000000000004e-05,
+      "loss": 0.5594,
+      "step": 242
+    },
+    {
+      "epoch": 0.2920673076923077,
+      "grad_norm": 0.4003139739124793,
+      "learning_rate": 4.86e-05,
+      "loss": 0.605,
+      "step": 243
+    },
+    {
+      "epoch": 0.2932692307692308,
+      "grad_norm": 0.43401176635858596,
+      "learning_rate": 4.88e-05,
+      "loss": 0.6351,
+      "step": 244
+    },
+    {
+      "epoch": 0.29447115384615385,
+      "grad_norm": 0.39592381920402375,
+      "learning_rate": 4.9e-05,
+      "loss": 0.5962,
+      "step": 245
+    },
+    {
+      "epoch": 0.2956730769230769,
+      "grad_norm": 0.4854347171403406,
+      "learning_rate": 4.92e-05,
+      "loss": 0.6024,
+      "step": 246
+    },
+    {
+      "epoch": 0.296875,
+      "grad_norm": 0.46982829692442607,
+      "learning_rate": 4.94e-05,
+      "loss": 0.5661,
+      "step": 247
+    },
+    {
+      "epoch": 0.2980769230769231,
+      "grad_norm": 0.5920360804393904,
+      "learning_rate": 4.96e-05,
+      "loss": 0.584,
+      "step": 248
+    },
+    {
+      "epoch": 0.29927884615384615,
+      "grad_norm": 0.5380532082908318,
+      "learning_rate": 4.9800000000000004e-05,
+      "loss": 0.5918,
+      "step": 249
+    },
+    {
+      "epoch": 0.3004807692307692,
+      "grad_norm": 0.41092767762603133,
+      "learning_rate": 5e-05,
+      "loss": 0.5701,
+      "step": 250
+    },
+    {
+      "epoch": 0.3016826923076923,
+      "grad_norm": 0.538680138495851,
+      "learning_rate": 4.9977738201246665e-05,
+      "loss": 0.5554,
+      "step": 251
+    },
+    {
+      "epoch": 0.30288461538461536,
+      "grad_norm": 0.6541167472384556,
+      "learning_rate": 4.995547640249332e-05,
+      "loss": 0.6114,
+      "step": 252
+    },
+    {
+      "epoch": 0.30408653846153844,
+      "grad_norm": 0.7633644238470539,
+      "learning_rate": 4.993321460373999e-05,
+      "loss": 0.5931,
+      "step": 253
+    },
+    {
+      "epoch": 0.30528846153846156,
+      "grad_norm": 0.521894995513932,
+      "learning_rate": 4.9910952804986644e-05,
+      "loss": 0.5715,
+      "step": 254
+    },
+    {
+      "epoch": 0.30649038461538464,
+      "grad_norm": 0.5023456759955174,
+      "learning_rate": 4.9888691006233306e-05,
+      "loss": 0.5601,
+      "step": 255
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.6575998217044469,
+      "learning_rate": 4.986642920747996e-05,
+      "loss": 0.5847,
+      "step": 256
+    },
+    {
+      "epoch": 0.3088942307692308,
+      "grad_norm": 0.45629781645672896,
+      "learning_rate": 4.984416740872663e-05,
+      "loss": 0.5836,
+      "step": 257
+    },
+    {
+      "epoch": 0.31009615384615385,
+      "grad_norm": 0.44710825881751226,
+      "learning_rate": 4.982190560997329e-05,
+      "loss": 0.5895,
+      "step": 258
+    },
+    {
+      "epoch": 0.3112980769230769,
+      "grad_norm": 0.46872142692799096,
+      "learning_rate": 4.979964381121995e-05,
+      "loss": 0.5984,
+      "step": 259
+    },
+    {
+      "epoch": 0.3125,
+      "grad_norm": 0.5584578244426207,
+      "learning_rate": 4.977738201246661e-05,
+      "loss": 0.5866,
+      "step": 260
+    },
+    {
+      "epoch": 0.3137019230769231,
+      "grad_norm": 0.5680587188596463,
+      "learning_rate": 4.9755120213713266e-05,
+      "loss": 0.5824,
+      "step": 261
+    },
+    {
+      "epoch": 0.31490384615384615,
+      "grad_norm": 0.4231467132044837,
+      "learning_rate": 4.9732858414959934e-05,
+      "loss": 0.5995,
+      "step": 262
+    },
+    {
+      "epoch": 0.3161057692307692,
+      "grad_norm": 0.6000681393772778,
+      "learning_rate": 4.971059661620659e-05,
+      "loss": 0.5721,
+      "step": 263
+    },
+    {
+      "epoch": 0.3173076923076923,
+      "grad_norm": 0.6205540980123596,
+      "learning_rate": 4.968833481745325e-05,
+      "loss": 0.5663,
+      "step": 264
+    },
+    {
+      "epoch": 0.31850961538461536,
+      "grad_norm": 0.6720364445709425,
+      "learning_rate": 4.9666073018699914e-05,
+      "loss": 0.5755,
+      "step": 265
+    },
+    {
+      "epoch": 0.31971153846153844,
+      "grad_norm": 0.4067976595573186,
+      "learning_rate": 4.9643811219946576e-05,
+      "loss": 0.5788,
+      "step": 266
+    },
+    {
+      "epoch": 0.32091346153846156,
+      "grad_norm": 0.6221001392748503,
+      "learning_rate": 4.962154942119324e-05,
+      "loss": 0.6298,
+      "step": 267
+    },
+    {
+      "epoch": 0.32211538461538464,
+      "grad_norm": 0.5548317463330392,
+      "learning_rate": 4.9599287622439894e-05,
+      "loss": 0.5868,
+      "step": 268
+    },
+    {
+      "epoch": 0.3233173076923077,
+      "grad_norm": 0.46193546596369683,
+      "learning_rate": 4.9577025823686556e-05,
+      "loss": 0.5798,
+      "step": 269
+    },
+    {
+      "epoch": 0.3245192307692308,
+      "grad_norm": 0.6324701767824795,
+      "learning_rate": 4.955476402493322e-05,
+      "loss": 0.5922,
+      "step": 270
+    },
+    {
+      "epoch": 0.32572115384615385,
+      "grad_norm": 0.677671264749846,
+      "learning_rate": 4.953250222617988e-05,
+      "loss": 0.5508,
+      "step": 271
+    },
+    {
+      "epoch": 0.3269230769230769,
+      "grad_norm": 0.43027871362663866,
+      "learning_rate": 4.951024042742654e-05,
+      "loss": 0.5799,
+      "step": 272
+    },
+    {
+      "epoch": 0.328125,
+      "grad_norm": 0.590355176222183,
+      "learning_rate": 4.94879786286732e-05,
+      "loss": 0.6073,
+      "step": 273
+    },
+    {
+      "epoch": 0.3293269230769231,
+      "grad_norm": 0.6580473849030944,
+      "learning_rate": 4.946571682991986e-05,
+      "loss": 0.5752,
+      "step": 274
+    },
+    {
+      "epoch": 0.33052884615384615,
+      "grad_norm": 0.49378098733347614,
+      "learning_rate": 4.944345503116652e-05,
+      "loss": 0.5982,
+      "step": 275
+    },
+    {
+      "epoch": 0.3317307692307692,
+      "grad_norm": 0.5588003846099084,
+      "learning_rate": 4.9421193232413184e-05,
+      "loss": 0.5621,
+      "step": 276
+    },
+    {
+      "epoch": 0.3329326923076923,
+      "grad_norm": 0.5124858197595443,
+      "learning_rate": 4.939893143365984e-05,
+      "loss": 0.582,
+      "step": 277
+    },
+    {
+      "epoch": 0.33413461538461536,
+      "grad_norm": 0.44876128961718775,
+      "learning_rate": 4.93766696349065e-05,
+      "loss": 0.528,
+      "step": 278
+    },
+    {
+      "epoch": 0.33533653846153844,
+      "grad_norm": 0.49682397546609514,
+      "learning_rate": 4.9354407836153164e-05,
+      "loss": 0.6164,
+      "step": 279
+    },
+    {
+      "epoch": 0.33653846153846156,
+      "grad_norm": 0.36072941148428345,
+      "learning_rate": 4.9332146037399826e-05,
+      "loss": 0.5791,
+      "step": 280
+    },
+    {
+      "epoch": 0.33774038461538464,
+      "grad_norm": 0.4105213620899732,
+      "learning_rate": 4.930988423864649e-05,
+      "loss": 0.6064,
+      "step": 281
+    },
+    {
+      "epoch": 0.3389423076923077,
+      "grad_norm": 0.3815471375439723,
+      "learning_rate": 4.928762243989314e-05,
+      "loss": 0.6041,
+      "step": 282
+    },
+    {
+      "epoch": 0.3401442307692308,
+      "grad_norm": 0.4075948523580099,
+      "learning_rate": 4.9265360641139805e-05,
+      "loss": 0.6148,
+      "step": 283
+    },
+    {
+      "epoch": 0.34134615384615385,
+      "grad_norm": 0.429413104582311,
+      "learning_rate": 4.924309884238647e-05,
+      "loss": 0.576,
+      "step": 284
+    },
+    {
+      "epoch": 0.3425480769230769,
+      "grad_norm": 0.41977805937372664,
+      "learning_rate": 4.922083704363313e-05,
+      "loss": 0.5511,
+      "step": 285
+    },
+    {
+      "epoch": 0.34375,
+      "grad_norm": 0.4753609711035092,
+      "learning_rate": 4.919857524487979e-05,
+      "loss": 0.5979,
+      "step": 286
+    },
+    {
+      "epoch": 0.3449519230769231,
+      "grad_norm": 0.4562994429762158,
+      "learning_rate": 4.917631344612645e-05,
+      "loss": 0.5938,
+      "step": 287
+    },
+    {
+      "epoch": 0.34615384615384615,
+      "grad_norm": 0.3981988504484483,
+      "learning_rate": 4.915405164737311e-05,
+      "loss": 0.6019,
+      "step": 288
+    },
+    {
+      "epoch": 0.3473557692307692,
+      "grad_norm": 0.43418585918505903,
+      "learning_rate": 4.913178984861977e-05,
+      "loss": 0.5777,
+      "step": 289
+    },
+    {
+      "epoch": 0.3485576923076923,
+      "grad_norm": 0.4394998354176906,
+      "learning_rate": 4.9109528049866433e-05,
+      "loss": 0.557,
+      "step": 290
+    },
+    {
+      "epoch": 0.34975961538461536,
+      "grad_norm": 0.4088302669528794,
+      "learning_rate": 4.908726625111309e-05,
+      "loss": 0.5833,
+      "step": 291
+    },
+    {
+      "epoch": 0.35096153846153844,
+      "grad_norm": 0.401313897529922,
+      "learning_rate": 4.906500445235975e-05,
+      "loss": 0.5658,
+      "step": 292
+    },
+    {
+      "epoch": 0.35216346153846156,
+      "grad_norm": 0.5113767910426895,
+      "learning_rate": 4.904274265360641e-05,
+      "loss": 0.572,
+      "step": 293
+    },
+    {
+      "epoch": 0.35336538461538464,
+      "grad_norm": 0.36498973365928106,
+      "learning_rate": 4.9020480854853075e-05,
+      "loss": 0.5703,
+      "step": 294
+    },
+    {
+      "epoch": 0.3545673076923077,
+      "grad_norm": 0.49744581242892416,
+      "learning_rate": 4.899821905609974e-05,
+      "loss": 0.5913,
+      "step": 295
+    },
+    {
+      "epoch": 0.3557692307692308,
+      "grad_norm": 0.44891165085738804,
+      "learning_rate": 4.897595725734639e-05,
+      "loss": 0.5999,
+      "step": 296
+    },
+    {
+      "epoch": 0.35697115384615385,
+      "grad_norm": 0.42652217879395266,
+      "learning_rate": 4.8953695458593055e-05,
+      "loss": 0.5804,
+      "step": 297
+    },
+    {
+      "epoch": 0.3581730769230769,
+      "grad_norm": 0.48319683095719074,
+      "learning_rate": 4.893143365983972e-05,
+      "loss": 0.5777,
+      "step": 298
+    },
+    {
+      "epoch": 0.359375,
+      "grad_norm": 0.3963405354358038,
+      "learning_rate": 4.890917186108638e-05,
+      "loss": 0.5614,
+      "step": 299
+    },
+    {
+      "epoch": 0.3605769230769231,
+      "grad_norm": 0.5324737717836817,
+      "learning_rate": 4.888691006233304e-05,
+      "loss": 0.5678,
+      "step": 300
+    },
+    {
+      "epoch": 0.36177884615384615,
+      "grad_norm": 0.385136739724319,
+      "learning_rate": 4.8864648263579697e-05,
+      "loss": 0.5567,
+      "step": 301
+    },
+    {
+      "epoch": 0.3629807692307692,
+      "grad_norm": 0.5202397934044365,
+      "learning_rate": 4.884238646482636e-05,
+      "loss": 0.5832,
+      "step": 302
+    },
+    {
+      "epoch": 0.3641826923076923,
+      "grad_norm": 0.5145234590859966,
+      "learning_rate": 4.882012466607302e-05,
+      "loss": 0.5762,
+      "step": 303
+    },
+    {
+      "epoch": 0.36538461538461536,
+      "grad_norm": 0.4375287404081779,
+      "learning_rate": 4.879786286731968e-05,
+      "loss": 0.5433,
+      "step": 304
+    },
+    {
+      "epoch": 0.36658653846153844,
+      "grad_norm": 0.43093731895364545,
+      "learning_rate": 4.877560106856634e-05,
+      "loss": 0.5381,
+      "step": 305
+    },
+    {
+      "epoch": 0.36778846153846156,
+      "grad_norm": 0.40316344663482906,
+      "learning_rate": 4.8753339269813e-05,
+      "loss": 0.5941,
+      "step": 306
+    },
+    {
+      "epoch": 0.36899038461538464,
+      "grad_norm": 0.39546551366494587,
+      "learning_rate": 4.873107747105967e-05,
+      "loss": 0.5287,
+      "step": 307
+    },
+    {
+      "epoch": 0.3701923076923077,
+      "grad_norm": 0.3721275166067844,
+      "learning_rate": 4.8708815672306325e-05,
+      "loss": 0.5896,
+      "step": 308
+    },
+    {
+      "epoch": 0.3713942307692308,
+      "grad_norm": 0.43673516759602793,
+      "learning_rate": 4.868655387355299e-05,
+      "loss": 0.547,
+      "step": 309
+    },
+    {
+      "epoch": 0.37259615384615385,
+      "grad_norm": 0.3925283179444157,
+      "learning_rate": 4.866429207479964e-05,
+      "loss": 0.5869,
+      "step": 310
+    },
+    {
+      "epoch": 0.3737980769230769,
+      "grad_norm": 0.4408904922516116,
+      "learning_rate": 4.8642030276046304e-05,
+      "loss": 0.5335,
+      "step": 311
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 0.5101541371332177,
+      "learning_rate": 4.8619768477292966e-05,
+      "loss": 0.5648,
+      "step": 312
+    },
+    {
+      "epoch": 0.3762019230769231,
+      "grad_norm": 0.4626241100257864,
+      "learning_rate": 4.859750667853963e-05,
+      "loss": 0.519,
+      "step": 313
+    },
+    {
+      "epoch": 0.37740384615384615,
+      "grad_norm": 0.46658727387191545,
+      "learning_rate": 4.857524487978629e-05,
+      "loss": 0.5605,
+      "step": 314
+    },
+    {
+      "epoch": 0.3786057692307692,
+      "grad_norm": 0.4373997619329127,
+      "learning_rate": 4.8552983081032946e-05,
+      "loss": 0.5691,
+      "step": 315
+    },
+    {
+      "epoch": 0.3798076923076923,
+      "grad_norm": 0.5253628926422157,
+      "learning_rate": 4.8530721282279615e-05,
+      "loss": 0.5443,
+      "step": 316
+    },
+    {
+      "epoch": 0.38100961538461536,
+      "grad_norm": 0.40478035087874104,
+      "learning_rate": 4.850845948352627e-05,
+      "loss": 0.5553,
+      "step": 317
+    },
+    {
+      "epoch": 0.38221153846153844,
+      "grad_norm": 0.5797760672600898,
+      "learning_rate": 4.848619768477293e-05,
+      "loss": 0.5769,
+      "step": 318
+    },
+    {
+      "epoch": 0.38341346153846156,
+      "grad_norm": 0.5226827352240917,
+      "learning_rate": 4.846393588601959e-05,
+      "loss": 0.6277,
+      "step": 319
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.5087510839763183,
+      "learning_rate": 4.844167408726625e-05,
+      "loss": 0.6067,
+      "step": 320
+    },
+    {
+      "epoch": 0.3858173076923077,
+      "grad_norm": 0.5740042586256213,
+      "learning_rate": 4.841941228851292e-05,
+      "loss": 0.5578,
+      "step": 321
+    },
+    {
+      "epoch": 0.3870192307692308,
+      "grad_norm": 0.5539535213379385,
+      "learning_rate": 4.8397150489759574e-05,
+      "loss": 0.582,
+      "step": 322
+    },
+    {
+      "epoch": 0.38822115384615385,
+      "grad_norm": 0.5074290544571007,
+      "learning_rate": 4.8374888691006236e-05,
+      "loss": 0.5806,
+      "step": 323
+    },
+    {
+      "epoch": 0.3894230769230769,
+      "grad_norm": 0.4670478557385501,
+      "learning_rate": 4.835262689225289e-05,
+      "loss": 0.5661,
+      "step": 324
+    },
+    {
+      "epoch": 0.390625,
+      "grad_norm": 0.570513702861374,
+      "learning_rate": 4.833036509349956e-05,
+      "loss": 0.5997,
+      "step": 325
+    },
+    {
+      "epoch": 0.3918269230769231,
+      "grad_norm": 0.4988170512047621,
+      "learning_rate": 4.8308103294746216e-05,
+      "loss": 0.5952,
+      "step": 326
+    },
+    {
+      "epoch": 0.39302884615384615,
+      "grad_norm": 0.566429546136247,
+      "learning_rate": 4.828584149599288e-05,
+      "loss": 0.5681,
+      "step": 327
+    },
+    {
+      "epoch": 0.3942307692307692,
+      "grad_norm": 0.48675921012029477,
+      "learning_rate": 4.826357969723954e-05,
+      "loss": 0.5553,
+      "step": 328
+    },
+    {
+      "epoch": 0.3954326923076923,
+      "grad_norm": 0.40726532563349344,
+      "learning_rate": 4.82413178984862e-05,
+      "loss": 0.5768,
+      "step": 329
+    },
+    {
+      "epoch": 0.39663461538461536,
+      "grad_norm": 0.43847042852246204,
+      "learning_rate": 4.8219056099732865e-05,
+      "loss": 0.5312,
+      "step": 330
+    },
+    {
+      "epoch": 0.39783653846153844,
+      "grad_norm": 0.4507181174726414,
+      "learning_rate": 4.819679430097952e-05,
+      "loss": 0.6147,
+      "step": 331
+    },
+    {
+      "epoch": 0.39903846153846156,
+      "grad_norm": 0.37222554669821756,
+      "learning_rate": 4.817453250222618e-05,
+      "loss": 0.5478,
+      "step": 332
+    },
+    {
+      "epoch": 0.40024038461538464,
+      "grad_norm": 0.5176797307353165,
+      "learning_rate": 4.815227070347284e-05,
+      "loss": 0.5879,
+      "step": 333
+    },
+    {
+      "epoch": 0.4014423076923077,
+      "grad_norm": 0.3986924546727592,
+      "learning_rate": 4.8130008904719506e-05,
+      "loss": 0.588,
+      "step": 334
+    },
+    {
+      "epoch": 0.4026442307692308,
+      "grad_norm": 0.4234862464852973,
+      "learning_rate": 4.810774710596617e-05,
+      "loss": 0.5635,
+      "step": 335
+    },
+    {
+      "epoch": 0.40384615384615385,
+      "grad_norm": 0.4012123552489672,
+      "learning_rate": 4.8085485307212824e-05,
+      "loss": 0.5368,
+      "step": 336
+    },
+    {
+      "epoch": 0.4050480769230769,
+      "grad_norm": 0.3984852071869459,
+      "learning_rate": 4.8063223508459486e-05,
+      "loss": 0.5732,
+      "step": 337
+    },
+    {
+      "epoch": 0.40625,
+      "grad_norm": 0.46110225802190136,
+      "learning_rate": 4.804096170970615e-05,
+      "loss": 0.5697,
+      "step": 338
+    },
+    {
+      "epoch": 0.4074519230769231,
+      "grad_norm": 0.3545403499877976,
+      "learning_rate": 4.801869991095281e-05,
+      "loss": 0.5821,
+      "step": 339
+    },
+    {
+      "epoch": 0.40865384615384615,
+      "grad_norm": 0.5270345889193206,
+      "learning_rate": 4.7996438112199466e-05,
+      "loss": 0.5823,
+      "step": 340
+    },
+    {
+      "epoch": 0.4098557692307692,
+      "grad_norm": 0.37705851883536823,
+      "learning_rate": 4.797417631344613e-05,
+      "loss": 0.5845,
+      "step": 341
+    },
+    {
+      "epoch": 0.4110576923076923,
+      "grad_norm": 0.5049003916904655,
+      "learning_rate": 4.795191451469279e-05,
+      "loss": 0.5737,
+      "step": 342
+    },
+    {
+      "epoch": 0.41225961538461536,
+      "grad_norm": 0.4492334696474428,
+      "learning_rate": 4.792965271593945e-05,
+      "loss": 0.5612,
+      "step": 343
+    },
+    {
+      "epoch": 0.41346153846153844,
+      "grad_norm": 0.5717696068712854,
+      "learning_rate": 4.7907390917186114e-05,
+      "loss": 0.5464,
+      "step": 344
+    },
+    {
+      "epoch": 0.41466346153846156,
+      "grad_norm": 0.4431003659041063,
+      "learning_rate": 4.788512911843277e-05,
+      "loss": 0.5562,
+      "step": 345
+    },
+    {
+      "epoch": 0.41586538461538464,
+      "grad_norm": 0.4957058582601219,
+      "learning_rate": 4.786286731967943e-05,
+      "loss": 0.6024,
+      "step": 346
+    },
+    {
+      "epoch": 0.4170673076923077,
+      "grad_norm": 0.5203314378507247,
+      "learning_rate": 4.7840605520926094e-05,
+      "loss": 0.5843,
+      "step": 347
+    },
+    {
+      "epoch": 0.4182692307692308,
+      "grad_norm": 0.4018859781309395,
+      "learning_rate": 4.7818343722172756e-05,
+      "loss": 0.564,
+      "step": 348
+    },
+    {
+      "epoch": 0.41947115384615385,
+      "grad_norm": 0.42264926816260506,
+      "learning_rate": 4.779608192341942e-05,
+      "loss": 0.519,
+      "step": 349
+    },
+    {
+      "epoch": 0.4206730769230769,
+      "grad_norm": 0.360101814697917,
+      "learning_rate": 4.777382012466607e-05,
+      "loss": 0.5757,
+      "step": 350
+    },
+    {
+      "epoch": 0.421875,
+      "grad_norm": 0.5014701186533777,
+      "learning_rate": 4.7751558325912735e-05,
+      "loss": 0.5729,
+      "step": 351
+    },
+    {
+      "epoch": 0.4230769230769231,
+      "grad_norm": 0.42902334887973526,
+      "learning_rate": 4.77292965271594e-05,
+      "loss": 0.582,
+      "step": 352
+    },
+    {
+      "epoch": 0.42427884615384615,
+      "grad_norm": 0.4773755093311051,
+      "learning_rate": 4.770703472840606e-05,
+      "loss": 0.5387,
+      "step": 353
+    },
+    {
+      "epoch": 0.4254807692307692,
+      "grad_norm": 0.5939451795141301,
+      "learning_rate": 4.7684772929652715e-05,
+      "loss": 0.5898,
+      "step": 354
+    },
+    {
+      "epoch": 0.4266826923076923,
+      "grad_norm": 0.6279042424403133,
+      "learning_rate": 4.766251113089938e-05,
+      "loss": 0.578,
+      "step": 355
+    },
+    {
+      "epoch": 0.42788461538461536,
+      "grad_norm": 0.4315856617331273,
+      "learning_rate": 4.764024933214604e-05,
+      "loss": 0.5562,
+      "step": 356
+    },
+    {
+      "epoch": 0.42908653846153844,
+      "grad_norm": 0.5538884975426492,
+      "learning_rate": 4.76179875333927e-05,
+      "loss": 0.5342,
+      "step": 357
+    },
+    {
+      "epoch": 0.43028846153846156,
+      "grad_norm": 0.4368336820940177,
+      "learning_rate": 4.7595725734639364e-05,
+      "loss": 0.5356,
+      "step": 358
+    },
+    {
+      "epoch": 0.43149038461538464,
+      "grad_norm": 0.6013441847730407,
+      "learning_rate": 4.757346393588602e-05,
+      "loss": 0.5447,
+      "step": 359
+    },
+    {
+      "epoch": 0.4326923076923077,
+      "grad_norm": 0.4070996140841394,
+      "learning_rate": 4.755120213713268e-05,
+      "loss": 0.5497,
+      "step": 360
+    },
+    {
+      "epoch": 0.4338942307692308,
+      "grad_norm": 0.580740021170733,
+      "learning_rate": 4.752894033837934e-05,
+      "loss": 0.5647,
+      "step": 361
+    },
+    {
+      "epoch": 0.43509615384615385,
+      "grad_norm": 0.4505076007316201,
+      "learning_rate": 4.7506678539626005e-05,
+      "loss": 0.5609,
+      "step": 362
+    },
+    {
+      "epoch": 0.4362980769230769,
+      "grad_norm": 0.48526049184822356,
+      "learning_rate": 4.748441674087267e-05,
+      "loss": 0.5332,
+      "step": 363
+    },
+    {
+      "epoch": 0.4375,
+      "grad_norm": 0.4930033981383507,
+      "learning_rate": 4.746215494211932e-05,
+      "loss": 0.5759,
+      "step": 364
+    },
+    {
+      "epoch": 0.4387019230769231,
+      "grad_norm": 0.5021106367080058,
+      "learning_rate": 4.7439893143365985e-05,
+      "loss": 0.555,
+      "step": 365
+    },
+    {
+      "epoch": 0.43990384615384615,
+      "grad_norm": 0.4688507262378052,
+      "learning_rate": 4.741763134461265e-05,
+      "loss": 0.5633,
+      "step": 366
+    },
+    {
+      "epoch": 0.4411057692307692,
+      "grad_norm": 0.4271266809835138,
+      "learning_rate": 4.739536954585931e-05,
+      "loss": 0.5694,
+      "step": 367
+    },
+    {
+      "epoch": 0.4423076923076923,
+      "grad_norm": 0.5606200554109524,
+      "learning_rate": 4.7373107747105965e-05,
+      "loss": 0.5826,
+      "step": 368
+    },
+    {
+      "epoch": 0.44350961538461536,
+      "grad_norm": 0.5355662148356647,
+      "learning_rate": 4.735084594835263e-05,
+      "loss": 0.5956,
+      "step": 369
+    },
+    {
+      "epoch": 0.44471153846153844,
+      "grad_norm": 0.41050308843092465,
+      "learning_rate": 4.732858414959929e-05,
+      "loss": 0.5182,
+      "step": 370
+    },
+    {
+      "epoch": 0.44591346153846156,
+      "grad_norm": 0.5501321711357368,
+      "learning_rate": 4.730632235084595e-05,
+      "loss": 0.5444,
+      "step": 371
+    },
+    {
+      "epoch": 0.44711538461538464,
+      "grad_norm": 0.47900272379285846,
+      "learning_rate": 4.728406055209261e-05,
+      "loss": 0.5481,
+      "step": 372
+    },
+    {
+      "epoch": 0.4483173076923077,
+      "grad_norm": 0.44294933082758625,
+      "learning_rate": 4.726179875333927e-05,
+      "loss": 0.5715,
+      "step": 373
+    },
+    {
+      "epoch": 0.4495192307692308,
+      "grad_norm": 0.49669286431798787,
+      "learning_rate": 4.723953695458593e-05,
+      "loss": 0.5857,
+      "step": 374
+    },
+    {
+      "epoch": 0.45072115384615385,
+      "grad_norm": 0.4499149339238857,
+      "learning_rate": 4.721727515583259e-05,
+      "loss": 0.5856,
+      "step": 375
+    },
+    {
+      "epoch": 0.4519230769230769,
+      "grad_norm": 0.5388312380700053,
+      "learning_rate": 4.7195013357079255e-05,
+      "loss": 0.5858,
+      "step": 376
+    },
+    {
+      "epoch": 0.453125,
+      "grad_norm": 0.38887903315295613,
+      "learning_rate": 4.717275155832592e-05,
+      "loss": 0.5431,
+      "step": 377
+    },
+    {
+      "epoch": 0.4543269230769231,
+      "grad_norm": 0.5603549856720417,
+      "learning_rate": 4.715048975957257e-05,
+      "loss": 0.5647,
+      "step": 378
+    },
+    {
+      "epoch": 0.45552884615384615,
+      "grad_norm": 0.39210976763899724,
+      "learning_rate": 4.7128227960819234e-05,
+      "loss": 0.5641,
+      "step": 379
+    },
+    {
+      "epoch": 0.4567307692307692,
+      "grad_norm": 0.4672610975995842,
+      "learning_rate": 4.7105966162065897e-05,
+      "loss": 0.5524,
+      "step": 380
+    },
+    {
+      "epoch": 0.4579326923076923,
+      "grad_norm": 0.38299549605841743,
+      "learning_rate": 4.708370436331256e-05,
+      "loss": 0.5469,
+      "step": 381
+    },
+    {
+      "epoch": 0.45913461538461536,
+      "grad_norm": 0.40796760198312465,
+      "learning_rate": 4.7061442564559214e-05,
+      "loss": 0.5426,
+      "step": 382
+    },
+    {
+      "epoch": 0.46033653846153844,
+      "grad_norm": 0.3902541903037137,
+      "learning_rate": 4.7039180765805876e-05,
+      "loss": 0.6007,
+      "step": 383
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.3227998896946205,
+      "learning_rate": 4.7016918967052545e-05,
+      "loss": 0.5397,
+      "step": 384
+    },
+    {
+      "epoch": 0.46274038461538464,
+      "grad_norm": 0.4433022520952517,
+      "learning_rate": 4.69946571682992e-05,
+      "loss": 0.6017,
+      "step": 385
+    },
+    {
+      "epoch": 0.4639423076923077,
+      "grad_norm": 0.3414405671665633,
+      "learning_rate": 4.697239536954586e-05,
+      "loss": 0.5544,
+      "step": 386
+    },
+    {
+      "epoch": 0.4651442307692308,
+      "grad_norm": 0.4311697473084504,
+      "learning_rate": 4.695013357079252e-05,
+      "loss": 0.5876,
+      "step": 387
+    },
+    {
+      "epoch": 0.46634615384615385,
+      "grad_norm": 0.35940916663826816,
+      "learning_rate": 4.692787177203919e-05,
+      "loss": 0.5905,
+      "step": 388
+    },
+    {
+      "epoch": 0.4675480769230769,
+      "grad_norm": 0.38552364211783885,
+      "learning_rate": 4.690560997328584e-05,
+      "loss": 0.5423,
+      "step": 389
+    },
+    {
+      "epoch": 0.46875,
+      "grad_norm": 0.36740086129944416,
+      "learning_rate": 4.6883348174532504e-05,
+      "loss": 0.5388,
+      "step": 390
+    },
+    {
+      "epoch": 0.4699519230769231,
+      "grad_norm": 0.4502902047938388,
+      "learning_rate": 4.6861086375779167e-05,
+      "loss": 0.5748,
+      "step": 391
+    },
+    {
+      "epoch": 0.47115384615384615,
+      "grad_norm": 0.4005232891694738,
+      "learning_rate": 4.683882457702582e-05,
+      "loss": 0.5384,
+      "step": 392
+    },
+    {
+      "epoch": 0.4723557692307692,
+      "grad_norm": 0.4234784400047729,
+      "learning_rate": 4.681656277827249e-05,
+      "loss": 0.5558,
+      "step": 393
+    },
+    {
+      "epoch": 0.4735576923076923,
+      "grad_norm": 0.4415670652875659,
+      "learning_rate": 4.6794300979519146e-05,
+      "loss": 0.5518,
+      "step": 394
+    },
+    {
+      "epoch": 0.47475961538461536,
+      "grad_norm": 0.4179021918748966,
+      "learning_rate": 4.677203918076581e-05,
+      "loss": 0.5597,
+      "step": 395
+    },
+    {
+      "epoch": 0.47596153846153844,
+      "grad_norm": 0.42306960408074124,
+      "learning_rate": 4.6749777382012464e-05,
+      "loss": 0.5691,
+      "step": 396
+    },
+    {
+      "epoch": 0.47716346153846156,
+      "grad_norm": 0.41389651059027266,
+      "learning_rate": 4.672751558325913e-05,
+      "loss": 0.5878,
+      "step": 397
+    },
+    {
+      "epoch": 0.47836538461538464,
+      "grad_norm": 0.41687041852994405,
+      "learning_rate": 4.6705253784505795e-05,
+      "loss": 0.5377,
+      "step": 398
+    },
+    {
+      "epoch": 0.4795673076923077,
+      "grad_norm": 0.45020926072565637,
+      "learning_rate": 4.668299198575245e-05,
+      "loss": 0.5386,
+      "step": 399
+    },
+    {
+      "epoch": 0.4807692307692308,
+      "grad_norm": 0.4309891675300111,
+      "learning_rate": 4.666073018699911e-05,
+      "loss": 0.5771,
+      "step": 400
+    },
+    {
+      "epoch": 0.48197115384615385,
+      "grad_norm": 0.41483364147367735,
+      "learning_rate": 4.663846838824577e-05,
+      "loss": 0.5409,
+      "step": 401
+    },
+    {
+      "epoch": 0.4831730769230769,
+      "grad_norm": 0.534436141342152,
+      "learning_rate": 4.6616206589492436e-05,
+      "loss": 0.5549,
+      "step": 402
+    },
+    {
+      "epoch": 0.484375,
+      "grad_norm": 0.3638450474172596,
+      "learning_rate": 4.659394479073909e-05,
+      "loss": 0.5484,
+      "step": 403
+    },
+    {
+      "epoch": 0.4855769230769231,
+      "grad_norm": 0.40636358021419994,
+      "learning_rate": 4.6571682991985754e-05,
+      "loss": 0.5375,
+      "step": 404
+    },
+    {
+      "epoch": 0.48677884615384615,
+      "grad_norm": 0.3679869399706649,
+      "learning_rate": 4.6549421193232416e-05,
+      "loss": 0.5367,
+      "step": 405
+    },
+    {
+      "epoch": 0.4879807692307692,
+      "grad_norm": 0.3509685255644028,
+      "learning_rate": 4.652715939447908e-05,
+      "loss": 0.567,
+      "step": 406
+    },
+    {
+      "epoch": 0.4891826923076923,
+      "grad_norm": 0.42241303724511325,
+      "learning_rate": 4.650489759572574e-05,
+      "loss": 0.5237,
+      "step": 407
+    },
+    {
+      "epoch": 0.49038461538461536,
+      "grad_norm": 0.41157781782992625,
+      "learning_rate": 4.6482635796972396e-05,
+      "loss": 0.5585,
+      "step": 408
+    },
+    {
+      "epoch": 0.49158653846153844,
+      "grad_norm": 0.4177849407546035,
+      "learning_rate": 4.646037399821906e-05,
+      "loss": 0.5658,
+      "step": 409
+    },
+    {
+      "epoch": 0.49278846153846156,
+      "grad_norm": 0.4885528424808206,
+      "learning_rate": 4.643811219946571e-05,
+      "loss": 0.5537,
+      "step": 410
+    },
+    {
+      "epoch": 0.49399038461538464,
+      "grad_norm": 1.4997665660896546,
+      "learning_rate": 4.641585040071238e-05,
+      "loss": 0.6051,
+      "step": 411
+    },
+    {
+      "epoch": 0.4951923076923077,
+      "grad_norm": 0.4248523554763195,
+      "learning_rate": 4.6393588601959044e-05,
+      "loss": 0.5584,
+      "step": 412
+    },
+    {
+      "epoch": 0.4963942307692308,
+      "grad_norm": 0.40976182019293367,
+      "learning_rate": 4.63713268032057e-05,
+      "loss": 0.556,
+      "step": 413
+    },
+    {
+      "epoch": 0.49759615384615385,
+      "grad_norm": 0.41347911232334844,
+      "learning_rate": 4.634906500445236e-05,
+      "loss": 0.5722,
+      "step": 414
+    },
+    {
+      "epoch": 0.4987980769230769,
+      "grad_norm": 0.49514199454310154,
+      "learning_rate": 4.6326803205699024e-05,
+      "loss": 0.5447,
+      "step": 415
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4512543632408064,
+      "learning_rate": 4.6304541406945686e-05,
+      "loss": 0.5688,
+      "step": 416
+    },
+    {
+      "epoch": 0.5012019230769231,
+      "grad_norm": 0.5070781032589782,
+      "learning_rate": 4.628227960819234e-05,
+      "loss": 0.573,
+      "step": 417
+    },
+    {
+      "epoch": 0.5024038461538461,
+      "grad_norm": 0.36252975768634677,
+      "learning_rate": 4.6260017809439003e-05,
+      "loss": 0.5501,
+      "step": 418
+    },
+    {
+      "epoch": 0.5036057692307693,
+      "grad_norm": 0.4847176256674956,
+      "learning_rate": 4.6237756010685666e-05,
+      "loss": 0.5693,
+      "step": 419
+    },
+    {
+      "epoch": 0.5048076923076923,
+      "grad_norm": 0.3561855536066967,
+      "learning_rate": 4.621549421193233e-05,
+      "loss": 0.5488,
+      "step": 420
+    },
+    {
+      "epoch": 0.5060096153846154,
+      "grad_norm": 0.40957287272768006,
+      "learning_rate": 4.619323241317899e-05,
+      "loss": 0.559,
+      "step": 421
+    },
+    {
+      "epoch": 0.5072115384615384,
+      "grad_norm": 0.38642040987148724,
+      "learning_rate": 4.6170970614425645e-05,
+      "loss": 0.5786,
+      "step": 422
+    },
+    {
+      "epoch": 0.5084134615384616,
+      "grad_norm": 0.4419407620447961,
+      "learning_rate": 4.614870881567231e-05,
+      "loss": 0.5353,
+      "step": 423
+    },
+    {
+      "epoch": 0.5096153846153846,
+      "grad_norm": 0.4318806228860499,
+      "learning_rate": 4.612644701691897e-05,
+      "loss": 0.5573,
+      "step": 424
+    },
+    {
+      "epoch": 0.5108173076923077,
+      "grad_norm": 0.4780218693737317,
+      "learning_rate": 4.610418521816563e-05,
+      "loss": 0.5481,
+      "step": 425
+    },
+    {
+      "epoch": 0.5120192307692307,
+      "grad_norm": 0.4107228186731172,
+      "learning_rate": 4.6081923419412294e-05,
+      "loss": 0.5581,
+      "step": 426
+    },
+    {
+      "epoch": 0.5132211538461539,
+      "grad_norm": 0.47284213825536786,
+      "learning_rate": 4.605966162065895e-05,
+      "loss": 0.5424,
+      "step": 427
+    },
+    {
+      "epoch": 0.5144230769230769,
+      "grad_norm": 0.3840042185239489,
+      "learning_rate": 4.603739982190561e-05,
+      "loss": 0.5705,
+      "step": 428
+    },
+    {
+      "epoch": 0.515625,
+      "grad_norm": 0.5533425405187559,
+      "learning_rate": 4.601513802315227e-05,
+      "loss": 0.5754,
+      "step": 429
+    },
+    {
+      "epoch": 0.5168269230769231,
+      "grad_norm": 0.49231692892603873,
+      "learning_rate": 4.5992876224398935e-05,
+      "loss": 0.5709,
+      "step": 430
+    },
+    {
+      "epoch": 0.5180288461538461,
+      "grad_norm": 0.3992488944571112,
+      "learning_rate": 4.597061442564559e-05,
+      "loss": 0.524,
+      "step": 431
+    },
+    {
+      "epoch": 0.5192307692307693,
+      "grad_norm": 0.527648208235614,
+      "learning_rate": 4.594835262689225e-05,
+      "loss": 0.5511,
+      "step": 432
+    },
+    {
+      "epoch": 0.5204326923076923,
+      "grad_norm": 0.33339865617316805,
+      "learning_rate": 4.5926090828138915e-05,
+      "loss": 0.4923,
+      "step": 433
+    },
+    {
+      "epoch": 0.5216346153846154,
+      "grad_norm": 0.507564159115053,
+      "learning_rate": 4.590382902938558e-05,
+      "loss": 0.5447,
+      "step": 434
+    },
+    {
+      "epoch": 0.5228365384615384,
+      "grad_norm": 0.48162392488599487,
+      "learning_rate": 4.588156723063224e-05,
+      "loss": 0.5349,
+      "step": 435
+    },
+    {
+      "epoch": 0.5240384615384616,
+      "grad_norm": 0.4305736742856398,
+      "learning_rate": 4.5859305431878895e-05,
+      "loss": 0.5417,
+      "step": 436
+    },
+    {
+      "epoch": 0.5252403846153846,
+      "grad_norm": 0.5273874348562133,
+      "learning_rate": 4.583704363312556e-05,
+      "loss": 0.5632,
+      "step": 437
+    },
+    {
+      "epoch": 0.5264423076923077,
+      "grad_norm": 0.4957128821579046,
+      "learning_rate": 4.581478183437222e-05,
+      "loss": 0.537,
+      "step": 438
+    },
+    {
+      "epoch": 0.5276442307692307,
+      "grad_norm": 0.4538446216794328,
+      "learning_rate": 4.579252003561888e-05,
+      "loss": 0.5534,
+      "step": 439
+    },
+    {
+      "epoch": 0.5288461538461539,
+      "grad_norm": 0.5844234980735714,
+      "learning_rate": 4.577025823686554e-05,
+      "loss": 0.5761,
+      "step": 440
+    },
+    {
+      "epoch": 0.5300480769230769,
+      "grad_norm": 0.4109303849613329,
+      "learning_rate": 4.57479964381122e-05,
+      "loss": 0.5525,
+      "step": 441
+    },
+    {
+      "epoch": 0.53125,
+      "grad_norm": 0.4583613311554167,
+      "learning_rate": 4.572573463935886e-05,
+      "loss": 0.5489,
+      "step": 442
+    },
+    {
+      "epoch": 0.5324519230769231,
+      "grad_norm": 0.3834741433547509,
+      "learning_rate": 4.570347284060552e-05,
+      "loss": 0.5106,
+      "step": 443
+    },
+    {
+      "epoch": 0.5336538461538461,
+      "grad_norm": 0.45866817139641197,
+      "learning_rate": 4.5681211041852185e-05,
+      "loss": 0.5917,
+      "step": 444
+    },
+    {
+      "epoch": 0.5348557692307693,
+      "grad_norm": 0.4792468623642278,
+      "learning_rate": 4.565894924309884e-05,
+      "loss": 0.5194,
+      "step": 445
+    },
+    {
+      "epoch": 0.5360576923076923,
+      "grad_norm": 0.3869326931351502,
+      "learning_rate": 4.56366874443455e-05,
+      "loss": 0.5635,
+      "step": 446
+    },
+    {
+      "epoch": 0.5372596153846154,
+      "grad_norm": 0.38239242665595496,
+      "learning_rate": 4.561442564559217e-05,
+      "loss": 0.5551,
+      "step": 447
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.3910531314568318,
+      "learning_rate": 4.559216384683883e-05,
+      "loss": 0.5294,
+      "step": 448
+    },
+    {
+      "epoch": 0.5396634615384616,
+      "grad_norm": 0.3648953281042955,
+      "learning_rate": 4.556990204808549e-05,
+      "loss": 0.5756,
+      "step": 449
+    },
+    {
+      "epoch": 0.5408653846153846,
+      "grad_norm": 0.34799855128264945,
+      "learning_rate": 4.5547640249332144e-05,
+      "loss": 0.5532,
+      "step": 450
+    },
+    {
+      "epoch": 0.5420673076923077,
+      "grad_norm": 0.3829621980857354,
+      "learning_rate": 4.5525378450578806e-05,
+      "loss": 0.5532,
+      "step": 451
+    },
+    {
+      "epoch": 0.5432692307692307,
+      "grad_norm": 0.37826846288731286,
+      "learning_rate": 4.550311665182547e-05,
+      "loss": 0.5362,
+      "step": 452
+    },
+    {
+      "epoch": 0.5444711538461539,
+      "grad_norm": 0.31291408165309126,
+      "learning_rate": 4.548085485307213e-05,
+      "loss": 0.4863,
+      "step": 453
+    },
+    {
+      "epoch": 0.5456730769230769,
+      "grad_norm": 0.39378253933434226,
+      "learning_rate": 4.545859305431879e-05,
+      "loss": 0.5909,
+      "step": 454
+    },
+    {
+      "epoch": 0.546875,
+      "grad_norm": 0.4018662321102967,
+      "learning_rate": 4.543633125556545e-05,
+      "loss": 0.5283,
+      "step": 455
+    },
+    {
+      "epoch": 0.5480769230769231,
+      "grad_norm": 0.44354276573029594,
+      "learning_rate": 4.541406945681212e-05,
+      "loss": 0.5845,
+      "step": 456
+    },
+    {
+      "epoch": 0.5492788461538461,
+      "grad_norm": 0.36089314003741857,
+      "learning_rate": 4.539180765805877e-05,
+      "loss": 0.5503,
+      "step": 457
+    },
+    {
+      "epoch": 0.5504807692307693,
+      "grad_norm": 0.4551264156265257,
+      "learning_rate": 4.5369545859305434e-05,
+      "loss": 0.5712,
+      "step": 458
+    },
+    {
+      "epoch": 0.5516826923076923,
+      "grad_norm": 0.34429396023084896,
+      "learning_rate": 4.534728406055209e-05,
+      "loss": 0.5313,
+      "step": 459
+    },
+    {
+      "epoch": 0.5528846153846154,
+      "grad_norm": 0.4347279987762241,
+      "learning_rate": 4.532502226179875e-05,
+      "loss": 0.5478,
+      "step": 460
+    },
+    {
+      "epoch": 0.5540865384615384,
+      "grad_norm": 0.3530251585979505,
+      "learning_rate": 4.530276046304542e-05,
+      "loss": 0.518,
+      "step": 461
+    },
+    {
+      "epoch": 0.5552884615384616,
+      "grad_norm": 0.36137576343098265,
+      "learning_rate": 4.5280498664292076e-05,
+      "loss": 0.5645,
+      "step": 462
+    },
+    {
+      "epoch": 0.5564903846153846,
+      "grad_norm": 0.37424422234355365,
+      "learning_rate": 4.525823686553874e-05,
+      "loss": 0.5013,
+      "step": 463
+    },
+    {
+      "epoch": 0.5576923076923077,
+      "grad_norm": 0.3520600216413509,
+      "learning_rate": 4.5235975066785394e-05,
+      "loss": 0.5708,
+      "step": 464
+    },
+    {
+      "epoch": 0.5588942307692307,
+      "grad_norm": 0.46559809660928864,
+      "learning_rate": 4.521371326803206e-05,
+      "loss": 0.5336,
+      "step": 465
+    },
+    {
+      "epoch": 0.5600961538461539,
+      "grad_norm": 0.41732238107357056,
+      "learning_rate": 4.519145146927872e-05,
+      "loss": 0.5501,
+      "step": 466
+    },
+    {
+      "epoch": 0.5612980769230769,
+      "grad_norm": 0.4253074567929444,
+      "learning_rate": 4.516918967052538e-05,
+      "loss": 0.5419,
+      "step": 467
+    },
+    {
+      "epoch": 0.5625,
+      "grad_norm": 0.43670678515240163,
+      "learning_rate": 4.514692787177204e-05,
+      "loss": 0.4984,
+      "step": 468
+    },
+    {
+      "epoch": 0.5637019230769231,
+      "grad_norm": 0.35231747677230163,
+      "learning_rate": 4.51246660730187e-05,
+      "loss": 0.5435,
+      "step": 469
+    },
+    {
+      "epoch": 0.5649038461538461,
+      "grad_norm": 0.4208187461147585,
+      "learning_rate": 4.5102404274265367e-05,
+      "loss": 0.5734,
+      "step": 470
+    },
+    {
+      "epoch": 0.5661057692307693,
+      "grad_norm": 0.3401108062236541,
+      "learning_rate": 4.508014247551202e-05,
+      "loss": 0.5602,
+      "step": 471
+    },
+    {
+      "epoch": 0.5673076923076923,
+      "grad_norm": 0.32759896662369553,
+      "learning_rate": 4.5057880676758684e-05,
+      "loss": 0.5037,
+      "step": 472
+    },
+    {
+      "epoch": 0.5685096153846154,
+      "grad_norm": 0.36264807775777563,
+      "learning_rate": 4.503561887800534e-05,
+      "loss": 0.5175,
+      "step": 473
+    },
+    {
+      "epoch": 0.5697115384615384,
+      "grad_norm": 0.39735558469436244,
+      "learning_rate": 4.501335707925201e-05,
+      "loss": 0.5706,
+      "step": 474
+    },
+    {
+      "epoch": 0.5709134615384616,
+      "grad_norm": 0.3215377547323047,
+      "learning_rate": 4.499109528049867e-05,
+      "loss": 0.5474,
+      "step": 475
+    },
+    {
+      "epoch": 0.5721153846153846,
+      "grad_norm": 0.39801551679650593,
+      "learning_rate": 4.4968833481745326e-05,
+      "loss": 0.5343,
+      "step": 476
+    },
+    {
+      "epoch": 0.5733173076923077,
+      "grad_norm": 0.382989175608452,
+      "learning_rate": 4.494657168299199e-05,
+      "loss": 0.5487,
+      "step": 477
+    },
+    {
+      "epoch": 0.5745192307692307,
+      "grad_norm": 0.378925790075167,
+      "learning_rate": 4.492430988423865e-05,
+      "loss": 0.5443,
+      "step": 478
+    },
+    {
+      "epoch": 0.5757211538461539,
+      "grad_norm": 0.3699565027324386,
+      "learning_rate": 4.490204808548531e-05,
+      "loss": 0.5694,
+      "step": 479
+    },
+    {
+      "epoch": 0.5769230769230769,
+      "grad_norm": 0.4027610791603482,
+      "learning_rate": 4.487978628673197e-05,
+      "loss": 0.5402,
+      "step": 480
+    },
+    {
+      "epoch": 0.578125,
+      "grad_norm": 0.4158014620833248,
+      "learning_rate": 4.485752448797863e-05,
+      "loss": 0.5731,
+      "step": 481
+    },
+    {
+      "epoch": 0.5793269230769231,
+      "grad_norm": 0.34933151723901024,
+      "learning_rate": 4.483526268922529e-05,
+      "loss": 0.5661,
+      "step": 482
+    },
+    {
+      "epoch": 0.5805288461538461,
+      "grad_norm": 0.4479371612664256,
+      "learning_rate": 4.4813000890471954e-05,
+      "loss": 0.53,
+      "step": 483
+    },
+    {
+      "epoch": 0.5817307692307693,
+      "grad_norm": 0.39208555485493035,
+      "learning_rate": 4.4790739091718616e-05,
+      "loss": 0.531,
+      "step": 484
+    },
+    {
+      "epoch": 0.5829326923076923,
+      "grad_norm": 0.3686890019453203,
+      "learning_rate": 4.476847729296527e-05,
+      "loss": 0.57,
+      "step": 485
+    },
+    {
+      "epoch": 0.5841346153846154,
+      "grad_norm": 0.35760871122927995,
+      "learning_rate": 4.4746215494211934e-05,
+      "loss": 0.5631,
+      "step": 486
+    },
+    {
+      "epoch": 0.5853365384615384,
+      "grad_norm": 0.3503088797800585,
+      "learning_rate": 4.4723953695458596e-05,
+      "loss": 0.5569,
+      "step": 487
+    },
+    {
+      "epoch": 0.5865384615384616,
+      "grad_norm": 0.37416666288252914,
+      "learning_rate": 4.470169189670526e-05,
+      "loss": 0.5684,
+      "step": 488
+    },
+    {
+      "epoch": 0.5877403846153846,
+      "grad_norm": 0.4139692447597723,
+      "learning_rate": 4.467943009795192e-05,
+      "loss": 0.5419,
+      "step": 489
+    },
+    {
+      "epoch": 0.5889423076923077,
+      "grad_norm": 0.3335119762210027,
+      "learning_rate": 4.4657168299198575e-05,
+      "loss": 0.5181,
+      "step": 490
+    },
+    {
+      "epoch": 0.5901442307692307,
+      "grad_norm": 0.4487375580416237,
+      "learning_rate": 4.463490650044524e-05,
+      "loss": 0.5126,
+      "step": 491
+    },
+    {
+      "epoch": 0.5913461538461539,
+      "grad_norm": 0.4474492537342872,
+      "learning_rate": 4.46126447016919e-05,
+      "loss": 0.5329,
+      "step": 492
+    },
+    {
+      "epoch": 0.5925480769230769,
+      "grad_norm": 0.41351958894846114,
+      "learning_rate": 4.459038290293856e-05,
+      "loss": 0.5688,
+      "step": 493
+    },
+    {
+      "epoch": 0.59375,
+      "grad_norm": 0.41938358443442253,
+      "learning_rate": 4.456812110418522e-05,
+      "loss": 0.5138,
+      "step": 494
+    },
+    {
+      "epoch": 0.5949519230769231,
+      "grad_norm": 0.3706663184065996,
+      "learning_rate": 4.454585930543188e-05,
+      "loss": 0.5291,
+      "step": 495
+    },
+    {
+      "epoch": 0.5961538461538461,
+      "grad_norm": 0.37837852869322364,
+      "learning_rate": 4.452359750667854e-05,
+      "loss": 0.5252,
+      "step": 496
+    },
+    {
+      "epoch": 0.5973557692307693,
+      "grad_norm": 0.3872261085827947,
+      "learning_rate": 4.4501335707925203e-05,
+      "loss": 0.5372,
+      "step": 497
+    },
+    {
+      "epoch": 0.5985576923076923,
+      "grad_norm": 0.3789299118554819,
+      "learning_rate": 4.4479073909171866e-05,
+      "loss": 0.5356,
+      "step": 498
+    },
+    {
+      "epoch": 0.5997596153846154,
+      "grad_norm": 0.36806798891937054,
+      "learning_rate": 4.445681211041852e-05,
+      "loss": 0.5114,
+      "step": 499
+    },
+    {
+      "epoch": 0.6009615384615384,
+      "grad_norm": 0.4073832033469708,
+      "learning_rate": 4.443455031166518e-05,
+      "loss": 0.5661,
+      "step": 500
+    },
+    {
+      "epoch": 0.6021634615384616,
+      "grad_norm": 0.3366048287789353,
+      "learning_rate": 4.4412288512911845e-05,
+      "loss": 0.5461,
+      "step": 501
+    },
+    {
+      "epoch": 0.6033653846153846,
+      "grad_norm": 0.3585921641911143,
+      "learning_rate": 4.439002671415851e-05,
+      "loss": 0.5485,
+      "step": 502
+    },
+    {
+      "epoch": 0.6045673076923077,
+      "grad_norm": 0.4322285770708658,
+      "learning_rate": 4.436776491540517e-05,
+      "loss": 0.5248,
+      "step": 503
+    },
+    {
+      "epoch": 0.6057692307692307,
+      "grad_norm": 0.40473772997583113,
+      "learning_rate": 4.4345503116651825e-05,
+      "loss": 0.5572,
+      "step": 504
+    },
+    {
+      "epoch": 0.6069711538461539,
+      "grad_norm": 0.36383403672240894,
+      "learning_rate": 4.432324131789849e-05,
+      "loss": 0.5331,
+      "step": 505
+    },
+    {
+      "epoch": 0.6081730769230769,
+      "grad_norm": 0.3580115333548413,
+      "learning_rate": 4.430097951914515e-05,
+      "loss": 0.5344,
+      "step": 506
+    },
+    {
+      "epoch": 0.609375,
+      "grad_norm": 0.44677519487047274,
+      "learning_rate": 4.427871772039181e-05,
+      "loss": 0.5356,
+      "step": 507
+    },
+    {
+      "epoch": 0.6105769230769231,
+      "grad_norm": 0.3057328954804058,
+      "learning_rate": 4.4256455921638467e-05,
+      "loss": 0.5236,
+      "step": 508
+    },
+    {
+      "epoch": 0.6117788461538461,
+      "grad_norm": 0.4155279431274658,
+      "learning_rate": 4.423419412288513e-05,
+      "loss": 0.5543,
+      "step": 509
+    },
+    {
+      "epoch": 0.6129807692307693,
+      "grad_norm": 0.3541978247014027,
+      "learning_rate": 4.421193232413179e-05,
+      "loss": 0.5275,
+      "step": 510
+    },
+    {
+      "epoch": 0.6141826923076923,
+      "grad_norm": 0.3712374772762189,
+      "learning_rate": 4.418967052537845e-05,
+      "loss": 0.5282,
+      "step": 511
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.4285631661487287,
+      "learning_rate": 4.4167408726625115e-05,
+      "loss": 0.5405,
+      "step": 512
+    },
+    {
+      "epoch": 0.6165865384615384,
+      "grad_norm": 0.40320934652030915,
+      "learning_rate": 4.414514692787177e-05,
+      "loss": 0.5429,
+      "step": 513
+    },
+    {
+      "epoch": 0.6177884615384616,
+      "grad_norm": 0.3679945978535122,
+      "learning_rate": 4.412288512911843e-05,
+      "loss": 0.5459,
+      "step": 514
+    },
+    {
+      "epoch": 0.6189903846153846,
+      "grad_norm": 0.4159569858991235,
+      "learning_rate": 4.4100623330365095e-05,
+      "loss": 0.5297,
+      "step": 515
+    },
+    {
+      "epoch": 0.6201923076923077,
+      "grad_norm": 0.3910627762143478,
+      "learning_rate": 4.407836153161176e-05,
+      "loss": 0.5339,
+      "step": 516
+    },
+    {
+      "epoch": 0.6213942307692307,
+      "grad_norm": 0.42318908466981653,
+      "learning_rate": 4.405609973285842e-05,
+      "loss": 0.5387,
+      "step": 517
+    },
+    {
+      "epoch": 0.6225961538461539,
+      "grad_norm": 0.30578675437201014,
+      "learning_rate": 4.4033837934105074e-05,
+      "loss": 0.5355,
+      "step": 518
+    },
+    {
+      "epoch": 0.6237980769230769,
+      "grad_norm": 0.40304068921213787,
+      "learning_rate": 4.4011576135351736e-05,
+      "loss": 0.5306,
+      "step": 519
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 0.34614680278073623,
+      "learning_rate": 4.39893143365984e-05,
+      "loss": 0.5079,
+      "step": 520
+    },
+    {
+      "epoch": 0.6262019230769231,
+      "grad_norm": 0.43925088433509085,
+      "learning_rate": 4.396705253784506e-05,
+      "loss": 0.5372,
+      "step": 521
+    },
+    {
+      "epoch": 0.6274038461538461,
+      "grad_norm": 0.33554514526838974,
+      "learning_rate": 4.3944790739091716e-05,
+      "loss": 0.5273,
+      "step": 522
+    },
+    {
+      "epoch": 0.6286057692307693,
+      "grad_norm": 0.40970173543601085,
+      "learning_rate": 4.392252894033838e-05,
+      "loss": 0.5336,
+      "step": 523
+    },
+    {
+      "epoch": 0.6298076923076923,
+      "grad_norm": 0.3727159006100873,
+      "learning_rate": 4.390026714158505e-05,
+      "loss": 0.5867,
+      "step": 524
+    },
+    {
+      "epoch": 0.6310096153846154,
+      "grad_norm": 0.4278402459513843,
+      "learning_rate": 4.38780053428317e-05,
+      "loss": 0.5475,
+      "step": 525
+    },
+    {
+      "epoch": 0.6322115384615384,
