diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,18438 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.999429549343982,
+  "eval_steps": 500,
+  "global_step": 2628,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011409013120365088,
+      "grad_norm": 54.65760974624293,
+      "learning_rate": 1.9011406844106465e-07,
+      "loss": 11.0934,
+      "step": 1
+    },
+    {
+      "epoch": 0.0022818026240730175,
+      "grad_norm": 54.77744512464696,
+      "learning_rate": 3.802281368821293e-07,
+      "loss": 11.0737,
+      "step": 2
+    },
+    {
+      "epoch": 0.0034227039361095267,
+      "grad_norm": 58.007163271645624,
+      "learning_rate": 5.70342205323194e-07,
+      "loss": 10.9735,
+      "step": 3
+    },
+    {
+      "epoch": 0.004563605248146035,
+      "grad_norm": 55.77970400435833,
+      "learning_rate": 7.604562737642586e-07,
+      "loss": 10.9883,
+      "step": 4
+    },
+    {
+      "epoch": 0.005704506560182544,
+      "grad_norm": 57.7425202963039,
+      "learning_rate": 9.505703422053232e-07,
+      "loss": 10.7121,
+      "step": 5
+    },
+    {
+      "epoch": 0.0068454078722190535,
+      "grad_norm": 56.50330125597672,
+      "learning_rate": 1.140684410646388e-06,
+      "loss": 10.9614,
+      "step": 6
+    },
+    {
+      "epoch": 0.007986309184255563,
+      "grad_norm": 54.6706710453405,
+      "learning_rate": 1.3307984790874525e-06,
+      "loss": 11.004,
+      "step": 7
+    },
+    {
+      "epoch": 0.00912721049629207,
+      "grad_norm": 60.82691342176186,
+      "learning_rate": 1.5209125475285172e-06,
+      "loss": 10.7417,
+      "step": 8
+    },
+    {
+      "epoch": 0.01026811180832858,
+      "grad_norm": 60.66911793811066,
+      "learning_rate": 1.711026615969582e-06,
+      "loss": 10.6853,
+      "step": 9
+    },
+    {
+      "epoch": 0.011409013120365089,
+      "grad_norm": 62.943469991829076,
+      "learning_rate": 1.9011406844106463e-06,
+      "loss": 10.5299,
+      "step": 10
+    },
+    {
+      "epoch": 0.012549914432401598,
+      "grad_norm": 87.13133317065872,
+      "learning_rate": 2.091254752851711e-06,
+      "loss": 9.2991,
+      "step": 11
+    },
+    {
+      "epoch": 0.013690815744438107,
+      "grad_norm": 92.23580150407606,
+      "learning_rate": 2.281368821292776e-06,
+      "loss": 9.1733,
+      "step": 12
+    },
+    {
+      "epoch": 0.014831717056474614,
+      "grad_norm": 98.2645411335883,
+      "learning_rate": 2.4714828897338406e-06,
+      "loss": 8.7304,
+      "step": 13
+    },
+    {
+      "epoch": 0.015972618368511125,
+      "grad_norm": 102.02071115957615,
+      "learning_rate": 2.661596958174905e-06,
+      "loss": 8.5508,
+      "step": 14
+    },
+    {
+      "epoch": 0.017113519680547633,
+      "grad_norm": 63.610623784408745,
+      "learning_rate": 2.8517110266159697e-06,
+      "loss": 3.6785,
+      "step": 15
+    },
+    {
+      "epoch": 0.01825442099258414,
+      "grad_norm": 56.24211488085604,
+      "learning_rate": 3.0418250950570345e-06,
+      "loss": 3.4128,
+      "step": 16
+    },
+    {
+      "epoch": 0.01939532230462065,
+      "grad_norm": 43.70694807512611,
+      "learning_rate": 3.2319391634980988e-06,
+      "loss": 2.8317,
+      "step": 17
+    },
+    {
+      "epoch": 0.02053622361665716,
+      "grad_norm": 33.75445660517821,
+      "learning_rate": 3.422053231939164e-06,
+      "loss": 2.4295,
+      "step": 18
+    },
+    {
+      "epoch": 0.02167712492869367,
+      "grad_norm": 25.615342965674877,
+      "learning_rate": 3.612167300380228e-06,
+      "loss": 2.1501,
+      "step": 19
+    },
+    {
+      "epoch": 0.022818026240730177,
+      "grad_norm": 6.639036637383511,
+      "learning_rate": 3.8022813688212926e-06,
+      "loss": 1.3864,
+      "step": 20
+    },
+    {
+      "epoch": 0.023958927552766685,
+      "grad_norm": 5.463351047298746,
+      "learning_rate": 3.992395437262358e-06,
+      "loss": 1.3348,
+      "step": 21
+    },
+    {
+      "epoch": 0.025099828864803195,
+      "grad_norm": 4.486017466274795,
+      "learning_rate": 4.182509505703422e-06,
+      "loss": 1.277,
+      "step": 22
+    },
+    {
+      "epoch": 0.026240730176839703,
+      "grad_norm": 3.403461105976068,
+      "learning_rate": 4.3726235741444865e-06,
+      "loss": 1.2209,
+      "step": 23
+    },
+    {
+      "epoch": 0.027381631488876214,
+      "grad_norm": 2.7431919860295952,
+      "learning_rate": 4.562737642585552e-06,
+      "loss": 1.1256,
+      "step": 24
+    },
+    {
+      "epoch": 0.02852253280091272,
+      "grad_norm": 2.1398348178705002,
+      "learning_rate": 4.752851711026616e-06,
+      "loss": 1.0774,
+      "step": 25
+    },
+    {
+      "epoch": 0.02966343411294923,
+      "grad_norm": 1.6629279927743938,
+      "learning_rate": 4.942965779467681e-06,
+      "loss": 1.0002,
+      "step": 26
+    },
+    {
+      "epoch": 0.03080433542498574,
+      "grad_norm": 5.096050154303005,
+      "learning_rate": 5.1330798479087455e-06,
+      "loss": 0.9648,
+      "step": 27
+    },
+    {
+      "epoch": 0.03194523673702225,
+      "grad_norm": 20.80863576938441,
+      "learning_rate": 5.32319391634981e-06,
+      "loss": 0.9545,
+      "step": 28
+    },
+    {
+      "epoch": 0.03308613804905876,
+      "grad_norm": 3.60709294861697,
+      "learning_rate": 5.513307984790875e-06,
+      "loss": 0.9197,
+      "step": 29
+    },
+    {
+      "epoch": 0.034227039361095266,
+      "grad_norm": 1.38157954773502,
+      "learning_rate": 5.703422053231939e-06,
+      "loss": 0.8682,
+      "step": 30
+    },
+    {
+      "epoch": 0.03536794067313177,
+      "grad_norm": 1.0909877915531283,
+      "learning_rate": 5.893536121673004e-06,
+      "loss": 0.8942,
+      "step": 31
+    },
+    {
+      "epoch": 0.03650884198516828,
+      "grad_norm": 0.9344301030076203,
+      "learning_rate": 6.083650190114069e-06,
+      "loss": 0.8461,
+      "step": 32
+    },
+    {
+      "epoch": 0.037649743297204795,
+      "grad_norm": 0.8591775118329756,
+      "learning_rate": 6.273764258555133e-06,
+      "loss": 0.8398,
+      "step": 33
+    },
+    {
+      "epoch": 0.0387906446092413,
+      "grad_norm": 0.7938493134058952,
+      "learning_rate": 6.4638783269961976e-06,
+      "loss": 0.8067,
+      "step": 34
+    },
+    {
+      "epoch": 0.03993154592127781,
+      "grad_norm": 0.7271814029051149,
+      "learning_rate": 6.653992395437263e-06,
+      "loss": 0.8076,
+      "step": 35
+    },
+    {
+      "epoch": 0.04107244723331432,
+      "grad_norm": 0.8987059398819114,
+      "learning_rate": 6.844106463878328e-06,
+      "loss": 0.813,
+      "step": 36
+    },
+    {
+      "epoch": 0.042213348545350825,
+      "grad_norm": 0.7348990443928746,
+      "learning_rate": 7.034220532319392e-06,
+      "loss": 0.7624,
+      "step": 37
+    },
+    {
+      "epoch": 0.04335424985738734,
+      "grad_norm": 0.6060862909294358,
+      "learning_rate": 7.224334600760456e-06,
+      "loss": 0.7888,
+      "step": 38
+    },
+    {
+      "epoch": 0.04449515116942385,
+      "grad_norm": 0.6572288005824576,
+      "learning_rate": 7.414448669201521e-06,
+      "loss": 0.7309,
+      "step": 39
+    },
+    {
+      "epoch": 0.045636052481460354,
+      "grad_norm": 0.7414983714652406,
+      "learning_rate": 7.604562737642585e-06,
+      "loss": 0.7613,
+      "step": 40
+    },
+    {
+      "epoch": 0.04677695379349686,
+      "grad_norm": 0.5647065598184654,
+      "learning_rate": 7.79467680608365e-06,
+      "loss": 0.714,
+      "step": 41
+    },
+    {
+      "epoch": 0.04791785510553337,
+      "grad_norm": 0.4897340721044095,
+      "learning_rate": 7.984790874524716e-06,
+      "loss": 0.6845,
+      "step": 42
+    },
+    {
+      "epoch": 0.04905875641756988,
+      "grad_norm": 0.47501108184275426,
+      "learning_rate": 8.17490494296578e-06,
+      "loss": 0.6938,
+      "step": 43
+    },
+    {
+      "epoch": 0.05019965772960639,
+      "grad_norm": 0.4100828588343714,
+      "learning_rate": 8.365019011406844e-06,
+      "loss": 0.661,
+      "step": 44
+    },
+    {
+      "epoch": 0.0513405590416429,
+      "grad_norm": 0.4600477852854328,
+      "learning_rate": 8.55513307984791e-06,
+      "loss": 0.6711,
+      "step": 45
+    },
+    {
+      "epoch": 0.052481460353679406,
+      "grad_norm": 0.49355119570935346,
+      "learning_rate": 8.745247148288973e-06,
+      "loss": 0.6621,
+      "step": 46
+    },
+    {
+      "epoch": 0.05362236166571591,
+      "grad_norm": 0.4895364232739456,
+      "learning_rate": 8.935361216730038e-06,
+      "loss": 0.6596,
+      "step": 47
+    },
+    {
+      "epoch": 0.05476326297775243,
+      "grad_norm": 0.380390266394673,
+      "learning_rate": 9.125475285171103e-06,
+      "loss": 0.6605,
+      "step": 48
+    },
+    {
+      "epoch": 0.055904164289788935,
+      "grad_norm": 0.4067888570775391,
+      "learning_rate": 9.315589353612169e-06,
+      "loss": 0.636,
+      "step": 49
+    },
+    {
+      "epoch": 0.05704506560182544,
+      "grad_norm": 0.41398686146929453,
+      "learning_rate": 9.505703422053232e-06,
+      "loss": 0.6569,
+      "step": 50
+    },
+    {
+      "epoch": 0.05818596691386195,
+      "grad_norm": 0.4159751891164936,
+      "learning_rate": 9.695817490494297e-06,
+      "loss": 0.6425,
+      "step": 51
+    },
+    {
+      "epoch": 0.05932686822589846,
+      "grad_norm": 0.4311602717231244,
+      "learning_rate": 9.885931558935362e-06,
+      "loss": 0.6617,
+      "step": 52
+    },
+    {
+      "epoch": 0.06046776953793497,
+      "grad_norm": 0.420579457115958,
+      "learning_rate": 1.0076045627376426e-05,
+      "loss": 0.6413,
+      "step": 53
+    },
+    {
+      "epoch": 0.06160867084997148,
+      "grad_norm": 0.32749957676716046,
+      "learning_rate": 1.0266159695817491e-05,
+      "loss": 0.6239,
+      "step": 54
+    },
+    {
+      "epoch": 0.06274957216200798,
+      "grad_norm": 0.33903263089513586,
+      "learning_rate": 1.0456273764258556e-05,
+      "loss": 0.6029,
+      "step": 55
+    },
+    {
+      "epoch": 0.0638904734740445,
+      "grad_norm": 0.3428426351133492,
+      "learning_rate": 1.064638783269962e-05,
+      "loss": 0.6329,
+      "step": 56
+    },
+    {
+      "epoch": 0.06503137478608101,
+      "grad_norm": 0.3028264313705029,
+      "learning_rate": 1.0836501901140685e-05,
+      "loss": 0.6237,
+      "step": 57
+    },
+    {
+      "epoch": 0.06617227609811752,
+      "grad_norm": 0.3196521894503953,
+      "learning_rate": 1.102661596958175e-05,
+      "loss": 0.6304,
+      "step": 58
+    },
+    {
+      "epoch": 0.06731317741015402,
+      "grad_norm": 0.31340766105880463,
+      "learning_rate": 1.1216730038022814e-05,
+      "loss": 0.6351,
+      "step": 59
+    },
+    {
+      "epoch": 0.06845407872219053,
+      "grad_norm": 0.2947895411613211,
+      "learning_rate": 1.1406844106463879e-05,
+      "loss": 0.6181,
+      "step": 60
+    },
+    {
+      "epoch": 0.06959498003422704,
+      "grad_norm": 0.3158134981188498,
+      "learning_rate": 1.1596958174904944e-05,
+      "loss": 0.6359,
+      "step": 61
+    },
+    {
+      "epoch": 0.07073588134626355,
+      "grad_norm": 0.2935907176760453,
+      "learning_rate": 1.1787072243346007e-05,
+      "loss": 0.6472,
+      "step": 62
+    },
+    {
+      "epoch": 0.07187678265830005,
+      "grad_norm": 0.2942195957589177,
+      "learning_rate": 1.1977186311787073e-05,
+      "loss": 0.6124,
+      "step": 63
+    },
+    {
+      "epoch": 0.07301768397033656,
+      "grad_norm": 0.31711851127434204,
+      "learning_rate": 1.2167300380228138e-05,
+      "loss": 0.6104,
+      "step": 64
+    },
+    {
+      "epoch": 0.07415858528237307,
+      "grad_norm": 0.27423197310063907,
+      "learning_rate": 1.2357414448669203e-05,
+      "loss": 0.6032,
+      "step": 65
+    },
+    {
+      "epoch": 0.07529948659440959,
+      "grad_norm": 0.32727870494920275,
+      "learning_rate": 1.2547528517110266e-05,
+      "loss": 0.5998,
+      "step": 66
+    },
+    {
+      "epoch": 0.0764403879064461,
+      "grad_norm": 0.268006371687303,
+      "learning_rate": 1.2737642585551332e-05,
+      "loss": 0.5977,
+      "step": 67
+    },
+    {
+      "epoch": 0.0775812892184826,
+      "grad_norm": 0.2691340842565907,
+      "learning_rate": 1.2927756653992395e-05,
+      "loss": 0.6017,
+      "step": 68
+    },
+    {
+      "epoch": 0.07872219053051911,
+      "grad_norm": 0.3215576928511998,
+      "learning_rate": 1.3117870722433462e-05,
+      "loss": 0.6111,
+      "step": 69
+    },
+    {
+      "epoch": 0.07986309184255562,
+      "grad_norm": 0.2524933357357243,
+      "learning_rate": 1.3307984790874526e-05,
+      "loss": 0.5856,
+      "step": 70
+    },
+    {
+      "epoch": 0.08100399315459213,
+      "grad_norm": 0.2976420763364573,
+      "learning_rate": 1.3498098859315589e-05,
+      "loss": 0.5745,
+      "step": 71
+    },
+    {
+      "epoch": 0.08214489446662863,
+      "grad_norm": 0.2721011406702133,
+      "learning_rate": 1.3688212927756656e-05,
+      "loss": 0.5956,
+      "step": 72
+    },
+    {
+      "epoch": 0.08328579577866514,
+      "grad_norm": 0.2602309725894347,
+      "learning_rate": 1.387832699619772e-05,
+      "loss": 0.5774,
+      "step": 73
+    },
+    {
+      "epoch": 0.08442669709070165,
+      "grad_norm": 0.2853820862748792,
+      "learning_rate": 1.4068441064638785e-05,
+      "loss": 0.5809,
+      "step": 74
+    },
+    {
+      "epoch": 0.08556759840273816,
+      "grad_norm": 0.27878528131213076,
+      "learning_rate": 1.4258555133079848e-05,
+      "loss": 0.5659,
+      "step": 75
+    },
+    {
+      "epoch": 0.08670849971477468,
+      "grad_norm": 0.2875430217961869,
+      "learning_rate": 1.4448669201520912e-05,
+      "loss": 0.5645,
+      "step": 76
+    },
+    {
+      "epoch": 0.08784940102681119,
+      "grad_norm": 0.3014267143867856,
+      "learning_rate": 1.4638783269961978e-05,
+      "loss": 0.5919,
+      "step": 77
+    },
+    {
+      "epoch": 0.0889903023388477,
+      "grad_norm": 0.2703052723297986,
+      "learning_rate": 1.4828897338403042e-05,
+      "loss": 0.5902,
+      "step": 78
+    },
+    {
+      "epoch": 0.0901312036508842,
+      "grad_norm": 0.303585390748195,
+      "learning_rate": 1.5019011406844107e-05,
+      "loss": 0.5772,
+      "step": 79
+    },
+    {
+      "epoch": 0.09127210496292071,
+      "grad_norm": 0.26905000684530966,
+      "learning_rate": 1.520912547528517e-05,
+      "loss": 0.5822,
+      "step": 80
+    },
+    {
+      "epoch": 0.09241300627495722,
+      "grad_norm": 0.28915135241524265,
+      "learning_rate": 1.5399239543726237e-05,
+      "loss": 0.5588,
+      "step": 81
+    },
+    {
+      "epoch": 0.09355390758699372,
+      "grad_norm": 0.29494358324783804,
+      "learning_rate": 1.55893536121673e-05,
+      "loss": 0.5909,
+      "step": 82
+    },
+    {
+      "epoch": 0.09469480889903023,
+      "grad_norm": 0.267960505652568,
+      "learning_rate": 1.5779467680608364e-05,
+      "loss": 0.5449,
+      "step": 83
+    },
+    {
+      "epoch": 0.09583571021106674,
+      "grad_norm": 0.29734284598046806,
+      "learning_rate": 1.596958174904943e-05,
+      "loss": 0.563,
+      "step": 84
+    },
+    {
+      "epoch": 0.09697661152310325,
+      "grad_norm": 0.2975566270568251,
+      "learning_rate": 1.6159695817490495e-05,
+      "loss": 0.5733,
+      "step": 85
+    },
+    {
+      "epoch": 0.09811751283513977,
+      "grad_norm": 0.3144496200827959,
+      "learning_rate": 1.634980988593156e-05,
+      "loss": 0.5393,
+      "step": 86
+    },
+    {
+      "epoch": 0.09925841414717627,
+      "grad_norm": 0.3439082551622902,
+      "learning_rate": 1.6539923954372625e-05,
+      "loss": 0.5801,
+      "step": 87
+    },
+    {
+      "epoch": 0.10039931545921278,
+      "grad_norm": 0.3272359309053644,
+      "learning_rate": 1.673003802281369e-05,
+      "loss": 0.595,
+      "step": 88
+    },
+    {
+      "epoch": 0.10154021677124929,
+      "grad_norm": 0.2806665105776375,
+      "learning_rate": 1.6920152091254756e-05,
+      "loss": 0.5539,
+      "step": 89
+    },
+    {
+      "epoch": 0.1026811180832858,
+      "grad_norm": 0.3583067110894645,
+      "learning_rate": 1.711026615969582e-05,
+      "loss": 0.5682,
+      "step": 90
+    },
+    {
+      "epoch": 0.1038220193953223,
+      "grad_norm": 0.3124598607304225,
+      "learning_rate": 1.7300380228136882e-05,
+      "loss": 0.5521,
+      "step": 91
+    },
+    {
+      "epoch": 0.10496292070735881,
+      "grad_norm": 0.30309929429920146,
+      "learning_rate": 1.7490494296577946e-05,
+      "loss": 0.5498,
+      "step": 92
+    },
+    {
+      "epoch": 0.10610382201939532,
+      "grad_norm": 0.3323899816536335,
+      "learning_rate": 1.7680608365019013e-05,
+      "loss": 0.5888,
+      "step": 93
+    },
+    {
+      "epoch": 0.10724472333143183,
+      "grad_norm": 0.2752652481956559,
+      "learning_rate": 1.7870722433460076e-05,
+      "loss": 0.5564,
+      "step": 94
+    },
+    {
+      "epoch": 0.10838562464346833,
+      "grad_norm": 0.31613768457039215,
+      "learning_rate": 1.806083650190114e-05,
+      "loss": 0.5526,
+      "step": 95
+    },
+    {
+      "epoch": 0.10952652595550486,
+      "grad_norm": 0.34846752285781013,
+      "learning_rate": 1.8250950570342207e-05,
+      "loss": 0.5605,
+      "step": 96
+    },
+    {
+      "epoch": 0.11066742726754136,
+      "grad_norm": 0.32422289472592464,
+      "learning_rate": 1.844106463878327e-05,
+      "loss": 0.5761,
+      "step": 97
+    },
+    {
+      "epoch": 0.11180832857957787,
+      "grad_norm": 0.2821646610453036,
+      "learning_rate": 1.8631178707224337e-05,
+      "loss": 0.538,
+      "step": 98
+    },
+    {
+      "epoch": 0.11294922989161438,
+      "grad_norm": 0.3363578624190279,
+      "learning_rate": 1.88212927756654e-05,
+      "loss": 0.5207,
+      "step": 99
+    },
+    {
+      "epoch": 0.11409013120365089,
+      "grad_norm": 0.33298699970382173,
+      "learning_rate": 1.9011406844106464e-05,
+      "loss": 0.5518,
+      "step": 100
+    },
+    {
+      "epoch": 0.11523103251568739,
+      "grad_norm": 0.34942931660045073,
+      "learning_rate": 1.920152091254753e-05,
+      "loss": 0.5472,
+      "step": 101
+    },
+    {
+      "epoch": 0.1163719338277239,
+      "grad_norm": 0.32312662063760633,
+      "learning_rate": 1.9391634980988594e-05,
+      "loss": 0.5673,
+      "step": 102
+    },
+    {
+      "epoch": 0.11751283513976041,
+      "grad_norm": 0.4126867118933167,
+      "learning_rate": 1.958174904942966e-05,
+      "loss": 0.5463,
+      "step": 103
+    },
+    {
+      "epoch": 0.11865373645179692,
+      "grad_norm": 0.32148007377648646,
+      "learning_rate": 1.9771863117870725e-05,
+      "loss": 0.5598,
+      "step": 104
+    },
+    {
+      "epoch": 0.11979463776383342,
+      "grad_norm": 0.3942522629458706,
+      "learning_rate": 1.9961977186311788e-05,
+      "loss": 0.5631,
+      "step": 105
+    },
+    {
+      "epoch": 0.12093553907586994,
+      "grad_norm": 0.3313247044983476,
+      "learning_rate": 2.0152091254752852e-05,
+      "loss": 0.5568,
+      "step": 106
+    },
+    {
+      "epoch": 0.12207644038790645,
+      "grad_norm": 0.4023164269134208,
+      "learning_rate": 2.0342205323193915e-05,
+      "loss": 0.5544,
+      "step": 107
+    },
+    {
+      "epoch": 0.12321734169994296,
+      "grad_norm": 0.3746260516214422,
+      "learning_rate": 2.0532319391634982e-05,
+      "loss": 0.5375,
+      "step": 108
+    },
+    {
+      "epoch": 0.12435824301197947,
+      "grad_norm": 0.31792875838737306,
+      "learning_rate": 2.0722433460076046e-05,
+      "loss": 0.5675,
+      "step": 109
+    },
+    {
+      "epoch": 0.12549914432401596,
+      "grad_norm": 0.36266589494136714,
+      "learning_rate": 2.0912547528517112e-05,
+      "loss": 0.5346,
+      "step": 110
+    },
+    {
+      "epoch": 0.1266400456360525,
+      "grad_norm": 0.35621216017329244,
+      "learning_rate": 2.1102661596958176e-05,
+      "loss": 0.5233,
+      "step": 111
+    },
+    {
+      "epoch": 0.127780946948089,
+      "grad_norm": 0.34992793415074863,
+      "learning_rate": 2.129277566539924e-05,
+      "loss": 0.5204,
+      "step": 112
+    },
+    {
+      "epoch": 0.1289218482601255,
+      "grad_norm": 0.43140776333761577,
+      "learning_rate": 2.1482889733840306e-05,
+      "loss": 0.5247,
+      "step": 113
+    },
+    {
+      "epoch": 0.13006274957216202,
+      "grad_norm": 0.30876595829241754,
+      "learning_rate": 2.167300380228137e-05,
+      "loss": 0.5366,
+      "step": 114
+    },
+    {
+      "epoch": 0.13120365088419852,
+      "grad_norm": 0.49393023217750726,
+      "learning_rate": 2.1863117870722437e-05,
+      "loss": 0.5543,
+      "step": 115
+    },
+    {
+      "epoch": 0.13234455219623503,
+      "grad_norm": 0.3174141199663179,
+      "learning_rate": 2.20532319391635e-05,
+      "loss": 0.5136,
+      "step": 116
+    },
+    {
+      "epoch": 0.13348545350827154,
+      "grad_norm": 0.43744514649905636,
+      "learning_rate": 2.2243346007604564e-05,
+      "loss": 0.562,
+      "step": 117
+    },
+    {
+      "epoch": 0.13462635482030805,
+      "grad_norm": 0.35548978897729616,
+      "learning_rate": 2.2433460076045627e-05,
+      "loss": 0.5324,
+      "step": 118
+    },
+    {
+      "epoch": 0.13576725613234455,
+      "grad_norm": 2.144134688500371,
+      "learning_rate": 2.262357414448669e-05,
+      "loss": 0.5417,
+      "step": 119
+    },
+    {
+      "epoch": 0.13690815744438106,
+      "grad_norm": 0.5181165703644828,
+      "learning_rate": 2.2813688212927758e-05,
+      "loss": 0.5439,
+      "step": 120
+    },
+    {
+      "epoch": 0.13804905875641757,
+      "grad_norm": 0.3562744580952984,
+      "learning_rate": 2.300380228136882e-05,
+      "loss": 0.4994,
+      "step": 121
+    },
+    {
+      "epoch": 0.13918996006845408,
+      "grad_norm": 0.47803128689037555,
+      "learning_rate": 2.3193916349809888e-05,
+      "loss": 0.5225,
+      "step": 122
+    },
+    {
+      "epoch": 0.14033086138049058,
+      "grad_norm": 0.43179878456026727,
+      "learning_rate": 2.338403041825095e-05,
+      "loss": 0.5173,
+      "step": 123
+    },
+    {
+      "epoch": 0.1414717626925271,
+      "grad_norm": 0.48797324922902563,
+      "learning_rate": 2.3574144486692015e-05,
+      "loss": 0.5331,
+      "step": 124
+    },
+    {
+      "epoch": 0.1426126640045636,
+      "grad_norm": 0.3713551353610012,
+      "learning_rate": 2.3764258555133082e-05,
+      "loss": 0.5399,
+      "step": 125
+    },
+    {
+      "epoch": 0.1437535653166001,
+      "grad_norm": 0.5169644283335567,
+      "learning_rate": 2.3954372623574145e-05,
+      "loss": 0.5357,
+      "step": 126
+    },
+    {
+      "epoch": 0.14489446662863661,
+      "grad_norm": 0.31629506927905415,
+      "learning_rate": 2.4144486692015212e-05,
+      "loss": 0.5239,
+      "step": 127
+    },
+    {
+      "epoch": 0.14603536794067312,
+      "grad_norm": 0.5278145043561342,
+      "learning_rate": 2.4334600760456276e-05,
+      "loss": 0.528,
+      "step": 128
+    },
+    {
+      "epoch": 0.14717626925270963,
+      "grad_norm": 0.3315425028798385,
+      "learning_rate": 2.452471482889734e-05,
+      "loss": 0.5495,
+      "step": 129
+    },
+    {
+      "epoch": 0.14831717056474614,
+      "grad_norm": 0.517192805192663,
+      "learning_rate": 2.4714828897338406e-05,
+      "loss": 0.5083,
+      "step": 130
+    },
+    {
+      "epoch": 0.14945807187678267,
+      "grad_norm": 0.36358628344754546,
+      "learning_rate": 2.490494296577947e-05,
+      "loss": 0.5132,
+      "step": 131
+    },
+    {
+      "epoch": 0.15059897318881918,
+      "grad_norm": 0.4243392518096659,
+      "learning_rate": 2.5095057034220533e-05,
+      "loss": 0.5239,
+      "step": 132
+    },
+    {
+      "epoch": 0.1517398745008557,
+      "grad_norm": 0.42938450026993014,
+      "learning_rate": 2.5285171102661596e-05,
+      "loss": 0.5222,
+      "step": 133
+    },
+    {
+      "epoch": 0.1528807758128922,
+      "grad_norm": 0.3341842284068171,
+      "learning_rate": 2.5475285171102663e-05,
+      "loss": 0.5186,
+      "step": 134
+    },
+    {
+      "epoch": 0.1540216771249287,
+      "grad_norm": 0.4541463176730268,
+      "learning_rate": 2.5665399239543723e-05,
+      "loss": 0.5455,
+      "step": 135
+    },
+    {
+      "epoch": 0.1551625784369652,
+      "grad_norm": 0.30644393922157925,
+      "learning_rate": 2.585551330798479e-05,
+      "loss": 0.5219,
+      "step": 136
+    },
+    {
+      "epoch": 0.15630347974900172,
+      "grad_norm": 0.4150075147103399,
+      "learning_rate": 2.6045627376425857e-05,
+      "loss": 0.516,
+      "step": 137
+    },
+    {
+      "epoch": 0.15744438106103822,
+      "grad_norm": 0.36139117065630944,
+      "learning_rate": 2.6235741444866924e-05,
+      "loss": 0.5293,
+      "step": 138
+    },
+    {
+      "epoch": 0.15858528237307473,
+      "grad_norm": 0.3803593895867164,
+      "learning_rate": 2.6425855513307984e-05,
+      "loss": 0.5197,
+      "step": 139
+    },
+    {
+      "epoch": 0.15972618368511124,
+      "grad_norm": 0.32396824952520176,
+      "learning_rate": 2.661596958174905e-05,
+      "loss": 0.5244,
+      "step": 140
+    },
+    {
+      "epoch": 0.16086708499714775,
+      "grad_norm": 0.3774873367059754,
+      "learning_rate": 2.6806083650190118e-05,
+      "loss": 0.5154,
+      "step": 141
+    },
+    {
+      "epoch": 0.16200798630918425,
+      "grad_norm": 0.34633191929954577,
+      "learning_rate": 2.6996197718631178e-05,
+      "loss": 0.5249,
+      "step": 142
+    },
+    {
+      "epoch": 0.16314888762122076,
+      "grad_norm": 0.3894815335476097,
+      "learning_rate": 2.7186311787072245e-05,
+      "loss": 0.5198,
+      "step": 143
+    },
+    {
+      "epoch": 0.16428978893325727,
+      "grad_norm": 0.36444762291566796,
+      "learning_rate": 2.7376425855513312e-05,
+      "loss": 0.5015,
+      "step": 144
+    },
+    {
+      "epoch": 0.16543069024529378,
+      "grad_norm": 0.423060528009577,
+      "learning_rate": 2.7566539923954375e-05,
+      "loss": 0.5218,
+      "step": 145
+    },
+    {
+      "epoch": 0.16657159155733028,
+      "grad_norm": 0.35615661027370654,
+      "learning_rate": 2.775665399239544e-05,
+      "loss": 0.5237,
+      "step": 146
+    },
+    {
+      "epoch": 0.1677124928693668,
+      "grad_norm": 0.46050626206133344,
+      "learning_rate": 2.7946768060836502e-05,
+      "loss": 0.5398,
+      "step": 147
+    },
+    {
+      "epoch": 0.1688533941814033,
+      "grad_norm": 0.36853471772573493,
+      "learning_rate": 2.813688212927757e-05,
+      "loss": 0.5095,
+      "step": 148
+    },
+    {
+      "epoch": 0.1699942954934398,
+      "grad_norm": 0.4932478981324432,
+      "learning_rate": 2.832699619771863e-05,
+      "loss": 0.5226,
+      "step": 149
+    },
+    {
+      "epoch": 0.17113519680547631,
+      "grad_norm": 0.37242675018432586,
+      "learning_rate": 2.8517110266159696e-05,
+      "loss": 0.5551,
+      "step": 150
+    },
+    {
+      "epoch": 0.17227609811751282,
+      "grad_norm": 0.37826249255692296,
+      "learning_rate": 2.8707224334600763e-05,
+      "loss": 0.4902,
+      "step": 151
+    },
+    {
+      "epoch": 0.17341699942954936,
+      "grad_norm": 0.39728450228224194,
+      "learning_rate": 2.8897338403041823e-05,
+      "loss": 0.5209,
+      "step": 152
+    },
+    {
+      "epoch": 0.17455790074158586,
+      "grad_norm": 0.40470187502226895,
+      "learning_rate": 2.908745247148289e-05,
+      "loss": 0.5359,
+      "step": 153
+    },
+    {
+      "epoch": 0.17569880205362237,
+      "grad_norm": 0.40620197495649407,
+      "learning_rate": 2.9277566539923957e-05,
+      "loss": 0.5242,
+      "step": 154
+    },
+    {
+      "epoch": 0.17683970336565888,
+      "grad_norm": 0.3727566620054441,
+      "learning_rate": 2.9467680608365024e-05,
+      "loss": 0.5055,
+      "step": 155
+    },
+    {
+      "epoch": 0.1779806046776954,
+      "grad_norm": 0.4398128842105725,
+      "learning_rate": 2.9657794676806084e-05,
+      "loss": 0.5408,
+      "step": 156
+    },
+    {
+      "epoch": 0.1791215059897319,
+      "grad_norm": 0.38490204332241523,
+      "learning_rate": 2.984790874524715e-05,
+      "loss": 0.5156,
+      "step": 157
+    },
+    {
+      "epoch": 0.1802624073017684,
+      "grad_norm": 0.4517639869302986,
+      "learning_rate": 3.0038022813688214e-05,
+      "loss": 0.528,
+      "step": 158
+    },
+    {
+      "epoch": 0.1814033086138049,
+      "grad_norm": 0.4954944774014952,
+      "learning_rate": 3.0228136882129278e-05,
+      "loss": 0.508,
+      "step": 159
+    },
+    {
+      "epoch": 0.18254420992584142,
+      "grad_norm": 0.3748332582197787,
+      "learning_rate": 3.041825095057034e-05,
+      "loss": 0.5167,
+      "step": 160
+    },
+    {
+      "epoch": 0.18368511123787792,
+      "grad_norm": 0.40942567421244364,
+      "learning_rate": 3.060836501901141e-05,
+      "loss": 0.4919,
+      "step": 161
+    },
+    {
+      "epoch": 0.18482601254991443,
+      "grad_norm": 0.37179984547777867,
+      "learning_rate": 3.0798479087452475e-05,
+      "loss": 0.5277,
+      "step": 162
+    },
+    {
+      "epoch": 0.18596691386195094,
+      "grad_norm": 0.4427813850278362,
+      "learning_rate": 3.098859315589354e-05,
+      "loss": 0.5084,
+      "step": 163
+    },
+    {
+      "epoch": 0.18710781517398745,
+      "grad_norm": 0.3953797053294692,
+      "learning_rate": 3.11787072243346e-05,
+      "loss": 0.4996,
+      "step": 164
+    },
+    {
+      "epoch": 0.18824871648602395,
+      "grad_norm": 0.4498502562903012,
+      "learning_rate": 3.1368821292775665e-05,
+      "loss": 0.5049,
+      "step": 165
+    },
+    {
+      "epoch": 0.18938961779806046,
+      "grad_norm": 0.4108040639817359,
+      "learning_rate": 3.155893536121673e-05,
+      "loss": 0.5328,
+      "step": 166
+    },
+    {
+      "epoch": 0.19053051911009697,
+      "grad_norm": 0.5410412527841703,
+      "learning_rate": 3.174904942965779e-05,
+      "loss": 0.5119,
+      "step": 167
+    },
+    {
+      "epoch": 0.19167142042213348,
+      "grad_norm": 0.4344065614283204,
+      "learning_rate": 3.193916349809886e-05,
+      "loss": 0.4837,
+      "step": 168
+    },
+    {
+      "epoch": 0.19281232173416998,
+      "grad_norm": 0.6626492419165108,
+      "learning_rate": 3.2129277566539926e-05,
+      "loss": 0.5132,
+      "step": 169
+    },
+    {
+      "epoch": 0.1939532230462065,
+      "grad_norm": 0.4038123054819114,
+      "learning_rate": 3.231939163498099e-05,
+      "loss": 0.5067,
+      "step": 170
+    },
+    {
+      "epoch": 0.195094124358243,
+      "grad_norm": 0.5636456551357991,
+      "learning_rate": 3.250950570342205e-05,
+      "loss": 0.4673,
+      "step": 171
+    },
+    {
+      "epoch": 0.19623502567027953,
+      "grad_norm": 0.5205483430986917,
+      "learning_rate": 3.269961977186312e-05,
+      "loss": 0.5265,
+      "step": 172
+    },
+    {
+      "epoch": 0.19737592698231604,
+      "grad_norm": 0.48211911743840075,
+      "learning_rate": 3.288973384030418e-05,
+      "loss": 0.507,
+      "step": 173
+    },
+    {
+      "epoch": 0.19851682829435255,
+      "grad_norm": 0.47908754790447056,
+      "learning_rate": 3.307984790874525e-05,
+      "loss": 0.5165,
+      "step": 174
+    },
+    {
+      "epoch": 0.19965772960638906,
+      "grad_norm": 0.4230015023005651,
+      "learning_rate": 3.3269961977186314e-05,
+      "loss": 0.5102,
+      "step": 175
+    },
+    {
+      "epoch": 0.20079863091842556,
+      "grad_norm": 0.5427071406265588,
+      "learning_rate": 3.346007604562738e-05,
+      "loss": 0.5054,
+      "step": 176
+    },
+    {
+      "epoch": 0.20193953223046207,
+      "grad_norm": 0.5099286314808494,
+      "learning_rate": 3.365019011406844e-05,
+      "loss": 0.5222,
+      "step": 177
+    },
+    {
+      "epoch": 0.20308043354249858,
+      "grad_norm": 0.6159657751770209,
+      "learning_rate": 3.384030418250951e-05,
+      "loss": 0.5431,
+      "step": 178
+    },
+    {
+      "epoch": 0.20422133485453509,
+      "grad_norm": 0.7193603225951907,
+      "learning_rate": 3.4030418250950574e-05,
+      "loss": 0.4963,
+      "step": 179
+    },
+    {
+      "epoch": 0.2053622361665716,
+      "grad_norm": 0.45033133846022033,
+      "learning_rate": 3.422053231939164e-05,
+      "loss": 0.5014,
+      "step": 180
+    },
+    {
+      "epoch": 0.2065031374786081,
+      "grad_norm": 0.7111256209057891,
+      "learning_rate": 3.44106463878327e-05,
+      "loss": 0.5059,
+      "step": 181
+    },
+    {
+      "epoch": 0.2076440387906446,
+      "grad_norm": 0.6493203071076293,
+      "learning_rate": 3.4600760456273765e-05,
+      "loss": 0.4941,
+      "step": 182
+    },
+    {
+      "epoch": 0.20878494010268112,
+      "grad_norm": 0.5283000924211072,
+      "learning_rate": 3.479087452471483e-05,
+      "loss": 0.5177,
+      "step": 183
+    },
+    {
+      "epoch": 0.20992584141471762,
+      "grad_norm": 0.653191879689384,
+      "learning_rate": 3.498098859315589e-05,
+      "loss": 0.513,
+      "step": 184
+    },
+    {
+      "epoch": 0.21106674272675413,
+      "grad_norm": 0.7341746099628614,
+      "learning_rate": 3.517110266159696e-05,
+      "loss": 0.4987,
+      "step": 185
+    },
+    {
+      "epoch": 0.21220764403879064,
+      "grad_norm": 0.4263764021054984,
+      "learning_rate": 3.5361216730038026e-05,
+      "loss": 0.4939,
+      "step": 186
+    },
+    {
+      "epoch": 0.21334854535082715,
+      "grad_norm": 0.6255115057204136,
+      "learning_rate": 3.555133079847909e-05,
+      "loss": 0.479,
+      "step": 187
+    },
+    {
+      "epoch": 0.21448944666286365,
+      "grad_norm": 0.5199063260718685,
+      "learning_rate": 3.574144486692015e-05,
+      "loss": 0.4973,
+      "step": 188
+    },
+    {
+      "epoch": 0.21563034797490016,
+      "grad_norm": 0.4976122263753569,
+      "learning_rate": 3.593155893536122e-05,
+      "loss": 0.4823,
+      "step": 189
+    },
+    {
+      "epoch": 0.21677124928693667,
+      "grad_norm": 0.5304914545222749,
+      "learning_rate": 3.612167300380228e-05,
+      "loss": 0.499,
+      "step": 190
+    },
+    {
+      "epoch": 0.21791215059897318,
+      "grad_norm": 0.4108657755861063,
+      "learning_rate": 3.631178707224335e-05,
+      "loss": 0.4968,
+      "step": 191
+    },
+    {
+      "epoch": 0.2190530519110097,
+      "grad_norm": 0.5596852694323317,
+      "learning_rate": 3.6501901140684413e-05,
+      "loss": 0.5001,
+      "step": 192
+    },
+    {
+      "epoch": 0.22019395322304622,
+      "grad_norm": 0.502896811074524,
+      "learning_rate": 3.669201520912548e-05,
+      "loss": 0.4846,
+      "step": 193
+    },
+    {
+      "epoch": 0.22133485453508273,
+      "grad_norm": 0.5091731389286583,
+      "learning_rate": 3.688212927756654e-05,
+      "loss": 0.5224,
+      "step": 194
+    },
+    {
+      "epoch": 0.22247575584711923,
+      "grad_norm": 0.5274679963561089,
+      "learning_rate": 3.7072243346007604e-05,
+      "loss": 0.5276,
+      "step": 195
+    },
+    {
+      "epoch": 0.22361665715915574,
+      "grad_norm": 0.3898202631701823,
+      "learning_rate": 3.7262357414448674e-05,
+      "loss": 0.4836,
+      "step": 196
+    },
+    {
+      "epoch": 0.22475755847119225,
+      "grad_norm": 0.42129635358682016,
+      "learning_rate": 3.745247148288973e-05,
+      "loss": 0.4678,
+      "step": 197
+    },
+    {
+      "epoch": 0.22589845978322876,
+      "grad_norm": 0.4618057419115625,
+      "learning_rate": 3.76425855513308e-05,
+      "loss": 0.5342,
+      "step": 198
+    },
+    {
+      "epoch": 0.22703936109526526,
+      "grad_norm": 0.4381869533215169,
+      "learning_rate": 3.7832699619771865e-05,
+      "loss": 0.4671,
+      "step": 199
+    },
+    {
+      "epoch": 0.22818026240730177,
+      "grad_norm": 0.5440837221051571,
+      "learning_rate": 3.802281368821293e-05,
+      "loss": 0.5059,
+      "step": 200
+    },
+    {
+      "epoch": 0.22932116371933828,
+      "grad_norm": 0.5334139765821885,
+      "learning_rate": 3.821292775665399e-05,
+      "loss": 0.5145,
+      "step": 201
+    },
+    {
+      "epoch": 0.23046206503137479,
+      "grad_norm": 0.3732495424879209,
+      "learning_rate": 3.840304182509506e-05,
+      "loss": 0.4838,
+      "step": 202
+    },
+    {
+      "epoch": 0.2316029663434113,
+      "grad_norm": 0.5988989154670815,
+      "learning_rate": 3.8593155893536125e-05,
+      "loss": 0.5037,
+      "step": 203
+    },
+    {
+      "epoch": 0.2327438676554478,
+      "grad_norm": 0.5200393200147204,
+      "learning_rate": 3.878326996197719e-05,
+      "loss": 0.4745,
+      "step": 204
+    },
+    {
+      "epoch": 0.2338847689674843,
+      "grad_norm": 0.4721161101938097,
+      "learning_rate": 3.897338403041825e-05,
+      "loss": 0.5197,
+      "step": 205
+    },
+    {
+      "epoch": 0.23502567027952082,
+      "grad_norm": 0.5749531215762707,
+      "learning_rate": 3.916349809885932e-05,
+      "loss": 0.4825,
+      "step": 206
+    },
+    {
+      "epoch": 0.23616657159155732,
+      "grad_norm": 0.40149594230609625,
+      "learning_rate": 3.935361216730038e-05,
+      "loss": 0.4597,
+      "step": 207
+    },
+    {
+      "epoch": 0.23730747290359383,
+      "grad_norm": 0.5486586047623457,
+      "learning_rate": 3.954372623574145e-05,
+      "loss": 0.4847,
+      "step": 208
+    },
+    {
+      "epoch": 0.23844837421563034,
+      "grad_norm": 0.5820602383783187,
+      "learning_rate": 3.973384030418251e-05,
+      "loss": 0.4877,
+      "step": 209
+    },
+    {
+      "epoch": 0.23958927552766685,
+      "grad_norm": 0.5489212619030648,
+      "learning_rate": 3.9923954372623577e-05,
+      "loss": 0.5013,
+      "step": 210
+    },
+    {
+      "epoch": 0.24073017683970335,
+      "grad_norm": 0.6147528782803471,
+      "learning_rate": 4.011406844106464e-05,
+      "loss": 0.4988,
+      "step": 211
+    },
+    {
+      "epoch": 0.2418710781517399,
+      "grad_norm": 0.5588776012384556,
+      "learning_rate": 4.0304182509505703e-05,
+      "loss": 0.5189,
+      "step": 212
+    },
+    {
+      "epoch": 0.2430119794637764,
+      "grad_norm": 0.4790644696477516,
+      "learning_rate": 4.0494296577946774e-05,
+      "loss": 0.5036,
+      "step": 213
+    },
+    {
+      "epoch": 0.2441528807758129,
+      "grad_norm": 0.4833204786443911,
+      "learning_rate": 4.068441064638783e-05,
+      "loss": 0.476,
+      "step": 214
+    },
+    {
+      "epoch": 0.2452937820878494,
+      "grad_norm": 0.47845223722709473,
+      "learning_rate": 4.08745247148289e-05,
+      "loss": 0.4965,
+      "step": 215
+    },
+    {
+      "epoch": 0.24643468339988592,
+      "grad_norm": 0.4481560825138665,
+      "learning_rate": 4.1064638783269964e-05,
+      "loss": 0.5017,
+      "step": 216
+    },
+    {
+      "epoch": 0.24757558471192243,
+      "grad_norm": 0.5086810495955193,
+      "learning_rate": 4.125475285171103e-05,
+      "loss": 0.4969,
+      "step": 217
+    },
+    {
+      "epoch": 0.24871648602395893,
+      "grad_norm": 0.5343756618348462,
+      "learning_rate": 4.144486692015209e-05,
+      "loss": 0.4821,
+      "step": 218
+    },
+    {
+      "epoch": 0.24985738733599544,
+      "grad_norm": 0.401711278686693,
+      "learning_rate": 4.163498098859316e-05,
+      "loss": 0.4985,
+      "step": 219
+    },
+    {
+      "epoch": 0.2509982886480319,
+      "grad_norm": 0.5346993959856637,
+      "learning_rate": 4.1825095057034225e-05,
+      "loss": 0.5037,
+      "step": 220
+    },
+    {
+      "epoch": 0.25213918996006845,
+      "grad_norm": 0.46140846706399685,
+      "learning_rate": 4.201520912547529e-05,
+      "loss": 0.4856,
+      "step": 221
+    },
+    {
+      "epoch": 0.253280091272105,
+      "grad_norm": 0.4839447665729109,
+      "learning_rate": 4.220532319391635e-05,
+      "loss": 0.4842,
+      "step": 222
+    },
+    {
+      "epoch": 0.25442099258414147,
+      "grad_norm": 0.3697667503798369,
+      "learning_rate": 4.2395437262357415e-05,
+      "loss": 0.4873,
+      "step": 223
+    },
+    {
+      "epoch": 0.255561893896178,
+      "grad_norm": 0.4427985621603716,
+      "learning_rate": 4.258555133079848e-05,
+      "loss": 0.4835,
+      "step": 224
+    },
+    {
+      "epoch": 0.2567027952082145,
+      "grad_norm": 0.39597098010473725,
+      "learning_rate": 4.277566539923954e-05,
+      "loss": 0.5005,
+      "step": 225
+    },
+    {
+      "epoch": 0.257843696520251,
+      "grad_norm": 0.4176905956944042,
+      "learning_rate": 4.296577946768061e-05,
+      "loss": 0.4986,
+      "step": 226
+    },
+    {
+      "epoch": 0.2589845978322875,
+      "grad_norm": 0.42126473910651496,
+      "learning_rate": 4.3155893536121676e-05,
+      "loss": 0.4788,
+      "step": 227
+    },
+    {
+      "epoch": 0.26012549914432403,
+      "grad_norm": 0.41188830182365,
+      "learning_rate": 4.334600760456274e-05,
+      "loss": 0.4925,
+      "step": 228
+    },
+    {
+      "epoch": 0.2612664004563605,
+      "grad_norm": 0.457545396221541,
+      "learning_rate": 4.35361216730038e-05,
+      "loss": 0.5031,
+      "step": 229
+    },
+    {
+      "epoch": 0.26240730176839705,
+      "grad_norm": 0.4116499547651067,
+      "learning_rate": 4.3726235741444873e-05,
+      "loss": 0.5025,
+      "step": 230
+    },
+    {
+      "epoch": 0.26354820308043353,
+      "grad_norm": 0.4760209285221355,
+      "learning_rate": 4.391634980988593e-05,
+      "loss": 0.4959,
+      "step": 231
+    },
+    {
+      "epoch": 0.26468910439247006,
+      "grad_norm": 0.3542120628745739,
+      "learning_rate": 4.4106463878327e-05,
+      "loss": 0.4936,
+      "step": 232
+    },
+    {
+      "epoch": 0.26583000570450654,
+      "grad_norm": 0.422037077717719,
+      "learning_rate": 4.4296577946768064e-05,
+      "loss": 0.4894,
+      "step": 233
+    },
+    {
+      "epoch": 0.2669709070165431,
+      "grad_norm": 0.3914300693793639,
+      "learning_rate": 4.448669201520913e-05,
+      "loss": 0.4931,
+      "step": 234
+    },
+    {
+      "epoch": 0.26811180832857956,
+      "grad_norm": 0.4320021869729976,
+      "learning_rate": 4.467680608365019e-05,
+      "loss": 0.4791,
+      "step": 235
+    },
+    {
+      "epoch": 0.2692527096406161,
+      "grad_norm": 0.4175577806166712,
+      "learning_rate": 4.4866920152091254e-05,
+      "loss": 0.4906,
+      "step": 236
+    },
+    {
+      "epoch": 0.2703936109526526,
+      "grad_norm": 0.41222578442948693,
+      "learning_rate": 4.5057034220532325e-05,
+      "loss": 0.4694,
+      "step": 237
+    },
+    {
+      "epoch": 0.2715345122646891,
+      "grad_norm": 0.43338867254807784,
+      "learning_rate": 4.524714828897338e-05,
+      "loss": 0.4922,
+      "step": 238
+    },
+    {
+      "epoch": 0.2726754135767256,
+      "grad_norm": 0.42813684712893474,
+      "learning_rate": 4.543726235741445e-05,
+      "loss": 0.4943,
+      "step": 239
+    },
+    {
+      "epoch": 0.2738163148887621,
+      "grad_norm": 0.4819004155531821,
+      "learning_rate": 4.5627376425855515e-05,
+      "loss": 0.511,
+      "step": 240
+    },
+    {
+      "epoch": 0.2749572162007986,
+      "grad_norm": 0.38917770762771053,
+      "learning_rate": 4.581749049429658e-05,
+      "loss": 0.4731,
+      "step": 241
+    },
+    {
+      "epoch": 0.27609811751283514,
+      "grad_norm": 0.4212713190618949,
+      "learning_rate": 4.600760456273764e-05,
+      "loss": 0.486,
+      "step": 242
+    },
+    {
+      "epoch": 0.2772390188248717,
+      "grad_norm": 0.48210207775472946,
+      "learning_rate": 4.619771863117871e-05,
+      "loss": 0.4701,
+      "step": 243
+    },
+    {
+      "epoch": 0.27837992013690815,
+      "grad_norm": 0.4440172173976686,
+      "learning_rate": 4.6387832699619776e-05,
+      "loss": 0.4807,
+      "step": 244
+    },
+    {
+      "epoch": 0.2795208214489447,
+      "grad_norm": 0.5697960117028452,
+      "learning_rate": 4.657794676806084e-05,