+      "grad_norm": 0.43541470305955526,
+      "learning_rate": 4.3855743544078365e-05,
+      "loss": 0.5544,
+      "step": 526
+    },
+    {
+      "epoch": 0.6334134615384616,
+      "grad_norm": 0.3497564911636232,
+      "learning_rate": 4.383348174532502e-05,
+      "loss": 0.5412,
+      "step": 527
+    },
+    {
+      "epoch": 0.6346153846153846,
+      "grad_norm": 0.3438848485943213,
+      "learning_rate": 4.381121994657169e-05,
+      "loss": 0.5182,
+      "step": 528
+    },
+    {
+      "epoch": 0.6358173076923077,
+      "grad_norm": 0.390481918768078,
+      "learning_rate": 4.3788958147818344e-05,
+      "loss": 0.5153,
+      "step": 529
+    },
+    {
+      "epoch": 0.6370192307692307,
+      "grad_norm": 0.3483416736669727,
+      "learning_rate": 4.3766696349065006e-05,
+      "loss": 0.5091,
+      "step": 530
+    },
+    {
+      "epoch": 0.6382211538461539,
+      "grad_norm": 0.4491897888094657,
+      "learning_rate": 4.374443455031167e-05,
+      "loss": 0.498,
+      "step": 531
+    },
+    {
+      "epoch": 0.6394230769230769,
+      "grad_norm": 0.3386199698801738,
+      "learning_rate": 4.3722172751558324e-05,
+      "loss": 0.5387,
+      "step": 532
+    },
+    {
+      "epoch": 0.640625,
+      "grad_norm": 0.4542256271670574,
+      "learning_rate": 4.369991095280499e-05,
+      "loss": 0.5198,
+      "step": 533
+    },
+    {
+      "epoch": 0.6418269230769231,
+      "grad_norm": 0.38993232519994747,
+      "learning_rate": 4.367764915405165e-05,
+      "loss": 0.5561,
+      "step": 534
+    },
+    {
+      "epoch": 0.6430288461538461,
+      "grad_norm": 0.46829960847865815,
+      "learning_rate": 4.365538735529831e-05,
+      "loss": 0.557,
+      "step": 535
+    },
+    {
+      "epoch": 0.6442307692307693,
+      "grad_norm": 0.4083258921524021,
+      "learning_rate": 4.363312555654497e-05,
+      "loss": 0.504,
+      "step": 536
+    },
+    {
+      "epoch": 0.6454326923076923,
+      "grad_norm": 0.4713218021525624,
+      "learning_rate": 4.3610863757791634e-05,
+      "loss": 0.5544,
+      "step": 537
+    },
+    {
+      "epoch": 0.6466346153846154,
+      "grad_norm": 0.34700891969187514,
+      "learning_rate": 4.35886019590383e-05,
+      "loss": 0.5552,
+      "step": 538
+    },
+    {
+      "epoch": 0.6478365384615384,
+      "grad_norm": 0.5399351173086345,
+      "learning_rate": 4.356634016028495e-05,
+      "loss": 0.5431,
+      "step": 539
+    },
+    {
+      "epoch": 0.6490384615384616,
+      "grad_norm": 0.37761964046624613,
+      "learning_rate": 4.3544078361531614e-05,
+      "loss": 0.5258,
+      "step": 540
+    },
+    {
+      "epoch": 0.6502403846153846,
+      "grad_norm": 0.5205759451623062,
+      "learning_rate": 4.352181656277827e-05,
+      "loss": 0.5162,
+      "step": 541
+    },
+    {
+      "epoch": 0.6514423076923077,
+      "grad_norm": 0.34077500051299714,
+      "learning_rate": 4.349955476402494e-05,
+      "loss": 0.5388,
+      "step": 542
+    },
+    {
+      "epoch": 0.6526442307692307,
+      "grad_norm": 0.513351086047026,
+      "learning_rate": 4.34772929652716e-05,
+      "loss": 0.5541,
+      "step": 543
+    },
+    {
+      "epoch": 0.6538461538461539,
+      "grad_norm": 0.4529597176350292,
+      "learning_rate": 4.3455031166518256e-05,
+      "loss": 0.5505,
+      "step": 544
+    },
+    {
+      "epoch": 0.6550480769230769,
+      "grad_norm": 0.4530853165646563,
+      "learning_rate": 4.343276936776492e-05,
+      "loss": 0.5607,
+      "step": 545
+    },
+    {
+      "epoch": 0.65625,
+      "grad_norm": 0.5418728727201232,
+      "learning_rate": 4.341050756901158e-05,
+      "loss": 0.568,
+      "step": 546
+    },
+    {
+      "epoch": 0.6574519230769231,
+      "grad_norm": 0.37506748838925313,
+      "learning_rate": 4.338824577025824e-05,
+      "loss": 0.5181,
+      "step": 547
+    },
+    {
+      "epoch": 0.6586538461538461,
+      "grad_norm": 0.39632834286596996,
+      "learning_rate": 4.33659839715049e-05,
+      "loss": 0.5587,
+      "step": 548
+    },
+    {
+      "epoch": 0.6598557692307693,
+      "grad_norm": 0.36932757912905145,
+      "learning_rate": 4.334372217275156e-05,
+      "loss": 0.5435,
+      "step": 549
+    },
+    {
+      "epoch": 0.6610576923076923,
+      "grad_norm": 0.3678980783008907,
+      "learning_rate": 4.332146037399822e-05,
+      "loss": 0.5471,
+      "step": 550
+    },
+    {
+      "epoch": 0.6622596153846154,
+      "grad_norm": 0.38824352059278117,
+      "learning_rate": 4.3299198575244884e-05,
+      "loss": 0.5317,
+      "step": 551
+    },
+    {
+      "epoch": 0.6634615384615384,
+      "grad_norm": 0.40700682374539526,
+      "learning_rate": 4.3276936776491546e-05,
+      "loss": 0.5747,
+      "step": 552
+    },
+    {
+      "epoch": 0.6646634615384616,
+      "grad_norm": 0.45841450469608785,
+      "learning_rate": 4.32546749777382e-05,
+      "loss": 0.4793,
+      "step": 553
+    },
+    {
+      "epoch": 0.6658653846153846,
+      "grad_norm": 0.34854712295706614,
+      "learning_rate": 4.3232413178984864e-05,
+      "loss": 0.5401,
+      "step": 554
+    },
+    {
+      "epoch": 0.6670673076923077,
+      "grad_norm": 0.4620257972601391,
+      "learning_rate": 4.3210151380231526e-05,
+      "loss": 0.5295,
+      "step": 555
+    },
+    {
+      "epoch": 0.6682692307692307,
+      "grad_norm": 0.4681276476691421,
+      "learning_rate": 4.318788958147819e-05,
+      "loss": 0.5366,
+      "step": 556
+    },
+    {
+      "epoch": 0.6694711538461539,
+      "grad_norm": 0.3728161001716294,
+      "learning_rate": 4.316562778272485e-05,
+      "loss": 0.5524,
+      "step": 557
+    },
+    {
+      "epoch": 0.6706730769230769,
+      "grad_norm": 0.6541028499436519,
+      "learning_rate": 4.3143365983971505e-05,
+      "loss": 0.5433,
+      "step": 558
+    },
+    {
+      "epoch": 0.671875,
+      "grad_norm": 0.3990779931171476,
+      "learning_rate": 4.312110418521817e-05,
+      "loss": 0.5193,
+      "step": 559
+    },
+    {
+      "epoch": 0.6730769230769231,
+      "grad_norm": 0.5345385808964015,
+      "learning_rate": 4.309884238646483e-05,
+      "loss": 0.5351,
+      "step": 560
+    },
+    {
+      "epoch": 0.6742788461538461,
+      "grad_norm": 0.42462412174779346,
+      "learning_rate": 4.307658058771149e-05,
+      "loss": 0.5306,
+      "step": 561
+    },
+    {
+      "epoch": 0.6754807692307693,
+      "grad_norm": 0.35521664317302737,
+      "learning_rate": 4.305431878895815e-05,
+      "loss": 0.5207,
+      "step": 562
+    },
+    {
+      "epoch": 0.6766826923076923,
+      "grad_norm": 0.4269498601669769,
+      "learning_rate": 4.303205699020481e-05,
+      "loss": 0.5535,
+      "step": 563
+    },
+    {
+      "epoch": 0.6778846153846154,
+      "grad_norm": 0.3606619486596656,
+      "learning_rate": 4.300979519145147e-05,
+      "loss": 0.5414,
+      "step": 564
+    },
+    {
+      "epoch": 0.6790865384615384,
+      "grad_norm": 0.40852562646483837,
+      "learning_rate": 4.2987533392698134e-05,
+      "loss": 0.5262,
+      "step": 565
+    },
+    {
+      "epoch": 0.6802884615384616,
+      "grad_norm": 0.3481570848549128,
+      "learning_rate": 4.2965271593944796e-05,
+      "loss": 0.5332,
+      "step": 566
+    },
+    {
+      "epoch": 0.6814903846153846,
+      "grad_norm": 0.3907035064574659,
+      "learning_rate": 4.294300979519145e-05,
+      "loss": 0.5194,
+      "step": 567
+    },
+    {
+      "epoch": 0.6826923076923077,
+      "grad_norm": 0.4161985700639464,
+      "learning_rate": 4.292074799643811e-05,
+      "loss": 0.5146,
+      "step": 568
+    },
+    {
+      "epoch": 0.6838942307692307,
+      "grad_norm": 0.3941424873352859,
+      "learning_rate": 4.2898486197684775e-05,
+      "loss": 0.5246,
+      "step": 569
+    },
+    {
+      "epoch": 0.6850961538461539,
+      "grad_norm": 0.4301914462747028,
+      "learning_rate": 4.287622439893144e-05,
+      "loss": 0.5319,
+      "step": 570
+    },
+    {
+      "epoch": 0.6862980769230769,
+      "grad_norm": 0.41221023242603316,
+      "learning_rate": 4.28539626001781e-05,
+      "loss": 0.5404,
+      "step": 571
+    },
+    {
+      "epoch": 0.6875,
+      "grad_norm": 0.538686304319232,
+      "learning_rate": 4.2831700801424755e-05,
+      "loss": 0.5187,
+      "step": 572
+    },
+    {
+      "epoch": 0.6887019230769231,
+      "grad_norm": 0.46448600221422726,
+      "learning_rate": 4.280943900267142e-05,
+      "loss": 0.4901,
+      "step": 573
+    },
+    {
+      "epoch": 0.6899038461538461,
+      "grad_norm": 0.4643193404282257,
+      "learning_rate": 4.278717720391808e-05,
+      "loss": 0.512,
+      "step": 574
+    },
+    {
+      "epoch": 0.6911057692307693,
+      "grad_norm": 0.566902042350752,
+      "learning_rate": 4.276491540516474e-05,
+      "loss": 0.5134,
+      "step": 575
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.33679768625141404,
+      "learning_rate": 4.27426536064114e-05,
+      "loss": 0.5223,
+      "step": 576
+    },
+    {
+      "epoch": 0.6935096153846154,
+      "grad_norm": 0.47746133634536303,
+      "learning_rate": 4.272039180765806e-05,
+      "loss": 0.5675,
+      "step": 577
+    },
+    {
+      "epoch": 0.6947115384615384,
+      "grad_norm": 0.4167526644264464,
+      "learning_rate": 4.269813000890472e-05,
+      "loss": 0.5455,
+      "step": 578
+    },
+    {
+      "epoch": 0.6959134615384616,
+      "grad_norm": 0.4204959164090705,
+      "learning_rate": 4.267586821015138e-05,
+      "loss": 0.5466,
+      "step": 579
+    },
+    {
+      "epoch": 0.6971153846153846,
+      "grad_norm": 0.500987677802682,
+      "learning_rate": 4.2653606411398045e-05,
+      "loss": 0.526,
+      "step": 580
+    },
+    {
+      "epoch": 0.6983173076923077,
+      "grad_norm": 0.4758487082159854,
+      "learning_rate": 4.26313446126447e-05,
+      "loss": 0.5583,
+      "step": 581
+    },
+    {
+      "epoch": 0.6995192307692307,
+      "grad_norm": 0.4651613711399878,
+      "learning_rate": 4.260908281389136e-05,
+      "loss": 0.5464,
+      "step": 582
+    },
+    {
+      "epoch": 0.7007211538461539,
+      "grad_norm": 0.547243083176506,
+      "learning_rate": 4.2586821015138025e-05,
+      "loss": 0.5541,
+      "step": 583
+    },
+    {
+      "epoch": 0.7019230769230769,
+      "grad_norm": 0.3958570236061405,
+      "learning_rate": 4.256455921638469e-05,
+      "loss": 0.5463,
+      "step": 584
+    },
+    {
+      "epoch": 0.703125,
+      "grad_norm": 0.5476151113674058,
+      "learning_rate": 4.254229741763135e-05,
+      "loss": 0.5741,
+      "step": 585
+    },
+    {
+      "epoch": 0.7043269230769231,
+      "grad_norm": 0.3677889127871356,
+      "learning_rate": 4.2520035618878004e-05,
+      "loss": 0.5357,
+      "step": 586
+    },
+    {
+      "epoch": 0.7055288461538461,
+      "grad_norm": 0.5629360440721651,
+      "learning_rate": 4.249777382012467e-05,
+      "loss": 0.5562,
+      "step": 587
+    },
+    {
+      "epoch": 0.7067307692307693,
+      "grad_norm": 0.5310581248501943,
+      "learning_rate": 4.247551202137133e-05,
+      "loss": 0.5367,
+      "step": 588
+    },
+    {
+      "epoch": 0.7079326923076923,
+      "grad_norm": 0.4210258201929359,
+      "learning_rate": 4.245325022261799e-05,
+      "loss": 0.5267,
+      "step": 589
+    },
+    {
+      "epoch": 0.7091346153846154,
+      "grad_norm": 0.5500921152060687,
+      "learning_rate": 4.2430988423864646e-05,
+      "loss": 0.5259,
+      "step": 590
+    },
+    {
+      "epoch": 0.7103365384615384,
+      "grad_norm": 0.35935211967360786,
+      "learning_rate": 4.240872662511131e-05,
+      "loss": 0.5281,
+      "step": 591
+    },
+    {
+      "epoch": 0.7115384615384616,
+      "grad_norm": 0.41475036932498577,
+      "learning_rate": 4.238646482635798e-05,
+      "loss": 0.5378,
+      "step": 592
+    },
+    {
+      "epoch": 0.7127403846153846,
+      "grad_norm": 0.3734269349998828,
+      "learning_rate": 4.236420302760463e-05,
+      "loss": 0.5561,
+      "step": 593
+    },
+    {
+      "epoch": 0.7139423076923077,
+      "grad_norm": 0.3566083975575042,
+      "learning_rate": 4.2341941228851295e-05,
+      "loss": 0.5567,
+      "step": 594
+    },
+    {
+      "epoch": 0.7151442307692307,
+      "grad_norm": 0.36646207015816584,
+      "learning_rate": 4.231967943009795e-05,
+      "loss": 0.5271,
+      "step": 595
+    },
+    {
+      "epoch": 0.7163461538461539,
+      "grad_norm": 0.37601938670490354,
+      "learning_rate": 4.229741763134462e-05,
+      "loss": 0.5209,
+      "step": 596
+    },
+    {
+      "epoch": 0.7175480769230769,
+      "grad_norm": 0.32696097999728185,
+      "learning_rate": 4.2275155832591274e-05,
+      "loss": 0.5556,
+      "step": 597
+    },
+    {
+      "epoch": 0.71875,
+      "grad_norm": 0.4407088379701345,
+      "learning_rate": 4.2252894033837936e-05,
+      "loss": 0.5613,
+      "step": 598
+    },
+    {
+      "epoch": 0.7199519230769231,
+      "grad_norm": 0.37638667564958544,
+      "learning_rate": 4.22306322350846e-05,
+      "loss": 0.5292,
+      "step": 599
+    },
+    {
+      "epoch": 0.7211538461538461,
+      "grad_norm": 0.42746441984770744,
+      "learning_rate": 4.2208370436331254e-05,
+      "loss": 0.5465,
+      "step": 600
+    },
+    {
+      "epoch": 0.7223557692307693,
+      "grad_norm": 0.3096964024645313,
+      "learning_rate": 4.218610863757792e-05,
+      "loss": 0.5177,
+      "step": 601
+    },
+    {
+      "epoch": 0.7235576923076923,
+      "grad_norm": 0.48976965575865844,
+      "learning_rate": 4.216384683882458e-05,
+      "loss": 0.4978,
+      "step": 602
+    },
+    {
+      "epoch": 0.7247596153846154,
+      "grad_norm": 0.38793404363844436,
+      "learning_rate": 4.214158504007124e-05,
+      "loss": 0.5204,
+      "step": 603
+    },
+    {
+      "epoch": 0.7259615384615384,
+      "grad_norm": 0.38966326833224213,
+      "learning_rate": 4.2119323241317896e-05,
+      "loss": 0.5258,
+      "step": 604
+    },
+    {
+      "epoch": 0.7271634615384616,
+      "grad_norm": 0.37313164816549194,
+      "learning_rate": 4.2097061442564565e-05,
+      "loss": 0.5173,
+      "step": 605
+    },
+    {
+      "epoch": 0.7283653846153846,
+      "grad_norm": 0.35693227935063165,
+      "learning_rate": 4.207479964381123e-05,
+      "loss": 0.5214,
+      "step": 606
+    },
+    {
+      "epoch": 0.7295673076923077,
+      "grad_norm": 0.3331684207718226,
+      "learning_rate": 4.205253784505788e-05,
+      "loss": 0.5205,
+      "step": 607
+    },
+    {
+      "epoch": 0.7307692307692307,
+      "grad_norm": 0.3768630491327925,
+      "learning_rate": 4.2030276046304544e-05,
+      "loss": 0.5398,
+      "step": 608
+    },
+    {
+      "epoch": 0.7319711538461539,
+      "grad_norm": 0.35011752841631383,
+      "learning_rate": 4.20080142475512e-05,
+      "loss": 0.5426,
+      "step": 609
+    },
+    {
+      "epoch": 0.7331730769230769,
+      "grad_norm": 0.40827249797703025,
+      "learning_rate": 4.198575244879787e-05,
+      "loss": 0.5409,
+      "step": 610
+    },
+    {
+      "epoch": 0.734375,
+      "grad_norm": 0.3564955265034016,
+      "learning_rate": 4.1963490650044524e-05,
+      "loss": 0.5192,
+      "step": 611
+    },
+    {
+      "epoch": 0.7355769230769231,
+      "grad_norm": 0.3681673250241195,
+      "learning_rate": 4.1941228851291186e-05,
+      "loss": 0.4886,
+      "step": 612
+    },
+    {
+      "epoch": 0.7367788461538461,
+      "grad_norm": 0.35640003089410405,
+      "learning_rate": 4.191896705253785e-05,
+      "loss": 0.5253,
+      "step": 613
+    },
+    {
+      "epoch": 0.7379807692307693,
+      "grad_norm": 0.4302717192894493,
+      "learning_rate": 4.189670525378451e-05,
+      "loss": 0.5352,
+      "step": 614
+    },
+    {
+      "epoch": 0.7391826923076923,
+      "grad_norm": 0.3403798719551935,
+      "learning_rate": 4.187444345503117e-05,
+      "loss": 0.5187,
+      "step": 615
+    },
+    {
+      "epoch": 0.7403846153846154,
+      "grad_norm": 0.3926870536494106,
+      "learning_rate": 4.185218165627783e-05,
+      "loss": 0.5322,
+      "step": 616
+    },
+    {
+      "epoch": 0.7415865384615384,
+      "grad_norm": 0.348579732261744,
+      "learning_rate": 4.182991985752449e-05,
+      "loss": 0.5414,
+      "step": 617
+    },
+    {
+      "epoch": 0.7427884615384616,
+      "grad_norm": 0.3633471449994058,
+      "learning_rate": 4.180765805877115e-05,
+      "loss": 0.5107,
+      "step": 618
+    },
+    {
+      "epoch": 0.7439903846153846,
+      "grad_norm": 0.32626885330987787,
+      "learning_rate": 4.1785396260017814e-05,
+      "loss": 0.5022,
+      "step": 619
+    },
+    {
+      "epoch": 0.7451923076923077,
+      "grad_norm": 0.39471828724451274,
+      "learning_rate": 4.1763134461264476e-05,
+      "loss": 0.5402,
+      "step": 620
+    },
+    {
+      "epoch": 0.7463942307692307,
+      "grad_norm": 0.33722627673239824,
+      "learning_rate": 4.174087266251113e-05,
+      "loss": 0.5326,
+      "step": 621
+    },
+    {
+      "epoch": 0.7475961538461539,
+      "grad_norm": 0.4143675955998991,
+      "learning_rate": 4.1718610863757794e-05,
+      "loss": 0.5446,
+      "step": 622
+    },
+    {
+      "epoch": 0.7487980769230769,
+      "grad_norm": 0.3153606863102826,
+      "learning_rate": 4.1696349065004456e-05,
+      "loss": 0.4956,
+      "step": 623
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.40340897291277716,
+      "learning_rate": 4.167408726625112e-05,
+      "loss": 0.526,
+      "step": 624
+    },
+    {
+      "epoch": 0.7512019230769231,
+      "grad_norm": 0.38800562294986,
+      "learning_rate": 4.165182546749777e-05,
+      "loss": 0.546,
+      "step": 625
+    },
+    {
+      "epoch": 0.7524038461538461,
+      "grad_norm": 0.35578014700784344,
+      "learning_rate": 4.1629563668744435e-05,
+      "loss": 0.5327,
+      "step": 626
+    },
+    {
+      "epoch": 0.7536057692307693,
+      "grad_norm": 0.7137819674150702,
+      "learning_rate": 4.16073018699911e-05,
+      "loss": 0.5426,
+      "step": 627
+    },
+    {
+      "epoch": 0.7548076923076923,
+      "grad_norm": 0.36791049960194894,
+      "learning_rate": 4.158504007123776e-05,
+      "loss": 0.5279,
+      "step": 628
+    },
+    {
+      "epoch": 0.7560096153846154,
+      "grad_norm": 0.4466692587729699,
+      "learning_rate": 4.156277827248442e-05,
+      "loss": 0.5372,
+      "step": 629
+    },
+    {
+      "epoch": 0.7572115384615384,
+      "grad_norm": 0.335052582165858,
+      "learning_rate": 4.154051647373108e-05,
+      "loss": 0.5517,
+      "step": 630
+    },
+    {
+      "epoch": 0.7584134615384616,
+      "grad_norm": 0.4164462683573313,
+      "learning_rate": 4.151825467497774e-05,
+      "loss": 0.549,
+      "step": 631
+    },
+    {
+      "epoch": 0.7596153846153846,
+      "grad_norm": 0.36747905886188237,
+      "learning_rate": 4.14959928762244e-05,
+      "loss": 0.4985,
+      "step": 632
+    },
+    {
+      "epoch": 0.7608173076923077,
+      "grad_norm": 0.3408132426869167,
+      "learning_rate": 4.1473731077471064e-05,
+      "loss": 0.5192,
+      "step": 633
+    },
+    {
+      "epoch": 0.7620192307692307,
+      "grad_norm": 0.3879772007202172,
+      "learning_rate": 4.1451469278717726e-05,
+      "loss": 0.5393,
+      "step": 634
+    },
+    {
+      "epoch": 0.7632211538461539,
+      "grad_norm": 0.36873473949769653,
+      "learning_rate": 4.142920747996438e-05,
+      "loss": 0.5393,
+      "step": 635
+    },
+    {
+      "epoch": 0.7644230769230769,
+      "grad_norm": 0.32779779547955107,
+      "learning_rate": 4.140694568121104e-05,
+      "loss": 0.5139,
+      "step": 636
+    },
+    {
+      "epoch": 0.765625,
+      "grad_norm": 0.36714826280394824,
+      "learning_rate": 4.1384683882457705e-05,
+      "loss": 0.5126,
+      "step": 637
+    },
+    {
+      "epoch": 0.7668269230769231,
+      "grad_norm": 0.3652905495092366,
+      "learning_rate": 4.136242208370437e-05,
+      "loss": 0.4882,
+      "step": 638
+    },
+    {
+      "epoch": 0.7680288461538461,
+      "grad_norm": 0.385351812872727,
+      "learning_rate": 4.134016028495102e-05,
+      "loss": 0.5128,
+      "step": 639
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.4038157379259219,
+      "learning_rate": 4.1317898486197685e-05,
+      "loss": 0.5395,
+      "step": 640
+    },
+    {
+      "epoch": 0.7704326923076923,
+      "grad_norm": 0.3348256417694082,
+      "learning_rate": 4.129563668744435e-05,
+      "loss": 0.512,
+      "step": 641
+    },
+    {
+      "epoch": 0.7716346153846154,
+      "grad_norm": 0.38832519806792265,
+      "learning_rate": 4.127337488869101e-05,
+      "loss": 0.5109,
+      "step": 642
+    },
+    {
+      "epoch": 0.7728365384615384,
+      "grad_norm": 0.338297600378126,
+      "learning_rate": 4.125111308993767e-05,
+      "loss": 0.5397,
+      "step": 643
+    },
+    {
+      "epoch": 0.7740384615384616,
+      "grad_norm": 0.3304632189493596,
+      "learning_rate": 4.122885129118433e-05,
+      "loss": 0.5234,
+      "step": 644
+    },
+    {
+      "epoch": 0.7752403846153846,
+      "grad_norm": 0.3699546768654847,
+      "learning_rate": 4.120658949243099e-05,
+      "loss": 0.5341,
+      "step": 645
+    },
+    {
+      "epoch": 0.7764423076923077,
+      "grad_norm": 0.3659628532289538,
+      "learning_rate": 4.118432769367765e-05,
+      "loss": 0.5197,
+      "step": 646
+    },
+    {
+      "epoch": 0.7776442307692307,
+      "grad_norm": 0.38853672495508096,
+      "learning_rate": 4.116206589492431e-05,
+      "loss": 0.5097,
+      "step": 647
+    },
+    {
+      "epoch": 0.7788461538461539,
+      "grad_norm": 0.3452893607976395,
+      "learning_rate": 4.1139804096170975e-05,
+      "loss": 0.5098,
+      "step": 648
+    },
+    {
+      "epoch": 0.7800480769230769,
+      "grad_norm": 0.39167961727656,
+      "learning_rate": 4.111754229741763e-05,
+      "loss": 0.5325,
+      "step": 649
+    },
+    {
+      "epoch": 0.78125,
+      "grad_norm": 0.4288427557920978,
+      "learning_rate": 4.109528049866429e-05,
+      "loss": 0.5377,
+      "step": 650
+    },
+    {
+      "epoch": 0.7824519230769231,
+      "grad_norm": 0.42806238763630955,
+      "learning_rate": 4.1073018699910955e-05,
+      "loss": 0.5175,
+      "step": 651
+    },
+    {
+      "epoch": 0.7836538461538461,
+      "grad_norm": 0.3885945417418187,
+      "learning_rate": 4.105075690115762e-05,
+      "loss": 0.5272,
+      "step": 652
+    },
+    {
+      "epoch": 0.7848557692307693,
+      "grad_norm": 0.4562562701347436,
+      "learning_rate": 4.102849510240427e-05,
+      "loss": 0.5103,
+      "step": 653
+    },
+    {
+      "epoch": 0.7860576923076923,
+      "grad_norm": 0.35852797708803497,
+      "learning_rate": 4.1006233303650935e-05,
+      "loss": 0.5533,
+      "step": 654
+    },
+    {
+      "epoch": 0.7872596153846154,
+      "grad_norm": 0.4245658906633076,
+      "learning_rate": 4.0983971504897603e-05,
+      "loss": 0.5223,
+      "step": 655
+    },
+    {
+      "epoch": 0.7884615384615384,
+      "grad_norm": 0.36656047402215847,
+      "learning_rate": 4.096170970614426e-05,
+      "loss": 0.5269,
+      "step": 656
+    },
+    {
+      "epoch": 0.7896634615384616,
+      "grad_norm": 0.37078338530050986,
+      "learning_rate": 4.093944790739092e-05,
+      "loss": 0.5062,
+      "step": 657
+    },
+    {
+      "epoch": 0.7908653846153846,
+      "grad_norm": 0.3619218035263405,
+      "learning_rate": 4.0917186108637576e-05,
+      "loss": 0.4884,
+      "step": 658
+    },
+    {
+      "epoch": 0.7920673076923077,
+      "grad_norm": 0.3831988953613949,
+      "learning_rate": 4.089492430988424e-05,
+      "loss": 0.5347,
+      "step": 659
+    },
+    {
+      "epoch": 0.7932692307692307,
+      "grad_norm": 0.47066219535466974,
+      "learning_rate": 4.08726625111309e-05,
+      "loss": 0.525,
+      "step": 660
+    },
+    {
+      "epoch": 0.7944711538461539,
+      "grad_norm": 0.33518008670215227,
+      "learning_rate": 4.085040071237756e-05,
+      "loss": 0.5173,
+      "step": 661
+    },
+    {
+      "epoch": 0.7956730769230769,
+      "grad_norm": 0.4347706816377338,
+      "learning_rate": 4.0828138913624225e-05,
+      "loss": 0.5115,
+      "step": 662
+    },
+    {
+      "epoch": 0.796875,
+      "grad_norm": 0.3718505820781983,
+      "learning_rate": 4.080587711487088e-05,
+      "loss": 0.53,
+      "step": 663
+    },
+    {
+      "epoch": 0.7980769230769231,
+      "grad_norm": 0.49814786971676445,
+      "learning_rate": 4.078361531611755e-05,
+      "loss": 0.5304,
+      "step": 664
+    },
+    {
+      "epoch": 0.7992788461538461,
+      "grad_norm": 0.34779487377212837,
+      "learning_rate": 4.0761353517364204e-05,
+      "loss": 0.5311,
+      "step": 665
+    },
+    {
+      "epoch": 0.8004807692307693,
+      "grad_norm": 0.4013064011659391,
+      "learning_rate": 4.0739091718610867e-05,
+      "loss": 0.53,
+      "step": 666
+    },
+    {
+      "epoch": 0.8016826923076923,
+      "grad_norm": 0.3439332199137704,
+      "learning_rate": 4.071682991985752e-05,
+      "loss": 0.5144,
+      "step": 667
+    },
+    {
+      "epoch": 0.8028846153846154,
+      "grad_norm": 0.34905082994266945,
+      "learning_rate": 4.0694568121104184e-05,
+      "loss": 0.4848,
+      "step": 668
+    },
+    {
+      "epoch": 0.8040865384615384,
+      "grad_norm": 0.40542219513119676,
+      "learning_rate": 4.067230632235085e-05,
+      "loss": 0.4992,
+      "step": 669
+    },
+    {
+      "epoch": 0.8052884615384616,
+      "grad_norm": 0.38371103027962356,
+      "learning_rate": 4.065004452359751e-05,
+      "loss": 0.5254,
+      "step": 670
+    },
+    {
+      "epoch": 0.8064903846153846,
+      "grad_norm": 0.40756692682715107,
+      "learning_rate": 4.062778272484417e-05,
+      "loss": 0.477,
+      "step": 671
+    },
+    {
+      "epoch": 0.8076923076923077,
+      "grad_norm": 0.3856488069402091,
+      "learning_rate": 4.0605520926090826e-05,
+      "loss": 0.5299,
+      "step": 672
+    },
+    {
+      "epoch": 0.8088942307692307,
+      "grad_norm": 0.3475406542334448,
+      "learning_rate": 4.0583259127337495e-05,
+      "loss": 0.5431,
+      "step": 673
+    },
+    {
+      "epoch": 0.8100961538461539,
+      "grad_norm": 0.4276001139288367,
+      "learning_rate": 4.056099732858415e-05,
+      "loss": 0.5121,
+      "step": 674
+    },
+    {
+      "epoch": 0.8112980769230769,
+      "grad_norm": 0.3357206608680462,
+      "learning_rate": 4.053873552983081e-05,
+      "loss": 0.5116,
+      "step": 675
+    },
+    {
+      "epoch": 0.8125,
+      "grad_norm": 0.5345797978678971,
+      "learning_rate": 4.0516473731077474e-05,
+      "loss": 0.5389,
+      "step": 676
+    },
+    {
+      "epoch": 0.8137019230769231,
+      "grad_norm": 0.37620262250924885,
+      "learning_rate": 4.0494211932324136e-05,
+      "loss": 0.5265,
+      "step": 677
+    },
+    {
+      "epoch": 0.8149038461538461,
+      "grad_norm": 0.5252578111520126,
+      "learning_rate": 4.04719501335708e-05,
+      "loss": 0.535,
+      "step": 678
+    },
+    {
+      "epoch": 0.8161057692307693,
+      "grad_norm": 0.389307636856047,
+      "learning_rate": 4.0449688334817454e-05,
+      "loss": 0.5223,
+      "step": 679
+    },
+    {
+      "epoch": 0.8173076923076923,
+      "grad_norm": 0.4117055603821147,
+      "learning_rate": 4.0427426536064116e-05,
+      "loss": 0.5328,
+      "step": 680
+    },
+    {
+      "epoch": 0.8185096153846154,
+      "grad_norm": 0.4539862057015515,
+      "learning_rate": 4.040516473731077e-05,
+      "loss": 0.5058,
+      "step": 681
+    },
+    {
+      "epoch": 0.8197115384615384,
+      "grad_norm": 0.4342546094534175,
+      "learning_rate": 4.038290293855744e-05,
+      "loss": 0.5047,
+      "step": 682
+    },
+    {
+      "epoch": 0.8209134615384616,
+      "grad_norm": 0.5093495390324291,
+      "learning_rate": 4.03606411398041e-05,
+      "loss": 0.5353,
+      "step": 683
+    },
+    {
+      "epoch": 0.8221153846153846,
+      "grad_norm": 0.3289724069625257,
+      "learning_rate": 4.033837934105076e-05,
+      "loss": 0.5456,
+      "step": 684
+    },
+    {
+      "epoch": 0.8233173076923077,
+      "grad_norm": 0.49644013213119326,
+      "learning_rate": 4.031611754229742e-05,
+      "loss": 0.5098,
+      "step": 685
+    },
+    {
+      "epoch": 0.8245192307692307,
+      "grad_norm": 0.3674519654027457,
+      "learning_rate": 4.029385574354408e-05,
+      "loss": 0.5529,
+      "step": 686
+    },
+    {
+      "epoch": 0.8257211538461539,
+      "grad_norm": 0.4328214380397096,
+      "learning_rate": 4.0271593944790744e-05,
+      "loss": 0.5499,
+      "step": 687
+    },
+    {
+      "epoch": 0.8269230769230769,
+      "grad_norm": 0.3407974749337339,
+      "learning_rate": 4.02493321460374e-05,
+      "loss": 0.5264,
+      "step": 688
+    },
+    {
+      "epoch": 0.828125,
+      "grad_norm": 0.4443666245207122,
+      "learning_rate": 4.022707034728406e-05,
+      "loss": 0.5165,
+      "step": 689
+    },
+    {
+      "epoch": 0.8293269230769231,
+      "grad_norm": 0.4022428781945724,
+      "learning_rate": 4.0204808548530724e-05,
+      "loss": 0.5196,
+      "step": 690
+    },
+    {
+      "epoch": 0.8305288461538461,
+      "grad_norm": 0.39833731684691104,
+      "learning_rate": 4.0182546749777386e-05,
+      "loss": 0.5264,
+      "step": 691
+    },
+    {
+      "epoch": 0.8317307692307693,
+      "grad_norm": 0.40349762557439584,
+      "learning_rate": 4.016028495102405e-05,
+      "loss": 0.5203,
+      "step": 692
+    },
+    {
+      "epoch": 0.8329326923076923,
+      "grad_norm": 0.3521244134680743,
+      "learning_rate": 4.0138023152270703e-05,
+      "loss": 0.5417,
+      "step": 693
+    },
+    {
+      "epoch": 0.8341346153846154,
+      "grad_norm": 0.4610011863438154,
+      "learning_rate": 4.0115761353517366e-05,
+      "loss": 0.5295,
+      "step": 694
+    },
+    {
+      "epoch": 0.8353365384615384,
+      "grad_norm": 0.37090157703792526,
+      "learning_rate": 4.009349955476403e-05,
+      "loss": 0.5462,
+      "step": 695
+    },
+    {
+      "epoch": 0.8365384615384616,
+      "grad_norm": 0.4219174777904086,
+      "learning_rate": 4.007123775601069e-05,
+      "loss": 0.5126,
+      "step": 696
+    },
+    {
+      "epoch": 0.8377403846153846,
+      "grad_norm": 0.4092655961554133,
+      "learning_rate": 4.004897595725735e-05,
+      "loss": 0.5489,
+      "step": 697
+    },
+    {
+      "epoch": 0.8389423076923077,
+      "grad_norm": 0.3740725452558885,
+      "learning_rate": 4.002671415850401e-05,
+      "loss": 0.5195,
+      "step": 698
+    },
+    {
+      "epoch": 0.8401442307692307,
+      "grad_norm": 0.3143727357619497,
+      "learning_rate": 4.000445235975067e-05,
+      "loss": 0.5347,
+      "step": 699
+    },
+    {
+      "epoch": 0.8413461538461539,
+      "grad_norm": 0.41514230534776253,
+      "learning_rate": 3.998219056099733e-05,
+      "loss": 0.5115,
+      "step": 700
+    },
+    {
+      "epoch": 0.8425480769230769,
+      "grad_norm": 0.3982254480065746,
+      "learning_rate": 3.9959928762243994e-05,
+      "loss": 0.5168,
+      "step": 701
+    },
+    {
+      "epoch": 0.84375,
+      "grad_norm": 0.4222163488909361,
+      "learning_rate": 3.993766696349065e-05,
+      "loss": 0.5575,
+      "step": 702
+    },
+    {
+      "epoch": 0.8449519230769231,
+      "grad_norm": 0.3610948802850605,
+      "learning_rate": 3.991540516473731e-05,
+      "loss": 0.5179,
+      "step": 703
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.3830417115252941,
+      "learning_rate": 3.989314336598397e-05,
+      "loss": 0.5194,
+      "step": 704
+    },
+    {
+      "epoch": 0.8473557692307693,
+      "grad_norm": 0.37341494448094337,
+      "learning_rate": 3.9870881567230636e-05,
+      "loss": 0.5278,
+      "step": 705
+    },
+    {
+      "epoch": 0.8485576923076923,
+      "grad_norm": 0.3513493767865045,
+      "learning_rate": 3.98486197684773e-05,
+      "loss": 0.5286,
+      "step": 706
+    },
+    {
+      "epoch": 0.8497596153846154,
+      "grad_norm": 0.3653348195273322,
+      "learning_rate": 3.982635796972395e-05,
+      "loss": 0.5516,
+      "step": 707
+    },
+    {
+      "epoch": 0.8509615384615384,
+      "grad_norm": 0.37495634260196026,
+      "learning_rate": 3.9804096170970615e-05,
+      "loss": 0.5423,
+      "step": 708
+    },
+    {
+      "epoch": 0.8521634615384616,
+      "grad_norm": 0.38174641750536087,
+      "learning_rate": 3.978183437221728e-05,
+      "loss": 0.5103,
+      "step": 709
+    },
+    {
+      "epoch": 0.8533653846153846,
+      "grad_norm": 0.37199340464173314,
+      "learning_rate": 3.975957257346394e-05,
+      "loss": 0.5089,
+      "step": 710
+    },
+    {
+      "epoch": 0.8545673076923077,
+      "grad_norm": 0.3305559138077109,
+      "learning_rate": 3.97373107747106e-05,
+      "loss": 0.5188,
+      "step": 711
+    },
+    {
+      "epoch": 0.8557692307692307,
+      "grad_norm": 0.3654169130250703,
+      "learning_rate": 3.971504897595726e-05,
+      "loss": 0.5377,
+      "step": 712
+    },
+    {
+      "epoch": 0.8569711538461539,
+      "grad_norm": 0.4086183552056209,
+      "learning_rate": 3.969278717720392e-05,
+      "loss": 0.5177,
+      "step": 713
+    },
+    {
+      "epoch": 0.8581730769230769,
+      "grad_norm": 0.3083758220253097,
+      "learning_rate": 3.967052537845058e-05,
+      "loss": 0.499,
+      "step": 714
+    },
+    {
+      "epoch": 0.859375,
+      "grad_norm": 0.42679327947661094,
+      "learning_rate": 3.964826357969724e-05,
+      "loss": 0.4998,
+      "step": 715
+    },
+    {
+      "epoch": 0.8605769230769231,
+      "grad_norm": 0.5392468571787924,
+      "learning_rate": 3.96260017809439e-05,
+      "loss": 0.5298,
+      "step": 716
+    },
+    {
+      "epoch": 0.8617788461538461,
+      "grad_norm": 0.44877291282615833,
+      "learning_rate": 3.960373998219056e-05,
+      "loss": 0.5351,
+      "step": 717
+    },
+    {
+      "epoch": 0.8629807692307693,
+      "grad_norm": 0.4073959081140446,
+      "learning_rate": 3.958147818343722e-05,
+      "loss": 0.5289,
+      "step": 718
+    },
+    {
+      "epoch": 0.8641826923076923,
+      "grad_norm": 0.44120298950786146,
+      "learning_rate": 3.9559216384683885e-05,
+      "loss": 0.5317,
+      "step": 719
+    },
+    {
+      "epoch": 0.8653846153846154,
+      "grad_norm": 0.32869324322309257,
+      "learning_rate": 3.953695458593055e-05,
+      "loss": 0.5255,
+      "step": 720
+    },
+    {
+      "epoch": 0.8665865384615384,
+      "grad_norm": 0.4382358919668934,
+      "learning_rate": 3.95146927871772e-05,
+      "loss": 0.5234,
+      "step": 721
+    },
+    {
+      "epoch": 0.8677884615384616,
+      "grad_norm": 0.4021258627831412,
+      "learning_rate": 3.9492430988423865e-05,
+      "loss": 0.5366,
+      "step": 722
+    },
+    {
+      "epoch": 0.8689903846153846,
+      "grad_norm": 0.4173725511711841,
+      "learning_rate": 3.947016918967053e-05,
+      "loss": 0.5296,
+      "step": 723
+    },
+    {
+      "epoch": 0.8701923076923077,
+      "grad_norm": 0.8403404016329585,
+      "learning_rate": 3.944790739091719e-05,
+      "loss": 0.5335,
+      "step": 724
+    },
+    {
+      "epoch": 0.8713942307692307,
+      "grad_norm": 0.3995268630213925,
+      "learning_rate": 3.942564559216385e-05,
+      "loss": 0.5451,
+      "step": 725
+    },
+    {
+      "epoch": 0.8725961538461539,
+      "grad_norm": 0.4049491981625935,
+      "learning_rate": 3.9403383793410506e-05,
+      "loss": 0.5365,
+      "step": 726
+    },
+    {
+      "epoch": 0.8737980769230769,
+      "grad_norm": 0.39798131415569543,
+      "learning_rate": 3.938112199465717e-05,
+      "loss": 0.5024,
+      "step": 727
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 0.4176743476379908,
+      "learning_rate": 3.935886019590383e-05,
+      "loss": 0.5351,
+      "step": 728
+    },
+    {
+      "epoch": 0.8762019230769231,
+      "grad_norm": 0.33512656691592035,
+      "learning_rate": 3.933659839715049e-05,
+      "loss": 0.537,
+      "step": 729
+    },
+    {
+      "epoch": 0.8774038461538461,
+      "grad_norm": 0.3370593854309884,
+      "learning_rate": 3.931433659839715e-05,
+      "loss": 0.5239,
+      "step": 730
+    },
+    {
+      "epoch": 0.8786057692307693,
+      "grad_norm": 0.37008462561565436,
+      "learning_rate": 3.929207479964381e-05,
+      "loss": 0.5561,
+      "step": 731
+    },
+    {
+      "epoch": 0.8798076923076923,
+      "grad_norm": 0.31853434776932765,
+      "learning_rate": 3.926981300089048e-05,
+      "loss": 0.5104,
+      "step": 732
+    },
+    {
+      "epoch": 0.8810096153846154,
+      "grad_norm": 0.5554555272408357,
+      "learning_rate": 3.9247551202137135e-05,
+      "loss": 0.5273,
+      "step": 733
+    },
+    {
+      "epoch": 0.8822115384615384,
+      "grad_norm": 0.3485657089475844,
+      "learning_rate": 3.92252894033838e-05,
+      "loss": 0.5418,
+      "step": 734
+    },
+    {
+      "epoch": 0.8834134615384616,
+      "grad_norm": 0.4417397042004106,
+      "learning_rate": 3.920302760463045e-05,
+      "loss": 0.5287,
+      "step": 735
+    },
+    {
+      "epoch": 0.8846153846153846,
+      "grad_norm": 0.37815288727778207,
+      "learning_rate": 3.918076580587712e-05,
+      "loss": 0.5081,
+      "step": 736
+    },
+    {
+      "epoch": 0.8858173076923077,
+      "grad_norm": 0.3559243438258194,
+      "learning_rate": 3.9158504007123776e-05,
+      "loss": 0.5253,
+      "step": 737
+    },
+    {
+      "epoch": 0.8870192307692307,
+      "grad_norm": 0.4011548835080678,
+      "learning_rate": 3.913624220837044e-05,
+      "loss": 0.5333,
+      "step": 738
+    },
+    {
+      "epoch": 0.8882211538461539,
+      "grad_norm": 0.38785663169621254,
+      "learning_rate": 3.91139804096171e-05,
+      "loss": 0.5202,
+      "step": 739
+    },
+    {
+      "epoch": 0.8894230769230769,
+      "grad_norm": 0.3517670367143401,
+      "learning_rate": 3.9091718610863756e-05,
+      "loss": 0.4953,
+      "step": 740
+    },
+    {
+      "epoch": 0.890625,
+      "grad_norm": 0.38860182277917654,
+      "learning_rate": 3.9069456812110425e-05,
+      "loss": 0.5057,
+      "step": 741
+    },
+    {
+      "epoch": 0.8918269230769231,
+      "grad_norm": 0.405701428722561,
+      "learning_rate": 3.904719501335708e-05,
+      "loss": 0.5599,
+      "step": 742
+    },
+    {
+      "epoch": 0.8930288461538461,
+      "grad_norm": 0.39030189571804186,
+      "learning_rate": 3.902493321460374e-05,
+      "loss": 0.5313,
+      "step": 743
+    },
+    {
+      "epoch": 0.8942307692307693,
+      "grad_norm": 0.48792364894116297,
+      "learning_rate": 3.90026714158504e-05,
+      "loss": 0.5396,
+      "step": 744
+    },
+    {
+      "epoch": 0.8954326923076923,
+      "grad_norm": 0.39506944907825736,
+      "learning_rate": 3.8980409617097067e-05,
+      "loss": 0.5365,
+      "step": 745
+    },
+    {
+      "epoch": 0.8966346153846154,
+      "grad_norm": 0.37649553967753563,
+      "learning_rate": 3.895814781834373e-05,
+      "loss": 0.4908,
+      "step": 746
+    },
+    {
+      "epoch": 0.8978365384615384,
+      "grad_norm": 0.35630941198536686,
+      "learning_rate": 3.8935886019590384e-05,
+      "loss": 0.5188,
+      "step": 747
+    },
+    {
+      "epoch": 0.8990384615384616,
+      "grad_norm": 0.3463079914804103,
+      "learning_rate": 3.8913624220837046e-05,
+      "loss": 0.5358,
+      "step": 748
+    },
+    {
+      "epoch": 0.9002403846153846,
+      "grad_norm": 0.3544561366919928,
+      "learning_rate": 3.88913624220837e-05,
+      "loss": 0.5066,
+      "step": 749
+    },
+    {
+      "epoch": 0.9014423076923077,
+      "grad_norm": 0.3077812999510648,
+      "learning_rate": 3.886910062333037e-05,
+      "loss": 0.5232,
+      "step": 750
+    },
+    {
+      "epoch": 0.9026442307692307,
+      "grad_norm": 0.33776361617663325,
+      "learning_rate": 3.8846838824577026e-05,
+      "loss": 0.5046,
+      "step": 751
+    },
+    {
+      "epoch": 0.9038461538461539,
+      "grad_norm": 0.3741576206202815,
+      "learning_rate": 3.882457702582369e-05,
+      "loss": 0.5561,
+      "step": 752
+    },
+    {
+      "epoch": 0.9050480769230769,
+      "grad_norm": 0.3505110661515964,
+      "learning_rate": 3.880231522707035e-05,
+      "loss": 0.511,
+      "step": 753
+    },
+    {
+      "epoch": 0.90625,
+      "grad_norm": 0.35529760635596114,
+      "learning_rate": 3.878005342831701e-05,
+      "loss": 0.5478,
+      "step": 754
+    },
+    {
+      "epoch": 0.9074519230769231,
+      "grad_norm": 0.335804276849097,
+      "learning_rate": 3.8757791629563674e-05,
+      "loss": 0.5069,
+      "step": 755
+    },
+    {
+      "epoch": 0.9086538461538461,
+      "grad_norm": 0.39854868998445375,
+      "learning_rate": 3.873552983081033e-05,
+      "loss": 0.5483,
+      "step": 756
+    },
+    {
+      "epoch": 0.9098557692307693,
+      "grad_norm": 0.37348285118930624,
+      "learning_rate": 3.871326803205699e-05,
+      "loss": 0.532,
+      "step": 757
+    },
+    {
+      "epoch": 0.9110576923076923,
+      "grad_norm": 0.3871406571346443,
+      "learning_rate": 3.869100623330365e-05,
+      "loss": 0.523,
+      "step": 758
+    },
+    {
+      "epoch": 0.9122596153846154,
+      "grad_norm": 0.33204941485570666,
+      "learning_rate": 3.8668744434550316e-05,
+      "loss": 0.5193,
+      "step": 759
+    },
+    {
+      "epoch": 0.9134615384615384,
+      "grad_norm": 0.37805191902105734,
+      "learning_rate": 3.864648263579698e-05,
+      "loss": 0.5009,
+      "step": 760
+    },
+    {
+      "epoch": 0.9146634615384616,
+      "grad_norm": 0.33615079366938816,
+      "learning_rate": 3.8624220837043634e-05,
+      "loss": 0.527,
+      "step": 761
+    },
+    {
+      "epoch": 0.9158653846153846,
+      "grad_norm": 0.3720985786900068,
+      "learning_rate": 3.8601959038290296e-05,
+      "loss": 0.5593,
+      "step": 762
+    },
+    {
+      "epoch": 0.9170673076923077,
+      "grad_norm": 0.3410412243544018,
+      "learning_rate": 3.857969723953696e-05,
+      "loss": 0.5202,
+      "step": 763
+    },
+    {
+      "epoch": 0.9182692307692307,
+      "grad_norm": 0.35690845467401394,
+      "learning_rate": 3.855743544078362e-05,
+      "loss": 0.5352,
+      "step": 764
+    },
+    {
+      "epoch": 0.9194711538461539,
+      "grad_norm": 0.34756033849841683,
+      "learning_rate": 3.8535173642030275e-05,
+      "loss": 0.4993,
+      "step": 765
+    },
+    {
+      "epoch": 0.9206730769230769,
+      "grad_norm": 0.3579257631271237,
+      "learning_rate": 3.851291184327694e-05,
+      "loss": 0.5511,
+      "step": 766
+    },
+    {
+      "epoch": 0.921875,
+      "grad_norm": 0.359159361830089,
+      "learning_rate": 3.84906500445236e-05,
+      "loss": 0.5257,
+      "step": 767
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.33669192952638,
+      "learning_rate": 3.846838824577026e-05,
+      "loss": 0.5088,
+      "step": 768
+    },
+    {
+      "epoch": 0.9242788461538461,
+      "grad_norm": 0.4042152802601286,
+      "learning_rate": 3.8446126447016924e-05,
+      "loss": 0.5066,
+      "step": 769
+    },
+    {
+      "epoch": 0.9254807692307693,
+      "grad_norm": 0.30463260533034053,
+      "learning_rate": 3.842386464826358e-05,
+      "loss": 0.505,
+      "step": 770
+    },
+    {
+      "epoch": 0.9266826923076923,
+      "grad_norm": 0.38364801307271956,
+      "learning_rate": 3.840160284951024e-05,
+      "loss": 0.5509,
+      "step": 771
+    },
+    {
+      "epoch": 0.9278846153846154,
+      "grad_norm": 0.34951258791972817,
+      "learning_rate": 3.8379341050756903e-05,
+      "loss": 0.4842,
+      "step": 772
+    },
+    {
+      "epoch": 0.9290865384615384,
+      "grad_norm": 0.4179672311198812,
+      "learning_rate": 3.8357079252003566e-05,
+      "loss": 0.5365,
+      "step": 773
+    },
+    {
+      "epoch": 0.9302884615384616,
+      "grad_norm": 0.3518912232734849,
+      "learning_rate": 3.833481745325023e-05,
+      "loss": 0.5701,
+      "step": 774
+    },
+    {
+      "epoch": 0.9314903846153846,
+      "grad_norm": 0.3814574868321844,
+      "learning_rate": 3.831255565449688e-05,
+      "loss": 0.5238,
+      "step": 775
+    },
+    {
+      "epoch": 0.9326923076923077,
+      "grad_norm": 0.35419066924371756,
+      "learning_rate": 3.8290293855743545e-05,
+      "loss": 0.497,
+      "step": 776
+    },
+    {
+      "epoch": 0.9338942307692307,
+      "grad_norm": 0.3768737749895518,
+      "learning_rate": 3.826803205699021e-05,
+      "loss": 0.5471,
+      "step": 777
+    },
+    {
+      "epoch": 0.9350961538461539,
+      "grad_norm": 0.47429684750497564,
+      "learning_rate": 3.824577025823687e-05,
+      "loss": 0.5604,
+      "step": 778
+    },
+    {
+      "epoch": 0.9362980769230769,
+      "grad_norm": 0.3194717433100731,
+      "learning_rate": 3.8223508459483525e-05,
+      "loss": 0.499,
+      "step": 779
+    },
+    {
+      "epoch": 0.9375,
+      "grad_norm": 0.46308775245196565,
+      "learning_rate": 3.820124666073019e-05,
+      "loss": 0.5147,
+      "step": 780
+    },
+    {
+      "epoch": 0.9387019230769231,
+      "grad_norm": 0.4402221122462486,
+      "learning_rate": 3.817898486197685e-05,
+      "loss": 0.5016,
+      "step": 781
+    },
+    {
+      "epoch": 0.9399038461538461,
+      "grad_norm": 0.4298190134180781,
+      "learning_rate": 3.815672306322351e-05,
+      "loss": 0.5172,
+      "step": 782
+    },
+    {
+      "epoch": 0.9411057692307693,
+      "grad_norm": 0.45860202192516886,
+      "learning_rate": 3.813446126447017e-05,
+      "loss": 0.5466,
+      "step": 783
+    },
+    {
+      "epoch": 0.9423076923076923,
+      "grad_norm": 0.3346556775742285,
+      "learning_rate": 3.811219946571683e-05,
+      "loss": 0.5221,
+      "step": 784
+    },
+    {
+      "epoch": 0.9435096153846154,
+      "grad_norm": 0.5561442989575147,
+      "learning_rate": 3.808993766696349e-05,
+      "loss": 0.5047,
+      "step": 785
+    },
+    {
+      "epoch": 0.9447115384615384,
+      "grad_norm": 0.3890101027979219,
+      "learning_rate": 3.806767586821015e-05,
+      "loss": 0.497,
+      "step": 786
+    },
+    {
+      "epoch": 0.9459134615384616,
+      "grad_norm": 0.4385600903682313,
+      "learning_rate": 3.8045414069456815e-05,
+      "loss": 0.4619,
+      "step": 787
+    },
+    {
+      "epoch": 0.9471153846153846,
+      "grad_norm": 0.37978901396509523,
+      "learning_rate": 3.802315227070348e-05,
+      "loss": 0.49,
+      "step": 788
+    },
+    {
+      "epoch": 0.9483173076923077,
+      "grad_norm": 0.3664204512227805,
+      "learning_rate": 3.800089047195013e-05,
+      "loss": 0.4946,
+      "step": 789
+    },
+    {
+      "epoch": 0.9495192307692307,
+      "grad_norm": 0.4382501268201123,
+      "learning_rate": 3.7978628673196795e-05,
+      "loss": 0.5024,
+      "step": 790
+    },
+    {
+      "epoch": 0.9507211538461539,
+      "grad_norm": 0.3799526413186672,
+      "learning_rate": 3.795636687444346e-05,
+      "loss": 0.512,
+      "step": 791
+    },
+    {
+      "epoch": 0.9519230769230769,
+      "grad_norm": 0.3994327651817104,
+      "learning_rate": 3.793410507569012e-05,
+      "loss": 0.4968,
+      "step": 792
+    },
+    {
+      "epoch": 0.953125,
+      "grad_norm": 0.4212696715593166,
+      "learning_rate": 3.7911843276936774e-05,
+      "loss": 0.5338,
+      "step": 793
+    },
+    {
+      "epoch": 0.9543269230769231,
+      "grad_norm": 0.4248130631599435,
+      "learning_rate": 3.7889581478183437e-05,
+      "loss": 0.5231,
+      "step": 794
+    },
+    {
+      "epoch": 0.9555288461538461,
+      "grad_norm": 0.3918492577639015,
+      "learning_rate": 3.7867319679430105e-05,
+      "loss": 0.5148,
+      "step": 795
+    },
+    {
+      "epoch": 0.9567307692307693,
+      "grad_norm": 0.4402386025584822,
+      "learning_rate": 3.784505788067676e-05,
+      "loss": 0.5032,
+      "step": 796
+    },
+    {
+      "epoch": 0.9579326923076923,
+      "grad_norm": 0.35325258473225907,
+      "learning_rate": 3.782279608192342e-05,
+      "loss": 0.4887,
+      "step": 797
+    },
+    {
+      "epoch": 0.9591346153846154,
+      "grad_norm": 0.37246807003400667,
+      "learning_rate": 3.780053428317008e-05,
+      "loss": 0.5127,
+      "step": 798
+    },
+    {
+      "epoch": 0.9603365384615384,
+      "grad_norm": 0.3869100382516642,
+      "learning_rate": 3.777827248441674e-05,
+      "loss": 0.5286,
+      "step": 799
+    },
+    {
+      "epoch": 0.9615384615384616,
+      "grad_norm": 0.38603908293442685,
+      "learning_rate": 3.77560106856634e-05,
+      "loss": 0.4869,
+      "step": 800
+    },
+    {
+      "epoch": 0.9627403846153846,
+      "grad_norm": 0.35516366012854106,
+      "learning_rate": 3.7733748886910065e-05,
+      "loss": 0.5123,
+      "step": 801
+    },
+    {
+      "epoch": 0.9639423076923077,
+      "grad_norm": 0.42087444672419183,
+      "learning_rate": 3.771148708815673e-05,
+      "loss": 0.5187,
+      "step": 802
+    },
+    {
+      "epoch": 0.9651442307692307,
+      "grad_norm": 0.45764940940436716,
+      "learning_rate": 3.768922528940338e-05,
+      "loss": 0.5413,
+      "step": 803
+    },
+    {
+      "epoch": 0.9663461538461539,
+      "grad_norm": 0.34752129392444486,
+      "learning_rate": 3.766696349065005e-05,
+      "loss": 0.4912,
+      "step": 804
+    },
+    {
+      "epoch": 0.9675480769230769,
+      "grad_norm": 0.4412505224339214,
+      "learning_rate": 3.7644701691896706e-05,
+      "loss": 0.4694,
+      "step": 805
+    },
+    {
+      "epoch": 0.96875,
+      "grad_norm": 0.4055305117262795,
+      "learning_rate": 3.762243989314337e-05,
+      "loss": 0.5458,
+      "step": 806
+    },
+    {
+      "epoch": 0.9699519230769231,
+      "grad_norm": 0.37299101964757325,
+      "learning_rate": 3.7600178094390024e-05,
+      "loss": 0.5166,
+      "step": 807
+    },
+    {
+      "epoch": 0.9711538461538461,
+      "grad_norm": 0.4873512038778861,
+      "learning_rate": 3.7577916295636686e-05,
+      "loss": 0.5301,
+      "step": 808
+    },
+    {
+      "epoch": 0.9723557692307693,
+      "grad_norm": 0.44087818176399846,
+      "learning_rate": 3.7555654496883355e-05,
+      "loss": 0.5132,
+      "step": 809
+    },
+    {
+      "epoch": 0.9735576923076923,
+      "grad_norm": 0.47727134211676786,
+      "learning_rate": 3.753339269813001e-05,
+      "loss": 0.5101,
+      "step": 810
+    },
+    {
+      "epoch": 0.9747596153846154,
+      "grad_norm": 0.5797091379020176,
+      "learning_rate": 3.751113089937667e-05,
+      "loss": 0.5187,
+      "step": 811
+    },
+    {
+      "epoch": 0.9759615384615384,
+      "grad_norm": 0.5097078781199755,
+      "learning_rate": 3.748886910062333e-05,
+      "loss": 0.5317,
+      "step": 812
+    },
+    {
+      "epoch": 0.9771634615384616,
+      "grad_norm": 0.42069930775895914,
+      "learning_rate": 3.746660730187e-05,
+      "loss": 0.5183,
+      "step": 813
+    },
+    {
+      "epoch": 0.9783653846153846,
+      "grad_norm": 0.42137623251050255,
+      "learning_rate": 3.744434550311665e-05,
+      "loss": 0.529,
+      "step": 814
+    },
+    {
+      "epoch": 0.9795673076923077,
+      "grad_norm": 0.4185440478578161,
+      "learning_rate": 3.7422083704363314e-05,
+      "loss": 0.5195,
+      "step": 815
+    },
+    {
+      "epoch": 0.9807692307692307,
+      "grad_norm": 0.38426383138686193,
+      "learning_rate": 3.7399821905609976e-05,
+      "loss": 0.5344,
+      "step": 816
+    },
+    {
+      "epoch": 0.9819711538461539,
+      "grad_norm": 0.5232104711243898,
+      "learning_rate": 3.737756010685663e-05,
+      "loss": 0.5314,
+      "step": 817
+    },
+    {
+      "epoch": 0.9831730769230769,
+      "grad_norm": 0.34036347561933294,
+      "learning_rate": 3.73552983081033e-05,
+      "loss": 0.5231,
+      "step": 818
+    },
+    {
+      "epoch": 0.984375,
+      "grad_norm": 0.45936446349141985,
+      "learning_rate": 3.7333036509349956e-05,
+      "loss": 0.5038,
+      "step": 819
+    },
+    {
+      "epoch": 0.9855769230769231,
+      "grad_norm": 0.35865239906487645,
+      "learning_rate": 3.731077471059662e-05,
+      "loss": 0.5285,
+      "step": 820
+    },
+    {
+      "epoch": 0.9867788461538461,
+      "grad_norm": 0.4209801626728313,
+      "learning_rate": 3.7288512911843273e-05,
+      "loss": 0.4847,
+      "step": 821
+    },
+    {
+      "epoch": 0.9879807692307693,
+      "grad_norm": 0.38745902044053215,
+      "learning_rate": 3.726625111308994e-05,
+      "loss": 0.4999,
+      "step": 822
+    },
+    {
+      "epoch": 0.9891826923076923,
+      "grad_norm": 0.31574782997188966,
+      "learning_rate": 3.7243989314336604e-05,
+      "loss": 0.4878,
+      "step": 823
+    },
+    {
+      "epoch": 0.9903846153846154,
+      "grad_norm": 0.40864850666466096,
+      "learning_rate": 3.722172751558326e-05,
+      "loss": 0.5025,
+      "step": 824
+    },
+    {
+      "epoch": 0.9915865384615384,
+      "grad_norm": 0.3214506135897434,
+      "learning_rate": 3.719946571682992e-05,
+      "loss": 0.4876,
+      "step": 825
+    },
+    {
+      "epoch": 0.9927884615384616,
+      "grad_norm": 0.3710711398185045,
+      "learning_rate": 3.7177203918076584e-05,
+      "loss": 0.5029,
+      "step": 826
+    },
+    {
+      "epoch": 0.9939903846153846,
+      "grad_norm": 0.34872343128863564,
+      "learning_rate": 3.7154942119323246e-05,
+      "loss": 0.4839,
+      "step": 827
+    },
+    {
+      "epoch": 0.9951923076923077,
+      "grad_norm": 0.2932771549371605,
+      "learning_rate": 3.71326803205699e-05,
+      "loss": 0.5025,
+      "step": 828
+    },
+    {
+      "epoch": 0.9963942307692307,
+      "grad_norm": 0.46121242485294556,
+      "learning_rate": 3.7110418521816564e-05,
+      "loss": 0.5276,
+      "step": 829
+    },
+    {
+      "epoch": 0.9975961538461539,
+      "grad_norm": 0.32318314781631596,
+      "learning_rate": 3.7088156723063226e-05,
+      "loss": 0.5003,
+      "step": 830
+    },
+    {
+      "epoch": 0.9987980769230769,
+      "grad_norm": 0.4853538995611778,
+      "learning_rate": 3.706589492430989e-05,
+      "loss": 0.494,
+      "step": 831
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.386410133614722,
+      "learning_rate": 3.704363312555655e-05,
+      "loss": 0.5411,
+      "step": 832
+    },
+    {
+      "epoch": 1.0012019230769231,
+      "grad_norm": 0.41312143727163403,
+      "learning_rate": 3.7021371326803205e-05,
+      "loss": 0.4483,
+      "step": 833
+    },
+    {
+      "epoch": 1.0024038461538463,
+      "grad_norm": 0.34924424125765363,
+      "learning_rate": 3.699910952804987e-05,
+      "loss": 0.4621,
+      "step": 834
+    },
+    {
+      "epoch": 1.0036057692307692,
+      "grad_norm": 0.33629205022022407,
+      "learning_rate": 3.697684772929653e-05,
+      "loss": 0.4312,
+      "step": 835
+    },
+    {
+      "epoch": 1.0048076923076923,
+      "grad_norm": 0.41986103616706294,
+      "learning_rate": 3.695458593054319e-05,
+      "loss": 0.4635,
+      "step": 836
+    },
+    {
+      "epoch": 1.0060096153846154,
+      "grad_norm": 0.2932160485485753,
+      "learning_rate": 3.6932324131789854e-05,
+      "loss": 0.4222,
+      "step": 837
+    },
+    {
+      "epoch": 1.0072115384615385,
+      "grad_norm": 0.3664862780600867,
+      "learning_rate": 3.691006233303651e-05,
+      "loss": 0.4358,
+      "step": 838
+    },
+    {
+      "epoch": 1.0084134615384615,
+      "grad_norm": 0.3169580305667028,
+      "learning_rate": 3.688780053428317e-05,
+      "loss": 0.435,
+      "step": 839
+    },
+    {
+      "epoch": 1.0096153846153846,
+      "grad_norm": 0.3598943307310428,
+      "learning_rate": 3.6865538735529834e-05,
+      "loss": 0.4545,
+      "step": 840
+    },
+    {
+      "epoch": 1.0108173076923077,
+      "grad_norm": 0.33297239649025945,
+      "learning_rate": 3.6843276936776496e-05,
+      "loss": 0.409,
+      "step": 841
+    },
+    {
+      "epoch": 1.0120192307692308,
+      "grad_norm": 0.37035374755763606,
+      "learning_rate": 3.682101513802315e-05,
+      "loss": 0.4655,
+      "step": 842
+    },
+    {
+      "epoch": 1.0132211538461537,
+      "grad_norm": 0.33616931731470284,
+      "learning_rate": 3.679875333926981e-05,
+      "loss": 0.4005,
+      "step": 843
+    },
+    {
+      "epoch": 1.0144230769230769,
+      "grad_norm": 0.37810871798000534,
+      "learning_rate": 3.6776491540516475e-05,
+      "loss": 0.4725,
+      "step": 844
+    },
+    {
+      "epoch": 1.015625,
+      "grad_norm": 0.35240476255138864,
+      "learning_rate": 3.675422974176314e-05,
+      "loss": 0.4208,
+      "step": 845
+    },
+    {
+      "epoch": 1.0168269230769231,
+      "grad_norm": 0.37597671986037157,
+      "learning_rate": 3.67319679430098e-05,
+      "loss": 0.4723,
+      "step": 846
+    },
+    {
+      "epoch": 1.0180288461538463,
+      "grad_norm": 0.32684387468599657,
+      "learning_rate": 3.6709706144256455e-05,
+      "loss": 0.4382,
+      "step": 847
+    },
+    {
+      "epoch": 1.0192307692307692,
+      "grad_norm": 0.33423900272738355,
+      "learning_rate": 3.668744434550312e-05,
+      "loss": 0.4603,
+      "step": 848
+    },
+    {
+      "epoch": 1.0204326923076923,
+      "grad_norm": 0.4014514965890662,
+      "learning_rate": 3.666518254674978e-05,
+      "loss": 0.4363,
+      "step": 849
+    },
+    {
+      "epoch": 1.0216346153846154,
+      "grad_norm": 0.33013753921061795,
+      "learning_rate": 3.664292074799644e-05,
+      "loss": 0.4196,
+      "step": 850
+    },
+    {
+      "epoch": 1.0228365384615385,
+      "grad_norm": 0.39196570864036795,
+      "learning_rate": 3.6620658949243103e-05,
+      "loss": 0.4306,
+      "step": 851
+    },
+    {
+      "epoch": 1.0240384615384615,
+      "grad_norm": 0.31873555693611527,
+      "learning_rate": 3.659839715048976e-05,
+      "loss": 0.4528,
+      "step": 852
+    },
+    {
+      "epoch": 1.0252403846153846,
+      "grad_norm": 0.4973135404681555,
+      "learning_rate": 3.657613535173642e-05,
+      "loss": 0.4717,
+      "step": 853
+    },
+    {
+      "epoch": 1.0264423076923077,
+      "grad_norm": 0.34671871920315095,
+      "learning_rate": 3.655387355298308e-05,
+      "loss": 0.4326,
+      "step": 854
+    },
+    {
+      "epoch": 1.0276442307692308,
+      "grad_norm": 0.4104750810611698,
+      "learning_rate": 3.6531611754229745e-05,
+      "loss": 0.4207,
+      "step": 855
+    },
+    {
+      "epoch": 1.0288461538461537,
+      "grad_norm": 0.3666616101448268,
+      "learning_rate": 3.65093499554764e-05,
+      "loss": 0.4522,
+      "step": 856
+    },
+    {
+      "epoch": 1.0300480769230769,
+      "grad_norm": 0.3994520142805233,
+      "learning_rate": 3.648708815672306e-05,
+      "loss": 0.471,
+      "step": 857
+    },
+    {
+      "epoch": 1.03125,
+      "grad_norm": 0.3816652462981472,
+      "learning_rate": 3.6464826357969725e-05,
+      "loss": 0.4484,
+      "step": 858
+    },
+    {
+      "epoch": 1.0324519230769231,
+      "grad_norm": 0.3689252375232736,
+      "learning_rate": 3.644256455921639e-05,
+      "loss": 0.4293,
+      "step": 859
+    },
+    {
+      "epoch": 1.0336538461538463,
+      "grad_norm": 0.3443155488065426,
+      "learning_rate": 3.642030276046305e-05,
+      "loss": 0.4409,
+      "step": 860
+    },
+    {
+      "epoch": 1.0348557692307692,
+      "grad_norm": 0.3170875073851732,
+      "learning_rate": 3.6398040961709704e-05,
+      "loss": 0.4306,
+      "step": 861
+    },
+    {
+      "epoch": 1.0360576923076923,
+      "grad_norm": 0.4208557072658073,
+      "learning_rate": 3.637577916295637e-05,
+      "loss": 0.4193,
+      "step": 862
+    },
+    {
+      "epoch": 1.0372596153846154,
+      "grad_norm": 0.3006275017442806,
+      "learning_rate": 3.635351736420303e-05,
+      "loss": 0.4186,
+      "step": 863
+    },
+    {
+      "epoch": 1.0384615384615385,
+      "grad_norm": 0.3883833332537646,
+      "learning_rate": 3.633125556544969e-05,
+      "loss": 0.4315,
+      "step": 864
+    },
+    {
+      "epoch": 1.0396634615384615,
+      "grad_norm": 0.3081478258343686,
+      "learning_rate": 3.630899376669635e-05,
+      "loss": 0.4379,
+      "step": 865
+    },
+    {
+      "epoch": 1.0408653846153846,
+      "grad_norm": 0.297708335300336,
+      "learning_rate": 3.628673196794301e-05,
+      "loss": 0.4481,
+      "step": 866
+    },
+    {
+      "epoch": 1.0420673076923077,
+      "grad_norm": 0.34326342630604234,
+      "learning_rate": 3.626447016918967e-05,
+      "loss": 0.4089,
+      "step": 867
+    },
+    {
+      "epoch": 1.0432692307692308,
+      "grad_norm": 0.33836110220653826,
+      "learning_rate": 3.624220837043633e-05,
+      "loss": 0.4675,
+      "step": 868
+    },
+    {
+      "epoch": 1.0444711538461537,
+      "grad_norm": 0.3677699893025392,
+      "learning_rate": 3.6219946571682995e-05,
+      "loss": 0.4166,
+      "step": 869
+    },
+    {
+      "epoch": 1.0456730769230769,
+      "grad_norm": 0.3131385594134888,
+      "learning_rate": 3.619768477292965e-05,
+      "loss": 0.4289,
+      "step": 870
+    },
+    {
+      "epoch": 1.046875,
+      "grad_norm": 0.32145761692528835,
+      "learning_rate": 3.617542297417631e-05,
+      "loss": 0.4378,
+      "step": 871
+    },
+    {
+      "epoch": 1.0480769230769231,
+      "grad_norm": 0.3323263906280892,
+      "learning_rate": 3.615316117542298e-05,
+      "loss": 0.4056,
+      "step": 872
+    },
+    {
+      "epoch": 1.0492788461538463,
+      "grad_norm": 0.3197905207319408,
+      "learning_rate": 3.6130899376669637e-05,
+      "loss": 0.3967,
+      "step": 873
+    },
+    {
+      "epoch": 1.0504807692307692,
+      "grad_norm": 0.3202972014905312,
+      "learning_rate": 3.61086375779163e-05,
+      "loss": 0.4293,
+      "step": 874
+    },
+    {
+      "epoch": 1.0516826923076923,
+      "grad_norm": 0.3222895073025427,
+      "learning_rate": 3.6086375779162954e-05,
+      "loss": 0.4137,
+      "step": 875
+    },
+    {
+      "epoch": 1.0528846153846154,
+      "grad_norm": 0.41840165617403524,
+      "learning_rate": 3.6064113980409616e-05,
+      "loss": 0.4759,
+      "step": 876
+    },
+    {
+      "epoch": 1.0540865384615385,
+      "grad_norm": 0.32025279309380555,
+      "learning_rate": 3.604185218165628e-05,
+      "loss": 0.4492,
+      "step": 877
+    },
+    {
+      "epoch": 1.0552884615384615,
+      "grad_norm": 0.3838437105863076,
+      "learning_rate": 3.601959038290294e-05,
+      "loss": 0.4496,
+      "step": 878
+    },
+    {
+      "epoch": 1.0564903846153846,
+      "grad_norm": 0.3586379751590531,
+      "learning_rate": 3.59973285841496e-05,
+      "loss": 0.4151,
+      "step": 879
+    },
+    {
+      "epoch": 1.0576923076923077,
+      "grad_norm": 0.3653565404633259,
+      "learning_rate": 3.597506678539626e-05,
+      "loss": 0.4303,
+      "step": 880
+    },
+    {
+      "epoch": 1.0588942307692308,
+      "grad_norm": 0.3565563798484146,
+      "learning_rate": 3.595280498664293e-05,
+      "loss": 0.4538,
+      "step": 881
+    },
+    {
+      "epoch": 1.0600961538461537,
+      "grad_norm": 0.3575439819706044,
+      "learning_rate": 3.593054318788958e-05,
+      "loss": 0.4282,
+      "step": 882
+    },
+    {
+      "epoch": 1.0612980769230769,
+      "grad_norm": 0.3161017695287641,
+      "learning_rate": 3.5908281389136244e-05,
+      "loss": 0.4504,
+      "step": 883
+    },
+    {
+      "epoch": 1.0625,
+      "grad_norm": 0.3369267911320925,
+      "learning_rate": 3.58860195903829e-05,
+      "loss": 0.4224,
+      "step": 884
+    },
+    {
+      "epoch": 1.0637019230769231,
+      "grad_norm": 0.3691250996841783,
+      "learning_rate": 3.586375779162957e-05,
+      "loss": 0.4415,
+      "step": 885
+    },
+    {
+      "epoch": 1.0649038461538463,
+      "grad_norm": 0.3709357518188861,
+      "learning_rate": 3.584149599287623e-05,
+      "loss": 0.4568,
+      "step": 886
+    },
+    {
+      "epoch": 1.0661057692307692,
+      "grad_norm": 0.35048050655770996,
+      "learning_rate": 3.5819234194122886e-05,
+      "loss": 0.44,
+      "step": 887
+    },
+    {
+      "epoch": 1.0673076923076923,
+      "grad_norm": 0.3674631549705021,
+      "learning_rate": 3.579697239536955e-05,
+      "loss": 0.4532,
+      "step": 888
+    },
+    {
+      "epoch": 1.0685096153846154,
+      "grad_norm": 0.3371345679545054,
+      "learning_rate": 3.5774710596616204e-05,
+      "loss": 0.4193,
+      "step": 889
+    },
+    {
+      "epoch": 1.0697115384615385,
+      "grad_norm": 0.3011152764747545,
+      "learning_rate": 3.575244879786287e-05,
+      "loss": 0.4214,
+      "step": 890
+    },
+    {
+      "epoch": 1.0709134615384615,
+      "grad_norm": 0.41030491795227575,
+      "learning_rate": 3.573018699910953e-05,
+      "loss": 0.4909,
+      "step": 891
+    },
+    {
+      "epoch": 1.0721153846153846,
+      "grad_norm": 0.3814689445541171,
+      "learning_rate": 3.570792520035619e-05,
+      "loss": 0.4434,
+      "step": 892
+    },
+    {
+      "epoch": 1.0733173076923077,
+      "grad_norm": 0.31139336551736807,
+      "learning_rate": 3.568566340160285e-05,
+      "loss": 0.4114,
+      "step": 893
+    },
+    {
+      "epoch": 1.0745192307692308,
+      "grad_norm": 0.3238374936261473,
+      "learning_rate": 3.5663401602849514e-05,
+      "loss": 0.4283,
+      "step": 894
+    },
+    {
+      "epoch": 1.0757211538461537,
+      "grad_norm": 0.3613082236250399,
+      "learning_rate": 3.5641139804096176e-05,
+      "loss": 0.4311,
+      "step": 895
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.3016092720539302,
+      "learning_rate": 3.561887800534283e-05,
+      "loss": 0.4505,
+      "step": 896
+    },
+    {
+      "epoch": 1.078125,
+      "grad_norm": 0.3521380615834303,
+      "learning_rate": 3.5596616206589494e-05,
+      "loss": 0.4448,
+      "step": 897
+    },
+    {
+      "epoch": 1.0793269230769231,
+      "grad_norm": 0.30738842038369646,
+      "learning_rate": 3.557435440783615e-05,
+      "loss": 0.4616,
+      "step": 898
+    },
+    {
+      "epoch": 1.0805288461538463,
+      "grad_norm": 0.3567785178540512,
+      "learning_rate": 3.555209260908282e-05,
+      "loss": 0.4835,
+      "step": 899
+    },
+    {
+      "epoch": 1.0817307692307692,
+      "grad_norm": 0.3423379230366388,
+      "learning_rate": 3.552983081032948e-05,
+      "loss": 0.4812,
+      "step": 900
+    },
+    {
+      "epoch": 1.0829326923076923,
+      "grad_norm": 0.332433339266352,
+      "learning_rate": 3.5507569011576136e-05,
+      "loss": 0.4127,
+      "step": 901
+    },
+    {
+      "epoch": 1.0841346153846154,
+      "grad_norm": 0.3376235422777881,
+      "learning_rate": 3.54853072128228e-05,
+      "loss": 0.4281,
+      "step": 902
+    },
+    {
+      "epoch": 1.0853365384615385,
+      "grad_norm": 0.3245121175686749,
+      "learning_rate": 3.546304541406946e-05,
+      "loss": 0.4184,
+      "step": 903
+    },
+    {
+      "epoch": 1.0865384615384615,
+      "grad_norm": 0.3355373000477208,
+      "learning_rate": 3.544078361531612e-05,
+      "loss": 0.4357,
+      "step": 904
+    },
+    {
+      "epoch": 1.0877403846153846,
+      "grad_norm": 0.3555587262287326,
+      "learning_rate": 3.541852181656278e-05,
+      "loss": 0.4274,
+      "step": 905
+    },
+    {
+      "epoch": 1.0889423076923077,
+      "grad_norm": 0.36734403343296834,
+      "learning_rate": 3.539626001780944e-05,
+      "loss": 0.4308,
+      "step": 906
+    },
+    {
+      "epoch": 1.0901442307692308,
+      "grad_norm": 0.39822896617591036,
+      "learning_rate": 3.53739982190561e-05,
+      "loss": 0.4612,
+      "step": 907
+    },
+    {
+      "epoch": 1.0913461538461537,
+      "grad_norm": 0.2978529363282579,
+      "learning_rate": 3.5351736420302764e-05,
+      "loss": 0.4618,
+      "step": 908
+    },
+    {
+      "epoch": 1.0925480769230769,
+      "grad_norm": 0.38482576495697246,
+      "learning_rate": 3.5329474621549426e-05,
+      "loss": 0.4436,
+      "step": 909
+    },
+    {
+      "epoch": 1.09375,
+      "grad_norm": 0.4117862059894906,
+      "learning_rate": 3.530721282279608e-05,
+      "loss": 0.468,
+      "step": 910
+    },
+    {
+      "epoch": 1.0949519230769231,
+      "grad_norm": 0.3330260854477405,
+      "learning_rate": 3.528495102404274e-05,
+      "loss": 0.4162,
+      "step": 911
+    },
+    {
+      "epoch": 1.0961538461538463,
+      "grad_norm": 0.33622502930975307,
+      "learning_rate": 3.5262689225289405e-05,
+      "loss": 0.4134,
+      "step": 912
+    },
+    {
+      "epoch": 1.0973557692307692,
+      "grad_norm": 0.43350961990160947,
+      "learning_rate": 3.524042742653607e-05,
+      "loss": 0.4336,
+      "step": 913
+    },
+    {
+      "epoch": 1.0985576923076923,
+      "grad_norm": 0.312499956725973,
+      "learning_rate": 3.521816562778273e-05,
+      "loss": 0.4358,
+      "step": 914
+    },
+    {
+      "epoch": 1.0997596153846154,
+      "grad_norm": 0.369097705107959,
+      "learning_rate": 3.5195903829029385e-05,
+      "loss": 0.4469,
+      "step": 915
+    },
+    {
+      "epoch": 1.1009615384615385,
+      "grad_norm": 0.31090546605883196,
+      "learning_rate": 3.517364203027605e-05,
+      "loss": 0.4485,
+      "step": 916
+    },
+    {
+      "epoch": 1.1021634615384615,
+      "grad_norm": 0.35022491933773037,
+      "learning_rate": 3.515138023152271e-05,
+      "loss": 0.4379,
+      "step": 917
+    },
+    {
+      "epoch": 1.1033653846153846,
+      "grad_norm": 0.32526174111730005,
+      "learning_rate": 3.512911843276937e-05,
+      "loss": 0.421,
+      "step": 918
+    },
+    {
+      "epoch": 1.1045673076923077,
+      "grad_norm": 0.2978300811857257,
+      "learning_rate": 3.510685663401603e-05,
+      "loss": 0.4445,
+      "step": 919
+    },
+    {
+      "epoch": 1.1057692307692308,
+      "grad_norm": 0.3111561246382403,
+      "learning_rate": 3.508459483526269e-05,
+      "loss": 0.4598,
+      "step": 920
+    },
+    {
+      "epoch": 1.1069711538461537,
+      "grad_norm": 0.3277156604096349,
+      "learning_rate": 3.506233303650935e-05,
+      "loss": 0.458,
+      "step": 921
+    },
+    {
+      "epoch": 1.1081730769230769,
+      "grad_norm": 0.27236760720680314,
+      "learning_rate": 3.504007123775601e-05,
+      "loss": 0.4577,
+      "step": 922
+    },
+    {
+      "epoch": 1.109375,
+      "grad_norm": 0.3154358263498169,
+      "learning_rate": 3.5017809439002675e-05,
+      "loss": 0.4144,
+      "step": 923
+    },
+    {
+      "epoch": 1.1105769230769231,
+      "grad_norm": 0.3149830918037447,
+      "learning_rate": 3.499554764024933e-05,
+      "loss": 0.4594,
+      "step": 924
+    },
+    {
+      "epoch": 1.1117788461538463,
+      "grad_norm": 0.3569449849017939,
+      "learning_rate": 3.497328584149599e-05,
+      "loss": 0.4515,
+      "step": 925
+    },
+    {
+      "epoch": 1.1129807692307692,
+      "grad_norm": 0.314807709297849,
+      "learning_rate": 3.4951024042742655e-05,
+      "loss": 0.4259,
+      "step": 926
+    },
+    {
+      "epoch": 1.1141826923076923,
+      "grad_norm": 0.324703625966616,
+      "learning_rate": 3.492876224398932e-05,
+      "loss": 0.4672,
+      "step": 927
+    },
+    {
+      "epoch": 1.1153846153846154,
+      "grad_norm": 0.31096050516155976,
+      "learning_rate": 3.490650044523598e-05,
+      "loss": 0.4451,
+      "step": 928
+    },
+    {
+      "epoch": 1.1165865384615385,
+      "grad_norm": 0.31496256223321856,
+      "learning_rate": 3.4884238646482635e-05,
+      "loss": 0.4887,
+      "step": 929
+    },
+    {
+      "epoch": 1.1177884615384615,
+      "grad_norm": 0.3369479232568071,
+      "learning_rate": 3.48619768477293e-05,
+      "loss": 0.4469,
+      "step": 930
+    },
+    {
+      "epoch": 1.1189903846153846,
+      "grad_norm": 0.28663479830598504,
+      "learning_rate": 3.483971504897596e-05,
+      "loss": 0.4489,
+      "step": 931
+    },
+    {
+      "epoch": 1.1201923076923077,
+      "grad_norm": 0.28773175875367496,
+      "learning_rate": 3.481745325022262e-05,
+      "loss": 0.4223,
+      "step": 932
+    },
+    {
+      "epoch": 1.1213942307692308,
+      "grad_norm": 0.3566044353349483,
+      "learning_rate": 3.4795191451469276e-05,
+      "loss": 0.461,
+      "step": 933
+    },
+    {
+      "epoch": 1.1225961538461537,
+      "grad_norm": 0.33153529027838713,
+      "learning_rate": 3.477292965271594e-05,
+      "loss": 0.4626,
+      "step": 934
+    },
+    {
+      "epoch": 1.1237980769230769,
+      "grad_norm": 0.3388819075800715,
+      "learning_rate": 3.475066785396261e-05,
+      "loss": 0.4508,
+      "step": 935
+    },
+    {
+      "epoch": 1.125,
+      "grad_norm": 0.32546619521140246,
+      "learning_rate": 3.472840605520926e-05,
+      "loss": 0.4123,
+      "step": 936
+    },
+    {
+      "epoch": 1.1262019230769231,
+      "grad_norm": 0.311888967846706,
+      "learning_rate": 3.4706144256455925e-05,
+      "loss": 0.4436,
+      "step": 937
+    },
+    {
+      "epoch": 1.1274038461538463,
+      "grad_norm": 0.31241440218009825,
+      "learning_rate": 3.468388245770258e-05,
+      "loss": 0.4538,
+      "step": 938
+    },
+    {
+      "epoch": 1.1286057692307692,
+      "grad_norm": 0.2808716922557873,
+      "learning_rate": 3.466162065894924e-05,
+      "loss": 0.4292,
+      "step": 939
+    },
+    {
+      "epoch": 1.1298076923076923,
+      "grad_norm": 0.302760777812588,
+      "learning_rate": 3.4639358860195904e-05,
+      "loss": 0.4119,
+      "step": 940
+    },
+    {
+      "epoch": 1.1310096153846154,
+      "grad_norm": 0.3429219970100684,
+      "learning_rate": 3.461709706144257e-05,
+      "loss": 0.4652,
+      "step": 941
+    },
+    {
+      "epoch": 1.1322115384615385,
+      "grad_norm": 0.37719324452637365,
+      "learning_rate": 3.459483526268923e-05,
+      "loss": 0.4247,
+      "step": 942
+    },
+    {
+      "epoch": 1.1334134615384615,
+      "grad_norm": 0.33339573879862966,
+      "learning_rate": 3.4572573463935884e-05,
+      "loss": 0.426,
+      "step": 943
+    },
+    {
+      "epoch": 1.1346153846153846,
+      "grad_norm": 0.3422951258467978,
+      "learning_rate": 3.455031166518255e-05,
+      "loss": 0.4384,
+      "step": 944
+    },
+    {
+      "epoch": 1.1358173076923077,
+      "grad_norm": 0.30977324704026277,
+      "learning_rate": 3.452804986642921e-05,
+      "loss": 0.4347,
+      "step": 945
+    },
+    {
+      "epoch": 1.1370192307692308,
+      "grad_norm": 0.37479023836196806,
+      "learning_rate": 3.450578806767587e-05,
+      "loss": 0.4485,
+      "step": 946
+    },
+    {
+      "epoch": 1.1382211538461537,
+      "grad_norm": 0.3496203769635772,
+      "learning_rate": 3.4483526268922526e-05,
+      "loss": 0.4186,
+      "step": 947
+    },
+    {
+      "epoch": 1.1394230769230769,
+      "grad_norm": 0.3349212883208834,
+      "learning_rate": 3.446126447016919e-05,
+      "loss": 0.4173,
+      "step": 948
+    },
+    {
+      "epoch": 1.140625,
+      "grad_norm": 0.3168647479367049,
+      "learning_rate": 3.443900267141586e-05,
+      "loss": 0.4683,
+      "step": 949
+    },
+    {
+      "epoch": 1.1418269230769231,
+      "grad_norm": 0.35485735871814944,
+      "learning_rate": 3.441674087266251e-05,
+      "loss": 0.4372,
+      "step": 950
+    },
+    {
+      "epoch": 1.1430288461538463,
+      "grad_norm": 0.3263671372355466,
+      "learning_rate": 3.4394479073909174e-05,
+      "loss": 0.4772,
+      "step": 951
+    },
+    {
+      "epoch": 1.1442307692307692,
+      "grad_norm": 0.3215363776128306,
+      "learning_rate": 3.437221727515583e-05,
+      "loss": 0.4389,
+      "step": 952
+    },
+    {
+      "epoch": 1.1454326923076923,
+      "grad_norm": 0.3859301598721572,
+      "learning_rate": 3.43499554764025e-05,
+      "loss": 0.4658,
+      "step": 953
+    },
+    {
+      "epoch": 1.1466346153846154,
+      "grad_norm": 0.3457197482655351,
+      "learning_rate": 3.4327693677649154e-05,
+      "loss": 0.4724,
+      "step": 954
+    },
+    {
+      "epoch": 1.1478365384615385,
+      "grad_norm": 0.36268388774376115,
+      "learning_rate": 3.4305431878895816e-05,
+      "loss": 0.4656,
+      "step": 955
+    },
+    {
+      "epoch": 1.1490384615384615,
+      "grad_norm": 0.30710125454947657,
+      "learning_rate": 3.428317008014248e-05,
+      "loss": 0.403,
+      "step": 956
+    },
+    {
+      "epoch": 1.1502403846153846,
+      "grad_norm": 0.30819055774956206,
+      "learning_rate": 3.4260908281389134e-05,
+      "loss": 0.4605,
+      "step": 957
+    },
+    {
+      "epoch": 1.1514423076923077,
+      "grad_norm": 0.3115268753576731,
+      "learning_rate": 3.42386464826358e-05,
+      "loss": 0.4152,
+      "step": 958
+    },
+    {
+      "epoch": 1.1526442307692308,
+      "grad_norm": 0.31067247872951953,
+      "learning_rate": 3.421638468388246e-05,
+      "loss": 0.4397,
+      "step": 959
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.3510527444723829,
+      "learning_rate": 3.419412288512912e-05,
+      "loss": 0.4639,
+      "step": 960
+    },
+    {
+      "epoch": 1.1550480769230769,
+      "grad_norm": 0.313625355149206,
+      "learning_rate": 3.4171861086375775e-05,
+      "loss": 0.4395,
+      "step": 961
+    },
+    {
+      "epoch": 1.15625,
+      "grad_norm": 0.31727639493636034,
+      "learning_rate": 3.4149599287622444e-05,
+      "loss": 0.4365,
+      "step": 962
+    },
+    {
+      "epoch": 1.1574519230769231,
+      "grad_norm": 0.2657756821373731,
+      "learning_rate": 3.4127337488869106e-05,
+      "loss": 0.4251,
+      "step": 963
+    },
+    {
+      "epoch": 1.1586538461538463,
+      "grad_norm": 0.29464995368303554,
+      "learning_rate": 3.410507569011576e-05,
+      "loss": 0.4243,
+      "step": 964
+    },
+    {
+      "epoch": 1.1598557692307692,
+      "grad_norm": 0.29201414121231584,
+      "learning_rate": 3.4082813891362424e-05,
+      "loss": 0.425,
+      "step": 965
+    },
+    {
+      "epoch": 1.1610576923076923,
+      "grad_norm": 0.28412185734626494,
+      "learning_rate": 3.406055209260908e-05,
+      "loss": 0.439,
+      "step": 966
+    },
+    {
+      "epoch": 1.1622596153846154,
+      "grad_norm": 0.3190267190556927,
+      "learning_rate": 3.403829029385575e-05,
+      "loss": 0.4396,
+      "step": 967
+    },
+    {
+      "epoch": 1.1634615384615385,
+      "grad_norm": 0.2990761581967904,
+      "learning_rate": 3.4016028495102404e-05,
+      "loss": 0.4199,
+      "step": 968
+    },
+    {
+      "epoch": 1.1646634615384615,
+      "grad_norm": 0.27519605122857627,
+      "learning_rate": 3.3993766696349066e-05,
+      "loss": 0.4218,
+      "step": 969
+    },
+    {
+      "epoch": 1.1658653846153846,
+      "grad_norm": 0.3461574156938599,
+      "learning_rate": 3.397150489759573e-05,
+      "loss": 0.4137,
+      "step": 970
+    },
+    {
+      "epoch": 1.1670673076923077,
+      "grad_norm": 0.28724531709124324,
+      "learning_rate": 3.394924309884239e-05,
+      "loss": 0.4518,
+      "step": 971
+    },
+    {
+      "epoch": 1.1682692307692308,
+      "grad_norm": 0.3282328509724695,
+      "learning_rate": 3.392698130008905e-05,
+      "loss": 0.4245,
+      "step": 972
+    },
+    {
+      "epoch": 1.1694711538461537,
+      "grad_norm": 0.3251361798270695,
+      "learning_rate": 3.390471950133571e-05,
+      "loss": 0.4335,
+      "step": 973
+    },
+    {
+      "epoch": 1.1706730769230769,
+      "grad_norm": 0.32869333491750286,
+      "learning_rate": 3.388245770258237e-05,
+      "loss": 0.4576,
+      "step": 974
+    },
+    {
+      "epoch": 1.171875,
+      "grad_norm": 0.3373200138741161,
+      "learning_rate": 3.386019590382903e-05,
+      "loss": 0.4375,
+      "step": 975
+    },
+    {
+      "epoch": 1.1730769230769231,
+      "grad_norm": 0.3064683124478583,
+      "learning_rate": 3.3837934105075694e-05,
+      "loss": 0.4411,
+      "step": 976
+    },
+    {
+      "epoch": 1.1742788461538463,
+      "grad_norm": 0.31007364625470935,
+      "learning_rate": 3.3815672306322356e-05,
+      "loss": 0.4246,
+      "step": 977
+    },
+    {
+      "epoch": 1.1754807692307692,
+      "grad_norm": 0.34584436497383614,
+      "learning_rate": 3.379341050756901e-05,
+      "loss": 0.4443,
+      "step": 978
+    },
+    {
+      "epoch": 1.1766826923076923,
+      "grad_norm": 0.29993831758901385,
+      "learning_rate": 3.3771148708815673e-05,
+      "loss": 0.4323,
+      "step": 979
+    },
+    {
+      "epoch": 1.1778846153846154,
+      "grad_norm": 0.31910283951977425,
+      "learning_rate": 3.3748886910062336e-05,
+      "loss": 0.4197,
+      "step": 980
+    },
+    {
+      "epoch": 1.1790865384615385,
+      "grad_norm": 0.36476236966102116,
+      "learning_rate": 3.3726625111309e-05,
+      "loss": 0.4572,
+      "step": 981
+    },
+    {
+      "epoch": 1.1802884615384615,
+      "grad_norm": 0.3297624655782387,
+      "learning_rate": 3.370436331255565e-05,
+      "loss": 0.4281,
+      "step": 982
+    },
+    {
+      "epoch": 1.1814903846153846,
+      "grad_norm": 0.3270117134588345,
+      "learning_rate": 3.3682101513802315e-05,
+      "loss": 0.4515,
+      "step": 983
+    },
+    {
+      "epoch": 1.1826923076923077,
+      "grad_norm": 0.2953849691164886,
+      "learning_rate": 3.365983971504898e-05,
+      "loss": 0.4739,
+      "step": 984
+    },
+    {
+      "epoch": 1.1838942307692308,
+      "grad_norm": 0.2909421368693807,
+      "learning_rate": 3.363757791629564e-05,
+      "loss": 0.421,
+      "step": 985
+    },
+    {
+      "epoch": 1.1850961538461537,
+      "grad_norm": 0.35785022101030567,
+      "learning_rate": 3.36153161175423e-05,
+      "loss": 0.4547,
+      "step": 986
+    },
+    {
+      "epoch": 1.1862980769230769,
+      "grad_norm": 0.27076560229535696,
+      "learning_rate": 3.359305431878896e-05,
+      "loss": 0.4322,
+      "step": 987
+    },
+    {
+      "epoch": 1.1875,
+      "grad_norm": 0.3904686478443894,
+      "learning_rate": 3.357079252003562e-05,
+      "loss": 0.4171,
+      "step": 988
+    },
+    {
+      "epoch": 1.1887019230769231,
+      "grad_norm": 0.36304085842437983,
+      "learning_rate": 3.354853072128228e-05,
+      "loss": 0.4518,
+      "step": 989
+    },
+    {
+      "epoch": 1.1899038461538463,
+      "grad_norm": 0.3138492078490953,
+      "learning_rate": 3.352626892252894e-05,
+      "loss": 0.4549,
+      "step": 990
+    },
+    {
+      "epoch": 1.1911057692307692,
+      "grad_norm": 0.40471253609174085,
+      "learning_rate": 3.3504007123775605e-05,
+      "loss": 0.4446,
+      "step": 991
+    },
+    {
+      "epoch": 1.1923076923076923,
+      "grad_norm": 0.3257330989779682,
+      "learning_rate": 3.348174532502226e-05,
+      "loss": 0.4398,
+      "step": 992
+    },
+    {
+      "epoch": 1.1935096153846154,
+      "grad_norm": 0.3457643975489633,
+      "learning_rate": 3.345948352626892e-05,
+      "loss": 0.4062,
+      "step": 993
+    },
+    {
+      "epoch": 1.1947115384615385,
+      "grad_norm": 0.3844228513477365,
+      "learning_rate": 3.3437221727515585e-05,
+      "loss": 0.4264,
+      "step": 994
+    },
+    {
+      "epoch": 1.1959134615384615,
+      "grad_norm": 0.2833638866620481,
+      "learning_rate": 3.341495992876225e-05,
+      "loss": 0.4102,
+      "step": 995
+    },
+    {
+      "epoch": 1.1971153846153846,
+      "grad_norm": 0.4032021281651122,
+      "learning_rate": 3.33926981300089e-05,
+      "loss": 0.4213,
+      "step": 996
+    },
+    {
+      "epoch": 1.1983173076923077,
+      "grad_norm": 0.33871434875323325,
+      "learning_rate": 3.3370436331255565e-05,
+      "loss": 0.4391,
+      "step": 997
+    },
+    {
+      "epoch": 1.1995192307692308,
+      "grad_norm": 0.3116516036748016,
+      "learning_rate": 3.334817453250223e-05,
+      "loss": 0.4256,
+      "step": 998
+    },
+    {
+      "epoch": 1.2007211538461537,
+      "grad_norm": 0.383977805094453,
+      "learning_rate": 3.332591273374889e-05,
+      "loss": 0.4392,
+      "step": 999
+    },
+    {
+      "epoch": 1.2019230769230769,
+      "grad_norm": 0.33578476632964266,
+      "learning_rate": 3.330365093499555e-05,
+      "loss": 0.4581,
+      "step": 1000
+    },
+    {
+      "epoch": 1.203125,
+      "grad_norm": 0.3421795680968129,
+      "learning_rate": 3.3281389136242206e-05,
+      "loss": 0.437,
+      "step": 1001
+    },
+    {
+      "epoch": 1.2043269230769231,
+      "grad_norm": 0.3819620806179971,
+      "learning_rate": 3.325912733748887e-05,
+      "loss": 0.4459,
+      "step": 1002
+    },
+    {
+      "epoch": 1.2055288461538463,
+      "grad_norm": 0.26762948036884593,
+      "learning_rate": 3.323686553873553e-05,
+      "loss": 0.4342,
+      "step": 1003
+    },
+    {
+      "epoch": 1.2067307692307692,
+      "grad_norm": 0.3533296405854599,
+      "learning_rate": 3.321460373998219e-05,
+      "loss": 0.4457,
+      "step": 1004
+    },
+    {
+      "epoch": 1.2079326923076923,
+      "grad_norm": 0.3041925888573904,
+      "learning_rate": 3.3192341941228855e-05,
+      "loss": 0.438,
+      "step": 1005
+    },
+    {
+      "epoch": 1.2091346153846154,
+      "grad_norm": 0.31718097034915416,
+      "learning_rate": 3.317008014247551e-05,
+      "loss": 0.4495,
+      "step": 1006
+    },
+    {
+      "epoch": 1.2103365384615385,
+      "grad_norm": 0.41063290167477173,
+      "learning_rate": 3.314781834372217e-05,
+      "loss": 0.4376,
+      "step": 1007
+    },
+    {
+      "epoch": 1.2115384615384615,
+      "grad_norm": 0.3662724735665931,
+      "learning_rate": 3.3125556544968835e-05,
+      "loss": 0.4604,
+      "step": 1008
+    },
+    {
+      "epoch": 1.2127403846153846,
+      "grad_norm": 0.26469965460702527,
+      "learning_rate": 3.31032947462155e-05,
+      "loss": 0.4244,
+      "step": 1009
+    },
+    {
+      "epoch": 1.2139423076923077,
+      "grad_norm": 0.3139247053509421,
+      "learning_rate": 3.308103294746215e-05,
+      "loss": 0.4628,
+      "step": 1010
+    },
+    {
+      "epoch": 1.2151442307692308,
+      "grad_norm": 0.29726657840946835,
+      "learning_rate": 3.3058771148708814e-05,
+      "loss": 0.4426,
+      "step": 1011
+    },
+    {
+      "epoch": 1.2163461538461537,
+      "grad_norm": 0.33361880041579856,
+      "learning_rate": 3.303650934995548e-05,
+      "loss": 0.4351,
+      "step": 1012
+    },
+    {
+      "epoch": 1.2175480769230769,
+      "grad_norm": 0.3146673915183789,
+      "learning_rate": 3.301424755120214e-05,
+      "loss": 0.4541,
+      "step": 1013
+    },
+    {
+      "epoch": 1.21875,
+      "grad_norm": 0.3278524372995347,
+      "learning_rate": 3.29919857524488e-05,
+      "loss": 0.4373,
+      "step": 1014
+    },
+    {
+      "epoch": 1.2199519230769231,
+      "grad_norm": 0.28637589564013005,
+      "learning_rate": 3.2969723953695456e-05,
+      "loss": 0.4336,
+      "step": 1015
+    },
+    {
+      "epoch": 1.2211538461538463,
+      "grad_norm": 0.36518818820486665,
+      "learning_rate": 3.294746215494212e-05,
+      "loss": 0.4451,
+      "step": 1016
+    },
+    {
+      "epoch": 1.2223557692307692,
+      "grad_norm": 0.30683653224723795,
+      "learning_rate": 3.292520035618878e-05,
+      "loss": 0.4317,
+      "step": 1017
+    },
+    {
+      "epoch": 1.2235576923076923,
+      "grad_norm": 0.2865214679759774,
+      "learning_rate": 3.290293855743544e-05,
+      "loss": 0.4144,
+      "step": 1018
+    },
+    {
+      "epoch": 1.2247596153846154,
+      "grad_norm": 0.29637291815219,
+      "learning_rate": 3.2880676758682105e-05,
+      "loss": 0.4426,
+      "step": 1019
+    },
+    {
+      "epoch": 1.2259615384615385,
+      "grad_norm": 0.2975172830529344,
+      "learning_rate": 3.285841495992876e-05,
+      "loss": 0.4119,
+      "step": 1020
+    },
+    {
+      "epoch": 1.2271634615384615,
+      "grad_norm": 0.28945410472187516,
+      "learning_rate": 3.283615316117543e-05,
+      "loss": 0.4389,
+      "step": 1021
+    },
+    {
+      "epoch": 1.2283653846153846,
+      "grad_norm": 0.29241495890042135,
+      "learning_rate": 3.2813891362422084e-05,
+      "loss": 0.4253,
+      "step": 1022
+    },
+    {
+      "epoch": 1.2295673076923077,
+      "grad_norm": 0.31430171275547597,
+      "learning_rate": 3.2791629563668746e-05,
+      "loss": 0.4407,
+      "step": 1023
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.33205055754455143,
+      "learning_rate": 3.27693677649154e-05,
+      "loss": 0.4255,
+      "step": 1024
+    },
+    {
+      "epoch": 1.2319711538461537,
+      "grad_norm": 0.2760794961664041,
+      "learning_rate": 3.274710596616207e-05,
+      "loss": 0.4484,
+      "step": 1025
+    },
+    {
+      "epoch": 1.2331730769230769,
+      "grad_norm": 0.35815529753969244,
+      "learning_rate": 3.272484416740873e-05,
+      "loss": 0.4448,
+      "step": 1026
+    },
+    {
+      "epoch": 1.234375,
+      "grad_norm": 0.3062396635561892,
+      "learning_rate": 3.270258236865539e-05,
+      "loss": 0.4502,
+      "step": 1027
+    },
+    {
+      "epoch": 1.2355769230769231,
+      "grad_norm": 0.34875449711916523,
+      "learning_rate": 3.268032056990205e-05,
+      "loss": 0.4306,
+      "step": 1028
+    },
+    {
+      "epoch": 1.2367788461538463,
+      "grad_norm": 0.33573917786737417,
+      "learning_rate": 3.2658058771148706e-05,
+      "loss": 0.4453,
+      "step": 1029
+    },
+    {
+      "epoch": 1.2379807692307692,
+      "grad_norm": 0.2921613819617989,
+      "learning_rate": 3.2635796972395374e-05,
+      "loss": 0.4361,
+      "step": 1030
+    },
+    {
+      "epoch": 1.2391826923076923,
+      "grad_norm": 0.3532544305368514,
+      "learning_rate": 3.261353517364203e-05,
+      "loss": 0.4316,
+      "step": 1031
+    },
+    {
+      "epoch": 1.2403846153846154,
+      "grad_norm": 0.2891449537393726,
+      "learning_rate": 3.259127337488869e-05,
+      "loss": 0.4547,
+      "step": 1032
+    },
+    {
+      "epoch": 1.2415865384615385,
+      "grad_norm": 0.3422324522552323,
+      "learning_rate": 3.2569011576135354e-05,
+      "loss": 0.4371,
+      "step": 1033
+    },
+    {
+      "epoch": 1.2427884615384615,
+      "grad_norm": 0.34730105746129397,
+      "learning_rate": 3.2546749777382016e-05,
+      "loss": 0.4557,
+      "step": 1034
+    },
+    {
+      "epoch": 1.2439903846153846,
+      "grad_norm": 0.3271513506520441,
+      "learning_rate": 3.252448797862868e-05,
+      "loss": 0.4672,
+      "step": 1035
+    },
+    {
+      "epoch": 1.2451923076923077,
+      "grad_norm": 0.33503885284038115,
+      "learning_rate": 3.2502226179875334e-05,
+      "loss": 0.4457,
+      "step": 1036
+    },
+    {
+      "epoch": 1.2463942307692308,
+      "grad_norm": 0.2953408129959925,
+      "learning_rate": 3.2479964381121996e-05,
+      "loss": 0.4192,
+      "step": 1037
+    },
+    {
+      "epoch": 1.2475961538461537,
+      "grad_norm": 0.3062665237883024,
+      "learning_rate": 3.245770258236865e-05,
+      "loss": 0.4257,
+      "step": 1038
+    },
+    {
+      "epoch": 1.2487980769230769,
+      "grad_norm": 0.3866186178009434,
+      "learning_rate": 3.243544078361532e-05,
+      "loss": 0.4298,
+      "step": 1039
+    },
+    {
+      "epoch": 1.25,
+      "grad_norm": 0.31575307005771464,
+      "learning_rate": 3.241317898486198e-05,
+      "loss": 0.441,
+      "step": 1040
+    },
+    {
+      "epoch": 1.2512019230769231,
+      "grad_norm": 0.3575280507333239,
+      "learning_rate": 3.239091718610864e-05,
+      "loss": 0.4427,
+      "step": 1041
+    },
+    {
+      "epoch": 1.2524038461538463,
+      "grad_norm": 0.42698512637425606,
+      "learning_rate": 3.23686553873553e-05,
+      "loss": 0.4339,
+      "step": 1042
+    },
+    {
+      "epoch": 1.2536057692307692,
+      "grad_norm": 0.29714018153202004,
+      "learning_rate": 3.234639358860196e-05,
+      "loss": 0.4574,
+      "step": 1043
+    },
+    {
+      "epoch": 1.2548076923076923,
+      "grad_norm": 0.3851376374393255,
+      "learning_rate": 3.2324131789848624e-05,
+      "loss": 0.4348,
+      "step": 1044
+    },
+    {
+      "epoch": 1.2560096153846154,
+      "grad_norm": 0.3579705063199461,
+      "learning_rate": 3.230186999109528e-05,
+      "loss": 0.4585,
+      "step": 1045
+    },
+    {
+      "epoch": 1.2572115384615383,
+      "grad_norm": 0.31873397258385605,
+      "learning_rate": 3.227960819234194e-05,
+      "loss": 0.4429,
+      "step": 1046
+    },
+    {
+      "epoch": 1.2584134615384617,
+      "grad_norm": 0.3665614615776611,
+      "learning_rate": 3.2257346393588604e-05,
+      "loss": 0.4378,
+      "step": 1047
+    },
+    {
+      "epoch": 1.2596153846153846,
+      "grad_norm": 0.3346982711358021,
+      "learning_rate": 3.2235084594835266e-05,
+      "loss": 0.426,
+      "step": 1048
+    },
+    {
+      "epoch": 1.2608173076923077,
+      "grad_norm": 0.3307122183271403,
+      "learning_rate": 3.221282279608193e-05,
+      "loss": 0.4459,
+      "step": 1049
+    },
+    {
+      "epoch": 1.2620192307692308,
+      "grad_norm": 0.36845401567164204,
+      "learning_rate": 3.219056099732858e-05,
+      "loss": 0.4231,
+      "step": 1050
+    },
+    {
+      "epoch": 1.2632211538461537,
+      "grad_norm": 0.7503759165802117,
+      "learning_rate": 3.2168299198575245e-05,
+      "loss": 0.43,
+      "step": 1051
+    },
+    {
+      "epoch": 1.2644230769230769,
+      "grad_norm": 0.35543427722954757,
+      "learning_rate": 3.214603739982191e-05,
+      "loss": 0.4462,
+      "step": 1052
+    },
+    {
+      "epoch": 1.265625,
+      "grad_norm": 0.31312787373607115,
+      "learning_rate": 3.212377560106857e-05,
+      "loss": 0.4374,
+      "step": 1053
+    },
+    {
+      "epoch": 1.2668269230769231,
+      "grad_norm": 0.3365010028477277,
+      "learning_rate": 3.210151380231523e-05,
+      "loss": 0.425,
+      "step": 1054
+    },
+    {
+      "epoch": 1.2680288461538463,
+      "grad_norm": 0.28561259247303394,
+      "learning_rate": 3.207925200356189e-05,
+      "loss": 0.4298,
+      "step": 1055
+    },
+    {
+      "epoch": 1.2692307692307692,
+      "grad_norm": 0.3213953885697151,
+      "learning_rate": 3.205699020480855e-05,
+      "loss": 0.4421,
+      "step": 1056
+    },
+    {
+      "epoch": 1.2704326923076923,
+      "grad_norm": 0.31812254533862844,
+      "learning_rate": 3.203472840605521e-05,
+      "loss": 0.4236,
+      "step": 1057
+    },
+    {
+      "epoch": 1.2716346153846154,
+      "grad_norm": 0.39155029903603933,
+      "learning_rate": 3.2012466607301873e-05,
+      "loss": 0.4259,
+      "step": 1058
+    },
+    {
+      "epoch": 1.2728365384615383,
+      "grad_norm": 0.32949772084056805,
+      "learning_rate": 3.199020480854853e-05,
+      "loss": 0.4549,
+      "step": 1059
+    },
+    {
+      "epoch": 1.2740384615384617,
+      "grad_norm": 0.3832084151794874,
+      "learning_rate": 3.196794300979519e-05,
+      "loss": 0.4332,
+      "step": 1060
+    },
+    {
+      "epoch": 1.2752403846153846,
+      "grad_norm": 0.2923572413170696,
+      "learning_rate": 3.194568121104185e-05,
+      "loss": 0.4105,
+      "step": 1061
+    },
+    {
+      "epoch": 1.2764423076923077,
+      "grad_norm": 0.33383806807769517,
+      "learning_rate": 3.1923419412288515e-05,
+      "loss": 0.4236,
+      "step": 1062
+    },
+    {
+      "epoch": 1.2776442307692308,
+      "grad_norm": 0.3609942199189663,
+      "learning_rate": 3.190115761353518e-05,
+      "loss": 0.4398,
+      "step": 1063
+    },
+    {
+      "epoch": 1.2788461538461537,
+      "grad_norm": 0.36402898639186776,
+      "learning_rate": 3.187889581478183e-05,
+      "loss": 0.431,
+      "step": 1064
+    },
+    {
+      "epoch": 1.2800480769230769,
+      "grad_norm": 0.3582260248255581,
+      "learning_rate": 3.1856634016028495e-05,
+      "loss": 0.4317,
+      "step": 1065
+    },
+    {
+      "epoch": 1.28125,
+      "grad_norm": 0.40387835537148276,
+      "learning_rate": 3.183437221727516e-05,
+      "loss": 0.448,
+      "step": 1066
+    },
+    {
+      "epoch": 1.2824519230769231,
+      "grad_norm": 0.31897640628321805,
+      "learning_rate": 3.181211041852182e-05,
+      "loss": 0.4207,
+      "step": 1067
+    },
+    {
+      "epoch": 1.2836538461538463,
+      "grad_norm": 0.2811614041382863,
+      "learning_rate": 3.178984861976848e-05,
+      "loss": 0.4189,
+      "step": 1068
+    },
+    {
+      "epoch": 1.2848557692307692,
+      "grad_norm": 0.33788771369144344,
+      "learning_rate": 3.1767586821015137e-05,
+      "loss": 0.4352,
+      "step": 1069
+    },
+    {
+      "epoch": 1.2860576923076923,
+      "grad_norm": 0.38261977042468,
+      "learning_rate": 3.17453250222618e-05,
+      "loss": 0.4498,
+      "step": 1070
+    },
+    {
+      "epoch": 1.2872596153846154,
+      "grad_norm": 0.3060072929173554,
+      "learning_rate": 3.172306322350846e-05,
+      "loss": 0.4487,
+      "step": 1071
+    },
+    {
+      "epoch": 1.2884615384615383,
+      "grad_norm": 0.35615169743880315,
+      "learning_rate": 3.170080142475512e-05,
+      "loss": 0.4139,
+      "step": 1072
+    },
+    {
+      "epoch": 1.2896634615384617,
+      "grad_norm": 0.3317470195903281,
+      "learning_rate": 3.167853962600178e-05,
+      "loss": 0.4303,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2908653846153846,
+      "grad_norm": 0.3981111633250077,
+      "learning_rate": 3.165627782724844e-05,
+      "loss": 0.4409,
+      "step": 1074
+    },
+    {
+      "epoch": 1.2920673076923077,
+      "grad_norm": 0.35072979814519284,
+      "learning_rate": 3.16340160284951e-05,
+      "loss": 0.4272,
+      "step": 1075
+    },
+    {
+      "epoch": 1.2932692307692308,
+      "grad_norm": 0.3671636434684682,
+      "learning_rate": 3.1611754229741765e-05,
+      "loss": 0.3975,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2944711538461537,
+      "grad_norm": 0.38841460803708616,
+      "learning_rate": 3.158949243098843e-05,
+      "loss": 0.4344,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2956730769230769,
+      "grad_norm": 0.363572401429653,
+      "learning_rate": 3.156723063223508e-05,
+      "loss": 0.4378,
+      "step": 1078
+    },
+    {
+      "epoch": 1.296875,
+      "grad_norm": 0.40651341422640935,
+      "learning_rate": 3.1544968833481744e-05,
+      "loss": 0.4301,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2980769230769231,
+      "grad_norm": 0.2986605220700124,
+      "learning_rate": 3.1522707034728406e-05,
+      "loss": 0.4034,
+      "step": 1080
+    },
+    {
+      "epoch": 1.2992788461538463,
+      "grad_norm": 0.48461404755323323,
+      "learning_rate": 3.150044523597507e-05,
+      "loss": 0.4459,
+      "step": 1081
+    },
+    {
+      "epoch": 1.3004807692307692,
+      "grad_norm": 0.3590508383619332,
+      "learning_rate": 3.147818343722173e-05,
+      "loss": 0.4368,
+      "step": 1082
+    },
+    {
+      "epoch": 1.3016826923076923,
+      "grad_norm": 0.350626749748603,
+      "learning_rate": 3.1455921638468386e-05,
+      "loss": 0.4263,
+      "step": 1083
+    },
+    {
+      "epoch": 1.3028846153846154,
+      "grad_norm": 0.29210892099930863,
+      "learning_rate": 3.1433659839715055e-05,
+      "loss": 0.403,
+      "step": 1084
+    },
+    {
+      "epoch": 1.3040865384615383,
+      "grad_norm": 0.3263269657319956,
+      "learning_rate": 3.141139804096171e-05,
+      "loss": 0.4298,
+      "step": 1085
+    },
+    {
+      "epoch": 1.3052884615384617,
+      "grad_norm": 0.3130057841969151,
+      "learning_rate": 3.138913624220837e-05,
+      "loss": 0.4496,
+      "step": 1086
+    },
+    {
+      "epoch": 1.3064903846153846,
+      "grad_norm": 0.3069689696312972,
+      "learning_rate": 3.136687444345503e-05,
+      "loss": 0.4203,
+      "step": 1087
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.38806488833109926,
+      "learning_rate": 3.134461264470169e-05,
+      "loss": 0.4489,
+      "step": 1088
+    },
+    {
+      "epoch": 1.3088942307692308,
+      "grad_norm": 0.3116953416322602,
+      "learning_rate": 3.132235084594836e-05,
+      "loss": 0.463,
+      "step": 1089
+    },
+    {
+      "epoch": 1.3100961538461537,
+      "grad_norm": 0.3192545903250187,
+      "learning_rate": 3.1300089047195014e-05,
+      "loss": 0.4344,
+      "step": 1090
+    },
+    {
+      "epoch": 1.3112980769230769,
+      "grad_norm": 0.33018235565487336,
+      "learning_rate": 3.1277827248441676e-05,
+      "loss": 0.4409,
+      "step": 1091
+    },
+    {
+      "epoch": 1.3125,
+      "grad_norm": 0.33431281246740574,