+      "loss": 0.477,
+      "step": 245
+    },
+    {
+      "epoch": 0.28066172276098117,
+      "grad_norm": 0.39282938347691765,
+      "learning_rate": 4.67680608365019e-05,
+      "loss": 0.4653,
+      "step": 246
+    },
+    {
+      "epoch": 0.2818026240730177,
+      "grad_norm": 0.49426639160977187,
+      "learning_rate": 4.695817490494297e-05,
+      "loss": 0.4757,
+      "step": 247
+    },
+    {
+      "epoch": 0.2829435253850542,
+      "grad_norm": 0.5392025850210608,
+      "learning_rate": 4.714828897338403e-05,
+      "loss": 0.489,
+      "step": 248
+    },
+    {
+      "epoch": 0.2840844266970907,
+      "grad_norm": 0.6749892159090881,
+      "learning_rate": 4.73384030418251e-05,
+      "loss": 0.4954,
+      "step": 249
+    },
+    {
+      "epoch": 0.2852253280091272,
+      "grad_norm": 0.45560583490235546,
+      "learning_rate": 4.7528517110266163e-05,
+      "loss": 0.4821,
+      "step": 250
+    },
+    {
+      "epoch": 0.28636622932116373,
+      "grad_norm": 0.5214154421651995,
+      "learning_rate": 4.771863117870723e-05,
+      "loss": 0.4849,
+      "step": 251
+    },
+    {
+      "epoch": 0.2875071306332002,
+      "grad_norm": 0.5633817781068714,
+      "learning_rate": 4.790874524714829e-05,
+      "loss": 0.4446,
+      "step": 252
+    },
+    {
+      "epoch": 0.28864803194523675,
+      "grad_norm": 0.48784588145190155,
+      "learning_rate": 4.8098859315589354e-05,
+      "loss": 0.4736,
+      "step": 253
+    },
+    {
+      "epoch": 0.28978893325727323,
+      "grad_norm": 0.5919296100354379,
+      "learning_rate": 4.8288973384030424e-05,
+      "loss": 0.4792,
+      "step": 254
+    },
+    {
+      "epoch": 0.29092983456930976,
+      "grad_norm": 0.5264307123179902,
+      "learning_rate": 4.847908745247148e-05,
+      "loss": 0.5007,
+      "step": 255
+    },
+    {
+      "epoch": 0.29207073588134624,
+      "grad_norm": 0.5852556002815498,
+      "learning_rate": 4.866920152091255e-05,
+      "loss": 0.4684,
+      "step": 256
+    },
+    {
+      "epoch": 0.2932116371933828,
+      "grad_norm": 0.5604440642451641,
+      "learning_rate": 4.8859315589353615e-05,
+      "loss": 0.4937,
+      "step": 257
+    },
+    {
+      "epoch": 0.29435253850541926,
+      "grad_norm": 0.5083827458606978,
+      "learning_rate": 4.904942965779468e-05,
+      "loss": 0.4761,
+      "step": 258
+    },
+    {
+      "epoch": 0.2954934398174558,
+      "grad_norm": 0.6035414465096012,
+      "learning_rate": 4.923954372623574e-05,
+      "loss": 0.4699,
+      "step": 259
+    },
+    {
+      "epoch": 0.2966343411294923,
+      "grad_norm": 0.45520013209105414,
+      "learning_rate": 4.942965779467681e-05,
+      "loss": 0.4703,
+      "step": 260
+    },
+    {
+      "epoch": 0.2977752424415288,
+      "grad_norm": 0.4981914121296809,
+      "learning_rate": 4.9619771863117875e-05,
+      "loss": 0.4711,
+      "step": 261
+    },
+    {
+      "epoch": 0.29891614375356534,
+      "grad_norm": 0.4843901405929016,
+      "learning_rate": 4.980988593155894e-05,
+      "loss": 0.49,
+      "step": 262
+    },
+    {
+      "epoch": 0.3000570450656018,
+      "grad_norm": 0.519532365295793,
+      "learning_rate": 5e-05,
+      "loss": 0.4549,
+      "step": 263
+    },
+    {
+      "epoch": 0.30119794637763836,
+      "grad_norm": 0.5085787597831526,
+      "learning_rate": 4.997885835095137e-05,
+      "loss": 0.4959,
+      "step": 264
+    },
+    {
+      "epoch": 0.30233884768967484,
+      "grad_norm": 0.5063823188387865,
+      "learning_rate": 4.995771670190275e-05,
+      "loss": 0.4817,
+      "step": 265
+    },
+    {
+      "epoch": 0.3034797490017114,
+      "grad_norm": 0.5183643098113566,
+      "learning_rate": 4.993657505285412e-05,
+      "loss": 0.4859,
+      "step": 266
+    },
+    {
+      "epoch": 0.30462065031374785,
+      "grad_norm": 0.726951443078331,
+      "learning_rate": 4.99154334038055e-05,
+      "loss": 0.508,
+      "step": 267
+    },
+    {
+      "epoch": 0.3057615516257844,
+      "grad_norm": 0.5979308391864225,
+      "learning_rate": 4.989429175475687e-05,
+      "loss": 0.4751,
+      "step": 268
+    },
+    {
+      "epoch": 0.30690245293782087,
+      "grad_norm": 0.5642531616180014,
+      "learning_rate": 4.987315010570825e-05,
+      "loss": 0.4618,
+      "step": 269
+    },
+    {
+      "epoch": 0.3080433542498574,
+      "grad_norm": 0.4528790393447231,
+      "learning_rate": 4.9852008456659624e-05,
+      "loss": 0.5006,
+      "step": 270
+    },
+    {
+      "epoch": 0.3091842555618939,
+      "grad_norm": 0.6608468146942221,
+      "learning_rate": 4.9830866807611e-05,
+      "loss": 0.4919,
+      "step": 271
+    },
+    {
+      "epoch": 0.3103251568739304,
+      "grad_norm": 0.5534909136528965,
+      "learning_rate": 4.980972515856237e-05,
+      "loss": 0.4787,
+      "step": 272
+    },
+    {
+      "epoch": 0.3114660581859669,
+      "grad_norm": 0.4855554939031897,
+      "learning_rate": 4.978858350951374e-05,
+      "loss": 0.4854,
+      "step": 273
+    },
+    {
+      "epoch": 0.31260695949800343,
+      "grad_norm": 0.6042118406817583,
+      "learning_rate": 4.976744186046512e-05,
+      "loss": 0.5032,
+      "step": 274
+    },
+    {
+      "epoch": 0.3137478608100399,
+      "grad_norm": 0.6301236427577049,
+      "learning_rate": 4.974630021141649e-05,
+      "loss": 0.504,
+      "step": 275
+    },
+    {
+      "epoch": 0.31488876212207645,
+      "grad_norm": 0.43962836238867853,
+      "learning_rate": 4.972515856236787e-05,
+      "loss": 0.4693,
+      "step": 276
+    },
+    {
+      "epoch": 0.31602966343411293,
+      "grad_norm": 0.6865074425382169,
+      "learning_rate": 4.970401691331924e-05,
+      "loss": 0.4984,
+      "step": 277
+    },
+    {
+      "epoch": 0.31717056474614946,
+      "grad_norm": 0.5707037542530051,
+      "learning_rate": 4.968287526427062e-05,
+      "loss": 0.4728,
+      "step": 278
+    },
+    {
+      "epoch": 0.31831146605818594,
+      "grad_norm": 0.48013638357531163,
+      "learning_rate": 4.966173361522199e-05,
+      "loss": 0.4897,
+      "step": 279
+    },
+    {
+      "epoch": 0.3194523673702225,
+      "grad_norm": 0.5553808724683528,
+      "learning_rate": 4.9640591966173365e-05,
+      "loss": 0.47,
+      "step": 280
+    },
+    {
+      "epoch": 0.32059326868225896,
+      "grad_norm": 0.6673573258804002,
+      "learning_rate": 4.9619450317124736e-05,
+      "loss": 0.4742,
+      "step": 281
+    },
+    {
+      "epoch": 0.3217341699942955,
+      "grad_norm": 0.4777963168027366,
+      "learning_rate": 4.959830866807611e-05,
+      "loss": 0.5114,
+      "step": 282
+    },
+    {
+      "epoch": 0.32287507130633203,
+      "grad_norm": 0.47238719588748307,
+      "learning_rate": 4.957716701902749e-05,
+      "loss": 0.4572,
+      "step": 283
+    },
+    {
+      "epoch": 0.3240159726183685,
+      "grad_norm": 0.5109087160306456,
+      "learning_rate": 4.955602536997886e-05,
+      "loss": 0.5017,
+      "step": 284
+    },
+    {
+      "epoch": 0.32515687393040504,
+      "grad_norm": 0.5913296622786067,
+      "learning_rate": 4.953488372093024e-05,
+      "loss": 0.4899,
+      "step": 285
+    },
+    {
+      "epoch": 0.3262977752424415,
+      "grad_norm": 0.4779423075939576,
+      "learning_rate": 4.951374207188161e-05,
+      "loss": 0.4728,
+      "step": 286
+    },
+    {
+      "epoch": 0.32743867655447806,
+      "grad_norm": 0.46538463444466194,
+      "learning_rate": 4.949260042283299e-05,
+      "loss": 0.4952,
+      "step": 287
+    },
+    {
+      "epoch": 0.32857957786651454,
+      "grad_norm": 0.595981118622069,
+      "learning_rate": 4.947145877378436e-05,
+      "loss": 0.4871,
+      "step": 288
+    },
+    {
+      "epoch": 0.3297204791785511,
+      "grad_norm": 0.4913472671952078,
+      "learning_rate": 4.9450317124735735e-05,
+      "loss": 0.4752,
+      "step": 289
+    },
+    {
+      "epoch": 0.33086138049058755,
+      "grad_norm": 0.41164661389934154,
+      "learning_rate": 4.9429175475687106e-05,
+      "loss": 0.482,
+      "step": 290
+    },
+    {
+      "epoch": 0.3320022818026241,
+      "grad_norm": 0.5609144536016337,
+      "learning_rate": 4.9408033826638476e-05,
+      "loss": 0.4645,
+      "step": 291
+    },
+    {
+      "epoch": 0.33314318311466057,
+      "grad_norm": 0.5121918191555825,
+      "learning_rate": 4.9386892177589854e-05,
+      "loss": 0.4984,
+      "step": 292
+    },
+    {
+      "epoch": 0.3342840844266971,
+      "grad_norm": 0.6135026464347259,
+      "learning_rate": 4.9365750528541225e-05,
+      "loss": 0.4792,
+      "step": 293
+    },
+    {
+      "epoch": 0.3354249857387336,
+      "grad_norm": 0.4892389294352073,
+      "learning_rate": 4.93446088794926e-05,
+      "loss": 0.4639,
+      "step": 294
+    },
+    {
+      "epoch": 0.3365658870507701,
+      "grad_norm": 0.7410520792633273,
+      "learning_rate": 4.932346723044397e-05,
+      "loss": 0.4808,
+      "step": 295
+    },
+    {
+      "epoch": 0.3377067883628066,
+      "grad_norm": 0.46888189797192714,
+      "learning_rate": 4.930232558139535e-05,
+      "loss": 0.4832,
+      "step": 296
+    },
+    {
+      "epoch": 0.33884768967484313,
+      "grad_norm": 0.6007793767794352,
+      "learning_rate": 4.928118393234673e-05,
+      "loss": 0.4716,
+      "step": 297
+    },
+    {
+      "epoch": 0.3399885909868796,
+      "grad_norm": 0.69216963559477,
+      "learning_rate": 4.9260042283298105e-05,
+      "loss": 0.5043,
+      "step": 298
+    },
+    {
+      "epoch": 0.34112949229891615,
+      "grad_norm": 0.4687938078596549,
+      "learning_rate": 4.9238900634249476e-05,
+      "loss": 0.4992,
+      "step": 299
+    },
+    {
+      "epoch": 0.34227039361095263,
+      "grad_norm": 0.7268590781702684,
+      "learning_rate": 4.9217758985200846e-05,
+      "loss": 0.4816,
+      "step": 300
+    },
+    {
+      "epoch": 0.34341129492298916,
+      "grad_norm": 0.45936413381359864,
+      "learning_rate": 4.9196617336152224e-05,
+      "loss": 0.4453,
+      "step": 301
+    },
+    {
+      "epoch": 0.34455219623502564,
+      "grad_norm": 0.4294250943742063,
+      "learning_rate": 4.9175475687103595e-05,
+      "loss": 0.468,
+      "step": 302
+    },
+    {
+      "epoch": 0.3456930975470622,
+      "grad_norm": 0.5527427692618347,
+      "learning_rate": 4.915433403805497e-05,
+      "loss": 0.4722,
+      "step": 303
+    },
+    {
+      "epoch": 0.3468339988590987,
+      "grad_norm": 0.5071440857492974,
+      "learning_rate": 4.913319238900634e-05,
+      "loss": 0.4816,
+      "step": 304
+    },
+    {
+      "epoch": 0.3479749001711352,
+      "grad_norm": 0.4686500913435551,
+      "learning_rate": 4.911205073995772e-05,
+      "loss": 0.4473,
+      "step": 305
+    },
+    {
+      "epoch": 0.34911580148317173,
+      "grad_norm": 0.5883985093169647,
+      "learning_rate": 4.909090909090909e-05,
+      "loss": 0.4721,
+      "step": 306
+    },
+    {
+      "epoch": 0.3502567027952082,
+      "grad_norm": 0.4591838303775839,
+      "learning_rate": 4.906976744186046e-05,
+      "loss": 0.4553,
+      "step": 307
+    },
+    {
+      "epoch": 0.35139760410724474,
+      "grad_norm": 0.5549203579767384,
+      "learning_rate": 4.904862579281184e-05,
+      "loss": 0.4806,
+      "step": 308
+    },
+    {
+      "epoch": 0.3525385054192812,
+      "grad_norm": 0.5707305006848794,
+      "learning_rate": 4.9027484143763217e-05,
+      "loss": 0.462,
+      "step": 309
+    },
+    {
+      "epoch": 0.35367940673131776,
+      "grad_norm": 0.5070136654928392,
+      "learning_rate": 4.9006342494714594e-05,
+      "loss": 0.4914,
+      "step": 310
+    },
+    {
+      "epoch": 0.35482030804335424,
+      "grad_norm": 0.513783859903657,
+      "learning_rate": 4.8985200845665965e-05,
+      "loss": 0.4563,
+      "step": 311
+    },
+    {
+      "epoch": 0.3559612093553908,
+      "grad_norm": 0.3903353048037027,
+      "learning_rate": 4.896405919661734e-05,
+      "loss": 0.4646,
+      "step": 312
+    },
+    {
+      "epoch": 0.35710211066742725,
+      "grad_norm": 0.4944624580259425,
+      "learning_rate": 4.894291754756871e-05,
+      "loss": 0.4794,
+      "step": 313
+    },
+    {
+      "epoch": 0.3582430119794638,
+      "grad_norm": 0.4650438379271747,
+      "learning_rate": 4.892177589852009e-05,
+      "loss": 0.4734,
+      "step": 314
+    },
+    {
+      "epoch": 0.35938391329150027,
+      "grad_norm": 0.40152357897892427,
+      "learning_rate": 4.890063424947146e-05,
+      "loss": 0.4558,
+      "step": 315
+    },
+    {
+      "epoch": 0.3605248146035368,
+      "grad_norm": 0.4062828773190652,
+      "learning_rate": 4.887949260042283e-05,
+      "loss": 0.4424,
+      "step": 316
+    },
+    {
+      "epoch": 0.3616657159155733,
+      "grad_norm": 0.4013001138970794,
+      "learning_rate": 4.885835095137421e-05,
+      "loss": 0.4799,
+      "step": 317
+    },
+    {
+      "epoch": 0.3628066172276098,
+      "grad_norm": 0.369004901568825,
+      "learning_rate": 4.883720930232558e-05,
+      "loss": 0.4653,
+      "step": 318
+    },
+    {
+      "epoch": 0.3639475185396463,
+      "grad_norm": 0.48032108696732456,
+      "learning_rate": 4.881606765327696e-05,
+      "loss": 0.469,
+      "step": 319
+    },
+    {
+      "epoch": 0.36508841985168283,
+      "grad_norm": 0.405515724241134,
+      "learning_rate": 4.879492600422833e-05,
+      "loss": 0.4697,
+      "step": 320
+    },
+    {
+      "epoch": 0.3662293211637193,
+      "grad_norm": 0.4030716686771247,
+      "learning_rate": 4.8773784355179705e-05,
+      "loss": 0.4743,
+      "step": 321
+    },
+    {
+      "epoch": 0.36737022247575585,
+      "grad_norm": 0.4611550051218631,
+      "learning_rate": 4.875264270613108e-05,
+      "loss": 0.4639,
+      "step": 322
+    },
+    {
+      "epoch": 0.3685111237877924,
+      "grad_norm": 0.49050018609555823,
+      "learning_rate": 4.873150105708246e-05,
+      "loss": 0.4709,
+      "step": 323
+    },
+    {
+      "epoch": 0.36965202509982886,
+      "grad_norm": 0.44515733384409556,
+      "learning_rate": 4.871035940803383e-05,
+      "loss": 0.4409,
+      "step": 324
+    },
+    {
+      "epoch": 0.3707929264118654,
+      "grad_norm": 0.405005368775498,
+      "learning_rate": 4.86892177589852e-05,
+      "loss": 0.4662,
+      "step": 325
+    },
+    {
+      "epoch": 0.3719338277239019,
+      "grad_norm": 0.44424805271869844,
+      "learning_rate": 4.866807610993658e-05,
+      "loss": 0.4919,
+      "step": 326
+    },
+    {
+      "epoch": 0.3730747290359384,
+      "grad_norm": 0.47916627855366895,
+      "learning_rate": 4.864693446088795e-05,
+      "loss": 0.4665,
+      "step": 327
+    },
+    {
+      "epoch": 0.3742156303479749,
+      "grad_norm": 0.4591787529396741,
+      "learning_rate": 4.862579281183933e-05,
+      "loss": 0.4751,
+      "step": 328
+    },
+    {
+      "epoch": 0.3753565316600114,
+      "grad_norm": 0.3863033452781703,
+      "learning_rate": 4.86046511627907e-05,
+      "loss": 0.4666,
+      "step": 329
+    },
+    {
+      "epoch": 0.3764974329720479,
+      "grad_norm": 0.4267031556202453,
+      "learning_rate": 4.8583509513742075e-05,
+      "loss": 0.4617,
+      "step": 330
+    },
+    {
+      "epoch": 0.37763833428408444,
+      "grad_norm": 0.4566590620989737,
+      "learning_rate": 4.8562367864693446e-05,
+      "loss": 0.4568,
+      "step": 331
+    },
+    {
+      "epoch": 0.3787792355961209,
+      "grad_norm": 0.3909866338741756,
+      "learning_rate": 4.8541226215644824e-05,
+      "loss": 0.4431,
+      "step": 332
+    },
+    {
+      "epoch": 0.37992013690815746,
+      "grad_norm": 0.4445867673454024,
+      "learning_rate": 4.8520084566596194e-05,
+      "loss": 0.4593,
+      "step": 333
+    },
+    {
+      "epoch": 0.38106103822019394,
+      "grad_norm": 0.4002731387396594,
+      "learning_rate": 4.849894291754757e-05,
+      "loss": 0.4411,
+      "step": 334
+    },
+    {
+      "epoch": 0.3822019395322305,
+      "grad_norm": 0.4133903450308309,
+      "learning_rate": 4.847780126849894e-05,
+      "loss": 0.4781,
+      "step": 335
+    },
+    {
+      "epoch": 0.38334284084426695,
+      "grad_norm": 0.47770252676093944,
+      "learning_rate": 4.845665961945032e-05,
+      "loss": 0.4883,
+      "step": 336
+    },
+    {
+      "epoch": 0.3844837421563035,
+      "grad_norm": 0.3666231185699412,
+      "learning_rate": 4.84355179704017e-05,
+      "loss": 0.4824,
+      "step": 337
+    },
+    {
+      "epoch": 0.38562464346833997,
+      "grad_norm": 0.48134048114822264,
+      "learning_rate": 4.841437632135307e-05,
+      "loss": 0.4586,
+      "step": 338
+    },
+    {
+      "epoch": 0.3867655447803765,
+      "grad_norm": 0.4186276272086947,
+      "learning_rate": 4.8393234672304445e-05,
+      "loss": 0.4743,
+      "step": 339
+    },
+    {
+      "epoch": 0.387906446092413,
+      "grad_norm": 0.4320564974138227,
+      "learning_rate": 4.8372093023255816e-05,
+      "loss": 0.4687,
+      "step": 340
+    },
+    {
+      "epoch": 0.3890473474044495,
+      "grad_norm": 0.4969726708200199,
+      "learning_rate": 4.8350951374207194e-05,
+      "loss": 0.4708,
+      "step": 341
+    },
+    {
+      "epoch": 0.390188248716486,
+      "grad_norm": 0.476437641405027,
+      "learning_rate": 4.8329809725158564e-05,
+      "loss": 0.4477,
+      "step": 342
+    },
+    {
+      "epoch": 0.39132915002852253,
+      "grad_norm": 0.4953362791687895,
+      "learning_rate": 4.8308668076109935e-05,
+      "loss": 0.4811,
+      "step": 343
+    },
+    {
+      "epoch": 0.39247005134055907,
+      "grad_norm": 0.551611928855792,
+      "learning_rate": 4.828752642706131e-05,
+      "loss": 0.4738,
+      "step": 344
+    },
+    {
+      "epoch": 0.39361095265259555,
+      "grad_norm": 0.40738026772406793,
+      "learning_rate": 4.826638477801268e-05,
+      "loss": 0.4615,
+      "step": 345
+    },
+    {
+      "epoch": 0.3947518539646321,
+      "grad_norm": 0.4782666268905305,
+      "learning_rate": 4.824524312896406e-05,
+      "loss": 0.4731,
+      "step": 346
+    },
+    {
+      "epoch": 0.39589275527666856,
+      "grad_norm": 0.44504504658811994,
+      "learning_rate": 4.822410147991543e-05,
+      "loss": 0.4677,
+      "step": 347
+    },
+    {
+      "epoch": 0.3970336565887051,
+      "grad_norm": 0.4533234538781712,
+      "learning_rate": 4.820295983086681e-05,
+      "loss": 0.4617,
+      "step": 348
+    },
+    {
+      "epoch": 0.3981745579007416,
+      "grad_norm": 0.46248079465133807,
+      "learning_rate": 4.8181818181818186e-05,
+      "loss": 0.4932,
+      "step": 349
+    },
+    {
+      "epoch": 0.3993154592127781,
+      "grad_norm": 0.4780267679080641,
+      "learning_rate": 4.8160676532769564e-05,
+      "loss": 0.4599,
+      "step": 350
+    },
+    {
+      "epoch": 0.4004563605248146,
+      "grad_norm": 0.4448999136385355,
+      "learning_rate": 4.8139534883720934e-05,
+      "loss": 0.487,
+      "step": 351
+    },
+    {
+      "epoch": 0.4015972618368511,
+      "grad_norm": 0.602985396056506,
+      "learning_rate": 4.8118393234672305e-05,
+      "loss": 0.4807,
+      "step": 352
+    },
+    {
+      "epoch": 0.4027381631488876,
+      "grad_norm": 0.4400835155350816,
+      "learning_rate": 4.809725158562368e-05,
+      "loss": 0.4614,
+      "step": 353
+    },
+    {
+      "epoch": 0.40387906446092414,
+      "grad_norm": 0.6103639448472573,
+      "learning_rate": 4.807610993657505e-05,
+      "loss": 0.4688,
+      "step": 354
+    },
+    {
+      "epoch": 0.4050199657729606,
+      "grad_norm": 0.48612080977613853,
+      "learning_rate": 4.805496828752643e-05,
+      "loss": 0.4554,
+      "step": 355
+    },
+    {
+      "epoch": 0.40616086708499716,
+      "grad_norm": 0.5220692500810554,
+      "learning_rate": 4.80338266384778e-05,
+      "loss": 0.4603,
+      "step": 356
+    },
+    {
+      "epoch": 0.40730176839703364,
+      "grad_norm": 0.5248213992380287,
+      "learning_rate": 4.801268498942918e-05,
+      "loss": 0.4623,
+      "step": 357
+    },
+    {
+      "epoch": 0.40844266970907017,
+      "grad_norm": 0.6420774579491392,
+      "learning_rate": 4.799154334038055e-05,
+      "loss": 0.4756,
+      "step": 358
+    },
+    {
+      "epoch": 0.40958357102110665,
+      "grad_norm": 0.4352271420359283,
+      "learning_rate": 4.797040169133193e-05,
+      "loss": 0.4516,
+      "step": 359
+    },
+    {
+      "epoch": 0.4107244723331432,
+      "grad_norm": 0.658584341212213,
+      "learning_rate": 4.79492600422833e-05,
+      "loss": 0.4763,
+      "step": 360
+    },
+    {
+      "epoch": 0.41186537364517967,
+      "grad_norm": 0.6084351588841551,
+      "learning_rate": 4.7928118393234675e-05,
+      "loss": 0.4798,
+      "step": 361
+    },
+    {
+      "epoch": 0.4130062749572162,
+      "grad_norm": 0.5771682284997688,
+      "learning_rate": 4.790697674418605e-05,
+      "loss": 0.481,
+      "step": 362
+    },
+    {
+      "epoch": 0.41414717626925274,
+      "grad_norm": 0.6043942596380553,
+      "learning_rate": 4.788583509513742e-05,
+      "loss": 0.4688,
+      "step": 363
+    },
+    {
+      "epoch": 0.4152880775812892,
+      "grad_norm": 0.45719477020974547,
+      "learning_rate": 4.78646934460888e-05,
+      "loss": 0.4501,
+      "step": 364
+    },
+    {
+      "epoch": 0.41642897889332575,
+      "grad_norm": 0.5741113368011214,
+      "learning_rate": 4.784355179704017e-05,
+      "loss": 0.4606,
+      "step": 365
+    },
+    {
+      "epoch": 0.41756988020536223,
+      "grad_norm": 0.6037611112612937,
+      "learning_rate": 4.782241014799155e-05,
+      "loss": 0.4681,
+      "step": 366
+    },
+    {
+      "epoch": 0.41871078151739877,
+      "grad_norm": 0.4385849776699442,
+      "learning_rate": 4.780126849894292e-05,
+      "loss": 0.4401,
+      "step": 367
+    },
+    {
+      "epoch": 0.41985168282943525,
+      "grad_norm": 0.4484383147175151,
+      "learning_rate": 4.77801268498943e-05,
+      "loss": 0.4737,
+      "step": 368
+    },
+    {
+      "epoch": 0.4209925841414718,
+      "grad_norm": 0.5756231728396068,
+      "learning_rate": 4.775898520084567e-05,
+      "loss": 0.4646,
+      "step": 369
+    },
+    {
+      "epoch": 0.42213348545350826,
+      "grad_norm": 0.39706507133213986,
+      "learning_rate": 4.773784355179704e-05,
+      "loss": 0.4645,
+      "step": 370
+    },
+    {
+      "epoch": 0.4232743867655448,
+      "grad_norm": 0.5924514871872969,
+      "learning_rate": 4.7716701902748416e-05,
+      "loss": 0.461,
+      "step": 371
+    },
+    {
+      "epoch": 0.4244152880775813,
+      "grad_norm": 0.41316719683932873,
+      "learning_rate": 4.7695560253699787e-05,
+      "loss": 0.4785,
+      "step": 372
+    },
+    {
+      "epoch": 0.4255561893896178,
+      "grad_norm": 0.39683025268881833,
+      "learning_rate": 4.7674418604651164e-05,
+      "loss": 0.4591,
+      "step": 373
+    },
+    {
+      "epoch": 0.4266970907016543,
+      "grad_norm": 0.48000498985785206,
+      "learning_rate": 4.765327695560254e-05,
+      "loss": 0.4414,
+      "step": 374
+    },
+    {
+      "epoch": 0.4278379920136908,
+      "grad_norm": 0.3591998859631829,
+      "learning_rate": 4.763213530655392e-05,
+      "loss": 0.471,
+      "step": 375
+    },
+    {
+      "epoch": 0.4289788933257273,
+      "grad_norm": 0.4431516881768443,
+      "learning_rate": 4.761099365750529e-05,
+      "loss": 0.4597,
+      "step": 376
+    },
+    {
+      "epoch": 0.43011979463776384,
+      "grad_norm": 0.36560648233369786,
+      "learning_rate": 4.758985200845667e-05,
+      "loss": 0.4682,
+      "step": 377
+    },
+    {
+      "epoch": 0.4312606959498003,
+      "grad_norm": 0.4558699153147223,
+      "learning_rate": 4.756871035940804e-05,
+      "loss": 0.4499,
+      "step": 378
+    },
+    {
+      "epoch": 0.43240159726183686,
+      "grad_norm": 0.5460346505513063,
+      "learning_rate": 4.754756871035941e-05,
+      "loss": 0.4539,
+      "step": 379
+    },
+    {
+      "epoch": 0.43354249857387334,
+      "grad_norm": 0.4945804014357001,
+      "learning_rate": 4.7526427061310786e-05,
+      "loss": 0.4668,
+      "step": 380
+    },
+    {
+      "epoch": 0.43468339988590987,
+      "grad_norm": 0.381982606728481,
+      "learning_rate": 4.7505285412262157e-05,
+      "loss": 0.4629,
+      "step": 381
+    },
+    {
+      "epoch": 0.43582430119794635,
+      "grad_norm": 0.5075561122741824,
+      "learning_rate": 4.7484143763213534e-05,
+      "loss": 0.4419,
+      "step": 382
+    },
+    {
+      "epoch": 0.4369652025099829,
+      "grad_norm": 0.4890820416949409,
+      "learning_rate": 4.7463002114164905e-05,
+      "loss": 0.474,
+      "step": 383
+    },
+    {
+      "epoch": 0.4381061038220194,
+      "grad_norm": 0.45099733935651065,
+      "learning_rate": 4.744186046511628e-05,
+      "loss": 0.4607,
+      "step": 384
+    },
+    {
+      "epoch": 0.4392470051340559,
+      "grad_norm": 0.46389402376612976,
+      "learning_rate": 4.742071881606765e-05,
+      "loss": 0.4529,
+      "step": 385
+    },
+    {
+      "epoch": 0.44038790644609244,
+      "grad_norm": 0.5176225091984493,
+      "learning_rate": 4.739957716701903e-05,
+      "loss": 0.4797,
+      "step": 386
+    },
+    {
+      "epoch": 0.4415288077581289,
+      "grad_norm": 0.5614444078753256,
+      "learning_rate": 4.73784355179704e-05,
+      "loss": 0.4415,
+      "step": 387
+    },
+    {
+      "epoch": 0.44266970907016545,
+      "grad_norm": 0.5620355202952496,
+      "learning_rate": 4.735729386892178e-05,
+      "loss": 0.4901,
+      "step": 388
+    },
+    {
+      "epoch": 0.44381061038220193,
+      "grad_norm": 0.3554047434699461,
+      "learning_rate": 4.7336152219873156e-05,
+      "loss": 0.4486,
+      "step": 389
+    },
+    {
+      "epoch": 0.44495151169423847,
+      "grad_norm": 0.565055930787776,
+      "learning_rate": 4.7315010570824527e-05,
+      "loss": 0.4794,
+      "step": 390
+    },
+    {
+      "epoch": 0.44609241300627495,
+      "grad_norm": 0.40372827032444986,
+      "learning_rate": 4.7293868921775904e-05,
+      "loss": 0.4549,
+      "step": 391
+    },
+    {
+      "epoch": 0.4472333143183115,
+      "grad_norm": 0.41175642665695783,
+      "learning_rate": 4.7272727272727275e-05,
+      "loss": 0.4515,
+      "step": 392
+    },
+    {
+      "epoch": 0.44837421563034796,
+      "grad_norm": 0.5428841648637599,
+      "learning_rate": 4.725158562367865e-05,
+      "loss": 0.4688,
+      "step": 393
+    },
+    {
+      "epoch": 0.4495151169423845,
+      "grad_norm": 0.3845089691572888,
+      "learning_rate": 4.723044397463002e-05,
+      "loss": 0.4599,
+      "step": 394
+    },
+    {
+      "epoch": 0.450656018254421,
+      "grad_norm": 0.5575460169748139,
+      "learning_rate": 4.7209302325581394e-05,
+      "loss": 0.46,
+      "step": 395
+    },
+    {
+      "epoch": 0.4517969195664575,
+      "grad_norm": 0.47319289284357274,
+      "learning_rate": 4.718816067653277e-05,
+      "loss": 0.4743,
+      "step": 396
+    },
+    {
+      "epoch": 0.452937820878494,
+      "grad_norm": 0.421430478800123,
+      "learning_rate": 4.716701902748414e-05,
+      "loss": 0.487,
+      "step": 397
+    },
+    {
+      "epoch": 0.4540787221905305,
+      "grad_norm": 0.6101298495936283,
+      "learning_rate": 4.714587737843552e-05,
+      "loss": 0.4225,
+      "step": 398
+    },
+    {
+      "epoch": 0.455219623502567,
+      "grad_norm": 0.40737340116351733,
+      "learning_rate": 4.712473572938689e-05,
+      "loss": 0.4783,
+      "step": 399
+    },
+    {
+      "epoch": 0.45636052481460354,
+      "grad_norm": 0.46178856188238293,
+      "learning_rate": 4.710359408033827e-05,
+      "loss": 0.5043,
+      "step": 400
+    },
+    {
+      "epoch": 0.45750142612664,
+      "grad_norm": 0.36605473430041985,
+      "learning_rate": 4.7082452431289645e-05,
+      "loss": 0.4727,
+      "step": 401
+    },
+    {
+      "epoch": 0.45864232743867656,
+      "grad_norm": 0.4601612072292561,
+      "learning_rate": 4.706131078224102e-05,
+      "loss": 0.4729,
+      "step": 402
+    },
+    {
+      "epoch": 0.4597832287507131,
+      "grad_norm": 0.39229637932652944,
+      "learning_rate": 4.704016913319239e-05,
+      "loss": 0.4318,
+      "step": 403
+    },
+    {
+      "epoch": 0.46092413006274957,
+      "grad_norm": 0.4978174882138765,
+      "learning_rate": 4.7019027484143764e-05,
+      "loss": 0.4593,
+      "step": 404
+    },
+    {
+      "epoch": 0.4620650313747861,
+      "grad_norm": 0.43613960694098686,
+      "learning_rate": 4.699788583509514e-05,
+      "loss": 0.4863,
+      "step": 405
+    },
+    {
+      "epoch": 0.4632059326868226,
+      "grad_norm": 0.4748040163303423,
+      "learning_rate": 4.697674418604651e-05,
+      "loss": 0.4736,
+      "step": 406
+    },
+    {
+      "epoch": 0.4643468339988591,
+      "grad_norm": 0.4507681010912006,
+      "learning_rate": 4.695560253699789e-05,
+      "loss": 0.4537,
+      "step": 407
+    },
+    {
+      "epoch": 0.4654877353108956,
+      "grad_norm": 0.3923537491739542,
+      "learning_rate": 4.693446088794926e-05,
+      "loss": 0.4417,
+      "step": 408
+    },
+    {
+      "epoch": 0.46662863662293214,
+      "grad_norm": 0.4819511355971476,
+      "learning_rate": 4.691331923890064e-05,
+      "loss": 0.4409,
+      "step": 409
+    },
+    {
+      "epoch": 0.4677695379349686,
+      "grad_norm": 0.46716431084892146,
+      "learning_rate": 4.689217758985201e-05,
+      "loss": 0.4361,
+      "step": 410
+    },
+    {
+      "epoch": 0.46891043924700515,
+      "grad_norm": 0.39268914274801164,
+      "learning_rate": 4.6871035940803386e-05,
+      "loss": 0.4527,
+      "step": 411
+    },
+    {
+      "epoch": 0.47005134055904163,
+      "grad_norm": 0.36742442071038406,
+      "learning_rate": 4.6849894291754756e-05,
+      "loss": 0.441,
+      "step": 412
+    },
+    {
+      "epoch": 0.47119224187107817,
+      "grad_norm": 0.3402798863969448,
+      "learning_rate": 4.6828752642706134e-05,
+      "loss": 0.4611,
+      "step": 413
+    },
+    {
+      "epoch": 0.47233314318311465,
+      "grad_norm": 0.4783227235848215,
+      "learning_rate": 4.680761099365751e-05,
+      "loss": 0.4468,
+      "step": 414
+    },
+    {
+      "epoch": 0.4734740444951512,
+      "grad_norm": 0.41760869890010066,
+      "learning_rate": 4.678646934460888e-05,
+      "loss": 0.419,
+      "step": 415
+    },
+    {
+      "epoch": 0.47461494580718766,
+      "grad_norm": 0.45492695505371195,
+      "learning_rate": 4.676532769556026e-05,
+      "loss": 0.4539,
+      "step": 416
+    },
+    {
+      "epoch": 0.4757558471192242,
+      "grad_norm": 0.4242418558060538,
+      "learning_rate": 4.674418604651163e-05,
+      "loss": 0.439,
+      "step": 417
+    },
+    {
+      "epoch": 0.4768967484312607,
+      "grad_norm": 0.44887147636919633,
+      "learning_rate": 4.672304439746301e-05,
+      "loss": 0.4681,
+      "step": 418
+    },
+    {
+      "epoch": 0.4780376497432972,
+      "grad_norm": 0.5450796324350122,
+      "learning_rate": 4.670190274841438e-05,
+      "loss": 0.4648,
+      "step": 419
+    },
+    {
+      "epoch": 0.4791785510553337,
+      "grad_norm": 0.4650341425698157,
+      "learning_rate": 4.6680761099365756e-05,
+      "loss": 0.4831,
+      "step": 420
+    },
+    {
+      "epoch": 0.4803194523673702,
+      "grad_norm": 0.48371809386595394,
+      "learning_rate": 4.6659619450317126e-05,
+      "loss": 0.45,
+      "step": 421
+    },
+    {
+      "epoch": 0.4814603536794067,
+      "grad_norm": 0.34547311035282235,
+      "learning_rate": 4.66384778012685e-05,
+      "loss": 0.4553,
+      "step": 422
+    },
+    {
+      "epoch": 0.48260125499144324,
+      "grad_norm": 0.4630732324267297,
+      "learning_rate": 4.6617336152219874e-05,
+      "loss": 0.4592,
+      "step": 423
+    },
+    {
+      "epoch": 0.4837421563034798,
+      "grad_norm": 0.3564440597661613,
+      "learning_rate": 4.6596194503171245e-05,
+      "loss": 0.4506,
+      "step": 424
+    },
+    {
+      "epoch": 0.48488305761551626,
+      "grad_norm": 0.39413931707375205,
+      "learning_rate": 4.657505285412262e-05,
+      "loss": 0.4722,
+      "step": 425
+    },
+    {
+      "epoch": 0.4860239589275528,
+      "grad_norm": 0.377062944638173,
+      "learning_rate": 4.6553911205074e-05,
+      "loss": 0.4587,
+      "step": 426
+    },
+    {
+      "epoch": 0.48716486023958927,
+      "grad_norm": 0.3624696717743631,
+      "learning_rate": 4.653276955602537e-05,
+      "loss": 0.4587,
+      "step": 427
+    },
+    {
+      "epoch": 0.4883057615516258,
+      "grad_norm": 0.3380749179957887,
+      "learning_rate": 4.651162790697675e-05,
+      "loss": 0.4456,
+      "step": 428
+    },
+    {
+      "epoch": 0.4894466628636623,
+      "grad_norm": 0.3928759661904698,
+      "learning_rate": 4.6490486257928126e-05,
+      "loss": 0.4789,
+      "step": 429
+    },
+    {
+      "epoch": 0.4905875641756988,
+      "grad_norm": 0.36642927400424125,
+      "learning_rate": 4.6469344608879496e-05,
+      "loss": 0.465,
+      "step": 430
+    },
+    {
+      "epoch": 0.4917284654877353,
+      "grad_norm": 0.4401740516737291,
+      "learning_rate": 4.644820295983087e-05,
+      "loss": 0.4472,
+      "step": 431
+    },
+    {
+      "epoch": 0.49286936679977184,
+      "grad_norm": 0.4635191449978593,
+      "learning_rate": 4.6427061310782244e-05,
+      "loss": 0.4637,
+      "step": 432
+    },
+    {
+      "epoch": 0.4940102681118083,
+      "grad_norm": 0.4321430146961908,
+      "learning_rate": 4.6405919661733615e-05,
+      "loss": 0.4705,
+      "step": 433
+    },
+    {
+      "epoch": 0.49515116942384485,
+      "grad_norm": 0.4583926087622166,
+      "learning_rate": 4.638477801268499e-05,
+      "loss": 0.45,
+      "step": 434
+    },
+    {
+      "epoch": 0.49629207073588133,
+      "grad_norm": 0.37924166776725926,
+      "learning_rate": 4.636363636363636e-05,
+      "loss": 0.4635,
+      "step": 435
+    },
+    {
+      "epoch": 0.49743297204791787,
+      "grad_norm": 0.3929198503783768,
+      "learning_rate": 4.634249471458774e-05,
+      "loss": 0.4592,
+      "step": 436
+    },
+    {
+      "epoch": 0.49857387335995434,
+      "grad_norm": 0.38813120210535945,
+      "learning_rate": 4.632135306553911e-05,
+      "loss": 0.4705,
+      "step": 437
+    },
+    {
+      "epoch": 0.4997147746719909,
+      "grad_norm": 0.4762099160364005,
+      "learning_rate": 4.630021141649049e-05,
+      "loss": 0.4792,
+      "step": 438
+    },
+    {
+      "epoch": 0.5008556759840274,
+      "grad_norm": 0.46547600756258845,
+      "learning_rate": 4.627906976744186e-05,
+      "loss": 0.4808,
+      "step": 439
+    },
+    {
+      "epoch": 0.5019965772960638,
+      "grad_norm": 0.4606755516881231,
+      "learning_rate": 4.625792811839324e-05,
+      "loss": 0.429,
+      "step": 440
+    },
+    {
+      "epoch": 0.5031374786081004,
+      "grad_norm": 0.4462475824492918,
+      "learning_rate": 4.6236786469344614e-05,
+      "loss": 0.4689,
+      "step": 441
+    },
+    {
+      "epoch": 0.5042783799201369,
+      "grad_norm": 0.47174621173295,
+      "learning_rate": 4.6215644820295985e-05,
+      "loss": 0.4489,
+      "step": 442
+    },
+    {
+      "epoch": 0.5054192812321734,
+      "grad_norm": 0.5015710382124482,
+      "learning_rate": 4.619450317124736e-05,
+      "loss": 0.4939,
+      "step": 443
+    },
+    {
+      "epoch": 0.50656018254421,
+      "grad_norm": 0.3818394201165467,
+      "learning_rate": 4.617336152219873e-05,
+      "loss": 0.4576,
+      "step": 444
+    },
+    {
+      "epoch": 0.5077010838562465,
+      "grad_norm": 0.4225972875649922,
+      "learning_rate": 4.615221987315011e-05,
+      "loss": 0.4654,
+      "step": 445
+    },
+    {
+      "epoch": 0.5088419851682829,
+      "grad_norm": 0.40882987228929407,
+      "learning_rate": 4.613107822410148e-05,
+      "loss": 0.4527,
+      "step": 446
+    },
+    {
+      "epoch": 0.5099828864803194,
+      "grad_norm": 0.5069671408649863,
+      "learning_rate": 4.610993657505286e-05,
+      "loss": 0.4576,
+      "step": 447
+    },
+    {
+      "epoch": 0.511123787792356,
+      "grad_norm": 0.4386417973131366,
+      "learning_rate": 4.608879492600423e-05,
+      "loss": 0.4552,
+      "step": 448
+    },
+    {
+      "epoch": 0.5122646891043925,
+      "grad_norm": 0.4747784151033817,
+      "learning_rate": 4.60676532769556e-05,
+      "loss": 0.4561,
+      "step": 449
+    },
+    {
+      "epoch": 0.513405590416429,
+      "grad_norm": 0.40932899748081414,
+      "learning_rate": 4.604651162790698e-05,
+      "loss": 0.467,
+      "step": 450
+    },
+    {
+      "epoch": 0.5145464917284654,
+      "grad_norm": 0.40171173472315463,
+      "learning_rate": 4.602536997885835e-05,
+      "loss": 0.4696,
+      "step": 451
+    },
+    {
+      "epoch": 0.515687393040502,
+      "grad_norm": 0.46145868641103904,
+      "learning_rate": 4.6004228329809726e-05,
+      "loss": 0.4737,
+      "step": 452
+    },
+    {
+      "epoch": 0.5168282943525385,
+      "grad_norm": 0.38520232110740626,
+      "learning_rate": 4.59830866807611e-05,
+      "loss": 0.4606,
+      "step": 453
+    },
+    {
+      "epoch": 0.517969195664575,
+      "grad_norm": 0.4728138540537639,
+      "learning_rate": 4.596194503171248e-05,
+      "loss": 0.4738,
+      "step": 454
+    },
+    {
+      "epoch": 0.5191100969766115,
+      "grad_norm": 0.40077436012399575,
+      "learning_rate": 4.594080338266385e-05,
+      "loss": 0.4459,
+      "step": 455
+    },
+    {
+      "epoch": 0.5202509982886481,
+      "grad_norm": 0.595822283116385,
+      "learning_rate": 4.591966173361523e-05,
+      "loss": 0.4549,
+      "step": 456
+    },
+    {
+      "epoch": 0.5213918996006845,
+      "grad_norm": 0.3754288530168292,
+      "learning_rate": 4.58985200845666e-05,
+      "loss": 0.441,
+      "step": 457
+    },
+    {
+      "epoch": 0.522532800912721,
+      "grad_norm": 0.4243930283188664,
+      "learning_rate": 4.587737843551797e-05,
+      "loss": 0.4319,
+      "step": 458
+    },
+    {
+      "epoch": 0.5236737022247575,
+      "grad_norm": 0.453192510555693,
+      "learning_rate": 4.585623678646935e-05,
+      "loss": 0.4745,
+      "step": 459
+    },
+    {
+      "epoch": 0.5248146035367941,
+      "grad_norm": 0.3976339179450837,
+      "learning_rate": 4.583509513742072e-05,
+      "loss": 0.458,
+      "step": 460
+    },
+    {
+      "epoch": 0.5259555048488306,
+      "grad_norm": 0.42197906453695416,
+      "learning_rate": 4.5813953488372096e-05,
+      "loss": 0.4591,
+      "step": 461
+    },
+    {
+      "epoch": 0.5270964061608671,
+      "grad_norm": 0.35718537422327623,
+      "learning_rate": 4.579281183932347e-05,
+      "loss": 0.4741,
+      "step": 462
+    },
+    {
+      "epoch": 0.5282373074729035,
+      "grad_norm": 0.409421659544212,
+      "learning_rate": 4.5771670190274844e-05,
+      "loss": 0.4496,
+      "step": 463
+    },
+    {
+      "epoch": 0.5293782087849401,
+      "grad_norm": 0.4168650729475048,
+      "learning_rate": 4.5750528541226215e-05,
+      "loss": 0.4372,
+      "step": 464
+    },
+    {
+      "epoch": 0.5305191100969766,
+      "grad_norm": 0.3801742006510912,
+      "learning_rate": 4.572938689217759e-05,
+      "loss": 0.4652,
+      "step": 465
+    },
+    {
+      "epoch": 0.5316600114090131,
+      "grad_norm": 0.42600016168123134,
+      "learning_rate": 4.570824524312897e-05,
+      "loss": 0.4416,
+      "step": 466
+    },
+    {
+      "epoch": 0.5328009127210497,
+      "grad_norm": 0.46602265019553896,
+      "learning_rate": 4.568710359408034e-05,
+      "loss": 0.4805,
+      "step": 467
+    },
+    {
+      "epoch": 0.5339418140330862,
+      "grad_norm": 0.48053850937804116,
+      "learning_rate": 4.566596194503172e-05,
+      "loss": 0.4575,
+      "step": 468
+    },
+    {
+      "epoch": 0.5350827153451226,
+      "grad_norm": 0.3986350097784419,
+      "learning_rate": 4.564482029598309e-05,
+      "loss": 0.4786,
+      "step": 469
+    },
+    {
+      "epoch": 0.5362236166571591,
+      "grad_norm": 0.4807794088581916,
+      "learning_rate": 4.5623678646934466e-05,
+      "loss": 0.4512,
+      "step": 470
+    },
+    {
+      "epoch": 0.5373645179691957,
+      "grad_norm": 0.433766824087754,
+      "learning_rate": 4.560253699788584e-05,
+      "loss": 0.4459,
+      "step": 471
+    },
+    {
+      "epoch": 0.5385054192812322,
+      "grad_norm": 0.4043451924904348,
+      "learning_rate": 4.5581395348837214e-05,
+      "loss": 0.4389,
+      "step": 472
+    },
+    {
+      "epoch": 0.5396463205932687,
+      "grad_norm": 0.4197291913156009,
+      "learning_rate": 4.5560253699788585e-05,
+      "loss": 0.4603,
+      "step": 473
+    },
+    {
+      "epoch": 0.5407872219053051,
+      "grad_norm": 0.4118103847467949,
+      "learning_rate": 4.553911205073996e-05,
+      "loss": 0.4201,
+      "step": 474
+    },
+    {
+      "epoch": 0.5419281232173417,
+      "grad_norm": 0.4168051352199324,
+      "learning_rate": 4.551797040169133e-05,
+      "loss": 0.4356,
+      "step": 475
+    },
+    {
+      "epoch": 0.5430690245293782,
+      "grad_norm": 0.41758694905260396,
+      "learning_rate": 4.5496828752642704e-05,
+      "loss": 0.4762,
+      "step": 476
+    },
+    {
+      "epoch": 0.5442099258414147,
+      "grad_norm": 0.3919808019373763,
+      "learning_rate": 4.547568710359408e-05,
+      "loss": 0.4476,
+      "step": 477
+    },
+    {
+      "epoch": 0.5453508271534512,
+      "grad_norm": 0.40308219490841923,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 0.4436,
+      "step": 478
+    },
+    {
+      "epoch": 0.5464917284654878,
+      "grad_norm": 0.39908396946636293,
+      "learning_rate": 4.543340380549683e-05,
+      "loss": 0.4541,
+      "step": 479
+    },
+    {
+      "epoch": 0.5476326297775242,
+      "grad_norm": 0.46934983540992264,
+      "learning_rate": 4.541226215644821e-05,
+      "loss": 0.4576,
+      "step": 480
+    },
+    {
+      "epoch": 0.5487735310895607,
+      "grad_norm": 0.3854077371954608,
+      "learning_rate": 4.5391120507399584e-05,
+      "loss": 0.4434,
+      "step": 481
+    },
+    {
+      "epoch": 0.5499144324015972,
+      "grad_norm": 0.5319242645223065,
+      "learning_rate": 4.5369978858350955e-05,
+      "loss": 0.4529,
+      "step": 482
+    },
+    {
+      "epoch": 0.5510553337136338,
+      "grad_norm": 0.44290253891327724,
+      "learning_rate": 4.5348837209302326e-05,
+      "loss": 0.4594,
+      "step": 483
+    },
+    {
+      "epoch": 0.5521962350256703,
+      "grad_norm": 0.4728615530783417,
+      "learning_rate": 4.53276955602537e-05,
+      "loss": 0.4448,
+      "step": 484
+    },
+    {
+      "epoch": 0.5533371363377068,
+      "grad_norm": 0.4404097761821421,
+      "learning_rate": 4.5306553911205074e-05,
+      "loss": 0.4426,
+      "step": 485
+    },
+    {
+      "epoch": 0.5544780376497433,
+      "grad_norm": 0.4197335397577139,
+      "learning_rate": 4.528541226215645e-05,
+      "loss": 0.4493,
+      "step": 486
+    },
+    {
+      "epoch": 0.5556189389617798,
+      "grad_norm": 0.477125915431258,
+      "learning_rate": 4.526427061310782e-05,
+      "loss": 0.438,
+      "step": 487
+    },
+    {
+      "epoch": 0.5567598402738163,
+      "grad_norm": 0.4696081174949721,
+      "learning_rate": 4.52431289640592e-05,
+      "loss": 0.455,
+      "step": 488
+    },
+    {
+      "epoch": 0.5579007415858528,
+      "grad_norm": 0.47219356202492724,
+      "learning_rate": 4.522198731501057e-05,
+      "loss": 0.4576,
+      "step": 489
+    },
+    {
+      "epoch": 0.5590416428978894,
+      "grad_norm": 0.446771837110723,
+      "learning_rate": 4.520084566596195e-05,
+      "loss": 0.4403,
+      "step": 490
+    },
+    {
+      "epoch": 0.5601825442099259,
+      "grad_norm": 0.45841656750886967,
+      "learning_rate": 4.517970401691332e-05,
+      "loss": 0.4703,
+      "step": 491
+    },
+    {