+      "learning_rate": 3.125556544968833e-05,
+      "loss": 0.4459,
+      "step": 1092
+    },
+    {
+      "epoch": 1.3137019230769231,
+      "grad_norm": 0.3653676329948962,
+      "learning_rate": 3.1233303650935e-05,
+      "loss": 0.4211,
+      "step": 1093
+    },
+    {
+      "epoch": 1.3149038461538463,
+      "grad_norm": 0.34736373602322906,
+      "learning_rate": 3.121104185218166e-05,
+      "loss": 0.4417,
+      "step": 1094
+    },
+    {
+      "epoch": 1.3161057692307692,
+      "grad_norm": 0.32416136572426124,
+      "learning_rate": 3.118878005342832e-05,
+      "loss": 0.4231,
+      "step": 1095
+    },
+    {
+      "epoch": 1.3173076923076923,
+      "grad_norm": 0.31431882791738475,
+      "learning_rate": 3.116651825467498e-05,
+      "loss": 0.4272,
+      "step": 1096
+    },
+    {
+      "epoch": 1.3185096153846154,
+      "grad_norm": 0.35379442995119226,
+      "learning_rate": 3.1144256455921636e-05,
+      "loss": 0.4532,
+      "step": 1097
+    },
+    {
+      "epoch": 1.3197115384615383,
+      "grad_norm": 0.3154707070219836,
+      "learning_rate": 3.1121994657168305e-05,
+      "loss": 0.4481,
+      "step": 1098
+    },
+    {
+      "epoch": 1.3209134615384617,
+      "grad_norm": 0.36135976412551474,
+      "learning_rate": 3.109973285841496e-05,
+      "loss": 0.4468,
+      "step": 1099
+    },
+    {
+      "epoch": 1.3221153846153846,
+      "grad_norm": 0.342969531125382,
+      "learning_rate": 3.107747105966162e-05,
+      "loss": 0.4731,
+      "step": 1100
+    },
+    {
+      "epoch": 1.3233173076923077,
+      "grad_norm": 0.30121239897500807,
+      "learning_rate": 3.1055209260908284e-05,
+      "loss": 0.4316,
+      "step": 1101
+    },
+    {
+      "epoch": 1.3245192307692308,
+      "grad_norm": 0.343316158627011,
+      "learning_rate": 3.1032947462154946e-05,
+      "loss": 0.4578,
+      "step": 1102
+    },
+    {
+      "epoch": 1.3257211538461537,
+      "grad_norm": 0.3134100205374526,
+      "learning_rate": 3.101068566340161e-05,
+      "loss": 0.4394,
+      "step": 1103
+    },
+    {
+      "epoch": 1.3269230769230769,
+      "grad_norm": 0.3149800078433672,
+      "learning_rate": 3.0988423864648264e-05,
+      "loss": 0.4128,
+      "step": 1104
+    },
+    {
+      "epoch": 1.328125,
+      "grad_norm": 0.3186567125729043,
+      "learning_rate": 3.0966162065894926e-05,
+      "loss": 0.4337,
+      "step": 1105
+    },
+    {
+      "epoch": 1.3293269230769231,
+      "grad_norm": 0.30155105209504085,
+      "learning_rate": 3.094390026714158e-05,
+      "loss": 0.4205,
+      "step": 1106
+    },
+    {
+      "epoch": 1.3305288461538463,
+      "grad_norm": 0.3370965230344429,
+      "learning_rate": 3.092163846838825e-05,
+      "loss": 0.4338,
+      "step": 1107
+    },
+    {
+      "epoch": 1.3317307692307692,
+      "grad_norm": 0.2880143755595478,
+      "learning_rate": 3.089937666963491e-05,
+      "loss": 0.4171,
+      "step": 1108
+    },
+    {
+      "epoch": 1.3329326923076923,
+      "grad_norm": 0.3057052601066619,
+      "learning_rate": 3.087711487088157e-05,
+      "loss": 0.4414,
+      "step": 1109
+    },
+    {
+      "epoch": 1.3341346153846154,
+      "grad_norm": 0.3030020526921385,
+      "learning_rate": 3.085485307212823e-05,
+      "loss": 0.4069,
+      "step": 1110
+    },
+    {
+      "epoch": 1.3353365384615383,
+      "grad_norm": 0.33550656901825376,
+      "learning_rate": 3.083259127337489e-05,
+      "loss": 0.4351,
+      "step": 1111
+    },
+    {
+      "epoch": 1.3365384615384617,
+      "grad_norm": 0.2909594287086215,
+      "learning_rate": 3.0810329474621554e-05,
+      "loss": 0.4254,
+      "step": 1112
+    },
+    {
+      "epoch": 1.3377403846153846,
+      "grad_norm": 0.31131134103452374,
+      "learning_rate": 3.078806767586821e-05,
+      "loss": 0.4398,
+      "step": 1113
+    },
+    {
+      "epoch": 1.3389423076923077,
+      "grad_norm": 0.3073374388298493,
+      "learning_rate": 3.076580587711487e-05,
+      "loss": 0.4074,
+      "step": 1114
+    },
+    {
+      "epoch": 1.3401442307692308,
+      "grad_norm": 0.29647852879616626,
+      "learning_rate": 3.0743544078361534e-05,
+      "loss": 0.4308,
+      "step": 1115
+    },
+    {
+      "epoch": 1.3413461538461537,
+      "grad_norm": 0.29222456666353064,
+      "learning_rate": 3.0721282279608196e-05,
+      "loss": 0.4225,
+      "step": 1116
+    },
+    {
+      "epoch": 1.3425480769230769,
+      "grad_norm": 0.291589122238424,
+      "learning_rate": 3.069902048085486e-05,
+      "loss": 0.4354,
+      "step": 1117
+    },
+    {
+      "epoch": 1.34375,
+      "grad_norm": 0.329044377773851,
+      "learning_rate": 3.067675868210151e-05,
+      "loss": 0.4285,
+      "step": 1118
+    },
+    {
+      "epoch": 1.3449519230769231,
+      "grad_norm": 0.2947212962217937,
+      "learning_rate": 3.0654496883348175e-05,
+      "loss": 0.4404,
+      "step": 1119
+    },
+    {
+      "epoch": 1.3461538461538463,
+      "grad_norm": 0.3177267901258307,
+      "learning_rate": 3.063223508459484e-05,
+      "loss": 0.4158,
+      "step": 1120
+    },
+    {
+      "epoch": 1.3473557692307692,
+      "grad_norm": 0.44266677756227435,
+      "learning_rate": 3.06099732858415e-05,
+      "loss": 0.4311,
+      "step": 1121
+    },
+    {
+      "epoch": 1.3485576923076923,
+      "grad_norm": 0.2725951302145634,
+      "learning_rate": 3.058771148708816e-05,
+      "loss": 0.4133,
+      "step": 1122
+    },
+    {
+      "epoch": 1.3497596153846154,
+      "grad_norm": 0.2989249117621078,
+      "learning_rate": 3.056544968833482e-05,
+      "loss": 0.4698,
+      "step": 1123
+    },
+    {
+      "epoch": 1.3509615384615383,
+      "grad_norm": 0.34135707859088443,
+      "learning_rate": 3.054318788958148e-05,
+      "loss": 0.4468,
+      "step": 1124
+    },
+    {
+      "epoch": 1.3521634615384617,
+      "grad_norm": 0.30428600076792256,
+      "learning_rate": 3.052092609082814e-05,
+      "loss": 0.4246,
+      "step": 1125
+    },
+    {
+      "epoch": 1.3533653846153846,
+      "grad_norm": 0.29040531752204396,
+      "learning_rate": 3.0498664292074804e-05,
+      "loss": 0.4108,
+      "step": 1126
+    },
+    {
+      "epoch": 1.3545673076923077,
+      "grad_norm": 0.31051281501425676,
+      "learning_rate": 3.047640249332146e-05,
+      "loss": 0.4561,
+      "step": 1127
+    },
+    {
+      "epoch": 1.3557692307692308,
+      "grad_norm": 0.29442274668862506,
+      "learning_rate": 3.0454140694568124e-05,
+      "loss": 0.4394,
+      "step": 1128
+    },
+    {
+      "epoch": 1.3569711538461537,
+      "grad_norm": 0.2856199542533373,
+      "learning_rate": 3.0431878895814787e-05,
+      "loss": 0.4194,
+      "step": 1129
+    },
+    {
+      "epoch": 1.3581730769230769,
+      "grad_norm": 0.3514126983047654,
+      "learning_rate": 3.0409617097061442e-05,
+      "loss": 0.4495,
+      "step": 1130
+    },
+    {
+      "epoch": 1.359375,
+      "grad_norm": 0.36821890995308776,
+      "learning_rate": 3.0387355298308107e-05,
+      "loss": 0.4468,
+      "step": 1131
+    },
+    {
+      "epoch": 1.3605769230769231,
+      "grad_norm": 0.2910047499657336,
+      "learning_rate": 3.0365093499554763e-05,
+      "loss": 0.4524,
+      "step": 1132
+    },
+    {
+      "epoch": 1.3617788461538463,
+      "grad_norm": 0.4052640171781396,
+      "learning_rate": 3.034283170080143e-05,
+      "loss": 0.4428,
+      "step": 1133
+    },
+    {
+      "epoch": 1.3629807692307692,
+      "grad_norm": 0.3111043932267442,
+      "learning_rate": 3.0320569902048084e-05,
+      "loss": 0.4367,
+      "step": 1134
+    },
+    {
+      "epoch": 1.3641826923076923,
+      "grad_norm": 0.4608689202409539,
+      "learning_rate": 3.029830810329475e-05,
+      "loss": 0.4344,
+      "step": 1135
+    },
+    {
+      "epoch": 1.3653846153846154,
+      "grad_norm": 0.3729976998266354,
+      "learning_rate": 3.027604630454141e-05,
+      "loss": 0.4242,
+      "step": 1136
+    },
+    {
+      "epoch": 1.3665865384615383,
+      "grad_norm": 0.4413806803839775,
+      "learning_rate": 3.025378450578807e-05,
+      "loss": 0.4609,
+      "step": 1137
+    },
+    {
+      "epoch": 1.3677884615384617,
+      "grad_norm": 0.34764446940004473,
+      "learning_rate": 3.0231522707034732e-05,
+      "loss": 0.4328,
+      "step": 1138
+    },
+    {
+      "epoch": 1.3689903846153846,
+      "grad_norm": 0.3397097801573298,
+      "learning_rate": 3.020926090828139e-05,
+      "loss": 0.4201,
+      "step": 1139
+    },
+    {
+      "epoch": 1.3701923076923077,
+      "grad_norm": 0.2961629819660706,
+      "learning_rate": 3.0186999109528053e-05,
+      "loss": 0.4173,
+      "step": 1140
+    },
+    {
+      "epoch": 1.3713942307692308,
+      "grad_norm": 0.38241606308286813,
+      "learning_rate": 3.016473731077471e-05,
+      "loss": 0.4655,
+      "step": 1141
+    },
+    {
+      "epoch": 1.3725961538461537,
+      "grad_norm": 0.33962396677100726,
+      "learning_rate": 3.0142475512021374e-05,
+      "loss": 0.4385,
+      "step": 1142
+    },
+    {
+      "epoch": 1.3737980769230769,
+      "grad_norm": 0.34510260309962776,
+      "learning_rate": 3.0120213713268036e-05,
+      "loss": 0.4487,
+      "step": 1143
+    },
+    {
+      "epoch": 1.375,
+      "grad_norm": 0.34484502085289126,
+      "learning_rate": 3.0097951914514695e-05,
+      "loss": 0.4579,
+      "step": 1144
+    },
+    {
+      "epoch": 1.3762019230769231,
+      "grad_norm": 0.32746666559388155,
+      "learning_rate": 3.0075690115761357e-05,
+      "loss": 0.4575,
+      "step": 1145
+    },
+    {
+      "epoch": 1.3774038461538463,
+      "grad_norm": 0.3292651089339387,
+      "learning_rate": 3.0053428317008016e-05,
+      "loss": 0.4333,
+      "step": 1146
+    },
+    {
+      "epoch": 1.3786057692307692,
+      "grad_norm": 0.2927953722639732,
+      "learning_rate": 3.0031166518254678e-05,
+      "loss": 0.4137,
+      "step": 1147
+    },
+    {
+      "epoch": 1.3798076923076923,
+      "grad_norm": 0.2990606741450613,
+      "learning_rate": 3.0008904719501337e-05,
+      "loss": 0.4338,
+      "step": 1148
+    },
+    {
+      "epoch": 1.3810096153846154,
+      "grad_norm": 0.2904583273208929,
+      "learning_rate": 2.9986642920748e-05,
+      "loss": 0.4636,
+      "step": 1149
+    },
+    {
+      "epoch": 1.3822115384615383,
+      "grad_norm": 0.2852373249883468,
+      "learning_rate": 2.996438112199466e-05,
+      "loss": 0.4259,
+      "step": 1150
+    },
+    {
+      "epoch": 1.3834134615384617,
+      "grad_norm": 0.2848052728533727,
+      "learning_rate": 2.994211932324132e-05,
+      "loss": 0.4516,
+      "step": 1151
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.28041237238575617,
+      "learning_rate": 2.9919857524487982e-05,
+      "loss": 0.4322,
+      "step": 1152
+    },
+    {
+      "epoch": 1.3858173076923077,
+      "grad_norm": 0.2547623804270529,
+      "learning_rate": 2.989759572573464e-05,
+      "loss": 0.415,
+      "step": 1153
+    },
+    {
+      "epoch": 1.3870192307692308,
+      "grad_norm": 0.3031734254817002,
+      "learning_rate": 2.9875333926981303e-05,
+      "loss": 0.4458,
+      "step": 1154
+    },
+    {
+      "epoch": 1.3882211538461537,
+      "grad_norm": 0.2940833653837902,
+      "learning_rate": 2.985307212822796e-05,
+      "loss": 0.4417,
+      "step": 1155
+    },
+    {
+      "epoch": 1.3894230769230769,
+      "grad_norm": 0.27908911277556453,
+      "learning_rate": 2.9830810329474623e-05,
+      "loss": 0.4162,
+      "step": 1156
+    },
+    {
+      "epoch": 1.390625,
+      "grad_norm": 0.30151709692256934,
+      "learning_rate": 2.9808548530721286e-05,
+      "loss": 0.4412,
+      "step": 1157
+    },
+    {
+      "epoch": 1.3918269230769231,
+      "grad_norm": 0.300800757173304,
+      "learning_rate": 2.9786286731967944e-05,
+      "loss": 0.4443,
+      "step": 1158
+    },
+    {
+      "epoch": 1.3930288461538463,
+      "grad_norm": 0.27307457253727974,
+      "learning_rate": 2.9764024933214606e-05,
+      "loss": 0.4566,
+      "step": 1159
+    },
+    {
+      "epoch": 1.3942307692307692,
+      "grad_norm": 0.3395399912613047,
+      "learning_rate": 2.9741763134461265e-05,
+      "loss": 0.4228,
+      "step": 1160
+    },
+    {
+      "epoch": 1.3954326923076923,
+      "grad_norm": 0.29350861688889457,
+      "learning_rate": 2.9719501335707927e-05,
+      "loss": 0.4306,
+      "step": 1161
+    },
+    {
+      "epoch": 1.3966346153846154,
+      "grad_norm": 0.2758412905354238,
+      "learning_rate": 2.9697239536954586e-05,
+      "loss": 0.4451,
+      "step": 1162
+    },
+    {
+      "epoch": 1.3978365384615383,
+      "grad_norm": 0.277789228540695,
+      "learning_rate": 2.9674977738201248e-05,
+      "loss": 0.4279,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3990384615384617,
+      "grad_norm": 0.29951318901047963,
+      "learning_rate": 2.965271593944791e-05,
+      "loss": 0.4023,
+      "step": 1164
+    },
+    {
+      "epoch": 1.4002403846153846,
+      "grad_norm": 0.3242707880523993,
+      "learning_rate": 2.963045414069457e-05,
+      "loss": 0.4401,
+      "step": 1165
+    },
+    {
+      "epoch": 1.4014423076923077,
+      "grad_norm": 0.29397875319775185,
+      "learning_rate": 2.960819234194123e-05,
+      "loss": 0.4144,
+      "step": 1166
+    },
+    {
+      "epoch": 1.4026442307692308,
+      "grad_norm": 0.34274467548074666,
+      "learning_rate": 2.958593054318789e-05,
+      "loss": 0.409,
+      "step": 1167
+    },
+    {
+      "epoch": 1.4038461538461537,
+      "grad_norm": 0.3305641210148041,
+      "learning_rate": 2.9563668744434552e-05,
+      "loss": 0.4391,
+      "step": 1168
+    },
+    {
+      "epoch": 1.4050480769230769,
+      "grad_norm": 0.30083208077997503,
+      "learning_rate": 2.954140694568121e-05,
+      "loss": 0.4259,
+      "step": 1169
+    },
+    {
+      "epoch": 1.40625,
+      "grad_norm": 0.3245029585163342,
+      "learning_rate": 2.9519145146927873e-05,
+      "loss": 0.4258,
+      "step": 1170
+    },
+    {
+      "epoch": 1.4074519230769231,
+      "grad_norm": 0.30915498127212965,
+      "learning_rate": 2.9496883348174535e-05,
+      "loss": 0.4319,
+      "step": 1171
+    },
+    {
+      "epoch": 1.4086538461538463,
+      "grad_norm": 0.3225788085005916,
+      "learning_rate": 2.9474621549421194e-05,
+      "loss": 0.4659,
+      "step": 1172
+    },
+    {
+      "epoch": 1.4098557692307692,
+      "grad_norm": 0.3511589343868233,
+      "learning_rate": 2.9452359750667856e-05,
+      "loss": 0.4666,
+      "step": 1173
+    },
+    {
+      "epoch": 1.4110576923076923,
+      "grad_norm": 0.29759436380668175,
+      "learning_rate": 2.9430097951914515e-05,
+      "loss": 0.4189,
+      "step": 1174
+    },
+    {
+      "epoch": 1.4122596153846154,
+      "grad_norm": 0.332234181313679,
+      "learning_rate": 2.9407836153161177e-05,
+      "loss": 0.4322,
+      "step": 1175
+    },
+    {
+      "epoch": 1.4134615384615383,
+      "grad_norm": 0.3021453930869036,
+      "learning_rate": 2.9385574354407836e-05,
+      "loss": 0.4118,
+      "step": 1176
+    },
+    {
+      "epoch": 1.4146634615384617,
+      "grad_norm": 0.36640600724659755,
+      "learning_rate": 2.9363312555654498e-05,
+      "loss": 0.4285,
+      "step": 1177
+    },
+    {
+      "epoch": 1.4158653846153846,
+      "grad_norm": 0.29832278178158667,
+      "learning_rate": 2.9341050756901163e-05,
+      "loss": 0.4381,
+      "step": 1178
+    },
+    {
+      "epoch": 1.4170673076923077,
+      "grad_norm": 0.32726997650002254,
+      "learning_rate": 2.931878895814782e-05,
+      "loss": 0.4188,
+      "step": 1179
+    },
+    {
+      "epoch": 1.4182692307692308,
+      "grad_norm": 0.30932907629322026,
+      "learning_rate": 2.929652715939448e-05,
+      "loss": 0.45,
+      "step": 1180
+    },
+    {
+      "epoch": 1.4194711538461537,
+      "grad_norm": 0.3442640555067361,
+      "learning_rate": 2.927426536064114e-05,
+      "loss": 0.4087,
+      "step": 1181
+    },
+    {
+      "epoch": 1.4206730769230769,
+      "grad_norm": 0.3493014204635096,
+      "learning_rate": 2.92520035618878e-05,
+      "loss": 0.4455,
+      "step": 1182
+    },
+    {
+      "epoch": 1.421875,
+      "grad_norm": 0.2822941891997059,
+      "learning_rate": 2.922974176313446e-05,
+      "loss": 0.4632,
+      "step": 1183
+    },
+    {
+      "epoch": 1.4230769230769231,
+      "grad_norm": 0.34930848117572827,
+      "learning_rate": 2.9207479964381123e-05,
+      "loss": 0.4467,
+      "step": 1184
+    },
+    {
+      "epoch": 1.4242788461538463,
+      "grad_norm": 0.32816895865607826,
+      "learning_rate": 2.9185218165627788e-05,
+      "loss": 0.4102,
+      "step": 1185
+    },
+    {
+      "epoch": 1.4254807692307692,
+      "grad_norm": 0.26926712701537986,
+      "learning_rate": 2.9162956366874443e-05,
+      "loss": 0.4204,
+      "step": 1186
+    },
+    {
+      "epoch": 1.4266826923076923,
+      "grad_norm": 0.3393038430218012,
+      "learning_rate": 2.914069456812111e-05,
+      "loss": 0.4226,
+      "step": 1187
+    },
+    {
+      "epoch": 1.4278846153846154,
+      "grad_norm": 0.4165345224491904,
+      "learning_rate": 2.9118432769367764e-05,
+      "loss": 0.4437,
+      "step": 1188
+    },
+    {
+      "epoch": 1.4290865384615383,
+      "grad_norm": 0.28542858673314614,
+      "learning_rate": 2.9096170970614426e-05,
+      "loss": 0.4442,
+      "step": 1189
+    },
+    {
+      "epoch": 1.4302884615384617,
+      "grad_norm": 0.33493085814858636,
+      "learning_rate": 2.9073909171861085e-05,
+      "loss": 0.4029,
+      "step": 1190
+    },
+    {
+      "epoch": 1.4314903846153846,
+      "grad_norm": 0.3415475573591766,
+      "learning_rate": 2.9051647373107747e-05,
+      "loss": 0.4058,
+      "step": 1191
+    },
+    {
+      "epoch": 1.4326923076923077,
+      "grad_norm": 0.3071858027193142,
+      "learning_rate": 2.9029385574354413e-05,
+      "loss": 0.4115,
+      "step": 1192
+    },
+    {
+      "epoch": 1.4338942307692308,
+      "grad_norm": 0.2955219403588275,
+      "learning_rate": 2.9007123775601068e-05,
+      "loss": 0.4504,
+      "step": 1193
+    },
+    {
+      "epoch": 1.4350961538461537,
+      "grad_norm": 0.3044399466063959,
+      "learning_rate": 2.8984861976847734e-05,
+      "loss": 0.4366,
+      "step": 1194
+    },
+    {
+      "epoch": 1.4362980769230769,
+      "grad_norm": 0.30240203681377337,
+      "learning_rate": 2.896260017809439e-05,
+      "loss": 0.4212,
+      "step": 1195
+    },
+    {
+      "epoch": 1.4375,
+      "grad_norm": 0.3031990120281369,
+      "learning_rate": 2.8940338379341055e-05,
+      "loss": 0.4407,
+      "step": 1196
+    },
+    {
+      "epoch": 1.4387019230769231,
+      "grad_norm": 0.336813516927253,
+      "learning_rate": 2.891807658058771e-05,
+      "loss": 0.4526,
+      "step": 1197
+    },
+    {
+      "epoch": 1.4399038461538463,
+      "grad_norm": 0.2584747193616262,
+      "learning_rate": 2.8895814781834375e-05,
+      "loss": 0.4419,
+      "step": 1198
+    },
+    {
+      "epoch": 1.4411057692307692,
+      "grad_norm": 0.3566857363943222,
+      "learning_rate": 2.8873552983081038e-05,
+      "loss": 0.4412,
+      "step": 1199
+    },
+    {
+      "epoch": 1.4423076923076923,
+      "grad_norm": 0.3156423861660561,
+      "learning_rate": 2.8851291184327693e-05,
+      "loss": 0.4683,
+      "step": 1200
+    },
+    {
+      "epoch": 1.4435096153846154,
+      "grad_norm": 0.31010922505462285,
+      "learning_rate": 2.882902938557436e-05,
+      "loss": 0.4214,
+      "step": 1201
+    },
+    {
+      "epoch": 1.4447115384615383,
+      "grad_norm": 0.28160997467333315,
+      "learning_rate": 2.8806767586821014e-05,
+      "loss": 0.4407,
+      "step": 1202
+    },
+    {
+      "epoch": 1.4459134615384617,
+      "grad_norm": 0.3287468285594419,
+      "learning_rate": 2.878450578806768e-05,
+      "loss": 0.4478,
+      "step": 1203
+    },
+    {
+      "epoch": 1.4471153846153846,
+      "grad_norm": 0.32304244448874714,
+      "learning_rate": 2.8762243989314335e-05,
+      "loss": 0.4121,
+      "step": 1204
+    },
+    {
+      "epoch": 1.4483173076923077,
+      "grad_norm": 0.3255090274239584,
+      "learning_rate": 2.8739982190561e-05,
+      "loss": 0.4562,
+      "step": 1205
+    },
+    {
+      "epoch": 1.4495192307692308,
+      "grad_norm": 0.3065717947263932,
+      "learning_rate": 2.8717720391807662e-05,
+      "loss": 0.4188,
+      "step": 1206
+    },
+    {
+      "epoch": 1.4507211538461537,
+      "grad_norm": 0.29327272943259164,
+      "learning_rate": 2.869545859305432e-05,
+      "loss": 0.4329,
+      "step": 1207
+    },
+    {
+      "epoch": 1.4519230769230769,
+      "grad_norm": 0.30133728406147237,
+      "learning_rate": 2.8673196794300983e-05,
+      "loss": 0.4288,
+      "step": 1208
+    },
+    {
+      "epoch": 1.453125,
+      "grad_norm": 0.2916216310400821,
+      "learning_rate": 2.8650934995547642e-05,
+      "loss": 0.4149,
+      "step": 1209
+    },
+    {
+      "epoch": 1.4543269230769231,
+      "grad_norm": 0.32148946230886905,
+      "learning_rate": 2.8628673196794304e-05,
+      "loss": 0.4518,
+      "step": 1210
+    },
+    {
+      "epoch": 1.4555288461538463,
+      "grad_norm": 0.2911579444198889,
+      "learning_rate": 2.860641139804096e-05,
+      "loss": 0.4288,
+      "step": 1211
+    },
+    {
+      "epoch": 1.4567307692307692,
+      "grad_norm": 0.3179493826778473,
+      "learning_rate": 2.8584149599287625e-05,
+      "loss": 0.4061,
+      "step": 1212
+    },
+    {
+      "epoch": 1.4579326923076923,
+      "grad_norm": 0.27150345185988883,
+      "learning_rate": 2.8561887800534287e-05,
+      "loss": 0.4182,
+      "step": 1213
+    },
+    {
+      "epoch": 1.4591346153846154,
+      "grad_norm": 0.27410749395888573,
+      "learning_rate": 2.8539626001780946e-05,
+      "loss": 0.4585,
+      "step": 1214
+    },
+    {
+      "epoch": 1.4603365384615383,
+      "grad_norm": 0.3423049901461306,
+      "learning_rate": 2.8517364203027608e-05,
+      "loss": 0.4309,
+      "step": 1215
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.25577688376504765,
+      "learning_rate": 2.8495102404274267e-05,
+      "loss": 0.4476,
+      "step": 1216
+    },
+    {
+      "epoch": 1.4627403846153846,
+      "grad_norm": 0.283737156797296,
+      "learning_rate": 2.847284060552093e-05,
+      "loss": 0.453,
+      "step": 1217
+    },
+    {
+      "epoch": 1.4639423076923077,
+      "grad_norm": 0.3292585162303327,
+      "learning_rate": 2.8450578806767588e-05,
+      "loss": 0.4339,
+      "step": 1218
+    },
+    {
+      "epoch": 1.4651442307692308,
+      "grad_norm": 0.29977962189345786,
+      "learning_rate": 2.842831700801425e-05,
+      "loss": 0.4485,
+      "step": 1219
+    },
+    {
+      "epoch": 1.4663461538461537,
+      "grad_norm": 0.2768877615623622,
+      "learning_rate": 2.8406055209260912e-05,
+      "loss": 0.4486,
+      "step": 1220
+    },
+    {
+      "epoch": 1.4675480769230769,
+      "grad_norm": 0.340349565826334,
+      "learning_rate": 2.838379341050757e-05,
+      "loss": 0.4444,
+      "step": 1221
+    },
+    {
+      "epoch": 1.46875,
+      "grad_norm": 0.2759582045963039,
+      "learning_rate": 2.8361531611754233e-05,
+      "loss": 0.4198,
+      "step": 1222
+    },
+    {
+      "epoch": 1.4699519230769231,
+      "grad_norm": 0.32899811951737823,
+      "learning_rate": 2.833926981300089e-05,
+      "loss": 0.471,
+      "step": 1223
+    },
+    {
+      "epoch": 1.4711538461538463,
+      "grad_norm": 0.31747509949450553,
+      "learning_rate": 2.8317008014247554e-05,
+      "loss": 0.441,
+      "step": 1224
+    },
+    {
+      "epoch": 1.4723557692307692,
+      "grad_norm": 0.33436626936537406,
+      "learning_rate": 2.8294746215494212e-05,
+      "loss": 0.4048,
+      "step": 1225
+    },
+    {
+      "epoch": 1.4735576923076923,
+      "grad_norm": 0.3362378910370095,
+      "learning_rate": 2.8272484416740874e-05,
+      "loss": 0.4235,
+      "step": 1226
+    },
+    {
+      "epoch": 1.4747596153846154,
+      "grad_norm": 0.3165501388577631,
+      "learning_rate": 2.8250222617987537e-05,
+      "loss": 0.4467,
+      "step": 1227
+    },
+    {
+      "epoch": 1.4759615384615383,
+      "grad_norm": 0.2930282913694221,
+      "learning_rate": 2.8227960819234195e-05,
+      "loss": 0.4338,
+      "step": 1228
+    },
+    {
+      "epoch": 1.4771634615384617,
+      "grad_norm": 0.29307576744465264,
+      "learning_rate": 2.8205699020480857e-05,
+      "loss": 0.4082,
+      "step": 1229
+    },
+    {
+      "epoch": 1.4783653846153846,
+      "grad_norm": 0.3631725254675727,
+      "learning_rate": 2.8183437221727516e-05,
+      "loss": 0.4102,
+      "step": 1230
+    },
+    {
+      "epoch": 1.4795673076923077,
+      "grad_norm": 0.32179570607862085,
+      "learning_rate": 2.816117542297418e-05,
+      "loss": 0.4448,
+      "step": 1231
+    },
+    {
+      "epoch": 1.4807692307692308,
+      "grad_norm": 0.33003341537429987,
+      "learning_rate": 2.8138913624220837e-05,
+      "loss": 0.4062,
+      "step": 1232
+    },
+    {
+      "epoch": 1.4819711538461537,
+      "grad_norm": 0.35491288138680704,
+      "learning_rate": 2.81166518254675e-05,
+      "loss": 0.4428,
+      "step": 1233
+    },
+    {
+      "epoch": 1.4831730769230769,
+      "grad_norm": 0.3254520625545464,
+      "learning_rate": 2.809439002671416e-05,
+      "loss": 0.4461,
+      "step": 1234
+    },
+    {
+      "epoch": 1.484375,
+      "grad_norm": 0.27288311383496333,
+      "learning_rate": 2.807212822796082e-05,
+      "loss": 0.418,
+      "step": 1235
+    },
+    {
+      "epoch": 1.4855769230769231,
+      "grad_norm": 0.3875792460943129,
+      "learning_rate": 2.8049866429207482e-05,
+      "loss": 0.4597,
+      "step": 1236
+    },
+    {
+      "epoch": 1.4867788461538463,
+      "grad_norm": 0.3994481286565239,
+      "learning_rate": 2.802760463045414e-05,
+      "loss": 0.427,
+      "step": 1237
+    },
+    {
+      "epoch": 1.4879807692307692,
+      "grad_norm": 0.3450324192842675,
+      "learning_rate": 2.8005342831700803e-05,
+      "loss": 0.4347,
+      "step": 1238
+    },
+    {
+      "epoch": 1.4891826923076923,
+      "grad_norm": 0.3423669191427863,
+      "learning_rate": 2.7983081032947462e-05,
+      "loss": 0.3947,
+      "step": 1239
+    },
+    {
+      "epoch": 1.4903846153846154,
+      "grad_norm": 0.4181878237474422,
+      "learning_rate": 2.7960819234194124e-05,
+      "loss": 0.4468,
+      "step": 1240
+    },
+    {
+      "epoch": 1.4915865384615383,
+      "grad_norm": 0.3089027537606574,
+      "learning_rate": 2.7938557435440786e-05,
+      "loss": 0.4195,
+      "step": 1241
+    },
+    {
+      "epoch": 1.4927884615384617,
+      "grad_norm": 0.4081352015746801,
+      "learning_rate": 2.7916295636687445e-05,
+      "loss": 0.4224,
+      "step": 1242
+    },
+    {
+      "epoch": 1.4939903846153846,
+      "grad_norm": 0.30749586925797223,
+      "learning_rate": 2.7894033837934107e-05,
+      "loss": 0.4421,
+      "step": 1243
+    },
+    {
+      "epoch": 1.4951923076923077,
+      "grad_norm": 0.31074586310161373,
+      "learning_rate": 2.7871772039180766e-05,
+      "loss": 0.4188,
+      "step": 1244
+    },
+    {
+      "epoch": 1.4963942307692308,
+      "grad_norm": 0.31664748564062,
+      "learning_rate": 2.7849510240427428e-05,
+      "loss": 0.4185,
+      "step": 1245
+    },
+    {
+      "epoch": 1.4975961538461537,
+      "grad_norm": 0.3481895638421032,
+      "learning_rate": 2.7827248441674087e-05,
+      "loss": 0.4193,
+      "step": 1246
+    },
+    {
+      "epoch": 1.4987980769230769,
+      "grad_norm": 0.29147090390018265,
+      "learning_rate": 2.780498664292075e-05,
+      "loss": 0.4141,
+      "step": 1247
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3286852691130483,
+      "learning_rate": 2.7782724844167414e-05,
+      "loss": 0.4498,
+      "step": 1248
+    },
+    {
+      "epoch": 1.5012019230769231,
+      "grad_norm": 0.3027425985983009,
+      "learning_rate": 2.776046304541407e-05,
+      "loss": 0.4197,
+      "step": 1249
+    },
+    {
+      "epoch": 1.5024038461538463,
+      "grad_norm": 0.3952222754231475,
+      "learning_rate": 2.7738201246660732e-05,
+      "loss": 0.4399,
+      "step": 1250
+    },
+    {
+      "epoch": 1.5036057692307692,
+      "grad_norm": 0.286967278241479,
+      "learning_rate": 2.771593944790739e-05,
+      "loss": 0.4357,
+      "step": 1251
+    },
+    {
+      "epoch": 1.5048076923076923,
+      "grad_norm": 0.28416640735594406,
+      "learning_rate": 2.7693677649154053e-05,
+      "loss": 0.4284,
+      "step": 1252
+    },
+    {
+      "epoch": 1.5060096153846154,
+      "grad_norm": 0.29152353516339796,
+      "learning_rate": 2.767141585040071e-05,
+      "loss": 0.4265,
+      "step": 1253
+    },
+    {
+      "epoch": 1.5072115384615383,
+      "grad_norm": 0.3021783116396041,
+      "learning_rate": 2.7649154051647374e-05,
+      "loss": 0.427,
+      "step": 1254
+    },
+    {
+      "epoch": 1.5084134615384617,
+      "grad_norm": 0.2578898315358026,
+      "learning_rate": 2.762689225289404e-05,
+      "loss": 0.435,
+      "step": 1255
+    },
+    {
+      "epoch": 1.5096153846153846,
+      "grad_norm": 0.30148057776062553,
+      "learning_rate": 2.7604630454140694e-05,
+      "loss": 0.4293,
+      "step": 1256
+    },
+    {
+      "epoch": 1.5108173076923077,
+      "grad_norm": 0.28225266958495016,
+      "learning_rate": 2.758236865538736e-05,
+      "loss": 0.4299,
+      "step": 1257
+    },
+    {
+      "epoch": 1.5120192307692308,
+      "grad_norm": 0.2814192371633315,
+      "learning_rate": 2.7560106856634015e-05,
+      "loss": 0.456,
+      "step": 1258
+    },
+    {
+      "epoch": 1.5132211538461537,
+      "grad_norm": 0.3086946877327328,
+      "learning_rate": 2.7537845057880677e-05,
+      "loss": 0.4379,
+      "step": 1259
+    },
+    {
+      "epoch": 1.5144230769230769,
+      "grad_norm": 0.2663157993207815,
+      "learning_rate": 2.7515583259127336e-05,
+      "loss": 0.4133,
+      "step": 1260
+    },
+    {
+      "epoch": 1.515625,
+      "grad_norm": 0.2881781612892015,
+      "learning_rate": 2.7493321460373998e-05,
+      "loss": 0.4132,
+      "step": 1261
+    },
+    {
+      "epoch": 1.5168269230769231,
+      "grad_norm": 0.3073395887537431,
+      "learning_rate": 2.7471059661620664e-05,
+      "loss": 0.4595,
+      "step": 1262
+    },
+    {
+      "epoch": 1.5180288461538463,
+      "grad_norm": 0.2660517482466701,
+      "learning_rate": 2.744879786286732e-05,
+      "loss": 0.4459,
+      "step": 1263
+    },
+    {
+      "epoch": 1.5192307692307692,
+      "grad_norm": 0.29013486259797816,
+      "learning_rate": 2.7426536064113985e-05,
+      "loss": 0.4332,
+      "step": 1264
+    },
+    {
+      "epoch": 1.5204326923076923,
+      "grad_norm": 0.2749754749935583,
+      "learning_rate": 2.740427426536064e-05,
+      "loss": 0.4384,
+      "step": 1265
+    },
+    {
+      "epoch": 1.5216346153846154,
+      "grad_norm": 0.2886297077314823,
+      "learning_rate": 2.7382012466607306e-05,
+      "loss": 0.4211,
+      "step": 1266
+    },
+    {
+      "epoch": 1.5228365384615383,
+      "grad_norm": 0.2792112183223728,
+      "learning_rate": 2.735975066785396e-05,
+      "loss": 0.4175,
+      "step": 1267
+    },
+    {
+      "epoch": 1.5240384615384617,
+      "grad_norm": 0.2985828971130069,
+      "learning_rate": 2.7337488869100626e-05,
+      "loss": 0.4269,
+      "step": 1268
+    },
+    {
+      "epoch": 1.5252403846153846,
+      "grad_norm": 0.3301726263635728,
+      "learning_rate": 2.731522707034729e-05,
+      "loss": 0.4399,
+      "step": 1269
+    },
+    {
+      "epoch": 1.5264423076923077,
+      "grad_norm": 0.31144091091174725,
+      "learning_rate": 2.7292965271593944e-05,
+      "loss": 0.4134,
+      "step": 1270
+    },
+    {
+      "epoch": 1.5276442307692308,
+      "grad_norm": 0.31962618611003835,
+      "learning_rate": 2.727070347284061e-05,
+      "loss": 0.435,
+      "step": 1271
+    },
+    {
+      "epoch": 1.5288461538461537,
+      "grad_norm": 0.2799094461698539,
+      "learning_rate": 2.7248441674087265e-05,
+      "loss": 0.4131,
+      "step": 1272
+    },
+    {
+      "epoch": 1.5300480769230769,
+      "grad_norm": 0.2890766227093793,
+      "learning_rate": 2.722617987533393e-05,
+      "loss": 0.4423,
+      "step": 1273
+    },
+    {
+      "epoch": 1.53125,
+      "grad_norm": 0.35054943674689104,
+      "learning_rate": 2.7203918076580586e-05,
+      "loss": 0.4124,
+      "step": 1274
+    },
+    {
+      "epoch": 1.5324519230769231,
+      "grad_norm": 0.29602078401392795,
+      "learning_rate": 2.718165627782725e-05,
+      "loss": 0.413,
+      "step": 1275
+    },
+    {
+      "epoch": 1.5336538461538463,
+      "grad_norm": 0.2879050922978971,
+      "learning_rate": 2.7159394479073913e-05,
+      "loss": 0.451,
+      "step": 1276
+    },
+    {
+      "epoch": 1.5348557692307692,
+      "grad_norm": 0.32700064777866206,
+      "learning_rate": 2.7137132680320572e-05,
+      "loss": 0.4339,
+      "step": 1277
+    },
+    {
+      "epoch": 1.5360576923076923,
+      "grad_norm": 0.2879818020502644,
+      "learning_rate": 2.7114870881567234e-05,
+      "loss": 0.4345,
+      "step": 1278
+    },
+    {
+      "epoch": 1.5372596153846154,
+      "grad_norm": 0.30694164205981367,
+      "learning_rate": 2.709260908281389e-05,
+      "loss": 0.4269,
+      "step": 1279
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.3155644385044049,
+      "learning_rate": 2.7070347284060555e-05,
+      "loss": 0.4038,
+      "step": 1280
+    },
+    {
+      "epoch": 1.5396634615384617,
+      "grad_norm": 0.30985149129492967,
+      "learning_rate": 2.704808548530721e-05,
+      "loss": 0.4267,
+      "step": 1281
+    },
+    {
+      "epoch": 1.5408653846153846,
+      "grad_norm": 0.30856651787014305,
+      "learning_rate": 2.7025823686553876e-05,
+      "loss": 0.4346,
+      "step": 1282
+    },
+    {
+      "epoch": 1.5420673076923077,
+      "grad_norm": 0.32156102711876344,
+      "learning_rate": 2.7003561887800538e-05,
+      "loss": 0.4344,
+      "step": 1283
+    },
+    {
+      "epoch": 1.5432692307692308,
+      "grad_norm": 0.3405733175960536,
+      "learning_rate": 2.6981300089047197e-05,
+      "loss": 0.4343,
+      "step": 1284
+    },
+    {
+      "epoch": 1.5444711538461537,
+      "grad_norm": 0.29954629361233187,
+      "learning_rate": 2.695903829029386e-05,
+      "loss": 0.4378,
+      "step": 1285
+    },
+    {
+      "epoch": 1.5456730769230769,
+      "grad_norm": 0.4068733170103639,
+      "learning_rate": 2.6936776491540518e-05,
+      "loss": 0.4244,
+      "step": 1286
+    },
+    {
+      "epoch": 1.546875,
+      "grad_norm": 0.27175712921865136,
+      "learning_rate": 2.691451469278718e-05,
+      "loss": 0.4104,
+      "step": 1287
+    },
+    {
+      "epoch": 1.5480769230769231,
+      "grad_norm": 0.3734187931242619,
+      "learning_rate": 2.689225289403384e-05,
+      "loss": 0.3978,
+      "step": 1288
+    },
+    {
+      "epoch": 1.5492788461538463,
+      "grad_norm": 0.4010620314042549,
+      "learning_rate": 2.68699910952805e-05,
+      "loss": 0.4334,
+      "step": 1289
+    },
+    {
+      "epoch": 1.5504807692307692,
+      "grad_norm": 0.2757316661772163,
+      "learning_rate": 2.6847729296527163e-05,
+      "loss": 0.4204,
+      "step": 1290
+    },
+    {
+      "epoch": 1.5516826923076923,
+      "grad_norm": 0.3729804552524698,
+      "learning_rate": 2.682546749777382e-05,
+      "loss": 0.44,
+      "step": 1291
+    },
+    {
+      "epoch": 1.5528846153846154,
+      "grad_norm": 0.35064633120067357,
+      "learning_rate": 2.6803205699020484e-05,
+      "loss": 0.4156,
+      "step": 1292
+    },
+    {
+      "epoch": 1.5540865384615383,
+      "grad_norm": 0.3061438592572993,
+      "learning_rate": 2.6780943900267142e-05,
+      "loss": 0.4208,
+      "step": 1293
+    },
+    {
+      "epoch": 1.5552884615384617,
+      "grad_norm": 0.3134931364295119,
+      "learning_rate": 2.6758682101513805e-05,
+      "loss": 0.4243,
+      "step": 1294
+    },
+    {
+      "epoch": 1.5564903846153846,
+      "grad_norm": 0.3277310181641551,
+      "learning_rate": 2.6736420302760463e-05,
+      "loss": 0.4344,
+      "step": 1295
+    },
+    {
+      "epoch": 1.5576923076923077,
+      "grad_norm": 0.2848692219808813,
+      "learning_rate": 2.6714158504007125e-05,
+      "loss": 0.4352,
+      "step": 1296
+    },
+    {
+      "epoch": 1.5588942307692308,
+      "grad_norm": 0.2825715876194632,
+      "learning_rate": 2.6691896705253788e-05,
+      "loss": 0.4243,
+      "step": 1297
+    },
+    {
+      "epoch": 1.5600961538461537,
+      "grad_norm": 0.3693356409924469,
+      "learning_rate": 2.6669634906500446e-05,
+      "loss": 0.4349,
+      "step": 1298
+    },
+    {
+      "epoch": 1.5612980769230769,
+      "grad_norm": 0.2821862191293658,
+      "learning_rate": 2.664737310774711e-05,
+      "loss": 0.4136,
+      "step": 1299
+    },
+    {
+      "epoch": 1.5625,
+      "grad_norm": 0.27746061961843826,
+      "learning_rate": 2.6625111308993767e-05,
+      "loss": 0.4171,
+      "step": 1300
+    },
+    {
+      "epoch": 1.5637019230769231,
+      "grad_norm": 0.2763962881657365,
+      "learning_rate": 2.660284951024043e-05,
+      "loss": 0.4407,
+      "step": 1301
+    },
+    {
+      "epoch": 1.5649038461538463,
+      "grad_norm": 0.34885169253504367,
+      "learning_rate": 2.6580587711487088e-05,
+      "loss": 0.4312,
+      "step": 1302
+    },
+    {
+      "epoch": 1.5661057692307692,
+      "grad_norm": 0.3501141429841623,
+      "learning_rate": 2.655832591273375e-05,
+      "loss": 0.4062,
+      "step": 1303
+    },
+    {
+      "epoch": 1.5673076923076923,
+      "grad_norm": 0.30760418667900946,
+      "learning_rate": 2.6536064113980412e-05,
+      "loss": 0.4232,
+      "step": 1304
+    },
+    {
+      "epoch": 1.5685096153846154,
+      "grad_norm": 0.43655373481488663,
+      "learning_rate": 2.651380231522707e-05,
+      "loss": 0.4293,
+      "step": 1305
+    },
+    {
+      "epoch": 1.5697115384615383,
+      "grad_norm": 0.36764588345244065,
+      "learning_rate": 2.6491540516473733e-05,
+      "loss": 0.4508,
+      "step": 1306
+    },
+    {
+      "epoch": 1.5709134615384617,
+      "grad_norm": 0.3909425444693062,
+      "learning_rate": 2.6469278717720392e-05,
+      "loss": 0.4436,
+      "step": 1307
+    },
+    {
+      "epoch": 1.5721153846153846,
+      "grad_norm": 0.326501977225342,
+      "learning_rate": 2.6447016918967054e-05,
+      "loss": 0.4396,
+      "step": 1308
+    },
+    {
+      "epoch": 1.5733173076923077,
+      "grad_norm": 0.32607104575581164,
+      "learning_rate": 2.6424755120213713e-05,
+      "loss": 0.409,
+      "step": 1309
+    },
+    {
+      "epoch": 1.5745192307692308,
+      "grad_norm": 0.3824369789511727,
+      "learning_rate": 2.6402493321460375e-05,
+      "loss": 0.4294,
+      "step": 1310
+    },
+    {
+      "epoch": 1.5757211538461537,
+      "grad_norm": 0.3620118276919301,
+      "learning_rate": 2.6380231522707037e-05,
+      "loss": 0.4281,
+      "step": 1311
+    },
+    {
+      "epoch": 1.5769230769230769,
+      "grad_norm": 0.2773460185244286,
+      "learning_rate": 2.6357969723953696e-05,
+      "loss": 0.4138,
+      "step": 1312
+    },
+    {
+      "epoch": 1.578125,
+      "grad_norm": 0.3319519036755422,
+      "learning_rate": 2.6335707925200358e-05,
+      "loss": 0.4403,
+      "step": 1313
+    },
+    {
+      "epoch": 1.5793269230769231,
+      "grad_norm": 0.3072477577249634,
+      "learning_rate": 2.6313446126447017e-05,
+      "loss": 0.4189,
+      "step": 1314
+    },
+    {
+      "epoch": 1.5805288461538463,
+      "grad_norm": 0.3068280496490125,
+      "learning_rate": 2.629118432769368e-05,
+      "loss": 0.453,
+      "step": 1315
+    },
+    {
+      "epoch": 1.5817307692307692,
+      "grad_norm": 0.32134628113681163,
+      "learning_rate": 2.6268922528940338e-05,
+      "loss": 0.4235,
+      "step": 1316
+    },
+    {
+      "epoch": 1.5829326923076923,
+      "grad_norm": 0.287421337834051,
+      "learning_rate": 2.6246660730187e-05,
+      "loss": 0.4379,
+      "step": 1317
+    },
+    {
+      "epoch": 1.5841346153846154,
+      "grad_norm": 0.2864650110455563,
+      "learning_rate": 2.6224398931433662e-05,
+      "loss": 0.4343,
+      "step": 1318
+    },
+    {
+      "epoch": 1.5853365384615383,
+      "grad_norm": 0.307384581153055,
+      "learning_rate": 2.620213713268032e-05,
+      "loss": 0.4067,
+      "step": 1319
+    },
+    {
+      "epoch": 1.5865384615384617,
+      "grad_norm": 0.27143152096021783,
+      "learning_rate": 2.6179875333926983e-05,
+      "loss": 0.4377,
+      "step": 1320
+    },
+    {
+      "epoch": 1.5877403846153846,
+      "grad_norm": 0.2826837643085764,
+      "learning_rate": 2.615761353517364e-05,
+      "loss": 0.4216,
+      "step": 1321
+    },
+    {
+      "epoch": 1.5889423076923077,
+      "grad_norm": 0.3091833790407276,
+      "learning_rate": 2.6135351736420304e-05,
+      "loss": 0.4146,
+      "step": 1322
+    },
+    {
+      "epoch": 1.5901442307692308,
+      "grad_norm": 0.27311201275703706,
+      "learning_rate": 2.6113089937666962e-05,
+      "loss": 0.4257,
+      "step": 1323
+    },
+    {
+      "epoch": 1.5913461538461537,
+      "grad_norm": 0.28584249915078963,
+      "learning_rate": 2.6090828138913624e-05,
+      "loss": 0.4409,
+      "step": 1324
+    },
+    {
+      "epoch": 1.5925480769230769,
+      "grad_norm": 0.26289502415363597,
+      "learning_rate": 2.606856634016029e-05,
+      "loss": 0.3889,
+      "step": 1325
+    },
+    {
+      "epoch": 1.59375,
+      "grad_norm": 0.28248612574836096,
+      "learning_rate": 2.6046304541406945e-05,
+      "loss": 0.4123,
+      "step": 1326
+    },
+    {
+      "epoch": 1.5949519230769231,
+      "grad_norm": 0.26718516096260037,
+      "learning_rate": 2.602404274265361e-05,
+      "loss": 0.4315,
+      "step": 1327
+    },
+    {
+      "epoch": 1.5961538461538463,
+      "grad_norm": 0.28715194249467657,
+      "learning_rate": 2.6001780943900266e-05,
+      "loss": 0.4646,
+      "step": 1328
+    },
+    {
+      "epoch": 1.5973557692307692,
+      "grad_norm": 0.30771001634491574,
+      "learning_rate": 2.597951914514693e-05,
+      "loss": 0.4355,
+      "step": 1329
+    },
+    {
+      "epoch": 1.5985576923076923,
+      "grad_norm": 0.3123745087452568,
+      "learning_rate": 2.5957257346393587e-05,
+      "loss": 0.4499,
+      "step": 1330
+    },
+    {
+      "epoch": 1.5997596153846154,
+      "grad_norm": 0.2616461757586271,
+      "learning_rate": 2.593499554764025e-05,
+      "loss": 0.4394,
+      "step": 1331
+    },
+    {
+      "epoch": 1.6009615384615383,
+      "grad_norm": 0.2834106649636486,
+      "learning_rate": 2.5912733748886915e-05,
+      "loss": 0.4588,
+      "step": 1332
+    },
+    {
+      "epoch": 1.6021634615384617,
+      "grad_norm": 0.26603745542873936,
+      "learning_rate": 2.589047195013357e-05,
+      "loss": 0.4592,
+      "step": 1333
+    },
+    {
+      "epoch": 1.6033653846153846,
+      "grad_norm": 0.27454551142081907,
+      "learning_rate": 2.5868210151380236e-05,
+      "loss": 0.4263,
+      "step": 1334
+    },
+    {
+      "epoch": 1.6045673076923077,
+      "grad_norm": 0.3179285667848208,
+      "learning_rate": 2.584594835262689e-05,
+      "loss": 0.4448,
+      "step": 1335
+    },
+    {
+      "epoch": 1.6057692307692308,
+      "grad_norm": 0.25047801351236515,
+      "learning_rate": 2.5823686553873557e-05,
+      "loss": 0.4199,
+      "step": 1336
+    },
+    {
+      "epoch": 1.6069711538461537,
+      "grad_norm": 0.2516608578793464,
+      "learning_rate": 2.5801424755120212e-05,
+      "loss": 0.4156,
+      "step": 1337
+    },
+    {
+      "epoch": 1.6081730769230769,
+      "grad_norm": 0.2813676122334217,
+      "learning_rate": 2.5779162956366877e-05,
+      "loss": 0.4361,
+      "step": 1338
+    },
+    {
+      "epoch": 1.609375,
+      "grad_norm": 0.2698484916496981,
+      "learning_rate": 2.575690115761354e-05,
+      "loss": 0.448,
+      "step": 1339
+    },
+    {
+      "epoch": 1.6105769230769231,
+      "grad_norm": 0.24845565556963542,
+      "learning_rate": 2.5734639358860195e-05,
+      "loss": 0.3829,
+      "step": 1340
+    },
+    {
+      "epoch": 1.6117788461538463,
+      "grad_norm": 0.2902389216520786,
+      "learning_rate": 2.571237756010686e-05,
+      "loss": 0.4489,
+      "step": 1341
+    },
+    {
+      "epoch": 1.6129807692307692,
+      "grad_norm": 0.31361557070095347,
+      "learning_rate": 2.5690115761353516e-05,
+      "loss": 0.4282,
+      "step": 1342
+    },
+    {
+      "epoch": 1.6141826923076923,
+      "grad_norm": 0.5918936650461452,
+      "learning_rate": 2.566785396260018e-05,
+      "loss": 0.4331,
+      "step": 1343
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.28070693769951044,
+      "learning_rate": 2.5645592163846837e-05,
+      "loss": 0.4534,
+      "step": 1344
+    },
+    {
+      "epoch": 1.6165865384615383,
+      "grad_norm": 0.3491366187977565,
+      "learning_rate": 2.5623330365093502e-05,
+      "loss": 0.4088,
+      "step": 1345
+    },
+    {
+      "epoch": 1.6177884615384617,
+      "grad_norm": 0.27969501360013105,
+      "learning_rate": 2.5601068566340164e-05,
+      "loss": 0.4385,
+      "step": 1346
+    },
+    {
+      "epoch": 1.6189903846153846,
+      "grad_norm": 0.3025525203828094,
+      "learning_rate": 2.5578806767586823e-05,
+      "loss": 0.4501,
+      "step": 1347
+    },
+    {
+      "epoch": 1.6201923076923077,
+      "grad_norm": 0.3337517525556035,
+      "learning_rate": 2.5556544968833485e-05,
+      "loss": 0.4265,
+      "step": 1348
+    },
+    {
+      "epoch": 1.6213942307692308,
+      "grad_norm": 0.3096245136067253,
+      "learning_rate": 2.553428317008014e-05,
+      "loss": 0.4244,
+      "step": 1349
+    },
+    {
+      "epoch": 1.6225961538461537,
+      "grad_norm": 0.3170107898675398,
+      "learning_rate": 2.5512021371326806e-05,
+      "loss": 0.42,
+      "step": 1350
+    },
+    {
+      "epoch": 1.6237980769230769,
+      "grad_norm": 0.3066146800746244,
+      "learning_rate": 2.548975957257346e-05,
+      "loss": 0.4296,
+      "step": 1351
+    },
+    {
+      "epoch": 1.625,
+      "grad_norm": 0.364925070388707,
+      "learning_rate": 2.5467497773820127e-05,
+      "loss": 0.4719,
+      "step": 1352
+    },
+    {
+      "epoch": 1.6262019230769231,
+      "grad_norm": 0.2665757820218757,
+      "learning_rate": 2.544523597506679e-05,
+      "loss": 0.3969,
+      "step": 1353
+    },
+    {
+      "epoch": 1.6274038461538463,
+      "grad_norm": 0.41900725765309954,
+      "learning_rate": 2.5422974176313448e-05,
+      "loss": 0.424,
+      "step": 1354
+    },
+    {
+      "epoch": 1.6286057692307692,
+      "grad_norm": 0.32221626431326333,
+      "learning_rate": 2.540071237756011e-05,
+      "loss": 0.4176,
+      "step": 1355
+    },
+    {
+      "epoch": 1.6298076923076923,
+      "grad_norm": 0.3295861654051597,
+      "learning_rate": 2.537845057880677e-05,
+      "loss": 0.3905,
+      "step": 1356
+    },
+    {
+      "epoch": 1.6310096153846154,
+      "grad_norm": 0.3117528502641533,
+      "learning_rate": 2.535618878005343e-05,
+      "loss": 0.4496,
+      "step": 1357
+    },
+    {
+      "epoch": 1.6322115384615383,
+      "grad_norm": 0.31160129705636774,
+      "learning_rate": 2.533392698130009e-05,
+      "loss": 0.3996,
+      "step": 1358
+    },
+    {
+      "epoch": 1.6334134615384617,
+      "grad_norm": 0.2815837262289157,
+      "learning_rate": 2.531166518254675e-05,
+      "loss": 0.3972,
+      "step": 1359
+    },
+    {
+      "epoch": 1.6346153846153846,
+      "grad_norm": 0.2958978887675882,
+      "learning_rate": 2.5289403383793414e-05,
+      "loss": 0.44,
+      "step": 1360
+    },
+    {
+      "epoch": 1.6358173076923077,
+      "grad_norm": 0.32917855956129194,
+      "learning_rate": 2.5267141585040073e-05,
+      "loss": 0.4108,
+      "step": 1361
+    },
+    {
+      "epoch": 1.6370192307692308,
+      "grad_norm": 0.28650893391379273,
+      "learning_rate": 2.5244879786286735e-05,
+      "loss": 0.4543,
+      "step": 1362
+    },
+    {
+      "epoch": 1.6382211538461537,
+      "grad_norm": 0.3428482796444057,
+      "learning_rate": 2.5222617987533393e-05,
+      "loss": 0.4297,
+      "step": 1363
+    },
+    {
+      "epoch": 1.6394230769230769,
+      "grad_norm": 0.30753627364742037,
+      "learning_rate": 2.5200356188780056e-05,
+      "loss": 0.4296,
+      "step": 1364
+    },
+    {
+      "epoch": 1.640625,
+      "grad_norm": 0.3434642747238102,
+      "learning_rate": 2.5178094390026714e-05,
+      "loss": 0.4357,
+      "step": 1365
+    },
+    {
+      "epoch": 1.6418269230769231,
+      "grad_norm": 0.32645890517339476,
+      "learning_rate": 2.5155832591273376e-05,
+      "loss": 0.4236,
+      "step": 1366
+    },
+    {
+      "epoch": 1.6430288461538463,
+      "grad_norm": 0.34499796489370904,
+      "learning_rate": 2.513357079252004e-05,
+      "loss": 0.4423,
+      "step": 1367
+    },
+    {
+      "epoch": 1.6442307692307692,
+      "grad_norm": 0.3337197080165474,
+      "learning_rate": 2.5111308993766697e-05,
+      "loss": 0.4057,
+      "step": 1368
+    },
+    {
+      "epoch": 1.6454326923076923,
+      "grad_norm": 0.28940746821962127,
+      "learning_rate": 2.508904719501336e-05,
+      "loss": 0.4173,
+      "step": 1369
+    },
+    {
+      "epoch": 1.6466346153846154,
+      "grad_norm": 0.3695433022638391,
+      "learning_rate": 2.5066785396260018e-05,
+      "loss": 0.4073,
+      "step": 1370
+    },
+    {
+      "epoch": 1.6478365384615383,
+      "grad_norm": 0.33831512159363053,
+      "learning_rate": 2.504452359750668e-05,
+      "loss": 0.4123,
+      "step": 1371
+    },
+    {
+      "epoch": 1.6490384615384617,
+      "grad_norm": 0.34969200443448983,
+      "learning_rate": 2.502226179875334e-05,
+      "loss": 0.4279,
+      "step": 1372
+    },
+    {
+      "epoch": 1.6502403846153846,
+      "grad_norm": 0.3950502090043345,
+      "learning_rate": 2.5e-05,
+      "loss": 0.4553,