+      "epoch": 0.5613234455219623,
+      "grad_norm": 0.45874829723391797,
+      "learning_rate": 4.5158562367864696e-05,
+      "loss": 0.4514,
+      "step": 492
+    },
+    {
+      "epoch": 0.5624643468339988,
+      "grad_norm": 0.37739002522158893,
+      "learning_rate": 4.513742071881607e-05,
+      "loss": 0.441,
+      "step": 493
+    },
+    {
+      "epoch": 0.5636052481460354,
+      "grad_norm": 0.43801802347474533,
+      "learning_rate": 4.5116279069767444e-05,
+      "loss": 0.4535,
+      "step": 494
+    },
+    {
+      "epoch": 0.5647461494580719,
+      "grad_norm": 0.41614070993796926,
+      "learning_rate": 4.509513742071882e-05,
+      "loss": 0.4629,
+      "step": 495
+    },
+    {
+      "epoch": 0.5658870507701084,
+      "grad_norm": 0.4865275595264474,
+      "learning_rate": 4.507399577167019e-05,
+      "loss": 0.4416,
+      "step": 496
+    },
+    {
+      "epoch": 0.5670279520821448,
+      "grad_norm": 0.40655687881400593,
+      "learning_rate": 4.505285412262157e-05,
+      "loss": 0.4574,
+      "step": 497
+    },
+    {
+      "epoch": 0.5681688533941814,
+      "grad_norm": 0.39883282407381804,
+      "learning_rate": 4.503171247357294e-05,
+      "loss": 0.4786,
+      "step": 498
+    },
+    {
+      "epoch": 0.5693097547062179,
+      "grad_norm": 0.37794544376846617,
+      "learning_rate": 4.501057082452432e-05,
+      "loss": 0.435,
+      "step": 499
+    },
+    {
+      "epoch": 0.5704506560182544,
+      "grad_norm": 0.4247854841899775,
+      "learning_rate": 4.498942917547569e-05,
+      "loss": 0.4476,
+      "step": 500
+    },
+    {
+      "epoch": 0.5715915573302909,
+      "grad_norm": 0.43645967584889545,
+      "learning_rate": 4.496828752642706e-05,
+      "loss": 0.4454,
+      "step": 501
+    },
+    {
+      "epoch": 0.5727324586423275,
+      "grad_norm": 0.5187989306988888,
+      "learning_rate": 4.4947145877378436e-05,
+      "loss": 0.4461,
+      "step": 502
+    },
+    {
+      "epoch": 0.573873359954364,
+      "grad_norm": 0.37308352823075,
+      "learning_rate": 4.492600422832981e-05,
+      "loss": 0.4341,
+      "step": 503
+    },
+    {
+      "epoch": 0.5750142612664004,
+      "grad_norm": 0.5932345231021587,
+      "learning_rate": 4.4904862579281184e-05,
+      "loss": 0.4484,
+      "step": 504
+    },
+    {
+      "epoch": 0.576155162578437,
+      "grad_norm": 0.3279197317008337,
+      "learning_rate": 4.488372093023256e-05,
+      "loss": 0.4491,
+      "step": 505
+    },
+    {
+      "epoch": 0.5772960638904735,
+      "grad_norm": 0.6100565413888734,
+      "learning_rate": 4.486257928118394e-05,
+      "loss": 0.4376,
+      "step": 506
+    },
+    {
+      "epoch": 0.57843696520251,
+      "grad_norm": 0.42409976663616067,
+      "learning_rate": 4.484143763213531e-05,
+      "loss": 0.4637,
+      "step": 507
+    },
+    {
+      "epoch": 0.5795778665145465,
+      "grad_norm": 0.47939120843371974,
+      "learning_rate": 4.482029598308669e-05,
+      "loss": 0.4512,
+      "step": 508
+    },
+    {
+      "epoch": 0.580718767826583,
+      "grad_norm": 0.6442312396725929,
+      "learning_rate": 4.479915433403806e-05,
+      "loss": 0.4742,
+      "step": 509
+    },
+    {
+      "epoch": 0.5818596691386195,
+      "grad_norm": 0.44420537485975053,
+      "learning_rate": 4.477801268498943e-05,
+      "loss": 0.4414,
+      "step": 510
+    },
+    {
+      "epoch": 0.583000570450656,
+      "grad_norm": 0.4742282726695118,
+      "learning_rate": 4.4756871035940806e-05,
+      "loss": 0.438,
+      "step": 511
+    },
+    {
+      "epoch": 0.5841414717626925,
+      "grad_norm": 0.38983875431488607,
+      "learning_rate": 4.473572938689218e-05,
+      "loss": 0.4615,
+      "step": 512
+    },
+    {
+      "epoch": 0.5852823730747291,
+      "grad_norm": 0.47490095846340535,
+      "learning_rate": 4.4714587737843555e-05,
+      "loss": 0.4709,
+      "step": 513
+    },
+    {
+      "epoch": 0.5864232743867656,
+      "grad_norm": 0.39696950465421044,
+      "learning_rate": 4.4693446088794925e-05,
+      "loss": 0.4367,
+      "step": 514
+    },
+    {
+      "epoch": 0.587564175698802,
+      "grad_norm": 0.45007345784469194,
+      "learning_rate": 4.46723044397463e-05,
+      "loss": 0.4566,
+      "step": 515
+    },
+    {
+      "epoch": 0.5887050770108385,
+      "grad_norm": 0.4198390996183237,
+      "learning_rate": 4.465116279069767e-05,
+      "loss": 0.4286,
+      "step": 516
+    },
+    {
+      "epoch": 0.5898459783228751,
+      "grad_norm": 0.3779400456562858,
+      "learning_rate": 4.463002114164905e-05,
+      "loss": 0.4437,
+      "step": 517
+    },
+    {
+      "epoch": 0.5909868796349116,
+      "grad_norm": 0.41728607717894356,
+      "learning_rate": 4.460887949260043e-05,
+      "loss": 0.4614,
+      "step": 518
+    },
+    {
+      "epoch": 0.5921277809469481,
+      "grad_norm": 0.39261690895532547,
+      "learning_rate": 4.45877378435518e-05,
+      "loss": 0.4332,
+      "step": 519
+    },
+    {
+      "epoch": 0.5932686822589845,
+      "grad_norm": 0.38020096492819416,
+      "learning_rate": 4.4566596194503176e-05,
+      "loss": 0.4549,
+      "step": 520
+    },
+    {
+      "epoch": 0.5944095835710211,
+      "grad_norm": 0.4517763945083566,
+      "learning_rate": 4.454545454545455e-05,
+      "loss": 0.45,
+      "step": 521
+    },
+    {
+      "epoch": 0.5955504848830576,
+      "grad_norm": 0.3645340466319383,
+      "learning_rate": 4.4524312896405925e-05,
+      "loss": 0.4473,
+      "step": 522
+    },
+    {
+      "epoch": 0.5966913861950941,
+      "grad_norm": 0.355065344453176,
+      "learning_rate": 4.4503171247357295e-05,
+      "loss": 0.4356,
+      "step": 523
+    },
+    {
+      "epoch": 0.5978322875071307,
+      "grad_norm": 0.37890494975171685,
+      "learning_rate": 4.448202959830867e-05,
+      "loss": 0.4399,
+      "step": 524
+    },
+    {
+      "epoch": 0.5989731888191672,
+      "grad_norm": 0.4555107224314838,
+      "learning_rate": 4.4460887949260043e-05,
+      "loss": 0.4521,
+      "step": 525
+    },
+    {
+      "epoch": 0.6001140901312036,
+      "grad_norm": 0.4362531451617325,
+      "learning_rate": 4.443974630021142e-05,
+      "loss": 0.4389,
+      "step": 526
+    },
+    {
+      "epoch": 0.6012549914432401,
+      "grad_norm": 0.4126012364452474,
+      "learning_rate": 4.441860465116279e-05,
+      "loss": 0.4392,
+      "step": 527
+    },
+    {
+      "epoch": 0.6023958927552767,
+      "grad_norm": 0.4250152394750049,
+      "learning_rate": 4.439746300211416e-05,
+      "loss": 0.4513,
+      "step": 528
+    },
+    {
+      "epoch": 0.6035367940673132,
+      "grad_norm": 0.5082523070659859,
+      "learning_rate": 4.437632135306554e-05,
+      "loss": 0.4879,
+      "step": 529
+    },
+    {
+      "epoch": 0.6046776953793497,
+      "grad_norm": 0.4238617257851682,
+      "learning_rate": 4.435517970401691e-05,
+      "loss": 0.4898,
+      "step": 530
+    },
+    {
+      "epoch": 0.6058185966913862,
+      "grad_norm": 0.5041686903889592,
+      "learning_rate": 4.433403805496829e-05,
+      "loss": 0.4484,
+      "step": 531
+    },
+    {
+      "epoch": 0.6069594980034227,
+      "grad_norm": 0.3551559352333077,
+      "learning_rate": 4.4312896405919665e-05,
+      "loss": 0.4463,
+      "step": 532
+    },
+    {
+      "epoch": 0.6081003993154592,
+      "grad_norm": 0.4417858792921517,
+      "learning_rate": 4.429175475687104e-05,
+      "loss": 0.4341,
+      "step": 533
+    },
+    {
+      "epoch": 0.6092413006274957,
+      "grad_norm": 0.36493597171615005,
+      "learning_rate": 4.4270613107822413e-05,
+      "loss": 0.4358,
+      "step": 534
+    },
+    {
+      "epoch": 0.6103822019395322,
+      "grad_norm": 0.4107588079345914,
+      "learning_rate": 4.424947145877379e-05,
+      "loss": 0.4407,
+      "step": 535
+    },
+    {
+      "epoch": 0.6115231032515688,
+      "grad_norm": 0.4463977858001759,
+      "learning_rate": 4.422832980972516e-05,
+      "loss": 0.4586,
+      "step": 536
+    },
+    {
+      "epoch": 0.6126640045636053,
+      "grad_norm": 0.3835290921335774,
+      "learning_rate": 4.420718816067653e-05,
+      "loss": 0.4686,
+      "step": 537
+    },
+    {
+      "epoch": 0.6138049058756417,
+      "grad_norm": 0.4235575449577358,
+      "learning_rate": 4.418604651162791e-05,
+      "loss": 0.4529,
+      "step": 538
+    },
+    {
+      "epoch": 0.6149458071876782,
+      "grad_norm": 0.37221690443646094,
+      "learning_rate": 4.416490486257928e-05,
+      "loss": 0.4449,
+      "step": 539
+    },
+    {
+      "epoch": 0.6160867084997148,
+      "grad_norm": 0.4607925845239333,
+      "learning_rate": 4.414376321353066e-05,
+      "loss": 0.4696,
+      "step": 540
+    },
+    {
+      "epoch": 0.6172276098117513,
+      "grad_norm": 0.4643989526274435,
+      "learning_rate": 4.412262156448203e-05,
+      "loss": 0.4588,
+      "step": 541
+    },
+    {
+      "epoch": 0.6183685111237878,
+      "grad_norm": 0.50099310470488,
+      "learning_rate": 4.4101479915433406e-05,
+      "loss": 0.4585,
+      "step": 542
+    },
+    {
+      "epoch": 0.6195094124358242,
+      "grad_norm": 0.41648393536142087,
+      "learning_rate": 4.408033826638478e-05,
+      "loss": 0.447,
+      "step": 543
+    },
+    {
+      "epoch": 0.6206503137478608,
+      "grad_norm": 0.46190569039774876,
+      "learning_rate": 4.4059196617336154e-05,
+      "loss": 0.4609,
+      "step": 544
+    },
+    {
+      "epoch": 0.6217912150598973,
+      "grad_norm": 0.4862977173283739,
+      "learning_rate": 4.403805496828753e-05,
+      "loss": 0.4375,
+      "step": 545
+    },
+    {
+      "epoch": 0.6229321163719338,
+      "grad_norm": 0.43294207299393156,
+      "learning_rate": 4.40169133192389e-05,
+      "loss": 0.4614,
+      "step": 546
+    },
+    {
+      "epoch": 0.6240730176839704,
+      "grad_norm": 0.43760535192587485,
+      "learning_rate": 4.399577167019028e-05,
+      "loss": 0.4528,
+      "step": 547
+    },
+    {
+      "epoch": 0.6252139189960069,
+      "grad_norm": 0.4247353277953283,
+      "learning_rate": 4.397463002114165e-05,
+      "loss": 0.4612,
+      "step": 548
+    },
+    {
+      "epoch": 0.6263548203080433,
+      "grad_norm": 0.3829123288668259,
+      "learning_rate": 4.395348837209303e-05,
+      "loss": 0.4559,
+      "step": 549
+    },
+    {
+      "epoch": 0.6274957216200798,
+      "grad_norm": 0.4363586414650476,
+      "learning_rate": 4.39323467230444e-05,
+      "loss": 0.4386,
+      "step": 550
+    },
+    {
+      "epoch": 0.6286366229321164,
+      "grad_norm": 0.43682485012615274,
+      "learning_rate": 4.3911205073995776e-05,
+      "loss": 0.4623,
+      "step": 551
+    },
+    {
+      "epoch": 0.6297775242441529,
+      "grad_norm": 0.4105313864106847,
+      "learning_rate": 4.389006342494715e-05,
+      "loss": 0.4538,
+      "step": 552
+    },
+    {
+      "epoch": 0.6309184255561894,
+      "grad_norm": 0.3949382109776014,
+      "learning_rate": 4.3868921775898524e-05,
+      "loss": 0.4633,
+      "step": 553
+    },
+    {
+      "epoch": 0.6320593268682259,
+      "grad_norm": 0.3787686846291949,
+      "learning_rate": 4.3847780126849895e-05,
+      "loss": 0.4492,
+      "step": 554
+    },
+    {
+      "epoch": 0.6332002281802624,
+      "grad_norm": 0.41141740644120794,
+      "learning_rate": 4.3826638477801266e-05,
+      "loss": 0.4499,
+      "step": 555
+    },
+    {
+      "epoch": 0.6343411294922989,
+      "grad_norm": 0.34586933918003787,
+      "learning_rate": 4.380549682875264e-05,
+      "loss": 0.4355,
+      "step": 556
+    },
+    {
+      "epoch": 0.6354820308043354,
+      "grad_norm": 0.3768323855768829,
+      "learning_rate": 4.378435517970402e-05,
+      "loss": 0.466,
+      "step": 557
+    },
+    {
+      "epoch": 0.6366229321163719,
+      "grad_norm": 0.3415297803284673,
+      "learning_rate": 4.37632135306554e-05,
+      "loss": 0.4622,
+      "step": 558
+    },
+    {
+      "epoch": 0.6377638334284085,
+      "grad_norm": 0.38672987304843115,
+      "learning_rate": 4.374207188160677e-05,
+      "loss": 0.4539,
+      "step": 559
+    },
+    {
+      "epoch": 0.638904734740445,
+      "grad_norm": 0.3891339211787994,
+      "learning_rate": 4.3720930232558146e-05,
+      "loss": 0.4655,
+      "step": 560
+    },
+    {
+      "epoch": 0.6400456360524814,
+      "grad_norm": 0.37804765686531433,
+      "learning_rate": 4.369978858350952e-05,
+      "loss": 0.4554,
+      "step": 561
+    },
+    {
+      "epoch": 0.6411865373645179,
+      "grad_norm": 0.41386853119656003,
+      "learning_rate": 4.367864693446089e-05,
+      "loss": 0.4563,
+      "step": 562
+    },
+    {
+      "epoch": 0.6423274386765545,
+      "grad_norm": 0.3575661475152617,
+      "learning_rate": 4.3657505285412265e-05,
+      "loss": 0.4614,
+      "step": 563
+    },
+    {
+      "epoch": 0.643468339988591,
+      "grad_norm": 0.3888937986353574,
+      "learning_rate": 4.3636363636363636e-05,
+      "loss": 0.4419,
+      "step": 564
+    },
+    {
+      "epoch": 0.6446092413006275,
+      "grad_norm": 0.3728029276862445,
+      "learning_rate": 4.361522198731501e-05,
+      "loss": 0.434,
+      "step": 565
+    },
+    {
+      "epoch": 0.6457501426126641,
+      "grad_norm": 0.3175142322947935,
+      "learning_rate": 4.3594080338266384e-05,
+      "loss": 0.4273,
+      "step": 566
+    },
+    {
+      "epoch": 0.6468910439247005,
+      "grad_norm": 0.3594464916075461,
+      "learning_rate": 4.357293868921776e-05,
+      "loss": 0.4175,
+      "step": 567
+    },
+    {
+      "epoch": 0.648031945236737,
+      "grad_norm": 0.31869903166119784,
+      "learning_rate": 4.355179704016913e-05,
+      "loss": 0.4376,
+      "step": 568
+    },
+    {
+      "epoch": 0.6491728465487735,
+      "grad_norm": 0.3775961482431801,
+      "learning_rate": 4.353065539112051e-05,
+      "loss": 0.4469,
+      "step": 569
+    },
+    {
+      "epoch": 0.6503137478608101,
+      "grad_norm": 0.35464274630677184,
+      "learning_rate": 4.350951374207189e-05,
+      "loss": 0.4393,
+      "step": 570
+    },
+    {
+      "epoch": 0.6514546491728466,
+      "grad_norm": 0.37213536538672665,
+      "learning_rate": 4.348837209302326e-05,
+      "loss": 0.4568,
+      "step": 571
+    },
+    {
+      "epoch": 0.652595550484883,
+      "grad_norm": 0.3781132343756537,
+      "learning_rate": 4.3467230443974635e-05,
+      "loss": 0.4535,
+      "step": 572
+    },
+    {
+      "epoch": 0.6537364517969195,
+      "grad_norm": 0.36912322283896176,
+      "learning_rate": 4.3446088794926006e-05,
+      "loss": 0.4391,
+      "step": 573
+    },
+    {
+      "epoch": 0.6548773531089561,
+      "grad_norm": 0.391374326479424,
+      "learning_rate": 4.342494714587738e-05,
+      "loss": 0.4449,
+      "step": 574
+    },
+    {
+      "epoch": 0.6560182544209926,
+      "grad_norm": 0.4036376938390525,
+      "learning_rate": 4.3403805496828754e-05,
+      "loss": 0.4497,
+      "step": 575
+    },
+    {
+      "epoch": 0.6571591557330291,
+      "grad_norm": 0.45153276772782586,
+      "learning_rate": 4.338266384778013e-05,
+      "loss": 0.4557,
+      "step": 576
+    },
+    {
+      "epoch": 0.6583000570450656,
+      "grad_norm": 0.364556752115259,
+      "learning_rate": 4.33615221987315e-05,
+      "loss": 0.437,
+      "step": 577
+    },
+    {
+      "epoch": 0.6594409583571021,
+      "grad_norm": 0.4115989364215942,
+      "learning_rate": 4.334038054968288e-05,
+      "loss": 0.4429,
+      "step": 578
+    },
+    {
+      "epoch": 0.6605818596691386,
+      "grad_norm": 0.34885336248603255,
+      "learning_rate": 4.331923890063425e-05,
+      "loss": 0.4508,
+      "step": 579
+    },
+    {
+      "epoch": 0.6617227609811751,
+      "grad_norm": 0.34730597454686063,
+      "learning_rate": 4.329809725158562e-05,
+      "loss": 0.4549,
+      "step": 580
+    },
+    {
+      "epoch": 0.6628636622932116,
+      "grad_norm": 0.3908862001840564,
+      "learning_rate": 4.3276955602537e-05,
+      "loss": 0.4316,
+      "step": 581
+    },
+    {
+      "epoch": 0.6640045636052482,
+      "grad_norm": 0.3632548361740951,
+      "learning_rate": 4.325581395348837e-05,
+      "loss": 0.4666,
+      "step": 582
+    },
+    {
+      "epoch": 0.6651454649172847,
+      "grad_norm": 0.32265295885901846,
+      "learning_rate": 4.3234672304439746e-05,
+      "loss": 0.416,
+      "step": 583
+    },
+    {
+      "epoch": 0.6662863662293211,
+      "grad_norm": 0.3505405594741678,
+      "learning_rate": 4.3213530655391124e-05,
+      "loss": 0.4508,
+      "step": 584
+    },
+    {
+      "epoch": 0.6674272675413577,
+      "grad_norm": 0.35717088625469834,
+      "learning_rate": 4.31923890063425e-05,
+      "loss": 0.4081,
+      "step": 585
+    },
+    {
+      "epoch": 0.6685681688533942,
+      "grad_norm": 0.3283024234495235,
+      "learning_rate": 4.317124735729387e-05,
+      "loss": 0.4228,
+      "step": 586
+    },
+    {
+      "epoch": 0.6697090701654307,
+      "grad_norm": 0.4160501781111346,
+      "learning_rate": 4.315010570824525e-05,
+      "loss": 0.4404,
+      "step": 587
+    },
+    {
+      "epoch": 0.6708499714774672,
+      "grad_norm": 0.3756742508086247,
+      "learning_rate": 4.312896405919662e-05,
+      "loss": 0.426,
+      "step": 588
+    },
+    {
+      "epoch": 0.6719908727895038,
+      "grad_norm": 0.40139465792011,
+      "learning_rate": 4.310782241014799e-05,
+      "loss": 0.4593,
+      "step": 589
+    },
+    {
+      "epoch": 0.6731317741015402,
+      "grad_norm": 0.4386921676603729,
+      "learning_rate": 4.308668076109937e-05,
+      "loss": 0.4357,
+      "step": 590
+    },
+    {
+      "epoch": 0.6742726754135767,
+      "grad_norm": 0.3823795585192541,
+      "learning_rate": 4.306553911205074e-05,
+      "loss": 0.4359,
+      "step": 591
+    },
+    {
+      "epoch": 0.6754135767256132,
+      "grad_norm": 0.43423365898558286,
+      "learning_rate": 4.3044397463002116e-05,
+      "loss": 0.429,
+      "step": 592
+    },
+    {
+      "epoch": 0.6765544780376498,
+      "grad_norm": 0.3641873671007196,
+      "learning_rate": 4.302325581395349e-05,
+      "loss": 0.4553,
+      "step": 593
+    },
+    {
+      "epoch": 0.6776953793496863,
+      "grad_norm": 0.4026856912219095,
+      "learning_rate": 4.3002114164904865e-05,
+      "loss": 0.4276,
+      "step": 594
+    },
+    {
+      "epoch": 0.6788362806617227,
+      "grad_norm": 0.3701040977337515,
+      "learning_rate": 4.2980972515856235e-05,
+      "loss": 0.4598,
+      "step": 595
+    },
+    {
+      "epoch": 0.6799771819737592,
+      "grad_norm": 0.36586205284253925,
+      "learning_rate": 4.295983086680761e-05,
+      "loss": 0.4536,
+      "step": 596
+    },
+    {
+      "epoch": 0.6811180832857958,
+      "grad_norm": 0.4076865049400923,
+      "learning_rate": 4.293868921775899e-05,
+      "loss": 0.4317,
+      "step": 597
+    },
+    {
+      "epoch": 0.6822589845978323,
+      "grad_norm": 0.3642152640530504,
+      "learning_rate": 4.291754756871036e-05,
+      "loss": 0.4408,
+      "step": 598
+    },
+    {
+      "epoch": 0.6833998859098688,
+      "grad_norm": 0.3863091802215573,
+      "learning_rate": 4.289640591966174e-05,
+      "loss": 0.4371,
+      "step": 599
+    },
+    {
+      "epoch": 0.6845407872219053,
+      "grad_norm": 0.3639715030666885,
+      "learning_rate": 4.287526427061311e-05,
+      "loss": 0.4774,
+      "step": 600
+    },
+    {
+      "epoch": 0.6856816885339418,
+      "grad_norm": 0.45375108529568486,
+      "learning_rate": 4.2854122621564486e-05,
+      "loss": 0.437,
+      "step": 601
+    },
+    {
+      "epoch": 0.6868225898459783,
+      "grad_norm": 0.3575285998180481,
+      "learning_rate": 4.283298097251586e-05,
+      "loss": 0.4373,
+      "step": 602
+    },
+    {
+      "epoch": 0.6879634911580148,
+      "grad_norm": 0.4316339298329872,
+      "learning_rate": 4.2811839323467235e-05,
+      "loss": 0.4435,
+      "step": 603
+    },
+    {
+      "epoch": 0.6891043924700513,
+      "grad_norm": 0.3982746161067394,
+      "learning_rate": 4.2790697674418605e-05,
+      "loss": 0.448,
+      "step": 604
+    },
+    {
+      "epoch": 0.6902452937820879,
+      "grad_norm": 0.4099138290097522,
+      "learning_rate": 4.276955602536998e-05,
+      "loss": 0.4492,
+      "step": 605
+    },
+    {
+      "epoch": 0.6913861950941244,
+      "grad_norm": 0.4411788104044321,
+      "learning_rate": 4.2748414376321353e-05,
+      "loss": 0.4558,
+      "step": 606
+    },
+    {
+      "epoch": 0.6925270964061608,
+      "grad_norm": 0.3529679384408867,
+      "learning_rate": 4.2727272727272724e-05,
+      "loss": 0.4384,
+      "step": 607
+    },
+    {
+      "epoch": 0.6936679977181974,
+      "grad_norm": 0.5228971823650882,
+      "learning_rate": 4.27061310782241e-05,
+      "loss": 0.4532,
+      "step": 608
+    },
+    {
+      "epoch": 0.6948088990302339,
+      "grad_norm": 0.4448630770296031,
+      "learning_rate": 4.268498942917548e-05,
+      "loss": 0.4329,
+      "step": 609
+    },
+    {
+      "epoch": 0.6959498003422704,
+      "grad_norm": 0.45059670019551196,
+      "learning_rate": 4.2663847780126857e-05,
+      "loss": 0.4423,
+      "step": 610
+    },
+    {
+      "epoch": 0.6970907016543069,
+      "grad_norm": 0.4150170695256607,
+      "learning_rate": 4.264270613107823e-05,
+      "loss": 0.4394,
+      "step": 611
+    },
+    {
+      "epoch": 0.6982316029663435,
+      "grad_norm": 0.4796338142973703,
+      "learning_rate": 4.2621564482029605e-05,
+      "loss": 0.4503,
+      "step": 612
+    },
+    {
+      "epoch": 0.6993725042783799,
+      "grad_norm": 0.44780277545991026,
+      "learning_rate": 4.2600422832980975e-05,
+      "loss": 0.4455,
+      "step": 613
+    },
+    {
+      "epoch": 0.7005134055904164,
+      "grad_norm": 0.3768482859123564,
+      "learning_rate": 4.257928118393235e-05,
+      "loss": 0.4379,
+      "step": 614
+    },
+    {
+      "epoch": 0.7016543069024529,
+      "grad_norm": 0.5435524540760096,
+      "learning_rate": 4.2558139534883724e-05,
+      "loss": 0.4411,
+      "step": 615
+    },
+    {
+      "epoch": 0.7027952082144895,
+      "grad_norm": 0.34397188124743944,
+      "learning_rate": 4.2536997885835094e-05,
+      "loss": 0.4418,
+      "step": 616
+    },
+    {
+      "epoch": 0.703936109526526,
+      "grad_norm": 0.40313198481577966,
+      "learning_rate": 4.251585623678647e-05,
+      "loss": 0.4481,
+      "step": 617
+    },
+    {
+      "epoch": 0.7050770108385624,
+      "grad_norm": 0.3711854195341639,
+      "learning_rate": 4.249471458773784e-05,
+      "loss": 0.4185,
+      "step": 618
+    },
+    {
+      "epoch": 0.7062179121505989,
+      "grad_norm": 0.437619206266504,
+      "learning_rate": 4.247357293868922e-05,
+      "loss": 0.4521,
+      "step": 619
+    },
+    {
+      "epoch": 0.7073588134626355,
+      "grad_norm": 0.3930220809028567,
+      "learning_rate": 4.245243128964059e-05,
+      "loss": 0.4405,
+      "step": 620
+    },
+    {
+      "epoch": 0.708499714774672,
+      "grad_norm": 0.38247345005779415,
+      "learning_rate": 4.243128964059197e-05,
+      "loss": 0.4208,
+      "step": 621
+    },
+    {
+      "epoch": 0.7096406160867085,
+      "grad_norm": 0.3437369531927041,
+      "learning_rate": 4.241014799154334e-05,
+      "loss": 0.4266,
+      "step": 622
+    },
+    {
+      "epoch": 0.710781517398745,
+      "grad_norm": 0.386284137063292,
+      "learning_rate": 4.2389006342494716e-05,
+      "loss": 0.4654,
+      "step": 623
+    },
+    {
+      "epoch": 0.7119224187107815,
+      "grad_norm": 0.41209070817450555,
+      "learning_rate": 4.2367864693446094e-05,
+      "loss": 0.4607,
+      "step": 624
+    },
+    {
+      "epoch": 0.713063320022818,
+      "grad_norm": 0.3597165445736412,
+      "learning_rate": 4.2346723044397464e-05,
+      "loss": 0.4375,
+      "step": 625
+    },
+    {
+      "epoch": 0.7142042213348545,
+      "grad_norm": 0.408138228545156,
+      "learning_rate": 4.232558139534884e-05,
+      "loss": 0.4502,
+      "step": 626
+    },
+    {
+      "epoch": 0.7153451226468911,
+      "grad_norm": 0.36219755317558133,
+      "learning_rate": 4.230443974630021e-05,
+      "loss": 0.4339,
+      "step": 627
+    },
+    {
+      "epoch": 0.7164860239589276,
+      "grad_norm": 0.46705886769040666,
+      "learning_rate": 4.228329809725159e-05,
+      "loss": 0.4471,
+      "step": 628
+    },
+    {
+      "epoch": 0.7176269252709641,
+      "grad_norm": 0.3737277765853678,
+      "learning_rate": 4.226215644820296e-05,
+      "loss": 0.4381,
+      "step": 629
+    },
+    {
+      "epoch": 0.7187678265830005,
+      "grad_norm": 0.45979676093246313,
+      "learning_rate": 4.224101479915434e-05,
+      "loss": 0.4127,
+      "step": 630
+    },
+    {
+      "epoch": 0.7199087278950371,
+      "grad_norm": 0.3527512994318668,
+      "learning_rate": 4.221987315010571e-05,
+      "loss": 0.4266,
+      "step": 631
+    },
+    {
+      "epoch": 0.7210496292070736,
+      "grad_norm": 0.45287617912713696,
+      "learning_rate": 4.2198731501057086e-05,
+      "loss": 0.4532,
+      "step": 632
+    },
+    {
+      "epoch": 0.7221905305191101,
+      "grad_norm": 0.4378159828504365,
+      "learning_rate": 4.217758985200846e-05,
+      "loss": 0.452,
+      "step": 633
+    },
+    {
+      "epoch": 0.7233314318311466,
+      "grad_norm": 0.3612319011735092,
+      "learning_rate": 4.215644820295983e-05,
+      "loss": 0.4402,
+      "step": 634
+    },
+    {
+      "epoch": 0.7244723331431832,
+      "grad_norm": 0.5319032554121892,
+      "learning_rate": 4.2135306553911205e-05,
+      "loss": 0.4515,
+      "step": 635
+    },
+    {
+      "epoch": 0.7256132344552196,
+      "grad_norm": 0.36897795859259513,
+      "learning_rate": 4.211416490486258e-05,
+      "loss": 0.4321,
+      "step": 636
+    },
+    {
+      "epoch": 0.7267541357672561,
+      "grad_norm": 0.6220522354879798,
+      "learning_rate": 4.209302325581396e-05,
+      "loss": 0.49,
+      "step": 637
+    },
+    {
+      "epoch": 0.7278950370792926,
+      "grad_norm": 0.34250504144074906,
+      "learning_rate": 4.207188160676533e-05,
+      "loss": 0.4517,
+      "step": 638
+    },
+    {
+      "epoch": 0.7290359383913292,
+      "grad_norm": 0.4460241276659019,
+      "learning_rate": 4.205073995771671e-05,
+      "loss": 0.442,
+      "step": 639
+    },
+    {
+      "epoch": 0.7301768397033657,
+      "grad_norm": 0.4474909489048968,
+      "learning_rate": 4.202959830866808e-05,
+      "loss": 0.444,
+      "step": 640
+    },
+    {
+      "epoch": 0.7313177410154021,
+      "grad_norm": 0.3577226115951208,
+      "learning_rate": 4.2008456659619456e-05,
+      "loss": 0.4335,
+      "step": 641
+    },
+    {
+      "epoch": 0.7324586423274386,
+      "grad_norm": 0.4436466435939107,
+      "learning_rate": 4.198731501057083e-05,
+      "loss": 0.4245,
+      "step": 642
+    },
+    {
+      "epoch": 0.7335995436394752,
+      "grad_norm": 0.3619666710909827,
+      "learning_rate": 4.19661733615222e-05,
+      "loss": 0.4435,
+      "step": 643
+    },
+    {
+      "epoch": 0.7347404449515117,
+      "grad_norm": 0.4557130869879469,
+      "learning_rate": 4.1945031712473575e-05,
+      "loss": 0.4601,
+      "step": 644
+    },
+    {
+      "epoch": 0.7358813462635482,
+      "grad_norm": 0.3990712259993655,
+      "learning_rate": 4.1923890063424946e-05,
+      "loss": 0.442,
+      "step": 645
+    },
+    {
+      "epoch": 0.7370222475755848,
+      "grad_norm": 0.35936180822504243,
+      "learning_rate": 4.190274841437632e-05,
+      "loss": 0.4258,
+      "step": 646
+    },
+    {
+      "epoch": 0.7381631488876212,
+      "grad_norm": 0.4412671459103769,
+      "learning_rate": 4.1881606765327694e-05,
+      "loss": 0.4366,
+      "step": 647
+    },
+    {
+      "epoch": 0.7393040501996577,
+      "grad_norm": 0.3480605482648498,
+      "learning_rate": 4.186046511627907e-05,
+      "loss": 0.424,
+      "step": 648
+    },
+    {
+      "epoch": 0.7404449515116942,
+      "grad_norm": 0.4545769918829186,
+      "learning_rate": 4.183932346723045e-05,
+      "loss": 0.4431,
+      "step": 649
+    },
+    {
+      "epoch": 0.7415858528237308,
+      "grad_norm": 0.38397067925303624,
+      "learning_rate": 4.181818181818182e-05,
+      "loss": 0.4333,
+      "step": 650
+    },
+    {
+      "epoch": 0.7427267541357673,
+      "grad_norm": 0.4006047262191059,
+      "learning_rate": 4.17970401691332e-05,
+      "loss": 0.4293,
+      "step": 651
+    },
+    {
+      "epoch": 0.7438676554478038,
+      "grad_norm": 0.33535629621116786,
+      "learning_rate": 4.177589852008457e-05,
+      "loss": 0.4369,
+      "step": 652
+    },
+    {
+      "epoch": 0.7450085567598402,
+      "grad_norm": 0.4430399993047538,
+      "learning_rate": 4.1754756871035945e-05,
+      "loss": 0.459,
+      "step": 653
+    },
+    {
+      "epoch": 0.7461494580718768,
+      "grad_norm": 0.48222782093736927,
+      "learning_rate": 4.1733615221987316e-05,
+      "loss": 0.457,
+      "step": 654
+    },
+    {
+      "epoch": 0.7472903593839133,
+      "grad_norm": 0.38117402510092147,
+      "learning_rate": 4.171247357293869e-05,
+      "loss": 0.4575,
+      "step": 655
+    },
+    {
+      "epoch": 0.7484312606959498,
+      "grad_norm": 0.5246051148892976,
+      "learning_rate": 4.1691331923890064e-05,
+      "loss": 0.4431,
+      "step": 656
+    },
+    {
+      "epoch": 0.7495721620079863,
+      "grad_norm": 0.34928037914201676,
+      "learning_rate": 4.167019027484144e-05,
+      "loss": 0.4386,
+      "step": 657
+    },
+    {
+      "epoch": 0.7507130633200229,
+      "grad_norm": 0.48475845301346415,
+      "learning_rate": 4.164904862579281e-05,
+      "loss": 0.4512,
+      "step": 658
+    },
+    {
+      "epoch": 0.7518539646320593,
+      "grad_norm": 0.33139730820658986,
+      "learning_rate": 4.162790697674418e-05,
+      "loss": 0.422,
+      "step": 659
+    },
+    {
+      "epoch": 0.7529948659440958,
+      "grad_norm": 0.46025984093444033,
+      "learning_rate": 4.160676532769556e-05,
+      "loss": 0.4546,
+      "step": 660
+    },
+    {
+      "epoch": 0.7541357672561323,
+      "grad_norm": 0.43742732367054127,
+      "learning_rate": 4.158562367864694e-05,
+      "loss": 0.4365,
+      "step": 661
+    },
+    {
+      "epoch": 0.7552766685681689,
+      "grad_norm": 0.39466973998333554,
+      "learning_rate": 4.1564482029598315e-05,
+      "loss": 0.447,
+      "step": 662
+    },
+    {
+      "epoch": 0.7564175698802054,
+      "grad_norm": 0.4164710504296488,
+      "learning_rate": 4.1543340380549686e-05,
+      "loss": 0.4341,
+      "step": 663
+    },
+    {
+      "epoch": 0.7575584711922418,
+      "grad_norm": 0.3445633470818866,
+      "learning_rate": 4.152219873150106e-05,
+      "loss": 0.461,
+      "step": 664
+    },
+    {
+      "epoch": 0.7586993725042784,
+      "grad_norm": 0.3580827508193947,
+      "learning_rate": 4.1501057082452434e-05,
+      "loss": 0.4172,
+      "step": 665
+    },
+    {
+      "epoch": 0.7598402738163149,
+      "grad_norm": 0.3676001265506075,
+      "learning_rate": 4.147991543340381e-05,
+      "loss": 0.4428,
+      "step": 666
+    },
+    {
+      "epoch": 0.7609811751283514,
+      "grad_norm": 0.3505942690631605,
+      "learning_rate": 4.145877378435518e-05,
+      "loss": 0.4432,
+      "step": 667
+    },
+    {
+      "epoch": 0.7621220764403879,
+      "grad_norm": 0.3891067229029948,
+      "learning_rate": 4.143763213530655e-05,
+      "loss": 0.4467,
+      "step": 668
+    },
+    {
+      "epoch": 0.7632629777524245,
+      "grad_norm": 0.3734083446553611,
+      "learning_rate": 4.141649048625793e-05,
+      "loss": 0.4269,
+      "step": 669
+    },
+    {
+      "epoch": 0.764403879064461,
+      "grad_norm": 0.39030324929612525,
+      "learning_rate": 4.13953488372093e-05,
+      "loss": 0.4421,
+      "step": 670
+    },
+    {
+      "epoch": 0.7655447803764974,
+      "grad_norm": 0.42958890638506364,
+      "learning_rate": 4.137420718816068e-05,
+      "loss": 0.4062,
+      "step": 671
+    },
+    {
+      "epoch": 0.7666856816885339,
+      "grad_norm": 0.30539972744468097,
+      "learning_rate": 4.135306553911205e-05,
+      "loss": 0.4479,
+      "step": 672
+    },
+    {
+      "epoch": 0.7678265830005705,
+      "grad_norm": 0.3702174124250884,
+      "learning_rate": 4.1331923890063427e-05,
+      "loss": 0.435,
+      "step": 673
+    },
+    {
+      "epoch": 0.768967484312607,
+      "grad_norm": 0.33009900205382475,
+      "learning_rate": 4.13107822410148e-05,
+      "loss": 0.4345,
+      "step": 674
+    },
+    {
+      "epoch": 0.7701083856246435,
+      "grad_norm": 0.3245007702409929,
+      "learning_rate": 4.1289640591966175e-05,
+      "loss": 0.4453,
+      "step": 675
+    },
+    {
+      "epoch": 0.7712492869366799,
+      "grad_norm": 0.40666300271812594,
+      "learning_rate": 4.126849894291755e-05,
+      "loss": 0.4605,
+      "step": 676
+    },
+    {
+      "epoch": 0.7723901882487165,
+      "grad_norm": 0.36635310340240823,
+      "learning_rate": 4.124735729386892e-05,
+      "loss": 0.4264,
+      "step": 677
+    },
+    {
+      "epoch": 0.773531089560753,
+      "grad_norm": 0.4299812969188949,
+      "learning_rate": 4.12262156448203e-05,
+      "loss": 0.4224,
+      "step": 678
+    },
+    {
+      "epoch": 0.7746719908727895,
+      "grad_norm": 0.359428711940889,
+      "learning_rate": 4.120507399577167e-05,
+      "loss": 0.4417,
+      "step": 679
+    },
+    {
+      "epoch": 0.775812892184826,
+      "grad_norm": 0.4677963722655903,
+      "learning_rate": 4.118393234672305e-05,
+      "loss": 0.4574,
+      "step": 680
+    },
+    {
+      "epoch": 0.7769537934968626,
+      "grad_norm": 0.4233752207812584,
+      "learning_rate": 4.116279069767442e-05,
+      "loss": 0.4481,
+      "step": 681
+    },
+    {
+      "epoch": 0.778094694808899,
+      "grad_norm": 0.3835874030237872,
+      "learning_rate": 4.1141649048625797e-05,
+      "loss": 0.4298,
+      "step": 682
+    },
+    {
+      "epoch": 0.7792355961209355,
+      "grad_norm": 0.37669645948536934,
+      "learning_rate": 4.112050739957717e-05,
+      "loss": 0.4513,
+      "step": 683
+    },
+    {
+      "epoch": 0.780376497432972,
+      "grad_norm": 0.36373603981263314,
+      "learning_rate": 4.1099365750528545e-05,
+      "loss": 0.443,
+      "step": 684
+    },
+    {
+      "epoch": 0.7815173987450086,
+      "grad_norm": 0.3753684899759581,
+      "learning_rate": 4.1078224101479915e-05,
+      "loss": 0.4269,
+      "step": 685
+    },
+    {
+      "epoch": 0.7826583000570451,
+      "grad_norm": 0.3605434844405242,
+      "learning_rate": 4.1057082452431286e-05,
+      "loss": 0.4323,
+      "step": 686
+    },
+    {
+      "epoch": 0.7837992013690815,
+      "grad_norm": 0.3462661479430446,
+      "learning_rate": 4.1035940803382664e-05,
+      "loss": 0.4326,
+      "step": 687
+    },
+    {
+      "epoch": 0.7849401026811181,
+      "grad_norm": 0.4204439751599984,
+      "learning_rate": 4.101479915433404e-05,
+      "loss": 0.4373,
+      "step": 688
+    },
+    {
+      "epoch": 0.7860810039931546,
+      "grad_norm": 0.3280914793709163,
+      "learning_rate": 4.099365750528542e-05,
+      "loss": 0.4535,
+      "step": 689
+    },
+    {
+      "epoch": 0.7872219053051911,
+      "grad_norm": 0.35041349522279136,
+      "learning_rate": 4.097251585623679e-05,
+      "loss": 0.4381,
+      "step": 690
+    },
+    {
+      "epoch": 0.7883628066172276,
+      "grad_norm": 0.3988862381741282,
+      "learning_rate": 4.0951374207188167e-05,
+      "loss": 0.4651,
+      "step": 691
+    },
+    {
+      "epoch": 0.7895037079292642,
+      "grad_norm": 0.3424394514285045,
+      "learning_rate": 4.093023255813954e-05,
+      "loss": 0.4441,
+      "step": 692
+    },
+    {
+      "epoch": 0.7906446092413006,
+      "grad_norm": 0.3490799047610466,
+      "learning_rate": 4.0909090909090915e-05,
+      "loss": 0.4404,
+      "step": 693
+    },
+    {
+      "epoch": 0.7917855105533371,
+      "grad_norm": 0.3503575821053399,
+      "learning_rate": 4.0887949260042285e-05,
+      "loss": 0.425,
+      "step": 694
+    },
+    {
+      "epoch": 0.7929264118653736,
+      "grad_norm": 0.365659614058789,
+      "learning_rate": 4.0866807610993656e-05,
+      "loss": 0.4339,
+      "step": 695
+    },
+    {
+      "epoch": 0.7940673131774102,
+      "grad_norm": 0.39539532500596763,
+      "learning_rate": 4.0845665961945034e-05,
+      "loss": 0.4446,
+      "step": 696
+    },
+    {
+      "epoch": 0.7952082144894467,
+      "grad_norm": 0.3250304775152391,
+      "learning_rate": 4.0824524312896404e-05,
+      "loss": 0.438,
+      "step": 697
+    },
+    {
+      "epoch": 0.7963491158014832,
+      "grad_norm": 0.4909285073320248,
+      "learning_rate": 4.080338266384778e-05,
+      "loss": 0.4463,
+      "step": 698
+    },
+    {
+      "epoch": 0.7974900171135196,
+      "grad_norm": 0.37590033225791325,
+      "learning_rate": 4.078224101479915e-05,
+      "loss": 0.4491,
+      "step": 699
+    },
+    {
+      "epoch": 0.7986309184255562,
+      "grad_norm": 0.45070180362214113,
+      "learning_rate": 4.076109936575053e-05,
+      "loss": 0.4299,
+      "step": 700
+    },
+    {
+      "epoch": 0.7997718197375927,
+      "grad_norm": 0.4251332388315139,
+      "learning_rate": 4.073995771670191e-05,
+      "loss": 0.4571,
+      "step": 701
+    },
+    {
+      "epoch": 0.8009127210496292,
+      "grad_norm": 0.48990786595017877,
+      "learning_rate": 4.0718816067653285e-05,
+      "loss": 0.4453,
+      "step": 702
+    },
+    {
+      "epoch": 0.8020536223616657,
+      "grad_norm": 0.5002547392375538,
+      "learning_rate": 4.0697674418604655e-05,
+      "loss": 0.4207,
+      "step": 703
+    },
+    {
+      "epoch": 0.8031945236737023,
+      "grad_norm": 0.4659070323210945,
+      "learning_rate": 4.0676532769556026e-05,
+      "loss": 0.4535,
+      "step": 704
+    },
+    {
+      "epoch": 0.8043354249857387,
+      "grad_norm": 0.5731835975203691,
+      "learning_rate": 4.0655391120507404e-05,
+      "loss": 0.4349,
+      "step": 705
+    },
+    {
+      "epoch": 0.8054763262977752,
+      "grad_norm": 0.4228304566829878,
+      "learning_rate": 4.0634249471458774e-05,
+      "loss": 0.4291,
+      "step": 706
+    },
+    {
+      "epoch": 0.8066172276098118,
+      "grad_norm": 0.5720554991869126,
+      "learning_rate": 4.061310782241015e-05,
+      "loss": 0.4554,
+      "step": 707
+    },
+    {
+      "epoch": 0.8077581289218483,
+      "grad_norm": 0.4853803231526458,
+      "learning_rate": 4.059196617336152e-05,
+      "loss": 0.4836,
+      "step": 708
+    },
+    {
+      "epoch": 0.8088990302338848,
+      "grad_norm": 0.4971208032161558,
+      "learning_rate": 4.05708245243129e-05,
+      "loss": 0.4413,
+      "step": 709
+    },
+    {
+      "epoch": 0.8100399315459212,
+      "grad_norm": 0.5127734596329273,
+      "learning_rate": 4.054968287526427e-05,
+      "loss": 0.4554,
+      "step": 710
+    },
+    {
+      "epoch": 0.8111808328579578,
+      "grad_norm": 0.6714943186997662,
+      "learning_rate": 4.052854122621565e-05,
+      "loss": 0.4447,
+      "step": 711
+    },
+    {
+      "epoch": 0.8123217341699943,
+      "grad_norm": 0.40895342997960105,
+      "learning_rate": 4.050739957716702e-05,
+      "loss": 0.4287,
+      "step": 712
+    },
+    {
+      "epoch": 0.8134626354820308,
+      "grad_norm": 0.5479757534782737,
+      "learning_rate": 4.0486257928118396e-05,
+      "loss": 0.4436,
+      "step": 713
+    },
+    {
+      "epoch": 0.8146035367940673,
+      "grad_norm": 0.5279155713449081,
+      "learning_rate": 4.046511627906977e-05,
+      "loss": 0.4381,
+      "step": 714
+    },
+    {
+      "epoch": 0.8157444381061039,
+      "grad_norm": 0.434094864002774,
+      "learning_rate": 4.0443974630021144e-05,
+      "loss": 0.4305,
+      "step": 715
+    },
+    {
+      "epoch": 0.8168853394181403,
+      "grad_norm": 0.6187261983111648,
+      "learning_rate": 4.042283298097252e-05,
+      "loss": 0.4264,
+      "step": 716
+    },
+    {
+      "epoch": 0.8180262407301768,
+      "grad_norm": 0.3828855462779502,
+      "learning_rate": 4.040169133192389e-05,
+      "loss": 0.4303,
+      "step": 717
+    },
+    {
+      "epoch": 0.8191671420422133,
+      "grad_norm": 0.5189377615161761,
+      "learning_rate": 4.038054968287527e-05,
+      "loss": 0.4384,
+      "step": 718
+    },
+    {
+      "epoch": 0.8203080433542499,
+      "grad_norm": 0.41329860126451856,
+      "learning_rate": 4.035940803382664e-05,
+      "loss": 0.4325,
+      "step": 719
+    },
+    {
+      "epoch": 0.8214489446662864,
+      "grad_norm": 0.3937265367654548,
+      "learning_rate": 4.033826638477802e-05,
+      "loss": 0.422,
+      "step": 720
+    },
+    {
+      "epoch": 0.8225898459783229,
+      "grad_norm": 0.43755932917102885,
+      "learning_rate": 4.031712473572939e-05,
+      "loss": 0.4028,
+      "step": 721
+    },
+    {
+      "epoch": 0.8237307472903593,
+      "grad_norm": 0.4157908856486112,
+      "learning_rate": 4.029598308668076e-05,
+      "loss": 0.438,
+      "step": 722
+    },
+    {
+      "epoch": 0.8248716486023959,
+      "grad_norm": 0.502065224922776,
+      "learning_rate": 4.027484143763214e-05,
+      "loss": 0.4181,
+      "step": 723
+    },
+    {
+      "epoch": 0.8260125499144324,
+      "grad_norm": 0.3780582332426594,
+      "learning_rate": 4.025369978858351e-05,
+      "loss": 0.4201,
+      "step": 724
+    },
+    {
+      "epoch": 0.8271534512264689,
+      "grad_norm": 0.34860648789791515,
+      "learning_rate": 4.0232558139534885e-05,
+      "loss": 0.4546,
+      "step": 725
+    },
+    {
+      "epoch": 0.8282943525385055,
+      "grad_norm": 0.43374943583927567,
+      "learning_rate": 4.0211416490486256e-05,
+      "loss": 0.4166,
+      "step": 726
+    },
+    {
+      "epoch": 0.829435253850542,
+      "grad_norm": 0.34017878511195426,
+      "learning_rate": 4.019027484143763e-05,
+      "loss": 0.4584,
+      "step": 727
+    },
+    {
+      "epoch": 0.8305761551625784,
+      "grad_norm": 0.4520337372094188,
+      "learning_rate": 4.016913319238901e-05,
+      "loss": 0.4515,
+      "step": 728
+    },
+    {
+      "epoch": 0.8317170564746149,
+      "grad_norm": 0.4484495076947604,
+      "learning_rate": 4.014799154334038e-05,
+      "loss": 0.4477,
+      "step": 729
+    },
+    {
+      "epoch": 0.8328579577866515,
+      "grad_norm": 0.39071842952373664,
+      "learning_rate": 4.012684989429176e-05,
+      "loss": 0.44,
+      "step": 730
+    },
+    {
+      "epoch": 0.833998859098688,
+      "grad_norm": 0.4709103934572106,
+      "learning_rate": 4.010570824524313e-05,
+      "loss": 0.4105,
+      "step": 731
+    },
+    {
+      "epoch": 0.8351397604107245,
+      "grad_norm": 0.3693181402492397,
+      "learning_rate": 4.008456659619451e-05,
+      "loss": 0.4423,
+      "step": 732
+    },
+    {
+      "epoch": 0.8362806617227609,
+      "grad_norm": 0.3438695738762642,
+      "learning_rate": 4.006342494714588e-05,
+      "loss": 0.4151,
+      "step": 733
+    },
+    {
+      "epoch": 0.8374215630347975,
+      "grad_norm": 0.3749841228080736,
+      "learning_rate": 4.0042283298097255e-05,
+      "loss": 0.4387,
+      "step": 734
+    },
+    {
+      "epoch": 0.838562464346834,
+      "grad_norm": 0.31093266972310163,
+      "learning_rate": 4.0021141649048626e-05,
+      "loss": 0.4245,
+      "step": 735
+    },
+    {
+      "epoch": 0.8397033656588705,
+      "grad_norm": 0.3842543200844472,
+      "learning_rate": 4e-05,
+      "loss": 0.4361,
+      "step": 736
+    },
+    {
+      "epoch": 0.840844266970907,
+      "grad_norm": 0.3766287686752745,
+      "learning_rate": 3.9978858350951374e-05,
+      "loss": 0.4228,
+      "step": 737
+    },
+    {
+      "epoch": 0.8419851682829436,
+      "grad_norm": 0.35517260008860085,
+      "learning_rate": 3.9957716701902745e-05,
+      "loss": 0.4359,