+      "step": 1373
+    },
+    {
+      "epoch": 1.6514423076923077,
+      "grad_norm": 0.38397410903861323,
+      "learning_rate": 2.497773820124666e-05,
+      "loss": 0.4126,
+      "step": 1374
+    },
+    {
+      "epoch": 1.6526442307692308,
+      "grad_norm": 0.27966308831874404,
+      "learning_rate": 2.4955476402493322e-05,
+      "loss": 0.4144,
+      "step": 1375
+    },
+    {
+      "epoch": 1.6538461538461537,
+      "grad_norm": 0.3426550783246446,
+      "learning_rate": 2.493321460373998e-05,
+      "loss": 0.4342,
+      "step": 1376
+    },
+    {
+      "epoch": 1.6550480769230769,
+      "grad_norm": 0.33247865296025647,
+      "learning_rate": 2.4910952804986646e-05,
+      "loss": 0.4161,
+      "step": 1377
+    },
+    {
+      "epoch": 1.65625,
+      "grad_norm": 0.30342665384338174,
+      "learning_rate": 2.4888691006233305e-05,
+      "loss": 0.4464,
+      "step": 1378
+    },
+    {
+      "epoch": 1.6574519230769231,
+      "grad_norm": 0.31838941753326877,
+      "learning_rate": 2.4866429207479967e-05,
+      "loss": 0.4307,
+      "step": 1379
+    },
+    {
+      "epoch": 1.6586538461538463,
+      "grad_norm": 0.3189492556465253,
+      "learning_rate": 2.4844167408726626e-05,
+      "loss": 0.4027,
+      "step": 1380
+    },
+    {
+      "epoch": 1.6598557692307692,
+      "grad_norm": 0.3493257280565083,
+      "learning_rate": 2.4821905609973288e-05,
+      "loss": 0.4228,
+      "step": 1381
+    },
+    {
+      "epoch": 1.6610576923076923,
+      "grad_norm": 0.30712968727640877,
+      "learning_rate": 2.4799643811219947e-05,
+      "loss": 0.4344,
+      "step": 1382
+    },
+    {
+      "epoch": 1.6622596153846154,
+      "grad_norm": 0.32494191590855986,
+      "learning_rate": 2.477738201246661e-05,
+      "loss": 0.4475,
+      "step": 1383
+    },
+    {
+      "epoch": 1.6634615384615383,
+      "grad_norm": 0.28601212570995926,
+      "learning_rate": 2.475512021371327e-05,
+      "loss": 0.4164,
+      "step": 1384
+    },
+    {
+      "epoch": 1.6646634615384617,
+      "grad_norm": 0.3341770497177679,
+      "learning_rate": 2.473285841495993e-05,
+      "loss": 0.4316,
+      "step": 1385
+    },
+    {
+      "epoch": 1.6658653846153846,
+      "grad_norm": 0.33160658406751764,
+      "learning_rate": 2.4710596616206592e-05,
+      "loss": 0.462,
+      "step": 1386
+    },
+    {
+      "epoch": 1.6670673076923077,
+      "grad_norm": 0.29169388980128064,
+      "learning_rate": 2.468833481745325e-05,
+      "loss": 0.4142,
+      "step": 1387
+    },
+    {
+      "epoch": 1.6682692307692308,
+      "grad_norm": 0.2991978868384951,
+      "learning_rate": 2.4666073018699913e-05,
+      "loss": 0.4302,
+      "step": 1388
+    },
+    {
+      "epoch": 1.6694711538461537,
+      "grad_norm": 0.33627114662039986,
+      "learning_rate": 2.464381121994657e-05,
+      "loss": 0.4311,
+      "step": 1389
+    },
+    {
+      "epoch": 1.6706730769230769,
+      "grad_norm": 0.3576767042718142,
+      "learning_rate": 2.4621549421193234e-05,
+      "loss": 0.4072,
+      "step": 1390
+    },
+    {
+      "epoch": 1.671875,
+      "grad_norm": 0.30378277633947387,
+      "learning_rate": 2.4599287622439896e-05,
+      "loss": 0.4594,
+      "step": 1391
+    },
+    {
+      "epoch": 1.6730769230769231,
+      "grad_norm": 0.3448484379869248,
+      "learning_rate": 2.4577025823686555e-05,
+      "loss": 0.4277,
+      "step": 1392
+    },
+    {
+      "epoch": 1.6742788461538463,
+      "grad_norm": 0.3107369692436639,
+      "learning_rate": 2.4554764024933217e-05,
+      "loss": 0.4155,
+      "step": 1393
+    },
+    {
+      "epoch": 1.6754807692307692,
+      "grad_norm": 0.30820269704479025,
+      "learning_rate": 2.4532502226179875e-05,
+      "loss": 0.4236,
+      "step": 1394
+    },
+    {
+      "epoch": 1.6766826923076923,
+      "grad_norm": 0.2834700962654255,
+      "learning_rate": 2.4510240427426538e-05,
+      "loss": 0.4197,
+      "step": 1395
+    },
+    {
+      "epoch": 1.6778846153846154,
+      "grad_norm": 0.3509975755532328,
+      "learning_rate": 2.4487978628673196e-05,
+      "loss": 0.4602,
+      "step": 1396
+    },
+    {
+      "epoch": 1.6790865384615383,
+      "grad_norm": 0.27465505697079035,
+      "learning_rate": 2.446571682991986e-05,
+      "loss": 0.3974,
+      "step": 1397
+    },
+    {
+      "epoch": 1.6802884615384617,
+      "grad_norm": 0.2834798380146236,
+      "learning_rate": 2.444345503116652e-05,
+      "loss": 0.4378,
+      "step": 1398
+    },
+    {
+      "epoch": 1.6814903846153846,
+      "grad_norm": 0.35226395132886607,
+      "learning_rate": 2.442119323241318e-05,
+      "loss": 0.4153,
+      "step": 1399
+    },
+    {
+      "epoch": 1.6826923076923077,
+      "grad_norm": 0.2626871365993348,
+      "learning_rate": 2.439893143365984e-05,
+      "loss": 0.4109,
+      "step": 1400
+    },
+    {
+      "epoch": 1.6838942307692308,
+      "grad_norm": 0.2930833119425354,
+      "learning_rate": 2.43766696349065e-05,
+      "loss": 0.4296,
+      "step": 1401
+    },
+    {
+      "epoch": 1.6850961538461537,
+      "grad_norm": 0.2870206146678828,
+      "learning_rate": 2.4354407836153162e-05,
+      "loss": 0.4132,
+      "step": 1402
+    },
+    {
+      "epoch": 1.6862980769230769,
+      "grad_norm": 0.27480399545256773,
+      "learning_rate": 2.433214603739982e-05,
+      "loss": 0.4198,
+      "step": 1403
+    },
+    {
+      "epoch": 1.6875,
+      "grad_norm": 0.2637662122761455,
+      "learning_rate": 2.4309884238646483e-05,
+      "loss": 0.4169,
+      "step": 1404
+    },
+    {
+      "epoch": 1.6887019230769231,
+      "grad_norm": 0.25535023909364846,
+      "learning_rate": 2.4287622439893145e-05,
+      "loss": 0.4065,
+      "step": 1405
+    },
+    {
+      "epoch": 1.6899038461538463,
+      "grad_norm": 0.28561240404339355,
+      "learning_rate": 2.4265360641139808e-05,
+      "loss": 0.4336,
+      "step": 1406
+    },
+    {
+      "epoch": 1.6911057692307692,
+      "grad_norm": 0.26163969328892844,
+      "learning_rate": 2.4243098842386466e-05,
+      "loss": 0.4154,
+      "step": 1407
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.2875756596169041,
+      "learning_rate": 2.4220837043633125e-05,
+      "loss": 0.3931,
+      "step": 1408
+    },
+    {
+      "epoch": 1.6935096153846154,
+      "grad_norm": 0.2838888348853831,
+      "learning_rate": 2.4198575244879787e-05,
+      "loss": 0.4654,
+      "step": 1409
+    },
+    {
+      "epoch": 1.6947115384615383,
+      "grad_norm": 0.25846067985454335,
+      "learning_rate": 2.4176313446126446e-05,
+      "loss": 0.4285,
+      "step": 1410
+    },
+    {
+      "epoch": 1.6959134615384617,
+      "grad_norm": 0.2883018433388527,
+      "learning_rate": 2.4154051647373108e-05,
+      "loss": 0.4174,
+      "step": 1411
+    },
+    {
+      "epoch": 1.6971153846153846,
+      "grad_norm": 0.26537005761307847,
+      "learning_rate": 2.413178984861977e-05,
+      "loss": 0.392,
+      "step": 1412
+    },
+    {
+      "epoch": 1.6983173076923077,
+      "grad_norm": 0.26151671398005844,
+      "learning_rate": 2.4109528049866432e-05,
+      "loss": 0.4054,
+      "step": 1413
+    },
+    {
+      "epoch": 1.6995192307692308,
+      "grad_norm": 0.26087157587526727,
+      "learning_rate": 2.408726625111309e-05,
+      "loss": 0.4126,
+      "step": 1414
+    },
+    {
+      "epoch": 1.7007211538461537,
+      "grad_norm": 0.30675874181185353,
+      "learning_rate": 2.4065004452359753e-05,
+      "loss": 0.4051,
+      "step": 1415
+    },
+    {
+      "epoch": 1.7019230769230769,
+      "grad_norm": 0.25903113345998197,
+      "learning_rate": 2.4042742653606412e-05,
+      "loss": 0.4061,
+      "step": 1416
+    },
+    {
+      "epoch": 1.703125,
+      "grad_norm": 0.28909242631546994,
+      "learning_rate": 2.4020480854853074e-05,
+      "loss": 0.4395,
+      "step": 1417
+    },
+    {
+      "epoch": 1.7043269230769231,
+      "grad_norm": 0.277905715463366,
+      "learning_rate": 2.3998219056099733e-05,
+      "loss": 0.4522,
+      "step": 1418
+    },
+    {
+      "epoch": 1.7055288461538463,
+      "grad_norm": 0.3028563881220706,
+      "learning_rate": 2.3975957257346395e-05,
+      "loss": 0.4373,
+      "step": 1419
+    },
+    {
+      "epoch": 1.7067307692307692,
+      "grad_norm": 0.25826846988371593,
+      "learning_rate": 2.3953695458593057e-05,
+      "loss": 0.4433,
+      "step": 1420
+    },
+    {
+      "epoch": 1.7079326923076923,
+      "grad_norm": 0.25586795169219034,
+      "learning_rate": 2.3931433659839716e-05,
+      "loss": 0.4062,
+      "step": 1421
+    },
+    {
+      "epoch": 1.7091346153846154,
+      "grad_norm": 0.27197484445494907,
+      "learning_rate": 2.3909171861086378e-05,
+      "loss": 0.4069,
+      "step": 1422
+    },
+    {
+      "epoch": 1.7103365384615383,
+      "grad_norm": 0.2801067062015412,
+      "learning_rate": 2.3886910062333037e-05,
+      "loss": 0.446,
+      "step": 1423
+    },
+    {
+      "epoch": 1.7115384615384617,
+      "grad_norm": 0.262853630307317,
+      "learning_rate": 2.38646482635797e-05,
+      "loss": 0.4089,
+      "step": 1424
+    },
+    {
+      "epoch": 1.7127403846153846,
+      "grad_norm": 0.3147658949078302,
+      "learning_rate": 2.3842386464826358e-05,
+      "loss": 0.4208,
+      "step": 1425
+    },
+    {
+      "epoch": 1.7139423076923077,
+      "grad_norm": 0.290727005800667,
+      "learning_rate": 2.382012466607302e-05,
+      "loss": 0.4319,
+      "step": 1426
+    },
+    {
+      "epoch": 1.7151442307692308,
+      "grad_norm": 0.29484878489171196,
+      "learning_rate": 2.3797862867319682e-05,
+      "loss": 0.428,
+      "step": 1427
+    },
+    {
+      "epoch": 1.7163461538461537,
+      "grad_norm": 0.2963372363128777,
+      "learning_rate": 2.377560106856634e-05,
+      "loss": 0.4146,
+      "step": 1428
+    },
+    {
+      "epoch": 1.7175480769230769,
+      "grad_norm": 0.30519282381894264,
+      "learning_rate": 2.3753339269813003e-05,
+      "loss": 0.4499,
+      "step": 1429
+    },
+    {
+      "epoch": 1.71875,
+      "grad_norm": 0.3137747859732027,
+      "learning_rate": 2.373107747105966e-05,
+      "loss": 0.3926,
+      "step": 1430
+    },
+    {
+      "epoch": 1.7199519230769231,
+      "grad_norm": 0.30516429399403233,
+      "learning_rate": 2.3708815672306324e-05,
+      "loss": 0.4118,
+      "step": 1431
+    },
+    {
+      "epoch": 1.7211538461538463,
+      "grad_norm": 0.3269344293647203,
+      "learning_rate": 2.3686553873552982e-05,
+      "loss": 0.4327,
+      "step": 1432
+    },
+    {
+      "epoch": 1.7223557692307692,
+      "grad_norm": 0.28054925693810734,
+      "learning_rate": 2.3664292074799644e-05,
+      "loss": 0.3957,
+      "step": 1433
+    },
+    {
+      "epoch": 1.7235576923076923,
+      "grad_norm": 0.25793000480599954,
+      "learning_rate": 2.3642030276046307e-05,
+      "loss": 0.4145,
+      "step": 1434
+    },
+    {
+      "epoch": 1.7247596153846154,
+      "grad_norm": 0.2840383367796164,
+      "learning_rate": 2.3619768477292965e-05,
+      "loss": 0.4364,
+      "step": 1435
+    },
+    {
+      "epoch": 1.7259615384615383,
+      "grad_norm": 0.3132671761048694,
+      "learning_rate": 2.3597506678539627e-05,
+      "loss": 0.4266,
+      "step": 1436
+    },
+    {
+      "epoch": 1.7271634615384617,
+      "grad_norm": 0.2695399758883694,
+      "learning_rate": 2.3575244879786286e-05,
+      "loss": 0.4167,
+      "step": 1437
+    },
+    {
+      "epoch": 1.7283653846153846,
+      "grad_norm": 0.27636717212908046,
+      "learning_rate": 2.3552983081032948e-05,
+      "loss": 0.4139,
+      "step": 1438
+    },
+    {
+      "epoch": 1.7295673076923077,
+      "grad_norm": 0.25383125938353646,
+      "learning_rate": 2.3530721282279607e-05,
+      "loss": 0.4342,
+      "step": 1439
+    },
+    {
+      "epoch": 1.7307692307692308,
+      "grad_norm": 0.27468335717456555,
+      "learning_rate": 2.3508459483526273e-05,
+      "loss": 0.4241,
+      "step": 1440
+    },
+    {
+      "epoch": 1.7319711538461537,
+      "grad_norm": 0.2641173945338099,
+      "learning_rate": 2.348619768477293e-05,
+      "loss": 0.4041,
+      "step": 1441
+    },
+    {
+      "epoch": 1.7331730769230769,
+      "grad_norm": 0.28391029294517034,
+      "learning_rate": 2.3463935886019593e-05,
+      "loss": 0.4363,
+      "step": 1442
+    },
+    {
+      "epoch": 1.734375,
+      "grad_norm": 0.26020380943824306,
+      "learning_rate": 2.3441674087266252e-05,
+      "loss": 0.4256,
+      "step": 1443
+    },
+    {
+      "epoch": 1.7355769230769231,
+      "grad_norm": 0.25962082207478393,
+      "learning_rate": 2.341941228851291e-05,
+      "loss": 0.4318,
+      "step": 1444
+    },
+    {
+      "epoch": 1.7367788461538463,
+      "grad_norm": 0.2871827928973887,
+      "learning_rate": 2.3397150489759573e-05,
+      "loss": 0.4071,
+      "step": 1445
+    },
+    {
+      "epoch": 1.7379807692307692,
+      "grad_norm": 0.2703979307557251,
+      "learning_rate": 2.3374888691006232e-05,
+      "loss": 0.4489,
+      "step": 1446
+    },
+    {
+      "epoch": 1.7391826923076923,
+      "grad_norm": 0.25868498475357316,
+      "learning_rate": 2.3352626892252897e-05,
+      "loss": 0.4125,
+      "step": 1447
+    },
+    {
+      "epoch": 1.7403846153846154,
+      "grad_norm": 0.254955005434906,
+      "learning_rate": 2.3330365093499556e-05,
+      "loss": 0.4103,
+      "step": 1448
+    },
+    {
+      "epoch": 1.7415865384615383,
+      "grad_norm": 0.27899046053641163,
+      "learning_rate": 2.3308103294746218e-05,
+      "loss": 0.4293,
+      "step": 1449
+    },
+    {
+      "epoch": 1.7427884615384617,
+      "grad_norm": 0.24324553901133536,
+      "learning_rate": 2.3285841495992877e-05,
+      "loss": 0.4155,
+      "step": 1450
+    },
+    {
+      "epoch": 1.7439903846153846,
+      "grad_norm": 0.3196453894278664,
+      "learning_rate": 2.326357969723954e-05,
+      "loss": 0.4228,
+      "step": 1451
+    },
+    {
+      "epoch": 1.7451923076923077,
+      "grad_norm": 0.2738058292397562,
+      "learning_rate": 2.3241317898486198e-05,
+      "loss": 0.4318,
+      "step": 1452
+    },
+    {
+      "epoch": 1.7463942307692308,
+      "grad_norm": 0.30793276199470915,
+      "learning_rate": 2.3219056099732857e-05,
+      "loss": 0.4258,
+      "step": 1453
+    },
+    {
+      "epoch": 1.7475961538461537,
+      "grad_norm": 0.29051274901633833,
+      "learning_rate": 2.3196794300979522e-05,
+      "loss": 0.4104,
+      "step": 1454
+    },
+    {
+      "epoch": 1.7487980769230769,
+      "grad_norm": 0.2662147860312565,
+      "learning_rate": 2.317453250222618e-05,
+      "loss": 0.4245,
+      "step": 1455
+    },
+    {
+      "epoch": 1.75,
+      "grad_norm": 0.3420563543344699,
+      "learning_rate": 2.3152270703472843e-05,
+      "loss": 0.4248,
+      "step": 1456
+    },
+    {
+      "epoch": 1.7512019230769231,
+      "grad_norm": 0.285626154641025,
+      "learning_rate": 2.3130008904719502e-05,
+      "loss": 0.4352,
+      "step": 1457
+    },
+    {
+      "epoch": 1.7524038461538463,
+      "grad_norm": 0.2956205419413504,
+      "learning_rate": 2.3107747105966164e-05,
+      "loss": 0.4178,
+      "step": 1458
+    },
+    {
+      "epoch": 1.7536057692307692,
+      "grad_norm": 0.3225013491097834,
+      "learning_rate": 2.3085485307212823e-05,
+      "loss": 0.4342,
+      "step": 1459
+    },
+    {
+      "epoch": 1.7548076923076923,
+      "grad_norm": 0.2647661429175042,
+      "learning_rate": 2.3063223508459485e-05,
+      "loss": 0.4475,
+      "step": 1460
+    },
+    {
+      "epoch": 1.7560096153846154,
+      "grad_norm": 0.28315449986267544,
+      "learning_rate": 2.3040961709706147e-05,
+      "loss": 0.4447,
+      "step": 1461
+    },
+    {
+      "epoch": 1.7572115384615383,
+      "grad_norm": 0.3329248285570044,
+      "learning_rate": 2.3018699910952806e-05,
+      "loss": 0.4499,
+      "step": 1462
+    },
+    {
+      "epoch": 1.7584134615384617,
+      "grad_norm": 0.270412272597286,
+      "learning_rate": 2.2996438112199468e-05,
+      "loss": 0.4338,
+      "step": 1463
+    },
+    {
+      "epoch": 1.7596153846153846,
+      "grad_norm": 0.3284321476967385,
+      "learning_rate": 2.2974176313446126e-05,
+      "loss": 0.4307,
+      "step": 1464
+    },
+    {
+      "epoch": 1.7608173076923077,
+      "grad_norm": 0.29662437121727686,
+      "learning_rate": 2.295191451469279e-05,
+      "loss": 0.4099,
+      "step": 1465
+    },
+    {
+      "epoch": 1.7620192307692308,
+      "grad_norm": 0.25044247215982773,
+      "learning_rate": 2.2929652715939447e-05,
+      "loss": 0.4033,
+      "step": 1466
+    },
+    {
+      "epoch": 1.7632211538461537,
+      "grad_norm": 0.31802259222382717,
+      "learning_rate": 2.290739091718611e-05,
+      "loss": 0.4172,
+      "step": 1467
+    },
+    {
+      "epoch": 1.7644230769230769,
+      "grad_norm": 0.27399001079507046,
+      "learning_rate": 2.288512911843277e-05,
+      "loss": 0.4235,
+      "step": 1468
+    },
+    {
+      "epoch": 1.765625,
+      "grad_norm": 0.33375871753343583,
+      "learning_rate": 2.286286731967943e-05,
+      "loss": 0.4192,
+      "step": 1469
+    },
+    {
+      "epoch": 1.7668269230769231,
+      "grad_norm": 0.32251513506482293,
+      "learning_rate": 2.2840605520926092e-05,
+      "loss": 0.4293,
+      "step": 1470
+    },
+    {
+      "epoch": 1.7680288461538463,
+      "grad_norm": 0.2909494010496952,
+      "learning_rate": 2.281834372217275e-05,
+      "loss": 0.4253,
+      "step": 1471
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.26645173469679945,
+      "learning_rate": 2.2796081923419413e-05,
+      "loss": 0.3949,
+      "step": 1472
+    },
+    {
+      "epoch": 1.7704326923076923,
+      "grad_norm": 0.31052490310858516,
+      "learning_rate": 2.2773820124666072e-05,
+      "loss": 0.4258,
+      "step": 1473
+    },
+    {
+      "epoch": 1.7716346153846154,
+      "grad_norm": 0.3425845212182336,
+      "learning_rate": 2.2751558325912734e-05,
+      "loss": 0.4236,
+      "step": 1474
+    },
+    {
+      "epoch": 1.7728365384615383,
+      "grad_norm": 0.2686652306107336,
+      "learning_rate": 2.2729296527159396e-05,
+      "loss": 0.4339,
+      "step": 1475
+    },
+    {
+      "epoch": 1.7740384615384617,
+      "grad_norm": 0.29013065438659696,
+      "learning_rate": 2.270703472840606e-05,
+      "loss": 0.3953,
+      "step": 1476
+    },
+    {
+      "epoch": 1.7752403846153846,
+      "grad_norm": 0.32465662027435765,
+      "learning_rate": 2.2684772929652717e-05,
+      "loss": 0.4211,
+      "step": 1477
+    },
+    {
+      "epoch": 1.7764423076923077,
+      "grad_norm": 0.2594679943819844,
+      "learning_rate": 2.2662511130899376e-05,
+      "loss": 0.4162,
+      "step": 1478
+    },
+    {
+      "epoch": 1.7776442307692308,
+      "grad_norm": 0.30221836508597943,
+      "learning_rate": 2.2640249332146038e-05,
+      "loss": 0.449,
+      "step": 1479
+    },
+    {
+      "epoch": 1.7788461538461537,
+      "grad_norm": 0.2980286565471681,
+      "learning_rate": 2.2617987533392697e-05,
+      "loss": 0.4314,
+      "step": 1480
+    },
+    {
+      "epoch": 1.7800480769230769,
+      "grad_norm": 0.2786250928219234,
+      "learning_rate": 2.259572573463936e-05,
+      "loss": 0.4191,
+      "step": 1481
+    },
+    {
+      "epoch": 1.78125,
+      "grad_norm": 0.2532999182734423,
+      "learning_rate": 2.257346393588602e-05,
+      "loss": 0.4083,
+      "step": 1482
+    },
+    {
+      "epoch": 1.7824519230769231,
+      "grad_norm": 0.31986816130601703,
+      "learning_rate": 2.2551202137132683e-05,
+      "loss": 0.449,
+      "step": 1483
+    },
+    {
+      "epoch": 1.7836538461538463,
+      "grad_norm": 0.2663131894097606,
+      "learning_rate": 2.2528940338379342e-05,
+      "loss": 0.4583,
+      "step": 1484
+    },
+    {
+      "epoch": 1.7848557692307692,
+      "grad_norm": 0.24241062361276103,
+      "learning_rate": 2.2506678539626004e-05,
+      "loss": 0.4148,
+      "step": 1485
+    },
+    {
+      "epoch": 1.7860576923076923,
+      "grad_norm": 0.2649731107148937,
+      "learning_rate": 2.2484416740872663e-05,
+      "loss": 0.4209,
+      "step": 1486
+    },
+    {
+      "epoch": 1.7872596153846154,
+      "grad_norm": 0.2626276232375506,
+      "learning_rate": 2.2462154942119325e-05,
+      "loss": 0.4347,
+      "step": 1487
+    },
+    {
+      "epoch": 1.7884615384615383,
+      "grad_norm": 0.2800700366961163,
+      "learning_rate": 2.2439893143365984e-05,
+      "loss": 0.4242,
+      "step": 1488
+    },
+    {
+      "epoch": 1.7896634615384617,
+      "grad_norm": 0.27270217175198164,
+      "learning_rate": 2.2417631344612646e-05,
+      "loss": 0.415,
+      "step": 1489
+    },
+    {
+      "epoch": 1.7908653846153846,
+      "grad_norm": 0.32689071128643754,
+      "learning_rate": 2.2395369545859308e-05,
+      "loss": 0.4227,
+      "step": 1490
+    },
+    {
+      "epoch": 1.7920673076923077,
+      "grad_norm": 0.2963628598013302,
+      "learning_rate": 2.2373107747105967e-05,
+      "loss": 0.4145,
+      "step": 1491
+    },
+    {
+      "epoch": 1.7932692307692308,
+      "grad_norm": 0.30047855124297496,
+      "learning_rate": 2.235084594835263e-05,
+      "loss": 0.4464,
+      "step": 1492
+    },
+    {
+      "epoch": 1.7944711538461537,
+      "grad_norm": 0.3077626928232181,
+      "learning_rate": 2.2328584149599288e-05,
+      "loss": 0.4279,
+      "step": 1493
+    },
+    {
+      "epoch": 1.7956730769230769,
+      "grad_norm": 0.3195567775175836,
+      "learning_rate": 2.230632235084595e-05,
+      "loss": 0.4388,
+      "step": 1494
+    },
+    {
+      "epoch": 1.796875,
+      "grad_norm": 0.2992262159742569,
+      "learning_rate": 2.228406055209261e-05,
+      "loss": 0.4173,
+      "step": 1495
+    },
+    {
+      "epoch": 1.7980769230769231,
+      "grad_norm": 0.32841983689597065,
+      "learning_rate": 2.226179875333927e-05,
+      "loss": 0.4144,
+      "step": 1496
+    },
+    {
+      "epoch": 1.7992788461538463,
+      "grad_norm": 0.2547146696079144,
+      "learning_rate": 2.2239536954585933e-05,
+      "loss": 0.4007,
+      "step": 1497
+    },
+    {
+      "epoch": 1.8004807692307692,
+      "grad_norm": 0.29637910499828996,
+      "learning_rate": 2.221727515583259e-05,
+      "loss": 0.4127,
+      "step": 1498
+    },
+    {
+      "epoch": 1.8016826923076923,
+      "grad_norm": 0.2855016188451794,
+      "learning_rate": 2.2195013357079254e-05,
+      "loss": 0.4052,
+      "step": 1499
+    },
+    {
+      "epoch": 1.8028846153846154,
+      "grad_norm": 0.30350256516697627,
+      "learning_rate": 2.2172751558325912e-05,
+      "loss": 0.4084,
+      "step": 1500
+    },
+    {
+      "epoch": 1.8040865384615383,
+      "grad_norm": 0.32973961164576676,
+      "learning_rate": 2.2150489759572575e-05,
+      "loss": 0.4041,
+      "step": 1501
+    },
+    {
+      "epoch": 1.8052884615384617,
+      "grad_norm": 0.29237366170194895,
+      "learning_rate": 2.2128227960819233e-05,
+      "loss": 0.427,
+      "step": 1502
+    },
+    {
+      "epoch": 1.8064903846153846,
+      "grad_norm": 0.28026383916405434,
+      "learning_rate": 2.2105966162065895e-05,
+      "loss": 0.3929,
+      "step": 1503
+    },
+    {
+      "epoch": 1.8076923076923077,
+      "grad_norm": 0.27580838197737984,
+      "learning_rate": 2.2083704363312558e-05,
+      "loss": 0.4065,
+      "step": 1504
+    },
+    {
+      "epoch": 1.8088942307692308,
+      "grad_norm": 0.31176072269131433,
+      "learning_rate": 2.2061442564559216e-05,
+      "loss": 0.4552,
+      "step": 1505
+    },
+    {
+      "epoch": 1.8100961538461537,
+      "grad_norm": 0.32876314501592363,
+      "learning_rate": 2.203918076580588e-05,
+      "loss": 0.435,
+      "step": 1506
+    },
+    {
+      "epoch": 1.8112980769230769,
+      "grad_norm": 0.27889063055415303,
+      "learning_rate": 2.2016918967052537e-05,
+      "loss": 0.4557,
+      "step": 1507
+    },
+    {
+      "epoch": 1.8125,
+      "grad_norm": 0.31780524098619867,
+      "learning_rate": 2.19946571682992e-05,
+      "loss": 0.4113,
+      "step": 1508
+    },
+    {
+      "epoch": 1.8137019230769231,
+      "grad_norm": 0.2818966219589243,
+      "learning_rate": 2.1972395369545858e-05,
+      "loss": 0.4239,
+      "step": 1509
+    },
+    {
+      "epoch": 1.8149038461538463,
+      "grad_norm": 0.37747474175989465,
+      "learning_rate": 2.1950133570792524e-05,
+      "loss": 0.412,
+      "step": 1510
+    },
+    {
+      "epoch": 1.8161057692307692,
+      "grad_norm": 0.2724616146886337,
+      "learning_rate": 2.1927871772039182e-05,
+      "loss": 0.4355,
+      "step": 1511
+    },
+    {
+      "epoch": 1.8173076923076923,
+      "grad_norm": 0.3848516202179862,
+      "learning_rate": 2.1905609973285844e-05,
+      "loss": 0.4559,
+      "step": 1512
+    },
+    {
+      "epoch": 1.8185096153846154,
+      "grad_norm": 0.3369831839532735,
+      "learning_rate": 2.1883348174532503e-05,
+      "loss": 0.4438,
+      "step": 1513
+    },
+    {
+      "epoch": 1.8197115384615383,
+      "grad_norm": 0.2817200289763952,
+      "learning_rate": 2.1861086375779162e-05,
+      "loss": 0.4121,
+      "step": 1514
+    },
+    {
+      "epoch": 1.8209134615384617,
+      "grad_norm": 0.327473874933598,
+      "learning_rate": 2.1838824577025824e-05,
+      "loss": 0.4343,
+      "step": 1515
+    },
+    {
+      "epoch": 1.8221153846153846,
+      "grad_norm": 0.3258813608137636,
+      "learning_rate": 2.1816562778272486e-05,
+      "loss": 0.4342,
+      "step": 1516
+    },
+    {
+      "epoch": 1.8233173076923077,
+      "grad_norm": 0.2956411626108861,
+      "learning_rate": 2.179430097951915e-05,
+      "loss": 0.4194,
+      "step": 1517
+    },
+    {
+      "epoch": 1.8245192307692308,
+      "grad_norm": 0.27234810299603407,
+      "learning_rate": 2.1772039180765807e-05,
+      "loss": 0.392,
+      "step": 1518
+    },
+    {
+      "epoch": 1.8257211538461537,
+      "grad_norm": 0.2486389316787076,
+      "learning_rate": 2.174977738201247e-05,
+      "loss": 0.4036,
+      "step": 1519
+    },
+    {
+      "epoch": 1.8269230769230769,
+      "grad_norm": 0.27145599711251356,
+      "learning_rate": 2.1727515583259128e-05,
+      "loss": 0.4028,
+      "step": 1520
+    },
+    {
+      "epoch": 1.828125,
+      "grad_norm": 0.27064289108894257,
+      "learning_rate": 2.170525378450579e-05,
+      "loss": 0.4378,
+      "step": 1521
+    },
+    {
+      "epoch": 1.8293269230769231,
+      "grad_norm": 0.2721680825951584,
+      "learning_rate": 2.168299198575245e-05,
+      "loss": 0.3768,
+      "step": 1522
+    },
+    {
+      "epoch": 1.8305288461538463,
+      "grad_norm": 0.2858943248957407,
+      "learning_rate": 2.166073018699911e-05,
+      "loss": 0.421,
+      "step": 1523
+    },
+    {
+      "epoch": 1.8317307692307692,
+      "grad_norm": 0.2834366032269098,
+      "learning_rate": 2.1638468388245773e-05,
+      "loss": 0.4207,
+      "step": 1524
+    },
+    {
+      "epoch": 1.8329326923076923,
+      "grad_norm": 0.28854001336822227,
+      "learning_rate": 2.1616206589492432e-05,
+      "loss": 0.415,
+      "step": 1525
+    },
+    {
+      "epoch": 1.8341346153846154,
+      "grad_norm": 0.2725705790677317,
+      "learning_rate": 2.1593944790739094e-05,
+      "loss": 0.4072,
+      "step": 1526
+    },
+    {
+      "epoch": 1.8353365384615383,
+      "grad_norm": 0.2827577057522399,
+      "learning_rate": 2.1571682991985753e-05,
+      "loss": 0.4369,
+      "step": 1527
+    },
+    {
+      "epoch": 1.8365384615384617,
+      "grad_norm": 0.2626034097376441,
+      "learning_rate": 2.1549421193232415e-05,
+      "loss": 0.3956,
+      "step": 1528
+    },
+    {
+      "epoch": 1.8377403846153846,
+      "grad_norm": 0.34943452902046407,
+      "learning_rate": 2.1527159394479074e-05,
+      "loss": 0.422,
+      "step": 1529
+    },
+    {
+      "epoch": 1.8389423076923077,
+      "grad_norm": 0.2985369198291767,
+      "learning_rate": 2.1504897595725736e-05,
+      "loss": 0.4031,
+      "step": 1530
+    },
+    {
+      "epoch": 1.8401442307692308,
+      "grad_norm": 0.28428451151777395,
+      "learning_rate": 2.1482635796972398e-05,
+      "loss": 0.4063,
+      "step": 1531
+    },
+    {
+      "epoch": 1.8413461538461537,
+      "grad_norm": 0.3249912732700625,
+      "learning_rate": 2.1460373998219057e-05,
+      "loss": 0.4041,
+      "step": 1532
+    },
+    {
+      "epoch": 1.8425480769230769,
+      "grad_norm": 0.30951868753127015,
+      "learning_rate": 2.143811219946572e-05,
+      "loss": 0.4336,
+      "step": 1533
+    },
+    {
+      "epoch": 1.84375,
+      "grad_norm": 0.2840890834099298,
+      "learning_rate": 2.1415850400712377e-05,
+      "loss": 0.4101,
+      "step": 1534
+    },
+    {
+      "epoch": 1.8449519230769231,
+      "grad_norm": 0.28944472590332654,
+      "learning_rate": 2.139358860195904e-05,
+      "loss": 0.4059,
+      "step": 1535
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.2944444966181839,
+      "learning_rate": 2.13713268032057e-05,
+      "loss": 0.421,
+      "step": 1536
+    },
+    {
+      "epoch": 1.8473557692307692,
+      "grad_norm": 0.28476057557977585,
+      "learning_rate": 2.134906500445236e-05,
+      "loss": 0.422,
+      "step": 1537
+    },
+    {
+      "epoch": 1.8485576923076923,
+      "grad_norm": 0.31367562258070186,
+      "learning_rate": 2.1326803205699023e-05,
+      "loss": 0.447,
+      "step": 1538
+    },
+    {
+      "epoch": 1.8497596153846154,
+      "grad_norm": 0.313066615737665,
+      "learning_rate": 2.130454140694568e-05,
+      "loss": 0.4246,
+      "step": 1539
+    },
+    {
+      "epoch": 1.8509615384615383,
+      "grad_norm": 0.2986958649009955,
+      "learning_rate": 2.1282279608192343e-05,
+      "loss": 0.4202,
+      "step": 1540
+    },
+    {
+      "epoch": 1.8521634615384617,
+      "grad_norm": 0.28794916435087853,
+      "learning_rate": 2.1260017809439002e-05,
+      "loss": 0.4251,
+      "step": 1541
+    },
+    {
+      "epoch": 1.8533653846153846,
+      "grad_norm": 0.3390772041859691,
+      "learning_rate": 2.1237756010685664e-05,
+      "loss": 0.4226,
+      "step": 1542
+    },
+    {
+      "epoch": 1.8545673076923077,
+      "grad_norm": 0.3202801974068068,
+      "learning_rate": 2.1215494211932323e-05,
+      "loss": 0.4416,
+      "step": 1543
+    },
+    {
+      "epoch": 1.8557692307692308,
+      "grad_norm": 0.26514461182960997,
+      "learning_rate": 2.119323241317899e-05,
+      "loss": 0.4131,
+      "step": 1544
+    },
+    {
+      "epoch": 1.8569711538461537,
+      "grad_norm": 0.30402360471323564,
+      "learning_rate": 2.1170970614425647e-05,
+      "loss": 0.4245,
+      "step": 1545
+    },
+    {
+      "epoch": 1.8581730769230769,
+      "grad_norm": 0.2840007735358458,
+      "learning_rate": 2.114870881567231e-05,
+      "loss": 0.4357,
+      "step": 1546
+    },
+    {
+      "epoch": 1.859375,
+      "grad_norm": 0.2642561678890017,
+      "learning_rate": 2.1126447016918968e-05,
+      "loss": 0.4254,
+      "step": 1547
+    },
+    {
+      "epoch": 1.8605769230769231,
+      "grad_norm": 0.2863532310816333,
+      "learning_rate": 2.1104185218165627e-05,
+      "loss": 0.4141,
+      "step": 1548
+    },
+    {
+      "epoch": 1.8617788461538463,
+      "grad_norm": 0.3449383858014571,
+      "learning_rate": 2.108192341941229e-05,
+      "loss": 0.4345,
+      "step": 1549
+    },
+    {
+      "epoch": 1.8629807692307692,
+      "grad_norm": 0.2633175397916455,
+      "learning_rate": 2.1059661620658948e-05,
+      "loss": 0.419,
+      "step": 1550
+    },
+    {
+      "epoch": 1.8641826923076923,
+      "grad_norm": 0.2877077183724589,
+      "learning_rate": 2.1037399821905613e-05,
+      "loss": 0.4232,
+      "step": 1551
+    },
+    {
+      "epoch": 1.8653846153846154,
+      "grad_norm": 0.2703628345527805,
+      "learning_rate": 2.1015138023152272e-05,
+      "loss": 0.4053,
+      "step": 1552
+    },
+    {
+      "epoch": 1.8665865384615383,
+      "grad_norm": 0.2991279582683904,
+      "learning_rate": 2.0992876224398934e-05,
+      "loss": 0.4039,
+      "step": 1553
+    },
+    {
+      "epoch": 1.8677884615384617,
+      "grad_norm": 0.27230889915681594,
+      "learning_rate": 2.0970614425645593e-05,
+      "loss": 0.4185,
+      "step": 1554
+    },
+    {
+      "epoch": 1.8689903846153846,
+      "grad_norm": 0.3747811101099768,
+      "learning_rate": 2.0948352626892255e-05,
+      "loss": 0.4331,
+      "step": 1555
+    },
+    {
+      "epoch": 1.8701923076923077,
+      "grad_norm": 0.2920384181991187,
+      "learning_rate": 2.0926090828138914e-05,
+      "loss": 0.4591,
+      "step": 1556
+    },
+    {
+      "epoch": 1.8713942307692308,
+      "grad_norm": 0.29777816399686646,
+      "learning_rate": 2.0903829029385576e-05,
+      "loss": 0.4481,
+      "step": 1557
+    },
+    {
+      "epoch": 1.8725961538461537,
+      "grad_norm": 0.3557909440302136,
+      "learning_rate": 2.0881567230632238e-05,
+      "loss": 0.4381,
+      "step": 1558
+    },
+    {
+      "epoch": 1.8737980769230769,
+      "grad_norm": 0.5221836105227674,
+      "learning_rate": 2.0859305431878897e-05,
+      "loss": 0.4165,
+      "step": 1559
+    },
+    {
+      "epoch": 1.875,
+      "grad_norm": 0.2867218659121369,
+      "learning_rate": 2.083704363312556e-05,
+      "loss": 0.4117,
+      "step": 1560
+    },
+    {
+      "epoch": 1.8762019230769231,
+      "grad_norm": 0.3593443899034351,
+      "learning_rate": 2.0814781834372218e-05,
+      "loss": 0.3883,
+      "step": 1561
+    },
+    {
+      "epoch": 1.8774038461538463,
+      "grad_norm": 0.3266848778370129,
+      "learning_rate": 2.079252003561888e-05,
+      "loss": 0.4099,
+      "step": 1562
+    },
+    {
+      "epoch": 1.8786057692307692,
+      "grad_norm": 0.26709838864644486,
+      "learning_rate": 2.077025823686554e-05,
+      "loss": 0.4295,
+      "step": 1563
+    },
+    {
+      "epoch": 1.8798076923076923,
+      "grad_norm": 0.3457854337453769,
+      "learning_rate": 2.07479964381122e-05,
+      "loss": 0.426,
+      "step": 1564
+    },
+    {
+      "epoch": 1.8810096153846154,
+      "grad_norm": 0.3451359018738587,
+      "learning_rate": 2.0725734639358863e-05,
+      "loss": 0.41,
+      "step": 1565
+    },
+    {
+      "epoch": 1.8822115384615383,
+      "grad_norm": 0.2686376145591865,
+      "learning_rate": 2.070347284060552e-05,
+      "loss": 0.42,
+      "step": 1566
+    },
+    {
+      "epoch": 1.8834134615384617,
+      "grad_norm": 0.6489571102701015,
+      "learning_rate": 2.0681211041852184e-05,
+      "loss": 0.4362,
+      "step": 1567
+    },
+    {
+      "epoch": 1.8846153846153846,
+      "grad_norm": 0.3136423597940074,
+      "learning_rate": 2.0658949243098843e-05,
+      "loss": 0.4344,
+      "step": 1568
+    },
+    {
+      "epoch": 1.8858173076923077,
+      "grad_norm": 0.33070611588093346,
+      "learning_rate": 2.0636687444345505e-05,
+      "loss": 0.4643,
+      "step": 1569
+    },
+    {
+      "epoch": 1.8870192307692308,
+      "grad_norm": 0.34738348622465154,
+      "learning_rate": 2.0614425645592163e-05,
+      "loss": 0.4523,
+      "step": 1570
+    },
+    {
+      "epoch": 1.8882211538461537,
+      "grad_norm": 0.3301202086188294,
+      "learning_rate": 2.0592163846838826e-05,
+      "loss": 0.4155,
+      "step": 1571
+    },
+    {
+      "epoch": 1.8894230769230769,
+      "grad_norm": 0.3175150565787592,
+      "learning_rate": 2.0569902048085488e-05,
+      "loss": 0.4168,
+      "step": 1572
+    },
+    {
+      "epoch": 1.890625,
+      "grad_norm": 0.33805890200792,
+      "learning_rate": 2.0547640249332146e-05,
+      "loss": 0.4287,
+      "step": 1573
+    },
+    {
+      "epoch": 1.8918269230769231,
+      "grad_norm": 0.3384043693853342,
+      "learning_rate": 2.052537845057881e-05,
+      "loss": 0.4191,
+      "step": 1574
+    },
+    {
+      "epoch": 1.8930288461538463,
+      "grad_norm": 0.2916925219716751,
+      "learning_rate": 2.0503116651825467e-05,
+      "loss": 0.4279,
+      "step": 1575
+    },
+    {
+      "epoch": 1.8942307692307692,
+      "grad_norm": 0.29719424832400654,
+      "learning_rate": 2.048085485307213e-05,
+      "loss": 0.4234,
+      "step": 1576
+    },
+    {
+      "epoch": 1.8954326923076923,
+      "grad_norm": 0.2904483756737338,
+      "learning_rate": 2.0458593054318788e-05,
+      "loss": 0.4335,
+      "step": 1577
+    },
+    {
+      "epoch": 1.8966346153846154,
+      "grad_norm": 0.3536249822828667,
+      "learning_rate": 2.043633125556545e-05,
+      "loss": 0.4126,
+      "step": 1578
+    },
+    {
+      "epoch": 1.8978365384615383,
+      "grad_norm": 0.3506374729321357,
+      "learning_rate": 2.0414069456812112e-05,
+      "loss": 0.406,
+      "step": 1579
+    },
+    {
+      "epoch": 1.8990384615384617,
+      "grad_norm": 0.32364669407804064,
+      "learning_rate": 2.0391807658058775e-05,
+      "loss": 0.4214,
+      "step": 1580
+    },
+    {
+      "epoch": 1.9002403846153846,
+      "grad_norm": 0.3234726420174138,
+      "learning_rate": 2.0369545859305433e-05,
+      "loss": 0.4143,
+      "step": 1581
+    },
+    {
+      "epoch": 1.9014423076923077,
+      "grad_norm": 0.3513460893932063,
+      "learning_rate": 2.0347284060552092e-05,
+      "loss": 0.4577,
+      "step": 1582
+    },
+    {
+      "epoch": 1.9026442307692308,
+      "grad_norm": 0.30830171115640065,
+      "learning_rate": 2.0325022261798754e-05,
+      "loss": 0.412,
+      "step": 1583
+    },
+    {
+      "epoch": 1.9038461538461537,
+      "grad_norm": 0.3358467023703407,
+      "learning_rate": 2.0302760463045413e-05,
+      "loss": 0.433,
+      "step": 1584
+    },
+    {
+      "epoch": 1.9050480769230769,
+      "grad_norm": 0.3193127266574012,
+      "learning_rate": 2.0280498664292075e-05,
+      "loss": 0.4523,
+      "step": 1585
+    },
+    {
+      "epoch": 1.90625,
+      "grad_norm": 0.7701635663443241,
+      "learning_rate": 2.0258236865538737e-05,
+      "loss": 0.4064,
+      "step": 1586
+    },
+    {
+      "epoch": 1.9074519230769231,
+      "grad_norm": 0.293637404192128,
+      "learning_rate": 2.02359750667854e-05,
+      "loss": 0.4119,
+      "step": 1587
+    },
+    {
+      "epoch": 1.9086538461538463,
+      "grad_norm": 0.27296939080722027,
+      "learning_rate": 2.0213713268032058e-05,
+      "loss": 0.4201,
+      "step": 1588
+    },
+    {
+      "epoch": 1.9098557692307692,
+      "grad_norm": 0.2749525057215862,
+      "learning_rate": 2.019145146927872e-05,
+      "loss": 0.4215,
+      "step": 1589
+    },
+    {
+      "epoch": 1.9110576923076923,
+      "grad_norm": 0.2863151186126506,
+      "learning_rate": 2.016918967052538e-05,
+      "loss": 0.4096,
+      "step": 1590
+    },
+    {
+      "epoch": 1.9122596153846154,
+      "grad_norm": 0.2632925122244369,
+      "learning_rate": 2.014692787177204e-05,
+      "loss": 0.4214,
+      "step": 1591
+    },
+    {
+      "epoch": 1.9134615384615383,
+      "grad_norm": 0.2913311978988049,
+      "learning_rate": 2.01246660730187e-05,
+      "loss": 0.4276,
+      "step": 1592
+    },
+    {
+      "epoch": 1.9146634615384617,
+      "grad_norm": 0.2838115543965109,
+      "learning_rate": 2.0102404274265362e-05,
+      "loss": 0.4296,
+      "step": 1593
+    },
+    {
+      "epoch": 1.9158653846153846,
+      "grad_norm": 0.28504890735773414,
+      "learning_rate": 2.0080142475512024e-05,
+      "loss": 0.4213,
+      "step": 1594
+    },
+    {
+      "epoch": 1.9170673076923077,
+      "grad_norm": 0.31177532378880246,
+      "learning_rate": 2.0057880676758683e-05,
+      "loss": 0.4144,
+      "step": 1595
+    },
+    {
+      "epoch": 1.9182692307692308,
+      "grad_norm": 0.24639150091001558,
+      "learning_rate": 2.0035618878005345e-05,
+      "loss": 0.4271,
+      "step": 1596
+    },
+    {
+      "epoch": 1.9194711538461537,
+      "grad_norm": 0.2704371212194619,
+      "learning_rate": 2.0013357079252004e-05,
+      "loss": 0.4252,
+      "step": 1597
+    },
+    {
+      "epoch": 1.9206730769230769,
+      "grad_norm": 0.2760927593114616,
+      "learning_rate": 1.9991095280498666e-05,
+      "loss": 0.4363,
+      "step": 1598
+    },
+    {
+      "epoch": 1.921875,
+      "grad_norm": 0.2748495651272052,
+      "learning_rate": 1.9968833481745325e-05,
+      "loss": 0.4075,
+      "step": 1599
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.27590993374096223,
+      "learning_rate": 1.9946571682991987e-05,
+      "loss": 0.3983,
+      "step": 1600
+    },
+    {
+      "epoch": 1.9242788461538463,
+      "grad_norm": 0.2559926388667504,
+      "learning_rate": 1.992430988423865e-05,
+      "loss": 0.4152,
+      "step": 1601
+    },
+    {
+      "epoch": 1.9254807692307692,
+      "grad_norm": 0.2809067784842958,
+      "learning_rate": 1.9902048085485308e-05,
+      "loss": 0.423,
+      "step": 1602
+    },
+    {
+      "epoch": 1.9266826923076923,
+      "grad_norm": 0.2541451393416145,
+      "learning_rate": 1.987978628673197e-05,
+      "loss": 0.4106,
+      "step": 1603
+    },
+    {
+      "epoch": 1.9278846153846154,
+      "grad_norm": 0.2820436751238914,
+      "learning_rate": 1.985752448797863e-05,
+      "loss": 0.4293,
+      "step": 1604
+    },
+    {
+      "epoch": 1.9290865384615383,
+      "grad_norm": 0.2781174297725503,
+      "learning_rate": 1.983526268922529e-05,
+      "loss": 0.4354,
+      "step": 1605
+    },
+    {
+      "epoch": 1.9302884615384617,
+      "grad_norm": 0.2774137288393877,
+      "learning_rate": 1.981300089047195e-05,
+      "loss": 0.4001,
+      "step": 1606
+    },
+    {
+      "epoch": 1.9314903846153846,
+      "grad_norm": 0.2496934369213698,
+      "learning_rate": 1.979073909171861e-05,
+      "loss": 0.4151,
+      "step": 1607
+    },
+    {
+      "epoch": 1.9326923076923077,
+      "grad_norm": 0.2621557156152525,
+      "learning_rate": 1.9768477292965274e-05,
+      "loss": 0.3988,
+      "step": 1608
+    },
+    {
+      "epoch": 1.9338942307692308,
+      "grad_norm": 0.2783767228035584,
+      "learning_rate": 1.9746215494211932e-05,
+      "loss": 0.436,
+      "step": 1609
+    },
+    {
+      "epoch": 1.9350961538461537,
+      "grad_norm": 0.2902766393952309,
+      "learning_rate": 1.9723953695458594e-05,
+      "loss": 0.4198,
+      "step": 1610
+    },
+    {
+      "epoch": 1.9362980769230769,
+      "grad_norm": 0.2732020302643002,
+      "learning_rate": 1.9701691896705253e-05,
+      "loss": 0.4078,
+      "step": 1611
+    },
+    {
+      "epoch": 1.9375,
+      "grad_norm": 0.2680254122995616,
+      "learning_rate": 1.9679430097951915e-05,
+      "loss": 0.4279,
+      "step": 1612
+    },
+    {
+      "epoch": 1.9387019230769231,
+      "grad_norm": 0.256815622503272,
+      "learning_rate": 1.9657168299198574e-05,
+      "loss": 0.4241,
+      "step": 1613
+    },
+    {
+      "epoch": 1.9399038461538463,
+      "grad_norm": 0.26310974311041846,
+      "learning_rate": 1.963490650044524e-05,
+      "loss": 0.4228,
+      "step": 1614
+    },
+    {
+      "epoch": 1.9411057692307692,
+      "grad_norm": 0.2609645285330759,
+      "learning_rate": 1.96126447016919e-05,
+      "loss": 0.392,
+      "step": 1615
+    },
+    {
+      "epoch": 1.9423076923076923,
+      "grad_norm": 0.28656332769456183,
+      "learning_rate": 1.959038290293856e-05,
+      "loss": 0.4169,
+      "step": 1616
+    },
+    {
+      "epoch": 1.9435096153846154,
+      "grad_norm": 0.24905862527176545,
+      "learning_rate": 1.956812110418522e-05,
+      "loss": 0.3951,
+      "step": 1617
+    },
+    {
+      "epoch": 1.9447115384615383,
+      "grad_norm": 0.25782836196329967,
+      "learning_rate": 1.9545859305431878e-05,
+      "loss": 0.4326,
+      "step": 1618
+    },
+    {
+      "epoch": 1.9459134615384617,
+      "grad_norm": 0.32872289216425443,
+      "learning_rate": 1.952359750667854e-05,
+      "loss": 0.4206,
+      "step": 1619
+    },
+    {
+      "epoch": 1.9471153846153846,
+      "grad_norm": 1.7007426367181955,
+      "learning_rate": 1.95013357079252e-05,
+      "loss": 0.4538,
+      "step": 1620
+    },
+    {
+      "epoch": 1.9483173076923077,
+      "grad_norm": 0.3168531687433226,
+      "learning_rate": 1.9479073909171864e-05,
+      "loss": 0.4159,
+      "step": 1621
+    },
+    {
+      "epoch": 1.9495192307692308,
+      "grad_norm": 0.37542783758380727,
+      "learning_rate": 1.9456812110418523e-05,
+      "loss": 0.4126,
+      "step": 1622
+    },
+    {
+      "epoch": 1.9507211538461537,
+      "grad_norm": 0.26581243758051976,
+      "learning_rate": 1.9434550311665185e-05,
+      "loss": 0.4051,
+      "step": 1623
+    },
+    {
+      "epoch": 1.9519230769230769,
+      "grad_norm": 0.3080916131531404,
+      "learning_rate": 1.9412288512911844e-05,
+      "loss": 0.4185,
+      "step": 1624
+    },
+    {
+      "epoch": 1.953125,
+      "grad_norm": 0.3137253777288739,
+      "learning_rate": 1.9390026714158506e-05,
+      "loss": 0.4264,
+      "step": 1625
+    },
+    {
+      "epoch": 1.9543269230769231,
+      "grad_norm": 0.32235976918950004,
+      "learning_rate": 1.9367764915405165e-05,
+      "loss": 0.4202,
+      "step": 1626
+    },
+    {
+      "epoch": 1.9555288461538463,
+      "grad_norm": 0.2809251357997884,
+      "learning_rate": 1.9345503116651824e-05,
+      "loss": 0.4151,
+      "step": 1627
+    },
+    {
+      "epoch": 1.9567307692307692,
+      "grad_norm": 0.27308589003722145,
+      "learning_rate": 1.932324131789849e-05,
+      "loss": 0.3941,
+      "step": 1628
+    },
+    {
+      "epoch": 1.9579326923076923,
+      "grad_norm": 0.3453982086346442,
+      "learning_rate": 1.9300979519145148e-05,
+      "loss": 0.44,
+      "step": 1629
+    },
+    {
+      "epoch": 1.9591346153846154,
+      "grad_norm": 0.25600053280502555,
+      "learning_rate": 1.927871772039181e-05,
+      "loss": 0.4112,
+      "step": 1630
+    },
+    {
+      "epoch": 1.9603365384615383,
+      "grad_norm": 0.29351571329237724,
+      "learning_rate": 1.925645592163847e-05,
+      "loss": 0.3925,
+      "step": 1631
+    },
+    {
+      "epoch": 1.9615384615384617,
+      "grad_norm": 0.27212170877982894,
+      "learning_rate": 1.923419412288513e-05,
+      "loss": 0.3948,
+      "step": 1632
+    },
+    {
+      "epoch": 1.9627403846153846,
+      "grad_norm": 0.25945289061279164,
+      "learning_rate": 1.921193232413179e-05,
+      "loss": 0.4075,
+      "step": 1633
+    },
+    {
+      "epoch": 1.9639423076923077,
+      "grad_norm": 0.3093465235964263,
+      "learning_rate": 1.9189670525378452e-05,
+      "loss": 0.4404,
+      "step": 1634
+    },
+    {
+      "epoch": 1.9651442307692308,
+      "grad_norm": 0.340511588849664,
+      "learning_rate": 1.9167408726625114e-05,
+      "loss": 0.4174,
+      "step": 1635
+    },
+    {
+      "epoch": 1.9663461538461537,
+      "grad_norm": 0.27444427863782345,
+      "learning_rate": 1.9145146927871773e-05,
+      "loss": 0.4263,
+      "step": 1636
+    },
+    {
+      "epoch": 1.9675480769230769,
+      "grad_norm": 0.3343801226582131,
+      "learning_rate": 1.9122885129118435e-05,
+      "loss": 0.4162,
+      "step": 1637
+    },
+    {
+      "epoch": 1.96875,
+      "grad_norm": 0.24258048411200678,
+      "learning_rate": 1.9100623330365093e-05,
+      "loss": 0.3814,
+      "step": 1638
+    },
+    {
+      "epoch": 1.9699519230769231,
+      "grad_norm": 0.2297529188008156,
+      "learning_rate": 1.9078361531611756e-05,
+      "loss": 0.3818,
+      "step": 1639
+    },
+    {
+      "epoch": 1.9711538461538463,
+      "grad_norm": 0.2737148320423315,
+      "learning_rate": 1.9056099732858414e-05,
+      "loss": 0.4192,
+      "step": 1640
+    },
+    {
+      "epoch": 1.9723557692307692,
+      "grad_norm": 0.32148865949906696,
+      "learning_rate": 1.9033837934105077e-05,
+      "loss": 0.4417,
+      "step": 1641
+    },
+    {
+      "epoch": 1.9735576923076923,
+      "grad_norm": 0.2505501301503861,
+      "learning_rate": 1.901157613535174e-05,
+      "loss": 0.4147,
+      "step": 1642
+    },
+    {
+      "epoch": 1.9747596153846154,
+      "grad_norm": 0.3109874339186675,
+      "learning_rate": 1.8989314336598397e-05,
+      "loss": 0.4159,
+      "step": 1643
+    },
+    {
+      "epoch": 1.9759615384615383,
+      "grad_norm": 0.3354028628631632,
+      "learning_rate": 1.896705253784506e-05,
+      "loss": 0.4136,
+      "step": 1644
+    },
+    {
+      "epoch": 1.9771634615384617,
+      "grad_norm": 0.27000099001001243,
+      "learning_rate": 1.8944790739091718e-05,
+      "loss": 0.4288,
+      "step": 1645
+    },
+    {
+      "epoch": 1.9783653846153846,
+      "grad_norm": 0.28516189897770255,
+      "learning_rate": 1.892252894033838e-05,
+      "loss": 0.3974,
+      "step": 1646
+    },
+    {
+      "epoch": 1.9795673076923077,
+      "grad_norm": 0.28919961695751095,
+      "learning_rate": 1.890026714158504e-05,
+      "loss": 0.409,
+      "step": 1647
+    },
+    {
+      "epoch": 1.9807692307692308,
+      "grad_norm": 0.29051687672237797,
+      "learning_rate": 1.88780053428317e-05,
+      "loss": 0.4194,
+      "step": 1648
+    },
+    {
+      "epoch": 1.9819711538461537,
+      "grad_norm": 0.2699828612950714,
+      "learning_rate": 1.8855743544078363e-05,
+      "loss": 0.4049,
+      "step": 1649
+    },
+    {
+      "epoch": 1.9831730769230769,
+      "grad_norm": 0.3207537195662113,
+      "learning_rate": 1.8833481745325026e-05,
+      "loss": 0.4152,
+      "step": 1650
+    },
+    {
+      "epoch": 1.984375,
+      "grad_norm": 0.30955019147611146,
+      "learning_rate": 1.8811219946571684e-05,
+      "loss": 0.4437,
+      "step": 1651
+    },
+    {
+      "epoch": 1.9855769230769231,
+      "grad_norm": 0.2807319449902102,
+      "learning_rate": 1.8788958147818343e-05,
+      "loss": 0.4082,
+      "step": 1652
+    },
+    {
+      "epoch": 1.9867788461538463,
+      "grad_norm": 0.39266417208279564,
+      "learning_rate": 1.8766696349065005e-05,
+      "loss": 0.4248,
+      "step": 1653
+    },
+    {
+      "epoch": 1.9879807692307692,
+      "grad_norm": 0.4070673519230354,
+      "learning_rate": 1.8744434550311664e-05,
+      "loss": 0.4325,
+      "step": 1654