+      "step": 738
+    },
+    {
+      "epoch": 0.84312606959498,
+      "grad_norm": 0.4419338843198511,
+      "learning_rate": 3.993657505285412e-05,
+      "loss": 0.4304,
+      "step": 739
+    },
+    {
+      "epoch": 0.8442669709070165,
+      "grad_norm": 0.513467064125959,
+      "learning_rate": 3.99154334038055e-05,
+      "loss": 0.4403,
+      "step": 740
+    },
+    {
+      "epoch": 0.845407872219053,
+      "grad_norm": 0.4810139092669358,
+      "learning_rate": 3.989429175475688e-05,
+      "loss": 0.4677,
+      "step": 741
+    },
+    {
+      "epoch": 0.8465487735310896,
+      "grad_norm": 0.39212605153330693,
+      "learning_rate": 3.987315010570825e-05,
+      "loss": 0.4425,
+      "step": 742
+    },
+    {
+      "epoch": 0.8476896748431261,
+      "grad_norm": 0.41040105526904663,
+      "learning_rate": 3.9852008456659625e-05,
+      "loss": 0.4298,
+      "step": 743
+    },
+    {
+      "epoch": 0.8488305761551626,
+      "grad_norm": 0.462499235826406,
+      "learning_rate": 3.9830866807610996e-05,
+      "loss": 0.4075,
+      "step": 744
+    },
+    {
+      "epoch": 0.849971477467199,
+      "grad_norm": 0.38462443248324996,
+      "learning_rate": 3.980972515856237e-05,
+      "loss": 0.4281,
+      "step": 745
+    },
+    {
+      "epoch": 0.8511123787792356,
+      "grad_norm": 0.38604978434276516,
+      "learning_rate": 3.9788583509513744e-05,
+      "loss": 0.4358,
+      "step": 746
+    },
+    {
+      "epoch": 0.8522532800912721,
+      "grad_norm": 0.3984294927193753,
+      "learning_rate": 3.9767441860465115e-05,
+      "loss": 0.4151,
+      "step": 747
+    },
+    {
+      "epoch": 0.8533941814033086,
+      "grad_norm": 0.36801620604521634,
+      "learning_rate": 3.974630021141649e-05,
+      "loss": 0.4293,
+      "step": 748
+    },
+    {
+      "epoch": 0.8545350827153452,
+      "grad_norm": 0.3374397580238608,
+      "learning_rate": 3.972515856236786e-05,
+      "loss": 0.4139,
+      "step": 749
+    },
+    {
+      "epoch": 0.8556759840273817,
+      "grad_norm": 0.42454353852197896,
+      "learning_rate": 3.970401691331924e-05,
+      "loss": 0.4351,
+      "step": 750
+    },
+    {
+      "epoch": 0.8568168853394181,
+      "grad_norm": 0.3940551073101426,
+      "learning_rate": 3.968287526427061e-05,
+      "loss": 0.4324,
+      "step": 751
+    },
+    {
+      "epoch": 0.8579577866514546,
+      "grad_norm": 0.40588237811060596,
+      "learning_rate": 3.966173361522199e-05,
+      "loss": 0.4194,
+      "step": 752
+    },
+    {
+      "epoch": 0.8590986879634912,
+      "grad_norm": 0.420807467022524,
+      "learning_rate": 3.9640591966173366e-05,
+      "loss": 0.4491,
+      "step": 753
+    },
+    {
+      "epoch": 0.8602395892755277,
+      "grad_norm": 0.4540159144239781,
+      "learning_rate": 3.961945031712474e-05,
+      "loss": 0.4092,
+      "step": 754
+    },
+    {
+      "epoch": 0.8613804905875642,
+      "grad_norm": 0.46954467807566896,
+      "learning_rate": 3.9598308668076114e-05,
+      "loss": 0.4504,
+      "step": 755
+    },
+    {
+      "epoch": 0.8625213918996006,
+      "grad_norm": 0.4221414094460434,
+      "learning_rate": 3.9577167019027485e-05,
+      "loss": 0.4364,
+      "step": 756
+    },
+    {
+      "epoch": 0.8636622932116372,
+      "grad_norm": 0.41911229551531937,
+      "learning_rate": 3.955602536997886e-05,
+      "loss": 0.4395,
+      "step": 757
+    },
+    {
+      "epoch": 0.8648031945236737,
+      "grad_norm": 0.4334933603326218,
+      "learning_rate": 3.953488372093023e-05,
+      "loss": 0.4341,
+      "step": 758
+    },
+    {
+      "epoch": 0.8659440958357102,
+      "grad_norm": 0.427182558681861,
+      "learning_rate": 3.951374207188161e-05,
+      "loss": 0.4234,
+      "step": 759
+    },
+    {
+      "epoch": 0.8670849971477467,
+      "grad_norm": 0.4679381773381296,
+      "learning_rate": 3.949260042283298e-05,
+      "loss": 0.4461,
+      "step": 760
+    },
+    {
+      "epoch": 0.8682258984597833,
+      "grad_norm": 0.3997765960963666,
+      "learning_rate": 3.947145877378436e-05,
+      "loss": 0.4553,
+      "step": 761
+    },
+    {
+      "epoch": 0.8693667997718197,
+      "grad_norm": 0.47893488084130953,
+      "learning_rate": 3.945031712473573e-05,
+      "loss": 0.4383,
+      "step": 762
+    },
+    {
+      "epoch": 0.8705077010838562,
+      "grad_norm": 0.387091880809383,
+      "learning_rate": 3.942917547568711e-05,
+      "loss": 0.4231,
+      "step": 763
+    },
+    {
+      "epoch": 0.8716486023958927,
+      "grad_norm": 0.3967940661402007,
+      "learning_rate": 3.940803382663848e-05,
+      "loss": 0.429,
+      "step": 764
+    },
+    {
+      "epoch": 0.8727895037079293,
+      "grad_norm": 0.45115954790258245,
+      "learning_rate": 3.9386892177589855e-05,
+      "loss": 0.4412,
+      "step": 765
+    },
+    {
+      "epoch": 0.8739304050199658,
+      "grad_norm": 0.31431831863394666,
+      "learning_rate": 3.9365750528541225e-05,
+      "loss": 0.4142,
+      "step": 766
+    },
+    {
+      "epoch": 0.8750713063320023,
+      "grad_norm": 0.5271261001088663,
+      "learning_rate": 3.93446088794926e-05,
+      "loss": 0.4582,
+      "step": 767
+    },
+    {
+      "epoch": 0.8762122076440388,
+      "grad_norm": 0.37960216759246174,
+      "learning_rate": 3.932346723044398e-05,
+      "loss": 0.4451,
+      "step": 768
+    },
+    {
+      "epoch": 0.8773531089560753,
+      "grad_norm": 0.5153918934168993,
+      "learning_rate": 3.930232558139535e-05,
+      "loss": 0.4421,
+      "step": 769
+    },
+    {
+      "epoch": 0.8784940102681118,
+      "grad_norm": 0.40777595518236365,
+      "learning_rate": 3.928118393234673e-05,
+      "loss": 0.4329,
+      "step": 770
+    },
+    {
+      "epoch": 0.8796349115801483,
+      "grad_norm": 0.4054358265499145,
+      "learning_rate": 3.92600422832981e-05,
+      "loss": 0.4492,
+      "step": 771
+    },
+    {
+      "epoch": 0.8807758128921849,
+      "grad_norm": 0.45758804362718547,
+      "learning_rate": 3.923890063424948e-05,
+      "loss": 0.4491,
+      "step": 772
+    },
+    {
+      "epoch": 0.8819167142042214,
+      "grad_norm": 0.4327636489834873,
+      "learning_rate": 3.921775898520085e-05,
+      "loss": 0.419,
+      "step": 773
+    },
+    {
+      "epoch": 0.8830576155162578,
+      "grad_norm": 0.5308983247497958,
+      "learning_rate": 3.919661733615222e-05,
+      "loss": 0.4195,
+      "step": 774
+    },
+    {
+      "epoch": 0.8841985168282943,
+      "grad_norm": 0.44186009558843903,
+      "learning_rate": 3.9175475687103596e-05,
+      "loss": 0.422,
+      "step": 775
+    },
+    {
+      "epoch": 0.8853394181403309,
+      "grad_norm": 0.49860412199875853,
+      "learning_rate": 3.9154334038054966e-05,
+      "loss": 0.4508,
+      "step": 776
+    },
+    {
+      "epoch": 0.8864803194523674,
+      "grad_norm": 0.5386256866496492,
+      "learning_rate": 3.9133192389006344e-05,
+      "loss": 0.4199,
+      "step": 777
+    },
+    {
+      "epoch": 0.8876212207644039,
+      "grad_norm": 0.42135087233790125,
+      "learning_rate": 3.9112050739957714e-05,
+      "loss": 0.439,
+      "step": 778
+    },
+    {
+      "epoch": 0.8887621220764403,
+      "grad_norm": 0.47429413048860186,
+      "learning_rate": 3.909090909090909e-05,
+      "loss": 0.4075,
+      "step": 779
+    },
+    {
+      "epoch": 0.8899030233884769,
+      "grad_norm": 0.507620116383454,
+      "learning_rate": 3.906976744186047e-05,
+      "loss": 0.4559,
+      "step": 780
+    },
+    {
+      "epoch": 0.8910439247005134,
+      "grad_norm": 0.4130323507865215,
+      "learning_rate": 3.904862579281185e-05,
+      "loss": 0.4335,
+      "step": 781
+    },
+    {
+      "epoch": 0.8921848260125499,
+      "grad_norm": 0.4919769248710124,
+      "learning_rate": 3.902748414376322e-05,
+      "loss": 0.4326,
+      "step": 782
+    },
+    {
+      "epoch": 0.8933257273245864,
+      "grad_norm": 0.3618438803737931,
+      "learning_rate": 3.900634249471459e-05,
+      "loss": 0.4149,
+      "step": 783
+    },
+    {
+      "epoch": 0.894466628636623,
+      "grad_norm": 0.3766229644972282,
+      "learning_rate": 3.8985200845665966e-05,
+      "loss": 0.4413,
+      "step": 784
+    },
+    {
+      "epoch": 0.8956075299486594,
+      "grad_norm": 0.40065923481523696,
+      "learning_rate": 3.8964059196617336e-05,
+      "loss": 0.4256,
+      "step": 785
+    },
+    {
+      "epoch": 0.8967484312606959,
+      "grad_norm": 0.38682593556698847,
+      "learning_rate": 3.8942917547568714e-05,
+      "loss": 0.4383,
+      "step": 786
+    },
+    {
+      "epoch": 0.8978893325727325,
+      "grad_norm": 0.36589625639407924,
+      "learning_rate": 3.8921775898520084e-05,
+      "loss": 0.3975,
+      "step": 787
+    },
+    {
+      "epoch": 0.899030233884769,
+      "grad_norm": 0.4992171026288852,
+      "learning_rate": 3.890063424947146e-05,
+      "loss": 0.4273,
+      "step": 788
+    },
+    {
+      "epoch": 0.9001711351968055,
+      "grad_norm": 0.34110261273539755,
+      "learning_rate": 3.887949260042283e-05,
+      "loss": 0.4338,
+      "step": 789
+    },
+    {
+      "epoch": 0.901312036508842,
+      "grad_norm": 0.5241888214017549,
+      "learning_rate": 3.885835095137421e-05,
+      "loss": 0.4486,
+      "step": 790
+    },
+    {
+      "epoch": 0.9024529378208785,
+      "grad_norm": 0.46893161994221105,
+      "learning_rate": 3.883720930232558e-05,
+      "loss": 0.4326,
+      "step": 791
+    },
+    {
+      "epoch": 0.903593839132915,
+      "grad_norm": 0.42273170967522355,
+      "learning_rate": 3.881606765327696e-05,
+      "loss": 0.4277,
+      "step": 792
+    },
+    {
+      "epoch": 0.9047347404449515,
+      "grad_norm": 0.4549507679298303,
+      "learning_rate": 3.8794926004228336e-05,
+      "loss": 0.4494,
+      "step": 793
+    },
+    {
+      "epoch": 0.905875641756988,
+      "grad_norm": 0.381434633502494,
+      "learning_rate": 3.8773784355179706e-05,
+      "loss": 0.4204,
+      "step": 794
+    },
+    {
+      "epoch": 0.9070165430690246,
+      "grad_norm": 0.449182413711188,
+      "learning_rate": 3.8752642706131084e-05,
+      "loss": 0.4151,
+      "step": 795
+    },
+    {
+      "epoch": 0.908157444381061,
+      "grad_norm": 0.39996145027480806,
+      "learning_rate": 3.8731501057082454e-05,
+      "loss": 0.4182,
+      "step": 796
+    },
+    {
+      "epoch": 0.9092983456930975,
+      "grad_norm": 0.4239131825637012,
+      "learning_rate": 3.871035940803383e-05,
+      "loss": 0.4333,
+      "step": 797
+    },
+    {
+      "epoch": 0.910439247005134,
+      "grad_norm": 0.38140086906401804,
+      "learning_rate": 3.86892177589852e-05,
+      "loss": 0.4263,
+      "step": 798
+    },
+    {
+      "epoch": 0.9115801483171706,
+      "grad_norm": 0.4151234690517119,
+      "learning_rate": 3.866807610993658e-05,
+      "loss": 0.3942,
+      "step": 799
+    },
+    {
+      "epoch": 0.9127210496292071,
+      "grad_norm": 0.48901591188294186,
+      "learning_rate": 3.864693446088795e-05,
+      "loss": 0.4268,
+      "step": 800
+    },
+    {
+      "epoch": 0.9138619509412436,
+      "grad_norm": 0.4229848535149157,
+      "learning_rate": 3.862579281183932e-05,
+      "loss": 0.4217,
+      "step": 801
+    },
+    {
+      "epoch": 0.91500285225328,
+      "grad_norm": 0.45973123150821005,
+      "learning_rate": 3.86046511627907e-05,
+      "loss": 0.458,
+      "step": 802
+    },
+    {
+      "epoch": 0.9161437535653166,
+      "grad_norm": 0.4609648365948717,
+      "learning_rate": 3.858350951374207e-05,
+      "loss": 0.4307,
+      "step": 803
+    },
+    {
+      "epoch": 0.9172846548773531,
+      "grad_norm": 0.36201542515906165,
+      "learning_rate": 3.856236786469345e-05,
+      "loss": 0.415,
+      "step": 804
+    },
+    {
+      "epoch": 0.9184255561893896,
+      "grad_norm": 0.4138868996856232,
+      "learning_rate": 3.8541226215644824e-05,
+      "loss": 0.4219,
+      "step": 805
+    },
+    {
+      "epoch": 0.9195664575014262,
+      "grad_norm": 0.425798724139904,
+      "learning_rate": 3.8520084566596195e-05,
+      "loss": 0.4324,
+      "step": 806
+    },
+    {
+      "epoch": 0.9207073588134627,
+      "grad_norm": 0.44617326124449147,
+      "learning_rate": 3.849894291754757e-05,
+      "loss": 0.4272,
+      "step": 807
+    },
+    {
+      "epoch": 0.9218482601254991,
+      "grad_norm": 0.3669146140612319,
+      "learning_rate": 3.847780126849895e-05,
+      "loss": 0.4435,
+      "step": 808
+    },
+    {
+      "epoch": 0.9229891614375356,
+      "grad_norm": 0.5009260847096676,
+      "learning_rate": 3.845665961945032e-05,
+      "loss": 0.4495,
+      "step": 809
+    },
+    {
+      "epoch": 0.9241300627495722,
+      "grad_norm": 0.304056760658916,
+      "learning_rate": 3.843551797040169e-05,
+      "loss": 0.4315,
+      "step": 810
+    },
+    {
+      "epoch": 0.9252709640616087,
+      "grad_norm": 0.4554386377496836,
+      "learning_rate": 3.841437632135307e-05,
+      "loss": 0.4365,
+      "step": 811
+    },
+    {
+      "epoch": 0.9264118653736452,
+      "grad_norm": 0.3152473341936204,
+      "learning_rate": 3.839323467230444e-05,
+      "loss": 0.4526,
+      "step": 812
+    },
+    {
+      "epoch": 0.9275527666856817,
+      "grad_norm": 0.3689287886745172,
+      "learning_rate": 3.837209302325582e-05,
+      "loss": 0.4277,
+      "step": 813
+    },
+    {
+      "epoch": 0.9286936679977182,
+      "grad_norm": 0.38617380995316614,
+      "learning_rate": 3.835095137420719e-05,
+      "loss": 0.4392,
+      "step": 814
+    },
+    {
+      "epoch": 0.9298345693097547,
+      "grad_norm": 0.31250501424875,
+      "learning_rate": 3.8329809725158565e-05,
+      "loss": 0.4252,
+      "step": 815
+    },
+    {
+      "epoch": 0.9309754706217912,
+      "grad_norm": 0.40260110774923935,
+      "learning_rate": 3.8308668076109936e-05,
+      "loss": 0.4337,
+      "step": 816
+    },
+    {
+      "epoch": 0.9321163719338277,
+      "grad_norm": 0.3434225940584785,
+      "learning_rate": 3.8287526427061307e-05,
+      "loss": 0.4188,
+      "step": 817
+    },
+    {
+      "epoch": 0.9332572732458643,
+      "grad_norm": 0.4583302147431869,
+      "learning_rate": 3.8266384778012684e-05,
+      "loss": 0.428,
+      "step": 818
+    },
+    {
+      "epoch": 0.9343981745579008,
+      "grad_norm": 0.4823100075277545,
+      "learning_rate": 3.824524312896406e-05,
+      "loss": 0.435,
+      "step": 819
+    },
+    {
+      "epoch": 0.9355390758699372,
+      "grad_norm": 0.3756749669723151,
+      "learning_rate": 3.822410147991544e-05,
+      "loss": 0.4361,
+      "step": 820
+    },
+    {
+      "epoch": 0.9366799771819737,
+      "grad_norm": 0.5240760743232675,
+      "learning_rate": 3.820295983086681e-05,
+      "loss": 0.4233,
+      "step": 821
+    },
+    {
+      "epoch": 0.9378208784940103,
+      "grad_norm": 0.39250208092232536,
+      "learning_rate": 3.818181818181819e-05,
+      "loss": 0.4221,
+      "step": 822
+    },
+    {
+      "epoch": 0.9389617798060468,
+      "grad_norm": 0.4739585449842317,
+      "learning_rate": 3.816067653276956e-05,
+      "loss": 0.4365,
+      "step": 823
+    },
+    {
+      "epoch": 0.9401026811180833,
+      "grad_norm": 0.4937231941861391,
+      "learning_rate": 3.8139534883720935e-05,
+      "loss": 0.4357,
+      "step": 824
+    },
+    {
+      "epoch": 0.9412435824301197,
+      "grad_norm": 0.449294685022375,
+      "learning_rate": 3.8118393234672306e-05,
+      "loss": 0.4595,
+      "step": 825
+    },
+    {
+      "epoch": 0.9423844837421563,
+      "grad_norm": 0.3384444427709971,
+      "learning_rate": 3.809725158562368e-05,
+      "loss": 0.4436,
+      "step": 826
+    },
+    {
+      "epoch": 0.9435253850541928,
+      "grad_norm": 0.48955378365349317,
+      "learning_rate": 3.8076109936575054e-05,
+      "loss": 0.4386,
+      "step": 827
+    },
+    {
+      "epoch": 0.9446662863662293,
+      "grad_norm": 0.3002513126100782,
+      "learning_rate": 3.8054968287526425e-05,
+      "loss": 0.428,
+      "step": 828
+    },
+    {
+      "epoch": 0.9458071876782659,
+      "grad_norm": 0.47686476251659693,
+      "learning_rate": 3.80338266384778e-05,
+      "loss": 0.4161,
+      "step": 829
+    },
+    {
+      "epoch": 0.9469480889903024,
+      "grad_norm": 0.3337526894068287,
+      "learning_rate": 3.801268498942917e-05,
+      "loss": 0.4491,
+      "step": 830
+    },
+    {
+      "epoch": 0.9480889903023388,
+      "grad_norm": 0.42874888426843116,
+      "learning_rate": 3.799154334038055e-05,
+      "loss": 0.4322,
+      "step": 831
+    },
+    {
+      "epoch": 0.9492298916143753,
+      "grad_norm": 0.3686621986757173,
+      "learning_rate": 3.797040169133193e-05,
+      "loss": 0.4452,
+      "step": 832
+    },
+    {
+      "epoch": 0.9503707929264119,
+      "grad_norm": 0.4427227801549552,
+      "learning_rate": 3.7949260042283305e-05,
+      "loss": 0.4409,
+      "step": 833
+    },
+    {
+      "epoch": 0.9515116942384484,
+      "grad_norm": 0.40716707620144105,
+      "learning_rate": 3.7928118393234676e-05,
+      "loss": 0.44,
+      "step": 834
+    },
+    {
+      "epoch": 0.9526525955504849,
+      "grad_norm": 0.4778332483885651,
+      "learning_rate": 3.790697674418605e-05,
+      "loss": 0.4321,
+      "step": 835
+    },
+    {
+      "epoch": 0.9537934968625214,
+      "grad_norm": 0.5215688673683716,
+      "learning_rate": 3.7885835095137424e-05,
+      "loss": 0.4139,
+      "step": 836
+    },
+    {
+      "epoch": 0.9549343981745579,
+      "grad_norm": 0.3973314649903041,
+      "learning_rate": 3.7864693446088795e-05,
+      "loss": 0.4481,
+      "step": 837
+    },
+    {
+      "epoch": 0.9560752994865944,
+      "grad_norm": 0.5913338031384034,
+      "learning_rate": 3.784355179704017e-05,
+      "loss": 0.441,
+      "step": 838
+    },
+    {
+      "epoch": 0.9572162007986309,
+      "grad_norm": 0.3517543605151038,
+      "learning_rate": 3.782241014799154e-05,
+      "loss": 0.449,
+      "step": 839
+    },
+    {
+      "epoch": 0.9583571021106674,
+      "grad_norm": 0.5008792785960227,
+      "learning_rate": 3.780126849894292e-05,
+      "loss": 0.4196,
+      "step": 840
+    },
+    {
+      "epoch": 0.959498003422704,
+      "grad_norm": 0.3855644570106595,
+      "learning_rate": 3.778012684989429e-05,
+      "loss": 0.4401,
+      "step": 841
+    },
+    {
+      "epoch": 0.9606389047347405,
+      "grad_norm": 0.46893696225907877,
+      "learning_rate": 3.775898520084567e-05,
+      "loss": 0.4411,
+      "step": 842
+    },
+    {
+      "epoch": 0.9617798060467769,
+      "grad_norm": 0.36157623560181484,
+      "learning_rate": 3.773784355179704e-05,
+      "loss": 0.4438,
+      "step": 843
+    },
+    {
+      "epoch": 0.9629207073588134,
+      "grad_norm": 0.44243231770872754,
+      "learning_rate": 3.771670190274842e-05,
+      "loss": 0.4406,
+      "step": 844
+    },
+    {
+      "epoch": 0.96406160867085,
+      "grad_norm": 0.3107496198874931,
+      "learning_rate": 3.7695560253699794e-05,
+      "loss": 0.4165,
+      "step": 845
+    },
+    {
+      "epoch": 0.9652025099828865,
+      "grad_norm": 0.4615834197410632,
+      "learning_rate": 3.7674418604651165e-05,
+      "loss": 0.426,
+      "step": 846
+    },
+    {
+      "epoch": 0.966343411294923,
+      "grad_norm": 0.349515582638473,
+      "learning_rate": 3.765327695560254e-05,
+      "loss": 0.4296,
+      "step": 847
+    },
+    {
+      "epoch": 0.9674843126069596,
+      "grad_norm": 0.4222707991171846,
+      "learning_rate": 3.763213530655391e-05,
+      "loss": 0.4452,
+      "step": 848
+    },
+    {
+      "epoch": 0.968625213918996,
+      "grad_norm": 0.39191168304626894,
+      "learning_rate": 3.761099365750529e-05,
+      "loss": 0.4399,
+      "step": 849
+    },
+    {
+      "epoch": 0.9697661152310325,
+      "grad_norm": 0.4116727814128068,
+      "learning_rate": 3.758985200845666e-05,
+      "loss": 0.4559,
+      "step": 850
+    },
+    {
+      "epoch": 0.970907016543069,
+      "grad_norm": 0.40681194079162225,
+      "learning_rate": 3.756871035940804e-05,
+      "loss": 0.4731,
+      "step": 851
+    },
+    {
+      "epoch": 0.9720479178551056,
+      "grad_norm": 0.44083783057627335,
+      "learning_rate": 3.754756871035941e-05,
+      "loss": 0.4335,
+      "step": 852
+    },
+    {
+      "epoch": 0.9731888191671421,
+      "grad_norm": 0.37049492685038155,
+      "learning_rate": 3.752642706131078e-05,
+      "loss": 0.4293,
+      "step": 853
+    },
+    {
+      "epoch": 0.9743297204791785,
+      "grad_norm": 0.4375755567317356,
+      "learning_rate": 3.750528541226216e-05,
+      "loss": 0.4354,
+      "step": 854
+    },
+    {
+      "epoch": 0.975470621791215,
+      "grad_norm": 0.33832049022487304,
+      "learning_rate": 3.748414376321353e-05,
+      "loss": 0.4293,
+      "step": 855
+    },
+    {
+      "epoch": 0.9766115231032516,
+      "grad_norm": 0.5121821917096232,
+      "learning_rate": 3.7463002114164906e-05,
+      "loss": 0.4537,
+      "step": 856
+    },
+    {
+      "epoch": 0.9777524244152881,
+      "grad_norm": 0.3818306200016373,
+      "learning_rate": 3.7441860465116276e-05,
+      "loss": 0.4541,
+      "step": 857
+    },
+    {
+      "epoch": 0.9788933257273246,
+      "grad_norm": 0.4215507470795194,
+      "learning_rate": 3.7420718816067654e-05,
+      "loss": 0.4173,
+      "step": 858
+    },
+    {
+      "epoch": 0.980034227039361,
+      "grad_norm": 0.396875209227274,
+      "learning_rate": 3.739957716701903e-05,
+      "loss": 0.4227,
+      "step": 859
+    },
+    {
+      "epoch": 0.9811751283513976,
+      "grad_norm": 0.44522795502012974,
+      "learning_rate": 3.737843551797041e-05,
+      "loss": 0.4254,
+      "step": 860
+    },
+    {
+      "epoch": 0.9823160296634341,
+      "grad_norm": 0.36390782196922894,
+      "learning_rate": 3.735729386892178e-05,
+      "loss": 0.435,
+      "step": 861
+    },
+    {
+      "epoch": 0.9834569309754706,
+      "grad_norm": 0.4051435067373246,
+      "learning_rate": 3.733615221987315e-05,
+      "loss": 0.4469,
+      "step": 862
+    },
+    {
+      "epoch": 0.9845978322875071,
+      "grad_norm": 0.4300788624635129,
+      "learning_rate": 3.731501057082453e-05,
+      "loss": 0.4337,
+      "step": 863
+    },
+    {
+      "epoch": 0.9857387335995437,
+      "grad_norm": 0.4518840799967234,
+      "learning_rate": 3.72938689217759e-05,
+      "loss": 0.431,
+      "step": 864
+    },
+    {
+      "epoch": 0.9868796349115802,
+      "grad_norm": 0.3917993799092396,
+      "learning_rate": 3.7272727272727276e-05,
+      "loss": 0.4052,
+      "step": 865
+    },
+    {
+      "epoch": 0.9880205362236166,
+      "grad_norm": 0.4311974148203897,
+      "learning_rate": 3.7251585623678646e-05,
+      "loss": 0.4479,
+      "step": 866
+    },
+    {
+      "epoch": 0.9891614375356532,
+      "grad_norm": 0.33663593761732574,
+      "learning_rate": 3.7230443974630024e-05,
+      "loss": 0.4426,
+      "step": 867
+    },
+    {
+      "epoch": 0.9903023388476897,
+      "grad_norm": 0.41764811028477467,
+      "learning_rate": 3.7209302325581394e-05,
+      "loss": 0.4231,
+      "step": 868
+    },
+    {
+      "epoch": 0.9914432401597262,
+      "grad_norm": 0.37959761334860076,
+      "learning_rate": 3.718816067653277e-05,
+      "loss": 0.437,
+      "step": 869
+    },
+    {
+      "epoch": 0.9925841414717627,
+      "grad_norm": 0.3751287923553439,
+      "learning_rate": 3.716701902748414e-05,
+      "loss": 0.4359,
+      "step": 870
+    },
+    {
+      "epoch": 0.9937250427837993,
+      "grad_norm": 0.39059717878139527,
+      "learning_rate": 3.714587737843552e-05,
+      "loss": 0.4266,
+      "step": 871
+    },
+    {
+      "epoch": 0.9948659440958357,
+      "grad_norm": 0.3668029831866769,
+      "learning_rate": 3.71247357293869e-05,
+      "loss": 0.4346,
+      "step": 872
+    },
+    {
+      "epoch": 0.9960068454078722,
+      "grad_norm": 0.3758670658619597,
+      "learning_rate": 3.710359408033827e-05,
+      "loss": 0.4247,
+      "step": 873
+    },
+    {
+      "epoch": 0.9971477467199087,
+      "grad_norm": 0.3831467505102513,
+      "learning_rate": 3.7082452431289646e-05,
+      "loss": 0.4424,
+      "step": 874
+    },
+    {
+      "epoch": 0.9982886480319453,
+      "grad_norm": 0.37818254454581546,
+      "learning_rate": 3.7061310782241016e-05,
+      "loss": 0.4441,
+      "step": 875
+    },
+    {
+      "epoch": 0.9994295493439818,
+      "grad_norm": 0.38315281097821646,
+      "learning_rate": 3.7040169133192394e-05,
+      "loss": 0.4294,
+      "step": 876
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.38315281097821646,
+      "learning_rate": 3.7019027484143764e-05,
+      "loss": 0.438,
+      "step": 877
+    },
+    {
+      "epoch": 1.0011409013120365,
+      "grad_norm": 0.5455887151600689,
+      "learning_rate": 3.699788583509514e-05,
+      "loss": 0.3733,
+      "step": 878
+    },
+    {
+      "epoch": 1.002281802624073,
+      "grad_norm": 0.36610896066989357,
+      "learning_rate": 3.697674418604651e-05,
+      "loss": 0.3699,
+      "step": 879
+    },
+    {
+      "epoch": 1.0034227039361094,
+      "grad_norm": 0.40501415591406764,
+      "learning_rate": 3.695560253699788e-05,
+      "loss": 0.381,
+      "step": 880
+    },
+    {
+      "epoch": 1.0045636052481461,
+      "grad_norm": 0.3825535046773347,
+      "learning_rate": 3.693446088794926e-05,
+      "loss": 0.3808,
+      "step": 881
+    },
+    {
+      "epoch": 1.0057045065601826,
+      "grad_norm": 0.35102076165474366,
+      "learning_rate": 3.691331923890063e-05,
+      "loss": 0.3537,
+      "step": 882
+    },
+    {
+      "epoch": 1.006845407872219,
+      "grad_norm": 0.34649503103224494,
+      "learning_rate": 3.689217758985201e-05,
+      "loss": 0.3736,
+      "step": 883
+    },
+    {
+      "epoch": 1.0079863091842556,
+      "grad_norm": 0.4681279678485136,
+      "learning_rate": 3.6871035940803386e-05,
+      "loss": 0.3548,
+      "step": 884
+    },
+    {
+      "epoch": 1.009127210496292,
+      "grad_norm": 0.37489320591183883,
+      "learning_rate": 3.6849894291754764e-05,
+      "loss": 0.3703,
+      "step": 885
+    },
+    {
+      "epoch": 1.0102681118083285,
+      "grad_norm": 0.42678389224608554,
+      "learning_rate": 3.6828752642706135e-05,
+      "loss": 0.381,
+      "step": 886
+    },
+    {
+      "epoch": 1.011409013120365,
+      "grad_norm": 0.4198303145341611,
+      "learning_rate": 3.680761099365751e-05,
+      "loss": 0.3937,
+      "step": 887
+    },
+    {
+      "epoch": 1.0125499144324015,
+      "grad_norm": 0.4041636421810587,
+      "learning_rate": 3.678646934460888e-05,
+      "loss": 0.3726,
+      "step": 888
+    },
+    {
+      "epoch": 1.0136908157444382,
+      "grad_norm": 0.3401962730269841,
+      "learning_rate": 3.676532769556025e-05,
+      "loss": 0.3655,
+      "step": 889
+    },
+    {
+      "epoch": 1.0148317170564747,
+      "grad_norm": 0.3770208662137992,
+      "learning_rate": 3.674418604651163e-05,
+      "loss": 0.3802,
+      "step": 890
+    },
+    {
+      "epoch": 1.0159726183685112,
+      "grad_norm": 0.46172099327646693,
+      "learning_rate": 3.6723044397463e-05,
+      "loss": 0.4032,
+      "step": 891
+    },
+    {
+      "epoch": 1.0171135196805476,
+      "grad_norm": 0.3596991346706822,
+      "learning_rate": 3.670190274841438e-05,
+      "loss": 0.3815,
+      "step": 892
+    },
+    {
+      "epoch": 1.0182544209925841,
+      "grad_norm": 0.49205467040562745,
+      "learning_rate": 3.668076109936575e-05,
+      "loss": 0.3497,
+      "step": 893
+    },
+    {
+      "epoch": 1.0193953223046206,
+      "grad_norm": 0.33360531637571633,
+      "learning_rate": 3.665961945031713e-05,
+      "loss": 0.3668,
+      "step": 894
+    },
+    {
+      "epoch": 1.020536223616657,
+      "grad_norm": 0.425656228941723,
+      "learning_rate": 3.66384778012685e-05,
+      "loss": 0.3691,
+      "step": 895
+    },
+    {
+      "epoch": 1.0216771249286936,
+      "grad_norm": 0.3723856412773373,
+      "learning_rate": 3.6617336152219875e-05,
+      "loss": 0.3686,
+      "step": 896
+    },
+    {
+      "epoch": 1.0228180262407303,
+      "grad_norm": 0.4885789857759674,
+      "learning_rate": 3.659619450317125e-05,
+      "loss": 0.3668,
+      "step": 897
+    },
+    {
+      "epoch": 1.0239589275527667,
+      "grad_norm": 0.3445642312108721,
+      "learning_rate": 3.6575052854122623e-05,
+      "loss": 0.3674,
+      "step": 898
+    },
+    {
+      "epoch": 1.0250998288648032,
+      "grad_norm": 0.5026457518007369,
+      "learning_rate": 3.6553911205074e-05,
+      "loss": 0.3539,
+      "step": 899
+    },
+    {
+      "epoch": 1.0262407301768397,
+      "grad_norm": 0.3074717187657618,
+      "learning_rate": 3.653276955602537e-05,
+      "loss": 0.3758,
+      "step": 900
+    },
+    {
+      "epoch": 1.0273816314888762,
+      "grad_norm": 0.39374297057459845,
+      "learning_rate": 3.651162790697675e-05,
+      "loss": 0.3406,
+      "step": 901
+    },
+    {
+      "epoch": 1.0285225328009127,
+      "grad_norm": 0.3715766019853829,
+      "learning_rate": 3.649048625792812e-05,
+      "loss": 0.3701,
+      "step": 902
+    },
+    {
+      "epoch": 1.0296634341129491,
+      "grad_norm": 0.36179693452859935,
+      "learning_rate": 3.64693446088795e-05,
+      "loss": 0.4046,
+      "step": 903
+    },
+    {
+      "epoch": 1.0308043354249858,
+      "grad_norm": 0.35887466882761154,
+      "learning_rate": 3.644820295983087e-05,
+      "loss": 0.373,
+      "step": 904
+    },
+    {
+      "epoch": 1.0319452367370223,
+      "grad_norm": 0.4627643573370423,
+      "learning_rate": 3.642706131078224e-05,
+      "loss": 0.3546,
+      "step": 905
+    },
+    {
+      "epoch": 1.0330861380490588,
+      "grad_norm": 0.3414800929229019,
+      "learning_rate": 3.6405919661733616e-05,
+      "loss": 0.382,
+      "step": 906
+    },
+    {
+      "epoch": 1.0342270393610953,
+      "grad_norm": 0.4167762624955359,
+      "learning_rate": 3.638477801268499e-05,
+      "loss": 0.3707,
+      "step": 907
+    },
+    {
+      "epoch": 1.0353679406731318,
+      "grad_norm": 0.3634328382063604,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 0.3892,
+      "step": 908
+    },
+    {
+      "epoch": 1.0365088419851682,
+      "grad_norm": 0.39913860764154596,
+      "learning_rate": 3.6342494714587735e-05,
+      "loss": 0.3689,
+      "step": 909
+    },
+    {
+      "epoch": 1.0376497432972047,
+      "grad_norm": 0.31150134389469925,
+      "learning_rate": 3.632135306553911e-05,
+      "loss": 0.3513,
+      "step": 910
+    },
+    {
+      "epoch": 1.0387906446092412,
+      "grad_norm": 0.44513217932997295,
+      "learning_rate": 3.630021141649049e-05,
+      "loss": 0.4083,
+      "step": 911
+    },
+    {
+      "epoch": 1.039931545921278,
+      "grad_norm": 0.29893701924629945,
+      "learning_rate": 3.627906976744187e-05,
+      "loss": 0.3476,
+      "step": 912
+    },
+    {
+      "epoch": 1.0410724472333144,
+      "grad_norm": 0.36380817324672665,
+      "learning_rate": 3.625792811839324e-05,
+      "loss": 0.3694,
+      "step": 913
+    },
+    {
+      "epoch": 1.0422133485453509,
+      "grad_norm": 0.3811952404266299,
+      "learning_rate": 3.623678646934461e-05,
+      "loss": 0.3512,
+      "step": 914
+    },
+    {
+      "epoch": 1.0433542498573873,
+      "grad_norm": 0.3487936521280187,
+      "learning_rate": 3.6215644820295986e-05,
+      "loss": 0.3605,
+      "step": 915
+    },
+    {
+      "epoch": 1.0444951511694238,
+      "grad_norm": 0.295589207624635,
+      "learning_rate": 3.619450317124736e-05,
+      "loss": 0.356,
+      "step": 916
+    },
+    {
+      "epoch": 1.0456360524814603,
+      "grad_norm": 0.3361959857373228,
+      "learning_rate": 3.6173361522198734e-05,
+      "loss": 0.361,
+      "step": 917
+    },
+    {
+      "epoch": 1.0467769537934968,
+      "grad_norm": 0.31037775502179515,
+      "learning_rate": 3.6152219873150105e-05,
+      "loss": 0.3599,
+      "step": 918
+    },
+    {
+      "epoch": 1.0479178551055335,
+      "grad_norm": 0.3480868223602575,
+      "learning_rate": 3.613107822410148e-05,
+      "loss": 0.3727,
+      "step": 919
+    },
+    {
+      "epoch": 1.04905875641757,
+      "grad_norm": 0.3759497242879828,
+      "learning_rate": 3.610993657505285e-05,
+      "loss": 0.3744,
+      "step": 920
+    },
+    {
+      "epoch": 1.0501996577296064,
+      "grad_norm": 0.44625648445941285,
+      "learning_rate": 3.608879492600423e-05,
+      "loss": 0.3982,
+      "step": 921
+    },
+    {
+      "epoch": 1.051340559041643,
+      "grad_norm": 0.3551571522322082,
+      "learning_rate": 3.60676532769556e-05,
+      "loss": 0.3683,
+      "step": 922
+    },
+    {
+      "epoch": 1.0524814603536794,
+      "grad_norm": 0.3958943914214428,
+      "learning_rate": 3.604651162790698e-05,
+      "loss": 0.3663,
+      "step": 923
+    },
+    {
+      "epoch": 1.0536223616657159,
+      "grad_norm": 0.39346013363138627,
+      "learning_rate": 3.6025369978858356e-05,
+      "loss": 0.3785,
+      "step": 924
+    },
+    {
+      "epoch": 1.0547632629777524,
+      "grad_norm": 0.35191600740827306,
+      "learning_rate": 3.600422832980973e-05,
+      "loss": 0.3897,
+      "step": 925
+    },
+    {
+      "epoch": 1.0559041642897888,
+      "grad_norm": 0.38676608726463646,
+      "learning_rate": 3.5983086680761104e-05,
+      "loss": 0.3729,
+      "step": 926
+    },
+    {
+      "epoch": 1.0570450656018255,
+      "grad_norm": 0.34962772536528997,
+      "learning_rate": 3.5961945031712475e-05,
+      "loss": 0.3795,
+      "step": 927
+    },
+    {
+      "epoch": 1.058185966913862,
+      "grad_norm": 0.3155978123445468,
+      "learning_rate": 3.594080338266385e-05,
+      "loss": 0.3533,
+      "step": 928
+    },
+    {
+      "epoch": 1.0593268682258985,
+      "grad_norm": 0.35223510579525524,
+      "learning_rate": 3.591966173361522e-05,
+      "loss": 0.3706,
+      "step": 929
+    },
+    {
+      "epoch": 1.060467769537935,
+      "grad_norm": 0.35708465819842106,
+      "learning_rate": 3.58985200845666e-05,
+      "loss": 0.3617,
+      "step": 930
+    },
+    {
+      "epoch": 1.0616086708499715,
+      "grad_norm": 0.3678947649013475,
+      "learning_rate": 3.587737843551797e-05,
+      "loss": 0.3913,
+      "step": 931
+    },
+    {
+      "epoch": 1.062749572162008,
+      "grad_norm": 0.37804755879302426,
+      "learning_rate": 3.585623678646934e-05,
+      "loss": 0.3643,
+      "step": 932
+    },
+    {
+      "epoch": 1.0638904734740444,
+      "grad_norm": 0.4209197871034825,
+      "learning_rate": 3.583509513742072e-05,
+      "loss": 0.383,
+      "step": 933
+    },
+    {
+      "epoch": 1.0650313747860811,
+      "grad_norm": 0.3875534348488064,
+      "learning_rate": 3.581395348837209e-05,
+      "loss": 0.3537,
+      "step": 934
+    },
+    {
+      "epoch": 1.0661722760981176,
+      "grad_norm": 0.339713004433371,
+      "learning_rate": 3.579281183932347e-05,
+      "loss": 0.3528,
+      "step": 935
+    },
+    {
+      "epoch": 1.067313177410154,
+      "grad_norm": 0.40452497385972475,
+      "learning_rate": 3.5771670190274845e-05,
+      "loss": 0.3799,
+      "step": 936
+    },
+    {
+      "epoch": 1.0684540787221906,
+      "grad_norm": 0.41269005352350124,
+      "learning_rate": 3.575052854122622e-05,
+      "loss": 0.3588,
+      "step": 937
+    },
+    {
+      "epoch": 1.069594980034227,
+      "grad_norm": 0.37102780996281465,
+      "learning_rate": 3.572938689217759e-05,
+      "loss": 0.3515,
+      "step": 938
+    },
+    {
+      "epoch": 1.0707358813462635,
+      "grad_norm": 0.35749919468134717,
+      "learning_rate": 3.570824524312897e-05,
+      "loss": 0.3922,
+      "step": 939
+    },
+    {
+      "epoch": 1.0718767826583,
+      "grad_norm": 0.449335797414619,
+      "learning_rate": 3.568710359408034e-05,
+      "loss": 0.3972,
+      "step": 940
+    },
+    {
+      "epoch": 1.0730176839703365,
+      "grad_norm": 0.3711962817643683,
+      "learning_rate": 3.566596194503171e-05,
+      "loss": 0.3781,
+      "step": 941
+    },
+    {
+      "epoch": 1.074158585282373,
+      "grad_norm": 0.39709697007286004,
+      "learning_rate": 3.564482029598309e-05,
+      "loss": 0.3747,
+      "step": 942
+    },
+    {
+      "epoch": 1.0752994865944097,
+      "grad_norm": 0.37530640138836563,
+      "learning_rate": 3.562367864693446e-05,
+      "loss": 0.3683,
+      "step": 943
+    },
+    {
+      "epoch": 1.0764403879064461,
+      "grad_norm": 0.42805517019254413,
+      "learning_rate": 3.560253699788584e-05,
+      "loss": 0.3743,
+      "step": 944
+    },
+    {
+      "epoch": 1.0775812892184826,
+      "grad_norm": 0.3788980861432765,
+      "learning_rate": 3.558139534883721e-05,
+      "loss": 0.3519,
+      "step": 945
+    },
+    {
+      "epoch": 1.078722190530519,
+      "grad_norm": 0.37575065677709124,
+      "learning_rate": 3.5560253699788586e-05,
+      "loss": 0.3635,
+      "step": 946
+    },
+    {
+      "epoch": 1.0798630918425556,
+      "grad_norm": 0.35473463562879565,
+      "learning_rate": 3.5539112050739956e-05,
+      "loss": 0.3917,
+      "step": 947
+    },
+    {
+      "epoch": 1.081003993154592,
+      "grad_norm": 0.5421488353080326,
+      "learning_rate": 3.5517970401691334e-05,
+      "loss": 0.3616,
+      "step": 948
+    },
+    {
+      "epoch": 1.0821448944666285,
+      "grad_norm": 0.322957416404556,
+      "learning_rate": 3.5496828752642705e-05,
+      "loss": 0.3765,
+      "step": 949
+    },
+    {
+      "epoch": 1.0832857957786652,
+      "grad_norm": 0.4798700548633995,
+      "learning_rate": 3.547568710359408e-05,
+      "loss": 0.3581,
+      "step": 950
+    },
+    {
+      "epoch": 1.0844266970907017,
+      "grad_norm": 0.31013380709522304,
+      "learning_rate": 3.545454545454546e-05,
+      "loss": 0.3625,
+      "step": 951
+    },
+    {
+      "epoch": 1.0855675984027382,
+      "grad_norm": 0.4206561403068851,
+      "learning_rate": 3.543340380549683e-05,
+      "loss": 0.3749,
+      "step": 952
+    },
+    {
+      "epoch": 1.0867084997147747,
+      "grad_norm": 0.39833895686913684,
+      "learning_rate": 3.541226215644821e-05,
+      "loss": 0.3822,
+      "step": 953
+    },
+    {
+      "epoch": 1.0878494010268112,
+      "grad_norm": 0.43256364250638124,
+      "learning_rate": 3.539112050739958e-05,
+      "loss": 0.3673,
+      "step": 954
+    },
+    {
+      "epoch": 1.0889903023388476,
+      "grad_norm": 0.3675468188829437,
+      "learning_rate": 3.5369978858350956e-05,
+      "loss": 0.3717,
+      "step": 955
+    },
+    {
+      "epoch": 1.0901312036508841,
+      "grad_norm": 0.3218252926229577,
+      "learning_rate": 3.5348837209302326e-05,
+      "loss": 0.375,
+      "step": 956
+    },
+    {
+      "epoch": 1.0912721049629206,
+      "grad_norm": 0.40359712151176336,
+      "learning_rate": 3.5327695560253704e-05,
+      "loss": 0.3895,
+      "step": 957
+    },
+    {
+      "epoch": 1.0924130062749573,
+      "grad_norm": 0.32814056788493573,
+      "learning_rate": 3.5306553911205075e-05,
+      "loss": 0.3496,
+      "step": 958
+    },
+    {
+      "epoch": 1.0935539075869938,
+      "grad_norm": 0.3700443887990979,
+      "learning_rate": 3.5285412262156445e-05,
+      "loss": 0.3813,
+      "step": 959
+    },
+    {
+      "epoch": 1.0946948088990303,
+      "grad_norm": 0.30296344478274184,
+      "learning_rate": 3.526427061310782e-05,
+      "loss": 0.3705,
+      "step": 960
+    },
+    {
+      "epoch": 1.0958357102110667,
+      "grad_norm": 0.35084001507574447,
+      "learning_rate": 3.5243128964059193e-05,
+      "loss": 0.356,
+      "step": 961
+    },
+    {
+      "epoch": 1.0969766115231032,
+      "grad_norm": 0.34046673230652147,
+      "learning_rate": 3.522198731501057e-05,
+      "loss": 0.3738,
+      "step": 962
+    },
+    {
+      "epoch": 1.0981175128351397,
+      "grad_norm": 0.3591654008324595,
+      "learning_rate": 3.520084566596195e-05,
+      "loss": 0.3641,
+      "step": 963
+    },
+    {
+      "epoch": 1.0992584141471762,
+      "grad_norm": 0.34788679397569927,
+      "learning_rate": 3.5179704016913326e-05,
+      "loss": 0.3861,
+      "step": 964
+    },
+    {
+      "epoch": 1.1003993154592129,
+      "grad_norm": 0.3489129692311338,
+      "learning_rate": 3.5158562367864696e-05,
+      "loss": 0.3963,
+      "step": 965
+    },
+    {
+      "epoch": 1.1015402167712494,
+      "grad_norm": 0.32115244707275425,
+      "learning_rate": 3.5137420718816074e-05,
+      "loss": 0.3772,
+      "step": 966
+    },
+    {
+      "epoch": 1.1026811180832858,
+      "grad_norm": 0.36405343778575244,
+      "learning_rate": 3.5116279069767445e-05,
+      "loss": 0.3593,
+      "step": 967
+    },
+    {
+      "epoch": 1.1038220193953223,
+      "grad_norm": 0.25934220146881914,
+      "learning_rate": 3.5095137420718815e-05,
+      "loss": 0.369,
+      "step": 968
+    },
+    {
+      "epoch": 1.1049629207073588,
+      "grad_norm": 0.3459709424202936,
+      "learning_rate": 3.507399577167019e-05,
+      "loss": 0.3605,
+      "step": 969
+    },
+    {
+      "epoch": 1.1061038220193953,
+      "grad_norm": 0.25530816000283274,
+      "learning_rate": 3.5052854122621563e-05,
+      "loss": 0.3702,
+      "step": 970
+    },
+    {
+      "epoch": 1.1072447233314318,
+      "grad_norm": 0.2977170590076077,
+      "learning_rate": 3.503171247357294e-05,
+      "loss": 0.3737,
+      "step": 971
+    },
+    {
+      "epoch": 1.1083856246434682,
+      "grad_norm": 0.2827441362997686,
+      "learning_rate": 3.501057082452431e-05,
+      "loss": 0.3759,
+      "step": 972
+    },
+    {
+      "epoch": 1.109526525955505,
+      "grad_norm": 0.3244608088210149,
+      "learning_rate": 3.498942917547569e-05,
+      "loss": 0.3682,
+      "step": 973
+    },
+    {
+      "epoch": 1.1106674272675414,
+      "grad_norm": 0.3347264076238722,
+      "learning_rate": 3.496828752642706e-05,
+      "loss": 0.3742,
+      "step": 974
+    },
+    {
+      "epoch": 1.111808328579578,
+      "grad_norm": 0.30216880984211525,
+      "learning_rate": 3.494714587737844e-05,
+      "loss": 0.3725,
+      "step": 975
+    },
+    {
+      "epoch": 1.1129492298916144,
+      "grad_norm": 0.3289978042150018,
+      "learning_rate": 3.4926004228329815e-05,
+      "loss": 0.3644,
+      "step": 976
+    },
+    {
+      "epoch": 1.1140901312036509,
+      "grad_norm": 0.3238466739022087,
+      "learning_rate": 3.4904862579281185e-05,
+      "loss": 0.3519,
+      "step": 977
+    },
+    {
+      "epoch": 1.1152310325156873,
+      "grad_norm": 0.31017493518179495,
+      "learning_rate": 3.488372093023256e-05,
+      "loss": 0.3626,
+      "step": 978
+    },
+    {
+      "epoch": 1.1163719338277238,
+      "grad_norm": 0.3038861691228129,
+      "learning_rate": 3.4862579281183933e-05,
+      "loss": 0.3588,
+      "step": 979
+    },
+    {
+      "epoch": 1.1175128351397605,
+      "grad_norm": 0.2929692138971038,
+      "learning_rate": 3.484143763213531e-05,
+      "loss": 0.3654,
+      "step": 980
+    },
+    {
+      "epoch": 1.118653736451797,
+      "grad_norm": 0.33957255149071514,
+      "learning_rate": 3.482029598308668e-05,
+      "loss": 0.3741,
+      "step": 981
+    },
+    {
+      "epoch": 1.1197946377638335,
+      "grad_norm": 0.37962287901659914,
+      "learning_rate": 3.479915433403806e-05,
+      "loss": 0.372,
+      "step": 982
+    },
+    {
+      "epoch": 1.12093553907587,
+      "grad_norm": 0.30917582487173056,
+      "learning_rate": 3.477801268498943e-05,
+      "loss": 0.3625,
+      "step": 983
+    },
+    {
+      "epoch": 1.1220764403879064,
+      "grad_norm": 0.43355534624442893,
+      "learning_rate": 3.47568710359408e-05,
+      "loss": 0.3857,
+      "step": 984
+    },
+    {
+      "epoch": 1.123217341699943,
+      "grad_norm": 0.3181982233605466,
+      "learning_rate": 3.473572938689218e-05,
+      "loss": 0.3809,
+      "step": 985
+    },
+    {
+      "epoch": 1.1243582430119794,
+      "grad_norm": 0.32622169252857547,
+      "learning_rate": 3.471458773784355e-05,
+      "loss": 0.3713,
+      "step": 986
+    },
+    {
+      "epoch": 1.1254991443240159,
+      "grad_norm": 0.3327070573503023,
+      "learning_rate": 3.4693446088794926e-05,
+      "loss": 0.3657,
+      "step": 987
+    },
+    {
+      "epoch": 1.1266400456360526,
+      "grad_norm": 0.3124938302800057,
+      "learning_rate": 3.4672304439746304e-05,
+      "loss": 0.3831,
+      "step": 988
+    },
+    {
+      "epoch": 1.127780946948089,
+      "grad_norm": 0.31952256373453464,
+      "learning_rate": 3.465116279069768e-05,
+      "loss": 0.3727,
+      "step": 989
+    },
+    {
+      "epoch": 1.1289218482601255,
+      "grad_norm": 0.3063299430772017,
+      "learning_rate": 3.463002114164905e-05,
+      "loss": 0.3891,
+      "step": 990
+    },
+    {
+      "epoch": 1.130062749572162,
+      "grad_norm": 0.31462507340453627,
+      "learning_rate": 3.460887949260043e-05,
+      "loss": 0.3541,
+      "step": 991
+    },
+    {
+      "epoch": 1.1312036508841985,
+      "grad_norm": 0.2890799992114355,
+      "learning_rate": 3.45877378435518e-05,
+      "loss": 0.3833,
+      "step": 992
+    },
+    {
+      "epoch": 1.132344552196235,
+      "grad_norm": 0.3615461671219346,
+      "learning_rate": 3.456659619450317e-05,
+      "loss": 0.3704,
+      "step": 993
+    },
+    {
+      "epoch": 1.1334854535082715,
+      "grad_norm": 0.3249056891163027,
+      "learning_rate": 3.454545454545455e-05,
+      "loss": 0.3745,
+      "step": 994
+    },
+    {
+      "epoch": 1.1346263548203082,
+      "grad_norm": 0.3137775773273121,
+      "learning_rate": 3.452431289640592e-05,
+      "loss": 0.3851,
+      "step": 995
+    },
+    {
+      "epoch": 1.1357672561323446,
+      "grad_norm": 0.29114940686367663,
+      "learning_rate": 3.4503171247357296e-05,
+      "loss": 0.3771,
+      "step": 996
+    },
+    {
+      "epoch": 1.1369081574443811,
+      "grad_norm": 0.37624711275686223,
+      "learning_rate": 3.448202959830867e-05,
+      "loss": 0.3526,
+      "step": 997
+    },
+    {
+      "epoch": 1.1380490587564176,
+      "grad_norm": 0.31260376745882296,
+      "learning_rate": 3.4460887949260044e-05,
+      "loss": 0.3741,
+      "step": 998
+    },
+    {
+      "epoch": 1.139189960068454,
+      "grad_norm": 0.33182211780658555,
+      "learning_rate": 3.4439746300211415e-05,
+      "loss": 0.3672,
+      "step": 999
+    },
+    {
+      "epoch": 1.1403308613804906,
+      "grad_norm": 0.25997411455628916,
+      "learning_rate": 3.441860465116279e-05,
+      "loss": 0.364,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1426126640045635,
+      "grad_norm": 0.33589476782435423,
+      "learning_rate": 3.439746300211416e-05,
+      "loss": 0.3859,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1437535653166,
+      "grad_norm": 0.323028968776872,
+      "learning_rate": 3.437632135306554e-05,
+      "loss": 0.3721,
+      "step": 1002
+    },
+    {
+      "epoch": 1.1448944666286367,
+      "grad_norm": 0.34849839129010746,
+      "learning_rate": 3.435517970401692e-05,
+      "loss": 0.3605,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1460353679406732,
+      "grad_norm": 0.3593724612628344,
+      "learning_rate": 3.433403805496829e-05,
+      "loss": 0.3712,
+      "step": 1004
+    },
+    {
+      "epoch": 1.1471762692527097,
+      "grad_norm": 0.3840174852626447,
+      "learning_rate": 3.4312896405919666e-05,
+      "loss": 0.3756,
+      "step": 1005
+    },
+    {
+      "epoch": 1.1483171705647461,
+      "grad_norm": 0.3786901323232798,
+      "learning_rate": 3.429175475687104e-05,
+      "loss": 0.3897,
+      "step": 1006
+    },
+    {
+      "epoch": 1.1494580718767826,
+      "grad_norm": 0.4736250302570152,
+      "learning_rate": 3.4270613107822414e-05,
+      "loss": 0.3835,
+      "step": 1007
+    },
+    {
+      "epoch": 1.150598973188819,
+      "grad_norm": 0.3205148314531088,
+      "learning_rate": 3.4249471458773785e-05,
+      "loss": 0.373,
+      "step": 1008
+    },
+    {
+      "epoch": 1.1517398745008558,
+      "grad_norm": 0.3674962877821635,
+      "learning_rate": 3.422832980972516e-05,
+      "loss": 0.3809,
+      "step": 1009
+    },
+    {
+      "epoch": 1.1528807758128923,
+      "grad_norm": 0.38716940285294676,
+      "learning_rate": 3.420718816067653e-05,
+      "loss": 0.3718,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1540216771249288,
+      "grad_norm": 0.34612933762885156,
+      "learning_rate": 3.4186046511627904e-05,
+      "loss": 0.3958,
+      "step": 1011
+    },
+    {
+      "epoch": 1.1551625784369652,
+      "grad_norm": 0.3261036228115381,
+      "learning_rate": 3.416490486257928e-05,
+      "loss": 0.3663,
+      "step": 1012
+    },
+    {
+      "epoch": 1.1563034797490017,
+      "grad_norm": 0.3798365061694364,
+      "learning_rate": 3.414376321353065e-05,
+      "loss": 0.371,
+      "step": 1013
+    },
+    {
+      "epoch": 1.1574443810610382,
+      "grad_norm": 0.36082089990308985,
+      "learning_rate": 3.412262156448203e-05,
+      "loss": 0.3807,
+      "step": 1014
+    },
+    {
+      "epoch": 1.1585852823730747,
+      "grad_norm": 0.3809948230684839,
+      "learning_rate": 3.410147991543341e-05,
+      "loss": 0.3696,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1597261836851112,
+      "grad_norm": 0.3946308893177107,
+      "learning_rate": 3.4080338266384784e-05,
+      "loss": 0.3773,
+      "step": 1016
+    },
+    {
+      "epoch": 1.1608670849971476,
+      "grad_norm": 0.364442606139928,
+      "learning_rate": 3.4059196617336155e-05,
+      "loss": 0.361,
+      "step": 1017
+    },
+    {
+      "epoch": 1.1620079863091843,
+      "grad_norm": 0.3609258874429056,
+      "learning_rate": 3.403805496828753e-05,
+      "loss": 0.3856,
+      "step": 1018
+    },
+    {
+      "epoch": 1.1631488876212208,
+      "grad_norm": 1.1416343579165587,
+      "learning_rate": 3.40169133192389e-05,
+      "loss": 0.375,
+      "step": 1019
+    },
+    {
+      "epoch": 1.1642897889332573,
+      "grad_norm": 0.45946054420108273,
+      "learning_rate": 3.3995771670190274e-05,
+      "loss": 0.3431,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1654306902452938,
+      "grad_norm": 0.3364729235961614,
+      "learning_rate": 3.397463002114165e-05,
+      "loss": 0.3501,
+      "step": 1021
+    },
+    {
+      "epoch": 1.1665715915573303,
+      "grad_norm": 0.4080943198746895,
+      "learning_rate": 3.395348837209302e-05,
+      "loss": 0.3789,
+      "step": 1022
+    },
+    {
+      "epoch": 1.1677124928693667,
+      "grad_norm": 0.44524984811221446,
+      "learning_rate": 3.39323467230444e-05,
+      "loss": 0.3677,
+      "step": 1023
+    },
+    {
+      "epoch": 1.1688533941814032,
+      "grad_norm": 0.41099661440760105,
+      "learning_rate": 3.391120507399577e-05,
+      "loss": 0.3556,
+      "step": 1024
+    },
+    {
+      "epoch": 1.16999429549344,
+      "grad_norm": 0.3574833952927246,
+      "learning_rate": 3.389006342494715e-05,
+      "loss": 0.3618,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1711351968054764,
+      "grad_norm": 0.49459507464416336,
+      "learning_rate": 3.386892177589852e-05,
+      "loss": 0.3549,
+      "step": 1026
+    },
+    {
+      "epoch": 1.1722760981175129,
+      "grad_norm": 0.322252465481915,
+      "learning_rate": 3.3847780126849896e-05,
+      "loss": 0.3683,
+      "step": 1027
+    },
+    {
+      "epoch": 1.1734169994295494,
+      "grad_norm": 0.3327748277535711,
+      "learning_rate": 3.382663847780127e-05,
+      "loss": 0.3587,
+      "step": 1028
+    },
+    {
+      "epoch": 1.1745579007415858,
+      "grad_norm": 0.30790560462013256,
+      "learning_rate": 3.3805496828752644e-05,
+      "loss": 0.3778,
+      "step": 1029
+    },
+    {
+      "epoch": 1.1756988020536223,
+      "grad_norm": 0.4146437969607725,
+      "learning_rate": 3.378435517970402e-05,
+      "loss": 0.3831,
+      "step": 1030
+    },
+    {
+      "epoch": 1.1768397033656588,
+      "grad_norm": 0.3029308530687607,
+      "learning_rate": 3.376321353065539e-05,
+      "loss": 0.376,
+      "step": 1031
+    },
+    {
+      "epoch": 1.1779806046776953,
+      "grad_norm": 0.3883708207491139,
+      "learning_rate": 3.374207188160677e-05,
+      "loss": 0.3827,
+      "step": 1032
+    },
+    {
+      "epoch": 1.179121505989732,
+      "grad_norm": 0.3264973115537315,
+      "learning_rate": 3.372093023255814e-05,
+      "loss": 0.3709,
+      "step": 1033
+    },
+    {
+      "epoch": 1.1802624073017685,
+      "grad_norm": 0.3224818782407752,
+      "learning_rate": 3.369978858350952e-05,
+      "loss": 0.3479,
+      "step": 1034
+    },
+    {
+      "epoch": 1.181403308613805,
+      "grad_norm": 0.35837329807660184,
+      "learning_rate": 3.367864693446089e-05,
+      "loss": 0.3537,
+      "step": 1035
+    },
+    {
+      "epoch": 1.1825442099258414,
+      "grad_norm": 0.32978113024365624,
+      "learning_rate": 3.3657505285412266e-05,
+      "loss": 0.3828,
+      "step": 1036
+    },
+    {
+      "epoch": 1.183685111237878,
+      "grad_norm": 0.30478082030156534,
+      "learning_rate": 3.3636363636363636e-05,
+      "loss": 0.3719,
+      "step": 1037
+    },
+    {
+      "epoch": 1.1848260125499144,
+      "grad_norm": 0.4039017489716658,
+      "learning_rate": 3.361522198731501e-05,
+      "loss": 0.3802,
+      "step": 1038
+    },
+    {
+      "epoch": 1.1859669138619509,
+      "grad_norm": 0.3137696235218554,
+      "learning_rate": 3.3594080338266385e-05,
+      "loss": 0.3497,
+      "step": 1039
+    },
+    {
+      "epoch": 1.1871078151739876,
+      "grad_norm": 0.3770758080250968,
+      "learning_rate": 3.357293868921776e-05,
+      "loss": 0.3857,
+      "step": 1040
+    },
+    {
+      "epoch": 1.188248716486024,
+      "grad_norm": 0.378178032731245,
+      "learning_rate": 3.355179704016913e-05,
+      "loss": 0.3825,
+      "step": 1041
+    },
+    {
+      "epoch": 1.1893896177980605,
+      "grad_norm": 0.3339076735915448,
+      "learning_rate": 3.353065539112051e-05,
+      "loss": 0.3554,
+      "step": 1042
+    },
+    {
+      "epoch": 1.190530519110097,
+      "grad_norm": 0.32202843560236183,
+      "learning_rate": 3.350951374207189e-05,
+      "loss": 0.385,
+      "step": 1043
+    },
+    {
+      "epoch": 1.1916714204221335,
+      "grad_norm": 0.39205899390410504,
+      "learning_rate": 3.348837209302326e-05,
+      "loss": 0.3939,
+      "step": 1044
+    },
+    {
+      "epoch": 1.19281232173417,
+      "grad_norm": 0.31585241143330434,
+      "learning_rate": 3.3467230443974636e-05,
+      "loss": 0.3697,
+      "step": 1045
+    },
+    {
+      "epoch": 1.1939532230462064,
+      "grad_norm": 0.34721707553982895,
+      "learning_rate": 3.3446088794926007e-05,
+      "loss": 0.3725,
+      "step": 1046
+    },
+    {
+      "epoch": 1.195094124358243,
+      "grad_norm": 0.3224483535503891,
+      "learning_rate": 3.342494714587738e-05,
+      "loss": 0.4002,
+      "step": 1047
+    },
+    {
+      "epoch": 1.1962350256702796,
+      "grad_norm": 0.34957765959120296,
+      "learning_rate": 3.3403805496828755e-05,
+      "loss": 0.3627,
+      "step": 1048
+    },
+    {
+      "epoch": 1.197375926982316,
+      "grad_norm": 0.38328812141174423,
+      "learning_rate": 3.3382663847780125e-05,
+      "loss": 0.4014,
+      "step": 1049
+    },
+    {
+      "epoch": 1.1985168282943526,
+      "grad_norm": 0.3758584671491964,
+      "learning_rate": 3.33615221987315e-05,
+      "loss": 0.3696,
+      "step": 1050
+    },
+    {
+      "epoch": 1.199657729606389,
+      "grad_norm": 0.39611799283306404,
+      "learning_rate": 3.3340380549682874e-05,
+      "loss": 0.387,
+      "step": 1051
+    },
+    {
+      "epoch": 1.2007986309184255,
+      "grad_norm": 0.2987773907456679,
+      "learning_rate": 3.331923890063425e-05,
+      "loss": 0.3879,
+      "step": 1052
+    },
+    {
+      "epoch": 1.201939532230462,
+      "grad_norm": 0.36426120265256484,
+      "learning_rate": 3.329809725158562e-05,
+      "loss": 0.362,
+      "step": 1053
+    },
+    {
+      "epoch": 1.2030804335424985,
+      "grad_norm": 0.36560634288948834,
+      "learning_rate": 3.3276955602537e-05,
+      "loss": 0.3509,
+      "step": 1054
+    },
+    {
+      "epoch": 1.2042213348545352,
+      "grad_norm": 0.3035450394151478,
+      "learning_rate": 3.3255813953488377e-05,
+      "loss": 0.376,
+      "step": 1055
+    },
+    {
+      "epoch": 1.2053622361665717,
+      "grad_norm": 0.39302963075707603,
+      "learning_rate": 3.323467230443975e-05,
+      "loss": 0.3699,
+      "step": 1056
+    },
+    {
+      "epoch": 1.2065031374786082,
+      "grad_norm": 0.35343270894560463,
+      "learning_rate": 3.3213530655391125e-05,
+      "loss": 0.3707,
+      "step": 1057
+    },
+    {
+      "epoch": 1.2076440387906446,
+      "grad_norm": 0.32922752073268385,
+      "learning_rate": 3.3192389006342495e-05,
+      "loss": 0.362,
+      "step": 1058
+    },
+    {
+      "epoch": 1.2087849401026811,
+      "grad_norm": 0.3497263627038085,
+      "learning_rate": 3.317124735729387e-05,
+      "loss": 0.3544,
+      "step": 1059
+    },
+    {
+      "epoch": 1.2099258414147176,
+      "grad_norm": 0.38176942411154147,
+      "learning_rate": 3.3150105708245244e-05,
+      "loss": 0.3786,
+      "step": 1060
+    },
+    {
+      "epoch": 1.211066742726754,
+      "grad_norm": 0.3655333255212552,
+      "learning_rate": 3.312896405919662e-05,
+      "loss": 0.3844,
+      "step": 1061
+    },
+    {
+      "epoch": 1.2122076440387906,
+      "grad_norm": 0.3562621090561445,
+      "learning_rate": 3.310782241014799e-05,
+      "loss": 0.3634,
+      "step": 1062
+    },
+    {
+      "epoch": 1.213348545350827,
+      "grad_norm": 0.3900734684566537,
+      "learning_rate": 3.308668076109937e-05,
+      "loss": 0.3837,
+      "step": 1063
+    },
+    {
+      "epoch": 1.2144894466628637,
+      "grad_norm": 0.3140299737074897,
+      "learning_rate": 3.306553911205074e-05,
+      "loss": 0.381,
+      "step": 1064
+    },
+    {
+      "epoch": 1.2156303479749002,
+      "grad_norm": 0.3167356658778117,
+      "learning_rate": 3.304439746300211e-05,
+      "loss": 0.3751,
+      "step": 1065
+    },
+    {
+      "epoch": 1.2167712492869367,
+      "grad_norm": 0.32986942860860147,
+      "learning_rate": 3.302325581395349e-05,
+      "loss": 0.3671,
+      "step": 1066
+    },
+    {
+      "epoch": 1.2179121505989732,
+      "grad_norm": 0.3980484794765463,
+      "learning_rate": 3.3002114164904865e-05,
+      "loss": 0.3922,
+      "step": 1067
+    },
+    {
+      "epoch": 1.2190530519110097,
+      "grad_norm": 0.31025070985969755,
+      "learning_rate": 3.298097251585624e-05,
+      "loss": 0.3612,
+      "step": 1068
+    },
+    {
+      "epoch": 1.2201939532230461,
+      "grad_norm": 0.3954022677035698,
+      "learning_rate": 3.2959830866807614e-05,
+      "loss": 0.3775,
+      "step": 1069
+    },
+    {
+      "epoch": 1.2213348545350828,
+      "grad_norm": 0.3398950566588672,
+      "learning_rate": 3.293868921775899e-05,
+      "loss": 0.3922,
+      "step": 1070
+    },
+    {
+      "epoch": 1.2224757558471193,
+      "grad_norm": 0.35576853153937077,
+      "learning_rate": 3.291754756871036e-05,
+      "loss": 0.3746,
+      "step": 1071
+    },
+    {
+      "epoch": 1.2236166571591558,
+      "grad_norm": 0.30802379229978166,
+      "learning_rate": 3.289640591966173e-05,
+      "loss": 0.3823,
+      "step": 1072
+    },
+    {
+      "epoch": 1.2247575584711923,
+      "grad_norm": 0.3493515239803427,
+      "learning_rate": 3.287526427061311e-05,
+      "loss": 0.3541,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2258984597832288,
+      "grad_norm": 0.27413733268524487,
+      "learning_rate": 3.285412262156448e-05,
+      "loss": 0.3654,
+      "step": 1074
+    },
+    {
+      "epoch": 1.2270393610952652,
+      "grad_norm": 0.30282716452139746,
+      "learning_rate": 3.283298097251586e-05,
+      "loss": 0.3711,
+      "step": 1075
+    },
+    {
+      "epoch": 1.2281802624073017,
+      "grad_norm": 0.31145240791005063,
+      "learning_rate": 3.281183932346723e-05,
+      "loss": 0.3583,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2293211637193382,
+      "grad_norm": 0.27129448563700514,
+      "learning_rate": 3.2790697674418606e-05,
+      "loss": 0.3643,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2304620650313747,
+      "grad_norm": 0.3295699887395353,
+      "learning_rate": 3.276955602536998e-05,
+      "loss": 0.347,
+      "step": 1078
+    },
+    {
+      "epoch": 1.2316029663434114,
+      "grad_norm": 0.3276186811143977,
+      "learning_rate": 3.2748414376321354e-05,
+      "loss": 0.3712,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2327438676554479,
+      "grad_norm": 0.34011289116421417,
+      "learning_rate": 3.272727272727273e-05,
+      "loss": 0.3789,
+      "step": 1080
+    },
+    {
+      "epoch": 1.2338847689674843,
+      "grad_norm": 0.3430954560333754,
+      "learning_rate": 3.27061310782241e-05,
+      "loss": 0.388,
+      "step": 1081
+    },
+    {
+      "epoch": 1.2350256702795208,
+      "grad_norm": 0.28511422528620944,
+      "learning_rate": 3.268498942917548e-05,
+      "loss": 0.3712,
+      "step": 1082
+    },
+    {
+      "epoch": 1.2361665715915573,
+      "grad_norm": 0.3469153679842275,
+      "learning_rate": 3.266384778012685e-05,
+      "loss": 0.3479,
+      "step": 1083
+    },
+    {
+      "epoch": 1.2373074729035938,
+      "grad_norm": 0.31225044583573147,
+      "learning_rate": 3.264270613107823e-05,
+      "loss": 0.366,
+      "step": 1084
+    },
+    {
+      "epoch": 1.2384483742156303,
+      "grad_norm": 0.3205379010988631,
+      "learning_rate": 3.26215644820296e-05,
+      "loss": 0.3818,
+      "step": 1085
+    },
+    {
+      "epoch": 1.239589275527667,
+      "grad_norm": 0.2840480563572048,
+      "learning_rate": 3.2600422832980976e-05,
+      "loss": 0.3624,
+      "step": 1086
+    },
+    {
+      "epoch": 1.2407301768397034,
+      "grad_norm": 0.3526143479765244,
+      "learning_rate": 3.257928118393235e-05,
+      "loss": 0.3703,
+      "step": 1087
+    },
+    {
+      "epoch": 1.24187107815174,
+      "grad_norm": 0.2684134070992963,
+      "learning_rate": 3.2558139534883724e-05,
+      "loss": 0.3577,
+      "step": 1088
+    },
+    {
+      "epoch": 1.2430119794637764,
+      "grad_norm": 0.32212969469475267,
+      "learning_rate": 3.2536997885835095e-05,
+      "loss": 0.3402,
+      "step": 1089
+    },
+    {
+      "epoch": 1.2441528807758129,
+      "grad_norm": 0.3240375629745611,
+      "learning_rate": 3.2515856236786466e-05,
+      "loss": 0.394,
+      "step": 1090
+    },
+    {
+      "epoch": 1.2452937820878494,
+      "grad_norm": 0.2829430362775589,
+      "learning_rate": 3.249471458773784e-05,
+      "loss": 0.3738,
+      "step": 1091
+    },
+    {
+      "epoch": 1.2464346833998858,
+      "grad_norm": 0.3415069305615634,
+      "learning_rate": 3.247357293868922e-05,
+      "loss": 0.3772,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2475755847119223,
+      "grad_norm": 0.35519434270985334,
+      "learning_rate": 3.245243128964059e-05,
+      "loss": 0.3958,
+      "step": 1093
+    },
+    {
+      "epoch": 1.248716486023959,
+      "grad_norm": 0.2887851058517861,
+      "learning_rate": 3.243128964059197e-05,
+      "loss": 0.3746,
+      "step": 1094
+    },
+    {
+      "epoch": 1.2498573873359955,
+      "grad_norm": 0.3456876409406876,
+      "learning_rate": 3.2410147991543346e-05,
+      "loss": 0.3647,
+      "step": 1095
+    },
+    {
+      "epoch": 1.250998288648032,
+      "grad_norm": 0.2923005180020809,
+      "learning_rate": 3.238900634249472e-05,
+      "loss": 0.3703,
+      "step": 1096
+    },
+    {
+      "epoch": 1.2521391899600685,
+      "grad_norm": 0.33057352412548374,
+      "learning_rate": 3.2367864693446094e-05,
+      "loss": 0.3643,
+      "step": 1097
+    },
+    {
+      "epoch": 1.253280091272105,
+      "grad_norm": 0.36994392272110005,
+      "learning_rate": 3.2346723044397465e-05,
+      "loss": 0.3662,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2544209925841414,
+      "grad_norm": 0.2935627570419225,
+      "learning_rate": 3.2325581395348836e-05,
+      "loss": 0.3835,
+      "step": 1099
+    },
+    {
+      "epoch": 1.2555618938961781,
+      "grad_norm": 0.35305608593147453,
+      "learning_rate": 3.230443974630021e-05,
+      "loss": 0.3741,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2567027952082146,
+      "grad_norm": 0.30256830350387487,
+      "learning_rate": 3.2283298097251584e-05,
+      "loss": 0.3688,
+      "step": 1101
+    },
+    {
+      "epoch": 1.257843696520251,
+      "grad_norm": 0.3460982440766267,
+      "learning_rate": 3.226215644820296e-05,
+      "loss": 0.3653,
+      "step": 1102
+    },
+    {
+      "epoch": 1.2589845978322876,
+      "grad_norm": 0.34403905347729535,
+      "learning_rate": 3.224101479915433e-05,
+      "loss": 0.3857,
+      "step": 1103
+    },
+    {
+      "epoch": 1.260125499144324,
+      "grad_norm": 0.3530369430889585,
+      "learning_rate": 3.221987315010571e-05,
+      "loss": 0.372,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2612664004563605,
+      "grad_norm": 0.3342161669182045,
+      "learning_rate": 3.219873150105708e-05,
+      "loss": 0.3725,
+      "step": 1105
+    },
+    {
+      "epoch": 1.262407301768397,
+      "grad_norm": 0.36362820655955347,
+      "learning_rate": 3.217758985200846e-05,
+      "loss": 0.3684,
+      "step": 1106
+    },
+    {
+      "epoch": 1.2635482030804335,
+      "grad_norm": 0.3241577589711961,
+      "learning_rate": 3.2156448202959835e-05,
+      "loss": 0.3756,
+      "step": 1107
+    },
+    {
+      "epoch": 1.26468910439247,
+      "grad_norm": 0.29636070346273397,
+      "learning_rate": 3.2135306553911206e-05,
+      "loss": 0.3747,
+      "step": 1108
+    },
+    {
+      "epoch": 1.2658300057045064,
+      "grad_norm": 0.3752679637351401,
+      "learning_rate": 3.211416490486258e-05,
+      "loss": 0.3628,
+      "step": 1109
+    },
+    {
+      "epoch": 1.2669709070165431,
+      "grad_norm": 0.36063049502355077,
+      "learning_rate": 3.2093023255813954e-05,
+      "loss": 0.3788,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2681118083285796,
+      "grad_norm": 0.36063293655342027,
+      "learning_rate": 3.207188160676533e-05,
+      "loss": 0.3647,
+      "step": 1111
+    },
+    {
+      "epoch": 1.269252709640616,
+      "grad_norm": 0.3876467932971257,
+      "learning_rate": 3.20507399577167e-05,
+      "loss": 0.3588,
+      "step": 1112
+    },
+    {
+      "epoch": 1.2703936109526526,
+      "grad_norm": 0.3877619528273815,
+      "learning_rate": 3.202959830866808e-05,
+      "loss": 0.3611,
+      "step": 1113
+    },
+    {
+      "epoch": 1.271534512264689,
+      "grad_norm": 0.3317695650930697,
+      "learning_rate": 3.200845665961945e-05,
+      "loss": 0.3613,
+      "step": 1114
+    },
+    {
+      "epoch": 1.2726754135767255,
+      "grad_norm": 0.3586599244171131,
+      "learning_rate": 3.198731501057083e-05,
+      "loss": 0.3581,
+      "step": 1115
+    },
+    {
+      "epoch": 1.2738163148887622,
+      "grad_norm": 0.32768476193965773,
+      "learning_rate": 3.19661733615222e-05,
+      "loss": 0.3624,
+      "step": 1116
+    },
+    {
+      "epoch": 1.2749572162007987,
+      "grad_norm": 0.342066085799505,
+      "learning_rate": 3.194503171247357e-05,
+      "loss": 0.3931,
+      "step": 1117
+    },
+    {
+      "epoch": 1.2760981175128352,
+      "grad_norm": 0.3212269324643488,
+      "learning_rate": 3.1923890063424947e-05,
+      "loss": 0.3824,
+      "step": 1118
+    },
+    {
+      "epoch": 1.2772390188248717,
+      "grad_norm": 0.34149971973938353,
+      "learning_rate": 3.1902748414376324e-05,
+      "loss": 0.3761,
+      "step": 1119
+    },
+    {
+      "epoch": 1.2783799201369082,
+      "grad_norm": 0.3708195020111092,
+      "learning_rate": 3.18816067653277e-05,
+      "loss": 0.3832,
+      "step": 1120
+    },
+    {
+      "epoch": 1.2795208214489446,
+      "grad_norm": 0.3662946332428756,
+      "learning_rate": 3.186046511627907e-05,
+      "loss": 0.3788,
+      "step": 1121
+    },
+    {
+      "epoch": 1.2806617227609811,
+      "grad_norm": 0.2718084249660657,
+      "learning_rate": 3.183932346723045e-05,
+      "loss": 0.3762,
+      "step": 1122
+    },
+    {
+      "epoch": 1.2818026240730176,
+      "grad_norm": 0.3549508122271783,
+      "learning_rate": 3.181818181818182e-05,
+      "loss": 0.3507,
+      "step": 1123
+    },
+    {
+      "epoch": 1.282943525385054,
+      "grad_norm": 0.3456469078705619,
+      "learning_rate": 3.17970401691332e-05,
+      "loss": 0.3788,
+      "step": 1124
+    },
+    {
+      "epoch": 1.2840844266970908,
+      "grad_norm": 0.29275976639218304,
+      "learning_rate": 3.177589852008457e-05,
+      "loss": 0.3637,
+      "step": 1125
+    },
+    {
+      "epoch": 1.2852253280091273,
+      "grad_norm": 0.35290128309001717,
+      "learning_rate": 3.175475687103594e-05,
+      "loss": 0.3851,
+      "step": 1126
+    },
+    {
+      "epoch": 1.2863662293211637,
+      "grad_norm": 0.31844354937645575,
+      "learning_rate": 3.173361522198732e-05,
+      "loss": 0.3786,
+      "step": 1127
+    },
+    {
+      "epoch": 1.2875071306332002,
+      "grad_norm": 0.3607279985421439,
+      "learning_rate": 3.171247357293869e-05,
+      "loss": 0.3584,
+      "step": 1128
+    },
+    {
+      "epoch": 1.2886480319452367,
+      "grad_norm": 0.3258151642831157,
+      "learning_rate": 3.1691331923890065e-05,
+      "loss": 0.3807,
+      "step": 1129
+    },
+    {
+      "epoch": 1.2897889332572732,
+      "grad_norm": 0.3584686282000858,
+      "learning_rate": 3.1670190274841435e-05,
+      "loss": 0.4076,
+      "step": 1130
+    },
+    {
+      "epoch": 1.2909298345693099,
+      "grad_norm": 0.34153478084703925,
+      "learning_rate": 3.164904862579281e-05,
+      "loss": 0.3538,
+      "step": 1131
+    },
+    {
+      "epoch": 1.2920707358813464,
+      "grad_norm": 0.3176502770934002,
+      "learning_rate": 3.162790697674419e-05,
+      "loss": 0.3794,
+      "step": 1132
+    },
+    {
+      "epoch": 1.2932116371933828,
+      "grad_norm": 0.3577524872996408,
+      "learning_rate": 3.160676532769556e-05,
+      "loss": 0.365,
+      "step": 1133
+    },
+    {
+      "epoch": 1.2943525385054193,
+      "grad_norm": 0.33076084058529676,
+      "learning_rate": 3.158562367864694e-05,
+      "loss": 0.3514,
+      "step": 1134
+    },
+    {
+      "epoch": 1.2954934398174558,
+      "grad_norm": 0.38879916007923476,
+      "learning_rate": 3.156448202959831e-05,
+      "loss": 0.3743,
+      "step": 1135
+    },
+    {
+      "epoch": 1.2966343411294923,
+      "grad_norm": 0.3635936827843266,
+      "learning_rate": 3.154334038054969e-05,
+      "loss": 0.3655,
+      "step": 1136
+    },
+    {
+      "epoch": 1.2977752424415288,
+      "grad_norm": 0.3476006034419538,
+      "learning_rate": 3.152219873150106e-05,
+      "loss": 0.3889,
+      "step": 1137
+    },
+    {
+      "epoch": 1.2989161437535652,
+      "grad_norm": 0.3667463941248487,
+      "learning_rate": 3.1501057082452435e-05,
+      "loss": 0.3971,
+      "step": 1138
+    },
+    {
+      "epoch": 1.3000570450656017,
+      "grad_norm": 0.310954947386153,
+      "learning_rate": 3.1479915433403805e-05,
+      "loss": 0.3753,
+      "step": 1139
+    },
+    {
+      "epoch": 1.3011979463776384,
+      "grad_norm": 0.36203472318239655,
+      "learning_rate": 3.145877378435518e-05,
+      "loss": 0.3613,
+      "step": 1140
+    },
+    {
+      "epoch": 1.302338847689675,
+      "grad_norm": 0.3494822072314133,
+      "learning_rate": 3.1437632135306554e-05,
+      "loss": 0.3851,
+      "step": 1141
+    },
+    {
+      "epoch": 1.3034797490017114,
+      "grad_norm": 0.3516976871249764,
+      "learning_rate": 3.141649048625793e-05,
+      "loss": 0.3574,
+      "step": 1142
+    },
+    {
+      "epoch": 1.3046206503137479,
+      "grad_norm": 0.26973236707033366,
+      "learning_rate": 3.13953488372093e-05,
+      "loss": 0.3791,
+      "step": 1143
+    },
+    {
+      "epoch": 1.3057615516257843,
+      "grad_norm": 0.3429365439599929,
+      "learning_rate": 3.137420718816067e-05,
+      "loss": 0.366,
+      "step": 1144
+    },
+    {
+      "epoch": 1.3069024529378208,
+      "grad_norm": 0.3165521363453188,
+      "learning_rate": 3.135306553911205e-05,
+      "loss": 0.3884,
+      "step": 1145
+    },
+    {
+      "epoch": 1.3080433542498575,
+      "grad_norm": 0.297830837920527,
+      "learning_rate": 3.133192389006343e-05,
+      "loss": 0.3801,
+      "step": 1146
+    },
+    {
+      "epoch": 1.309184255561894,
+      "grad_norm": 0.2876870489632023,
+      "learning_rate": 3.1310782241014805e-05,
+      "loss": 0.3721,
+      "step": 1147
+    },
+    {
+      "epoch": 1.3103251568739305,
+      "grad_norm": 0.3037731699134318,
+      "learning_rate": 3.1289640591966176e-05,
+      "loss": 0.3752,
+      "step": 1148
+    },
+    {
+      "epoch": 1.311466058185967,
+      "grad_norm": 0.32611078798800086,
+      "learning_rate": 3.126849894291755e-05,
+      "loss": 0.3733,
+      "step": 1149
+    },
+    {
+      "epoch": 1.3126069594980034,
+      "grad_norm": 0.2895325920295833,
+      "learning_rate": 3.1247357293868924e-05,
+      "loss": 0.3544,
+      "step": 1150
+    },
+    {
+      "epoch": 1.31374786081004,
+      "grad_norm": 0.3654499418349549,
+      "learning_rate": 3.1226215644820294e-05,
+      "loss": 0.3733,
+      "step": 1151
+    },
+    {
+      "epoch": 1.3148887621220764,
+      "grad_norm": 0.36567998053713596,
+      "learning_rate": 3.120507399577167e-05,
+      "loss": 0.3732,
+      "step": 1152
+    },
+    {
+      "epoch": 1.3160296634341129,
+      "grad_norm": 0.34435773909859874,
+      "learning_rate": 3.118393234672304e-05,
+      "loss": 0.3913,
+      "step": 1153
+    },
+    {
+      "epoch": 1.3171705647461494,
+      "grad_norm": 0.3957305592645215,
+      "learning_rate": 3.116279069767442e-05,
+      "loss": 0.3661,
+      "step": 1154
+    },
+    {
+      "epoch": 1.3183114660581858,
+      "grad_norm": 0.3374259528454154,
+      "learning_rate": 3.114164904862579e-05,
+      "loss": 0.3574,
+      "step": 1155
+    },
+    {
+      "epoch": 1.3194523673702225,
+      "grad_norm": 0.2937491447581557,
+      "learning_rate": 3.112050739957717e-05,
+      "loss": 0.3655,
+      "step": 1156
+    },
+    {
+      "epoch": 1.320593268682259,
+      "grad_norm": 0.36541534067539616,
+      "learning_rate": 3.109936575052854e-05,
+      "loss": 0.3967,
+      "step": 1157
+    },
+    {
+      "epoch": 1.3217341699942955,
+      "grad_norm": 0.32171994431469614,
+      "learning_rate": 3.1078224101479916e-05,
+      "loss": 0.3623,
+      "step": 1158
+    },
+    {
+      "epoch": 1.322875071306332,
+      "grad_norm": 0.32768929811903286,
+      "learning_rate": 3.1057082452431294e-05,
+      "loss": 0.381,
+      "step": 1159
+    },
+    {
+      "epoch": 1.3240159726183685,
+      "grad_norm": 0.28688520786573474,
+      "learning_rate": 3.1035940803382664e-05,
+      "loss": 0.3614,
+      "step": 1160
+    },
+    {
+      "epoch": 1.3251568739304052,
+      "grad_norm": 0.3447454760055943,
+      "learning_rate": 3.101479915433404e-05,
+      "loss": 0.3835,
+      "step": 1161
+    },
+    {
+      "epoch": 1.3262977752424416,
+      "grad_norm": 0.30834089409956156,
+      "learning_rate": 3.099365750528541e-05,
+      "loss": 0.3676,
+      "step": 1162
+    },
+    {
+      "epoch": 1.3274386765544781,
+      "grad_norm": 0.3769175020167887,
+      "learning_rate": 3.097251585623679e-05,
+      "loss": 0.3812,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3285795778665146,
+      "grad_norm": 0.3603398260692364,
+      "learning_rate": 3.095137420718816e-05,
+      "loss": 0.3885,
+      "step": 1164
+    },
+    {
+      "epoch": 1.329720479178551,
+      "grad_norm": 0.3818452964334466,
+      "learning_rate": 3.093023255813954e-05,
+      "loss": 0.3666,
+      "step": 1165
+    },
+    {
+      "epoch": 1.3308613804905876,
+      "grad_norm": 0.341137726154606,
+      "learning_rate": 3.090909090909091e-05,
+      "loss": 0.3626,
+      "step": 1166
+    },
+    {
+      "epoch": 1.332002281802624,
+      "grad_norm": 0.3677916501891261,
+      "learning_rate": 3.0887949260042286e-05,
+      "loss": 0.3742,
+      "step": 1167
+    },
+    {
+      "epoch": 1.3331431831146605,
+      "grad_norm": 0.35238646235818044,
+      "learning_rate": 3.086680761099366e-05,
+      "loss": 0.3776,
+      "step": 1168
+    },
+    {
+      "epoch": 1.334284084426697,
+      "grad_norm": 0.3241309831420008,
+      "learning_rate": 3.084566596194503e-05,
+      "loss": 0.4018,
+      "step": 1169
+    },
+    {
+      "epoch": 1.3354249857387335,
+      "grad_norm": 0.4211931838361165,
+      "learning_rate": 3.0824524312896405e-05,
+      "loss": 0.3666,
+      "step": 1170
+    },
+    {
+      "epoch": 1.3365658870507702,
+      "grad_norm": 0.3432763006800476,
+      "learning_rate": 3.080338266384778e-05,
+      "loss": 0.3815,
+      "step": 1171
+    },
+    {
+      "epoch": 1.3377067883628067,
+      "grad_norm": 0.3527700767350808,
+      "learning_rate": 3.078224101479916e-05,
+      "loss": 0.3583,
+      "step": 1172
+    },
+    {
+      "epoch": 1.3388476896748431,
+      "grad_norm": 0.37964149901308397,
+      "learning_rate": 3.076109936575053e-05,
+      "loss": 0.3579,
+      "step": 1173
+    },
+    {
+      "epoch": 1.3399885909868796,
+      "grad_norm": 0.28730602914506537,
+      "learning_rate": 3.073995771670191e-05,
+      "loss": 0.3468,
+      "step": 1174
+    },
+    {
+      "epoch": 1.341129492298916,
+      "grad_norm": 0.3002903212997977,
+      "learning_rate": 3.071881606765328e-05,
+      "loss": 0.3559,
+      "step": 1175
+    },
+    {
+      "epoch": 1.3422703936109526,
+      "grad_norm": 0.3297601544213069,
+      "learning_rate": 3.0697674418604656e-05,
+      "loss": 0.3726,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3434112949229893,
+      "grad_norm": 0.2746332939414886,
+      "learning_rate": 3.067653276955603e-05,
+      "loss": 0.3701,
+      "step": 1177
+    },
+    {
+      "epoch": 1.3445521962350258,
+      "grad_norm": 0.260651600604371,
+      "learning_rate": 3.06553911205074e-05,
+      "loss": 0.3519,
+      "step": 1178
+    },
+    {
+      "epoch": 1.3456930975470622,
+      "grad_norm": 0.30888849266481555,
+      "learning_rate": 3.0634249471458775e-05,
+      "loss": 0.4036,
+      "step": 1179
+    },
+    {
+      "epoch": 1.3468339988590987,
+      "grad_norm": 0.2709838548905517,
+      "learning_rate": 3.0613107822410146e-05,
+      "loss": 0.3742,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3479749001711352,
+      "grad_norm": 0.28410459677729055,
+      "learning_rate": 3.059196617336152e-05,
+      "loss": 0.3694,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3491158014831717,
+      "grad_norm": 0.27240207939222316,
+      "learning_rate": 3.0570824524312894e-05,
+      "loss": 0.3541,
+      "step": 1182
+    },
+    {
+      "epoch": 1.3502567027952082,
+      "grad_norm": 0.28022840245663483,
+      "learning_rate": 3.054968287526427e-05,
+      "loss": 0.3556,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3513976041072446,
+      "grad_norm": 0.3026651192991682,
+      "learning_rate": 3.052854122621565e-05,
+      "loss": 0.3787,
+      "step": 1184
+    },
+    {
+      "epoch": 1.3525385054192811,
+      "grad_norm": 0.3221082586985108,
+      "learning_rate": 3.0507399577167023e-05,
+      "loss": 0.3677,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3536794067313178,
+      "grad_norm": 0.2993724031412757,
+      "learning_rate": 3.0486257928118394e-05,
+      "loss": 0.3961,
+      "step": 1186
+    },
+    {
+      "epoch": 1.3548203080433543,
+      "grad_norm": 0.36575229654879793,
+      "learning_rate": 3.0465116279069768e-05,
+      "loss": 0.3806,
+      "step": 1187
+    },
+    {
+      "epoch": 1.3559612093553908,
+      "grad_norm": 0.2618840843711902,
+      "learning_rate": 3.0443974630021145e-05,
+      "loss": 0.3809,
+      "step": 1188
+    },
+    {
+      "epoch": 1.3571021106674273,
+      "grad_norm": 0.33259999621304476,
+      "learning_rate": 3.0422832980972516e-05,
+      "loss": 0.3664,
+      "step": 1189
+    },
+    {
+      "epoch": 1.3582430119794637,
+      "grad_norm": 0.2906615103162758,
+      "learning_rate": 3.0401691331923893e-05,
+      "loss": 0.3631,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3593839132915002,
+      "grad_norm": 0.30230407067896337,
+      "learning_rate": 3.0380549682875264e-05,
+      "loss": 0.3699,
+      "step": 1191
+    },
+    {
+      "epoch": 1.360524814603537,
+      "grad_norm": 0.31210182393997005,
+      "learning_rate": 3.035940803382664e-05,
+      "loss": 0.3714,
+      "step": 1192
+    },
+    {
+      "epoch": 1.3616657159155734,
+      "grad_norm": 0.2808150657733752,
+      "learning_rate": 3.0338266384778012e-05,
+      "loss": 0.3594,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3628066172276099,
+      "grad_norm": 0.3136491102076318,
+      "learning_rate": 3.031712473572939e-05,
+      "loss": 0.3671,
+      "step": 1194
+    },
+    {
+      "epoch": 1.3639475185396464,
+      "grad_norm": 0.32544208291676757,
+      "learning_rate": 3.0295983086680764e-05,
+      "loss": 0.3636,
+      "step": 1195
+    },
+    {
+      "epoch": 1.3650884198516828,
+      "grad_norm": 0.29240635927645753,
+      "learning_rate": 3.0274841437632134e-05,
+      "loss": 0.3688,
+      "step": 1196
+    },
+    {
+      "epoch": 1.3662293211637193,
+      "grad_norm": 0.3045047299850035,
+      "learning_rate": 3.0253699788583512e-05,
+      "loss": 0.3695,
+      "step": 1197
+    },
+    {
+      "epoch": 1.3673702224757558,
+      "grad_norm": 0.3038195498825899,
+      "learning_rate": 3.0232558139534883e-05,
+      "loss": 0.3655,
+      "step": 1198
+    },
+    {
+      "epoch": 1.3685111237877923,
+      "grad_norm": 0.34280384948848547,
+      "learning_rate": 3.021141649048626e-05,
+      "loss": 0.3787,
+      "step": 1199
+    },
+    {
+      "epoch": 1.3696520250998288,
+      "grad_norm": 0.29537008807280474,
+      "learning_rate": 3.019027484143763e-05,
+      "loss": 0.3649,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3707929264118655,
+      "grad_norm": 0.2988844933069723,
+      "learning_rate": 3.0169133192389008e-05,
+      "loss": 0.3577,
+      "step": 1201
+    },
+    {
+      "epoch": 1.371933827723902,
+      "grad_norm": 0.3038068470192822,
+      "learning_rate": 3.0147991543340382e-05,
+      "loss": 0.3624,
+      "step": 1202
+    },
+    {
+      "epoch": 1.3730747290359384,
+      "grad_norm": 0.3195670739387878,
+      "learning_rate": 3.012684989429176e-05,
+      "loss": 0.3622,
+      "step": 1203
+    },
+    {
+      "epoch": 1.374215630347975,
+      "grad_norm": 0.33884359703682376,
+      "learning_rate": 3.010570824524313e-05,
+      "loss": 0.3712,
+      "step": 1204
+    },
+    {
+      "epoch": 1.3753565316600114,
+      "grad_norm": 0.30921551673694203,
+      "learning_rate": 3.00845665961945e-05,
+      "loss": 0.3871,
+      "step": 1205
+    },
+    {
+      "epoch": 1.3764974329720479,
+      "grad_norm": 0.38097121845936377,
+      "learning_rate": 3.006342494714588e-05,
+      "loss": 0.3861,
+      "step": 1206
+    },
+    {
+      "epoch": 1.3776383342840846,
+      "grad_norm": 0.2684259417323635,
+      "learning_rate": 3.0042283298097253e-05,
+      "loss": 0.3782,
+      "step": 1207
+    },
+    {
+      "epoch": 1.378779235596121,
+      "grad_norm": 0.2956290222982241,
+      "learning_rate": 3.002114164904863e-05,
+      "loss": 0.3713,
+      "step": 1208
+    },
+    {
+      "epoch": 1.3799201369081575,
+      "grad_norm": 0.370180759036463,
+      "learning_rate": 3e-05,
+      "loss": 0.3828,
+      "step": 1209
+    },
+    {
+      "epoch": 1.381061038220194,
+      "grad_norm": 0.25836075503652833,
+      "learning_rate": 2.9978858350951378e-05,
+      "loss": 0.3647,
+      "step": 1210
+    },
+    {
+      "epoch": 1.3822019395322305,
+      "grad_norm": 0.3468462978558045,
+      "learning_rate": 2.995771670190275e-05,
+      "loss": 0.3765,
+      "step": 1211
+    },
+    {
+      "epoch": 1.383342840844267,
+      "grad_norm": 0.2824173533303826,
+      "learning_rate": 2.9936575052854126e-05,
+      "loss": 0.3681,
+      "step": 1212
+    },
+    {
+      "epoch": 1.3844837421563034,
+      "grad_norm": 0.28736316984448956,
+      "learning_rate": 2.9915433403805497e-05,
+      "loss": 0.3625,
+      "step": 1213
+    },
+    {
+      "epoch": 1.38562464346834,
+      "grad_norm": 0.33703904092494497,
+      "learning_rate": 2.989429175475687e-05,
+      "loss": 0.3777,
+      "step": 1214
+    },
+    {
+      "epoch": 1.3867655447803764,
+      "grad_norm": 0.29634799772730275,
+      "learning_rate": 2.987315010570825e-05,
+      "loss": 0.3608,
+      "step": 1215
+    },
+    {
+      "epoch": 1.3879064460924129,
+      "grad_norm": 0.3384257392542635,
+      "learning_rate": 2.985200845665962e-05,
+      "loss": 0.3602,
+      "step": 1216
+    },
+    {
+      "epoch": 1.3890473474044496,
+      "grad_norm": 0.27944486323405615,
+      "learning_rate": 2.9830866807610997e-05,
+      "loss": 0.3464,
+      "step": 1217
+    },
+    {
+      "epoch": 1.390188248716486,
+      "grad_norm": 0.32014182507961697,
+      "learning_rate": 2.9809725158562367e-05,
+      "loss": 0.3739,
+      "step": 1218
+    },
+    {
+      "epoch": 1.3913291500285225,
+      "grad_norm": 0.33271962982884457,
+      "learning_rate": 2.9788583509513745e-05,
+      "loss": 0.3969,
+      "step": 1219
+    },
+    {
+      "epoch": 1.392470051340559,
+      "grad_norm": 0.30683561080527283,
+      "learning_rate": 2.976744186046512e-05,
+      "loss": 0.3896,
+      "step": 1220
+    },
+    {
+      "epoch": 1.3936109526525955,
+      "grad_norm": 0.31727387176204047,
+      "learning_rate": 2.9746300211416493e-05,
+      "loss": 0.3706,
+      "step": 1221
+    },
+    {
+      "epoch": 1.3947518539646322,
+      "grad_norm": 0.35232887418187936,
+      "learning_rate": 2.9725158562367867e-05,
+      "loss": 0.357,
+      "step": 1222
+    },
+    {
+      "epoch": 1.3958927552766687,
+      "grad_norm": 0.2941768931758356,
+      "learning_rate": 2.9704016913319238e-05,
+      "loss": 0.3916,
+      "step": 1223
+    },
+    {
+      "epoch": 1.3970336565887052,
+      "grad_norm": 0.38849939713177317,
+      "learning_rate": 2.9682875264270615e-05,
+      "loss": 0.3731,
+      "step": 1224
+    },
+    {
+      "epoch": 1.3981745579007416,
+      "grad_norm": 0.2733119638416779,
+      "learning_rate": 2.9661733615221986e-05,
+      "loss": 0.3508,
+      "step": 1225
+    },
+    {
+      "epoch": 1.3993154592127781,
+      "grad_norm": 0.3141094815641651,
+      "learning_rate": 2.9640591966173363e-05,
+      "loss": 0.3747,
+      "step": 1226
+    },
+    {
+      "epoch": 1.4004563605248146,
+      "grad_norm": 0.25985281656595854,
+      "learning_rate": 2.9619450317124737e-05,
+      "loss": 0.3561,
+      "step": 1227
+    },
+    {
+      "epoch": 1.401597261836851,
+      "grad_norm": 0.30023051357215597,
+      "learning_rate": 2.9598308668076115e-05,
+      "loss": 0.3632,
+      "step": 1228
+    },
+    {
+      "epoch": 1.4027381631488876,
+      "grad_norm": 0.3000718719373125,
+      "learning_rate": 2.9577167019027486e-05,
+      "loss": 0.3793,
+      "step": 1229
+    },
+    {
+      "epoch": 1.403879064460924,
+      "grad_norm": 0.2784816189918396,
+      "learning_rate": 2.9556025369978863e-05,
+      "loss": 0.3622,
+      "step": 1230
+    },
+    {