+    },
+    {
+      "epoch": 1.9891826923076923,
+      "grad_norm": 0.30563351069281414,
+      "learning_rate": 1.8722172751558326e-05,
+      "loss": 0.3939,
+      "step": 1655
+    },
+    {
+      "epoch": 1.9903846153846154,
+      "grad_norm": 0.27019515393926286,
+      "learning_rate": 1.8699910952804988e-05,
+      "loss": 0.3997,
+      "step": 1656
+    },
+    {
+      "epoch": 1.9915865384615383,
+      "grad_norm": 0.3045939194173128,
+      "learning_rate": 1.867764915405165e-05,
+      "loss": 0.4393,
+      "step": 1657
+    },
+    {
+      "epoch": 1.9927884615384617,
+      "grad_norm": 0.27056621338479614,
+      "learning_rate": 1.865538735529831e-05,
+      "loss": 0.4171,
+      "step": 1658
+    },
+    {
+      "epoch": 1.9939903846153846,
+      "grad_norm": 0.24760662669640907,
+      "learning_rate": 1.863312555654497e-05,
+      "loss": 0.4295,
+      "step": 1659
+    },
+    {
+      "epoch": 1.9951923076923077,
+      "grad_norm": 0.26265937681556956,
+      "learning_rate": 1.861086375779163e-05,
+      "loss": 0.4181,
+      "step": 1660
+    },
+    {
+      "epoch": 1.9963942307692308,
+      "grad_norm": 0.2883165542837707,
+      "learning_rate": 1.8588601959038292e-05,
+      "loss": 0.434,
+      "step": 1661
+    },
+    {
+      "epoch": 1.9975961538461537,
+      "grad_norm": 0.2640441724185461,
+      "learning_rate": 1.856634016028495e-05,
+      "loss": 0.3965,
+      "step": 1662
+    },
+    {
+      "epoch": 1.9987980769230769,
+      "grad_norm": 0.28600014460004214,
+      "learning_rate": 1.8544078361531613e-05,
+      "loss": 0.4111,
+      "step": 1663
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.2825488214675188,
+      "learning_rate": 1.8521816562778275e-05,
+      "loss": 0.4189,
+      "step": 1664
+    },
+    {
+      "epoch": 2.001201923076923,
+      "grad_norm": 0.3209665674765138,
+      "learning_rate": 1.8499554764024934e-05,
+      "loss": 0.3205,
+      "step": 1665
+    },
+    {
+      "epoch": 2.0024038461538463,
+      "grad_norm": 0.2950228504121999,
+      "learning_rate": 1.8477292965271596e-05,
+      "loss": 0.3634,
+      "step": 1666
+    },
+    {
+      "epoch": 2.003605769230769,
+      "grad_norm": 0.26873455175452565,
+      "learning_rate": 1.8455031166518255e-05,
+      "loss": 0.3223,
+      "step": 1667
+    },
+    {
+      "epoch": 2.0048076923076925,
+      "grad_norm": 0.2938956913595507,
+      "learning_rate": 1.8432769367764917e-05,
+      "loss": 0.3214,
+      "step": 1668
+    },
+    {
+      "epoch": 2.0060096153846154,
+      "grad_norm": 0.2747582209648464,
+      "learning_rate": 1.8410507569011576e-05,
+      "loss": 0.3388,
+      "step": 1669
+    },
+    {
+      "epoch": 2.0072115384615383,
+      "grad_norm": 0.2908616374773169,
+      "learning_rate": 1.8388245770258238e-05,
+      "loss": 0.3284,
+      "step": 1670
+    },
+    {
+      "epoch": 2.0084134615384617,
+      "grad_norm": 0.2913687477723326,
+      "learning_rate": 1.83659839715049e-05,
+      "loss": 0.3075,
+      "step": 1671
+    },
+    {
+      "epoch": 2.0096153846153846,
+      "grad_norm": 0.3015921229712734,
+      "learning_rate": 1.834372217275156e-05,
+      "loss": 0.3287,
+      "step": 1672
+    },
+    {
+      "epoch": 2.0108173076923075,
+      "grad_norm": 0.2588011753352397,
+      "learning_rate": 1.832146037399822e-05,
+      "loss": 0.3154,
+      "step": 1673
+    },
+    {
+      "epoch": 2.012019230769231,
+      "grad_norm": 0.2903227714545422,
+      "learning_rate": 1.829919857524488e-05,
+      "loss": 0.3156,
+      "step": 1674
+    },
+    {
+      "epoch": 2.0132211538461537,
+      "grad_norm": 0.29421277586463573,
+      "learning_rate": 1.827693677649154e-05,
+      "loss": 0.3229,
+      "step": 1675
+    },
+    {
+      "epoch": 2.014423076923077,
+      "grad_norm": 0.27034693529663,
+      "learning_rate": 1.82546749777382e-05,
+      "loss": 0.319,
+      "step": 1676
+    },
+    {
+      "epoch": 2.015625,
+      "grad_norm": 0.2827638960781193,
+      "learning_rate": 1.8232413178984862e-05,
+      "loss": 0.3358,
+      "step": 1677
+    },
+    {
+      "epoch": 2.016826923076923,
+      "grad_norm": 0.258248965974793,
+      "learning_rate": 1.8210151380231525e-05,
+      "loss": 0.323,
+      "step": 1678
+    },
+    {
+      "epoch": 2.0180288461538463,
+      "grad_norm": 0.26667748572070665,
+      "learning_rate": 1.8187889581478183e-05,
+      "loss": 0.3187,
+      "step": 1679
+    },
+    {
+      "epoch": 2.019230769230769,
+      "grad_norm": 0.267146676431635,
+      "learning_rate": 1.8165627782724845e-05,
+      "loss": 0.3239,
+      "step": 1680
+    },
+    {
+      "epoch": 2.0204326923076925,
+      "grad_norm": 0.26552849248911864,
+      "learning_rate": 1.8143365983971504e-05,
+      "loss": 0.3345,
+      "step": 1681
+    },
+    {
+      "epoch": 2.0216346153846154,
+      "grad_norm": 0.2371208928130976,
+      "learning_rate": 1.8121104185218166e-05,
+      "loss": 0.3192,
+      "step": 1682
+    },
+    {
+      "epoch": 2.0228365384615383,
+      "grad_norm": 0.28166147823854276,
+      "learning_rate": 1.8098842386464825e-05,
+      "loss": 0.3208,
+      "step": 1683
+    },
+    {
+      "epoch": 2.0240384615384617,
+      "grad_norm": 0.2734203496856039,
+      "learning_rate": 1.807658058771149e-05,
+      "loss": 0.3414,
+      "step": 1684
+    },
+    {
+      "epoch": 2.0252403846153846,
+      "grad_norm": 0.28176328624947977,
+      "learning_rate": 1.805431878895815e-05,
+      "loss": 0.3467,
+      "step": 1685
+    },
+    {
+      "epoch": 2.0264423076923075,
+      "grad_norm": 0.2968559942130102,
+      "learning_rate": 1.8032056990204808e-05,
+      "loss": 0.3394,
+      "step": 1686
+    },
+    {
+      "epoch": 2.027644230769231,
+      "grad_norm": 0.22705775834052064,
+      "learning_rate": 1.800979519145147e-05,
+      "loss": 0.3276,
+      "step": 1687
+    },
+    {
+      "epoch": 2.0288461538461537,
+      "grad_norm": 0.30081158350718873,
+      "learning_rate": 1.798753339269813e-05,
+      "loss": 0.3035,
+      "step": 1688
+    },
+    {
+      "epoch": 2.030048076923077,
+      "grad_norm": 0.26551117669250174,
+      "learning_rate": 1.796527159394479e-05,
+      "loss": 0.3335,
+      "step": 1689
+    },
+    {
+      "epoch": 2.03125,
+      "grad_norm": 0.25167860937454856,
+      "learning_rate": 1.794300979519145e-05,
+      "loss": 0.3163,
+      "step": 1690
+    },
+    {
+      "epoch": 2.032451923076923,
+      "grad_norm": 0.2727082265417413,
+      "learning_rate": 1.7920747996438115e-05,
+      "loss": 0.3411,
+      "step": 1691
+    },
+    {
+      "epoch": 2.0336538461538463,
+      "grad_norm": 0.2383009083826183,
+      "learning_rate": 1.7898486197684774e-05,
+      "loss": 0.311,
+      "step": 1692
+    },
+    {
+      "epoch": 2.034855769230769,
+      "grad_norm": 0.2435415181918318,
+      "learning_rate": 1.7876224398931436e-05,
+      "loss": 0.3199,
+      "step": 1693
+    },
+    {
+      "epoch": 2.0360576923076925,
+      "grad_norm": 0.23373037011973744,
+      "learning_rate": 1.7853962600178095e-05,
+      "loss": 0.3222,
+      "step": 1694
+    },
+    {
+      "epoch": 2.0372596153846154,
+      "grad_norm": 0.24913560019314743,
+      "learning_rate": 1.7831700801424757e-05,
+      "loss": 0.3363,
+      "step": 1695
+    },
+    {
+      "epoch": 2.0384615384615383,
+      "grad_norm": 0.261397076033721,
+      "learning_rate": 1.7809439002671416e-05,
+      "loss": 0.3465,
+      "step": 1696
+    },
+    {
+      "epoch": 2.0396634615384617,
+      "grad_norm": 0.2525057584377757,
+      "learning_rate": 1.7787177203918075e-05,
+      "loss": 0.3309,
+      "step": 1697
+    },
+    {
+      "epoch": 2.0408653846153846,
+      "grad_norm": 0.24097375710408384,
+      "learning_rate": 1.776491540516474e-05,
+      "loss": 0.3394,
+      "step": 1698
+    },
+    {
+      "epoch": 2.0420673076923075,
+      "grad_norm": 0.22658863516447383,
+      "learning_rate": 1.77426536064114e-05,
+      "loss": 0.3126,
+      "step": 1699
+    },
+    {
+      "epoch": 2.043269230769231,
+      "grad_norm": 0.2799731633577757,
+      "learning_rate": 1.772039180765806e-05,
+      "loss": 0.3271,
+      "step": 1700
+    },
+    {
+      "epoch": 2.0444711538461537,
+      "grad_norm": 0.2980640463622482,
+      "learning_rate": 1.769813000890472e-05,
+      "loss": 0.3337,
+      "step": 1701
+    },
+    {
+      "epoch": 2.045673076923077,
+      "grad_norm": 0.2370911837163832,
+      "learning_rate": 1.7675868210151382e-05,
+      "loss": 0.3148,
+      "step": 1702
+    },
+    {
+      "epoch": 2.046875,
+      "grad_norm": 0.23863627612777982,
+      "learning_rate": 1.765360641139804e-05,
+      "loss": 0.3115,
+      "step": 1703
+    },
+    {
+      "epoch": 2.048076923076923,
+      "grad_norm": 0.2512073627344605,
+      "learning_rate": 1.7631344612644703e-05,
+      "loss": 0.3287,
+      "step": 1704
+    },
+    {
+      "epoch": 2.0492788461538463,
+      "grad_norm": 0.26105966107727563,
+      "learning_rate": 1.7609082813891365e-05,
+      "loss": 0.3102,
+      "step": 1705
+    },
+    {
+      "epoch": 2.050480769230769,
+      "grad_norm": 0.24926806358302359,
+      "learning_rate": 1.7586821015138024e-05,
+      "loss": 0.3201,
+      "step": 1706
+    },
+    {
+      "epoch": 2.0516826923076925,
+      "grad_norm": 0.24478379928435495,
+      "learning_rate": 1.7564559216384686e-05,
+      "loss": 0.3404,
+      "step": 1707
+    },
+    {
+      "epoch": 2.0528846153846154,
+      "grad_norm": 0.21821189483055253,
+      "learning_rate": 1.7542297417631344e-05,
+      "loss": 0.3282,
+      "step": 1708
+    },
+    {
+      "epoch": 2.0540865384615383,
+      "grad_norm": 0.2656904200321035,
+      "learning_rate": 1.7520035618878007e-05,
+      "loss": 0.3391,
+      "step": 1709
+    },
+    {
+      "epoch": 2.0552884615384617,
+      "grad_norm": 0.23077111059278663,
+      "learning_rate": 1.7497773820124665e-05,
+      "loss": 0.3079,
+      "step": 1710
+    },
+    {
+      "epoch": 2.0564903846153846,
+      "grad_norm": 0.2523093399747255,
+      "learning_rate": 1.7475512021371327e-05,
+      "loss": 0.3311,
+      "step": 1711
+    },
+    {
+      "epoch": 2.0576923076923075,
+      "grad_norm": 0.2487255592144454,
+      "learning_rate": 1.745325022261799e-05,
+      "loss": 0.2956,
+      "step": 1712
+    },
+    {
+      "epoch": 2.058894230769231,
+      "grad_norm": 0.23297901781125888,
+      "learning_rate": 1.743098842386465e-05,
+      "loss": 0.3206,
+      "step": 1713
+    },
+    {
+      "epoch": 2.0600961538461537,
+      "grad_norm": 0.25361831626948106,
+      "learning_rate": 1.740872662511131e-05,
+      "loss": 0.3274,
+      "step": 1714
+    },
+    {
+      "epoch": 2.061298076923077,
+      "grad_norm": 0.2881122041298559,
+      "learning_rate": 1.738646482635797e-05,
+      "loss": 0.3178,
+      "step": 1715
+    },
+    {
+      "epoch": 2.0625,
+      "grad_norm": 0.2573697253124712,
+      "learning_rate": 1.736420302760463e-05,
+      "loss": 0.3123,
+      "step": 1716
+    },
+    {
+      "epoch": 2.063701923076923,
+      "grad_norm": 0.241335622758871,
+      "learning_rate": 1.734194122885129e-05,
+      "loss": 0.3148,
+      "step": 1717
+    },
+    {
+      "epoch": 2.0649038461538463,
+      "grad_norm": 0.2386916403630111,
+      "learning_rate": 1.7319679430097952e-05,
+      "loss": 0.3,
+      "step": 1718
+    },
+    {
+      "epoch": 2.066105769230769,
+      "grad_norm": 0.24330664604300378,
+      "learning_rate": 1.7297417631344614e-05,
+      "loss": 0.32,
+      "step": 1719
+    },
+    {
+      "epoch": 2.0673076923076925,
+      "grad_norm": 0.284014574458192,
+      "learning_rate": 1.7275155832591277e-05,
+      "loss": 0.3403,
+      "step": 1720
+    },
+    {
+      "epoch": 2.0685096153846154,
+      "grad_norm": 0.2531338313556435,
+      "learning_rate": 1.7252894033837935e-05,
+      "loss": 0.3168,
+      "step": 1721
+    },
+    {
+      "epoch": 2.0697115384615383,
+      "grad_norm": 0.23772831859440244,
+      "learning_rate": 1.7230632235084594e-05,
+      "loss": 0.3087,
+      "step": 1722
+    },
+    {
+      "epoch": 2.0709134615384617,
+      "grad_norm": 0.25118470405380966,
+      "learning_rate": 1.7208370436331256e-05,
+      "loss": 0.319,
+      "step": 1723
+    },
+    {
+      "epoch": 2.0721153846153846,
+      "grad_norm": 0.24549453047595804,
+      "learning_rate": 1.7186108637577915e-05,
+      "loss": 0.3244,
+      "step": 1724
+    },
+    {
+      "epoch": 2.0733173076923075,
+      "grad_norm": 0.25057698544081247,
+      "learning_rate": 1.7163846838824577e-05,
+      "loss": 0.308,
+      "step": 1725
+    },
+    {
+      "epoch": 2.074519230769231,
+      "grad_norm": 0.2408057809389488,
+      "learning_rate": 1.714158504007124e-05,
+      "loss": 0.3189,
+      "step": 1726
+    },
+    {
+      "epoch": 2.0757211538461537,
+      "grad_norm": 0.2891232695278914,
+      "learning_rate": 1.71193232413179e-05,
+      "loss": 0.3187,
+      "step": 1727
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.2586535970073632,
+      "learning_rate": 1.709706144256456e-05,
+      "loss": 0.3551,
+      "step": 1728
+    },
+    {
+      "epoch": 2.078125,
+      "grad_norm": 0.29409412652085576,
+      "learning_rate": 1.7074799643811222e-05,
+      "loss": 0.3466,
+      "step": 1729
+    },
+    {
+      "epoch": 2.079326923076923,
+      "grad_norm": 0.24679365190495126,
+      "learning_rate": 1.705253784505788e-05,
+      "loss": 0.3321,
+      "step": 1730
+    },
+    {
+      "epoch": 2.0805288461538463,
+      "grad_norm": 0.2399613699354062,
+      "learning_rate": 1.703027604630454e-05,
+      "loss": 0.3131,
+      "step": 1731
+    },
+    {
+      "epoch": 2.081730769230769,
+      "grad_norm": 0.30009648111767656,
+      "learning_rate": 1.7008014247551202e-05,
+      "loss": 0.3409,
+      "step": 1732
+    },
+    {
+      "epoch": 2.0829326923076925,
+      "grad_norm": 0.2506872991583116,
+      "learning_rate": 1.6985752448797864e-05,
+      "loss": 0.3365,
+      "step": 1733
+    },
+    {
+      "epoch": 2.0841346153846154,
+      "grad_norm": 0.284202340801502,
+      "learning_rate": 1.6963490650044526e-05,
+      "loss": 0.3243,
+      "step": 1734
+    },
+    {
+      "epoch": 2.0853365384615383,
+      "grad_norm": 0.3111993686782253,
+      "learning_rate": 1.6941228851291185e-05,
+      "loss": 0.3425,
+      "step": 1735
+    },
+    {
+      "epoch": 2.0865384615384617,
+      "grad_norm": 0.25631061665580374,
+      "learning_rate": 1.6918967052537847e-05,
+      "loss": 0.3284,
+      "step": 1736
+    },
+    {
+      "epoch": 2.0877403846153846,
+      "grad_norm": 0.23892892279813943,
+      "learning_rate": 1.6896705253784506e-05,
+      "loss": 0.3402,
+      "step": 1737
+    },
+    {
+      "epoch": 2.0889423076923075,
+      "grad_norm": 0.3194477422448464,
+      "learning_rate": 1.6874443455031168e-05,
+      "loss": 0.3053,
+      "step": 1738
+    },
+    {
+      "epoch": 2.090144230769231,
+      "grad_norm": 0.2568698038042167,
+      "learning_rate": 1.6852181656277827e-05,
+      "loss": 0.3191,
+      "step": 1739
+    },
+    {
+      "epoch": 2.0913461538461537,
+      "grad_norm": 0.2465299174854252,
+      "learning_rate": 1.682991985752449e-05,
+      "loss": 0.318,
+      "step": 1740
+    },
+    {
+      "epoch": 2.092548076923077,
+      "grad_norm": 0.26568938418564936,
+      "learning_rate": 1.680765805877115e-05,
+      "loss": 0.3408,
+      "step": 1741
+    },
+    {
+      "epoch": 2.09375,
+      "grad_norm": 0.2607476565695497,
+      "learning_rate": 1.678539626001781e-05,
+      "loss": 0.3037,
+      "step": 1742
+    },
+    {
+      "epoch": 2.094951923076923,
+      "grad_norm": 0.27318264342918913,
+      "learning_rate": 1.676313446126447e-05,
+      "loss": 0.331,
+      "step": 1743
+    },
+    {
+      "epoch": 2.0961538461538463,
+      "grad_norm": 0.24820092335796176,
+      "learning_rate": 1.674087266251113e-05,
+      "loss": 0.3169,
+      "step": 1744
+    },
+    {
+      "epoch": 2.097355769230769,
+      "grad_norm": 0.2580751811036428,
+      "learning_rate": 1.6718610863757793e-05,
+      "loss": 0.3044,
+      "step": 1745
+    },
+    {
+      "epoch": 2.0985576923076925,
+      "grad_norm": 0.2569052280752823,
+      "learning_rate": 1.669634906500445e-05,
+      "loss": 0.3243,
+      "step": 1746
+    },
+    {
+      "epoch": 2.0997596153846154,
+      "grad_norm": 0.25138030177008,
+      "learning_rate": 1.6674087266251113e-05,
+      "loss": 0.328,
+      "step": 1747
+    },
+    {
+      "epoch": 2.1009615384615383,
+      "grad_norm": 0.24412032341763742,
+      "learning_rate": 1.6651825467497776e-05,
+      "loss": 0.3261,
+      "step": 1748
+    },
+    {
+      "epoch": 2.1021634615384617,
+      "grad_norm": 0.2672524687592169,
+      "learning_rate": 1.6629563668744434e-05,
+      "loss": 0.3227,
+      "step": 1749
+    },
+    {
+      "epoch": 2.1033653846153846,
+      "grad_norm": 0.23209355043862034,
+      "learning_rate": 1.6607301869991096e-05,
+      "loss": 0.3222,
+      "step": 1750
+    },
+    {
+      "epoch": 2.1045673076923075,
+      "grad_norm": 0.23690705134589157,
+      "learning_rate": 1.6585040071237755e-05,
+      "loss": 0.3296,
+      "step": 1751
+    },
+    {
+      "epoch": 2.105769230769231,
+      "grad_norm": 0.24477261382796453,
+      "learning_rate": 1.6562778272484417e-05,
+      "loss": 0.3343,
+      "step": 1752
+    },
+    {
+      "epoch": 2.1069711538461537,
+      "grad_norm": 0.2581393063502429,
+      "learning_rate": 1.6540516473731076e-05,
+      "loss": 0.3265,
+      "step": 1753
+    },
+    {
+      "epoch": 2.108173076923077,
+      "grad_norm": 0.2520323872811074,
+      "learning_rate": 1.651825467497774e-05,
+      "loss": 0.3397,
+      "step": 1754
+    },
+    {
+      "epoch": 2.109375,
+      "grad_norm": 0.2398742505473097,
+      "learning_rate": 1.64959928762244e-05,
+      "loss": 0.3206,
+      "step": 1755
+    },
+    {
+      "epoch": 2.110576923076923,
+      "grad_norm": 0.2687800410343687,
+      "learning_rate": 1.647373107747106e-05,
+      "loss": 0.3116,
+      "step": 1756
+    },
+    {
+      "epoch": 2.1117788461538463,
+      "grad_norm": 0.27104059748718823,
+      "learning_rate": 1.645146927871772e-05,
+      "loss": 0.3426,
+      "step": 1757
+    },
+    {
+      "epoch": 2.112980769230769,
+      "grad_norm": 0.2765621138225596,
+      "learning_rate": 1.642920747996438e-05,
+      "loss": 0.3263,
+      "step": 1758
+    },
+    {
+      "epoch": 2.1141826923076925,
+      "grad_norm": 0.28295074523853125,
+      "learning_rate": 1.6406945681211042e-05,
+      "loss": 0.3323,
+      "step": 1759
+    },
+    {
+      "epoch": 2.1153846153846154,
+      "grad_norm": 0.26536447642613537,
+      "learning_rate": 1.63846838824577e-05,
+      "loss": 0.318,
+      "step": 1760
+    },
+    {
+      "epoch": 2.1165865384615383,
+      "grad_norm": 0.2996632206645658,
+      "learning_rate": 1.6362422083704366e-05,
+      "loss": 0.3207,
+      "step": 1761
+    },
+    {
+      "epoch": 2.1177884615384617,
+      "grad_norm": 0.3107839545310344,
+      "learning_rate": 1.6340160284951025e-05,
+      "loss": 0.3322,
+      "step": 1762
+    },
+    {
+      "epoch": 2.1189903846153846,
+      "grad_norm": 0.25689892859627245,
+      "learning_rate": 1.6317898486197687e-05,
+      "loss": 0.3307,
+      "step": 1763
+    },
+    {
+      "epoch": 2.1201923076923075,
+      "grad_norm": 0.2893176561519215,
+      "learning_rate": 1.6295636687444346e-05,
+      "loss": 0.3408,
+      "step": 1764
+    },
+    {
+      "epoch": 2.121394230769231,
+      "grad_norm": 0.28857457767036193,
+      "learning_rate": 1.6273374888691008e-05,
+      "loss": 0.3296,
+      "step": 1765
+    },
+    {
+      "epoch": 2.1225961538461537,
+      "grad_norm": 0.2304979431578095,
+      "learning_rate": 1.6251113089937667e-05,
+      "loss": 0.3167,
+      "step": 1766
+    },
+    {
+      "epoch": 2.123798076923077,
+      "grad_norm": 0.29852062242496524,
+      "learning_rate": 1.6228851291184326e-05,
+      "loss": 0.3147,
+      "step": 1767
+    },
+    {
+      "epoch": 2.125,
+      "grad_norm": 0.3196871832848296,
+      "learning_rate": 1.620658949243099e-05,
+      "loss": 0.3186,
+      "step": 1768
+    },
+    {
+      "epoch": 2.126201923076923,
+      "grad_norm": 0.27752257703055744,
+      "learning_rate": 1.618432769367765e-05,
+      "loss": 0.3306,
+      "step": 1769
+    },
+    {
+      "epoch": 2.1274038461538463,
+      "grad_norm": 0.3005190272453843,
+      "learning_rate": 1.6162065894924312e-05,
+      "loss": 0.3256,
+      "step": 1770
+    },
+    {
+      "epoch": 2.128605769230769,
+      "grad_norm": 0.30648398792633397,
+      "learning_rate": 1.613980409617097e-05,
+      "loss": 0.3268,
+      "step": 1771
+    },
+    {
+      "epoch": 2.1298076923076925,
+      "grad_norm": 0.32522230988273804,
+      "learning_rate": 1.6117542297417633e-05,
+      "loss": 0.3092,
+      "step": 1772
+    },
+    {
+      "epoch": 2.1310096153846154,
+      "grad_norm": 0.2746907218489806,
+      "learning_rate": 1.609528049866429e-05,
+      "loss": 0.3193,
+      "step": 1773
+    },
+    {
+      "epoch": 2.1322115384615383,
+      "grad_norm": 0.25209101524987354,
+      "learning_rate": 1.6073018699910954e-05,
+      "loss": 0.3205,
+      "step": 1774
+    },
+    {
+      "epoch": 2.1334134615384617,
+      "grad_norm": 0.26057378716353663,
+      "learning_rate": 1.6050756901157616e-05,
+      "loss": 0.3392,
+      "step": 1775
+    },
+    {
+      "epoch": 2.1346153846153846,
+      "grad_norm": 0.2859648154582435,
+      "learning_rate": 1.6028495102404275e-05,
+      "loss": 0.3443,
+      "step": 1776
+    },
+    {
+      "epoch": 2.1358173076923075,
+      "grad_norm": 0.2467208522048817,
+      "learning_rate": 1.6006233303650937e-05,
+      "loss": 0.3383,
+      "step": 1777
+    },
+    {
+      "epoch": 2.137019230769231,
+      "grad_norm": 0.24207002011080467,
+      "learning_rate": 1.5983971504897595e-05,
+      "loss": 0.3174,
+      "step": 1778
+    },
+    {
+      "epoch": 2.1382211538461537,
+      "grad_norm": 0.2584765599390294,
+      "learning_rate": 1.5961709706144258e-05,
+      "loss": 0.3247,
+      "step": 1779
+    },
+    {
+      "epoch": 2.139423076923077,
+      "grad_norm": 0.2602781539425154,
+      "learning_rate": 1.5939447907390916e-05,
+      "loss": 0.3369,
+      "step": 1780
+    },
+    {
+      "epoch": 2.140625,
+      "grad_norm": 0.25226550122801783,
+      "learning_rate": 1.591718610863758e-05,
+      "loss": 0.3104,
+      "step": 1781
+    },
+    {
+      "epoch": 2.141826923076923,
+      "grad_norm": 0.27976218634514616,
+      "learning_rate": 1.589492430988424e-05,
+      "loss": 0.343,
+      "step": 1782
+    },
+    {
+      "epoch": 2.1430288461538463,
+      "grad_norm": 0.23583995976574862,
+      "learning_rate": 1.58726625111309e-05,
+      "loss": 0.3291,
+      "step": 1783
+    },
+    {
+      "epoch": 2.144230769230769,
+      "grad_norm": 0.23709679203566417,
+      "learning_rate": 1.585040071237756e-05,
+      "loss": 0.3303,
+      "step": 1784
+    },
+    {
+      "epoch": 2.1454326923076925,
+      "grad_norm": 0.2722852426871755,
+      "learning_rate": 1.582813891362422e-05,
+      "loss": 0.3234,
+      "step": 1785
+    },
+    {
+      "epoch": 2.1466346153846154,
+      "grad_norm": 0.23964979001274983,
+      "learning_rate": 1.5805877114870882e-05,
+      "loss": 0.3197,
+      "step": 1786
+    },
+    {
+      "epoch": 2.1478365384615383,
+      "grad_norm": 0.23929498126292098,
+      "learning_rate": 1.578361531611754e-05,
+      "loss": 0.3238,
+      "step": 1787
+    },
+    {
+      "epoch": 2.1490384615384617,
+      "grad_norm": 0.2569519845285522,
+      "learning_rate": 1.5761353517364203e-05,
+      "loss": 0.3329,
+      "step": 1788
+    },
+    {
+      "epoch": 2.1502403846153846,
+      "grad_norm": 0.3076547675999885,
+      "learning_rate": 1.5739091718610865e-05,
+      "loss": 0.3354,
+      "step": 1789
+    },
+    {
+      "epoch": 2.1514423076923075,
+      "grad_norm": 0.2582724363510655,
+      "learning_rate": 1.5716829919857528e-05,
+      "loss": 0.3234,
+      "step": 1790
+    },
+    {
+      "epoch": 2.152644230769231,
+      "grad_norm": 0.22719891161088324,
+      "learning_rate": 1.5694568121104186e-05,
+      "loss": 0.314,
+      "step": 1791
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.2458264947774297,
+      "learning_rate": 1.5672306322350845e-05,
+      "loss": 0.3351,
+      "step": 1792
+    },
+    {
+      "epoch": 2.155048076923077,
+      "grad_norm": 0.2551501557665136,
+      "learning_rate": 1.5650044523597507e-05,
+      "loss": 0.322,
+      "step": 1793
+    },
+    {
+      "epoch": 2.15625,
+      "grad_norm": 0.23182778985447478,
+      "learning_rate": 1.5627782724844166e-05,
+      "loss": 0.335,
+      "step": 1794
+    },
+    {
+      "epoch": 2.157451923076923,
+      "grad_norm": 0.29231276059356875,
+      "learning_rate": 1.560552092609083e-05,
+      "loss": 0.3386,
+      "step": 1795
+    },
+    {
+      "epoch": 2.1586538461538463,
+      "grad_norm": 0.2364634840856608,
+      "learning_rate": 1.558325912733749e-05,
+      "loss": 0.3163,
+      "step": 1796
+    },
+    {
+      "epoch": 2.159855769230769,
+      "grad_norm": 0.80704274780889,
+      "learning_rate": 1.5560997328584152e-05,
+      "loss": 0.3373,
+      "step": 1797
+    },
+    {
+      "epoch": 2.1610576923076925,
+      "grad_norm": 0.22923395004821165,
+      "learning_rate": 1.553873552983081e-05,
+      "loss": 0.3184,
+      "step": 1798
+    },
+    {
+      "epoch": 2.1622596153846154,
+      "grad_norm": 0.3046614000821981,
+      "learning_rate": 1.5516473731077473e-05,
+      "loss": 0.3493,
+      "step": 1799
+    },
+    {
+      "epoch": 2.1634615384615383,
+      "grad_norm": 0.23459020431581834,
+      "learning_rate": 1.5494211932324132e-05,
+      "loss": 0.3276,
+      "step": 1800
+    },
+    {
+      "epoch": 2.1646634615384617,
+      "grad_norm": 0.2700889367425373,
+      "learning_rate": 1.547195013357079e-05,
+      "loss": 0.3257,
+      "step": 1801
+    },
+    {
+      "epoch": 2.1658653846153846,
+      "grad_norm": 0.24752336528790309,
+      "learning_rate": 1.5449688334817456e-05,
+      "loss": 0.3342,
+      "step": 1802
+    },
+    {
+      "epoch": 2.1670673076923075,
+      "grad_norm": 0.2510308371540042,
+      "learning_rate": 1.5427426536064115e-05,
+      "loss": 0.3341,
+      "step": 1803
+    },
+    {
+      "epoch": 2.168269230769231,
+      "grad_norm": 0.26100488702019353,
+      "learning_rate": 1.5405164737310777e-05,
+      "loss": 0.3214,
+      "step": 1804
+    },
+    {
+      "epoch": 2.1694711538461537,
+      "grad_norm": 0.2528235525809164,
+      "learning_rate": 1.5382902938557436e-05,
+      "loss": 0.3318,
+      "step": 1805
+    },
+    {
+      "epoch": 2.170673076923077,
+      "grad_norm": 0.26320750432962303,
+      "learning_rate": 1.5360641139804098e-05,
+      "loss": 0.3148,
+      "step": 1806
+    },
+    {
+      "epoch": 2.171875,
+      "grad_norm": 0.290894332023697,
+      "learning_rate": 1.5338379341050757e-05,
+      "loss": 0.3087,
+      "step": 1807
+    },
+    {
+      "epoch": 2.173076923076923,
+      "grad_norm": 0.29004796771012903,
+      "learning_rate": 1.531611754229742e-05,
+      "loss": 0.3404,
+      "step": 1808
+    },
+    {
+      "epoch": 2.1742788461538463,
+      "grad_norm": 0.24473494406173127,
+      "learning_rate": 1.529385574354408e-05,
+      "loss": 0.3302,
+      "step": 1809
+    },
+    {
+      "epoch": 2.175480769230769,
+      "grad_norm": 0.2945444781354667,
+      "learning_rate": 1.527159394479074e-05,
+      "loss": 0.3261,
+      "step": 1810
+    },
+    {
+      "epoch": 2.1766826923076925,
+      "grad_norm": 0.26809359063967125,
+      "learning_rate": 1.5249332146037402e-05,
+      "loss": 0.3299,
+      "step": 1811
+    },
+    {
+      "epoch": 2.1778846153846154,
+      "grad_norm": 0.2514094143619762,
+      "learning_rate": 1.5227070347284062e-05,
+      "loss": 0.3224,
+      "step": 1812
+    },
+    {
+      "epoch": 2.1790865384615383,
+      "grad_norm": 0.2604316134696773,
+      "learning_rate": 1.5204808548530721e-05,
+      "loss": 0.3147,
+      "step": 1813
+    },
+    {
+      "epoch": 2.1802884615384617,
+      "grad_norm": 0.2812473744394753,
+      "learning_rate": 1.5182546749777381e-05,
+      "loss": 0.3189,
+      "step": 1814
+    },
+    {
+      "epoch": 2.1814903846153846,
+      "grad_norm": 0.2725674103734274,
+      "learning_rate": 1.5160284951024042e-05,
+      "loss": 0.3328,
+      "step": 1815
+    },
+    {
+      "epoch": 2.1826923076923075,
+      "grad_norm": 0.26205777323826157,
+      "learning_rate": 1.5138023152270706e-05,
+      "loss": 0.3357,
+      "step": 1816
+    },
+    {
+      "epoch": 2.183894230769231,
+      "grad_norm": 0.2686761392124138,
+      "learning_rate": 1.5115761353517366e-05,
+      "loss": 0.3258,
+      "step": 1817
+    },
+    {
+      "epoch": 2.1850961538461537,
+      "grad_norm": 0.24582281864411373,
+      "learning_rate": 1.5093499554764027e-05,
+      "loss": 0.3096,
+      "step": 1818
+    },
+    {
+      "epoch": 2.186298076923077,
+      "grad_norm": 0.23393961153192067,
+      "learning_rate": 1.5071237756010687e-05,
+      "loss": 0.3152,
+      "step": 1819
+    },
+    {
+      "epoch": 2.1875,
+      "grad_norm": 0.2503124896415827,
+      "learning_rate": 1.5048975957257347e-05,
+      "loss": 0.3307,
+      "step": 1820
+    },
+    {
+      "epoch": 2.188701923076923,
+      "grad_norm": 0.2567707912122752,
+      "learning_rate": 1.5026714158504008e-05,
+      "loss": 0.326,
+      "step": 1821
+    },
+    {
+      "epoch": 2.1899038461538463,
+      "grad_norm": 0.23725422538804466,
+      "learning_rate": 1.5004452359750668e-05,
+      "loss": 0.3241,
+      "step": 1822
+    },
+    {
+      "epoch": 2.191105769230769,
+      "grad_norm": 0.22234333662558678,
+      "learning_rate": 1.498219056099733e-05,
+      "loss": 0.3095,
+      "step": 1823
+    },
+    {
+      "epoch": 2.1923076923076925,
+      "grad_norm": 0.24029157000685208,
+      "learning_rate": 1.4959928762243991e-05,
+      "loss": 0.3313,
+      "step": 1824
+    },
+    {
+      "epoch": 2.1935096153846154,
+      "grad_norm": 0.30695255842893016,
+      "learning_rate": 1.4937666963490651e-05,
+      "loss": 0.3388,
+      "step": 1825
+    },
+    {
+      "epoch": 2.1947115384615383,
+      "grad_norm": 0.23492902150734063,
+      "learning_rate": 1.4915405164737312e-05,
+      "loss": 0.3154,
+      "step": 1826
+    },
+    {
+      "epoch": 2.1959134615384617,
+      "grad_norm": 0.22975416775907595,
+      "learning_rate": 1.4893143365983972e-05,
+      "loss": 0.3394,
+      "step": 1827
+    },
+    {
+      "epoch": 2.1971153846153846,
+      "grad_norm": 1.1261206150437395,
+      "learning_rate": 1.4870881567230633e-05,
+      "loss": 0.358,
+      "step": 1828
+    },
+    {
+      "epoch": 2.1983173076923075,
+      "grad_norm": 0.22637064873362425,
+      "learning_rate": 1.4848619768477293e-05,
+      "loss": 0.3166,
+      "step": 1829
+    },
+    {
+      "epoch": 2.199519230769231,
+      "grad_norm": 0.2783692255456279,
+      "learning_rate": 1.4826357969723955e-05,
+      "loss": 0.3374,
+      "step": 1830
+    },
+    {
+      "epoch": 2.2007211538461537,
+      "grad_norm": 0.2294219871916797,
+      "learning_rate": 1.4804096170970616e-05,
+      "loss": 0.3348,
+      "step": 1831
+    },
+    {
+      "epoch": 2.201923076923077,
+      "grad_norm": 0.28049361966937914,
+      "learning_rate": 1.4781834372217276e-05,
+      "loss": 0.3315,
+      "step": 1832
+    },
+    {
+      "epoch": 2.203125,
+      "grad_norm": 0.2465010830075706,
+      "learning_rate": 1.4759572573463936e-05,
+      "loss": 0.308,
+      "step": 1833
+    },
+    {
+      "epoch": 2.204326923076923,
+      "grad_norm": 0.23342461795742311,
+      "learning_rate": 1.4737310774710597e-05,
+      "loss": 0.3314,
+      "step": 1834
+    },
+    {
+      "epoch": 2.2055288461538463,
+      "grad_norm": 0.2537730198780229,
+      "learning_rate": 1.4715048975957257e-05,
+      "loss": 0.3286,
+      "step": 1835
+    },
+    {
+      "epoch": 2.206730769230769,
+      "grad_norm": 0.24936045605189108,
+      "learning_rate": 1.4692787177203918e-05,
+      "loss": 0.3252,
+      "step": 1836
+    },
+    {
+      "epoch": 2.2079326923076925,
+      "grad_norm": 0.243863356243492,
+      "learning_rate": 1.4670525378450582e-05,
+      "loss": 0.3362,
+      "step": 1837
+    },
+    {
+      "epoch": 2.2091346153846154,
+      "grad_norm": 0.2360670228653486,
+      "learning_rate": 1.464826357969724e-05,
+      "loss": 0.3178,
+      "step": 1838
+    },
+    {
+      "epoch": 2.2103365384615383,
+      "grad_norm": 0.2356090550142569,
+      "learning_rate": 1.46260017809439e-05,
+      "loss": 0.3219,
+      "step": 1839
+    },
+    {
+      "epoch": 2.2115384615384617,
+      "grad_norm": 0.23906699470138132,
+      "learning_rate": 1.4603739982190561e-05,
+      "loss": 0.3162,
+      "step": 1840
+    },
+    {
+      "epoch": 2.2127403846153846,
+      "grad_norm": 0.2638069283450429,
+      "learning_rate": 1.4581478183437222e-05,
+      "loss": 0.3141,
+      "step": 1841
+    },
+    {
+      "epoch": 2.2139423076923075,
+      "grad_norm": 0.2518642845084144,
+      "learning_rate": 1.4559216384683882e-05,
+      "loss": 0.3002,
+      "step": 1842
+    },
+    {
+      "epoch": 2.215144230769231,
+      "grad_norm": 0.25760503976068666,
+      "learning_rate": 1.4536954585930543e-05,
+      "loss": 0.317,
+      "step": 1843
+    },
+    {
+      "epoch": 2.2163461538461537,
+      "grad_norm": 0.24746325042344505,
+      "learning_rate": 1.4514692787177206e-05,
+      "loss": 0.3283,
+      "step": 1844
+    },
+    {
+      "epoch": 2.217548076923077,
+      "grad_norm": 0.26461097140039697,
+      "learning_rate": 1.4492430988423867e-05,
+      "loss": 0.3373,
+      "step": 1845
+    },
+    {
+      "epoch": 2.21875,
+      "grad_norm": 0.24566470564380619,
+      "learning_rate": 1.4470169189670527e-05,
+      "loss": 0.3186,
+      "step": 1846
+    },
+    {
+      "epoch": 2.219951923076923,
+      "grad_norm": 0.2556337470682149,
+      "learning_rate": 1.4447907390917188e-05,
+      "loss": 0.3243,
+      "step": 1847
+    },
+    {
+      "epoch": 2.2211538461538463,
+      "grad_norm": 0.25935244235624916,
+      "learning_rate": 1.4425645592163846e-05,
+      "loss": 0.3142,
+      "step": 1848
+    },
+    {
+      "epoch": 2.222355769230769,
+      "grad_norm": 0.23155917085166808,
+      "learning_rate": 1.4403383793410507e-05,
+      "loss": 0.3329,
+      "step": 1849
+    },
+    {
+      "epoch": 2.2235576923076925,
+      "grad_norm": 0.292089016793862,
+      "learning_rate": 1.4381121994657167e-05,
+      "loss": 0.3334,
+      "step": 1850
+    },
+    {
+      "epoch": 2.2247596153846154,
+      "grad_norm": 0.2463683271961105,
+      "learning_rate": 1.4358860195903831e-05,
+      "loss": 0.3236,
+      "step": 1851
+    },
+    {
+      "epoch": 2.2259615384615383,
+      "grad_norm": 0.22975832951272282,
+      "learning_rate": 1.4336598397150492e-05,
+      "loss": 0.3403,
+      "step": 1852
+    },
+    {
+      "epoch": 2.2271634615384617,
+      "grad_norm": 0.23802951454789065,
+      "learning_rate": 1.4314336598397152e-05,
+      "loss": 0.329,
+      "step": 1853
+    },
+    {
+      "epoch": 2.2283653846153846,
+      "grad_norm": 0.2197015518983584,
+      "learning_rate": 1.4292074799643812e-05,
+      "loss": 0.331,
+      "step": 1854
+    },
+    {
+      "epoch": 2.2295673076923075,
+      "grad_norm": 0.23630826153019324,
+      "learning_rate": 1.4269813000890473e-05,
+      "loss": 0.3334,
+      "step": 1855
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.2421531692110706,
+      "learning_rate": 1.4247551202137133e-05,
+      "loss": 0.3086,
+      "step": 1856
+    },
+    {
+      "epoch": 2.2319711538461537,
+      "grad_norm": 0.2532012265809135,
+      "learning_rate": 1.4225289403383794e-05,
+      "loss": 0.3242,
+      "step": 1857
+    },
+    {
+      "epoch": 2.233173076923077,
+      "grad_norm": 0.23589281488408492,
+      "learning_rate": 1.4203027604630456e-05,
+      "loss": 0.3282,
+      "step": 1858
+    },
+    {
+      "epoch": 2.234375,
+      "grad_norm": 0.2322791070771825,
+      "learning_rate": 1.4180765805877116e-05,
+      "loss": 0.3137,
+      "step": 1859
+    },
+    {
+      "epoch": 2.235576923076923,
+      "grad_norm": 0.22470064384043817,
+      "learning_rate": 1.4158504007123777e-05,
+      "loss": 0.2998,
+      "step": 1860
+    },
+    {
+      "epoch": 2.2367788461538463,
+      "grad_norm": 0.24856480856286245,
+      "learning_rate": 1.4136242208370437e-05,
+      "loss": 0.3222,
+      "step": 1861
+    },
+    {
+      "epoch": 2.237980769230769,
+      "grad_norm": 0.22287974472455002,
+      "learning_rate": 1.4113980409617098e-05,
+      "loss": 0.3264,
+      "step": 1862
+    },
+    {
+      "epoch": 2.2391826923076925,
+      "grad_norm": 0.21521772071488843,
+      "learning_rate": 1.4091718610863758e-05,
+      "loss": 0.3161,
+      "step": 1863
+    },
+    {
+      "epoch": 2.2403846153846154,
+      "grad_norm": 0.24425893596282805,
+      "learning_rate": 1.4069456812110419e-05,
+      "loss": 0.3209,
+      "step": 1864
+    },
+    {
+      "epoch": 2.2415865384615383,
+      "grad_norm": 0.24952223440271837,
+      "learning_rate": 1.404719501335708e-05,
+      "loss": 0.3297,
+      "step": 1865
+    },
+    {
+      "epoch": 2.2427884615384617,
+      "grad_norm": 0.2243364930928503,
+      "learning_rate": 1.4024933214603741e-05,
+      "loss": 0.3228,
+      "step": 1866
+    },
+    {
+      "epoch": 2.2439903846153846,
+      "grad_norm": 0.21845028136483546,
+      "learning_rate": 1.4002671415850402e-05,
+      "loss": 0.3256,
+      "step": 1867
+    },
+    {
+      "epoch": 2.2451923076923075,
+      "grad_norm": 0.23162773961847113,
+      "learning_rate": 1.3980409617097062e-05,
+      "loss": 0.3261,
+      "step": 1868
+    },
+    {
+      "epoch": 2.246394230769231,
+      "grad_norm": 0.24485807936100232,
+      "learning_rate": 1.3958147818343722e-05,
+      "loss": 0.326,
+      "step": 1869
+    },
+    {
+      "epoch": 2.2475961538461537,
+      "grad_norm": 0.21695542281151148,
+      "learning_rate": 1.3935886019590383e-05,
+      "loss": 0.3133,
+      "step": 1870
+    },
+    {
+      "epoch": 2.248798076923077,
+      "grad_norm": 1.5069909806636488,
+      "learning_rate": 1.3913624220837043e-05,
+      "loss": 0.364,
+      "step": 1871
+    },
+    {
+      "epoch": 2.25,
+      "grad_norm": 0.24318524249658552,
+      "learning_rate": 1.3891362422083707e-05,
+      "loss": 0.3339,
+      "step": 1872
+    },
+    {
+      "epoch": 2.251201923076923,
+      "grad_norm": 0.26162408234543444,
+      "learning_rate": 1.3869100623330366e-05,
+      "loss": 0.3375,
+      "step": 1873
+    },
+    {
+      "epoch": 2.2524038461538463,
+      "grad_norm": 0.26385777461225796,
+      "learning_rate": 1.3846838824577026e-05,
+      "loss": 0.3285,
+      "step": 1874
+    },
+    {
+      "epoch": 2.253605769230769,
+      "grad_norm": 0.27203197093927056,
+      "learning_rate": 1.3824577025823687e-05,
+      "loss": 0.3446,
+      "step": 1875
+    },
+    {
+      "epoch": 2.2548076923076925,
+      "grad_norm": 0.25946280850893444,
+      "learning_rate": 1.3802315227070347e-05,
+      "loss": 0.3301,
+      "step": 1876
+    },
+    {
+      "epoch": 2.2560096153846154,
+      "grad_norm": 0.23340643251974943,
+      "learning_rate": 1.3780053428317008e-05,
+      "loss": 0.3335,
+      "step": 1877
+    },
+    {
+      "epoch": 2.2572115384615383,
+      "grad_norm": 0.25722949777433823,
+      "learning_rate": 1.3757791629563668e-05,
+      "loss": 0.3292,
+      "step": 1878
+    },
+    {
+      "epoch": 2.2584134615384617,
+      "grad_norm": 0.2529374930533914,
+      "learning_rate": 1.3735529830810332e-05,
+      "loss": 0.3309,
+      "step": 1879
+    },
+    {
+      "epoch": 2.2596153846153846,
+      "grad_norm": 0.2205773370430708,
+      "learning_rate": 1.3713268032056992e-05,
+      "loss": 0.3211,
+      "step": 1880
+    },
+    {
+      "epoch": 2.2608173076923075,
+      "grad_norm": 0.21466784871280326,
+      "learning_rate": 1.3691006233303653e-05,
+      "loss": 0.3071,
+      "step": 1881
+    },
+    {
+      "epoch": 2.262019230769231,
+      "grad_norm": 0.23947967320391464,
+      "learning_rate": 1.3668744434550313e-05,
+      "loss": 0.323,
+      "step": 1882
+    },
+    {
+      "epoch": 2.2632211538461537,
+      "grad_norm": 0.24143046455066333,
+      "learning_rate": 1.3646482635796972e-05,
+      "loss": 0.3179,
+      "step": 1883
+    },
+    {
+      "epoch": 2.264423076923077,
+      "grad_norm": 0.22770555824264302,
+      "learning_rate": 1.3624220837043632e-05,
+      "loss": 0.3365,
+      "step": 1884
+    },
+    {
+      "epoch": 2.265625,
+      "grad_norm": 0.23360264606470268,
+      "learning_rate": 1.3601959038290293e-05,
+      "loss": 0.3182,
+      "step": 1885
+    },
+    {
+      "epoch": 2.266826923076923,
+      "grad_norm": 0.23144541961583498,
+      "learning_rate": 1.3579697239536957e-05,
+      "loss": 0.3142,
+      "step": 1886
+    },
+    {
+      "epoch": 2.2680288461538463,
+      "grad_norm": 0.2362221558231656,
+      "learning_rate": 1.3557435440783617e-05,
+      "loss": 0.3153,
+      "step": 1887
+    },
+    {
+      "epoch": 2.269230769230769,
+      "grad_norm": 0.24042854012303025,
+      "learning_rate": 1.3535173642030278e-05,
+      "loss": 0.3449,
+      "step": 1888
+    },
+    {
+      "epoch": 2.2704326923076925,
+      "grad_norm": 0.2376840329551465,
+      "learning_rate": 1.3512911843276938e-05,
+      "loss": 0.3566,
+      "step": 1889
+    },
+    {
+      "epoch": 2.2716346153846154,
+      "grad_norm": 0.23408779314527883,
+      "learning_rate": 1.3490650044523598e-05,
+      "loss": 0.3433,
+      "step": 1890
+    },
+    {
+      "epoch": 2.2728365384615383,
+      "grad_norm": 0.24407519672126185,
+      "learning_rate": 1.3468388245770259e-05,
+      "loss": 0.3283,
+      "step": 1891
+    },
+    {
+      "epoch": 2.2740384615384617,
+      "grad_norm": 0.2333085720315056,
+      "learning_rate": 1.344612644701692e-05,
+      "loss": 0.3049,
+      "step": 1892
+    },
+    {
+      "epoch": 2.2752403846153846,
+      "grad_norm": 0.23409040386005217,
+      "learning_rate": 1.3423864648263581e-05,
+      "loss": 0.3229,
+      "step": 1893
+    },
+    {
+      "epoch": 2.2764423076923075,
+      "grad_norm": 0.24234526377841725,
+      "learning_rate": 1.3401602849510242e-05,
+      "loss": 0.3131,
+      "step": 1894
+    },
+    {
+      "epoch": 2.277644230769231,
+      "grad_norm": 0.22623672816120494,
+      "learning_rate": 1.3379341050756902e-05,
+      "loss": 0.3192,
+      "step": 1895
+    },
+    {
+      "epoch": 2.2788461538461537,
+      "grad_norm": 0.2283508115791479,
+      "learning_rate": 1.3357079252003563e-05,
+      "loss": 0.3228,
+      "step": 1896
+    },
+    {
+      "epoch": 2.280048076923077,
+      "grad_norm": 0.22116447377424353,
+      "learning_rate": 1.3334817453250223e-05,
+      "loss": 0.3195,
+      "step": 1897
+    },
+    {
+      "epoch": 2.28125,
+      "grad_norm": 0.2686061896843144,
+      "learning_rate": 1.3312555654496884e-05,
+      "loss": 0.3273,
+      "step": 1898
+    },
+    {
+      "epoch": 2.282451923076923,
+      "grad_norm": 0.24059429377674654,
+      "learning_rate": 1.3290293855743544e-05,
+      "loss": 0.3235,
+      "step": 1899
+    },
+    {
+      "epoch": 2.2836538461538463,
+      "grad_norm": 0.2559580504896581,
+      "learning_rate": 1.3268032056990206e-05,
+      "loss": 0.3459,
+      "step": 1900
+    },
+    {
+      "epoch": 2.284855769230769,
+      "grad_norm": 0.23061995754748565,
+      "learning_rate": 1.3245770258236867e-05,
+      "loss": 0.3244,
+      "step": 1901
+    },
+    {
+      "epoch": 2.2860576923076925,
+      "grad_norm": 0.24057157765116782,
+      "learning_rate": 1.3223508459483527e-05,
+      "loss": 0.3216,
+      "step": 1902
+    },
+    {
+      "epoch": 2.2872596153846154,
+      "grad_norm": 0.22647009818839947,
+      "learning_rate": 1.3201246660730187e-05,
+      "loss": 0.3199,
+      "step": 1903
+    },
+    {
+      "epoch": 2.2884615384615383,
+      "grad_norm": 0.2323173132695458,
+      "learning_rate": 1.3178984861976848e-05,
+      "loss": 0.3059,
+      "step": 1904
+    },
+    {
+      "epoch": 2.2896634615384617,
+      "grad_norm": 0.2517534028294729,
+      "learning_rate": 1.3156723063223508e-05,
+      "loss": 0.3089,
+      "step": 1905
+    },
+    {
+      "epoch": 2.2908653846153846,
+      "grad_norm": 0.23890059713901884,
+      "learning_rate": 1.3134461264470169e-05,
+      "loss": 0.316,
+      "step": 1906
+    },
+    {
+      "epoch": 2.2920673076923075,
+      "grad_norm": 0.2118074667122507,
+      "learning_rate": 1.3112199465716831e-05,
+      "loss": 0.3006,
+      "step": 1907
+    },
+    {
+      "epoch": 2.293269230769231,
+      "grad_norm": 0.23492003988937696,
+      "learning_rate": 1.3089937666963491e-05,
+      "loss": 0.3497,
+      "step": 1908
+    },
+    {
+      "epoch": 2.2944711538461537,
+      "grad_norm": 0.22459938919270764,
+      "learning_rate": 1.3067675868210152e-05,
+      "loss": 0.3086,
+      "step": 1909
+    },
+    {
+      "epoch": 2.295673076923077,
+      "grad_norm": 0.24338486110000823,
+      "learning_rate": 1.3045414069456812e-05,
+      "loss": 0.325,
+      "step": 1910
+    },
+    {
+      "epoch": 2.296875,
+      "grad_norm": 0.262095925235346,
+      "learning_rate": 1.3023152270703473e-05,
+      "loss": 0.3744,
+      "step": 1911
+    },
+    {
+      "epoch": 2.298076923076923,
+      "grad_norm": 0.24330228258681583,
+      "learning_rate": 1.3000890471950133e-05,
+      "loss": 0.3234,
+      "step": 1912
+    },
+    {
+      "epoch": 2.2992788461538463,
+      "grad_norm": 0.26431874945097766,
+      "learning_rate": 1.2978628673196794e-05,
+      "loss": 0.3201,
+      "step": 1913
+    },
+    {
+      "epoch": 2.300480769230769,
+      "grad_norm": 0.247012162806937,
+      "learning_rate": 1.2956366874443457e-05,
+      "loss": 0.3238,
+      "step": 1914
+    },
+    {
+      "epoch": 2.3016826923076925,
+      "grad_norm": 0.25040497806511114,
+      "learning_rate": 1.2934105075690118e-05,
+      "loss": 0.351,
+      "step": 1915
+    },
+    {
+      "epoch": 2.3028846153846154,
+      "grad_norm": 0.23863599672113203,
+      "learning_rate": 1.2911843276936778e-05,
+      "loss": 0.3242,
+      "step": 1916
+    },
+    {
+      "epoch": 2.3040865384615383,
+      "grad_norm": 0.28091115079621576,
+      "learning_rate": 1.2889581478183439e-05,
+      "loss": 0.3115,
+      "step": 1917
+    },
+    {
+      "epoch": 2.3052884615384617,
+      "grad_norm": 0.26443239972190297,
+      "learning_rate": 1.2867319679430097e-05,
+      "loss": 0.3024,
+      "step": 1918
+    },
+    {
+      "epoch": 2.3064903846153846,
+      "grad_norm": 0.21495869975885362,
+      "learning_rate": 1.2845057880676758e-05,
+      "loss": 0.3148,
+      "step": 1919
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.25341811684075427,
+      "learning_rate": 1.2822796081923418e-05,
+      "loss": 0.31,
+      "step": 1920
+    },
+    {
+      "epoch": 2.308894230769231,
+      "grad_norm": 0.2326150503624291,
+      "learning_rate": 1.2800534283170082e-05,
+      "loss": 0.3374,
+      "step": 1921
+    },
+    {
+      "epoch": 2.3100961538461537,
+      "grad_norm": 0.2239939926865567,
+      "learning_rate": 1.2778272484416743e-05,
+      "loss": 0.3198,
+      "step": 1922
+    },
+    {
+      "epoch": 2.311298076923077,
+      "grad_norm": 0.2152793762645785,
+      "learning_rate": 1.2756010685663403e-05,
+      "loss": 0.3214,
+      "step": 1923
+    },
+    {
+      "epoch": 2.3125,
+      "grad_norm": 0.2245180962752487,
+      "learning_rate": 1.2733748886910063e-05,
+      "loss": 0.3069,
+      "step": 1924
+    },
+    {
+      "epoch": 2.313701923076923,
+      "grad_norm": 0.22884175642034582,
+      "learning_rate": 1.2711487088156724e-05,
+      "loss": 0.3208,
+      "step": 1925
+    },
+    {
+      "epoch": 2.3149038461538463,
+      "grad_norm": 0.2141595781869691,
+      "learning_rate": 1.2689225289403384e-05,
+      "loss": 0.325,
+      "step": 1926
+    },
+    {
+      "epoch": 2.316105769230769,
+      "grad_norm": 0.22194517254840962,
+      "learning_rate": 1.2666963490650045e-05,
+      "loss": 0.317,
+      "step": 1927
+    },
+    {
+      "epoch": 2.3173076923076925,
+      "grad_norm": 0.2409704065274669,
+      "learning_rate": 1.2644701691896707e-05,
+      "loss": 0.2997,
+      "step": 1928
+    },
+    {
+      "epoch": 2.3185096153846154,
+      "grad_norm": 0.23234842462938912,
+      "learning_rate": 1.2622439893143367e-05,
+      "loss": 0.3271,
+      "step": 1929
+    },
+    {
+      "epoch": 2.3197115384615383,
+      "grad_norm": 0.22711207288736082,
+      "learning_rate": 1.2600178094390028e-05,
+      "loss": 0.3216,
+      "step": 1930
+    },
+    {
+      "epoch": 2.3209134615384617,
+      "grad_norm": 0.2606141152586289,
+      "learning_rate": 1.2577916295636688e-05,
+      "loss": 0.3411,
+      "step": 1931
+    },
+    {
+      "epoch": 2.3221153846153846,
+      "grad_norm": 0.2288890594463383,
+      "learning_rate": 1.2555654496883349e-05,
+      "loss": 0.3256,
+      "step": 1932
+    },
+    {
+      "epoch": 2.3233173076923075,
+      "grad_norm": 0.2217254437082552,
+      "learning_rate": 1.2533392698130009e-05,
+      "loss": 0.3093,
+      "step": 1933
+    },
+    {
+      "epoch": 2.324519230769231,
+      "grad_norm": 0.24816958773783251,
+      "learning_rate": 1.251113089937667e-05,
+      "loss": 0.318,
+      "step": 1934
+    },
+    {
+      "epoch": 2.3257211538461537,
+      "grad_norm": 0.22472439921017798,
+      "learning_rate": 1.248886910062333e-05,
+      "loss": 0.3101,
+      "step": 1935
+    },
+    {
+      "epoch": 2.326923076923077,