+      "epoch": 1.4050199657729605,
+      "grad_norm": 0.33685900817290626,
+      "learning_rate": 2.9534883720930234e-05,
+      "loss": 0.3782,
+      "step": 1231
+    },
+    {
+      "epoch": 1.4061608670849972,
+      "grad_norm": 0.27572068061345933,
+      "learning_rate": 2.9513742071881604e-05,
+      "loss": 0.3622,
+      "step": 1232
+    },
+    {
+      "epoch": 1.4073017683970337,
+      "grad_norm": 0.31338924633780607,
+      "learning_rate": 2.9492600422832982e-05,
+      "loss": 0.381,
+      "step": 1233
+    },
+    {
+      "epoch": 1.4084426697090702,
+      "grad_norm": 0.2734014953931813,
+      "learning_rate": 2.9471458773784356e-05,
+      "loss": 0.3546,
+      "step": 1234
+    },
+    {
+      "epoch": 1.4095835710211067,
+      "grad_norm": 0.3026690395161937,
+      "learning_rate": 2.9450317124735733e-05,
+      "loss": 0.3628,
+      "step": 1235
+    },
+    {
+      "epoch": 1.4107244723331431,
+      "grad_norm": 0.3562181622146391,
+      "learning_rate": 2.9429175475687104e-05,
+      "loss": 0.3561,
+      "step": 1236
+    },
+    {
+      "epoch": 1.4118653736451796,
+      "grad_norm": 0.2768764956360295,
+      "learning_rate": 2.940803382663848e-05,
+      "loss": 0.3779,
+      "step": 1237
+    },
+    {
+      "epoch": 1.4130062749572163,
+      "grad_norm": 0.3441377334550278,
+      "learning_rate": 2.9386892177589852e-05,
+      "loss": 0.3666,
+      "step": 1238
+    },
+    {
+      "epoch": 1.4141471762692528,
+      "grad_norm": 0.3387001258999047,
+      "learning_rate": 2.9365750528541226e-05,
+      "loss": 0.3789,
+      "step": 1239
+    },
+    {
+      "epoch": 1.4152880775812893,
+      "grad_norm": 0.30476509136714863,
+      "learning_rate": 2.9344608879492604e-05,
+      "loss": 0.3645,
+      "step": 1240
+    },
+    {
+      "epoch": 1.4164289788933258,
+      "grad_norm": 0.37908310081130964,
+      "learning_rate": 2.9323467230443974e-05,
+      "loss": 0.3774,
+      "step": 1241
+    },
+    {
+      "epoch": 1.4175698802053622,
+      "grad_norm": 0.32793422438669423,
+      "learning_rate": 2.9302325581395352e-05,
+      "loss": 0.3734,
+      "step": 1242
+    },
+    {
+      "epoch": 1.4187107815173987,
+      "grad_norm": 0.3277132324333076,
+      "learning_rate": 2.9281183932346723e-05,
+      "loss": 0.3904,
+      "step": 1243
+    },
+    {
+      "epoch": 1.4198516828294352,
+      "grad_norm": 0.3312953337286729,
+      "learning_rate": 2.92600422832981e-05,
+      "loss": 0.3692,
+      "step": 1244
+    },
+    {
+      "epoch": 1.4209925841414717,
+      "grad_norm": 0.2715055972763672,
+      "learning_rate": 2.923890063424947e-05,
+      "loss": 0.3655,
+      "step": 1245
+    },
+    {
+      "epoch": 1.4221334854535082,
+      "grad_norm": 0.2953587172579261,
+      "learning_rate": 2.9217758985200848e-05,
+      "loss": 0.3766,
+      "step": 1246
+    },
+    {
+      "epoch": 1.4232743867655449,
+      "grad_norm": 0.31388222966291707,
+      "learning_rate": 2.9196617336152222e-05,
+      "loss": 0.3674,
+      "step": 1247
+    },
+    {
+      "epoch": 1.4244152880775813,
+      "grad_norm": 0.35965122052908766,
+      "learning_rate": 2.9175475687103593e-05,
+      "loss": 0.3782,
+      "step": 1248
+    },
+    {
+      "epoch": 1.4255561893896178,
+      "grad_norm": 0.30486150383607696,
+      "learning_rate": 2.915433403805497e-05,
+      "loss": 0.368,
+      "step": 1249
+    },
+    {
+      "epoch": 1.4266970907016543,
+      "grad_norm": 0.39677093495998783,
+      "learning_rate": 2.913319238900634e-05,
+      "loss": 0.3683,
+      "step": 1250
+    },
+    {
+      "epoch": 1.4278379920136908,
+      "grad_norm": 0.3223810641459737,
+      "learning_rate": 2.911205073995772e-05,
+      "loss": 0.3689,
+      "step": 1251
+    },
+    {
+      "epoch": 1.4289788933257273,
+      "grad_norm": 0.30507032725118605,
+      "learning_rate": 2.909090909090909e-05,
+      "loss": 0.3705,
+      "step": 1252
+    },
+    {
+      "epoch": 1.430119794637764,
+      "grad_norm": 0.4194926217364012,
+      "learning_rate": 2.9069767441860467e-05,
+      "loss": 0.3687,
+      "step": 1253
+    },
+    {
+      "epoch": 1.4312606959498004,
+      "grad_norm": 0.2954199664136386,
+      "learning_rate": 2.904862579281184e-05,
+      "loss": 0.3728,
+      "step": 1254
+    },
+    {
+      "epoch": 1.432401597261837,
+      "grad_norm": 0.41519643808435247,
+      "learning_rate": 2.9027484143763218e-05,
+      "loss": 0.3616,
+      "step": 1255
+    },
+    {
+      "epoch": 1.4335424985738734,
+      "grad_norm": 0.37013820480001497,
+      "learning_rate": 2.900634249471459e-05,
+      "loss": 0.3631,
+      "step": 1256
+    },
+    {
+      "epoch": 1.4346833998859099,
+      "grad_norm": 0.3637646288174245,
+      "learning_rate": 2.898520084566596e-05,
+      "loss": 0.3705,
+      "step": 1257
+    },
+    {
+      "epoch": 1.4358243011979464,
+      "grad_norm": 0.3994429680732055,
+      "learning_rate": 2.8964059196617337e-05,
+      "loss": 0.3603,
+      "step": 1258
+    },
+    {
+      "epoch": 1.4369652025099828,
+      "grad_norm": 0.3261584175677973,
+      "learning_rate": 2.894291754756871e-05,
+      "loss": 0.3783,
+      "step": 1259
+    },
+    {
+      "epoch": 1.4381061038220193,
+      "grad_norm": 0.3535660656201677,
+      "learning_rate": 2.892177589852009e-05,
+      "loss": 0.3722,
+      "step": 1260
+    },
+    {
+      "epoch": 1.4392470051340558,
+      "grad_norm": 0.3547559486950115,
+      "learning_rate": 2.890063424947146e-05,
+      "loss": 0.385,
+      "step": 1261
+    },
+    {
+      "epoch": 1.4403879064460925,
+      "grad_norm": 0.27982112384758195,
+      "learning_rate": 2.8879492600422837e-05,
+      "loss": 0.3651,
+      "step": 1262
+    },
+    {
+      "epoch": 1.441528807758129,
+      "grad_norm": 0.33283939892409153,
+      "learning_rate": 2.8858350951374207e-05,
+      "loss": 0.3657,
+      "step": 1263
+    },
+    {
+      "epoch": 1.4426697090701655,
+      "grad_norm": 0.32733079528541736,
+      "learning_rate": 2.8837209302325585e-05,
+      "loss": 0.3787,
+      "step": 1264
+    },
+    {
+      "epoch": 1.443810610382202,
+      "grad_norm": 0.23586854044772235,
+      "learning_rate": 2.8816067653276956e-05,
+      "loss": 0.3585,
+      "step": 1265
+    },
+    {
+      "epoch": 1.4449515116942384,
+      "grad_norm": 0.3178604473656048,
+      "learning_rate": 2.879492600422833e-05,
+      "loss": 0.3617,
+      "step": 1266
+    },
+    {
+      "epoch": 1.446092413006275,
+      "grad_norm": 0.340401195862735,
+      "learning_rate": 2.8773784355179707e-05,
+      "loss": 0.3569,
+      "step": 1267
+    },
+    {
+      "epoch": 1.4472333143183116,
+      "grad_norm": 0.2832755699487134,
+      "learning_rate": 2.8752642706131078e-05,
+      "loss": 0.3752,
+      "step": 1268
+    },
+    {
+      "epoch": 1.448374215630348,
+      "grad_norm": 0.2816782125944015,
+      "learning_rate": 2.8731501057082455e-05,
+      "loss": 0.3694,
+      "step": 1269
+    },
+    {
+      "epoch": 1.4495151169423846,
+      "grad_norm": 0.2962764604229308,
+      "learning_rate": 2.8710359408033826e-05,
+      "loss": 0.3631,
+      "step": 1270
+    },
+    {
+      "epoch": 1.450656018254421,
+      "grad_norm": 0.3123088681993934,
+      "learning_rate": 2.8689217758985203e-05,
+      "loss": 0.3982,
+      "step": 1271
+    },
+    {
+      "epoch": 1.4517969195664575,
+      "grad_norm": 0.2589384687492925,
+      "learning_rate": 2.8668076109936574e-05,
+      "loss": 0.3663,
+      "step": 1272
+    },
+    {
+      "epoch": 1.452937820878494,
+      "grad_norm": 0.39661680089050527,
+      "learning_rate": 2.864693446088795e-05,
+      "loss": 0.3854,
+      "step": 1273
+    },
+    {
+      "epoch": 1.4540787221905305,
+      "grad_norm": 0.25225760585377377,
+      "learning_rate": 2.8625792811839326e-05,
+      "loss": 0.346,
+      "step": 1274
+    },
+    {
+      "epoch": 1.455219623502567,
+      "grad_norm": 0.3721747355722219,
+      "learning_rate": 2.8604651162790696e-05,
+      "loss": 0.3704,
+      "step": 1275
+    },
+    {
+      "epoch": 1.4563605248146034,
+      "grad_norm": 0.36112828697634103,
+      "learning_rate": 2.8583509513742074e-05,
+      "loss": 0.3779,
+      "step": 1276
+    },
+    {
+      "epoch": 1.45750142612664,
+      "grad_norm": 0.3074224938411391,
+      "learning_rate": 2.8562367864693444e-05,
+      "loss": 0.376,
+      "step": 1277
+    },
+    {
+      "epoch": 1.4586423274386766,
+      "grad_norm": 0.30091086011362805,
+      "learning_rate": 2.8541226215644822e-05,
+      "loss": 0.3551,
+      "step": 1278
+    },
+    {
+      "epoch": 1.459783228750713,
+      "grad_norm": 0.29332910598868295,
+      "learning_rate": 2.8520084566596196e-05,
+      "loss": 0.3588,
+      "step": 1279
+    },
+    {
+      "epoch": 1.4609241300627496,
+      "grad_norm": 0.27911592405959207,
+      "learning_rate": 2.8498942917547573e-05,
+      "loss": 0.3878,
+      "step": 1280
+    },
+    {
+      "epoch": 1.462065031374786,
+      "grad_norm": 0.26555424161809305,
+      "learning_rate": 2.8477801268498944e-05,
+      "loss": 0.3643,
+      "step": 1281
+    },
+    {
+      "epoch": 1.4632059326868225,
+      "grad_norm": 0.3114698661350436,
+      "learning_rate": 2.845665961945032e-05,
+      "loss": 0.3891,
+      "step": 1282
+    },
+    {
+      "epoch": 1.4643468339988592,
+      "grad_norm": 0.3417032679196954,
+      "learning_rate": 2.8435517970401692e-05,
+      "loss": 0.378,
+      "step": 1283
+    },
+    {
+      "epoch": 1.4654877353108957,
+      "grad_norm": 0.31269026412846357,
+      "learning_rate": 2.8414376321353063e-05,
+      "loss": 0.3632,
+      "step": 1284
+    },
+    {
+      "epoch": 1.4666286366229322,
+      "grad_norm": 0.2818282401954472,
+      "learning_rate": 2.839323467230444e-05,
+      "loss": 0.3857,
+      "step": 1285
+    },
+    {
+      "epoch": 1.4677695379349687,
+      "grad_norm": 0.3242138101510427,
+      "learning_rate": 2.8372093023255815e-05,
+      "loss": 0.3671,
+      "step": 1286
+    },
+    {
+      "epoch": 1.4689104392470052,
+      "grad_norm": 0.2899834506946572,
+      "learning_rate": 2.8350951374207192e-05,
+      "loss": 0.3672,
+      "step": 1287
+    },
+    {
+      "epoch": 1.4700513405590416,
+      "grad_norm": 0.3605761691098186,
+      "learning_rate": 2.8329809725158563e-05,
+      "loss": 0.3763,
+      "step": 1288
+    },
+    {
+      "epoch": 1.471192241871078,
+      "grad_norm": 0.33147697016869365,
+      "learning_rate": 2.830866807610994e-05,
+      "loss": 0.3525,
+      "step": 1289
+    },
+    {
+      "epoch": 1.4723331431831146,
+      "grad_norm": 0.325133970011816,
+      "learning_rate": 2.828752642706131e-05,
+      "loss": 0.3796,
+      "step": 1290
+    },
+    {
+      "epoch": 1.473474044495151,
+      "grad_norm": 0.33180923891881353,
+      "learning_rate": 2.8266384778012688e-05,
+      "loss": 0.3526,
+      "step": 1291
+    },
+    {
+      "epoch": 1.4746149458071875,
+      "grad_norm": 0.30830437851470766,
+      "learning_rate": 2.824524312896406e-05,
+      "loss": 0.3619,
+      "step": 1292
+    },
+    {
+      "epoch": 1.4757558471192243,
+      "grad_norm": 0.34255116892939635,
+      "learning_rate": 2.8224101479915433e-05,
+      "loss": 0.3557,
+      "step": 1293
+    },
+    {
+      "epoch": 1.4768967484312607,
+      "grad_norm": 0.3676335419437175,
+      "learning_rate": 2.820295983086681e-05,
+      "loss": 0.3826,
+      "step": 1294
+    },
+    {
+      "epoch": 1.4780376497432972,
+      "grad_norm": 0.3473952158913241,
+      "learning_rate": 2.818181818181818e-05,
+      "loss": 0.3696,
+      "step": 1295
+    },
+    {
+      "epoch": 1.4791785510553337,
+      "grad_norm": 0.3437354048268917,
+      "learning_rate": 2.816067653276956e-05,
+      "loss": 0.3697,
+      "step": 1296
+    },
+    {
+      "epoch": 1.4803194523673702,
+      "grad_norm": 0.34931348276289553,
+      "learning_rate": 2.813953488372093e-05,
+      "loss": 0.3867,
+      "step": 1297
+    },
+    {
+      "epoch": 1.4814603536794066,
+      "grad_norm": 0.3516920235276441,
+      "learning_rate": 2.8118393234672307e-05,
+      "loss": 0.3883,
+      "step": 1298
+    },
+    {
+      "epoch": 1.4826012549914434,
+      "grad_norm": 0.3369724056209684,
+      "learning_rate": 2.809725158562368e-05,
+      "loss": 0.3651,
+      "step": 1299
+    },
+    {
+      "epoch": 1.4837421563034798,
+      "grad_norm": 0.40382857751436074,
+      "learning_rate": 2.807610993657506e-05,
+      "loss": 0.3782,
+      "step": 1300
+    },
+    {
+      "epoch": 1.4848830576155163,
+      "grad_norm": 0.3095489263656417,
+      "learning_rate": 2.805496828752643e-05,
+      "loss": 0.3676,
+      "step": 1301
+    },
+    {
+      "epoch": 1.4860239589275528,
+      "grad_norm": 0.3503334326036625,
+      "learning_rate": 2.80338266384778e-05,
+      "loss": 0.3537,
+      "step": 1302
+    },
+    {
+      "epoch": 1.4871648602395893,
+      "grad_norm": 0.3311850238618065,
+      "learning_rate": 2.8012684989429177e-05,
+      "loss": 0.3624,
+      "step": 1303
+    },
+    {
+      "epoch": 1.4883057615516257,
+      "grad_norm": 0.3844061940542019,
+      "learning_rate": 2.7991543340380548e-05,
+      "loss": 0.365,
+      "step": 1304
+    },
+    {
+      "epoch": 1.4894466628636622,
+      "grad_norm": 0.30280777632485556,
+      "learning_rate": 2.7970401691331925e-05,
+      "loss": 0.3782,
+      "step": 1305
+    },
+    {
+      "epoch": 1.4905875641756987,
+      "grad_norm": 0.38759228805496665,
+      "learning_rate": 2.79492600422833e-05,
+      "loss": 0.3681,
+      "step": 1306
+    },
+    {
+      "epoch": 1.4917284654877352,
+      "grad_norm": 0.30175396840705576,
+      "learning_rate": 2.7928118393234677e-05,
+      "loss": 0.3587,
+      "step": 1307
+    },
+    {
+      "epoch": 1.492869366799772,
+      "grad_norm": 0.38149939268945615,
+      "learning_rate": 2.7906976744186048e-05,
+      "loss": 0.3876,
+      "step": 1308
+    },
+    {
+      "epoch": 1.4940102681118084,
+      "grad_norm": 0.3287914417483383,
+      "learning_rate": 2.7885835095137425e-05,
+      "loss": 0.3665,
+      "step": 1309
+    },
+    {
+      "epoch": 1.4951511694238449,
+      "grad_norm": 0.34723730837713296,
+      "learning_rate": 2.7864693446088796e-05,
+      "loss": 0.3799,
+      "step": 1310
+    },
+    {
+      "epoch": 1.4962920707358813,
+      "grad_norm": 0.3538733140564447,
+      "learning_rate": 2.784355179704017e-05,
+      "loss": 0.3705,
+      "step": 1311
+    },
+    {
+      "epoch": 1.4974329720479178,
+      "grad_norm": 0.28754610115885065,
+      "learning_rate": 2.7822410147991547e-05,
+      "loss": 0.3685,
+      "step": 1312
+    },
+    {
+      "epoch": 1.4985738733599543,
+      "grad_norm": 0.35941620116530376,
+      "learning_rate": 2.7801268498942918e-05,
+      "loss": 0.357,
+      "step": 1313
+    },
+    {
+      "epoch": 1.499714774671991,
+      "grad_norm": 0.28384788457688803,
+      "learning_rate": 2.7780126849894295e-05,
+      "loss": 0.3873,
+      "step": 1314
+    },
+    {
+      "epoch": 1.5008556759840275,
+      "grad_norm": 0.2921901422746972,
+      "learning_rate": 2.7758985200845666e-05,
+      "loss": 0.3576,
+      "step": 1315
+    },
+    {
+      "epoch": 1.501996577296064,
+      "grad_norm": 0.3069192169786753,
+      "learning_rate": 2.7737843551797044e-05,
+      "loss": 0.3912,
+      "step": 1316
+    },
+    {
+      "epoch": 1.5031374786081004,
+      "grad_norm": 0.28324765644412975,
+      "learning_rate": 2.7716701902748414e-05,
+      "loss": 0.3649,
+      "step": 1317
+    },
+    {
+      "epoch": 1.504278379920137,
+      "grad_norm": 0.2961812995454121,
+      "learning_rate": 2.769556025369979e-05,
+      "loss": 0.3482,
+      "step": 1318
+    },
+    {
+      "epoch": 1.5054192812321734,
+      "grad_norm": 0.32061195189305464,
+      "learning_rate": 2.7674418604651166e-05,
+      "loss": 0.3935,
+      "step": 1319
+    },
+    {
+      "epoch": 1.5065601825442099,
+      "grad_norm": 0.32516542840350554,
+      "learning_rate": 2.7653276955602536e-05,
+      "loss": 0.3589,
+      "step": 1320
+    },
+    {
+      "epoch": 1.5077010838562463,
+      "grad_norm": 0.3089964135811039,
+      "learning_rate": 2.7632135306553914e-05,
+      "loss": 0.3827,
+      "step": 1321
+    },
+    {
+      "epoch": 1.5088419851682828,
+      "grad_norm": 0.32133331637060614,
+      "learning_rate": 2.7610993657505285e-05,
+      "loss": 0.4043,
+      "step": 1322
+    },
+    {
+      "epoch": 1.5099828864803193,
+      "grad_norm": 0.31101837387755843,
+      "learning_rate": 2.7589852008456662e-05,
+      "loss": 0.3734,
+      "step": 1323
+    },
+    {
+      "epoch": 1.511123787792356,
+      "grad_norm": 0.3158550217117038,
+      "learning_rate": 2.7568710359408033e-05,
+      "loss": 0.3665,
+      "step": 1324
+    },
+    {
+      "epoch": 1.5122646891043925,
+      "grad_norm": 0.31029750240919934,
+      "learning_rate": 2.754756871035941e-05,
+      "loss": 0.3787,
+      "step": 1325
+    },
+    {
+      "epoch": 1.513405590416429,
+      "grad_norm": 0.3044823379693363,
+      "learning_rate": 2.7526427061310784e-05,
+      "loss": 0.3621,
+      "step": 1326
+    },
+    {
+      "epoch": 1.5145464917284654,
+      "grad_norm": 0.31325599526226944,
+      "learning_rate": 2.7505285412262155e-05,
+      "loss": 0.382,
+      "step": 1327
+    },
+    {
+      "epoch": 1.5156873930405022,
+      "grad_norm": 0.3126617606799847,
+      "learning_rate": 2.7484143763213532e-05,
+      "loss": 0.3404,
+      "step": 1328
+    },
+    {
+      "epoch": 1.5168282943525386,
+      "grad_norm": 0.3676906021498789,
+      "learning_rate": 2.7463002114164903e-05,
+      "loss": 0.3776,
+      "step": 1329
+    },
+    {
+      "epoch": 1.517969195664575,
+      "grad_norm": 0.32710198071826563,
+      "learning_rate": 2.744186046511628e-05,
+      "loss": 0.3789,
+      "step": 1330
+    },
+    {
+      "epoch": 1.5191100969766116,
+      "grad_norm": 0.29538279939957807,
+      "learning_rate": 2.7420718816067655e-05,
+      "loss": 0.3805,
+      "step": 1331
+    },
+    {
+      "epoch": 1.520250998288648,
+      "grad_norm": 0.4323552211327581,
+      "learning_rate": 2.7399577167019032e-05,
+      "loss": 0.3979,
+      "step": 1332
+    },
+    {
+      "epoch": 1.5213918996006845,
+      "grad_norm": 0.30159103255087294,
+      "learning_rate": 2.7378435517970403e-05,
+      "loss": 0.3609,
+      "step": 1333
+    },
+    {
+      "epoch": 1.522532800912721,
+      "grad_norm": 0.34952476702805724,
+      "learning_rate": 2.735729386892178e-05,
+      "loss": 0.392,
+      "step": 1334
+    },
+    {
+      "epoch": 1.5236737022247575,
+      "grad_norm": 0.4043924554170926,
+      "learning_rate": 2.733615221987315e-05,
+      "loss": 0.3681,
+      "step": 1335
+    },
+    {
+      "epoch": 1.524814603536794,
+      "grad_norm": 0.33704014908802854,
+      "learning_rate": 2.731501057082452e-05,
+      "loss": 0.3544,
+      "step": 1336
+    },
+    {
+      "epoch": 1.5259555048488305,
+      "grad_norm": 0.3811213990480285,
+      "learning_rate": 2.72938689217759e-05,
+      "loss": 0.3867,
+      "step": 1337
+    },
+    {
+      "epoch": 1.527096406160867,
+      "grad_norm": 0.3717796810852278,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 0.3931,
+      "step": 1338
+    },
+    {
+      "epoch": 1.5282373074729034,
+      "grad_norm": 0.3356101639570807,
+      "learning_rate": 2.725158562367865e-05,
+      "loss": 0.3809,
+      "step": 1339
+    },
+    {
+      "epoch": 1.5293782087849401,
+      "grad_norm": 0.3543267000421066,
+      "learning_rate": 2.723044397463002e-05,
+      "loss": 0.3603,
+      "step": 1340
+    },
+    {
+      "epoch": 1.5305191100969766,
+      "grad_norm": 0.41995732245076695,
+      "learning_rate": 2.72093023255814e-05,
+      "loss": 0.3655,
+      "step": 1341
+    },
+    {
+      "epoch": 1.531660011409013,
+      "grad_norm": 0.30788330372687395,
+      "learning_rate": 2.718816067653277e-05,
+      "loss": 0.353,
+      "step": 1342
+    },
+    {
+      "epoch": 1.5328009127210498,
+      "grad_norm": 0.4323021854865316,
+      "learning_rate": 2.7167019027484147e-05,
+      "loss": 0.3737,
+      "step": 1343
+    },
+    {
+      "epoch": 1.5339418140330863,
+      "grad_norm": 0.2932504556713375,
+      "learning_rate": 2.7145877378435518e-05,
+      "loss": 0.3725,
+      "step": 1344
+    },
+    {
+      "epoch": 1.5350827153451228,
+      "grad_norm": 0.38646332196884464,
+      "learning_rate": 2.712473572938689e-05,
+      "loss": 0.3836,
+      "step": 1345
+    },
+    {
+      "epoch": 1.5362236166571592,
+      "grad_norm": 0.3318875060177627,
+      "learning_rate": 2.710359408033827e-05,
+      "loss": 0.3587,
+      "step": 1346
+    },
+    {
+      "epoch": 1.5373645179691957,
+      "grad_norm": 0.39425050106556736,
+      "learning_rate": 2.708245243128964e-05,
+      "loss": 0.356,
+      "step": 1347
+    },
+    {
+      "epoch": 1.5385054192812322,
+      "grad_norm": 0.3312786798058442,
+      "learning_rate": 2.7061310782241017e-05,
+      "loss": 0.3819,
+      "step": 1348
+    },
+    {
+      "epoch": 1.5396463205932687,
+      "grad_norm": 0.3751189043301931,
+      "learning_rate": 2.7040169133192388e-05,
+      "loss": 0.358,
+      "step": 1349
+    },
+    {
+      "epoch": 1.5407872219053051,
+      "grad_norm": 0.45866033683350615,
+      "learning_rate": 2.7019027484143765e-05,
+      "loss": 0.3623,
+      "step": 1350
+    },
+    {
+      "epoch": 1.5419281232173416,
+      "grad_norm": 0.293269475978877,
+      "learning_rate": 2.699788583509514e-05,
+      "loss": 0.3675,
+      "step": 1351
+    },
+    {
+      "epoch": 1.543069024529378,
+      "grad_norm": 0.5264603659706322,
+      "learning_rate": 2.6976744186046517e-05,
+      "loss": 0.3801,
+      "step": 1352
+    },
+    {
+      "epoch": 1.5442099258414146,
+      "grad_norm": 0.3169556373264047,
+      "learning_rate": 2.6955602536997888e-05,
+      "loss": 0.3632,
+      "step": 1353
+    },
+    {
+      "epoch": 1.545350827153451,
+      "grad_norm": 0.49559534834221575,
+      "learning_rate": 2.6934460887949258e-05,
+      "loss": 0.3695,
+      "step": 1354
+    },
+    {
+      "epoch": 1.5464917284654878,
+      "grad_norm": 0.2853680856246618,
+      "learning_rate": 2.6913319238900636e-05,
+      "loss": 0.3505,
+      "step": 1355
+    },
+    {
+      "epoch": 1.5476326297775242,
+      "grad_norm": 0.35237363707507235,
+      "learning_rate": 2.6892177589852006e-05,
+      "loss": 0.363,
+      "step": 1356
+    },
+    {
+      "epoch": 1.5487735310895607,
+      "grad_norm": 0.34812240762886904,
+      "learning_rate": 2.6871035940803384e-05,
+      "loss": 0.356,
+      "step": 1357
+    },
+    {
+      "epoch": 1.5499144324015972,
+      "grad_norm": 0.3249639412396184,
+      "learning_rate": 2.6849894291754758e-05,
+      "loss": 0.3676,
+      "step": 1358
+    },
+    {
+      "epoch": 1.551055333713634,
+      "grad_norm": 0.34991069473657654,
+      "learning_rate": 2.6828752642706135e-05,
+      "loss": 0.359,
+      "step": 1359
+    },
+    {
+      "epoch": 1.5521962350256704,
+      "grad_norm": 0.29576634258933143,
+      "learning_rate": 2.6807610993657506e-05,
+      "loss": 0.3671,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5533371363377069,
+      "grad_norm": 0.34606104122898584,
+      "learning_rate": 2.6786469344608884e-05,
+      "loss": 0.371,
+      "step": 1361
+    },
+    {
+      "epoch": 1.5544780376497433,
+      "grad_norm": 0.2775582193569763,
+      "learning_rate": 2.6765327695560254e-05,
+      "loss": 0.3629,
+      "step": 1362
+    },
+    {
+      "epoch": 1.5556189389617798,
+      "grad_norm": 0.3247050074902207,
+      "learning_rate": 2.674418604651163e-05,
+      "loss": 0.3842,
+      "step": 1363
+    },
+    {
+      "epoch": 1.5567598402738163,
+      "grad_norm": 0.2723391641667623,
+      "learning_rate": 2.6723044397463002e-05,
+      "loss": 0.3604,
+      "step": 1364
+    },
+    {
+      "epoch": 1.5579007415858528,
+      "grad_norm": 0.36316336391236725,
+      "learning_rate": 2.6701902748414376e-05,
+      "loss": 0.3947,
+      "step": 1365
+    },
+    {
+      "epoch": 1.5590416428978893,
+      "grad_norm": 0.275443180804818,
+      "learning_rate": 2.6680761099365754e-05,
+      "loss": 0.3961,
+      "step": 1366
+    },
+    {
+      "epoch": 1.5601825442099257,
+      "grad_norm": 0.35579725890937663,
+      "learning_rate": 2.6659619450317125e-05,
+      "loss": 0.3846,
+      "step": 1367
+    },
+    {
+      "epoch": 1.5613234455219622,
+      "grad_norm": 0.26830941897973276,
+      "learning_rate": 2.6638477801268502e-05,
+      "loss": 0.3835,
+      "step": 1368
+    },
+    {
+      "epoch": 1.5624643468339987,
+      "grad_norm": 0.30243736502540497,
+      "learning_rate": 2.6617336152219873e-05,
+      "loss": 0.3481,
+      "step": 1369
+    },
+    {
+      "epoch": 1.5636052481460354,
+      "grad_norm": 0.31682892993854983,
+      "learning_rate": 2.659619450317125e-05,
+      "loss": 0.3684,
+      "step": 1370
+    },
+    {
+      "epoch": 1.5647461494580719,
+      "grad_norm": 0.27334699163010084,
+      "learning_rate": 2.6575052854122624e-05,
+      "loss": 0.358,
+      "step": 1371
+    },
+    {
+      "epoch": 1.5658870507701084,
+      "grad_norm": 0.2895943256047113,
+      "learning_rate": 2.6553911205073995e-05,
+      "loss": 0.3601,
+      "step": 1372
+    },
+    {
+      "epoch": 1.5670279520821448,
+      "grad_norm": 0.28755144042185043,
+      "learning_rate": 2.6532769556025372e-05,
+      "loss": 0.3717,
+      "step": 1373
+    },
+    {
+      "epoch": 1.5681688533941815,
+      "grad_norm": 0.25968518258420775,
+      "learning_rate": 2.6511627906976743e-05,
+      "loss": 0.3626,
+      "step": 1374
+    },
+    {
+      "epoch": 1.569309754706218,
+      "grad_norm": 0.32179811901000466,
+      "learning_rate": 2.649048625792812e-05,
+      "loss": 0.3673,
+      "step": 1375
+    },
+    {
+      "epoch": 1.5704506560182545,
+      "grad_norm": 0.2766533809177961,
+      "learning_rate": 2.646934460887949e-05,
+      "loss": 0.3691,
+      "step": 1376
+    },
+    {
+      "epoch": 1.571591557330291,
+      "grad_norm": 0.302638177000013,
+      "learning_rate": 2.644820295983087e-05,
+      "loss": 0.3672,
+      "step": 1377
+    },
+    {
+      "epoch": 1.5727324586423275,
+      "grad_norm": 0.296654847118796,
+      "learning_rate": 2.6427061310782243e-05,
+      "loss": 0.3812,
+      "step": 1378
+    },
+    {
+      "epoch": 1.573873359954364,
+      "grad_norm": 0.33824246630089,
+      "learning_rate": 2.640591966173362e-05,
+      "loss": 0.3756,
+      "step": 1379
+    },
+    {
+      "epoch": 1.5750142612664004,
+      "grad_norm": 0.3351383592040746,
+      "learning_rate": 2.638477801268499e-05,
+      "loss": 0.3766,
+      "step": 1380
+    },
+    {
+      "epoch": 1.576155162578437,
+      "grad_norm": 0.3024600320447683,
+      "learning_rate": 2.636363636363636e-05,
+      "loss": 0.3662,
+      "step": 1381
+    },
+    {
+      "epoch": 1.5772960638904734,
+      "grad_norm": 0.32753567856330584,
+      "learning_rate": 2.634249471458774e-05,
+      "loss": 0.3565,
+      "step": 1382
+    },
+    {
+      "epoch": 1.5784369652025099,
+      "grad_norm": 0.345187119181163,
+      "learning_rate": 2.6321353065539113e-05,
+      "loss": 0.3799,
+      "step": 1383
+    },
+    {
+      "epoch": 1.5795778665145463,
+      "grad_norm": 0.3331299603253503,
+      "learning_rate": 2.6300211416490487e-05,
+      "loss": 0.3987,
+      "step": 1384
+    },
+    {
+      "epoch": 1.580718767826583,
+      "grad_norm": 0.28725692061029895,
+      "learning_rate": 2.627906976744186e-05,
+      "loss": 0.3383,
+      "step": 1385
+    },
+    {
+      "epoch": 1.5818596691386195,
+      "grad_norm": 0.3111425790157666,
+      "learning_rate": 2.625792811839324e-05,
+      "loss": 0.3617,
+      "step": 1386
+    },
+    {
+      "epoch": 1.583000570450656,
+      "grad_norm": 0.299296343009948,
+      "learning_rate": 2.623678646934461e-05,
+      "loss": 0.3644,
+      "step": 1387
+    },
+    {
+      "epoch": 1.5841414717626925,
+      "grad_norm": 0.3347632168210099,
+      "learning_rate": 2.6215644820295987e-05,
+      "loss": 0.3723,
+      "step": 1388
+    },
+    {
+      "epoch": 1.5852823730747292,
+      "grad_norm": 0.31376561832951744,
+      "learning_rate": 2.6194503171247358e-05,
+      "loss": 0.3658,
+      "step": 1389
+    },
+    {
+      "epoch": 1.5864232743867657,
+      "grad_norm": 0.31027997164647236,
+      "learning_rate": 2.617336152219873e-05,
+      "loss": 0.3743,
+      "step": 1390
+    },
+    {
+      "epoch": 1.5875641756988021,
+      "grad_norm": 0.3526552435594915,
+      "learning_rate": 2.615221987315011e-05,
+      "loss": 0.3599,
+      "step": 1391
+    },
+    {
+      "epoch": 1.5887050770108386,
+      "grad_norm": 0.29310310962923514,
+      "learning_rate": 2.613107822410148e-05,
+      "loss": 0.3459,
+      "step": 1392
+    },
+    {
+      "epoch": 1.589845978322875,
+      "grad_norm": 0.300313890699162,
+      "learning_rate": 2.6109936575052857e-05,
+      "loss": 0.3638,
+      "step": 1393
+    },
+    {
+      "epoch": 1.5909868796349116,
+      "grad_norm": 0.35597737339719526,
+      "learning_rate": 2.6088794926004228e-05,
+      "loss": 0.3629,
+      "step": 1394
+    },
+    {
+      "epoch": 1.592127780946948,
+      "grad_norm": 0.29485050155881004,
+      "learning_rate": 2.6067653276955605e-05,
+      "loss": 0.3495,
+      "step": 1395
+    },
+    {
+      "epoch": 1.5932686822589845,
+      "grad_norm": 0.3083586629060851,
+      "learning_rate": 2.6046511627906976e-05,
+      "loss": 0.3682,
+      "step": 1396
+    },
+    {
+      "epoch": 1.594409583571021,
+      "grad_norm": 0.3087204856725598,
+      "learning_rate": 2.6025369978858354e-05,
+      "loss": 0.3516,
+      "step": 1397
+    },
+    {
+      "epoch": 1.5955504848830575,
+      "grad_norm": 0.28096277736373027,
+      "learning_rate": 2.6004228329809728e-05,
+      "loss": 0.365,
+      "step": 1398
+    },
+    {
+      "epoch": 1.596691386195094,
+      "grad_norm": 0.30757276234849873,
+      "learning_rate": 2.59830866807611e-05,
+      "loss": 0.3732,
+      "step": 1399
+    },
+    {
+      "epoch": 1.5978322875071307,
+      "grad_norm": 0.3333668786223208,
+      "learning_rate": 2.5961945031712476e-05,
+      "loss": 0.375,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5989731888191672,
+      "grad_norm": 0.2651214513170331,
+      "learning_rate": 2.5940803382663846e-05,
+      "loss": 0.3599,
+      "step": 1401
+    },
+    {
+      "epoch": 1.6001140901312036,
+      "grad_norm": 0.2973829631907778,
+      "learning_rate": 2.5919661733615224e-05,
+      "loss": 0.3605,
+      "step": 1402
+    },
+    {
+      "epoch": 1.6012549914432401,
+      "grad_norm": 0.2831085318391084,
+      "learning_rate": 2.5898520084566598e-05,
+      "loss": 0.3491,
+      "step": 1403
+    },
+    {
+      "epoch": 1.6023958927552768,
+      "grad_norm": 0.32151978953955745,
+      "learning_rate": 2.5877378435517975e-05,
+      "loss": 0.3752,
+      "step": 1404
+    },
+    {
+      "epoch": 1.6035367940673133,
+      "grad_norm": 0.2821353356916035,
+      "learning_rate": 2.5856236786469346e-05,
+      "loss": 0.3588,
+      "step": 1405
+    },
+    {
+      "epoch": 1.6046776953793498,
+      "grad_norm": 0.3125896600678469,
+      "learning_rate": 2.5835095137420717e-05,
+      "loss": 0.3708,
+      "step": 1406
+    },
+    {
+      "epoch": 1.6058185966913863,
+      "grad_norm": 0.29262459892442727,
+      "learning_rate": 2.5813953488372094e-05,
+      "loss": 0.3893,
+      "step": 1407
+    },
+    {
+      "epoch": 1.6069594980034227,
+      "grad_norm": 0.3339011655139542,
+      "learning_rate": 2.5792811839323465e-05,
+      "loss": 0.3824,
+      "step": 1408
+    },
+    {
+      "epoch": 1.6081003993154592,
+      "grad_norm": 0.3015512370604447,
+      "learning_rate": 2.5771670190274842e-05,
+      "loss": 0.361,
+      "step": 1409
+    },
+    {
+      "epoch": 1.6092413006274957,
+      "grad_norm": 0.2682503512944308,
+      "learning_rate": 2.5750528541226217e-05,
+      "loss": 0.3606,
+      "step": 1410
+    },
+    {
+      "epoch": 1.6103822019395322,
+      "grad_norm": 0.2939224257467947,
+      "learning_rate": 2.5729386892177594e-05,
+      "loss": 0.3706,
+      "step": 1411
+    },
+    {
+      "epoch": 1.6115231032515687,
+      "grad_norm": 0.29144618672695827,
+      "learning_rate": 2.5708245243128965e-05,
+      "loss": 0.364,
+      "step": 1412
+    },
+    {
+      "epoch": 1.6126640045636051,
+      "grad_norm": 0.2905815562354008,
+      "learning_rate": 2.5687103594080342e-05,
+      "loss": 0.3565,
+      "step": 1413
+    },
+    {
+      "epoch": 1.6138049058756416,
+      "grad_norm": 0.30246721937468807,
+      "learning_rate": 2.5665961945031713e-05,
+      "loss": 0.3718,
+      "step": 1414
+    },
+    {
+      "epoch": 1.614945807187678,
+      "grad_norm": 0.2714015002455204,
+      "learning_rate": 2.5644820295983087e-05,
+      "loss": 0.358,
+      "step": 1415
+    },
+    {
+      "epoch": 1.6160867084997148,
+      "grad_norm": 0.29518888476520655,
+      "learning_rate": 2.562367864693446e-05,
+      "loss": 0.3659,
+      "step": 1416
+    },
+    {
+      "epoch": 1.6172276098117513,
+      "grad_norm": 0.32113379666355385,
+      "learning_rate": 2.5602536997885835e-05,
+      "loss": 0.3523,
+      "step": 1417
+    },
+    {
+      "epoch": 1.6183685111237878,
+      "grad_norm": 0.3084244799881773,
+      "learning_rate": 2.5581395348837212e-05,
+      "loss": 0.3632,
+      "step": 1418
+    },
+    {
+      "epoch": 1.6195094124358242,
+      "grad_norm": 0.3112848716857277,
+      "learning_rate": 2.5560253699788583e-05,
+      "loss": 0.3604,
+      "step": 1419
+    },
+    {
+      "epoch": 1.620650313747861,
+      "grad_norm": 0.297659167096412,
+      "learning_rate": 2.553911205073996e-05,
+      "loss": 0.372,
+      "step": 1420
+    },
+    {
+      "epoch": 1.6217912150598974,
+      "grad_norm": 0.38804953050492647,
+      "learning_rate": 2.551797040169133e-05,
+      "loss": 0.3778,
+      "step": 1421
+    },
+    {
+      "epoch": 1.622932116371934,
+      "grad_norm": 0.3422019172941039,
+      "learning_rate": 2.549682875264271e-05,
+      "loss": 0.378,
+      "step": 1422
+    },
+    {
+      "epoch": 1.6240730176839704,
+      "grad_norm": 0.291456523422963,
+      "learning_rate": 2.5475687103594083e-05,
+      "loss": 0.3592,
+      "step": 1423
+    },
+    {
+      "epoch": 1.6252139189960069,
+      "grad_norm": 0.3890506166879392,
+      "learning_rate": 2.5454545454545454e-05,
+      "loss": 0.3775,
+      "step": 1424
+    },
+    {
+      "epoch": 1.6263548203080433,
+      "grad_norm": 0.32715486375630815,
+      "learning_rate": 2.543340380549683e-05,
+      "loss": 0.3721,
+      "step": 1425
+    },
+    {
+      "epoch": 1.6274957216200798,
+      "grad_norm": 0.3095339401874632,
+      "learning_rate": 2.54122621564482e-05,
+      "loss": 0.3543,
+      "step": 1426
+    },
+    {
+      "epoch": 1.6286366229321163,
+      "grad_norm": 0.2968567670795314,
+      "learning_rate": 2.539112050739958e-05,
+      "loss": 0.368,
+      "step": 1427
+    },
+    {
+      "epoch": 1.6297775242441528,
+      "grad_norm": 0.3009801689582668,
+      "learning_rate": 2.536997885835095e-05,
+      "loss": 0.3443,
+      "step": 1428
+    },
+    {
+      "epoch": 1.6309184255561893,
+      "grad_norm": 0.25882579033711073,
+      "learning_rate": 2.5348837209302327e-05,
+      "loss": 0.3706,
+      "step": 1429
+    },
+    {
+      "epoch": 1.6320593268682257,
+      "grad_norm": 0.27518164377492776,
+      "learning_rate": 2.53276955602537e-05,
+      "loss": 0.3717,
+      "step": 1430
+    },
+    {
+      "epoch": 1.6332002281802624,
+      "grad_norm": 0.32265443241190234,
+      "learning_rate": 2.530655391120508e-05,
+      "loss": 0.3933,
+      "step": 1431
+    },
+    {
+      "epoch": 1.634341129492299,
+      "grad_norm": 0.30698302763604135,
+      "learning_rate": 2.528541226215645e-05,
+      "loss": 0.362,
+      "step": 1432
+    },
+    {
+      "epoch": 1.6354820308043354,
+      "grad_norm": 0.2715331709115899,
+      "learning_rate": 2.526427061310782e-05,
+      "loss": 0.3749,
+      "step": 1433
+    },
+    {
+      "epoch": 1.6366229321163719,
+      "grad_norm": 0.3091602334662984,
+      "learning_rate": 2.5243128964059198e-05,
+      "loss": 0.3524,
+      "step": 1434
+    },
+    {
+      "epoch": 1.6377638334284086,
+      "grad_norm": 0.34223110608310403,
+      "learning_rate": 2.5221987315010572e-05,
+      "loss": 0.3891,
+      "step": 1435
+    },
+    {
+      "epoch": 1.638904734740445,
+      "grad_norm": 0.28760443573042516,
+      "learning_rate": 2.5200845665961946e-05,
+      "loss": 0.3681,
+      "step": 1436
+    },
+    {
+      "epoch": 1.6400456360524815,
+      "grad_norm": 0.30154305625963906,
+      "learning_rate": 2.517970401691332e-05,
+      "loss": 0.3641,
+      "step": 1437
+    },
+    {
+      "epoch": 1.641186537364518,
+      "grad_norm": 0.3720716239663011,
+      "learning_rate": 2.5158562367864697e-05,
+      "loss": 0.3991,
+      "step": 1438
+    },
+    {
+      "epoch": 1.6423274386765545,
+      "grad_norm": 0.28530367783404104,
+      "learning_rate": 2.5137420718816068e-05,
+      "loss": 0.3808,
+      "step": 1439
+    },
+    {
+      "epoch": 1.643468339988591,
+      "grad_norm": 0.27013626954015546,
+      "learning_rate": 2.5116279069767445e-05,
+      "loss": 0.3688,
+      "step": 1440
+    },
+    {
+      "epoch": 1.6446092413006275,
+      "grad_norm": 0.31682127904243484,
+      "learning_rate": 2.5095137420718816e-05,
+      "loss": 0.3541,
+      "step": 1441
+    },
+    {
+      "epoch": 1.645750142612664,
+      "grad_norm": 0.2841931208482495,
+      "learning_rate": 2.507399577167019e-05,
+      "loss": 0.3602,
+      "step": 1442
+    },
+    {
+      "epoch": 1.6468910439247004,
+      "grad_norm": 0.276074554952914,
+      "learning_rate": 2.5052854122621568e-05,
+      "loss": 0.3632,
+      "step": 1443
+    },
+    {
+      "epoch": 1.648031945236737,
+      "grad_norm": 0.3045129394868517,
+      "learning_rate": 2.503171247357294e-05,
+      "loss": 0.3612,
+      "step": 1444
+    },
+    {
+      "epoch": 1.6491728465487734,
+      "grad_norm": 0.2711753572169419,
+      "learning_rate": 2.5010570824524316e-05,
+      "loss": 0.3695,
+      "step": 1445
+    },
+    {
+      "epoch": 1.65031374786081,
+      "grad_norm": 0.31184353074715726,
+      "learning_rate": 2.4989429175475687e-05,
+      "loss": 0.3577,
+      "step": 1446
+    },
+    {
+      "epoch": 1.6514546491728466,
+      "grad_norm": 0.2736196705378409,
+      "learning_rate": 2.496828752642706e-05,
+      "loss": 0.3538,
+      "step": 1447
+    },
+    {
+      "epoch": 1.652595550484883,
+      "grad_norm": 0.2829200724593569,
+      "learning_rate": 2.4947145877378435e-05,
+      "loss": 0.3703,
+      "step": 1448
+    },
+    {
+      "epoch": 1.6537364517969195,
+      "grad_norm": 0.296167668333878,
+      "learning_rate": 2.4926004228329812e-05,
+      "loss": 0.3682,
+      "step": 1449
+    },
+    {
+      "epoch": 1.6548773531089562,
+      "grad_norm": 0.2624154269631874,
+      "learning_rate": 2.4904862579281186e-05,
+      "loss": 0.3637,
+      "step": 1450
+    },
+    {
+      "epoch": 1.6560182544209927,
+      "grad_norm": 0.29443772973012555,
+      "learning_rate": 2.488372093023256e-05,
+      "loss": 0.3474,
+      "step": 1451
+    },
+    {
+      "epoch": 1.6571591557330292,
+      "grad_norm": 0.308700169528517,
+      "learning_rate": 2.4862579281183934e-05,
+      "loss": 0.3827,
+      "step": 1452
+    },
+    {
+      "epoch": 1.6583000570450657,
+      "grad_norm": 0.29493785751514123,
+      "learning_rate": 2.484143763213531e-05,
+      "loss": 0.3421,
+      "step": 1453
+    },
+    {
+      "epoch": 1.6594409583571021,
+      "grad_norm": 0.2713838260766285,
+      "learning_rate": 2.4820295983086683e-05,
+      "loss": 0.351,
+      "step": 1454
+    },
+    {
+      "epoch": 1.6605818596691386,
+      "grad_norm": 0.28390750706850665,
+      "learning_rate": 2.4799154334038057e-05,
+      "loss": 0.3493,
+      "step": 1455
+    },
+    {
+      "epoch": 1.661722760981175,
+      "grad_norm": 0.2990257269654475,
+      "learning_rate": 2.477801268498943e-05,
+      "loss": 0.3612,
+      "step": 1456
+    },
+    {
+      "epoch": 1.6628636622932116,
+      "grad_norm": 0.2665546960167287,
+      "learning_rate": 2.4756871035940805e-05,
+      "loss": 0.3576,
+      "step": 1457
+    },
+    {
+      "epoch": 1.664004563605248,
+      "grad_norm": 0.30439565309196187,
+      "learning_rate": 2.473572938689218e-05,
+      "loss": 0.3576,
+      "step": 1458
+    },
+    {
+      "epoch": 1.6651454649172845,
+      "grad_norm": 0.33814380783466985,
+      "learning_rate": 2.4714587737843553e-05,
+      "loss": 0.384,
+      "step": 1459
+    },
+    {
+      "epoch": 1.666286366229321,
+      "grad_norm": 0.30098003633111947,
+      "learning_rate": 2.4693446088794927e-05,
+      "loss": 0.3613,
+      "step": 1460
+    },
+    {
+      "epoch": 1.6674272675413577,
+      "grad_norm": 0.294974905406365,
+      "learning_rate": 2.46723044397463e-05,
+      "loss": 0.3707,
+      "step": 1461
+    },
+    {
+      "epoch": 1.6685681688533942,
+      "grad_norm": 0.3708465734189445,
+      "learning_rate": 2.4651162790697675e-05,
+      "loss": 0.3733,
+      "step": 1462
+    },
+    {
+      "epoch": 1.6697090701654307,
+      "grad_norm": 0.3843774740323064,
+      "learning_rate": 2.4630021141649053e-05,
+      "loss": 0.3677,
+      "step": 1463
+    },
+    {
+      "epoch": 1.6708499714774672,
+      "grad_norm": 0.2862491424432163,
+      "learning_rate": 2.4608879492600423e-05,
+      "loss": 0.3644,
+      "step": 1464
+    },
+    {
+      "epoch": 1.6719908727895039,
+      "grad_norm": 0.3820186373619291,
+      "learning_rate": 2.4587737843551797e-05,
+      "loss": 0.3683,
+      "step": 1465
+    },
+    {
+      "epoch": 1.6731317741015403,
+      "grad_norm": 0.3605477948643615,
+      "learning_rate": 2.456659619450317e-05,
+      "loss": 0.3643,
+      "step": 1466
+    },
+    {
+      "epoch": 1.6742726754135768,
+      "grad_norm": 0.3745644891120593,
+      "learning_rate": 2.4545454545454545e-05,
+      "loss": 0.3963,
+      "step": 1467
+    },
+    {
+      "epoch": 1.6754135767256133,
+      "grad_norm": 0.2843263524287354,
+      "learning_rate": 2.452431289640592e-05,
+      "loss": 0.3573,
+      "step": 1468
+    },
+    {
+      "epoch": 1.6765544780376498,
+      "grad_norm": 0.3095584589918876,
+      "learning_rate": 2.4503171247357297e-05,
+      "loss": 0.3672,
+      "step": 1469
+    },
+    {
+      "epoch": 1.6776953793496863,
+      "grad_norm": 0.33775878021951095,
+      "learning_rate": 2.448202959830867e-05,
+      "loss": 0.3644,
+      "step": 1470
+    },
+    {
+      "epoch": 1.6788362806617227,
+      "grad_norm": 0.2798731836487535,
+      "learning_rate": 2.4460887949260045e-05,
+      "loss": 0.3904,
+      "step": 1471
+    },
+    {
+      "epoch": 1.6799771819737592,
+      "grad_norm": 0.3176793770462291,
+      "learning_rate": 2.4439746300211416e-05,
+      "loss": 0.3503,
+      "step": 1472
+    },
+    {
+      "epoch": 1.6811180832857957,
+      "grad_norm": 0.2817311269242639,
+      "learning_rate": 2.441860465116279e-05,
+      "loss": 0.3507,
+      "step": 1473
+    },
+    {
+      "epoch": 1.6822589845978322,
+      "grad_norm": 0.332496204883691,
+      "learning_rate": 2.4397463002114164e-05,
+      "loss": 0.3701,
+      "step": 1474
+    },
+    {
+      "epoch": 1.6833998859098687,
+      "grad_norm": 0.31023078517004654,
+      "learning_rate": 2.437632135306554e-05,
+      "loss": 0.3793,
+      "step": 1475
+    },