+      "grad_norm": 0.26411994831470537,
+      "learning_rate": 1.246660730186999e-05,
+      "loss": 0.3182,
+      "step": 1936
+    },
+    {
+      "epoch": 2.328125,
+      "grad_norm": 0.22114125314075841,
+      "learning_rate": 1.2444345503116653e-05,
+      "loss": 0.3346,
+      "step": 1937
+    },
+    {
+      "epoch": 2.329326923076923,
+      "grad_norm": 0.22612614877499917,
+      "learning_rate": 1.2422083704363313e-05,
+      "loss": 0.3321,
+      "step": 1938
+    },
+    {
+      "epoch": 2.3305288461538463,
+      "grad_norm": 0.22690644714636513,
+      "learning_rate": 1.2399821905609973e-05,
+      "loss": 0.3165,
+      "step": 1939
+    },
+    {
+      "epoch": 2.331730769230769,
+      "grad_norm": 0.2212790486298027,
+      "learning_rate": 1.2377560106856636e-05,
+      "loss": 0.3094,
+      "step": 1940
+    },
+    {
+      "epoch": 2.3329326923076925,
+      "grad_norm": 0.24928430201450635,
+      "learning_rate": 1.2355298308103296e-05,
+      "loss": 0.3245,
+      "step": 1941
+    },
+    {
+      "epoch": 2.3341346153846154,
+      "grad_norm": 0.2463040713732894,
+      "learning_rate": 1.2333036509349956e-05,
+      "loss": 0.3236,
+      "step": 1942
+    },
+    {
+      "epoch": 2.3353365384615383,
+      "grad_norm": 0.23877717475216842,
+      "learning_rate": 1.2310774710596617e-05,
+      "loss": 0.3314,
+      "step": 1943
+    },
+    {
+      "epoch": 2.3365384615384617,
+      "grad_norm": 0.233481416912988,
+      "learning_rate": 1.2288512911843277e-05,
+      "loss": 0.3303,
+      "step": 1944
+    },
+    {
+      "epoch": 2.3377403846153846,
+      "grad_norm": 0.2674549079080054,
+      "learning_rate": 1.2266251113089938e-05,
+      "loss": 0.3344,
+      "step": 1945
+    },
+    {
+      "epoch": 2.3389423076923075,
+      "grad_norm": 0.2580071787038362,
+      "learning_rate": 1.2243989314336598e-05,
+      "loss": 0.3058,
+      "step": 1946
+    },
+    {
+      "epoch": 2.340144230769231,
+      "grad_norm": 0.23273106467184615,
+      "learning_rate": 1.222172751558326e-05,
+      "loss": 0.3457,
+      "step": 1947
+    },
+    {
+      "epoch": 2.3413461538461537,
+      "grad_norm": 0.2315236470722281,
+      "learning_rate": 1.219946571682992e-05,
+      "loss": 0.3174,
+      "step": 1948
+    },
+    {
+      "epoch": 2.342548076923077,
+      "grad_norm": 0.23007715312576152,
+      "learning_rate": 1.2177203918076581e-05,
+      "loss": 0.3154,
+      "step": 1949
+    },
+    {
+      "epoch": 2.34375,
+      "grad_norm": 0.26124598278687317,
+      "learning_rate": 1.2154942119323242e-05,
+      "loss": 0.3234,
+      "step": 1950
+    },
+    {
+      "epoch": 2.344951923076923,
+      "grad_norm": 0.2381868030978186,
+      "learning_rate": 1.2132680320569904e-05,
+      "loss": 0.3142,
+      "step": 1951
+    },
+    {
+      "epoch": 2.3461538461538463,
+      "grad_norm": 0.24189323694759005,
+      "learning_rate": 1.2110418521816562e-05,
+      "loss": 0.3218,
+      "step": 1952
+    },
+    {
+      "epoch": 2.347355769230769,
+      "grad_norm": 0.22443788124034567,
+      "learning_rate": 1.2088156723063223e-05,
+      "loss": 0.3367,
+      "step": 1953
+    },
+    {
+      "epoch": 2.3485576923076925,
+      "grad_norm": 0.23860729010357978,
+      "learning_rate": 1.2065894924309885e-05,
+      "loss": 0.3233,
+      "step": 1954
+    },
+    {
+      "epoch": 2.3497596153846154,
+      "grad_norm": 0.24391533078400182,
+      "learning_rate": 1.2043633125556546e-05,
+      "loss": 0.3181,
+      "step": 1955
+    },
+    {
+      "epoch": 2.3509615384615383,
+      "grad_norm": 0.24801191483594848,
+      "learning_rate": 1.2021371326803206e-05,
+      "loss": 0.3378,
+      "step": 1956
+    },
+    {
+      "epoch": 2.3521634615384617,
+      "grad_norm": 0.22520143368354759,
+      "learning_rate": 1.1999109528049866e-05,
+      "loss": 0.3078,
+      "step": 1957
+    },
+    {
+      "epoch": 2.3533653846153846,
+      "grad_norm": 0.25520961309931567,
+      "learning_rate": 1.1976847729296529e-05,
+      "loss": 0.3313,
+      "step": 1958
+    },
+    {
+      "epoch": 2.3545673076923075,
+      "grad_norm": 0.23544219083668033,
+      "learning_rate": 1.1954585930543189e-05,
+      "loss": 0.3306,
+      "step": 1959
+    },
+    {
+      "epoch": 2.355769230769231,
+      "grad_norm": 0.23386836880735848,
+      "learning_rate": 1.193232413178985e-05,
+      "loss": 0.3161,
+      "step": 1960
+    },
+    {
+      "epoch": 2.3569711538461537,
+      "grad_norm": 0.24582308898269128,
+      "learning_rate": 1.191006233303651e-05,
+      "loss": 0.3372,
+      "step": 1961
+    },
+    {
+      "epoch": 2.358173076923077,
+      "grad_norm": 0.23803541840973452,
+      "learning_rate": 1.188780053428317e-05,
+      "loss": 0.33,
+      "step": 1962
+    },
+    {
+      "epoch": 2.359375,
+      "grad_norm": 0.23483668824605317,
+      "learning_rate": 1.186553873552983e-05,
+      "loss": 0.3559,
+      "step": 1963
+    },
+    {
+      "epoch": 2.360576923076923,
+      "grad_norm": 0.23633284820043474,
+      "learning_rate": 1.1843276936776491e-05,
+      "loss": 0.3199,
+      "step": 1964
+    },
+    {
+      "epoch": 2.3617788461538463,
+      "grad_norm": 0.24901457544575095,
+      "learning_rate": 1.1821015138023153e-05,
+      "loss": 0.3167,
+      "step": 1965
+    },
+    {
+      "epoch": 2.362980769230769,
+      "grad_norm": 0.24355992935922013,
+      "learning_rate": 1.1798753339269814e-05,
+      "loss": 0.3298,
+      "step": 1966
+    },
+    {
+      "epoch": 2.3641826923076925,
+      "grad_norm": 0.22242882465807168,
+      "learning_rate": 1.1776491540516474e-05,
+      "loss": 0.3202,
+      "step": 1967
+    },
+    {
+      "epoch": 2.3653846153846154,
+      "grad_norm": 0.23441749601763376,
+      "learning_rate": 1.1754229741763136e-05,
+      "loss": 0.3441,
+      "step": 1968
+    },
+    {
+      "epoch": 2.3665865384615383,
+      "grad_norm": 0.2288441873521735,
+      "learning_rate": 1.1731967943009797e-05,
+      "loss": 0.3252,
+      "step": 1969
+    },
+    {
+      "epoch": 2.3677884615384617,
+      "grad_norm": 0.22751072735246297,
+      "learning_rate": 1.1709706144256455e-05,
+      "loss": 0.3386,
+      "step": 1970
+    },
+    {
+      "epoch": 2.3689903846153846,
+      "grad_norm": 0.2230785829665385,
+      "learning_rate": 1.1687444345503116e-05,
+      "loss": 0.3103,
+      "step": 1971
+    },
+    {
+      "epoch": 2.3701923076923075,
+      "grad_norm": 0.23082655716934825,
+      "learning_rate": 1.1665182546749778e-05,
+      "loss": 0.3182,
+      "step": 1972
+    },
+    {
+      "epoch": 2.371394230769231,
+      "grad_norm": 0.23555175052961774,
+      "learning_rate": 1.1642920747996438e-05,
+      "loss": 0.3185,
+      "step": 1973
+    },
+    {
+      "epoch": 2.3725961538461537,
+      "grad_norm": 0.23268206803295557,
+      "learning_rate": 1.1620658949243099e-05,
+      "loss": 0.3179,
+      "step": 1974
+    },
+    {
+      "epoch": 2.373798076923077,
+      "grad_norm": 0.22983202731952376,
+      "learning_rate": 1.1598397150489761e-05,
+      "loss": 0.3281,
+      "step": 1975
+    },
+    {
+      "epoch": 2.375,
+      "grad_norm": 0.24415899445140818,
+      "learning_rate": 1.1576135351736421e-05,
+      "loss": 0.3376,
+      "step": 1976
+    },
+    {
+      "epoch": 2.376201923076923,
+      "grad_norm": 0.23610301885991092,
+      "learning_rate": 1.1553873552983082e-05,
+      "loss": 0.3252,
+      "step": 1977
+    },
+    {
+      "epoch": 2.3774038461538463,
+      "grad_norm": 0.23240943634461025,
+      "learning_rate": 1.1531611754229742e-05,
+      "loss": 0.323,
+      "step": 1978
+    },
+    {
+      "epoch": 2.378605769230769,
+      "grad_norm": 0.21207736617051845,
+      "learning_rate": 1.1509349955476403e-05,
+      "loss": 0.3105,
+      "step": 1979
+    },
+    {
+      "epoch": 2.3798076923076925,
+      "grad_norm": 0.24416701343714584,
+      "learning_rate": 1.1487088156723063e-05,
+      "loss": 0.3246,
+      "step": 1980
+    },
+    {
+      "epoch": 2.3810096153846154,
+      "grad_norm": 0.23331594590728538,
+      "learning_rate": 1.1464826357969724e-05,
+      "loss": 0.3351,
+      "step": 1981
+    },
+    {
+      "epoch": 2.3822115384615383,
+      "grad_norm": 0.2357124829608915,
+      "learning_rate": 1.1442564559216386e-05,
+      "loss": 0.3424,
+      "step": 1982
+    },
+    {
+      "epoch": 2.3834134615384617,
+      "grad_norm": 0.2191158747889211,
+      "learning_rate": 1.1420302760463046e-05,
+      "loss": 0.3127,
+      "step": 1983
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.2537741671836198,
+      "learning_rate": 1.1398040961709707e-05,
+      "loss": 0.32,
+      "step": 1984
+    },
+    {
+      "epoch": 2.3858173076923075,
+      "grad_norm": 0.22202793233450652,
+      "learning_rate": 1.1375779162956367e-05,
+      "loss": 0.308,
+      "step": 1985
+    },
+    {
+      "epoch": 2.387019230769231,
+      "grad_norm": 0.2109280200099059,
+      "learning_rate": 1.135351736420303e-05,
+      "loss": 0.3066,
+      "step": 1986
+    },
+    {
+      "epoch": 2.3882211538461537,
+      "grad_norm": 0.25578064257352207,
+      "learning_rate": 1.1331255565449688e-05,
+      "loss": 0.3391,
+      "step": 1987
+    },
+    {
+      "epoch": 2.389423076923077,
+      "grad_norm": 0.23859776379825964,
+      "learning_rate": 1.1308993766696348e-05,
+      "loss": 0.3142,
+      "step": 1988
+    },
+    {
+      "epoch": 2.390625,
+      "grad_norm": 0.24897298168156876,
+      "learning_rate": 1.128673196794301e-05,
+      "loss": 0.3363,
+      "step": 1989
+    },
+    {
+      "epoch": 2.391826923076923,
+      "grad_norm": 0.24040152193087405,
+      "learning_rate": 1.1264470169189671e-05,
+      "loss": 0.3141,
+      "step": 1990
+    },
+    {
+      "epoch": 2.3930288461538463,
+      "grad_norm": 0.22376892592739003,
+      "learning_rate": 1.1242208370436331e-05,
+      "loss": 0.3181,
+      "step": 1991
+    },
+    {
+      "epoch": 2.394230769230769,
+      "grad_norm": 0.23387832311148946,
+      "learning_rate": 1.1219946571682992e-05,
+      "loss": 0.3198,
+      "step": 1992
+    },
+    {
+      "epoch": 2.3954326923076925,
+      "grad_norm": 0.2173126435277069,
+      "learning_rate": 1.1197684772929654e-05,
+      "loss": 0.3308,
+      "step": 1993
+    },
+    {
+      "epoch": 2.3966346153846154,
+      "grad_norm": 0.2331025271541905,
+      "learning_rate": 1.1175422974176314e-05,
+      "loss": 0.3274,
+      "step": 1994
+    },
+    {
+      "epoch": 2.3978365384615383,
+      "grad_norm": 0.21978635826981122,
+      "learning_rate": 1.1153161175422975e-05,
+      "loss": 0.3172,
+      "step": 1995
+    },
+    {
+      "epoch": 2.3990384615384617,
+      "grad_norm": 0.21859484489439274,
+      "learning_rate": 1.1130899376669635e-05,
+      "loss": 0.3473,
+      "step": 1996
+    },
+    {
+      "epoch": 2.4002403846153846,
+      "grad_norm": 0.22045696261433406,
+      "learning_rate": 1.1108637577916296e-05,
+      "loss": 0.3262,
+      "step": 1997
+    },
+    {
+      "epoch": 2.4014423076923075,
+      "grad_norm": 0.22267005840640147,
+      "learning_rate": 1.1086375779162956e-05,
+      "loss": 0.3312,
+      "step": 1998
+    },
+    {
+      "epoch": 2.402644230769231,
+      "grad_norm": 0.21785580724087686,
+      "learning_rate": 1.1064113980409617e-05,
+      "loss": 0.3172,
+      "step": 1999
+    },
+    {
+      "epoch": 2.4038461538461537,
+      "grad_norm": 0.219211171780588,
+      "learning_rate": 1.1041852181656279e-05,
+      "loss": 0.3309,
+      "step": 2000
+    },
+    {
+      "epoch": 2.405048076923077,
+      "grad_norm": 0.25084143458093733,
+      "learning_rate": 1.101959038290294e-05,
+      "loss": 0.3186,
+      "step": 2001
+    },
+    {
+      "epoch": 2.40625,
+      "grad_norm": 0.2242520764841523,
+      "learning_rate": 1.09973285841496e-05,
+      "loss": 0.32,
+      "step": 2002
+    },
+    {
+      "epoch": 2.407451923076923,
+      "grad_norm": 0.22223721363046967,
+      "learning_rate": 1.0975066785396262e-05,
+      "loss": 0.3288,
+      "step": 2003
+    },
+    {
+      "epoch": 2.4086538461538463,
+      "grad_norm": 0.2456094545643119,
+      "learning_rate": 1.0952804986642922e-05,
+      "loss": 0.3398,
+      "step": 2004
+    },
+    {
+      "epoch": 2.409855769230769,
+      "grad_norm": 0.23504949797447683,
+      "learning_rate": 1.0930543187889581e-05,
+      "loss": 0.3249,
+      "step": 2005
+    },
+    {
+      "epoch": 2.4110576923076925,
+      "grad_norm": 0.47797349291586255,
+      "learning_rate": 1.0908281389136243e-05,
+      "loss": 0.3171,
+      "step": 2006
+    },
+    {
+      "epoch": 2.4122596153846154,
+      "grad_norm": 0.22276698128121,
+      "learning_rate": 1.0886019590382904e-05,
+      "loss": 0.3149,
+      "step": 2007
+    },
+    {
+      "epoch": 2.4134615384615383,
+      "grad_norm": 0.2604704621172367,
+      "learning_rate": 1.0863757791629564e-05,
+      "loss": 0.3239,
+      "step": 2008
+    },
+    {
+      "epoch": 2.4146634615384617,
+      "grad_norm": 0.21521752938068084,
+      "learning_rate": 1.0841495992876224e-05,
+      "loss": 0.3332,
+      "step": 2009
+    },
+    {
+      "epoch": 2.4158653846153846,
+      "grad_norm": 0.2305416522285085,
+      "learning_rate": 1.0819234194122887e-05,
+      "loss": 0.3384,
+      "step": 2010
+    },
+    {
+      "epoch": 2.4170673076923075,
+      "grad_norm": 0.21907701739374413,
+      "learning_rate": 1.0796972395369547e-05,
+      "loss": 0.3193,
+      "step": 2011
+    },
+    {
+      "epoch": 2.418269230769231,
+      "grad_norm": 0.23018286726890202,
+      "learning_rate": 1.0774710596616207e-05,
+      "loss": 0.3279,
+      "step": 2012
+    },
+    {
+      "epoch": 2.4194711538461537,
+      "grad_norm": 0.23513307680201295,
+      "learning_rate": 1.0752448797862868e-05,
+      "loss": 0.3125,
+      "step": 2013
+    },
+    {
+      "epoch": 2.420673076923077,
+      "grad_norm": 0.22481896156370007,
+      "learning_rate": 1.0730186999109528e-05,
+      "loss": 0.3434,
+      "step": 2014
+    },
+    {
+      "epoch": 2.421875,
+      "grad_norm": 0.22431977071690004,
+      "learning_rate": 1.0707925200356189e-05,
+      "loss": 0.3278,
+      "step": 2015
+    },
+    {
+      "epoch": 2.423076923076923,
+      "grad_norm": 0.22312862025391605,
+      "learning_rate": 1.068566340160285e-05,
+      "loss": 0.3414,
+      "step": 2016
+    },
+    {
+      "epoch": 2.4242788461538463,
+      "grad_norm": 0.23236197448197415,
+      "learning_rate": 1.0663401602849511e-05,
+      "loss": 0.3465,
+      "step": 2017
+    },
+    {
+      "epoch": 2.425480769230769,
+      "grad_norm": 0.227587573276689,
+      "learning_rate": 1.0641139804096172e-05,
+      "loss": 0.3268,
+      "step": 2018
+    },
+    {
+      "epoch": 2.4266826923076925,
+      "grad_norm": 0.22484904225781682,
+      "learning_rate": 1.0618878005342832e-05,
+      "loss": 0.3288,
+      "step": 2019
+    },
+    {
+      "epoch": 2.4278846153846154,
+      "grad_norm": 0.2101336824144019,
+      "learning_rate": 1.0596616206589494e-05,
+      "loss": 0.3027,
+      "step": 2020
+    },
+    {
+      "epoch": 2.4290865384615383,
+      "grad_norm": 0.2174989835430156,
+      "learning_rate": 1.0574354407836155e-05,
+      "loss": 0.3185,
+      "step": 2021
+    },
+    {
+      "epoch": 2.4302884615384617,
+      "grad_norm": 0.21695148862569852,
+      "learning_rate": 1.0552092609082813e-05,
+      "loss": 0.308,
+      "step": 2022
+    },
+    {
+      "epoch": 2.4314903846153846,
+      "grad_norm": 0.21926569928561274,
+      "learning_rate": 1.0529830810329474e-05,
+      "loss": 0.3333,
+      "step": 2023
+    },
+    {
+      "epoch": 2.4326923076923075,
+      "grad_norm": 0.21640246663104565,
+      "learning_rate": 1.0507569011576136e-05,
+      "loss": 0.3136,
+      "step": 2024
+    },
+    {
+      "epoch": 2.433894230769231,
+      "grad_norm": 0.22055537577469958,
+      "learning_rate": 1.0485307212822796e-05,
+      "loss": 0.3448,
+      "step": 2025
+    },
+    {
+      "epoch": 2.4350961538461537,
+      "grad_norm": 0.21365433686873322,
+      "learning_rate": 1.0463045414069457e-05,
+      "loss": 0.3163,
+      "step": 2026
+    },
+    {
+      "epoch": 2.436298076923077,
+      "grad_norm": 0.22375261579091954,
+      "learning_rate": 1.0440783615316119e-05,
+      "loss": 0.3289,
+      "step": 2027
+    },
+    {
+      "epoch": 2.4375,
+      "grad_norm": 0.23001398748850183,
+      "learning_rate": 1.041852181656278e-05,
+      "loss": 0.321,
+      "step": 2028
+    },
+    {
+      "epoch": 2.438701923076923,
+      "grad_norm": 0.21851596735019987,
+      "learning_rate": 1.039626001780944e-05,
+      "loss": 0.3151,
+      "step": 2029
+    },
+    {
+      "epoch": 2.4399038461538463,
+      "grad_norm": 0.22542525169919417,
+      "learning_rate": 1.03739982190561e-05,
+      "loss": 0.3035,
+      "step": 2030
+    },
+    {
+      "epoch": 2.441105769230769,
+      "grad_norm": 0.22364872281254827,
+      "learning_rate": 1.035173642030276e-05,
+      "loss": 0.3279,
+      "step": 2031
+    },
+    {
+      "epoch": 2.4423076923076925,
+      "grad_norm": 0.21569756511949473,
+      "learning_rate": 1.0329474621549421e-05,
+      "loss": 0.3242,
+      "step": 2032
+    },
+    {
+      "epoch": 2.4435096153846154,
+      "grad_norm": 0.267794142647368,
+      "learning_rate": 1.0307212822796082e-05,
+      "loss": 0.326,
+      "step": 2033
+    },
+    {
+      "epoch": 2.4447115384615383,
+      "grad_norm": 0.22845533184519104,
+      "learning_rate": 1.0284951024042744e-05,
+      "loss": 0.3365,
+      "step": 2034
+    },
+    {
+      "epoch": 2.4459134615384617,
+      "grad_norm": 0.22739776815935442,
+      "learning_rate": 1.0262689225289404e-05,
+      "loss": 0.3221,
+      "step": 2035
+    },
+    {
+      "epoch": 2.4471153846153846,
+      "grad_norm": 0.24731626898190742,
+      "learning_rate": 1.0240427426536065e-05,
+      "loss": 0.3398,
+      "step": 2036
+    },
+    {
+      "epoch": 2.4483173076923075,
+      "grad_norm": 0.21763063774208513,
+      "learning_rate": 1.0218165627782725e-05,
+      "loss": 0.3386,
+      "step": 2037
+    },
+    {
+      "epoch": 2.449519230769231,
+      "grad_norm": 0.2708249175721318,
+      "learning_rate": 1.0195903829029387e-05,
+      "loss": 0.3468,
+      "step": 2038
+    },
+    {
+      "epoch": 2.4507211538461537,
+      "grad_norm": 0.24760872751484775,
+      "learning_rate": 1.0173642030276046e-05,
+      "loss": 0.3398,
+      "step": 2039
+    },
+    {
+      "epoch": 2.451923076923077,
+      "grad_norm": 0.24739224397066623,
+      "learning_rate": 1.0151380231522706e-05,
+      "loss": 0.3193,
+      "step": 2040
+    },
+    {
+      "epoch": 2.453125,
+      "grad_norm": 0.21837016955260197,
+      "learning_rate": 1.0129118432769369e-05,
+      "loss": 0.3303,
+      "step": 2041
+    },
+    {
+      "epoch": 2.454326923076923,
+      "grad_norm": 0.22944744245610524,
+      "learning_rate": 1.0106856634016029e-05,
+      "loss": 0.3225,
+      "step": 2042
+    },
+    {
+      "epoch": 2.4555288461538463,
+      "grad_norm": 0.23688410939376267,
+      "learning_rate": 1.008459483526269e-05,
+      "loss": 0.3144,
+      "step": 2043
+    },
+    {
+      "epoch": 2.456730769230769,
+      "grad_norm": 0.21402586033585264,
+      "learning_rate": 1.006233303650935e-05,
+      "loss": 0.3159,
+      "step": 2044
+    },
+    {
+      "epoch": 2.4579326923076925,
+      "grad_norm": 0.2338576767031622,
+      "learning_rate": 1.0040071237756012e-05,
+      "loss": 0.3336,
+      "step": 2045
+    },
+    {
+      "epoch": 2.4591346153846154,
+      "grad_norm": 0.22323731248334192,
+      "learning_rate": 1.0017809439002672e-05,
+      "loss": 0.3233,
+      "step": 2046
+    },
+    {
+      "epoch": 2.4603365384615383,
+      "grad_norm": 0.22596219071225418,
+      "learning_rate": 9.995547640249333e-06,
+      "loss": 0.34,
+      "step": 2047
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.21902256636135797,
+      "learning_rate": 9.973285841495993e-06,
+      "loss": 0.3248,
+      "step": 2048
+    },
+    {
+      "epoch": 2.4627403846153846,
+      "grad_norm": 0.23354606294775337,
+      "learning_rate": 9.951024042742654e-06,
+      "loss": 0.3254,
+      "step": 2049
+    },
+    {
+      "epoch": 2.4639423076923075,
+      "grad_norm": 0.2499823842629235,
+      "learning_rate": 9.928762243989314e-06,
+      "loss": 0.3432,
+      "step": 2050
+    },
+    {
+      "epoch": 2.465144230769231,
+      "grad_norm": 0.22017458890365238,
+      "learning_rate": 9.906500445235975e-06,
+      "loss": 0.3232,
+      "step": 2051
+    },
+    {
+      "epoch": 2.4663461538461537,
+      "grad_norm": 0.2409217361449599,
+      "learning_rate": 9.884238646482637e-06,
+      "loss": 0.3219,
+      "step": 2052
+    },
+    {
+      "epoch": 2.467548076923077,
+      "grad_norm": 0.22194983808349106,
+      "learning_rate": 9.861976847729297e-06,
+      "loss": 0.2959,
+      "step": 2053
+    },
+    {
+      "epoch": 2.46875,
+      "grad_norm": 0.24487519220103046,
+      "learning_rate": 9.839715048975958e-06,
+      "loss": 0.3325,
+      "step": 2054
+    },
+    {
+      "epoch": 2.469951923076923,
+      "grad_norm": 0.23902102198005953,
+      "learning_rate": 9.81745325022262e-06,
+      "loss": 0.3077,
+      "step": 2055
+    },
+    {
+      "epoch": 2.4711538461538463,
+      "grad_norm": 0.23175485734087906,
+      "learning_rate": 9.79519145146928e-06,
+      "loss": 0.3232,
+      "step": 2056
+    },
+    {
+      "epoch": 2.472355769230769,
+      "grad_norm": 0.2246390177251172,
+      "learning_rate": 9.772929652715939e-06,
+      "loss": 0.343,
+      "step": 2057
+    },
+    {
+      "epoch": 2.4735576923076925,
+      "grad_norm": 0.27944432710807227,
+      "learning_rate": 9.7506678539626e-06,
+      "loss": 0.3226,
+      "step": 2058
+    },
+    {
+      "epoch": 2.4747596153846154,
+      "grad_norm": 0.2546513674617655,
+      "learning_rate": 9.728406055209262e-06,
+      "loss": 0.3238,
+      "step": 2059
+    },
+    {
+      "epoch": 2.4759615384615383,
+      "grad_norm": 0.21855030201703077,
+      "learning_rate": 9.706144256455922e-06,
+      "loss": 0.3159,
+      "step": 2060
+    },
+    {
+      "epoch": 2.4771634615384617,
+      "grad_norm": 0.21972514952096994,
+      "learning_rate": 9.683882457702582e-06,
+      "loss": 0.3006,
+      "step": 2061
+    },
+    {
+      "epoch": 2.4783653846153846,
+      "grad_norm": 0.2660165054703141,
+      "learning_rate": 9.661620658949245e-06,
+      "loss": 0.3219,
+      "step": 2062
+    },
+    {
+      "epoch": 2.4795673076923075,
+      "grad_norm": 0.24125171717589772,
+      "learning_rate": 9.639358860195905e-06,
+      "loss": 0.3234,
+      "step": 2063
+    },
+    {
+      "epoch": 2.480769230769231,
+      "grad_norm": 0.23340596890122398,
+      "learning_rate": 9.617097061442565e-06,
+      "loss": 0.3194,
+      "step": 2064
+    },
+    {
+      "epoch": 2.4819711538461537,
+      "grad_norm": 0.23225258464102283,
+      "learning_rate": 9.594835262689226e-06,
+      "loss": 0.3146,
+      "step": 2065
+    },
+    {
+      "epoch": 2.483173076923077,
+      "grad_norm": 0.2504758252758769,
+      "learning_rate": 9.572573463935886e-06,
+      "loss": 0.3292,
+      "step": 2066
+    },
+    {
+      "epoch": 2.484375,
+      "grad_norm": 0.22907607364622068,
+      "learning_rate": 9.550311665182547e-06,
+      "loss": 0.332,
+      "step": 2067
+    },
+    {
+      "epoch": 2.485576923076923,
+      "grad_norm": 0.2374450360504777,
+      "learning_rate": 9.528049866429207e-06,
+      "loss": 0.3144,
+      "step": 2068
+    },
+    {
+      "epoch": 2.4867788461538463,
+      "grad_norm": 0.23163217079600296,
+      "learning_rate": 9.50578806767587e-06,
+      "loss": 0.342,
+      "step": 2069
+    },
+    {
+      "epoch": 2.487980769230769,
+      "grad_norm": 0.24694336521162716,
+      "learning_rate": 9.48352626892253e-06,
+      "loss": 0.3249,
+      "step": 2070
+    },
+    {
+      "epoch": 2.4891826923076925,
+      "grad_norm": 0.2414845773586669,
+      "learning_rate": 9.46126447016919e-06,
+      "loss": 0.339,
+      "step": 2071
+    },
+    {
+      "epoch": 2.4903846153846154,
+      "grad_norm": 0.2360462652747458,
+      "learning_rate": 9.43900267141585e-06,
+      "loss": 0.326,
+      "step": 2072
+    },
+    {
+      "epoch": 2.4915865384615383,
+      "grad_norm": 0.2570814852253227,
+      "learning_rate": 9.416740872662513e-06,
+      "loss": 0.3191,
+      "step": 2073
+    },
+    {
+      "epoch": 2.4927884615384617,
+      "grad_norm": 0.2366402805471971,
+      "learning_rate": 9.394479073909172e-06,
+      "loss": 0.3203,
+      "step": 2074
+    },
+    {
+      "epoch": 2.4939903846153846,
+      "grad_norm": 0.2406316743850866,
+      "learning_rate": 9.372217275155832e-06,
+      "loss": 0.3101,
+      "step": 2075
+    },
+    {
+      "epoch": 2.4951923076923075,
+      "grad_norm": 0.25571858432787264,
+      "learning_rate": 9.349955476402494e-06,
+      "loss": 0.3153,
+      "step": 2076
+    },
+    {
+      "epoch": 2.496394230769231,
+      "grad_norm": 0.231426276063418,
+      "learning_rate": 9.327693677649155e-06,
+      "loss": 0.3168,
+      "step": 2077
+    },
+    {
+      "epoch": 2.4975961538461537,
+      "grad_norm": 0.23003151397175906,
+      "learning_rate": 9.305431878895815e-06,
+      "loss": 0.3019,
+      "step": 2078
+    },
+    {
+      "epoch": 2.498798076923077,
+      "grad_norm": 0.23604372952979305,
+      "learning_rate": 9.283170080142475e-06,
+      "loss": 0.3182,
+      "step": 2079
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.24517677094421592,
+      "learning_rate": 9.260908281389138e-06,
+      "loss": 0.3346,
+      "step": 2080
+    },
+    {
+      "epoch": 2.5012019230769234,
+      "grad_norm": 0.23632847166693005,
+      "learning_rate": 9.238646482635798e-06,
+      "loss": 0.3396,
+      "step": 2081
+    },
+    {
+      "epoch": 2.5024038461538463,
+      "grad_norm": 0.22594825767783666,
+      "learning_rate": 9.216384683882458e-06,
+      "loss": 0.3071,
+      "step": 2082
+    },
+    {
+      "epoch": 2.503605769230769,
+      "grad_norm": 0.20833243943307456,
+      "learning_rate": 9.194122885129119e-06,
+      "loss": 0.3215,
+      "step": 2083
+    },
+    {
+      "epoch": 2.5048076923076925,
+      "grad_norm": 0.2653003708804367,
+      "learning_rate": 9.17186108637578e-06,
+      "loss": 0.3198,
+      "step": 2084
+    },
+    {
+      "epoch": 2.5060096153846154,
+      "grad_norm": 0.23671087139099173,
+      "learning_rate": 9.14959928762244e-06,
+      "loss": 0.3186,
+      "step": 2085
+    },
+    {
+      "epoch": 2.5072115384615383,
+      "grad_norm": 0.22852510296070594,
+      "learning_rate": 9.1273374888691e-06,
+      "loss": 0.3538,
+      "step": 2086
+    },
+    {
+      "epoch": 2.5084134615384617,
+      "grad_norm": 0.22623274479156757,
+      "learning_rate": 9.105075690115762e-06,
+      "loss": 0.3291,
+      "step": 2087
+    },
+    {
+      "epoch": 2.5096153846153846,
+      "grad_norm": 0.2594064708800619,
+      "learning_rate": 9.082813891362423e-06,
+      "loss": 0.3314,
+      "step": 2088
+    },
+    {
+      "epoch": 2.5108173076923075,
+      "grad_norm": 0.23562388793546782,
+      "learning_rate": 9.060552092609083e-06,
+      "loss": 0.3409,
+      "step": 2089
+    },
+    {
+      "epoch": 2.512019230769231,
+      "grad_norm": 0.21558505013006535,
+      "learning_rate": 9.038290293855745e-06,
+      "loss": 0.3271,
+      "step": 2090
+    },
+    {
+      "epoch": 2.5132211538461537,
+      "grad_norm": 0.23200863709052646,
+      "learning_rate": 9.016028495102404e-06,
+      "loss": 0.3283,
+      "step": 2091
+    },
+    {
+      "epoch": 2.5144230769230766,
+      "grad_norm": 0.2241166994972684,
+      "learning_rate": 8.993766696349064e-06,
+      "loss": 0.3312,
+      "step": 2092
+    },
+    {
+      "epoch": 2.515625,
+      "grad_norm": 0.2334521252773408,
+      "learning_rate": 8.971504897595725e-06,
+      "loss": 0.3404,
+      "step": 2093
+    },
+    {
+      "epoch": 2.5168269230769234,
+      "grad_norm": 0.2379059013186444,
+      "learning_rate": 8.949243098842387e-06,
+      "loss": 0.3253,
+      "step": 2094
+    },
+    {
+      "epoch": 2.5180288461538463,
+      "grad_norm": 0.23832998774635222,
+      "learning_rate": 8.926981300089047e-06,
+      "loss": 0.3421,
+      "step": 2095
+    },
+    {
+      "epoch": 2.519230769230769,
+      "grad_norm": 0.20200967392533375,
+      "learning_rate": 8.904719501335708e-06,
+      "loss": 0.3089,
+      "step": 2096
+    },
+    {
+      "epoch": 2.5204326923076925,
+      "grad_norm": 0.28903875813494584,
+      "learning_rate": 8.88245770258237e-06,
+      "loss": 0.3286,
+      "step": 2097
+    },
+    {
+      "epoch": 2.5216346153846154,
+      "grad_norm": 0.2408756404322587,
+      "learning_rate": 8.86019590382903e-06,
+      "loss": 0.3175,
+      "step": 2098
+    },
+    {
+      "epoch": 2.5228365384615383,
+      "grad_norm": 0.23847103196270084,
+      "learning_rate": 8.837934105075691e-06,
+      "loss": 0.3332,
+      "step": 2099
+    },
+    {
+      "epoch": 2.5240384615384617,
+      "grad_norm": 0.22039619819549833,
+      "learning_rate": 8.815672306322351e-06,
+      "loss": 0.3446,
+      "step": 2100
+    },
+    {
+      "epoch": 2.5252403846153846,
+      "grad_norm": 0.22588983783465438,
+      "learning_rate": 8.793410507569012e-06,
+      "loss": 0.3382,
+      "step": 2101
+    },
+    {
+      "epoch": 2.5264423076923075,
+      "grad_norm": 0.2648743231174083,
+      "learning_rate": 8.771148708815672e-06,
+      "loss": 0.3561,
+      "step": 2102
+    },
+    {
+      "epoch": 2.527644230769231,
+      "grad_norm": 0.2553718640824547,
+      "learning_rate": 8.748886910062333e-06,
+      "loss": 0.3306,
+      "step": 2103
+    },
+    {
+      "epoch": 2.5288461538461537,
+      "grad_norm": 0.2079334433461607,
+      "learning_rate": 8.726625111308995e-06,
+      "loss": 0.3326,
+      "step": 2104
+    },
+    {
+      "epoch": 2.5300480769230766,
+      "grad_norm": 0.25851598637089135,
+      "learning_rate": 8.704363312555655e-06,
+      "loss": 0.3294,
+      "step": 2105
+    },
+    {
+      "epoch": 2.53125,
+      "grad_norm": 0.22682275911085864,
+      "learning_rate": 8.682101513802316e-06,
+      "loss": 0.2961,
+      "step": 2106
+    },
+    {
+      "epoch": 2.5324519230769234,
+      "grad_norm": 0.2296591586693542,
+      "learning_rate": 8.659839715048976e-06,
+      "loss": 0.3153,
+      "step": 2107
+    },
+    {
+      "epoch": 2.5336538461538463,
+      "grad_norm": 0.23496454180139584,
+      "learning_rate": 8.637577916295638e-06,
+      "loss": 0.3342,
+      "step": 2108
+    },
+    {
+      "epoch": 2.534855769230769,
+      "grad_norm": 0.2475381804343742,
+      "learning_rate": 8.615316117542297e-06,
+      "loss": 0.3322,
+      "step": 2109
+    },
+    {
+      "epoch": 2.5360576923076925,
+      "grad_norm": 0.21544970027471966,
+      "learning_rate": 8.593054318788957e-06,
+      "loss": 0.3345,
+      "step": 2110
+    },
+    {
+      "epoch": 2.5372596153846154,
+      "grad_norm": 0.2309143465980592,
+      "learning_rate": 8.57079252003562e-06,
+      "loss": 0.3193,
+      "step": 2111
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.21317201520850318,
+      "learning_rate": 8.54853072128228e-06,
+      "loss": 0.3049,
+      "step": 2112
+    },
+    {
+      "epoch": 2.5396634615384617,
+      "grad_norm": 0.22056028050754845,
+      "learning_rate": 8.52626892252894e-06,
+      "loss": 0.3222,
+      "step": 2113
+    },
+    {
+      "epoch": 2.5408653846153846,
+      "grad_norm": 0.23050342151373834,
+      "learning_rate": 8.504007123775601e-06,
+      "loss": 0.3271,
+      "step": 2114
+    },
+    {
+      "epoch": 2.5420673076923075,
+      "grad_norm": 0.221853365750641,
+      "learning_rate": 8.481745325022263e-06,
+      "loss": 0.3295,
+      "step": 2115
+    },
+    {
+      "epoch": 2.543269230769231,
+      "grad_norm": 0.2212073837768712,
+      "learning_rate": 8.459483526268923e-06,
+      "loss": 0.3178,
+      "step": 2116
+    },
+    {
+      "epoch": 2.5444711538461537,
+      "grad_norm": 0.2440609529086934,
+      "learning_rate": 8.437221727515584e-06,
+      "loss": 0.3183,
+      "step": 2117
+    },
+    {
+      "epoch": 2.5456730769230766,
+      "grad_norm": 0.21633137329805074,
+      "learning_rate": 8.414959928762244e-06,
+      "loss": 0.3392,
+      "step": 2118
+    },
+    {
+      "epoch": 2.546875,
+      "grad_norm": 0.2116022230907213,
+      "learning_rate": 8.392698130008905e-06,
+      "loss": 0.3278,
+      "step": 2119
+    },
+    {
+      "epoch": 2.5480769230769234,
+      "grad_norm": 0.2252731366861999,
+      "learning_rate": 8.370436331255565e-06,
+      "loss": 0.3289,
+      "step": 2120
+    },
+    {
+      "epoch": 2.5492788461538463,
+      "grad_norm": 0.2206376084371446,
+      "learning_rate": 8.348174532502226e-06,
+      "loss": 0.3202,
+      "step": 2121
+    },
+    {
+      "epoch": 2.550480769230769,
+      "grad_norm": 0.2495415530253815,
+      "learning_rate": 8.325912733748888e-06,
+      "loss": 0.3255,
+      "step": 2122
+    },
+    {
+      "epoch": 2.5516826923076925,
+      "grad_norm": 0.24932420525466142,
+      "learning_rate": 8.303650934995548e-06,
+      "loss": 0.3284,
+      "step": 2123
+    },
+    {
+      "epoch": 2.5528846153846154,
+      "grad_norm": 0.2121236533216481,
+      "learning_rate": 8.281389136242209e-06,
+      "loss": 0.3129,
+      "step": 2124
+    },
+    {
+      "epoch": 2.5540865384615383,
+      "grad_norm": 0.2091178162215918,
+      "learning_rate": 8.25912733748887e-06,
+      "loss": 0.3236,
+      "step": 2125
+    },
+    {
+      "epoch": 2.5552884615384617,
+      "grad_norm": 0.24384072957122746,
+      "learning_rate": 8.23686553873553e-06,
+      "loss": 0.3234,
+      "step": 2126
+    },
+    {
+      "epoch": 2.5564903846153846,
+      "grad_norm": 0.24115527514249815,
+      "learning_rate": 8.21460373998219e-06,
+      "loss": 0.3484,
+      "step": 2127
+    },
+    {
+      "epoch": 2.5576923076923075,
+      "grad_norm": 0.23508850081165036,
+      "learning_rate": 8.19234194122885e-06,
+      "loss": 0.3215,
+      "step": 2128
+    },
+    {
+      "epoch": 2.558894230769231,
+      "grad_norm": 0.22251360878437565,
+      "learning_rate": 8.170080142475513e-06,
+      "loss": 0.3337,
+      "step": 2129
+    },
+    {
+      "epoch": 2.5600961538461537,
+      "grad_norm": 0.20601123039413594,
+      "learning_rate": 8.147818343722173e-06,
+      "loss": 0.3016,
+      "step": 2130
+    },
+    {
+      "epoch": 2.5612980769230766,
+      "grad_norm": 0.21347506692881166,
+      "learning_rate": 8.125556544968833e-06,
+      "loss": 0.3079,
+      "step": 2131
+    },
+    {
+      "epoch": 2.5625,
+      "grad_norm": 0.22897236719121378,
+      "learning_rate": 8.103294746215496e-06,
+      "loss": 0.3322,
+      "step": 2132
+    },
+    {
+      "epoch": 2.5637019230769234,
+      "grad_norm": 0.22099345623761696,
+      "learning_rate": 8.081032947462156e-06,
+      "loss": 0.3262,
+      "step": 2133
+    },
+    {
+      "epoch": 2.5649038461538463,
+      "grad_norm": 0.20503917354984813,
+      "learning_rate": 8.058771148708816e-06,
+      "loss": 0.3003,
+      "step": 2134
+    },
+    {
+      "epoch": 2.566105769230769,
+      "grad_norm": 0.21410414856513366,
+      "learning_rate": 8.036509349955477e-06,
+      "loss": 0.3294,
+      "step": 2135
+    },
+    {
+      "epoch": 2.5673076923076925,
+      "grad_norm": 0.23057059773309285,
+      "learning_rate": 8.014247551202137e-06,
+      "loss": 0.3266,
+      "step": 2136
+    },
+    {
+      "epoch": 2.5685096153846154,
+      "grad_norm": 0.23712936413571886,
+      "learning_rate": 7.991985752448798e-06,
+      "loss": 0.321,
+      "step": 2137
+    },
+    {
+      "epoch": 2.5697115384615383,
+      "grad_norm": 0.22502873647049823,
+      "learning_rate": 7.969723953695458e-06,
+      "loss": 0.312,
+      "step": 2138
+    },
+    {
+      "epoch": 2.5709134615384617,
+      "grad_norm": 0.24240153139416945,
+      "learning_rate": 7.94746215494212e-06,
+      "loss": 0.3196,
+      "step": 2139
+    },
+    {
+      "epoch": 2.5721153846153846,
+      "grad_norm": 0.22597326666503564,
+      "learning_rate": 7.92520035618878e-06,
+      "loss": 0.3243,
+      "step": 2140
+    },
+    {
+      "epoch": 2.5733173076923075,
+      "grad_norm": 0.22838911352081157,
+      "learning_rate": 7.902938557435441e-06,
+      "loss": 0.3336,
+      "step": 2141
+    },
+    {
+      "epoch": 2.574519230769231,
+      "grad_norm": 0.2185381709420268,
+      "learning_rate": 7.880676758682102e-06,
+      "loss": 0.336,
+      "step": 2142
+    },
+    {
+      "epoch": 2.5757211538461537,
+      "grad_norm": 0.23875200381949505,
+      "learning_rate": 7.858414959928764e-06,
+      "loss": 0.3124,
+      "step": 2143
+    },
+    {
+      "epoch": 2.5769230769230766,
+      "grad_norm": 0.2134611264316966,
+      "learning_rate": 7.836153161175422e-06,
+      "loss": 0.3276,
+      "step": 2144
+    },
+    {
+      "epoch": 2.578125,
+      "grad_norm": 0.20851781564781038,
+      "learning_rate": 7.813891362422083e-06,
+      "loss": 0.3116,
+      "step": 2145
+    },
+    {
+      "epoch": 2.5793269230769234,
+      "grad_norm": 0.22429371393946593,
+      "learning_rate": 7.791629563668745e-06,
+      "loss": 0.3504,
+      "step": 2146
+    },
+    {
+      "epoch": 2.5805288461538463,
+      "grad_norm": 0.2466395849553161,
+      "learning_rate": 7.769367764915406e-06,
+      "loss": 0.3175,
+      "step": 2147
+    },
+    {
+      "epoch": 2.581730769230769,
+      "grad_norm": 0.21868847865421237,
+      "learning_rate": 7.747105966162066e-06,
+      "loss": 0.3318,
+      "step": 2148
+    },
+    {
+      "epoch": 2.5829326923076925,
+      "grad_norm": 0.23203699002285277,
+      "learning_rate": 7.724844167408728e-06,
+      "loss": 0.3175,
+      "step": 2149
+    },
+    {
+      "epoch": 2.5841346153846154,
+      "grad_norm": 0.22622806837614803,
+      "learning_rate": 7.702582368655389e-06,
+      "loss": 0.3178,
+      "step": 2150
+    },
+    {
+      "epoch": 2.5853365384615383,
+      "grad_norm": 0.21366644656203743,
+      "learning_rate": 7.680320569902049e-06,
+      "loss": 0.3365,
+      "step": 2151
+    },
+    {
+      "epoch": 2.5865384615384617,
+      "grad_norm": 0.22056104086052486,
+      "learning_rate": 7.65805877114871e-06,
+      "loss": 0.3227,
+      "step": 2152
+    },
+    {
+      "epoch": 2.5877403846153846,
+      "grad_norm": 0.21336201502581176,
+      "learning_rate": 7.63579697239537e-06,
+      "loss": 0.3081,
+      "step": 2153
+    },
+    {
+      "epoch": 2.5889423076923075,
+      "grad_norm": 0.22509127091153638,
+      "learning_rate": 7.613535173642031e-06,
+      "loss": 0.3161,
+      "step": 2154
+    },
+    {
+      "epoch": 2.590144230769231,
+      "grad_norm": 0.21909287922631274,
+      "learning_rate": 7.591273374888691e-06,
+      "loss": 0.3201,
+      "step": 2155
+    },
+    {
+      "epoch": 2.5913461538461537,
+      "grad_norm": 0.20767569874546338,
+      "learning_rate": 7.569011576135353e-06,
+      "loss": 0.3169,
+      "step": 2156
+    },
+    {
+      "epoch": 2.5925480769230766,
+      "grad_norm": 0.2437772364098172,
+      "learning_rate": 7.546749777382013e-06,
+      "loss": 0.3362,
+      "step": 2157
+    },
+    {
+      "epoch": 2.59375,
+      "grad_norm": 0.22377769513109538,
+      "learning_rate": 7.524487978628674e-06,
+      "loss": 0.3202,
+      "step": 2158
+    },
+    {
+      "epoch": 2.5949519230769234,
+      "grad_norm": 0.22491952195085158,
+      "learning_rate": 7.502226179875334e-06,
+      "loss": 0.3229,
+      "step": 2159
+    },
+    {
+      "epoch": 2.5961538461538463,
+      "grad_norm": 0.22310783535368764,
+      "learning_rate": 7.4799643811219954e-06,
+      "loss": 0.3338,
+      "step": 2160
+    },
+    {
+      "epoch": 2.597355769230769,
+      "grad_norm": 0.23126118066924387,
+      "learning_rate": 7.457702582368656e-06,
+      "loss": 0.3158,
+      "step": 2161
+    },
+    {
+      "epoch": 2.5985576923076925,
+      "grad_norm": 0.2242437557669824,
+      "learning_rate": 7.435440783615316e-06,
+      "loss": 0.3326,
+      "step": 2162
+    },
+    {
+      "epoch": 2.5997596153846154,
+      "grad_norm": 0.23181858054042714,
+      "learning_rate": 7.413178984861978e-06,
+      "loss": 0.3295,
+      "step": 2163
+    },
+    {
+      "epoch": 2.6009615384615383,
+      "grad_norm": 0.2192256945463143,
+      "learning_rate": 7.390917186108638e-06,
+      "loss": 0.3346,
+      "step": 2164
+    },
+    {
+      "epoch": 2.6021634615384617,
+      "grad_norm": 0.21728560112557982,
+      "learning_rate": 7.3686553873552985e-06,
+      "loss": 0.3283,
+      "step": 2165
+    },
+    {
+      "epoch": 2.6033653846153846,
+      "grad_norm": 0.21621059046541638,
+      "learning_rate": 7.346393588601959e-06,
+      "loss": 0.3325,
+      "step": 2166
+    },
+    {
+      "epoch": 2.6045673076923075,
+      "grad_norm": 0.24669486577280522,
+      "learning_rate": 7.32413178984862e-06,
+      "loss": 0.3231,
+      "step": 2167
+    },
+    {
+      "epoch": 2.605769230769231,
+      "grad_norm": 0.2100690243243088,
+      "learning_rate": 7.301869991095281e-06,
+      "loss": 0.2992,
+      "step": 2168
+    },
+    {
+      "epoch": 2.6069711538461537,
+      "grad_norm": 0.24808103487500627,
+      "learning_rate": 7.279608192341941e-06,
+      "loss": 0.3211,
+      "step": 2169
+    },
+    {
+      "epoch": 2.6081730769230766,
+      "grad_norm": 0.4040960356985032,
+      "learning_rate": 7.257346393588603e-06,
+      "loss": 0.3304,
+      "step": 2170
+    },
+    {
+      "epoch": 2.609375,
+      "grad_norm": 0.21581744668063943,
+      "learning_rate": 7.235084594835264e-06,
+      "loss": 0.321,
+      "step": 2171
+    },
+    {
+      "epoch": 2.6105769230769234,
+      "grad_norm": 0.25119633642126726,
+      "learning_rate": 7.212822796081923e-06,
+      "loss": 0.3225,
+      "step": 2172
+    },
+    {
+      "epoch": 2.6117788461538463,
+      "grad_norm": 0.23932567995099135,
+      "learning_rate": 7.190560997328584e-06,
+      "loss": 0.3263,
+      "step": 2173
+    },
+    {
+      "epoch": 2.612980769230769,
+      "grad_norm": 0.2144005516939886,
+      "learning_rate": 7.168299198575246e-06,
+      "loss": 0.3127,
+      "step": 2174
+    },
+    {
+      "epoch": 2.6141826923076925,
+      "grad_norm": 0.2260206779289561,
+      "learning_rate": 7.146037399821906e-06,
+      "loss": 0.3352,
+      "step": 2175
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.20612866367993116,
+      "learning_rate": 7.123775601068567e-06,
+      "loss": 0.3051,
+      "step": 2176
+    },
+    {
+      "epoch": 2.6165865384615383,
+      "grad_norm": 0.24181574173120637,
+      "learning_rate": 7.101513802315228e-06,
+      "loss": 0.3263,
+      "step": 2177
+    },
+    {
+      "epoch": 2.6177884615384617,
+      "grad_norm": 0.2281971273402615,
+      "learning_rate": 7.079252003561888e-06,
+      "loss": 0.3192,
+      "step": 2178
+    },
+    {
+      "epoch": 2.6189903846153846,
+      "grad_norm": 0.2282456160615323,
+      "learning_rate": 7.056990204808549e-06,
+      "loss": 0.3399,
+      "step": 2179
+    },
+    {
+      "epoch": 2.6201923076923075,
+      "grad_norm": 0.2173797683506271,
+      "learning_rate": 7.034728406055209e-06,
+      "loss": 0.331,
+      "step": 2180
+    },
+    {
+      "epoch": 2.621394230769231,
+      "grad_norm": 0.247846061683121,
+      "learning_rate": 7.0124666073018706e-06,
+      "loss": 0.3444,
+      "step": 2181
+    },
+    {
+      "epoch": 2.6225961538461537,
+      "grad_norm": 0.3274859463526463,
+      "learning_rate": 6.990204808548531e-06,
+      "loss": 0.3235,
+      "step": 2182
+    },
+    {
+      "epoch": 2.6237980769230766,
+      "grad_norm": 0.2504571820321685,
+      "learning_rate": 6.9679430097951914e-06,
+      "loss": 0.341,
+      "step": 2183
+    },
+    {
+      "epoch": 2.625,
+      "grad_norm": 0.20724197545167133,
+      "learning_rate": 6.9456812110418536e-06,
+      "loss": 0.3328,
+      "step": 2184
+    },
+    {
+      "epoch": 2.6262019230769234,
+      "grad_norm": 0.2367672255217757,
+      "learning_rate": 6.923419412288513e-06,
+      "loss": 0.311,
+      "step": 2185
+    },
+    {
+      "epoch": 2.6274038461538463,
+      "grad_norm": 0.24103208902154835,
+      "learning_rate": 6.901157613535174e-06,
+      "loss": 0.3394,
+      "step": 2186
+    },
+    {
+      "epoch": 2.628605769230769,
+      "grad_norm": 0.22419934411166256,
+      "learning_rate": 6.878895814781834e-06,
+      "loss": 0.3027,
+      "step": 2187
+    },
+    {
+      "epoch": 2.6298076923076925,
+      "grad_norm": 0.20626405162335654,
+      "learning_rate": 6.856634016028496e-06,
+      "loss": 0.3179,
+      "step": 2188
+    },
+    {
+      "epoch": 2.6310096153846154,
+      "grad_norm": 0.2295213783310608,
+      "learning_rate": 6.834372217275157e-06,
+      "loss": 0.3366,
+      "step": 2189
+    },
+    {
+      "epoch": 2.6322115384615383,
+      "grad_norm": 0.22496574122406368,
+      "learning_rate": 6.812110418521816e-06,
+      "loss": 0.326,
+      "step": 2190
+    },
+    {
+      "epoch": 2.6334134615384617,
+      "grad_norm": 0.22820735542097537,
+      "learning_rate": 6.789848619768478e-06,
+      "loss": 0.3311,
+      "step": 2191
+    },
+    {
+      "epoch": 2.6346153846153846,
+      "grad_norm": 0.21454506522624467,
+      "learning_rate": 6.767586821015139e-06,
+      "loss": 0.3177,
+      "step": 2192
+    },
+    {
+      "epoch": 2.6358173076923075,
+      "grad_norm": 0.21284222387247084,
+      "learning_rate": 6.745325022261799e-06,
+      "loss": 0.3023,
+      "step": 2193
+    },
+    {
+      "epoch": 2.637019230769231,
+      "grad_norm": 0.24096731838101654,
+      "learning_rate": 6.72306322350846e-06,
+      "loss": 0.3209,
+      "step": 2194
+    },
+    {
+      "epoch": 2.6382211538461537,
+      "grad_norm": 0.26171502613025094,
+      "learning_rate": 6.700801424755121e-06,
+      "loss": 0.3298,
+      "step": 2195
+    },
+    {
+      "epoch": 2.6394230769230766,
+      "grad_norm": 0.20943484132782844,
+      "learning_rate": 6.678539626001781e-06,
+      "loss": 0.3304,
+      "step": 2196
+    },
+    {
+      "epoch": 2.640625,
+      "grad_norm": 0.21166392786292137,
+      "learning_rate": 6.656277827248442e-06,
+      "loss": 0.3469,
+      "step": 2197
+    },
+    {
+      "epoch": 2.6418269230769234,
+      "grad_norm": 0.2227722958729637,
+      "learning_rate": 6.634016028495103e-06,
+      "loss": 0.3147,
+      "step": 2198
+    },
+    {
+      "epoch": 2.6430288461538463,
+      "grad_norm": 0.2200545332808951,
+      "learning_rate": 6.6117542297417635e-06,
+      "loss": 0.3113,
+      "step": 2199
+    },
+    {
+      "epoch": 2.644230769230769,
+      "grad_norm": 0.22879491369805308,
+      "learning_rate": 6.589492430988424e-06,
+      "loss": 0.3416,
+      "step": 2200
+    },
+    {
+      "epoch": 2.6454326923076925,
+      "grad_norm": 0.21399432805154145,
+      "learning_rate": 6.567230632235084e-06,
+      "loss": 0.3248,
+      "step": 2201
+    },
+    {
+      "epoch": 2.6466346153846154,
+      "grad_norm": 0.23800999734169384,
+      "learning_rate": 6.544968833481746e-06,
+      "loss": 0.3302,
+      "step": 2202
+    },
+    {
+      "epoch": 2.6478365384615383,
+      "grad_norm": 0.24087652588673183,
+      "learning_rate": 6.522707034728406e-06,
+      "loss": 0.3151,
+      "step": 2203
+    },
+    {
+      "epoch": 2.6490384615384617,
+      "grad_norm": 0.2277283801063856,
+      "learning_rate": 6.5004452359750666e-06,
+      "loss": 0.305,
+      "step": 2204
+    },
+    {
+      "epoch": 2.6502403846153846,
+      "grad_norm": 0.22601584009180414,
+      "learning_rate": 6.478183437221729e-06,
+      "loss": 0.3346,
+      "step": 2205
+    },
+    {
+      "epoch": 2.6514423076923075,
+      "grad_norm": 0.20353307931624298,
+      "learning_rate": 6.455921638468389e-06,
+      "loss": 0.3277,
+      "step": 2206
+    },
+    {
+      "epoch": 2.652644230769231,
+      "grad_norm": 0.20428102978199747,
+      "learning_rate": 6.433659839715049e-06,
+      "loss": 0.3161,
+      "step": 2207
+    },
+    {
+      "epoch": 2.6538461538461537,
+      "grad_norm": 0.23252852180056516,
+      "learning_rate": 6.411398040961709e-06,
+      "loss": 0.3239,
+      "step": 2208
+    },
+    {
+      "epoch": 2.6550480769230766,
+      "grad_norm": 0.2146502240082318,
+      "learning_rate": 6.389136242208371e-06,
+      "loss": 0.3413,
+      "step": 2209
+    },
+    {
+      "epoch": 2.65625,
+      "grad_norm": 0.21269142353939338,
+      "learning_rate": 6.366874443455032e-06,
+      "loss": 0.3303,
+      "step": 2210
+    },
+    {
+      "epoch": 2.6574519230769234,
+      "grad_norm": 0.20850980475101014,
+      "learning_rate": 6.344612644701692e-06,
+      "loss": 0.3235,
+      "step": 2211
+    },
+    {
+      "epoch": 2.6586538461538463,
+      "grad_norm": 0.2468248305495981,
+      "learning_rate": 6.3223508459483535e-06,
+      "loss": 0.3325,
+      "step": 2212
+    },
+    {
+      "epoch": 2.659855769230769,
+      "grad_norm": 0.21168295031453613,
+      "learning_rate": 6.300089047195014e-06,
+      "loss": 0.3403,
+      "step": 2213
+    },
+    {
+      "epoch": 2.6610576923076925,
+      "grad_norm": 0.24668379149146577,
+      "learning_rate": 6.277827248441674e-06,
+      "loss": 0.3279,
+      "step": 2214
+    },
+    {
+      "epoch": 2.6622596153846154,
+      "grad_norm": 0.21192710607534254,
+      "learning_rate": 6.255565449688335e-06,
+      "loss": 0.3017,
+      "step": 2215
+    },
+    {
+      "epoch": 2.6634615384615383,
+      "grad_norm": 0.22117589875011423,
+      "learning_rate": 6.233303650934995e-06,
+      "loss": 0.31,
+      "step": 2216
+    },
+    {
+      "epoch": 2.6646634615384617,
+      "grad_norm": 0.21934422110699972,
+      "learning_rate": 6.2110418521816565e-06,
+      "loss": 0.3472,