+    {
+      "epoch": 1.6845407872219051,
+      "grad_norm": 0.3074921350102049,
+      "learning_rate": 2.4355179704016916e-05,
+      "loss": 0.3587,
+      "step": 1476
+    },
+    {
+      "epoch": 1.6856816885339418,
+      "grad_norm": 0.31490984904000224,
+      "learning_rate": 2.433403805496829e-05,
+      "loss": 0.366,
+      "step": 1477
+    },
+    {
+      "epoch": 1.6868225898459783,
+      "grad_norm": 0.3316060402389124,
+      "learning_rate": 2.4312896405919664e-05,
+      "loss": 0.3741,
+      "step": 1478
+    },
+    {
+      "epoch": 1.6879634911580148,
+      "grad_norm": 0.28937031810952313,
+      "learning_rate": 2.4291754756871038e-05,
+      "loss": 0.3583,
+      "step": 1479
+    },
+    {
+      "epoch": 1.6891043924700513,
+      "grad_norm": 0.29230363513203816,
+      "learning_rate": 2.4270613107822412e-05,
+      "loss": 0.3444,
+      "step": 1480
+    },
+    {
+      "epoch": 1.690245293782088,
+      "grad_norm": 0.3506573605117981,
+      "learning_rate": 2.4249471458773786e-05,
+      "loss": 0.3626,
+      "step": 1481
+    },
+    {
+      "epoch": 1.6913861950941245,
+      "grad_norm": 0.26141179784062163,
+      "learning_rate": 2.422832980972516e-05,
+      "loss": 0.3528,
+      "step": 1482
+    },
+    {
+      "epoch": 1.692527096406161,
+      "grad_norm": 0.3330228413197761,
+      "learning_rate": 2.4207188160676534e-05,
+      "loss": 0.379,
+      "step": 1483
+    },
+    {
+      "epoch": 1.6936679977181974,
+      "grad_norm": 0.29655254496962724,
+      "learning_rate": 2.4186046511627908e-05,
+      "loss": 0.3869,
+      "step": 1484
+    },
+    {
+      "epoch": 1.694808899030234,
+      "grad_norm": 0.2732954777447921,
+      "learning_rate": 2.4164904862579282e-05,
+      "loss": 0.3745,
+      "step": 1485
+    },
+    {
+      "epoch": 1.6959498003422704,
+      "grad_norm": 0.325475009094081,
+      "learning_rate": 2.4143763213530656e-05,
+      "loss": 0.3748,
+      "step": 1486
+    },
+    {
+      "epoch": 1.6970907016543069,
+      "grad_norm": 0.2993471282332693,
+      "learning_rate": 2.412262156448203e-05,
+      "loss": 0.3668,
+      "step": 1487
+    },
+    {
+      "epoch": 1.6982316029663433,
+      "grad_norm": 0.31603236693685227,
+      "learning_rate": 2.4101479915433404e-05,
+      "loss": 0.3727,
+      "step": 1488
+    },
+    {
+      "epoch": 1.6993725042783798,
+      "grad_norm": 0.2817814174465588,
+      "learning_rate": 2.4080338266384782e-05,
+      "loss": 0.3554,
+      "step": 1489
+    },
+    {
+      "epoch": 1.7005134055904163,
+      "grad_norm": 0.32120059999814593,
+      "learning_rate": 2.4059196617336153e-05,
+      "loss": 0.3659,
+      "step": 1490
+    },
+    {
+      "epoch": 1.7016543069024528,
+      "grad_norm": 0.28924927945350043,
+      "learning_rate": 2.4038054968287527e-05,
+      "loss": 0.3731,
+      "step": 1491
+    },
+    {
+      "epoch": 1.7027952082144895,
+      "grad_norm": 0.284643173875121,
+      "learning_rate": 2.40169133192389e-05,
+      "loss": 0.3884,
+      "step": 1492
+    },
+    {
+      "epoch": 1.703936109526526,
+      "grad_norm": 0.2665022764695233,
+      "learning_rate": 2.3995771670190275e-05,
+      "loss": 0.3547,
+      "step": 1493
+    },
+    {
+      "epoch": 1.7050770108385624,
+      "grad_norm": 0.33643098630695556,
+      "learning_rate": 2.397463002114165e-05,
+      "loss": 0.3753,
+      "step": 1494
+    },
+    {
+      "epoch": 1.706217912150599,
+      "grad_norm": 0.25812980136170705,
+      "learning_rate": 2.3953488372093026e-05,
+      "loss": 0.3671,
+      "step": 1495
+    },
+    {
+      "epoch": 1.7073588134626356,
+      "grad_norm": 0.2876613549689138,
+      "learning_rate": 2.39323467230444e-05,
+      "loss": 0.3714,
+      "step": 1496
+    },
+    {
+      "epoch": 1.708499714774672,
+      "grad_norm": 0.31197661099028084,
+      "learning_rate": 2.3911205073995774e-05,
+      "loss": 0.3718,
+      "step": 1497
+    },
+    {
+      "epoch": 1.7096406160867086,
+      "grad_norm": 0.29188522368743014,
+      "learning_rate": 2.389006342494715e-05,
+      "loss": 0.3757,
+      "step": 1498
+    },
+    {
+      "epoch": 1.710781517398745,
+      "grad_norm": 0.2592921193768068,
+      "learning_rate": 2.386892177589852e-05,
+      "loss": 0.361,
+      "step": 1499
+    },
+    {
+      "epoch": 1.7119224187107815,
+      "grad_norm": 0.28717400651849345,
+      "learning_rate": 2.3847780126849893e-05,
+      "loss": 0.3923,
+      "step": 1500
+    },
+    {
+      "epoch": 1.713063320022818,
+      "grad_norm": 0.2837339672529112,
+      "learning_rate": 2.382663847780127e-05,
+      "loss": 0.3788,
+      "step": 1501
+    },
+    {
+      "epoch": 1.7142042213348545,
+      "grad_norm": 0.25399925608301915,
+      "learning_rate": 2.3805496828752645e-05,
+      "loss": 0.3634,
+      "step": 1502
+    },
+    {
+      "epoch": 1.715345122646891,
+      "grad_norm": 0.2505460201463145,
+      "learning_rate": 2.378435517970402e-05,
+      "loss": 0.3614,
+      "step": 1503
+    },
+    {
+      "epoch": 1.7164860239589275,
+      "grad_norm": 0.25947337418651406,
+      "learning_rate": 2.3763213530655393e-05,
+      "loss": 0.3512,
+      "step": 1504
+    },
+    {
+      "epoch": 1.717626925270964,
+      "grad_norm": 0.30519573039102393,
+      "learning_rate": 2.3742071881606767e-05,
+      "loss": 0.3713,
+      "step": 1505
+    },
+    {
+      "epoch": 1.7187678265830004,
+      "grad_norm": 0.23254546868546908,
+      "learning_rate": 2.372093023255814e-05,
+      "loss": 0.3413,
+      "step": 1506
+    },
+    {
+      "epoch": 1.7199087278950371,
+      "grad_norm": 0.36817083989520405,
+      "learning_rate": 2.3699788583509515e-05,
+      "loss": 0.3749,
+      "step": 1507
+    },
+    {
+      "epoch": 1.7210496292070736,
+      "grad_norm": 0.3186865756330951,
+      "learning_rate": 2.367864693446089e-05,
+      "loss": 0.3495,
+      "step": 1508
+    },
+    {
+      "epoch": 1.72219053051911,
+      "grad_norm": 0.2808903990515687,
+      "learning_rate": 2.3657505285412263e-05,
+      "loss": 0.3558,
+      "step": 1509
+    },
+    {
+      "epoch": 1.7233314318311466,
+      "grad_norm": 0.2972003145721708,
+      "learning_rate": 2.3636363636363637e-05,
+      "loss": 0.3684,
+      "step": 1510
+    },
+    {
+      "epoch": 1.7244723331431833,
+      "grad_norm": 0.32992144660872674,
+      "learning_rate": 2.361522198731501e-05,
+      "loss": 0.3901,
+      "step": 1511
+    },
+    {
+      "epoch": 1.7256132344552197,
+      "grad_norm": 0.304645651405577,
+      "learning_rate": 2.3594080338266386e-05,
+      "loss": 0.3643,
+      "step": 1512
+    },
+    {
+      "epoch": 1.7267541357672562,
+      "grad_norm": 0.40363172426881777,
+      "learning_rate": 2.357293868921776e-05,
+      "loss": 0.3692,
+      "step": 1513
+    },
+    {
+      "epoch": 1.7278950370792927,
+      "grad_norm": 0.2735019850601522,
+      "learning_rate": 2.3551797040169134e-05,
+      "loss": 0.3787,
+      "step": 1514
+    },
+    {
+      "epoch": 1.7290359383913292,
+      "grad_norm": 0.3442056801706256,
+      "learning_rate": 2.353065539112051e-05,
+      "loss": 0.3945,
+      "step": 1515
+    },
+    {
+      "epoch": 1.7301768397033657,
+      "grad_norm": 0.3955918249097106,
+      "learning_rate": 2.3509513742071882e-05,
+      "loss": 0.3833,
+      "step": 1516
+    },
+    {
+      "epoch": 1.7313177410154021,
+      "grad_norm": 0.31123018585787954,
+      "learning_rate": 2.3488372093023256e-05,
+      "loss": 0.3535,
+      "step": 1517
+    },
+    {
+      "epoch": 1.7324586423274386,
+      "grad_norm": 0.33153086678958815,
+      "learning_rate": 2.346723044397463e-05,
+      "loss": 0.3763,
+      "step": 1518
+    },
+    {
+      "epoch": 1.733599543639475,
+      "grad_norm": 0.3480404455336306,
+      "learning_rate": 2.3446088794926004e-05,
+      "loss": 0.3673,
+      "step": 1519
+    },
+    {
+      "epoch": 1.7347404449515116,
+      "grad_norm": 0.3149346024288447,
+      "learning_rate": 2.3424947145877378e-05,
+      "loss": 0.3595,
+      "step": 1520
+    },
+    {
+      "epoch": 1.735881346263548,
+      "grad_norm": 0.30056431949640394,
+      "learning_rate": 2.3403805496828756e-05,
+      "loss": 0.3656,
+      "step": 1521
+    },
+    {
+      "epoch": 1.7370222475755848,
+      "grad_norm": 0.3041892875716995,
+      "learning_rate": 2.338266384778013e-05,
+      "loss": 0.3599,
+      "step": 1522
+    },
+    {
+      "epoch": 1.7381631488876212,
+      "grad_norm": 0.33025855308799845,
+      "learning_rate": 2.3361522198731504e-05,
+      "loss": 0.3674,
+      "step": 1523
+    },
+    {
+      "epoch": 1.7393040501996577,
+      "grad_norm": 0.33179918502041716,
+      "learning_rate": 2.3340380549682878e-05,
+      "loss": 0.3786,
+      "step": 1524
+    },
+    {
+      "epoch": 1.7404449515116942,
+      "grad_norm": 0.2757882558131415,
+      "learning_rate": 2.331923890063425e-05,
+      "loss": 0.3621,
+      "step": 1525
+    },
+    {
+      "epoch": 1.741585852823731,
+      "grad_norm": 0.3240423708410654,
+      "learning_rate": 2.3298097251585623e-05,
+      "loss": 0.3502,
+      "step": 1526
+    },
+    {
+      "epoch": 1.7427267541357674,
+      "grad_norm": 0.3355411788126555,
+      "learning_rate": 2.3276955602537e-05,
+      "loss": 0.3823,
+      "step": 1527
+    },
+    {
+      "epoch": 1.7438676554478039,
+      "grad_norm": 0.29410966517590625,
+      "learning_rate": 2.3255813953488374e-05,
+      "loss": 0.3612,
+      "step": 1528
+    },
+    {
+      "epoch": 1.7450085567598403,
+      "grad_norm": 0.3129468551760911,
+      "learning_rate": 2.3234672304439748e-05,
+      "loss": 0.3811,
+      "step": 1529
+    },
+    {
+      "epoch": 1.7461494580718768,
+      "grad_norm": 0.37196565677390886,
+      "learning_rate": 2.3213530655391122e-05,
+      "loss": 0.3649,
+      "step": 1530
+    },
+    {
+      "epoch": 1.7472903593839133,
+      "grad_norm": 0.27482339581092213,
+      "learning_rate": 2.3192389006342496e-05,
+      "loss": 0.3702,
+      "step": 1531
+    },
+    {
+      "epoch": 1.7484312606959498,
+      "grad_norm": 0.32354120499237055,
+      "learning_rate": 2.317124735729387e-05,
+      "loss": 0.3696,
+      "step": 1532
+    },
+    {
+      "epoch": 1.7495721620079863,
+      "grad_norm": 0.344333434322942,
+      "learning_rate": 2.3150105708245244e-05,
+      "loss": 0.3775,
+      "step": 1533
+    },
+    {
+      "epoch": 1.7507130633200227,
+      "grad_norm": 0.2501858892242347,
+      "learning_rate": 2.312896405919662e-05,
+      "loss": 0.3673,
+      "step": 1534
+    },
+    {
+      "epoch": 1.7518539646320592,
+      "grad_norm": 0.2619318350318475,
+      "learning_rate": 2.3107822410147993e-05,
+      "loss": 0.371,
+      "step": 1535
+    },
+    {
+      "epoch": 1.7529948659440957,
+      "grad_norm": 0.28267590669718795,
+      "learning_rate": 2.3086680761099367e-05,
+      "loss": 0.346,
+      "step": 1536
+    },
+    {
+      "epoch": 1.7541357672561322,
+      "grad_norm": 0.2822911644558842,
+      "learning_rate": 2.306553911205074e-05,
+      "loss": 0.3611,
+      "step": 1537
+    },
+    {
+      "epoch": 1.7552766685681689,
+      "grad_norm": 0.24913100106359873,
+      "learning_rate": 2.3044397463002115e-05,
+      "loss": 0.3587,
+      "step": 1538
+    },
+    {
+      "epoch": 1.7564175698802054,
+      "grad_norm": 0.2782277920002323,
+      "learning_rate": 2.302325581395349e-05,
+      "loss": 0.3589,
+      "step": 1539
+    },
+    {
+      "epoch": 1.7575584711922418,
+      "grad_norm": 0.29310960832535715,
+      "learning_rate": 2.3002114164904863e-05,
+      "loss": 0.3534,
+      "step": 1540
+    },
+    {
+      "epoch": 1.7586993725042785,
+      "grad_norm": 0.26490888346701547,
+      "learning_rate": 2.298097251585624e-05,
+      "loss": 0.3625,
+      "step": 1541
+    },
+    {
+      "epoch": 1.759840273816315,
+      "grad_norm": 0.3124532199065016,
+      "learning_rate": 2.2959830866807614e-05,
+      "loss": 0.3533,
+      "step": 1542
+    },
+    {
+      "epoch": 1.7609811751283515,
+      "grad_norm": 0.28873529134829373,
+      "learning_rate": 2.2938689217758985e-05,
+      "loss": 0.3846,
+      "step": 1543
+    },
+    {
+      "epoch": 1.762122076440388,
+      "grad_norm": 0.2907772323127595,
+      "learning_rate": 2.291754756871036e-05,
+      "loss": 0.3279,
+      "step": 1544
+    },
+    {
+      "epoch": 1.7632629777524245,
+      "grad_norm": 0.29886052812500113,
+      "learning_rate": 2.2896405919661733e-05,
+      "loss": 0.3585,
+      "step": 1545
+    },
+    {
+      "epoch": 1.764403879064461,
+      "grad_norm": 0.29662624280244543,
+      "learning_rate": 2.2875264270613107e-05,
+      "loss": 0.3646,
+      "step": 1546
+    },
+    {
+      "epoch": 1.7655447803764974,
+      "grad_norm": 0.3141679404141358,
+      "learning_rate": 2.2854122621564485e-05,
+      "loss": 0.374,
+      "step": 1547
+    },
+    {
+      "epoch": 1.766685681688534,
+      "grad_norm": 0.2951349996054108,
+      "learning_rate": 2.283298097251586e-05,
+      "loss": 0.3668,
+      "step": 1548
+    },
+    {
+      "epoch": 1.7678265830005704,
+      "grad_norm": 0.29342310695343443,
+      "learning_rate": 2.2811839323467233e-05,
+      "loss": 0.3495,
+      "step": 1549
+    },
+    {
+      "epoch": 1.7689674843126069,
+      "grad_norm": 0.25105234676118393,
+      "learning_rate": 2.2790697674418607e-05,
+      "loss": 0.3439,
+      "step": 1550
+    },
+    {
+      "epoch": 1.7701083856246433,
+      "grad_norm": 0.28202846143657684,
+      "learning_rate": 2.276955602536998e-05,
+      "loss": 0.3622,
+      "step": 1551
+    },
+    {
+      "epoch": 1.7712492869366798,
+      "grad_norm": 0.2577324776105485,
+      "learning_rate": 2.2748414376321352e-05,
+      "loss": 0.3674,
+      "step": 1552
+    },
+    {
+      "epoch": 1.7723901882487165,
+      "grad_norm": 0.26427649352790966,
+      "learning_rate": 2.272727272727273e-05,
+      "loss": 0.3915,
+      "step": 1553
+    },
+    {
+      "epoch": 1.773531089560753,
+      "grad_norm": 0.2975077391703283,
+      "learning_rate": 2.2706131078224103e-05,
+      "loss": 0.3719,
+      "step": 1554
+    },
+    {
+      "epoch": 1.7746719908727895,
+      "grad_norm": 0.30371902180036175,
+      "learning_rate": 2.2684989429175477e-05,
+      "loss": 0.3625,
+      "step": 1555
+    },
+    {
+      "epoch": 1.775812892184826,
+      "grad_norm": 0.2978704874115724,
+      "learning_rate": 2.266384778012685e-05,
+      "loss": 0.3492,
+      "step": 1556
+    },
+    {
+      "epoch": 1.7769537934968627,
+      "grad_norm": 0.3424710072192493,
+      "learning_rate": 2.2642706131078226e-05,
+      "loss": 0.376,
+      "step": 1557
+    },
+    {
+      "epoch": 1.7780946948088991,
+      "grad_norm": 0.4304707541450273,
+      "learning_rate": 2.26215644820296e-05,
+      "loss": 0.3716,
+      "step": 1558
+    },
+    {
+      "epoch": 1.7792355961209356,
+      "grad_norm": 0.2888383768672417,
+      "learning_rate": 2.2600422832980974e-05,
+      "loss": 0.3657,
+      "step": 1559
+    },
+    {
+      "epoch": 1.780376497432972,
+      "grad_norm": 0.3424825882345746,
+      "learning_rate": 2.2579281183932348e-05,
+      "loss": 0.3886,
+      "step": 1560
+    },
+    {
+      "epoch": 1.7815173987450086,
+      "grad_norm": 0.3321929869237414,
+      "learning_rate": 2.2558139534883722e-05,
+      "loss": 0.3675,
+      "step": 1561
+    },
+    {
+      "epoch": 1.782658300057045,
+      "grad_norm": 0.3593851391701366,
+      "learning_rate": 2.2536997885835096e-05,
+      "loss": 0.3589,
+      "step": 1562
+    },
+    {
+      "epoch": 1.7837992013690815,
+      "grad_norm": 0.3095036515987366,
+      "learning_rate": 2.251585623678647e-05,
+      "loss": 0.3628,
+      "step": 1563
+    },
+    {
+      "epoch": 1.784940102681118,
+      "grad_norm": 0.31974279496437447,
+      "learning_rate": 2.2494714587737844e-05,
+      "loss": 0.3642,
+      "step": 1564
+    },
+    {
+      "epoch": 1.7860810039931545,
+      "grad_norm": 0.356612501455481,
+      "learning_rate": 2.2473572938689218e-05,
+      "loss": 0.3601,
+      "step": 1565
+    },
+    {
+      "epoch": 1.787221905305191,
+      "grad_norm": 0.3569612897898255,
+      "learning_rate": 2.2452431289640592e-05,
+      "loss": 0.3815,
+      "step": 1566
+    },
+    {
+      "epoch": 1.7883628066172275,
+      "grad_norm": 0.30994755489776965,
+      "learning_rate": 2.243128964059197e-05,
+      "loss": 0.3647,
+      "step": 1567
+    },
+    {
+      "epoch": 1.7895037079292642,
+      "grad_norm": 0.3737275403511739,
+      "learning_rate": 2.2410147991543344e-05,
+      "loss": 0.3712,
+      "step": 1568
+    },
+    {
+      "epoch": 1.7906446092413006,
+      "grad_norm": 0.407808772953075,
+      "learning_rate": 2.2389006342494714e-05,
+      "loss": 0.3809,
+      "step": 1569
+    },
+    {
+      "epoch": 1.7917855105533371,
+      "grad_norm": 0.3169286511807649,
+      "learning_rate": 2.236786469344609e-05,
+      "loss": 0.3699,
+      "step": 1570
+    },
+    {
+      "epoch": 1.7929264118653736,
+      "grad_norm": 0.3650292862468508,
+      "learning_rate": 2.2346723044397463e-05,
+      "loss": 0.3661,
+      "step": 1571
+    },
+    {
+      "epoch": 1.7940673131774103,
+      "grad_norm": 0.3566495908380766,
+      "learning_rate": 2.2325581395348837e-05,
+      "loss": 0.3502,
+      "step": 1572
+    },
+    {
+      "epoch": 1.7952082144894468,
+      "grad_norm": 0.2875990542081684,
+      "learning_rate": 2.2304439746300214e-05,
+      "loss": 0.3495,
+      "step": 1573
+    },
+    {
+      "epoch": 1.7963491158014833,
+      "grad_norm": 0.32014895991533876,
+      "learning_rate": 2.2283298097251588e-05,
+      "loss": 0.3836,
+      "step": 1574
+    },
+    {
+      "epoch": 1.7974900171135197,
+      "grad_norm": 0.4105886211909479,
+      "learning_rate": 2.2262156448202962e-05,
+      "loss": 0.3775,
+      "step": 1575
+    },
+    {
+      "epoch": 1.7986309184255562,
+      "grad_norm": 0.3737630383761046,
+      "learning_rate": 2.2241014799154336e-05,
+      "loss": 0.3892,
+      "step": 1576
+    },
+    {
+      "epoch": 1.7997718197375927,
+      "grad_norm": 0.3247924558161578,
+      "learning_rate": 2.221987315010571e-05,
+      "loss": 0.3536,
+      "step": 1577
+    },
+    {
+      "epoch": 1.8009127210496292,
+      "grad_norm": 0.3077737827513184,
+      "learning_rate": 2.219873150105708e-05,
+      "loss": 0.3513,
+      "step": 1578
+    },
+    {
+      "epoch": 1.8020536223616657,
+      "grad_norm": 0.30397617493723933,
+      "learning_rate": 2.2177589852008455e-05,
+      "loss": 0.367,
+      "step": 1579
+    },
+    {
+      "epoch": 1.8031945236737021,
+      "grad_norm": 0.2899719233093763,
+      "learning_rate": 2.2156448202959833e-05,
+      "loss": 0.371,
+      "step": 1580
+    },
+    {
+      "epoch": 1.8043354249857386,
+      "grad_norm": 0.250887784452979,
+      "learning_rate": 2.2135306553911207e-05,
+      "loss": 0.3731,
+      "step": 1581
+    },
+    {
+      "epoch": 1.805476326297775,
+      "grad_norm": 0.2988653720689675,
+      "learning_rate": 2.211416490486258e-05,
+      "loss": 0.355,
+      "step": 1582
+    },
+    {
+      "epoch": 1.8066172276098118,
+      "grad_norm": 0.27970342221002986,
+      "learning_rate": 2.2093023255813955e-05,
+      "loss": 0.379,
+      "step": 1583
+    },
+    {
+      "epoch": 1.8077581289218483,
+      "grad_norm": 0.2685564185888628,
+      "learning_rate": 2.207188160676533e-05,
+      "loss": 0.3785,
+      "step": 1584
+    },
+    {
+      "epoch": 1.8088990302338848,
+      "grad_norm": 0.3217391026492225,
+      "learning_rate": 2.2050739957716703e-05,
+      "loss": 0.387,
+      "step": 1585
+    },
+    {
+      "epoch": 1.8100399315459212,
+      "grad_norm": 0.2922819981265088,
+      "learning_rate": 2.2029598308668077e-05,
+      "loss": 0.3792,
+      "step": 1586
+    },
+    {
+      "epoch": 1.811180832857958,
+      "grad_norm": 0.27230409663403116,
+      "learning_rate": 2.200845665961945e-05,
+      "loss": 0.3661,
+      "step": 1587
+    },
+    {
+      "epoch": 1.8123217341699944,
+      "grad_norm": 0.32291579528376907,
+      "learning_rate": 2.1987315010570825e-05,
+      "loss": 0.3796,
+      "step": 1588
+    },
+    {
+      "epoch": 1.813462635482031,
+      "grad_norm": 0.25901863105280115,
+      "learning_rate": 2.19661733615222e-05,
+      "loss": 0.3605,
+      "step": 1589
+    },
+    {
+      "epoch": 1.8146035367940674,
+      "grad_norm": 0.31352888990371003,
+      "learning_rate": 2.1945031712473573e-05,
+      "loss": 0.3766,
+      "step": 1590
+    },
+    {
+      "epoch": 1.8157444381061039,
+      "grad_norm": 0.28862685464242693,
+      "learning_rate": 2.1923890063424947e-05,
+      "loss": 0.3574,
+      "step": 1591
+    },
+    {
+      "epoch": 1.8168853394181403,
+      "grad_norm": 0.2753939853521965,
+      "learning_rate": 2.190274841437632e-05,
+      "loss": 0.3627,
+      "step": 1592
+    },
+    {
+      "epoch": 1.8180262407301768,
+      "grad_norm": 0.27928625456853756,
+      "learning_rate": 2.18816067653277e-05,
+      "loss": 0.3828,
+      "step": 1593
+    },
+    {
+      "epoch": 1.8191671420422133,
+      "grad_norm": 0.25294551894757467,
+      "learning_rate": 2.1860465116279073e-05,
+      "loss": 0.3374,
+      "step": 1594
+    },
+    {
+      "epoch": 1.8203080433542498,
+      "grad_norm": 0.2616026047197691,
+      "learning_rate": 2.1839323467230444e-05,
+      "loss": 0.3416,
+      "step": 1595
+    },
+    {
+      "epoch": 1.8214489446662863,
+      "grad_norm": 0.2409961331311988,
+      "learning_rate": 2.1818181818181818e-05,
+      "loss": 0.3494,
+      "step": 1596
+    },
+    {
+      "epoch": 1.8225898459783227,
+      "grad_norm": 0.26493946117653555,
+      "learning_rate": 2.1797040169133192e-05,
+      "loss": 0.3644,
+      "step": 1597
+    },
+    {
+      "epoch": 1.8237307472903592,
+      "grad_norm": 0.28798554747908567,
+      "learning_rate": 2.1775898520084566e-05,
+      "loss": 0.3624,
+      "step": 1598
+    },
+    {
+      "epoch": 1.824871648602396,
+      "grad_norm": 0.26437027686258674,
+      "learning_rate": 2.1754756871035943e-05,
+      "loss": 0.3572,
+      "step": 1599
+    },
+    {
+      "epoch": 1.8260125499144324,
+      "grad_norm": 0.24691020921569914,
+      "learning_rate": 2.1733615221987317e-05,
+      "loss": 0.3862,
+      "step": 1600
+    },
+    {
+      "epoch": 1.8271534512264689,
+      "grad_norm": 0.2649171577055507,
+      "learning_rate": 2.171247357293869e-05,
+      "loss": 0.3527,
+      "step": 1601
+    },
+    {
+      "epoch": 1.8282943525385056,
+      "grad_norm": 0.28647906553531793,
+      "learning_rate": 2.1691331923890066e-05,
+      "loss": 0.379,
+      "step": 1602
+    },
+    {
+      "epoch": 1.829435253850542,
+      "grad_norm": 0.27888222041894295,
+      "learning_rate": 2.167019027484144e-05,
+      "loss": 0.3841,
+      "step": 1603
+    },
+    {
+      "epoch": 1.8305761551625785,
+      "grad_norm": 0.25803288173345323,
+      "learning_rate": 2.164904862579281e-05,
+      "loss": 0.3496,
+      "step": 1604
+    },
+    {
+      "epoch": 1.831717056474615,
+      "grad_norm": 0.3152294889154458,
+      "learning_rate": 2.1627906976744184e-05,
+      "loss": 0.3796,
+      "step": 1605
+    },
+    {
+      "epoch": 1.8328579577866515,
+      "grad_norm": 0.2837028459205526,
+      "learning_rate": 2.1606765327695562e-05,
+      "loss": 0.3718,
+      "step": 1606
+    },
+    {
+      "epoch": 1.833998859098688,
+      "grad_norm": 0.2942487206676281,
+      "learning_rate": 2.1585623678646936e-05,
+      "loss": 0.3672,
+      "step": 1607
+    },
+    {
+      "epoch": 1.8351397604107245,
+      "grad_norm": 0.27094002473416,
+      "learning_rate": 2.156448202959831e-05,
+      "loss": 0.3825,
+      "step": 1608
+    },
+    {
+      "epoch": 1.836280661722761,
+      "grad_norm": 0.3080124085299322,
+      "learning_rate": 2.1543340380549684e-05,
+      "loss": 0.3768,
+      "step": 1609
+    },
+    {
+      "epoch": 1.8374215630347974,
+      "grad_norm": 0.29458146796584034,
+      "learning_rate": 2.1522198731501058e-05,
+      "loss": 0.3826,
+      "step": 1610
+    },
+    {
+      "epoch": 1.838562464346834,
+      "grad_norm": 0.2709545289583168,
+      "learning_rate": 2.1501057082452432e-05,
+      "loss": 0.3492,
+      "step": 1611
+    },
+    {
+      "epoch": 1.8397033656588704,
+      "grad_norm": 0.2626488042697663,
+      "learning_rate": 2.1479915433403806e-05,
+      "loss": 0.3576,
+      "step": 1612
+    },
+    {
+      "epoch": 1.8408442669709069,
+      "grad_norm": 0.29505118436293093,
+      "learning_rate": 2.145877378435518e-05,
+      "loss": 0.3679,
+      "step": 1613
+    },
+    {
+      "epoch": 1.8419851682829436,
+      "grad_norm": 0.2811956966393874,
+      "learning_rate": 2.1437632135306555e-05,
+      "loss": 0.3706,
+      "step": 1614
+    },
+    {
+      "epoch": 1.84312606959498,
+      "grad_norm": 0.2857476313553613,
+      "learning_rate": 2.141649048625793e-05,
+      "loss": 0.363,
+      "step": 1615
+    },
+    {
+      "epoch": 1.8442669709070165,
+      "grad_norm": 0.26379976558838814,
+      "learning_rate": 2.1395348837209303e-05,
+      "loss": 0.3656,
+      "step": 1616
+    },
+    {
+      "epoch": 1.845407872219053,
+      "grad_norm": 0.30310468777418886,
+      "learning_rate": 2.1374207188160677e-05,
+      "loss": 0.3766,
+      "step": 1617
+    },
+    {
+      "epoch": 1.8465487735310897,
+      "grad_norm": 0.29055054780933476,
+      "learning_rate": 2.135306553911205e-05,
+      "loss": 0.3768,
+      "step": 1618
+    },
+    {
+      "epoch": 1.8476896748431262,
+      "grad_norm": 0.26275911199095003,
+      "learning_rate": 2.1331923890063428e-05,
+      "loss": 0.3655,
+      "step": 1619
+    },
+    {
+      "epoch": 1.8488305761551627,
+      "grad_norm": 0.3075089578022422,
+      "learning_rate": 2.1310782241014802e-05,
+      "loss": 0.3671,
+      "step": 1620
+    },
+    {
+      "epoch": 1.8499714774671991,
+      "grad_norm": 0.24011215531163518,
+      "learning_rate": 2.1289640591966176e-05,
+      "loss": 0.3392,
+      "step": 1621
+    },
+    {
+      "epoch": 1.8511123787792356,
+      "grad_norm": 0.3003670940065247,
+      "learning_rate": 2.1268498942917547e-05,
+      "loss": 0.3745,
+      "step": 1622
+    },
+    {
+      "epoch": 1.852253280091272,
+      "grad_norm": 0.32867835263049694,
+      "learning_rate": 2.124735729386892e-05,
+      "loss": 0.374,
+      "step": 1623
+    },
+    {
+      "epoch": 1.8533941814033086,
+      "grad_norm": 0.2736149864201345,
+      "learning_rate": 2.1226215644820295e-05,
+      "loss": 0.3866,
+      "step": 1624
+    },
+    {
+      "epoch": 1.854535082715345,
+      "grad_norm": 0.32004051514876714,
+      "learning_rate": 2.120507399577167e-05,
+      "loss": 0.3499,
+      "step": 1625
+    },
+    {
+      "epoch": 1.8556759840273815,
+      "grad_norm": 0.3104696099218829,
+      "learning_rate": 2.1183932346723047e-05,
+      "loss": 0.3821,
+      "step": 1626
+    },
+    {
+      "epoch": 1.856816885339418,
+      "grad_norm": 0.3296864222385294,
+      "learning_rate": 2.116279069767442e-05,
+      "loss": 0.3764,
+      "step": 1627
+    },
+    {
+      "epoch": 1.8579577866514545,
+      "grad_norm": 0.35311586083947655,
+      "learning_rate": 2.1141649048625795e-05,
+      "loss": 0.3784,
+      "step": 1628
+    },
+    {
+      "epoch": 1.8590986879634912,
+      "grad_norm": 0.2645700372496905,
+      "learning_rate": 2.112050739957717e-05,
+      "loss": 0.3465,
+      "step": 1629
+    },
+    {
+      "epoch": 1.8602395892755277,
+      "grad_norm": 0.34980987571760414,
+      "learning_rate": 2.1099365750528543e-05,
+      "loss": 0.3638,
+      "step": 1630
+    },
+    {
+      "epoch": 1.8613804905875642,
+      "grad_norm": 0.279728784438425,
+      "learning_rate": 2.1078224101479914e-05,
+      "loss": 0.3711,
+      "step": 1631
+    },
+    {
+      "epoch": 1.8625213918996006,
+      "grad_norm": 0.2753861016019802,
+      "learning_rate": 2.105708245243129e-05,
+      "loss": 0.3683,
+      "step": 1632
+    },
+    {
+      "epoch": 1.8636622932116373,
+      "grad_norm": 0.2979160433487785,
+      "learning_rate": 2.1035940803382665e-05,
+      "loss": 0.3649,
+      "step": 1633
+    },
+    {
+      "epoch": 1.8648031945236738,
+      "grad_norm": 0.28213788188185157,
+      "learning_rate": 2.101479915433404e-05,
+      "loss": 0.3551,
+      "step": 1634
+    },
+    {
+      "epoch": 1.8659440958357103,
+      "grad_norm": 0.29960423038143713,
+      "learning_rate": 2.0993657505285413e-05,
+      "loss": 0.3468,
+      "step": 1635
+    },
+    {
+      "epoch": 1.8670849971477468,
+      "grad_norm": 0.27966434696237047,
+      "learning_rate": 2.0972515856236788e-05,
+      "loss": 0.3631,
+      "step": 1636
+    },
+    {
+      "epoch": 1.8682258984597833,
+      "grad_norm": 0.35759260629909057,
+      "learning_rate": 2.095137420718816e-05,
+      "loss": 0.3776,
+      "step": 1637
+    },
+    {
+      "epoch": 1.8693667997718197,
+      "grad_norm": 0.24413963286231924,
+      "learning_rate": 2.0930232558139536e-05,
+      "loss": 0.357,
+      "step": 1638
+    },
+    {
+      "epoch": 1.8705077010838562,
+      "grad_norm": 0.2942994597300611,
+      "learning_rate": 2.090909090909091e-05,
+      "loss": 0.3606,
+      "step": 1639
+    },
+    {
+      "epoch": 1.8716486023958927,
+      "grad_norm": 0.36190249258733775,
+      "learning_rate": 2.0887949260042284e-05,
+      "loss": 0.3752,
+      "step": 1640
+    },
+    {
+      "epoch": 1.8727895037079292,
+      "grad_norm": 0.2608033194653284,
+      "learning_rate": 2.0866807610993658e-05,
+      "loss": 0.3452,
+      "step": 1641
+    },
+    {
+      "epoch": 1.8739304050199657,
+      "grad_norm": 0.26918936097256435,
+      "learning_rate": 2.0845665961945032e-05,
+      "loss": 0.3739,
+      "step": 1642
+    },
+    {
+      "epoch": 1.8750713063320021,
+      "grad_norm": 0.31655255219302714,
+      "learning_rate": 2.0824524312896406e-05,
+      "loss": 0.3906,
+      "step": 1643
+    },
+    {
+      "epoch": 1.8762122076440388,
+      "grad_norm": 0.22432722104783256,
+      "learning_rate": 2.080338266384778e-05,
+      "loss": 0.3452,
+      "step": 1644
+    },
+    {
+      "epoch": 1.8773531089560753,
+      "grad_norm": 0.27099992542216245,
+      "learning_rate": 2.0782241014799158e-05,
+      "loss": 0.3562,
+      "step": 1645
+    },
+    {
+      "epoch": 1.8784940102681118,
+      "grad_norm": 0.2687749493001164,
+      "learning_rate": 2.076109936575053e-05,
+      "loss": 0.3839,
+      "step": 1646
+    },
+    {
+      "epoch": 1.8796349115801483,
+      "grad_norm": 0.2745458632822797,
+      "learning_rate": 2.0739957716701906e-05,
+      "loss": 0.3585,
+      "step": 1647
+    },
+    {
+      "epoch": 1.880775812892185,
+      "grad_norm": 0.2649532173440124,
+      "learning_rate": 2.0718816067653276e-05,
+      "loss": 0.3643,
+      "step": 1648
+    },
+    {
+      "epoch": 1.8819167142042215,
+      "grad_norm": 0.2926959708581756,
+      "learning_rate": 2.069767441860465e-05,
+      "loss": 0.362,
+      "step": 1649
+    },
+    {
+      "epoch": 1.883057615516258,
+      "grad_norm": 0.28133283588264074,
+      "learning_rate": 2.0676532769556025e-05,
+      "loss": 0.3672,
+      "step": 1650
+    },
+    {
+      "epoch": 1.8841985168282944,
+      "grad_norm": 0.26954237691342353,
+      "learning_rate": 2.06553911205074e-05,
+      "loss": 0.3632,
+      "step": 1651
+    },
+    {
+      "epoch": 1.885339418140331,
+      "grad_norm": 0.28736812482531,
+      "learning_rate": 2.0634249471458776e-05,
+      "loss": 0.3413,
+      "step": 1652
+    },
+    {
+      "epoch": 1.8864803194523674,
+      "grad_norm": 0.2743819590826279,
+      "learning_rate": 2.061310782241015e-05,
+      "loss": 0.378,
+      "step": 1653
+    },
+    {
+      "epoch": 1.8876212207644039,
+      "grad_norm": 0.30624176993565166,
+      "learning_rate": 2.0591966173361524e-05,
+      "loss": 0.3663,
+      "step": 1654
+    },
+    {
+      "epoch": 1.8887621220764403,
+      "grad_norm": 0.2589219913447823,
+      "learning_rate": 2.0570824524312898e-05,
+      "loss": 0.3492,
+      "step": 1655
+    },
+    {
+      "epoch": 1.8899030233884768,
+      "grad_norm": 0.2768282482738376,
+      "learning_rate": 2.0549682875264272e-05,
+      "loss": 0.3618,
+      "step": 1656
+    },
+    {
+      "epoch": 1.8910439247005133,
+      "grad_norm": 0.2739420056242106,
+      "learning_rate": 2.0528541226215643e-05,
+      "loss": 0.3555,
+      "step": 1657
+    },
+    {
+      "epoch": 1.8921848260125498,
+      "grad_norm": 0.26928054226917636,
+      "learning_rate": 2.050739957716702e-05,
+      "loss": 0.3558,
+      "step": 1658
+    },
+    {
+      "epoch": 1.8933257273245863,
+      "grad_norm": 0.3207377769023388,
+      "learning_rate": 2.0486257928118395e-05,
+      "loss": 0.3798,
+      "step": 1659
+    },
+    {
+      "epoch": 1.894466628636623,
+      "grad_norm": 0.24989405415599858,
+      "learning_rate": 2.046511627906977e-05,
+      "loss": 0.3613,
+      "step": 1660
+    },
+    {
+      "epoch": 1.8956075299486594,
+      "grad_norm": 0.2607195645529972,
+      "learning_rate": 2.0443974630021143e-05,
+      "loss": 0.3709,
+      "step": 1661
+    },
+    {
+      "epoch": 1.896748431260696,
+      "grad_norm": 0.2733555160670365,
+      "learning_rate": 2.0422832980972517e-05,
+      "loss": 0.3614,
+      "step": 1662
+    },
+    {
+      "epoch": 1.8978893325727326,
+      "grad_norm": 0.27871241601853286,
+      "learning_rate": 2.040169133192389e-05,
+      "loss": 0.3648,
+      "step": 1663
+    },
+    {
+      "epoch": 1.899030233884769,
+      "grad_norm": 0.31234592073077655,
+      "learning_rate": 2.0380549682875265e-05,
+      "loss": 0.3683,
+      "step": 1664
+    },
+    {
+      "epoch": 1.9001711351968056,
+      "grad_norm": 0.27741519189889535,
+      "learning_rate": 2.0359408033826642e-05,
+      "loss": 0.373,
+      "step": 1665
+    },
+    {
+      "epoch": 1.901312036508842,
+      "grad_norm": 0.28054435993083715,
+      "learning_rate": 2.0338266384778013e-05,
+      "loss": 0.3557,
+      "step": 1666
+    },
+    {
+      "epoch": 1.9024529378208785,
+      "grad_norm": 0.27814457755739297,
+      "learning_rate": 2.0317124735729387e-05,
+      "loss": 0.3858,
+      "step": 1667
+    },
+    {
+      "epoch": 1.903593839132915,
+      "grad_norm": 0.2884539606763188,
+      "learning_rate": 2.029598308668076e-05,
+      "loss": 0.3741,
+      "step": 1668
+    },
+    {
+      "epoch": 1.9047347404449515,
+      "grad_norm": 0.2618620359150727,
+      "learning_rate": 2.0274841437632135e-05,
+      "loss": 0.3649,
+      "step": 1669
+    },
+    {
+      "epoch": 1.905875641756988,
+      "grad_norm": 0.25113408676713106,
+      "learning_rate": 2.025369978858351e-05,
+      "loss": 0.3622,
+      "step": 1670
+    },
+    {
+      "epoch": 1.9070165430690245,
+      "grad_norm": 0.2784177351306522,
+      "learning_rate": 2.0232558139534883e-05,
+      "loss": 0.3517,
+      "step": 1671
+    },
+    {
+      "epoch": 1.908157444381061,
+      "grad_norm": 0.2856685467357065,
+      "learning_rate": 2.021141649048626e-05,
+      "loss": 0.3683,
+      "step": 1672
+    },
+    {
+      "epoch": 1.9092983456930974,
+      "grad_norm": 0.26172116445181576,
+      "learning_rate": 2.0190274841437635e-05,
+      "loss": 0.354,
+      "step": 1673
+    },
+    {
+      "epoch": 1.910439247005134,
+      "grad_norm": 0.26335022190661506,
+      "learning_rate": 2.016913319238901e-05,
+      "loss": 0.3507,
+      "step": 1674
+    },
+    {
+      "epoch": 1.9115801483171706,
+      "grad_norm": 0.25622287485822953,
+      "learning_rate": 2.014799154334038e-05,
+      "loss": 0.3731,
+      "step": 1675
+    },
+    {
+      "epoch": 1.912721049629207,
+      "grad_norm": 0.2601225492512362,
+      "learning_rate": 2.0126849894291754e-05,
+      "loss": 0.3632,
+      "step": 1676
+    },
+    {
+      "epoch": 1.9138619509412436,
+      "grad_norm": 0.2604903931429644,
+      "learning_rate": 2.0105708245243128e-05,
+      "loss": 0.3519,
+      "step": 1677
+    },
+    {
+      "epoch": 1.91500285225328,
+      "grad_norm": 0.24755000365197305,
+      "learning_rate": 2.0084566596194505e-05,
+      "loss": 0.3564,
+      "step": 1678
+    },
+    {
+      "epoch": 1.9161437535653167,
+      "grad_norm": 0.2638138653479009,
+      "learning_rate": 2.006342494714588e-05,
+      "loss": 0.3626,
+      "step": 1679
+    },
+    {
+      "epoch": 1.9172846548773532,
+      "grad_norm": 0.25681513482848295,
+      "learning_rate": 2.0042283298097253e-05,
+      "loss": 0.3581,
+      "step": 1680
+    },
+    {
+      "epoch": 1.9184255561893897,
+      "grad_norm": 0.2557089013629714,
+      "learning_rate": 2.0021141649048628e-05,
+      "loss": 0.3369,
+      "step": 1681
+    },
+    {
+      "epoch": 1.9195664575014262,
+      "grad_norm": 0.30474490842943003,
+      "learning_rate": 2e-05,
+      "loss": 0.3586,
+      "step": 1682
+    },
+    {
+      "epoch": 1.9207073588134627,
+      "grad_norm": 0.23362258404659542,
+      "learning_rate": 1.9978858350951372e-05,
+      "loss": 0.3548,
+      "step": 1683
+    },
+    {
+      "epoch": 1.9218482601254991,
+      "grad_norm": 0.282854268486239,
+      "learning_rate": 1.995771670190275e-05,
+      "loss": 0.3833,
+      "step": 1684
+    },
+    {
+      "epoch": 1.9229891614375356,
+      "grad_norm": 0.28664344992691926,
+      "learning_rate": 1.9936575052854124e-05,
+      "loss": 0.3568,
+      "step": 1685
+    },
+    {
+      "epoch": 1.924130062749572,
+      "grad_norm": 0.2671099649115389,
+      "learning_rate": 1.9915433403805498e-05,
+      "loss": 0.3646,
+      "step": 1686
+    },
+    {
+      "epoch": 1.9252709640616086,
+      "grad_norm": 0.2954713254197816,
+      "learning_rate": 1.9894291754756872e-05,
+      "loss": 0.3643,
+      "step": 1687
+    },
+    {
+      "epoch": 1.926411865373645,
+      "grad_norm": 0.33930978902768366,
+      "learning_rate": 1.9873150105708246e-05,
+      "loss": 0.3769,
+      "step": 1688
+    },
+    {
+      "epoch": 1.9275527666856815,
+      "grad_norm": 0.29824577630447163,
+      "learning_rate": 1.985200845665962e-05,
+      "loss": 0.3768,
+      "step": 1689
+    },
+    {
+      "epoch": 1.9286936679977182,
+      "grad_norm": 0.29176022977395993,
+      "learning_rate": 1.9830866807610994e-05,
+      "loss": 0.3507,
+      "step": 1690
+    },
+    {
+      "epoch": 1.9298345693097547,
+      "grad_norm": 0.24390654056177216,
+      "learning_rate": 1.980972515856237e-05,
+      "loss": 0.3506,
+      "step": 1691
+    },
+    {
+      "epoch": 1.9309754706217912,
+      "grad_norm": 0.2590795747683378,
+      "learning_rate": 1.9788583509513742e-05,
+      "loss": 0.3501,
+      "step": 1692
+    },
+    {
+      "epoch": 1.9321163719338277,
+      "grad_norm": 0.28164134893382536,
+      "learning_rate": 1.9767441860465116e-05,
+      "loss": 0.3608,
+      "step": 1693
+    },
+    {
+      "epoch": 1.9332572732458644,
+      "grad_norm": 0.2647748061214524,
+      "learning_rate": 1.974630021141649e-05,
+      "loss": 0.3606,
+      "step": 1694
+    },
+    {
+      "epoch": 1.9343981745579009,
+      "grad_norm": 0.28010445382267146,
+      "learning_rate": 1.9725158562367865e-05,
+      "loss": 0.3581,
+      "step": 1695
+    },
+    {
+      "epoch": 1.9355390758699373,
+      "grad_norm": 0.2533929878448776,
+      "learning_rate": 1.970401691331924e-05,
+      "loss": 0.3706,
+      "step": 1696
+    },
+    {
+      "epoch": 1.9366799771819738,
+      "grad_norm": 0.28611878053557527,
+      "learning_rate": 1.9682875264270613e-05,
+      "loss": 0.3728,
+      "step": 1697
+    },
+    {
+      "epoch": 1.9378208784940103,
+      "grad_norm": 0.28041130653252294,
+      "learning_rate": 1.966173361522199e-05,
+      "loss": 0.3622,
+      "step": 1698
+    },
+    {
+      "epoch": 1.9389617798060468,
+      "grad_norm": 0.28497047972994977,
+      "learning_rate": 1.9640591966173364e-05,
+      "loss": 0.3563,
+      "step": 1699
+    },
+    {
+      "epoch": 1.9401026811180833,
+      "grad_norm": 0.2918310757848588,
+      "learning_rate": 1.961945031712474e-05,
+      "loss": 0.3766,
+      "step": 1700
+    },
+    {
+      "epoch": 1.9412435824301197,
+      "grad_norm": 0.2369182789719308,
+      "learning_rate": 1.959830866807611e-05,
+      "loss": 0.3478,
+      "step": 1701
+    },
+    {
+      "epoch": 1.9423844837421562,
+      "grad_norm": 0.2647060419296417,
+      "learning_rate": 1.9577167019027483e-05,
+      "loss": 0.3582,
+      "step": 1702
+    },
+    {
+      "epoch": 1.9435253850541927,
+      "grad_norm": 0.2920307858064205,
+      "learning_rate": 1.9556025369978857e-05,
+      "loss": 0.369,
+      "step": 1703
+    },
+    {
+      "epoch": 1.9446662863662292,
+      "grad_norm": 0.3227097667935105,
+      "learning_rate": 1.9534883720930235e-05,
+      "loss": 0.368,
+      "step": 1704
+    },
+    {
+      "epoch": 1.9458071876782659,
+      "grad_norm": 0.2667720664080837,
+      "learning_rate": 1.951374207188161e-05,
+      "loss": 0.3637,
+      "step": 1705
+    },
+    {
+      "epoch": 1.9469480889903024,
+      "grad_norm": 0.3062658322537471,
+      "learning_rate": 1.9492600422832983e-05,
+      "loss": 0.3663,
+      "step": 1706
+    },
+    {
+      "epoch": 1.9480889903023388,
+      "grad_norm": 0.27305673090240135,
+      "learning_rate": 1.9471458773784357e-05,
+      "loss": 0.3622,
+      "step": 1707
+    },
+    {
+      "epoch": 1.9492298916143753,
+      "grad_norm": 0.2641687358716487,
+      "learning_rate": 1.945031712473573e-05,
+      "loss": 0.355,
+      "step": 1708
+    },
+    {
+      "epoch": 1.950370792926412,
+      "grad_norm": 0.2580431501007941,
+      "learning_rate": 1.9429175475687105e-05,
+      "loss": 0.3652,
+      "step": 1709
+    },
+    {
+      "epoch": 1.9515116942384485,
+      "grad_norm": 0.2344794936808685,
+      "learning_rate": 1.940803382663848e-05,
+      "loss": 0.3468,
+      "step": 1710
+    },
+    {
+      "epoch": 1.952652595550485,
+      "grad_norm": 0.2886415993544298,
+      "learning_rate": 1.9386892177589853e-05,
+      "loss": 0.3757,
+      "step": 1711
+    },
+    {
+      "epoch": 1.9537934968625215,
+      "grad_norm": 0.3049085901486973,
+      "learning_rate": 1.9365750528541227e-05,
+      "loss": 0.3737,
+      "step": 1712
+    },
+    {
+      "epoch": 1.954934398174558,
+      "grad_norm": 0.26596397633339325,
+      "learning_rate": 1.93446088794926e-05,
+      "loss": 0.3642,
+      "step": 1713
+    },
+    {
+      "epoch": 1.9560752994865944,
+      "grad_norm": 0.30294710633352423,
+      "learning_rate": 1.9323467230443975e-05,
+      "loss": 0.3668,
+      "step": 1714
+    },
+    {
+      "epoch": 1.957216200798631,
+      "grad_norm": 0.26349235851430564,
+      "learning_rate": 1.930232558139535e-05,
+      "loss": 0.3559,
+      "step": 1715
+    },
+    {
+      "epoch": 1.9583571021106674,
+      "grad_norm": 0.2685783037927963,
+      "learning_rate": 1.9281183932346724e-05,
+      "loss": 0.3774,
+      "step": 1716
+    },
+    {
+      "epoch": 1.9594980034227039,
+      "grad_norm": 0.264292192361419,
+      "learning_rate": 1.9260042283298098e-05,
+      "loss": 0.3735,
+      "step": 1717
+    },
+    {
+      "epoch": 1.9606389047347403,
+      "grad_norm": 0.3554897307494108,
+      "learning_rate": 1.9238900634249475e-05,
+      "loss": 0.371,
+      "step": 1718
+    },
+    {
+      "epoch": 1.9617798060467768,
+      "grad_norm": 0.31087400630355955,
+      "learning_rate": 1.9217758985200846e-05,
+      "loss": 0.3661,
+      "step": 1719
+    },
+    {
+      "epoch": 1.9629207073588133,
+      "grad_norm": 0.3095791509449235,
+      "learning_rate": 1.919661733615222e-05,
+      "loss": 0.3706,