+      "step": 2217
+    },
+    {
+      "epoch": 2.6658653846153846,
+      "grad_norm": 0.23998488806173077,
+      "learning_rate": 6.188780053428318e-06,
+      "loss": 0.3108,
+      "step": 2218
+    },
+    {
+      "epoch": 2.6670673076923075,
+      "grad_norm": 0.245430688846935,
+      "learning_rate": 6.166518254674978e-06,
+      "loss": 0.3359,
+      "step": 2219
+    },
+    {
+      "epoch": 2.668269230769231,
+      "grad_norm": 0.22599849093129082,
+      "learning_rate": 6.144256455921639e-06,
+      "loss": 0.3156,
+      "step": 2220
+    },
+    {
+      "epoch": 2.6694711538461537,
+      "grad_norm": 0.2358303042028186,
+      "learning_rate": 6.121994657168299e-06,
+      "loss": 0.3522,
+      "step": 2221
+    },
+    {
+      "epoch": 2.6706730769230766,
+      "grad_norm": 0.22780109218376918,
+      "learning_rate": 6.09973285841496e-06,
+      "loss": 0.3466,
+      "step": 2222
+    },
+    {
+      "epoch": 2.671875,
+      "grad_norm": 0.2283039885823596,
+      "learning_rate": 6.077471059661621e-06,
+      "loss": 0.3259,
+      "step": 2223
+    },
+    {
+      "epoch": 2.6730769230769234,
+      "grad_norm": 0.21234311003151393,
+      "learning_rate": 6.055209260908281e-06,
+      "loss": 0.32,
+      "step": 2224
+    },
+    {
+      "epoch": 2.6742788461538463,
+      "grad_norm": 0.2117227245400787,
+      "learning_rate": 6.0329474621549425e-06,
+      "loss": 0.3101,
+      "step": 2225
+    },
+    {
+      "epoch": 2.675480769230769,
+      "grad_norm": 0.23407019959090672,
+      "learning_rate": 6.010685663401603e-06,
+      "loss": 0.3183,
+      "step": 2226
+    },
+    {
+      "epoch": 2.6766826923076925,
+      "grad_norm": 0.23837596990010557,
+      "learning_rate": 5.988423864648264e-06,
+      "loss": 0.3245,
+      "step": 2227
+    },
+    {
+      "epoch": 2.6778846153846154,
+      "grad_norm": 0.20972182484178853,
+      "learning_rate": 5.966162065894925e-06,
+      "loss": 0.3404,
+      "step": 2228
+    },
+    {
+      "epoch": 2.6790865384615383,
+      "grad_norm": 0.2134100599334659,
+      "learning_rate": 5.943900267141585e-06,
+      "loss": 0.3154,
+      "step": 2229
+    },
+    {
+      "epoch": 2.6802884615384617,
+      "grad_norm": 0.21032559412635568,
+      "learning_rate": 5.9216384683882456e-06,
+      "loss": 0.3298,
+      "step": 2230
+    },
+    {
+      "epoch": 2.6814903846153846,
+      "grad_norm": 0.2244942290915701,
+      "learning_rate": 5.899376669634907e-06,
+      "loss": 0.3226,
+      "step": 2231
+    },
+    {
+      "epoch": 2.6826923076923075,
+      "grad_norm": 0.22717013674795722,
+      "learning_rate": 5.877114870881568e-06,
+      "loss": 0.3311,
+      "step": 2232
+    },
+    {
+      "epoch": 2.683894230769231,
+      "grad_norm": 0.20710520691652248,
+      "learning_rate": 5.854853072128228e-06,
+      "loss": 0.3055,
+      "step": 2233
+    },
+    {
+      "epoch": 2.6850961538461537,
+      "grad_norm": 0.19990110290870772,
+      "learning_rate": 5.832591273374889e-06,
+      "loss": 0.3265,
+      "step": 2234
+    },
+    {
+      "epoch": 2.6862980769230766,
+      "grad_norm": 0.2084282987463783,
+      "learning_rate": 5.8103294746215495e-06,
+      "loss": 0.2991,
+      "step": 2235
+    },
+    {
+      "epoch": 2.6875,
+      "grad_norm": 0.22122216607479642,
+      "learning_rate": 5.788067675868211e-06,
+      "loss": 0.3261,
+      "step": 2236
+    },
+    {
+      "epoch": 2.6887019230769234,
+      "grad_norm": 0.20307439846287556,
+      "learning_rate": 5.765805877114871e-06,
+      "loss": 0.3321,
+      "step": 2237
+    },
+    {
+      "epoch": 2.6899038461538463,
+      "grad_norm": 0.22208796335635297,
+      "learning_rate": 5.743544078361532e-06,
+      "loss": 0.312,
+      "step": 2238
+    },
+    {
+      "epoch": 2.691105769230769,
+      "grad_norm": 0.2173087433664763,
+      "learning_rate": 5.721282279608193e-06,
+      "loss": 0.3222,
+      "step": 2239
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.20805333462784334,
+      "learning_rate": 5.699020480854853e-06,
+      "loss": 0.3334,
+      "step": 2240
+    },
+    {
+      "epoch": 2.6935096153846154,
+      "grad_norm": 0.2139176009230089,
+      "learning_rate": 5.676758682101515e-06,
+      "loss": 0.3226,
+      "step": 2241
+    },
+    {
+      "epoch": 2.6947115384615383,
+      "grad_norm": 0.2377160869270565,
+      "learning_rate": 5.654496883348174e-06,
+      "loss": 0.3182,
+      "step": 2242
+    },
+    {
+      "epoch": 2.6959134615384617,
+      "grad_norm": 0.21482257667229684,
+      "learning_rate": 5.6322350845948355e-06,
+      "loss": 0.3461,
+      "step": 2243
+    },
+    {
+      "epoch": 2.6971153846153846,
+      "grad_norm": 0.2248845632515664,
+      "learning_rate": 5.609973285841496e-06,
+      "loss": 0.3167,
+      "step": 2244
+    },
+    {
+      "epoch": 2.6983173076923075,
+      "grad_norm": 0.2081530364860066,
+      "learning_rate": 5.587711487088157e-06,
+      "loss": 0.313,
+      "step": 2245
+    },
+    {
+      "epoch": 2.699519230769231,
+      "grad_norm": 0.22877955718880372,
+      "learning_rate": 5.565449688334818e-06,
+      "loss": 0.3206,
+      "step": 2246
+    },
+    {
+      "epoch": 2.7007211538461537,
+      "grad_norm": 0.2289909517533349,
+      "learning_rate": 5.543187889581478e-06,
+      "loss": 0.3466,
+      "step": 2247
+    },
+    {
+      "epoch": 2.7019230769230766,
+      "grad_norm": 0.2108599457383338,
+      "learning_rate": 5.520926090828139e-06,
+      "loss": 0.3172,
+      "step": 2248
+    },
+    {
+      "epoch": 2.703125,
+      "grad_norm": 0.25281139340138,
+      "learning_rate": 5.4986642920748e-06,
+      "loss": 0.3211,
+      "step": 2249
+    },
+    {
+      "epoch": 2.7043269230769234,
+      "grad_norm": 0.2308316114601775,
+      "learning_rate": 5.476402493321461e-06,
+      "loss": 0.3179,
+      "step": 2250
+    },
+    {
+      "epoch": 2.7055288461538463,
+      "grad_norm": 0.2397937007392328,
+      "learning_rate": 5.4541406945681215e-06,
+      "loss": 0.3353,
+      "step": 2251
+    },
+    {
+      "epoch": 2.706730769230769,
+      "grad_norm": 0.2173111469242044,
+      "learning_rate": 5.431878895814782e-06,
+      "loss": 0.3226,
+      "step": 2252
+    },
+    {
+      "epoch": 2.7079326923076925,
+      "grad_norm": 0.22940631280822393,
+      "learning_rate": 5.409617097061443e-06,
+      "loss": 0.3283,
+      "step": 2253
+    },
+    {
+      "epoch": 2.7091346153846154,
+      "grad_norm": 0.22728340435910252,
+      "learning_rate": 5.387355298308104e-06,
+      "loss": 0.3251,
+      "step": 2254
+    },
+    {
+      "epoch": 2.7103365384615383,
+      "grad_norm": 0.20946273359316464,
+      "learning_rate": 5.365093499554764e-06,
+      "loss": 0.3099,
+      "step": 2255
+    },
+    {
+      "epoch": 2.7115384615384617,
+      "grad_norm": 0.21969144865731774,
+      "learning_rate": 5.342831700801425e-06,
+      "loss": 0.3303,
+      "step": 2256
+    },
+    {
+      "epoch": 2.7127403846153846,
+      "grad_norm": 0.22480563243079218,
+      "learning_rate": 5.320569902048086e-06,
+      "loss": 0.325,
+      "step": 2257
+    },
+    {
+      "epoch": 2.7139423076923075,
+      "grad_norm": 0.22919460341853934,
+      "learning_rate": 5.298308103294747e-06,
+      "loss": 0.3335,
+      "step": 2258
+    },
+    {
+      "epoch": 2.715144230769231,
+      "grad_norm": 0.2388532846412133,
+      "learning_rate": 5.276046304541407e-06,
+      "loss": 0.3324,
+      "step": 2259
+    },
+    {
+      "epoch": 2.7163461538461537,
+      "grad_norm": 0.21297804907582732,
+      "learning_rate": 5.253784505788068e-06,
+      "loss": 0.3332,
+      "step": 2260
+    },
+    {
+      "epoch": 2.7175480769230766,
+      "grad_norm": 0.20918075464492483,
+      "learning_rate": 5.2315227070347285e-06,
+      "loss": 0.3182,
+      "step": 2261
+    },
+    {
+      "epoch": 2.71875,
+      "grad_norm": 0.22429350535604153,
+      "learning_rate": 5.20926090828139e-06,
+      "loss": 0.3103,
+      "step": 2262
+    },
+    {
+      "epoch": 2.7199519230769234,
+      "grad_norm": 0.20945456543181853,
+      "learning_rate": 5.18699910952805e-06,
+      "loss": 0.3352,
+      "step": 2263
+    },
+    {
+      "epoch": 2.7211538461538463,
+      "grad_norm": 0.23945173982243803,
+      "learning_rate": 5.164737310774711e-06,
+      "loss": 0.3273,
+      "step": 2264
+    },
+    {
+      "epoch": 2.722355769230769,
+      "grad_norm": 0.21483118383137453,
+      "learning_rate": 5.142475512021372e-06,
+      "loss": 0.3224,
+      "step": 2265
+    },
+    {
+      "epoch": 2.7235576923076925,
+      "grad_norm": 0.20401699305154908,
+      "learning_rate": 5.120213713268032e-06,
+      "loss": 0.3261,
+      "step": 2266
+    },
+    {
+      "epoch": 2.7247596153846154,
+      "grad_norm": 0.24248434936392596,
+      "learning_rate": 5.097951914514694e-06,
+      "loss": 0.3208,
+      "step": 2267
+    },
+    {
+      "epoch": 2.7259615384615383,
+      "grad_norm": 0.22988328606110228,
+      "learning_rate": 5.075690115761353e-06,
+      "loss": 0.3236,
+      "step": 2268
+    },
+    {
+      "epoch": 2.7271634615384617,
+      "grad_norm": 0.20200974621985965,
+      "learning_rate": 5.0534283170080145e-06,
+      "loss": 0.3096,
+      "step": 2269
+    },
+    {
+      "epoch": 2.7283653846153846,
+      "grad_norm": 0.21043114963121134,
+      "learning_rate": 5.031166518254675e-06,
+      "loss": 0.3253,
+      "step": 2270
+    },
+    {
+      "epoch": 2.7295673076923075,
+      "grad_norm": 0.2118677596797426,
+      "learning_rate": 5.008904719501336e-06,
+      "loss": 0.3242,
+      "step": 2271
+    },
+    {
+      "epoch": 2.730769230769231,
+      "grad_norm": 0.20755889214098322,
+      "learning_rate": 4.986642920747997e-06,
+      "loss": 0.3208,
+      "step": 2272
+    },
+    {
+      "epoch": 2.7319711538461537,
+      "grad_norm": 0.22896685943450365,
+      "learning_rate": 4.964381121994657e-06,
+      "loss": 0.3094,
+      "step": 2273
+    },
+    {
+      "epoch": 2.7331730769230766,
+      "grad_norm": 0.2112548546924289,
+      "learning_rate": 4.942119323241318e-06,
+      "loss": 0.316,
+      "step": 2274
+    },
+    {
+      "epoch": 2.734375,
+      "grad_norm": 0.2092836458058015,
+      "learning_rate": 4.919857524487979e-06,
+      "loss": 0.3256,
+      "step": 2275
+    },
+    {
+      "epoch": 2.7355769230769234,
+      "grad_norm": 0.21222150968911263,
+      "learning_rate": 4.89759572573464e-06,
+      "loss": 0.3145,
+      "step": 2276
+    },
+    {
+      "epoch": 2.7367788461538463,
+      "grad_norm": 0.20575650269184637,
+      "learning_rate": 4.8753339269813e-06,
+      "loss": 0.3156,
+      "step": 2277
+    },
+    {
+      "epoch": 2.737980769230769,
+      "grad_norm": 0.20602552842904018,
+      "learning_rate": 4.853072128227961e-06,
+      "loss": 0.3129,
+      "step": 2278
+    },
+    {
+      "epoch": 2.7391826923076925,
+      "grad_norm": 0.219633129319683,
+      "learning_rate": 4.830810329474622e-06,
+      "loss": 0.3232,
+      "step": 2279
+    },
+    {
+      "epoch": 2.7403846153846154,
+      "grad_norm": 0.22455002996201173,
+      "learning_rate": 4.808548530721283e-06,
+      "loss": 0.329,
+      "step": 2280
+    },
+    {
+      "epoch": 2.7415865384615383,
+      "grad_norm": 0.2056240914613487,
+      "learning_rate": 4.786286731967943e-06,
+      "loss": 0.3192,
+      "step": 2281
+    },
+    {
+      "epoch": 2.7427884615384617,
+      "grad_norm": 0.20880933636997542,
+      "learning_rate": 4.764024933214604e-06,
+      "loss": 0.3302,
+      "step": 2282
+    },
+    {
+      "epoch": 2.7439903846153846,
+      "grad_norm": 0.21918509773081804,
+      "learning_rate": 4.741763134461265e-06,
+      "loss": 0.3205,
+      "step": 2283
+    },
+    {
+      "epoch": 2.7451923076923075,
+      "grad_norm": 0.20973680457090974,
+      "learning_rate": 4.719501335707925e-06,
+      "loss": 0.3219,
+      "step": 2284
+    },
+    {
+      "epoch": 2.746394230769231,
+      "grad_norm": 0.20009620554612975,
+      "learning_rate": 4.697239536954586e-06,
+      "loss": 0.3146,
+      "step": 2285
+    },
+    {
+      "epoch": 2.7475961538461537,
+      "grad_norm": 0.2258784675848955,
+      "learning_rate": 4.674977738201247e-06,
+      "loss": 0.3121,
+      "step": 2286
+    },
+    {
+      "epoch": 2.7487980769230766,
+      "grad_norm": 0.33022611951621006,
+      "learning_rate": 4.6527159394479075e-06,
+      "loss": 0.3118,
+      "step": 2287
+    },
+    {
+      "epoch": 2.75,
+      "grad_norm": 0.21380213032530812,
+      "learning_rate": 4.630454140694569e-06,
+      "loss": 0.336,
+      "step": 2288
+    },
+    {
+      "epoch": 2.7512019230769234,
+      "grad_norm": 0.20861312661887402,
+      "learning_rate": 4.608192341941229e-06,
+      "loss": 0.3335,
+      "step": 2289
+    },
+    {
+      "epoch": 2.7524038461538463,
+      "grad_norm": 0.2305456033904115,
+      "learning_rate": 4.58593054318789e-06,
+      "loss": 0.3224,
+      "step": 2290
+    },
+    {
+      "epoch": 2.753605769230769,
+      "grad_norm": 0.201764511385519,
+      "learning_rate": 4.56366874443455e-06,
+      "loss": 0.3166,
+      "step": 2291
+    },
+    {
+      "epoch": 2.7548076923076925,
+      "grad_norm": 0.21186565835838075,
+      "learning_rate": 4.541406945681211e-06,
+      "loss": 0.342,
+      "step": 2292
+    },
+    {
+      "epoch": 2.7560096153846154,
+      "grad_norm": 0.23038927737644163,
+      "learning_rate": 4.519145146927873e-06,
+      "loss": 0.3223,
+      "step": 2293
+    },
+    {
+      "epoch": 2.7572115384615383,
+      "grad_norm": 0.2098794738093442,
+      "learning_rate": 4.496883348174532e-06,
+      "loss": 0.3202,
+      "step": 2294
+    },
+    {
+      "epoch": 2.7584134615384617,
+      "grad_norm": 0.20710222162604877,
+      "learning_rate": 4.4746215494211935e-06,
+      "loss": 0.3218,
+      "step": 2295
+    },
+    {
+      "epoch": 2.7596153846153846,
+      "grad_norm": 0.21103479963693356,
+      "learning_rate": 4.452359750667854e-06,
+      "loss": 0.33,
+      "step": 2296
+    },
+    {
+      "epoch": 2.7608173076923075,
+      "grad_norm": 0.2144889622059913,
+      "learning_rate": 4.430097951914515e-06,
+      "loss": 0.3426,
+      "step": 2297
+    },
+    {
+      "epoch": 2.762019230769231,
+      "grad_norm": 0.2095491759882215,
+      "learning_rate": 4.407836153161176e-06,
+      "loss": 0.3254,
+      "step": 2298
+    },
+    {
+      "epoch": 2.7632211538461537,
+      "grad_norm": 0.20074799103910956,
+      "learning_rate": 4.385574354407836e-06,
+      "loss": 0.3152,
+      "step": 2299
+    },
+    {
+      "epoch": 2.7644230769230766,
+      "grad_norm": 0.21803956442742098,
+      "learning_rate": 4.363312555654497e-06,
+      "loss": 0.32,
+      "step": 2300
+    },
+    {
+      "epoch": 2.765625,
+      "grad_norm": 0.2174083010298193,
+      "learning_rate": 4.341050756901158e-06,
+      "loss": 0.3288,
+      "step": 2301
+    },
+    {
+      "epoch": 2.7668269230769234,
+      "grad_norm": 0.1978607855207011,
+      "learning_rate": 4.318788958147819e-06,
+      "loss": 0.3104,
+      "step": 2302
+    },
+    {
+      "epoch": 2.7680288461538463,
+      "grad_norm": 0.20508823723481617,
+      "learning_rate": 4.296527159394479e-06,
+      "loss": 0.3266,
+      "step": 2303
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.21355147824145185,
+      "learning_rate": 4.27426536064114e-06,
+      "loss": 0.2914,
+      "step": 2304
+    },
+    {
+      "epoch": 2.7704326923076925,
+      "grad_norm": 0.23416974330413057,
+      "learning_rate": 4.2520035618878004e-06,
+      "loss": 0.3342,
+      "step": 2305
+    },
+    {
+      "epoch": 2.7716346153846154,
+      "grad_norm": 0.21955634924914832,
+      "learning_rate": 4.229741763134462e-06,
+      "loss": 0.3344,
+      "step": 2306
+    },
+    {
+      "epoch": 2.7728365384615383,
+      "grad_norm": 0.20026038282427802,
+      "learning_rate": 4.207479964381122e-06,
+      "loss": 0.3033,
+      "step": 2307
+    },
+    {
+      "epoch": 2.7740384615384617,
+      "grad_norm": 0.21391773868618041,
+      "learning_rate": 4.185218165627783e-06,
+      "loss": 0.3126,
+      "step": 2308
+    },
+    {
+      "epoch": 2.7752403846153846,
+      "grad_norm": 0.23000336469664467,
+      "learning_rate": 4.162956366874444e-06,
+      "loss": 0.3324,
+      "step": 2309
+    },
+    {
+      "epoch": 2.7764423076923075,
+      "grad_norm": 0.19880892483948476,
+      "learning_rate": 4.140694568121104e-06,
+      "loss": 0.3043,
+      "step": 2310
+    },
+    {
+      "epoch": 2.777644230769231,
+      "grad_norm": 0.22114249289245647,
+      "learning_rate": 4.118432769367765e-06,
+      "loss": 0.3474,
+      "step": 2311
+    },
+    {
+      "epoch": 2.7788461538461537,
+      "grad_norm": 0.20511194409941744,
+      "learning_rate": 4.096170970614425e-06,
+      "loss": 0.3272,
+      "step": 2312
+    },
+    {
+      "epoch": 2.7800480769230766,
+      "grad_norm": 0.21656670697408426,
+      "learning_rate": 4.0739091718610865e-06,
+      "loss": 0.3109,
+      "step": 2313
+    },
+    {
+      "epoch": 2.78125,
+      "grad_norm": 0.20693133069281144,
+      "learning_rate": 4.051647373107748e-06,
+      "loss": 0.3091,
+      "step": 2314
+    },
+    {
+      "epoch": 2.7824519230769234,
+      "grad_norm": 0.21445249975019987,
+      "learning_rate": 4.029385574354408e-06,
+      "loss": 0.3343,
+      "step": 2315
+    },
+    {
+      "epoch": 2.7836538461538463,
+      "grad_norm": 0.20115538365867563,
+      "learning_rate": 4.007123775601069e-06,
+      "loss": 0.3357,
+      "step": 2316
+    },
+    {
+      "epoch": 2.784855769230769,
+      "grad_norm": 0.21562704990824155,
+      "learning_rate": 3.984861976847729e-06,
+      "loss": 0.3005,
+      "step": 2317
+    },
+    {
+      "epoch": 2.7860576923076925,
+      "grad_norm": 0.20583226192467022,
+      "learning_rate": 3.96260017809439e-06,
+      "loss": 0.319,
+      "step": 2318
+    },
+    {
+      "epoch": 2.7872596153846154,
+      "grad_norm": 0.21054062713145322,
+      "learning_rate": 3.940338379341051e-06,
+      "loss": 0.3416,
+      "step": 2319
+    },
+    {
+      "epoch": 2.7884615384615383,
+      "grad_norm": 0.19310564235369945,
+      "learning_rate": 3.918076580587711e-06,
+      "loss": 0.3016,
+      "step": 2320
+    },
+    {
+      "epoch": 2.7896634615384617,
+      "grad_norm": 0.21861951236883737,
+      "learning_rate": 3.8958147818343725e-06,
+      "loss": 0.3161,
+      "step": 2321
+    },
+    {
+      "epoch": 2.7908653846153846,
+      "grad_norm": 0.21155406320334913,
+      "learning_rate": 3.873552983081033e-06,
+      "loss": 0.3377,
+      "step": 2322
+    },
+    {
+      "epoch": 2.7920673076923075,
+      "grad_norm": 0.19164754489935343,
+      "learning_rate": 3.851291184327694e-06,
+      "loss": 0.3184,
+      "step": 2323
+    },
+    {
+      "epoch": 2.793269230769231,
+      "grad_norm": 0.22898832413609965,
+      "learning_rate": 3.829029385574355e-06,
+      "loss": 0.3155,
+      "step": 2324
+    },
+    {
+      "epoch": 2.7944711538461537,
+      "grad_norm": 0.22439456840486902,
+      "learning_rate": 3.8067675868210156e-06,
+      "loss": 0.328,
+      "step": 2325
+    },
+    {
+      "epoch": 2.7956730769230766,
+      "grad_norm": 0.21987711104649962,
+      "learning_rate": 3.7845057880676764e-06,
+      "loss": 0.3248,
+      "step": 2326
+    },
+    {
+      "epoch": 2.796875,
+      "grad_norm": 0.22791158956683705,
+      "learning_rate": 3.762243989314337e-06,
+      "loss": 0.3342,
+      "step": 2327
+    },
+    {
+      "epoch": 2.7980769230769234,
+      "grad_norm": 0.22185522281384268,
+      "learning_rate": 3.7399821905609977e-06,
+      "loss": 0.3263,
+      "step": 2328
+    },
+    {
+      "epoch": 2.7992788461538463,
+      "grad_norm": 0.206765590489316,
+      "learning_rate": 3.717720391807658e-06,
+      "loss": 0.3168,
+      "step": 2329
+    },
+    {
+      "epoch": 2.800480769230769,
+      "grad_norm": 0.2112599463796515,
+      "learning_rate": 3.695458593054319e-06,
+      "loss": 0.3285,
+      "step": 2330
+    },
+    {
+      "epoch": 2.8016826923076925,
+      "grad_norm": 0.210515297233265,
+      "learning_rate": 3.6731967943009795e-06,
+      "loss": 0.3234,
+      "step": 2331
+    },
+    {
+      "epoch": 2.8028846153846154,
+      "grad_norm": 0.20316394543479674,
+      "learning_rate": 3.6509349955476403e-06,
+      "loss": 0.3199,
+      "step": 2332
+    },
+    {
+      "epoch": 2.8040865384615383,
+      "grad_norm": 0.2285020677830581,
+      "learning_rate": 3.6286731967943016e-06,
+      "loss": 0.3383,
+      "step": 2333
+    },
+    {
+      "epoch": 2.8052884615384617,
+      "grad_norm": 0.2089341023338367,
+      "learning_rate": 3.6064113980409616e-06,
+      "loss": 0.317,
+      "step": 2334
+    },
+    {
+      "epoch": 2.8064903846153846,
+      "grad_norm": 0.21009332550950274,
+      "learning_rate": 3.584149599287623e-06,
+      "loss": 0.3209,
+      "step": 2335
+    },
+    {
+      "epoch": 2.8076923076923075,
+      "grad_norm": 0.21927669263942265,
+      "learning_rate": 3.5618878005342833e-06,
+      "loss": 0.317,
+      "step": 2336
+    },
+    {
+      "epoch": 2.808894230769231,
+      "grad_norm": 0.21822534134288263,
+      "learning_rate": 3.539626001780944e-06,
+      "loss": 0.3187,
+      "step": 2337
+    },
+    {
+      "epoch": 2.8100961538461537,
+      "grad_norm": 0.21626894255393714,
+      "learning_rate": 3.5173642030276046e-06,
+      "loss": 0.3384,
+      "step": 2338
+    },
+    {
+      "epoch": 2.8112980769230766,
+      "grad_norm": 0.22319802982804515,
+      "learning_rate": 3.4951024042742655e-06,
+      "loss": 0.3335,
+      "step": 2339
+    },
+    {
+      "epoch": 2.8125,
+      "grad_norm": 0.2091019740728171,
+      "learning_rate": 3.4728406055209268e-06,
+      "loss": 0.3106,
+      "step": 2340
+    },
+    {
+      "epoch": 2.8137019230769234,
+      "grad_norm": 0.21169983270344953,
+      "learning_rate": 3.450578806767587e-06,
+      "loss": 0.3313,
+      "step": 2341
+    },
+    {
+      "epoch": 2.8149038461538463,
+      "grad_norm": 0.21253745088380727,
+      "learning_rate": 3.428317008014248e-06,
+      "loss": 0.3081,
+      "step": 2342
+    },
+    {
+      "epoch": 2.816105769230769,
+      "grad_norm": 0.1981249222798371,
+      "learning_rate": 3.406055209260908e-06,
+      "loss": 0.3112,
+      "step": 2343
+    },
+    {
+      "epoch": 2.8173076923076925,
+      "grad_norm": 0.2036028506981652,
+      "learning_rate": 3.3837934105075694e-06,
+      "loss": 0.3293,
+      "step": 2344
+    },
+    {
+      "epoch": 2.8185096153846154,
+      "grad_norm": 0.21210195994216338,
+      "learning_rate": 3.36153161175423e-06,
+      "loss": 0.3313,
+      "step": 2345
+    },
+    {
+      "epoch": 2.8197115384615383,
+      "grad_norm": 0.21153124064379739,
+      "learning_rate": 3.3392698130008907e-06,
+      "loss": 0.3289,
+      "step": 2346
+    },
+    {
+      "epoch": 2.8209134615384617,
+      "grad_norm": 0.1931739661211508,
+      "learning_rate": 3.3170080142475515e-06,
+      "loss": 0.3188,
+      "step": 2347
+    },
+    {
+      "epoch": 2.8221153846153846,
+      "grad_norm": 0.2106293100163903,
+      "learning_rate": 3.294746215494212e-06,
+      "loss": 0.3174,
+      "step": 2348
+    },
+    {
+      "epoch": 2.8233173076923075,
+      "grad_norm": 0.19497040841352126,
+      "learning_rate": 3.272484416740873e-06,
+      "loss": 0.3108,
+      "step": 2349
+    },
+    {
+      "epoch": 2.824519230769231,
+      "grad_norm": 0.19975190566254578,
+      "learning_rate": 3.2502226179875333e-06,
+      "loss": 0.3081,
+      "step": 2350
+    },
+    {
+      "epoch": 2.8257211538461537,
+      "grad_norm": 0.2126150621722713,
+      "learning_rate": 3.2279608192341946e-06,
+      "loss": 0.3192,
+      "step": 2351
+    },
+    {
+      "epoch": 2.8269230769230766,
+      "grad_norm": 0.21787944912501767,
+      "learning_rate": 3.2056990204808546e-06,
+      "loss": 0.3172,
+      "step": 2352
+    },
+    {
+      "epoch": 2.828125,
+      "grad_norm": 0.20962794988372987,
+      "learning_rate": 3.183437221727516e-06,
+      "loss": 0.3246,
+      "step": 2353
+    },
+    {
+      "epoch": 2.8293269230769234,
+      "grad_norm": 0.2051883017531654,
+      "learning_rate": 3.1611754229741767e-06,
+      "loss": 0.3088,
+      "step": 2354
+    },
+    {
+      "epoch": 2.8305288461538463,
+      "grad_norm": 0.21157851380894024,
+      "learning_rate": 3.138913624220837e-06,
+      "loss": 0.3399,
+      "step": 2355
+    },
+    {
+      "epoch": 2.831730769230769,
+      "grad_norm": 0.1973144009652116,
+      "learning_rate": 3.1166518254674976e-06,
+      "loss": 0.3174,
+      "step": 2356
+    },
+    {
+      "epoch": 2.8329326923076925,
+      "grad_norm": 0.20394363958911063,
+      "learning_rate": 3.094390026714159e-06,
+      "loss": 0.3373,
+      "step": 2357
+    },
+    {
+      "epoch": 2.8341346153846154,
+      "grad_norm": 0.2116711327926234,
+      "learning_rate": 3.0721282279608193e-06,
+      "loss": 0.3214,
+      "step": 2358
+    },
+    {
+      "epoch": 2.8353365384615383,
+      "grad_norm": 0.2257414590520089,
+      "learning_rate": 3.04986642920748e-06,
+      "loss": 0.3172,
+      "step": 2359
+    },
+    {
+      "epoch": 2.8365384615384617,
+      "grad_norm": 0.2079787047392415,
+      "learning_rate": 3.0276046304541406e-06,
+      "loss": 0.3218,
+      "step": 2360
+    },
+    {
+      "epoch": 2.8377403846153846,
+      "grad_norm": 0.22134262413374645,
+      "learning_rate": 3.0053428317008015e-06,
+      "loss": 0.3108,
+      "step": 2361
+    },
+    {
+      "epoch": 2.8389423076923075,
+      "grad_norm": 0.208526116596908,
+      "learning_rate": 2.9830810329474623e-06,
+      "loss": 0.3328,
+      "step": 2362
+    },
+    {
+      "epoch": 2.840144230769231,
+      "grad_norm": 0.21288942602010738,
+      "learning_rate": 2.9608192341941228e-06,
+      "loss": 0.3537,
+      "step": 2363
+    },
+    {
+      "epoch": 2.8413461538461537,
+      "grad_norm": 0.20997818988628725,
+      "learning_rate": 2.938557435440784e-06,
+      "loss": 0.3178,
+      "step": 2364
+    },
+    {
+      "epoch": 2.8425480769230766,
+      "grad_norm": 0.2206997791886747,
+      "learning_rate": 2.9162956366874445e-06,
+      "loss": 0.319,
+      "step": 2365
+    },
+    {
+      "epoch": 2.84375,
+      "grad_norm": 0.20860562375039354,
+      "learning_rate": 2.8940338379341054e-06,
+      "loss": 0.3138,
+      "step": 2366
+    },
+    {
+      "epoch": 2.8449519230769234,
+      "grad_norm": 0.20762254959493393,
+      "learning_rate": 2.871772039180766e-06,
+      "loss": 0.3143,
+      "step": 2367
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.20922218191888492,
+      "learning_rate": 2.8495102404274267e-06,
+      "loss": 0.3242,
+      "step": 2368
+    },
+    {
+      "epoch": 2.847355769230769,
+      "grad_norm": 0.22790844169890573,
+      "learning_rate": 2.827248441674087e-06,
+      "loss": 0.3427,
+      "step": 2369
+    },
+    {
+      "epoch": 2.8485576923076925,
+      "grad_norm": 0.2249176752156991,
+      "learning_rate": 2.804986642920748e-06,
+      "loss": 0.3342,
+      "step": 2370
+    },
+    {
+      "epoch": 2.8497596153846154,
+      "grad_norm": 0.2038721216539552,
+      "learning_rate": 2.782724844167409e-06,
+      "loss": 0.3295,
+      "step": 2371
+    },
+    {
+      "epoch": 2.8509615384615383,
+      "grad_norm": 0.205558745543784,
+      "learning_rate": 2.7604630454140697e-06,
+      "loss": 0.3227,
+      "step": 2372
+    },
+    {
+      "epoch": 2.8521634615384617,
+      "grad_norm": 0.2139449293519989,
+      "learning_rate": 2.7382012466607306e-06,
+      "loss": 0.314,
+      "step": 2373
+    },
+    {
+      "epoch": 2.8533653846153846,
+      "grad_norm": 0.20502225523861384,
+      "learning_rate": 2.715939447907391e-06,
+      "loss": 0.3288,
+      "step": 2374
+    },
+    {
+      "epoch": 2.8545673076923075,
+      "grad_norm": 0.20767674270806918,
+      "learning_rate": 2.693677649154052e-06,
+      "loss": 0.3098,
+      "step": 2375
+    },
+    {
+      "epoch": 2.855769230769231,
+      "grad_norm": 0.22432601775001454,
+      "learning_rate": 2.6714158504007123e-06,
+      "loss": 0.3135,
+      "step": 2376
+    },
+    {
+      "epoch": 2.8569711538461537,
+      "grad_norm": 0.21614392425692963,
+      "learning_rate": 2.6491540516473736e-06,
+      "loss": 0.3362,
+      "step": 2377
+    },
+    {
+      "epoch": 2.8581730769230766,
+      "grad_norm": 0.20693410908803517,
+      "learning_rate": 2.626892252894034e-06,
+      "loss": 0.3187,
+      "step": 2378
+    },
+    {
+      "epoch": 2.859375,
+      "grad_norm": 0.2140753744901032,
+      "learning_rate": 2.604630454140695e-06,
+      "loss": 0.3301,
+      "step": 2379
+    },
+    {
+      "epoch": 2.8605769230769234,
+      "grad_norm": 0.21801264274613252,
+      "learning_rate": 2.5823686553873553e-06,
+      "loss": 0.3242,
+      "step": 2380
+    },
+    {
+      "epoch": 2.8617788461538463,
+      "grad_norm": 0.20419238202137194,
+      "learning_rate": 2.560106856634016e-06,
+      "loss": 0.3181,
+      "step": 2381
+    },
+    {
+      "epoch": 2.862980769230769,
+      "grad_norm": 0.20370357032660633,
+      "learning_rate": 2.5378450578806766e-06,
+      "loss": 0.3049,
+      "step": 2382
+    },
+    {
+      "epoch": 2.8641826923076925,
+      "grad_norm": 0.20891223636171613,
+      "learning_rate": 2.5155832591273375e-06,
+      "loss": 0.3146,
+      "step": 2383
+    },
+    {
+      "epoch": 2.8653846153846154,
+      "grad_norm": 0.2105465857979953,
+      "learning_rate": 2.4933214603739983e-06,
+      "loss": 0.3071,
+      "step": 2384
+    },
+    {
+      "epoch": 2.8665865384615383,
+      "grad_norm": 0.20967475931578147,
+      "learning_rate": 2.471059661620659e-06,
+      "loss": 0.3299,
+      "step": 2385
+    },
+    {
+      "epoch": 2.8677884615384617,
+      "grad_norm": 0.21980832881499712,
+      "learning_rate": 2.44879786286732e-06,
+      "loss": 0.32,
+      "step": 2386
+    },
+    {
+      "epoch": 2.8689903846153846,
+      "grad_norm": 0.2174833797345961,
+      "learning_rate": 2.4265360641139805e-06,
+      "loss": 0.3175,
+      "step": 2387
+    },
+    {
+      "epoch": 2.8701923076923075,
+      "grad_norm": 0.20104976811564237,
+      "learning_rate": 2.4042742653606414e-06,
+      "loss": 0.3218,
+      "step": 2388
+    },
+    {
+      "epoch": 2.871394230769231,
+      "grad_norm": 0.2018900817917325,
+      "learning_rate": 2.382012466607302e-06,
+      "loss": 0.314,
+      "step": 2389
+    },
+    {
+      "epoch": 2.8725961538461537,
+      "grad_norm": 0.20131967348438862,
+      "learning_rate": 2.3597506678539627e-06,
+      "loss": 0.3112,
+      "step": 2390
+    },
+    {
+      "epoch": 2.8737980769230766,
+      "grad_norm": 0.2277340900809495,
+      "learning_rate": 2.3374888691006235e-06,
+      "loss": 0.331,
+      "step": 2391
+    },
+    {
+      "epoch": 2.875,
+      "grad_norm": 0.200676758571091,
+      "learning_rate": 2.3152270703472844e-06,
+      "loss": 0.299,
+      "step": 2392
+    },
+    {
+      "epoch": 2.8762019230769234,
+      "grad_norm": 0.20813321889241176,
+      "learning_rate": 2.292965271593945e-06,
+      "loss": 0.2916,
+      "step": 2393
+    },
+    {
+      "epoch": 2.8774038461538463,
+      "grad_norm": 0.19860008327742518,
+      "learning_rate": 2.2707034728406057e-06,
+      "loss": 0.3176,
+      "step": 2394
+    },
+    {
+      "epoch": 2.878605769230769,
+      "grad_norm": 0.20107296808557198,
+      "learning_rate": 2.248441674087266e-06,
+      "loss": 0.3106,
+      "step": 2395
+    },
+    {
+      "epoch": 2.8798076923076925,
+      "grad_norm": 0.20732463277636706,
+      "learning_rate": 2.226179875333927e-06,
+      "loss": 0.3139,
+      "step": 2396
+    },
+    {
+      "epoch": 2.8810096153846154,
+      "grad_norm": 0.20860623884277918,
+      "learning_rate": 2.203918076580588e-06,
+      "loss": 0.3192,
+      "step": 2397
+    },
+    {
+      "epoch": 2.8822115384615383,
+      "grad_norm": 0.20132721162852218,
+      "learning_rate": 2.1816562778272487e-06,
+      "loss": 0.3122,
+      "step": 2398
+    },
+    {
+      "epoch": 2.8834134615384617,
+      "grad_norm": 0.19069226612151358,
+      "learning_rate": 2.1593944790739096e-06,
+      "loss": 0.3159,
+      "step": 2399
+    },
+    {
+      "epoch": 2.8846153846153846,
+      "grad_norm": 0.22047796618748602,
+      "learning_rate": 2.13713268032057e-06,
+      "loss": 0.3108,
+      "step": 2400
+    },
+    {
+      "epoch": 2.8858173076923075,
+      "grad_norm": 0.20360679615126653,
+      "learning_rate": 2.114870881567231e-06,
+      "loss": 0.3243,
+      "step": 2401
+    },
+    {
+      "epoch": 2.887019230769231,
+      "grad_norm": 0.21002597325343858,
+      "learning_rate": 2.0926090828138913e-06,
+      "loss": 0.3133,
+      "step": 2402
+    },
+    {
+      "epoch": 2.8882211538461537,
+      "grad_norm": 0.21434759299105915,
+      "learning_rate": 2.070347284060552e-06,
+      "loss": 0.3283,
+      "step": 2403
+    },
+    {
+      "epoch": 2.8894230769230766,
+      "grad_norm": 0.201439768739117,
+      "learning_rate": 2.0480854853072126e-06,
+      "loss": 0.3106,
+      "step": 2404
+    },
+    {
+      "epoch": 2.890625,
+      "grad_norm": 0.20943371958856688,
+      "learning_rate": 2.025823686553874e-06,
+      "loss": 0.3218,
+      "step": 2405
+    },
+    {
+      "epoch": 2.8918269230769234,
+      "grad_norm": 0.20873061255804443,
+      "learning_rate": 2.0035618878005343e-06,
+      "loss": 0.3409,
+      "step": 2406
+    },
+    {
+      "epoch": 2.8930288461538463,
+      "grad_norm": 0.20226493739815904,
+      "learning_rate": 1.981300089047195e-06,
+      "loss": 0.3132,
+      "step": 2407
+    },
+    {
+      "epoch": 2.894230769230769,
+      "grad_norm": 0.19820274968679683,
+      "learning_rate": 1.9590382902938556e-06,
+      "loss": 0.322,
+      "step": 2408
+    },
+    {
+      "epoch": 2.8954326923076925,
+      "grad_norm": 0.20991103840706454,
+      "learning_rate": 1.9367764915405165e-06,
+      "loss": 0.3153,
+      "step": 2409
+    },
+    {
+      "epoch": 2.8966346153846154,
+      "grad_norm": 0.20585038492144797,
+      "learning_rate": 1.9145146927871773e-06,
+      "loss": 0.3002,
+      "step": 2410
+    },
+    {
+      "epoch": 2.8978365384615383,
+      "grad_norm": 0.20638745843167686,
+      "learning_rate": 1.8922528940338382e-06,
+      "loss": 0.3218,
+      "step": 2411
+    },
+    {
+      "epoch": 2.8990384615384617,
+      "grad_norm": 0.20169514732743912,
+      "learning_rate": 1.8699910952804989e-06,
+      "loss": 0.3135,
+      "step": 2412
+    },
+    {
+      "epoch": 2.9002403846153846,
+      "grad_norm": 0.21346856397858902,
+      "learning_rate": 1.8477292965271595e-06,
+      "loss": 0.3261,
+      "step": 2413
+    },
+    {
+      "epoch": 2.9014423076923075,
+      "grad_norm": 0.20111659465319962,
+      "learning_rate": 1.8254674977738202e-06,
+      "loss": 0.3376,
+      "step": 2414
+    },
+    {
+      "epoch": 2.902644230769231,
+      "grad_norm": 0.2159996521841501,
+      "learning_rate": 1.8032056990204808e-06,
+      "loss": 0.3515,
+      "step": 2415
+    },
+    {
+      "epoch": 2.9038461538461537,
+      "grad_norm": 0.1938820205657798,
+      "learning_rate": 1.7809439002671417e-06,
+      "loss": 0.2999,
+      "step": 2416
+    },
+    {
+      "epoch": 2.9050480769230766,
+      "grad_norm": 0.19994183379130534,
+      "learning_rate": 1.7586821015138023e-06,
+      "loss": 0.3189,
+      "step": 2417
+    },
+    {
+      "epoch": 2.90625,
+      "grad_norm": 0.20637857600259568,
+      "learning_rate": 1.7364203027604634e-06,
+      "loss": 0.3353,
+      "step": 2418
+    },
+    {
+      "epoch": 2.9074519230769234,
+      "grad_norm": 0.27781401061656363,
+      "learning_rate": 1.714158504007124e-06,
+      "loss": 0.3256,
+      "step": 2419
+    },
+    {
+      "epoch": 2.9086538461538463,
+      "grad_norm": 0.1991777594442277,
+      "learning_rate": 1.6918967052537847e-06,
+      "loss": 0.3235,
+      "step": 2420
+    },
+    {
+      "epoch": 2.909855769230769,
+      "grad_norm": 0.20097902117567257,
+      "learning_rate": 1.6696349065004453e-06,
+      "loss": 0.3173,
+      "step": 2421
+    },
+    {
+      "epoch": 2.9110576923076925,
+      "grad_norm": 0.21366247259062351,
+      "learning_rate": 1.647373107747106e-06,
+      "loss": 0.3262,
+      "step": 2422
+    },
+    {
+      "epoch": 2.9122596153846154,
+      "grad_norm": 0.20476695024136493,
+      "learning_rate": 1.6251113089937666e-06,
+      "loss": 0.3173,
+      "step": 2423
+    },
+    {
+      "epoch": 2.9134615384615383,
+      "grad_norm": 0.2012871120934599,
+      "learning_rate": 1.6028495102404273e-06,
+      "loss": 0.3387,
+      "step": 2424
+    },
+    {
+      "epoch": 2.9146634615384617,
+      "grad_norm": 0.18911138496681312,
+      "learning_rate": 1.5805877114870884e-06,
+      "loss": 0.3046,
+      "step": 2425
+    },
+    {
+      "epoch": 2.9158653846153846,
+      "grad_norm": 0.20175320126557547,
+      "learning_rate": 1.5583259127337488e-06,
+      "loss": 0.3208,
+      "step": 2426
+    },
+    {
+      "epoch": 2.9170673076923075,
+      "grad_norm": 0.2040526268829482,
+      "learning_rate": 1.5360641139804097e-06,
+      "loss": 0.3209,
+      "step": 2427
+    },
+    {
+      "epoch": 2.918269230769231,
+      "grad_norm": 0.19679660201381438,
+      "learning_rate": 1.5138023152270703e-06,
+      "loss": 0.3234,
+      "step": 2428
+    },
+    {
+      "epoch": 2.9194711538461537,
+      "grad_norm": 0.22365474641772404,
+      "learning_rate": 1.4915405164737312e-06,
+      "loss": 0.3325,
+      "step": 2429
+    },
+    {
+      "epoch": 2.9206730769230766,
+      "grad_norm": 0.211629048886138,
+      "learning_rate": 1.469278717720392e-06,
+      "loss": 0.3155,
+      "step": 2430
+    },
+    {
+      "epoch": 2.921875,
+      "grad_norm": 0.20496427788211086,
+      "learning_rate": 1.4470169189670527e-06,
+      "loss": 0.3038,
+      "step": 2431
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.20451077054713418,
+      "learning_rate": 1.4247551202137133e-06,
+      "loss": 0.3283,
+      "step": 2432
+    },
+    {
+      "epoch": 2.9242788461538463,
+      "grad_norm": 0.19618301533586857,
+      "learning_rate": 1.402493321460374e-06,
+      "loss": 0.3194,
+      "step": 2433
+    },
+    {
+      "epoch": 2.925480769230769,
+      "grad_norm": 0.2365824184516574,
+      "learning_rate": 1.3802315227070348e-06,
+      "loss": 0.3297,
+      "step": 2434
+    },
+    {
+      "epoch": 2.9266826923076925,
+      "grad_norm": 0.20687093368094947,
+      "learning_rate": 1.3579697239536955e-06,
+      "loss": 0.3139,
+      "step": 2435
+    },
+    {
+      "epoch": 2.9278846153846154,
+      "grad_norm": 0.2011479706240221,
+      "learning_rate": 1.3357079252003561e-06,
+      "loss": 0.3277,
+      "step": 2436
+    },
+    {
+      "epoch": 2.9290865384615383,
+      "grad_norm": 0.19748031355953227,
+      "learning_rate": 1.313446126447017e-06,
+      "loss": 0.312,
+      "step": 2437
+    },
+    {
+      "epoch": 2.9302884615384617,
+      "grad_norm": 0.21295846485576952,
+      "learning_rate": 1.2911843276936777e-06,
+      "loss": 0.3298,
+      "step": 2438
+    },
+    {
+      "epoch": 2.9314903846153846,
+      "grad_norm": 0.21182732318876513,
+      "learning_rate": 1.2689225289403383e-06,
+      "loss": 0.3376,
+      "step": 2439
+    },
+    {
+      "epoch": 2.9326923076923075,
+      "grad_norm": 0.19554923787948136,
+      "learning_rate": 1.2466607301869992e-06,
+      "loss": 0.3051,
+      "step": 2440
+    },
+    {
+      "epoch": 2.933894230769231,
+      "grad_norm": 0.2241359286374354,
+      "learning_rate": 1.22439893143366e-06,
+      "loss": 0.3156,
+      "step": 2441
+    },
+    {
+      "epoch": 2.9350961538461537,
+      "grad_norm": 0.19728894366548966,
+      "learning_rate": 1.2021371326803207e-06,
+      "loss": 0.3154,
+      "step": 2442
+    },
+    {
+      "epoch": 2.9362980769230766,
+      "grad_norm": 0.22153868471688312,
+      "learning_rate": 1.1798753339269813e-06,
+      "loss": 0.3165,
+      "step": 2443
+    },
+    {
+      "epoch": 2.9375,
+      "grad_norm": 0.20154225533572526,
+      "learning_rate": 1.1576135351736422e-06,
+      "loss": 0.3085,
+      "step": 2444
+    },
+    {
+      "epoch": 2.9387019230769234,
+      "grad_norm": 0.20080171909732408,
+      "learning_rate": 1.1353517364203028e-06,
+      "loss": 0.3205,
+      "step": 2445
+    },
+    {
+      "epoch": 2.9399038461538463,
+      "grad_norm": 0.20324575874173634,
+      "learning_rate": 1.1130899376669635e-06,
+      "loss": 0.3142,
+      "step": 2446
+    },
+    {
+      "epoch": 2.941105769230769,
+      "grad_norm": 0.19065665376570026,
+      "learning_rate": 1.0908281389136244e-06,
+      "loss": 0.2978,
+      "step": 2447
+    },
+    {
+      "epoch": 2.9423076923076925,
+      "grad_norm": 0.19750478755848652,
+      "learning_rate": 1.068566340160285e-06,
+      "loss": 0.3245,
+      "step": 2448
+    },
+    {
+      "epoch": 2.9435096153846154,
+      "grad_norm": 0.20442034313212323,
+      "learning_rate": 1.0463045414069457e-06,
+      "loss": 0.3288,
+      "step": 2449
+    },
+    {
+      "epoch": 2.9447115384615383,
+      "grad_norm": 0.20344855450625435,
+      "learning_rate": 1.0240427426536063e-06,
+      "loss": 0.3288,
+      "step": 2450
+    },
+    {
+      "epoch": 2.9459134615384617,
+      "grad_norm": 0.21728067033775617,
+      "learning_rate": 1.0017809439002672e-06,
+      "loss": 0.322,
+      "step": 2451
+    },
+    {
+      "epoch": 2.9471153846153846,
+      "grad_norm": 0.1925873077820777,
+      "learning_rate": 9.795191451469278e-07,
+      "loss": 0.3116,
+      "step": 2452
+    },
+    {
+      "epoch": 2.9483173076923075,
+      "grad_norm": 0.20017998490266767,
+      "learning_rate": 9.572573463935887e-07,
+      "loss": 0.3143,
+      "step": 2453
+    },
+    {
+      "epoch": 2.949519230769231,
+      "grad_norm": 0.19545915076388307,
+      "learning_rate": 9.349955476402494e-07,
+      "loss": 0.3145,
+      "step": 2454
+    },
+    {
+      "epoch": 2.9507211538461537,
+      "grad_norm": 0.19451292322268096,
+      "learning_rate": 9.127337488869101e-07,
+      "loss": 0.3183,
+      "step": 2455
+    },
+    {
+      "epoch": 2.9519230769230766,
+      "grad_norm": 0.20043814749258262,
+      "learning_rate": 8.904719501335708e-07,
+      "loss": 0.3325,
+      "step": 2456
+    },
+    {
+      "epoch": 2.953125,
+      "grad_norm": 0.20613046756970607,
+      "learning_rate": 8.682101513802317e-07,
+      "loss": 0.3208,
+      "step": 2457
+    },
+    {
+      "epoch": 2.9543269230769234,
+      "grad_norm": 0.19534664344024533,
+      "learning_rate": 8.459483526268923e-07,
+      "loss": 0.3123,
+      "step": 2458
+    },
+    {
+      "epoch": 2.9555288461538463,
+      "grad_norm": 0.20600734618451222,
+      "learning_rate": 8.23686553873553e-07,
+      "loss": 0.3226,
+      "step": 2459
+    },
+    {
+      "epoch": 2.956730769230769,
+      "grad_norm": 0.1938553612504044,
+      "learning_rate": 8.014247551202136e-07,
+      "loss": 0.3338,
+      "step": 2460
+    },
+    {
+      "epoch": 2.9579326923076925,
+      "grad_norm": 0.20158582807877923,
+      "learning_rate": 7.791629563668744e-07,
+      "loss": 0.3133,
+      "step": 2461
+    },
+    {
+      "epoch": 2.9591346153846154,
+      "grad_norm": 0.20438067663244727,
+      "learning_rate": 7.569011576135352e-07,
+      "loss": 0.3319,
+      "step": 2462
+    },
+    {
+      "epoch": 2.9603365384615383,
+      "grad_norm": 0.2022667062674894,
+      "learning_rate": 7.34639358860196e-07,
+      "loss": 0.3101,
+      "step": 2463
+    },
+    {
+      "epoch": 2.9615384615384617,
+      "grad_norm": 0.19629980956875406,
+      "learning_rate": 7.123775601068567e-07,
+      "loss": 0.3199,
+      "step": 2464
+    },
+    {
+      "epoch": 2.9627403846153846,
+      "grad_norm": 0.21574458823807213,
+      "learning_rate": 6.901157613535174e-07,
+      "loss": 0.3394,
+      "step": 2465
+    },
+    {
+      "epoch": 2.9639423076923075,
+      "grad_norm": 0.19923828313626818,
+      "learning_rate": 6.678539626001781e-07,
+      "loss": 0.3114,
+      "step": 2466
+    },
+    {
+      "epoch": 2.965144230769231,
+      "grad_norm": 0.20055937816799171,
+      "learning_rate": 6.455921638468388e-07,
+      "loss": 0.3135,
+      "step": 2467
+    },
+    {
+      "epoch": 2.9663461538461537,
+      "grad_norm": 0.1961734557526551,
+      "learning_rate": 6.233303650934996e-07,
+      "loss": 0.3159,
+      "step": 2468
+    },
+    {
+      "epoch": 2.9675480769230766,
+      "grad_norm": 0.19109640687757581,
+      "learning_rate": 6.010685663401603e-07,
+      "loss": 0.3052,
+      "step": 2469
+    },
+    {
+      "epoch": 2.96875,
+      "grad_norm": 0.20350837148410403,
+      "learning_rate": 5.788067675868211e-07,
+      "loss": 0.3327,
+      "step": 2470
+    },
+    {
+      "epoch": 2.9699519230769234,
+      "grad_norm": 0.20047795356019152,
+      "learning_rate": 5.565449688334817e-07,
+      "loss": 0.3156,
+      "step": 2471
+    },
+    {
+      "epoch": 2.9711538461538463,
+      "grad_norm": 0.19805868140868124,
+      "learning_rate": 5.342831700801425e-07,
+      "loss": 0.3146,
+      "step": 2472
+    },
+    {
+      "epoch": 2.972355769230769,
+      "grad_norm": 0.18809435028405155,
+      "learning_rate": 5.120213713268032e-07,
+      "loss": 0.3105,
+      "step": 2473
+    },
+    {
+      "epoch": 2.9735576923076925,
+      "grad_norm": 0.19591437716894264,
+      "learning_rate": 4.897595725734639e-07,
+      "loss": 0.3278,
+      "step": 2474
+    },
+    {
+      "epoch": 2.9747596153846154,
+      "grad_norm": 0.218634662523432,
+      "learning_rate": 4.674977738201247e-07,
+      "loss": 0.3259,
+      "step": 2475
+    },
+    {
+      "epoch": 2.9759615384615383,
+      "grad_norm": 0.2021035471311907,
+      "learning_rate": 4.452359750667854e-07,
+      "loss": 0.3104,
+      "step": 2476
+    },
+    {
+      "epoch": 2.9771634615384617,
+      "grad_norm": 0.20091445833065733,
+      "learning_rate": 4.2297417631344617e-07,
+      "loss": 0.316,
+      "step": 2477
+    },
+    {
+      "epoch": 2.9783653846153846,
+      "grad_norm": 0.19298658189715018,
+      "learning_rate": 4.007123775601068e-07,
+      "loss": 0.3211,
+      "step": 2478
+    },
+    {
+      "epoch": 2.9795673076923075,
+      "grad_norm": 0.20017714171791726,
+      "learning_rate": 3.784505788067676e-07,
+      "loss": 0.3101,
+      "step": 2479
+    },
+    {
+      "epoch": 2.980769230769231,
+      "grad_norm": 0.1977819098813659,
+      "learning_rate": 3.5618878005342833e-07,
+      "loss": 0.3141,
+      "step": 2480
+    },
+    {
+      "epoch": 2.9819711538461537,
+      "grad_norm": 0.22104150283914262,
+      "learning_rate": 3.3392698130008904e-07,
+      "loss": 0.3237,
+      "step": 2481
+    },
+    {
+      "epoch": 2.9831730769230766,
+      "grad_norm": 0.2003996058400876,
+      "learning_rate": 3.116651825467498e-07,
+      "loss": 0.3237,
+      "step": 2482
+    },
+    {
+      "epoch": 2.984375,
+      "grad_norm": 0.2087065256331076,
+      "learning_rate": 2.8940338379341055e-07,
+      "loss": 0.3305,
+      "step": 2483
+    },
+    {
+      "epoch": 2.9855769230769234,
+      "grad_norm": 0.1987750834430871,
+      "learning_rate": 2.6714158504007125e-07,
+      "loss": 0.3228,
+      "step": 2484
+    },
+    {
+      "epoch": 2.9867788461538463,
+      "grad_norm": 0.2019797460427991,
+      "learning_rate": 2.4487978628673195e-07,
+      "loss": 0.317,
+      "step": 2485
+    },
+    {
+      "epoch": 2.987980769230769,
+      "grad_norm": 0.20114919875656193,
+      "learning_rate": 2.226179875333927e-07,
+      "loss": 0.3141,
+      "step": 2486
+    },
+    {
+      "epoch": 2.9891826923076925,
+      "grad_norm": 0.21059202675105237,
+      "learning_rate": 2.003561887800534e-07,
+      "loss": 0.3298,
+      "step": 2487
+    },
+    {
+      "epoch": 2.9903846153846154,
+      "grad_norm": 0.19801890311274234,
+      "learning_rate": 1.7809439002671417e-07,
+      "loss": 0.3105,
+      "step": 2488
+    },
+    {
+      "epoch": 2.9915865384615383,
+      "grad_norm": 0.19253558375093824,
+      "learning_rate": 1.558325912733749e-07,
+      "loss": 0.3069,
+      "step": 2489
+    },
+    {
+      "epoch": 2.9927884615384617,
+      "grad_norm": 0.19553706710255356,
+      "learning_rate": 1.3357079252003563e-07,
+      "loss": 0.323,
+      "step": 2490
+    },
+    {
+      "epoch": 2.9939903846153846,
+      "grad_norm": 0.20505944104997983,
+      "learning_rate": 1.1130899376669635e-07,
+      "loss": 0.342,
+      "step": 2491
+    },
+    {
+      "epoch": 2.9951923076923075,
+      "grad_norm": 0.21096898628417732,
+      "learning_rate": 8.904719501335708e-08,
+      "loss": 0.3319,
+      "step": 2492
+    },
+    {
+      "epoch": 2.996394230769231,
+      "grad_norm": 0.22616232984039394,
+      "learning_rate": 6.678539626001781e-08,
+      "loss": 0.3215,
+      "step": 2493
+    },
+    {
+      "epoch": 2.9975961538461537,
+      "grad_norm": 0.19732308240200605,
+      "learning_rate": 4.452359750667854e-08,
+      "loss": 0.3216,
+      "step": 2494
+    },
+    {
+      "epoch": 2.9987980769230766,
+      "grad_norm": 0.20657671320594906,
+      "learning_rate": 2.226179875333927e-08,
+      "loss": 0.3375,
+      "step": 2495
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.19579786694915596,
+      "learning_rate": 0.0,
+      "loss": 0.327,
+      "step": 2496
+    },
+    {
+      "epoch": 3.0,
+      "step": 2496,
+      "total_flos": 2.134279390798807e+18,
+      "train_loss": 0.5064016570313237,
+      "train_runtime": 146267.2004,
+      "train_samples_per_second": 0.273,
+      "train_steps_per_second": 0.017
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2496,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.134279390798807e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}