+      "step": 1720
+    },
+    {
+      "epoch": 1.96406160867085,
+      "grad_norm": 0.2738969954461718,
+      "learning_rate": 1.9175475687103594e-05,
+      "loss": 0.3876,
+      "step": 1721
+    },
+    {
+      "epoch": 1.9652025099828865,
+      "grad_norm": 0.34003420604038476,
+      "learning_rate": 1.9154334038054968e-05,
+      "loss": 0.3956,
+      "step": 1722
+    },
+    {
+      "epoch": 1.966343411294923,
+      "grad_norm": 0.28565166124104413,
+      "learning_rate": 1.9133192389006342e-05,
+      "loss": 0.3478,
+      "step": 1723
+    },
+    {
+      "epoch": 1.9674843126069597,
+      "grad_norm": 0.2632293762228427,
+      "learning_rate": 1.911205073995772e-05,
+      "loss": 0.3672,
+      "step": 1724
+    },
+    {
+      "epoch": 1.9686252139189961,
+      "grad_norm": 0.2973825583942189,
+      "learning_rate": 1.9090909090909094e-05,
+      "loss": 0.3726,
+      "step": 1725
+    },
+    {
+      "epoch": 1.9697661152310326,
+      "grad_norm": 0.2525267995321032,
+      "learning_rate": 1.9069767441860468e-05,
+      "loss": 0.374,
+      "step": 1726
+    },
+    {
+      "epoch": 1.970907016543069,
+      "grad_norm": 0.24811016517447812,
+      "learning_rate": 1.904862579281184e-05,
+      "loss": 0.3956,
+      "step": 1727
+    },
+    {
+      "epoch": 1.9720479178551056,
+      "grad_norm": 0.3270553538748556,
+      "learning_rate": 1.9027484143763212e-05,
+      "loss": 0.3662,
+      "step": 1728
+    },
+    {
+      "epoch": 1.973188819167142,
+      "grad_norm": 0.2541293268170106,
+      "learning_rate": 1.9006342494714586e-05,
+      "loss": 0.3403,
+      "step": 1729
+    },
+    {
+      "epoch": 1.9743297204791785,
+      "grad_norm": 0.27007055492054866,
+      "learning_rate": 1.8985200845665964e-05,
+      "loss": 0.3598,
+      "step": 1730
+    },
+    {
+      "epoch": 1.975470621791215,
+      "grad_norm": 0.2789281069327813,
+      "learning_rate": 1.8964059196617338e-05,
+      "loss": 0.3708,
+      "step": 1731
+    },
+    {
+      "epoch": 1.9766115231032515,
+      "grad_norm": 0.335240345320772,
+      "learning_rate": 1.8942917547568712e-05,
+      "loss": 0.3759,
+      "step": 1732
+    },
+    {
+      "epoch": 1.977752424415288,
+      "grad_norm": 0.3014749329262505,
+      "learning_rate": 1.8921775898520086e-05,
+      "loss": 0.3743,
+      "step": 1733
+    },
+    {
+      "epoch": 1.9788933257273245,
+      "grad_norm": 0.2642588744547289,
+      "learning_rate": 1.890063424947146e-05,
+      "loss": 0.3742,
+      "step": 1734
+    },
+    {
+      "epoch": 1.980034227039361,
+      "grad_norm": 0.3913575419827154,
+      "learning_rate": 1.8879492600422834e-05,
+      "loss": 0.3467,
+      "step": 1735
+    },
+    {
+      "epoch": 1.9811751283513976,
+      "grad_norm": 0.3105592595689725,
+      "learning_rate": 1.885835095137421e-05,
+      "loss": 0.3998,
+      "step": 1736
+    },
+    {
+      "epoch": 1.9823160296634341,
+      "grad_norm": 0.294507885586603,
+      "learning_rate": 1.8837209302325582e-05,
+      "loss": 0.3636,
+      "step": 1737
+    },
+    {
+      "epoch": 1.9834569309754706,
+      "grad_norm": 0.32337663354893215,
+      "learning_rate": 1.8816067653276956e-05,
+      "loss": 0.3502,
+      "step": 1738
+    },
+    {
+      "epoch": 1.984597832287507,
+      "grad_norm": 0.3285424012955404,
+      "learning_rate": 1.879492600422833e-05,
+      "loss": 0.3856,
+      "step": 1739
+    },
+    {
+      "epoch": 1.9857387335995438,
+      "grad_norm": 0.2904265186501213,
+      "learning_rate": 1.8773784355179705e-05,
+      "loss": 0.371,
+      "step": 1740
+    },
+    {
+      "epoch": 1.9868796349115803,
+      "grad_norm": 0.2837758477482779,
+      "learning_rate": 1.875264270613108e-05,
+      "loss": 0.3556,
+      "step": 1741
+    },
+    {
+      "epoch": 1.9880205362236167,
+      "grad_norm": 0.3199499304677922,
+      "learning_rate": 1.8731501057082453e-05,
+      "loss": 0.3781,
+      "step": 1742
+    },
+    {
+      "epoch": 1.9891614375356532,
+      "grad_norm": 0.2620483932749284,
+      "learning_rate": 1.8710359408033827e-05,
+      "loss": 0.347,
+      "step": 1743
+    },
+    {
+      "epoch": 1.9903023388476897,
+      "grad_norm": 0.31940723204059734,
+      "learning_rate": 1.8689217758985204e-05,
+      "loss": 0.3531,
+      "step": 1744
+    },
+    {
+      "epoch": 1.9914432401597262,
+      "grad_norm": 0.2623270110349358,
+      "learning_rate": 1.8668076109936575e-05,
+      "loss": 0.3779,
+      "step": 1745
+    },
+    {
+      "epoch": 1.9925841414717627,
+      "grad_norm": 0.2877974818385608,
+      "learning_rate": 1.864693446088795e-05,
+      "loss": 0.3796,
+      "step": 1746
+    },
+    {
+      "epoch": 1.9937250427837991,
+      "grad_norm": 0.30045186550520864,
+      "learning_rate": 1.8625792811839323e-05,
+      "loss": 0.3884,
+      "step": 1747
+    },
+    {
+      "epoch": 1.9948659440958356,
+      "grad_norm": 0.2869930913760931,
+      "learning_rate": 1.8604651162790697e-05,
+      "loss": 0.3359,
+      "step": 1748
+    },
+    {
+      "epoch": 1.996006845407872,
+      "grad_norm": 0.2916651153608028,
+      "learning_rate": 1.858350951374207e-05,
+      "loss": 0.359,
+      "step": 1749
+    },
+    {
+      "epoch": 1.9971477467199086,
+      "grad_norm": 0.32125796964460684,
+      "learning_rate": 1.856236786469345e-05,
+      "loss": 0.391,
+      "step": 1750
+    },
+    {
+      "epoch": 1.9982886480319453,
+      "grad_norm": 0.28157904329195294,
+      "learning_rate": 1.8541226215644823e-05,
+      "loss": 0.3531,
+      "step": 1751
+    },
+    {
+      "epoch": 1.9994295493439818,
+      "grad_norm": 0.28627744809790545,
+      "learning_rate": 1.8520084566596197e-05,
+      "loss": 0.3655,
+      "step": 1752
+    },
+    {
+      "epoch": 2.0011409013120365,
+      "grad_norm": 0.4487577079137177,
+      "learning_rate": 1.849894291754757e-05,
+      "loss": 0.6555,
+      "step": 1753
+    },
+    {
+      "epoch": 2.002281802624073,
+      "grad_norm": 0.28088289953032464,
+      "learning_rate": 1.847780126849894e-05,
+      "loss": 0.2958,
+      "step": 1754
+    },
+    {
+      "epoch": 2.0034227039361094,
+      "grad_norm": 0.2891272992850902,
+      "learning_rate": 1.8456659619450316e-05,
+      "loss": 0.3185,
+      "step": 1755
+    },
+    {
+      "epoch": 2.004563605248146,
+      "grad_norm": 0.4112554650528418,
+      "learning_rate": 1.8435517970401693e-05,
+      "loss": 0.3052,
+      "step": 1756
+    },
+    {
+      "epoch": 2.0057045065601824,
+      "grad_norm": 0.26946760277582954,
+      "learning_rate": 1.8414376321353067e-05,
+      "loss": 0.3075,
+      "step": 1757
+    },
+    {
+      "epoch": 2.006845407872219,
+      "grad_norm": 0.3043281070992764,
+      "learning_rate": 1.839323467230444e-05,
+      "loss": 0.2972,
+      "step": 1758
+    },
+    {
+      "epoch": 2.0079863091842554,
+      "grad_norm": 0.32883104084724435,
+      "learning_rate": 1.8372093023255815e-05,
+      "loss": 0.2994,
+      "step": 1759
+    },
+    {
+      "epoch": 2.0091272104962923,
+      "grad_norm": 0.30200499224952815,
+      "learning_rate": 1.835095137420719e-05,
+      "loss": 0.3,
+      "step": 1760
+    },
+    {
+      "epoch": 2.0102681118083288,
+      "grad_norm": 0.2828662075516682,
+      "learning_rate": 1.8329809725158564e-05,
+      "loss": 0.2955,
+      "step": 1761
+    },
+    {
+      "epoch": 2.0114090131203652,
+      "grad_norm": 0.2992366159862029,
+      "learning_rate": 1.8308668076109938e-05,
+      "loss": 0.3035,
+      "step": 1762
+    },
+    {
+      "epoch": 2.0125499144324017,
+      "grad_norm": 0.3371880577990774,
+      "learning_rate": 1.8287526427061312e-05,
+      "loss": 0.2949,
+      "step": 1763
+    },
+    {
+      "epoch": 2.013690815744438,
+      "grad_norm": 0.28344316901737787,
+      "learning_rate": 1.8266384778012686e-05,
+      "loss": 0.3048,
+      "step": 1764
+    },
+    {
+      "epoch": 2.0148317170564747,
+      "grad_norm": 0.27179491325048666,
+      "learning_rate": 1.824524312896406e-05,
+      "loss": 0.2888,
+      "step": 1765
+    },
+    {
+      "epoch": 2.015972618368511,
+      "grad_norm": 0.2727511224813464,
+      "learning_rate": 1.8224101479915434e-05,
+      "loss": 0.2915,
+      "step": 1766
+    },
+    {
+      "epoch": 2.0171135196805476,
+      "grad_norm": 0.35663638818735105,
+      "learning_rate": 1.8202959830866808e-05,
+      "loss": 0.2928,
+      "step": 1767
+    },
+    {
+      "epoch": 2.018254420992584,
+      "grad_norm": 0.28095383665244567,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 0.2927,
+      "step": 1768
+    },
+    {
+      "epoch": 2.0193953223046206,
+      "grad_norm": 0.2711800724825952,
+      "learning_rate": 1.8160676532769556e-05,
+      "loss": 0.3033,
+      "step": 1769
+    },
+    {
+      "epoch": 2.020536223616657,
+      "grad_norm": 0.30322673551396767,
+      "learning_rate": 1.8139534883720934e-05,
+      "loss": 0.2953,
+      "step": 1770
+    },
+    {
+      "epoch": 2.0216771249286936,
+      "grad_norm": 0.2750325414334585,
+      "learning_rate": 1.8118393234672304e-05,
+      "loss": 0.2815,
+      "step": 1771
+    },
+    {
+      "epoch": 2.02281802624073,
+      "grad_norm": 0.2472532660983106,
+      "learning_rate": 1.809725158562368e-05,
+      "loss": 0.2921,
+      "step": 1772
+    },
+    {
+      "epoch": 2.0239589275527665,
+      "grad_norm": 0.2927848622819586,
+      "learning_rate": 1.8076109936575052e-05,
+      "loss": 0.3057,
+      "step": 1773
+    },
+    {
+      "epoch": 2.025099828864803,
+      "grad_norm": 0.2294021321585527,
+      "learning_rate": 1.8054968287526427e-05,
+      "loss": 0.2974,
+      "step": 1774
+    },
+    {
+      "epoch": 2.02624073017684,
+      "grad_norm": 0.26484271651685803,
+      "learning_rate": 1.80338266384778e-05,
+      "loss": 0.2848,
+      "step": 1775
+    },
+    {
+      "epoch": 2.0273816314888764,
+      "grad_norm": 0.2538624793787531,
+      "learning_rate": 1.8012684989429178e-05,
+      "loss": 0.2997,
+      "step": 1776
+    },
+    {
+      "epoch": 2.028522532800913,
+      "grad_norm": 0.24532737131358523,
+      "learning_rate": 1.7991543340380552e-05,
+      "loss": 0.2934,
+      "step": 1777
+    },
+    {
+      "epoch": 2.0296634341129494,
+      "grad_norm": 0.33354789343265456,
+      "learning_rate": 1.7970401691331926e-05,
+      "loss": 0.2949,
+      "step": 1778
+    },
+    {
+      "epoch": 2.030804335424986,
+      "grad_norm": 0.2463122343168031,
+      "learning_rate": 1.79492600422833e-05,
+      "loss": 0.2914,
+      "step": 1779
+    },
+    {
+      "epoch": 2.0319452367370223,
+      "grad_norm": 0.2723099135506961,
+      "learning_rate": 1.792811839323467e-05,
+      "loss": 0.3127,
+      "step": 1780
+    },
+    {
+      "epoch": 2.033086138049059,
+      "grad_norm": 0.2641672662067301,
+      "learning_rate": 1.7906976744186045e-05,
+      "loss": 0.2877,
+      "step": 1781
+    },
+    {
+      "epoch": 2.0342270393610953,
+      "grad_norm": 0.2590296004276685,
+      "learning_rate": 1.7885835095137422e-05,
+      "loss": 0.2825,
+      "step": 1782
+    },
+    {
+      "epoch": 2.0353679406731318,
+      "grad_norm": 0.23856790203525635,
+      "learning_rate": 1.7864693446088797e-05,
+      "loss": 0.2809,
+      "step": 1783
+    },
+    {
+      "epoch": 2.0365088419851682,
+      "grad_norm": 0.25679769430683064,
+      "learning_rate": 1.784355179704017e-05,
+      "loss": 0.2668,
+      "step": 1784
+    },
+    {
+      "epoch": 2.0376497432972047,
+      "grad_norm": 0.23357636555828296,
+      "learning_rate": 1.7822410147991545e-05,
+      "loss": 0.2934,
+      "step": 1785
+    },
+    {
+      "epoch": 2.038790644609241,
+      "grad_norm": 0.2635889031071584,
+      "learning_rate": 1.780126849894292e-05,
+      "loss": 0.3032,
+      "step": 1786
+    },
+    {
+      "epoch": 2.0399315459212777,
+      "grad_norm": 0.24980277535986198,
+      "learning_rate": 1.7780126849894293e-05,
+      "loss": 0.2977,
+      "step": 1787
+    },
+    {
+      "epoch": 2.041072447233314,
+      "grad_norm": 0.295692136350144,
+      "learning_rate": 1.7758985200845667e-05,
+      "loss": 0.2952,
+      "step": 1788
+    },
+    {
+      "epoch": 2.0422133485453506,
+      "grad_norm": 0.23935600190392897,
+      "learning_rate": 1.773784355179704e-05,
+      "loss": 0.2964,
+      "step": 1789
+    },
+    {
+      "epoch": 2.043354249857387,
+      "grad_norm": 0.30965446159470705,
+      "learning_rate": 1.7716701902748415e-05,
+      "loss": 0.2739,
+      "step": 1790
+    },
+    {
+      "epoch": 2.044495151169424,
+      "grad_norm": 0.24093476355063664,
+      "learning_rate": 1.769556025369979e-05,
+      "loss": 0.303,
+      "step": 1791
+    },
+    {
+      "epoch": 2.0456360524814605,
+      "grad_norm": 0.25136517438617395,
+      "learning_rate": 1.7674418604651163e-05,
+      "loss": 0.2871,
+      "step": 1792
+    },
+    {
+      "epoch": 2.046776953793497,
+      "grad_norm": 0.2751162964138566,
+      "learning_rate": 1.7653276955602537e-05,
+      "loss": 0.2935,
+      "step": 1793
+    },
+    {
+      "epoch": 2.0479178551055335,
+      "grad_norm": 0.26971281309270306,
+      "learning_rate": 1.763213530655391e-05,
+      "loss": 0.2928,
+      "step": 1794
+    },
+    {
+      "epoch": 2.04905875641757,
+      "grad_norm": 0.2312877390629995,
+      "learning_rate": 1.7610993657505285e-05,
+      "loss": 0.2898,
+      "step": 1795
+    },
+    {
+      "epoch": 2.0501996577296064,
+      "grad_norm": 0.2712479419818717,
+      "learning_rate": 1.7589852008456663e-05,
+      "loss": 0.2859,
+      "step": 1796
+    },
+    {
+      "epoch": 2.051340559041643,
+      "grad_norm": 0.310067258301682,
+      "learning_rate": 1.7568710359408037e-05,
+      "loss": 0.326,
+      "step": 1797
+    },
+    {
+      "epoch": 2.0524814603536794,
+      "grad_norm": 0.2526105961518339,
+      "learning_rate": 1.7547568710359408e-05,
+      "loss": 0.3056,
+      "step": 1798
+    },
+    {
+      "epoch": 2.053622361665716,
+      "grad_norm": 0.2579468875301229,
+      "learning_rate": 1.7526427061310782e-05,
+      "loss": 0.3039,
+      "step": 1799
+    },
+    {
+      "epoch": 2.0547632629777524,
+      "grad_norm": 0.24865148191408976,
+      "learning_rate": 1.7505285412262156e-05,
+      "loss": 0.299,
+      "step": 1800
+    },
+    {
+      "epoch": 2.055904164289789,
+      "grad_norm": 0.27482713881750265,
+      "learning_rate": 1.748414376321353e-05,
+      "loss": 0.2828,
+      "step": 1801
+    },
+    {
+      "epoch": 2.0570450656018253,
+      "grad_norm": 0.23944885286193957,
+      "learning_rate": 1.7463002114164907e-05,
+      "loss": 0.2802,
+      "step": 1802
+    },
+    {
+      "epoch": 2.058185966913862,
+      "grad_norm": 0.23387539270157587,
+      "learning_rate": 1.744186046511628e-05,
+      "loss": 0.2833,
+      "step": 1803
+    },
+    {
+      "epoch": 2.0593268682258983,
+      "grad_norm": 0.2507361486702264,
+      "learning_rate": 1.7420718816067655e-05,
+      "loss": 0.2912,
+      "step": 1804
+    },
+    {
+      "epoch": 2.0604677695379348,
+      "grad_norm": 0.27028591396300616,
+      "learning_rate": 1.739957716701903e-05,
+      "loss": 0.2753,
+      "step": 1805
+    },
+    {
+      "epoch": 2.0616086708499717,
+      "grad_norm": 0.2619452920223977,
+      "learning_rate": 1.73784355179704e-05,
+      "loss": 0.2842,
+      "step": 1806
+    },
+    {
+      "epoch": 2.062749572162008,
+      "grad_norm": 0.2607113990408732,
+      "learning_rate": 1.7357293868921774e-05,
+      "loss": 0.292,
+      "step": 1807
+    },
+    {
+      "epoch": 2.0638904734740446,
+      "grad_norm": 0.30100822829080276,
+      "learning_rate": 1.7336152219873152e-05,
+      "loss": 0.3064,
+      "step": 1808
+    },
+    {
+      "epoch": 2.065031374786081,
+      "grad_norm": 0.2568094886960586,
+      "learning_rate": 1.7315010570824526e-05,
+      "loss": 0.2769,
+      "step": 1809
+    },
+    {
+      "epoch": 2.0661722760981176,
+      "grad_norm": 0.2668816139835237,
+      "learning_rate": 1.72938689217759e-05,
+      "loss": 0.2893,
+      "step": 1810
+    },
+    {
+      "epoch": 2.067313177410154,
+      "grad_norm": 0.275437215316109,
+      "learning_rate": 1.7272727272727274e-05,
+      "loss": 0.2889,
+      "step": 1811
+    },
+    {
+      "epoch": 2.0684540787221906,
+      "grad_norm": 0.2416657979758825,
+      "learning_rate": 1.7251585623678648e-05,
+      "loss": 0.2943,
+      "step": 1812
+    },
+    {
+      "epoch": 2.069594980034227,
+      "grad_norm": 0.28532369815909503,
+      "learning_rate": 1.7230443974630022e-05,
+      "loss": 0.2953,
+      "step": 1813
+    },
+    {
+      "epoch": 2.0707358813462635,
+      "grad_norm": 0.23963135477526373,
+      "learning_rate": 1.7209302325581396e-05,
+      "loss": 0.2967,
+      "step": 1814
+    },
+    {
+      "epoch": 2.0718767826583,
+      "grad_norm": 0.27869751780119484,
+      "learning_rate": 1.718816067653277e-05,
+      "loss": 0.2963,
+      "step": 1815
+    },
+    {
+      "epoch": 2.0730176839703365,
+      "grad_norm": 0.2662336852675231,
+      "learning_rate": 1.7167019027484144e-05,
+      "loss": 0.2972,
+      "step": 1816
+    },
+    {
+      "epoch": 2.074158585282373,
+      "grad_norm": 0.2588271856116877,
+      "learning_rate": 1.714587737843552e-05,
+      "loss": 0.2945,
+      "step": 1817
+    },
+    {
+      "epoch": 2.0752994865944094,
+      "grad_norm": 0.2862172344229437,
+      "learning_rate": 1.7124735729386892e-05,
+      "loss": 0.2994,
+      "step": 1818
+    },
+    {
+      "epoch": 2.076440387906446,
+      "grad_norm": 0.24831537941992662,
+      "learning_rate": 1.7103594080338267e-05,
+      "loss": 0.2834,
+      "step": 1819
+    },
+    {
+      "epoch": 2.0775812892184824,
+      "grad_norm": 0.22516742916788882,
+      "learning_rate": 1.708245243128964e-05,
+      "loss": 0.2971,
+      "step": 1820
+    },
+    {
+      "epoch": 2.0787221905305193,
+      "grad_norm": 0.28169017677827185,
+      "learning_rate": 1.7061310782241015e-05,
+      "loss": 0.2861,
+      "step": 1821
+    },
+    {
+      "epoch": 2.079863091842556,
+      "grad_norm": 0.23806396697312238,
+      "learning_rate": 1.7040169133192392e-05,
+      "loss": 0.2834,
+      "step": 1822
+    },
+    {
+      "epoch": 2.0810039931545923,
+      "grad_norm": 0.2555243399385284,
+      "learning_rate": 1.7019027484143766e-05,
+      "loss": 0.304,
+      "step": 1823
+    },
+    {
+      "epoch": 2.0821448944666288,
+      "grad_norm": 0.2578938428402628,
+      "learning_rate": 1.6997885835095137e-05,
+      "loss": 0.2895,
+      "step": 1824
+    },
+    {
+      "epoch": 2.0832857957786652,
+      "grad_norm": 0.23149122222114324,
+      "learning_rate": 1.697674418604651e-05,
+      "loss": 0.2806,
+      "step": 1825
+    },
+    {
+      "epoch": 2.0844266970907017,
+      "grad_norm": 0.25122516172954884,
+      "learning_rate": 1.6955602536997885e-05,
+      "loss": 0.2906,
+      "step": 1826
+    },
+    {
+      "epoch": 2.085567598402738,
+      "grad_norm": 0.22496719907018478,
+      "learning_rate": 1.693446088794926e-05,
+      "loss": 0.268,
+      "step": 1827
+    },
+    {
+      "epoch": 2.0867084997147747,
+      "grad_norm": 0.24917040442115526,
+      "learning_rate": 1.6913319238900637e-05,
+      "loss": 0.3072,
+      "step": 1828
+    },
+    {
+      "epoch": 2.087849401026811,
+      "grad_norm": 0.2461073133402563,
+      "learning_rate": 1.689217758985201e-05,
+      "loss": 0.3023,
+      "step": 1829
+    },
+    {
+      "epoch": 2.0889903023388476,
+      "grad_norm": 0.24067405395346075,
+      "learning_rate": 1.6871035940803385e-05,
+      "loss": 0.2873,
+      "step": 1830
+    },
+    {
+      "epoch": 2.090131203650884,
+      "grad_norm": 0.23945458238615078,
+      "learning_rate": 1.684989429175476e-05,
+      "loss": 0.2902,
+      "step": 1831
+    },
+    {
+      "epoch": 2.0912721049629206,
+      "grad_norm": 0.2382053179878838,
+      "learning_rate": 1.6828752642706133e-05,
+      "loss": 0.3078,
+      "step": 1832
+    },
+    {
+      "epoch": 2.092413006274957,
+      "grad_norm": 0.23429092759412315,
+      "learning_rate": 1.6807610993657504e-05,
+      "loss": 0.3189,
+      "step": 1833
+    },
+    {
+      "epoch": 2.0935539075869936,
+      "grad_norm": 0.24756030196541584,
+      "learning_rate": 1.678646934460888e-05,
+      "loss": 0.3011,
+      "step": 1834
+    },
+    {
+      "epoch": 2.09469480889903,
+      "grad_norm": 0.25175437537430273,
+      "learning_rate": 1.6765327695560255e-05,
+      "loss": 0.2838,
+      "step": 1835
+    },
+    {
+      "epoch": 2.095835710211067,
+      "grad_norm": 0.23741898688224317,
+      "learning_rate": 1.674418604651163e-05,
+      "loss": 0.2942,
+      "step": 1836
+    },
+    {
+      "epoch": 2.0969766115231034,
+      "grad_norm": 0.24003294023247418,
+      "learning_rate": 1.6723044397463003e-05,
+      "loss": 0.2993,
+      "step": 1837
+    },
+    {
+      "epoch": 2.09811751283514,
+      "grad_norm": 0.23478472744981252,
+      "learning_rate": 1.6701902748414377e-05,
+      "loss": 0.2907,
+      "step": 1838
+    },
+    {
+      "epoch": 2.0992584141471764,
+      "grad_norm": 0.23365403996013454,
+      "learning_rate": 1.668076109936575e-05,
+      "loss": 0.2772,
+      "step": 1839
+    },
+    {
+      "epoch": 2.100399315459213,
+      "grad_norm": 0.24999253956880316,
+      "learning_rate": 1.6659619450317125e-05,
+      "loss": 0.3015,
+      "step": 1840
+    },
+    {
+      "epoch": 2.1015402167712494,
+      "grad_norm": 0.23580829687884494,
+      "learning_rate": 1.66384778012685e-05,
+      "loss": 0.289,
+      "step": 1841
+    },
+    {
+      "epoch": 2.102681118083286,
+      "grad_norm": 0.2605491108678695,
+      "learning_rate": 1.6617336152219874e-05,
+      "loss": 0.2987,
+      "step": 1842
+    },
+    {
+      "epoch": 2.1038220193953223,
+      "grad_norm": 0.24526608405712047,
+      "learning_rate": 1.6596194503171248e-05,
+      "loss": 0.2911,
+      "step": 1843
+    },
+    {
+      "epoch": 2.104962920707359,
+      "grad_norm": 0.23507549939657288,
+      "learning_rate": 1.6575052854122622e-05,
+      "loss": 0.2935,
+      "step": 1844
+    },
+    {
+      "epoch": 2.1061038220193953,
+      "grad_norm": 0.23017170444286497,
+      "learning_rate": 1.6553911205073996e-05,
+      "loss": 0.293,
+      "step": 1845
+    },
+    {
+      "epoch": 2.1072447233314318,
+      "grad_norm": 0.23209560845918595,
+      "learning_rate": 1.653276955602537e-05,
+      "loss": 0.2893,
+      "step": 1846
+    },
+    {
+      "epoch": 2.1083856246434682,
+      "grad_norm": 0.23595690639390743,
+      "learning_rate": 1.6511627906976744e-05,
+      "loss": 0.2805,
+      "step": 1847
+    },
+    {
+      "epoch": 2.1095265259555047,
+      "grad_norm": 0.21983931290432804,
+      "learning_rate": 1.649048625792812e-05,
+      "loss": 0.2902,
+      "step": 1848
+    },
+    {
+      "epoch": 2.110667427267541,
+      "grad_norm": 0.21796041933539737,
+      "learning_rate": 1.6469344608879496e-05,
+      "loss": 0.2823,
+      "step": 1849
+    },
+    {
+      "epoch": 2.1118083285795777,
+      "grad_norm": 0.24121642344375474,
+      "learning_rate": 1.6448202959830866e-05,
+      "loss": 0.2942,
+      "step": 1850
+    },
+    {
+      "epoch": 2.1129492298916146,
+      "grad_norm": 0.24186049324723294,
+      "learning_rate": 1.642706131078224e-05,
+      "loss": 0.2868,
+      "step": 1851
+    },
+    {
+      "epoch": 2.114090131203651,
+      "grad_norm": 0.22393730692732144,
+      "learning_rate": 1.6405919661733614e-05,
+      "loss": 0.3086,
+      "step": 1852
+    },
+    {
+      "epoch": 2.1152310325156876,
+      "grad_norm": 0.2617315914853435,
+      "learning_rate": 1.638477801268499e-05,
+      "loss": 0.2875,
+      "step": 1853
+    },
+    {
+      "epoch": 2.116371933827724,
+      "grad_norm": 0.23203734158676664,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 0.2716,
+      "step": 1854
+    },
+    {
+      "epoch": 2.1175128351397605,
+      "grad_norm": 0.286581008935328,
+      "learning_rate": 1.634249471458774e-05,
+      "loss": 0.2991,
+      "step": 1855
+    },
+    {
+      "epoch": 2.118653736451797,
+      "grad_norm": 0.24041373152409004,
+      "learning_rate": 1.6321353065539114e-05,
+      "loss": 0.292,
+      "step": 1856
+    },
+    {
+      "epoch": 2.1197946377638335,
+      "grad_norm": 0.29656734889505126,
+      "learning_rate": 1.6300211416490488e-05,
+      "loss": 0.2864,
+      "step": 1857
+    },
+    {
+      "epoch": 2.12093553907587,
+      "grad_norm": 0.25003488021762793,
+      "learning_rate": 1.6279069767441862e-05,
+      "loss": 0.3022,
+      "step": 1858
+    },
+    {
+      "epoch": 2.1220764403879064,
+      "grad_norm": 0.22864083847140418,
+      "learning_rate": 1.6257928118393233e-05,
+      "loss": 0.2952,
+      "step": 1859
+    },
+    {
+      "epoch": 2.123217341699943,
+      "grad_norm": 0.26581686925779857,
+      "learning_rate": 1.623678646934461e-05,
+      "loss": 0.2856,
+      "step": 1860
+    },
+    {
+      "epoch": 2.1243582430119794,
+      "grad_norm": 0.26086638331564543,
+      "learning_rate": 1.6215644820295984e-05,
+      "loss": 0.2935,
+      "step": 1861
+    },
+    {
+      "epoch": 2.125499144324016,
+      "grad_norm": 0.23337841700522194,
+      "learning_rate": 1.619450317124736e-05,
+      "loss": 0.2852,
+      "step": 1862
+    },
+    {
+      "epoch": 2.1266400456360524,
+      "grad_norm": 0.2769523730953264,
+      "learning_rate": 1.6173361522198733e-05,
+      "loss": 0.2781,
+      "step": 1863
+    },
+    {
+      "epoch": 2.127780946948089,
+      "grad_norm": 0.2718124892758348,
+      "learning_rate": 1.6152219873150107e-05,
+      "loss": 0.2962,
+      "step": 1864
+    },
+    {
+      "epoch": 2.1289218482601253,
+      "grad_norm": 0.22143722786026926,
+      "learning_rate": 1.613107822410148e-05,
+      "loss": 0.3111,
+      "step": 1865
+    },
+    {
+      "epoch": 2.1300627495721622,
+      "grad_norm": 0.27102920563649024,
+      "learning_rate": 1.6109936575052855e-05,
+      "loss": 0.2935,
+      "step": 1866
+    },
+    {
+      "epoch": 2.1312036508841987,
+      "grad_norm": 0.27336499438269957,
+      "learning_rate": 1.608879492600423e-05,
+      "loss": 0.2993,
+      "step": 1867
+    },
+    {
+      "epoch": 2.132344552196235,
+      "grad_norm": 0.2178279407621219,
+      "learning_rate": 1.6067653276955603e-05,
+      "loss": 0.2765,
+      "step": 1868
+    },
+    {
+      "epoch": 2.1334854535082717,
+      "grad_norm": 0.23210951727420406,
+      "learning_rate": 1.6046511627906977e-05,
+      "loss": 0.3076,
+      "step": 1869
+    },
+    {
+      "epoch": 2.134626354820308,
+      "grad_norm": 0.24938828827661355,
+      "learning_rate": 1.602536997885835e-05,
+      "loss": 0.2999,
+      "step": 1870
+    },
+    {
+      "epoch": 2.1357672561323446,
+      "grad_norm": 0.28143320894936363,
+      "learning_rate": 1.6004228329809725e-05,
+      "loss": 0.2875,
+      "step": 1871
+    },
+    {
+      "epoch": 2.136908157444381,
+      "grad_norm": 0.2363884317835549,
+      "learning_rate": 1.59830866807611e-05,
+      "loss": 0.2818,
+      "step": 1872
+    },
+    {
+      "epoch": 2.1380490587564176,
+      "grad_norm": 0.23142000312910235,
+      "learning_rate": 1.5961945031712473e-05,
+      "loss": 0.2886,
+      "step": 1873
+    },
+    {
+      "epoch": 2.139189960068454,
+      "grad_norm": 0.2967980449947715,
+      "learning_rate": 1.594080338266385e-05,
+      "loss": 0.287,
+      "step": 1874
+    },
+    {
+      "epoch": 2.1403308613804906,
+      "grad_norm": 0.2523233122910311,
+      "learning_rate": 1.5919661733615225e-05,
+      "loss": 0.3046,
+      "step": 1875
+    },
+    {
+      "epoch": 2.141471762692527,
+      "grad_norm": 0.253248683782715,
+      "learning_rate": 1.58985200845666e-05,
+      "loss": 0.299,
+      "step": 1876
+    },
+    {
+      "epoch": 2.1426126640045635,
+      "grad_norm": 0.25140721281143225,
+      "learning_rate": 1.587737843551797e-05,
+      "loss": 0.3008,
+      "step": 1877
+    },
+    {
+      "epoch": 2.1437535653166,
+      "grad_norm": 0.2661164691317505,
+      "learning_rate": 1.5856236786469344e-05,
+      "loss": 0.2956,
+      "step": 1878
+    },
+    {
+      "epoch": 2.1448944666286365,
+      "grad_norm": 0.23762095264944688,
+      "learning_rate": 1.5835095137420718e-05,
+      "loss": 0.2934,
+      "step": 1879
+    },
+    {
+      "epoch": 2.146035367940673,
+      "grad_norm": 0.2699960211507811,
+      "learning_rate": 1.5813953488372095e-05,
+      "loss": 0.3012,
+      "step": 1880
+    },
+    {
+      "epoch": 2.1471762692527094,
+      "grad_norm": 0.25416412797833793,
+      "learning_rate": 1.579281183932347e-05,
+      "loss": 0.2997,
+      "step": 1881
+    },
+    {
+      "epoch": 2.148317170564746,
+      "grad_norm": 0.2603044805803768,
+      "learning_rate": 1.5771670190274843e-05,
+      "loss": 0.311,
+      "step": 1882
+    },
+    {
+      "epoch": 2.149458071876783,
+      "grad_norm": 0.24861515965118985,
+      "learning_rate": 1.5750528541226217e-05,
+      "loss": 0.2749,
+      "step": 1883
+    },
+    {
+      "epoch": 2.1505989731888193,
+      "grad_norm": 0.2266673641657746,
+      "learning_rate": 1.572938689217759e-05,
+      "loss": 0.2851,
+      "step": 1884
+    },
+    {
+      "epoch": 2.151739874500856,
+      "grad_norm": 0.2542824263685328,
+      "learning_rate": 1.5708245243128966e-05,
+      "loss": 0.2778,
+      "step": 1885
+    },
+    {
+      "epoch": 2.1528807758128923,
+      "grad_norm": 0.23241437414135388,
+      "learning_rate": 1.5687103594080336e-05,
+      "loss": 0.3021,
+      "step": 1886
+    },
+    {
+      "epoch": 2.1540216771249288,
+      "grad_norm": 0.2439766144777705,
+      "learning_rate": 1.5665961945031714e-05,
+      "loss": 0.2918,
+      "step": 1887
+    },
+    {
+      "epoch": 2.1551625784369652,
+      "grad_norm": 0.22410145683828062,
+      "learning_rate": 1.5644820295983088e-05,
+      "loss": 0.2934,
+      "step": 1888
+    },
+    {
+      "epoch": 2.1563034797490017,
+      "grad_norm": 0.21422379309473735,
+      "learning_rate": 1.5623678646934462e-05,
+      "loss": 0.2804,
+      "step": 1889
+    },
+    {
+      "epoch": 2.157444381061038,
+      "grad_norm": 0.23352117964227453,
+      "learning_rate": 1.5602536997885836e-05,
+      "loss": 0.2767,
+      "step": 1890
+    },
+    {
+      "epoch": 2.1585852823730747,
+      "grad_norm": 0.2202691337690104,
+      "learning_rate": 1.558139534883721e-05,
+      "loss": 0.2938,
+      "step": 1891
+    },
+    {
+      "epoch": 2.159726183685111,
+      "grad_norm": 0.23851514096715734,
+      "learning_rate": 1.5560253699788584e-05,
+      "loss": 0.2825,
+      "step": 1892
+    },
+    {
+      "epoch": 2.1608670849971476,
+      "grad_norm": 0.24720474182268987,
+      "learning_rate": 1.5539112050739958e-05,
+      "loss": 0.2968,
+      "step": 1893
+    },
+    {
+      "epoch": 2.162007986309184,
+      "grad_norm": 0.23232992455363097,
+      "learning_rate": 1.5517970401691332e-05,
+      "loss": 0.2905,
+      "step": 1894
+    },
+    {
+      "epoch": 2.1631488876212206,
+      "grad_norm": 0.26117342795271153,
+      "learning_rate": 1.5496828752642706e-05,
+      "loss": 0.2905,
+      "step": 1895
+    },
+    {
+      "epoch": 2.164289788933257,
+      "grad_norm": 0.22523236320973763,
+      "learning_rate": 1.547568710359408e-05,
+      "loss": 0.2862,
+      "step": 1896
+    },
+    {
+      "epoch": 2.1654306902452936,
+      "grad_norm": 0.2516240346631564,
+      "learning_rate": 1.5454545454545454e-05,
+      "loss": 0.3104,
+      "step": 1897
+    },
+    {
+      "epoch": 2.1665715915573305,
+      "grad_norm": 0.23518703025428958,
+      "learning_rate": 1.543340380549683e-05,
+      "loss": 0.3012,
+      "step": 1898
+    },
+    {
+      "epoch": 2.167712492869367,
+      "grad_norm": 0.2373995553413098,
+      "learning_rate": 1.5412262156448203e-05,
+      "loss": 0.2817,
+      "step": 1899
+    },
+    {
+      "epoch": 2.1688533941814034,
+      "grad_norm": 0.2215128863227245,
+      "learning_rate": 1.539112050739958e-05,
+      "loss": 0.3211,
+      "step": 1900
+    },
+    {
+      "epoch": 2.16999429549344,
+      "grad_norm": 0.24566821893965865,
+      "learning_rate": 1.5369978858350954e-05,
+      "loss": 0.2903,
+      "step": 1901
+    },
+    {
+      "epoch": 2.1711351968054764,
+      "grad_norm": 0.23907763184389938,
+      "learning_rate": 1.5348837209302328e-05,
+      "loss": 0.2917,
+      "step": 1902
+    },
+    {
+      "epoch": 2.172276098117513,
+      "grad_norm": 0.24613813251538924,
+      "learning_rate": 1.53276955602537e-05,
+      "loss": 0.2983,
+      "step": 1903
+    },
+    {
+      "epoch": 2.1734169994295494,
+      "grad_norm": 0.27660756080669346,
+      "learning_rate": 1.5306553911205073e-05,
+      "loss": 0.3074,
+      "step": 1904
+    },
+    {
+      "epoch": 2.174557900741586,
+      "grad_norm": 0.24073203570701351,
+      "learning_rate": 1.5285412262156447e-05,
+      "loss": 0.3047,
+      "step": 1905
+    },
+    {
+      "epoch": 2.1756988020536223,
+      "grad_norm": 0.24942911547741597,
+      "learning_rate": 1.5264270613107824e-05,
+      "loss": 0.2895,
+      "step": 1906
+    },
+    {
+      "epoch": 2.176839703365659,
+      "grad_norm": 0.2634505861930869,
+      "learning_rate": 1.5243128964059197e-05,
+      "loss": 0.2874,
+      "step": 1907
+    },
+    {
+      "epoch": 2.1779806046776953,
+      "grad_norm": 0.25700128710560166,
+      "learning_rate": 1.5221987315010573e-05,
+      "loss": 0.2921,
+      "step": 1908
+    },
+    {
+      "epoch": 2.1791215059897318,
+      "grad_norm": 0.2504882028942424,
+      "learning_rate": 1.5200845665961947e-05,
+      "loss": 0.2945,
+      "step": 1909
+    },
+    {
+      "epoch": 2.1802624073017682,
+      "grad_norm": 0.25717922096384405,
+      "learning_rate": 1.517970401691332e-05,
+      "loss": 0.2996,
+      "step": 1910
+    },
+    {
+      "epoch": 2.1814033086138047,
+      "grad_norm": 0.24086404886837165,
+      "learning_rate": 1.5158562367864695e-05,
+      "loss": 0.3071,
+      "step": 1911
+    },
+    {
+      "epoch": 2.182544209925841,
+      "grad_norm": 0.25023284208061214,
+      "learning_rate": 1.5137420718816067e-05,
+      "loss": 0.2834,
+      "step": 1912
+    },
+    {
+      "epoch": 2.183685111237878,
+      "grad_norm": 0.24495771528295845,
+      "learning_rate": 1.5116279069767441e-05,
+      "loss": 0.2846,
+      "step": 1913
+    },
+    {
+      "epoch": 2.1848260125499146,
+      "grad_norm": 0.23837446613226718,
+      "learning_rate": 1.5095137420718815e-05,
+      "loss": 0.2985,
+      "step": 1914
+    },
+    {
+      "epoch": 2.185966913861951,
+      "grad_norm": 0.25371400523552884,
+      "learning_rate": 1.5073995771670191e-05,
+      "loss": 0.2935,
+      "step": 1915
+    },
+    {
+      "epoch": 2.1871078151739876,
+      "grad_norm": 0.22958447884607433,
+      "learning_rate": 1.5052854122621565e-05,
+      "loss": 0.3015,
+      "step": 1916
+    },
+    {
+      "epoch": 2.188248716486024,
+      "grad_norm": 0.24419362474562023,
+      "learning_rate": 1.503171247357294e-05,
+      "loss": 0.2794,
+      "step": 1917
+    },
+    {
+      "epoch": 2.1893896177980605,
+      "grad_norm": 0.2482930512404429,
+      "learning_rate": 1.5010570824524315e-05,
+      "loss": 0.2881,
+      "step": 1918
+    },
+    {
+      "epoch": 2.190530519110097,
+      "grad_norm": 0.2518793397866365,
+      "learning_rate": 1.4989429175475689e-05,
+      "loss": 0.3022,
+      "step": 1919
+    },
+    {
+      "epoch": 2.1916714204221335,
+      "grad_norm": 0.23310730322180817,
+      "learning_rate": 1.4968287526427063e-05,
+      "loss": 0.3002,
+      "step": 1920
+    },
+    {
+      "epoch": 2.19281232173417,
+      "grad_norm": 0.24918138942890583,
+      "learning_rate": 1.4947145877378436e-05,
+      "loss": 0.3062,
+      "step": 1921
+    },
+    {
+      "epoch": 2.1939532230462064,
+      "grad_norm": 0.24663388333295883,
+      "learning_rate": 1.492600422832981e-05,
+      "loss": 0.2856,
+      "step": 1922
+    },
+    {
+      "epoch": 2.195094124358243,
+      "grad_norm": 0.2513589833120337,
+      "learning_rate": 1.4904862579281184e-05,
+      "loss": 0.3004,
+      "step": 1923
+    },
+    {
+      "epoch": 2.1962350256702794,
+      "grad_norm": 0.26311683859831203,
+      "learning_rate": 1.488372093023256e-05,
+      "loss": 0.2826,
+      "step": 1924
+    },
+    {
+      "epoch": 2.197375926982316,
+      "grad_norm": 0.22385424880185895,
+      "learning_rate": 1.4862579281183934e-05,
+      "loss": 0.2839,
+      "step": 1925
+    },
+    {
+      "epoch": 2.1985168282943524,
+      "grad_norm": 0.2159732264542219,
+      "learning_rate": 1.4841437632135308e-05,
+      "loss": 0.2959,
+      "step": 1926
+    },
+    {
+      "epoch": 2.199657729606389,
+      "grad_norm": 0.2162278305808399,
+      "learning_rate": 1.4820295983086682e-05,
+      "loss": 0.2891,
+      "step": 1927
+    },
+    {
+      "epoch": 2.2007986309184258,
+      "grad_norm": 0.24339108189160494,
+      "learning_rate": 1.4799154334038057e-05,
+      "loss": 0.3227,
+      "step": 1928
+    },
+    {
+      "epoch": 2.2019395322304622,
+      "grad_norm": 0.23716291707294404,
+      "learning_rate": 1.4778012684989432e-05,
+      "loss": 0.3111,
+      "step": 1929
+    },
+    {
+      "epoch": 2.2030804335424987,
+      "grad_norm": 0.2224056361602292,
+      "learning_rate": 1.4756871035940802e-05,
+      "loss": 0.2855,
+      "step": 1930
+    },
+    {
+      "epoch": 2.204221334854535,
+      "grad_norm": 0.22391025870440956,
+      "learning_rate": 1.4735729386892178e-05,
+      "loss": 0.2788,
+      "step": 1931
+    },
+    {
+      "epoch": 2.2053622361665717,
+      "grad_norm": 0.2044482302165388,
+      "learning_rate": 1.4714587737843552e-05,
+      "loss": 0.2935,
+      "step": 1932
+    },
+    {
+      "epoch": 2.206503137478608,
+      "grad_norm": 0.21986998112087408,
+      "learning_rate": 1.4693446088794926e-05,
+      "loss": 0.2848,
+      "step": 1933
+    },
+    {
+      "epoch": 2.2076440387906446,
+      "grad_norm": 0.23802896070685578,
+      "learning_rate": 1.4672304439746302e-05,
+      "loss": 0.2836,
+      "step": 1934
+    },
+    {
+      "epoch": 2.208784940102681,
+      "grad_norm": 0.22630388090138429,
+      "learning_rate": 1.4651162790697676e-05,
+      "loss": 0.3023,
+      "step": 1935
+    },
+    {
+      "epoch": 2.2099258414147176,
+      "grad_norm": 0.24389924649877973,
+      "learning_rate": 1.463002114164905e-05,
+      "loss": 0.266,
+      "step": 1936
+    },
+    {
+      "epoch": 2.211066742726754,
+      "grad_norm": 0.23067707246778427,
+      "learning_rate": 1.4608879492600424e-05,
+      "loss": 0.2921,
+      "step": 1937
+    },
+    {
+      "epoch": 2.2122076440387906,
+      "grad_norm": 0.24602098091273714,
+      "learning_rate": 1.4587737843551796e-05,
+      "loss": 0.2675,
+      "step": 1938
+    },
+    {
+      "epoch": 2.213348545350827,
+      "grad_norm": 0.22413934497837779,
+      "learning_rate": 1.456659619450317e-05,
+      "loss": 0.307,
+      "step": 1939
+    },
+    {
+      "epoch": 2.2144894466628635,
+      "grad_norm": 0.235601927149484,
+      "learning_rate": 1.4545454545454545e-05,
+      "loss": 0.3009,
+      "step": 1940
+    },
+    {
+      "epoch": 2.2156303479749,
+      "grad_norm": 0.23907967781378392,
+      "learning_rate": 1.452431289640592e-05,
+      "loss": 0.3082,
+      "step": 1941
+    },
+    {
+      "epoch": 2.2167712492869365,
+      "grad_norm": 0.24293175224705252,
+      "learning_rate": 1.4503171247357294e-05,
+      "loss": 0.2992,
+      "step": 1942
+    },
+    {
+      "epoch": 2.2179121505989734,
+      "grad_norm": 0.2552656926793362,
+      "learning_rate": 1.4482029598308669e-05,
+      "loss": 0.2947,
+      "step": 1943
+    },
+    {
+      "epoch": 2.21905305191101,
+      "grad_norm": 0.22778782163505898,
+      "learning_rate": 1.4460887949260044e-05,
+      "loss": 0.2996,
+      "step": 1944
+    },
+    {
+      "epoch": 2.2201939532230464,
+      "grad_norm": 0.23183846405236744,
+      "learning_rate": 1.4439746300211418e-05,
+      "loss": 0.2862,
+      "step": 1945
+    },
+    {
+      "epoch": 2.221334854535083,
+      "grad_norm": 0.2566338595557142,
+      "learning_rate": 1.4418604651162792e-05,
+      "loss": 0.2971,
+      "step": 1946
+    },
+    {
+      "epoch": 2.2224757558471193,
+      "grad_norm": 0.25108060135481586,
+      "learning_rate": 1.4397463002114165e-05,
+      "loss": 0.3039,
+      "step": 1947
+    },
+    {
+      "epoch": 2.223616657159156,
+      "grad_norm": 0.21702286457482928,
+      "learning_rate": 1.4376321353065539e-05,
+      "loss": 0.291,
+      "step": 1948
+    },
+    {
+      "epoch": 2.2247575584711923,
+      "grad_norm": 0.2281526549309785,
+      "learning_rate": 1.4355179704016913e-05,
+      "loss": 0.2869,
+      "step": 1949
+    },
+    {
+      "epoch": 2.2258984597832288,
+      "grad_norm": 0.2297634774448516,
+      "learning_rate": 1.4334038054968287e-05,
+      "loss": 0.2862,
+      "step": 1950
+    },
+    {
+      "epoch": 2.2270393610952652,
+      "grad_norm": 0.224251286343074,
+      "learning_rate": 1.4312896405919663e-05,
+      "loss": 0.2978,
+      "step": 1951
+    },
+    {
+      "epoch": 2.2281802624073017,
+      "grad_norm": 0.22372187364380428,
+      "learning_rate": 1.4291754756871037e-05,
+      "loss": 0.2829,
+      "step": 1952
+    },
+    {
+      "epoch": 2.229321163719338,
+      "grad_norm": 0.2582197182895316,
+      "learning_rate": 1.4270613107822411e-05,
+      "loss": 0.3214,
+      "step": 1953
+    },
+    {
+      "epoch": 2.2304620650313747,
+      "grad_norm": 0.22771491528249513,
+      "learning_rate": 1.4249471458773787e-05,
+      "loss": 0.2929,
+      "step": 1954
+    },
+    {
+      "epoch": 2.231602966343411,
+      "grad_norm": 0.22136843625693173,
+      "learning_rate": 1.422832980972516e-05,
+      "loss": 0.295,
+      "step": 1955
+    },
+    {
+      "epoch": 2.2327438676554476,
+      "grad_norm": 0.26906645511646826,
+      "learning_rate": 1.4207188160676532e-05,
+      "loss": 0.2893,
+      "step": 1956
+    },
+    {
+      "epoch": 2.233884768967484,
+      "grad_norm": 0.2215278537540451,
+      "learning_rate": 1.4186046511627907e-05,
+      "loss": 0.288,
+      "step": 1957
+    },
+    {
+      "epoch": 2.235025670279521,
+      "grad_norm": 0.23908694956372106,
+      "learning_rate": 1.4164904862579281e-05,
+      "loss": 0.2938,
+      "step": 1958
+    },
+    {
+      "epoch": 2.2361665715915575,
+      "grad_norm": 0.229217654612187,
+      "learning_rate": 1.4143763213530655e-05,
+      "loss": 0.2942,
+      "step": 1959
+    },
+    {
+      "epoch": 2.237307472903594,
+      "grad_norm": 0.23519915486491552,
+      "learning_rate": 1.412262156448203e-05,
+      "loss": 0.2873,
+      "step": 1960
+    },
+    {
+      "epoch": 2.2384483742156305,
+      "grad_norm": 0.2403014255500897,
+      "learning_rate": 1.4101479915433405e-05,
+      "loss": 0.3018,
+      "step": 1961
+    },
+    {
+      "epoch": 2.239589275527667,
+      "grad_norm": 0.22939951535091463,
+      "learning_rate": 1.408033826638478e-05,
+      "loss": 0.2882,
+      "step": 1962
+    },
+    {
+      "epoch": 2.2407301768397034,
+      "grad_norm": 0.24287675445152404,
+      "learning_rate": 1.4059196617336153e-05,
+      "loss": 0.3037,
+      "step": 1963
+    },
+    {
+      "epoch": 2.24187107815174,
+      "grad_norm": 0.2345981608383373,
+      "learning_rate": 1.403805496828753e-05,
+      "loss": 0.3208,
+      "step": 1964
+    },
+    {
+      "epoch": 2.2430119794637764,
+      "grad_norm": 0.24968054421359923,
+      "learning_rate": 1.40169133192389e-05,
+      "loss": 0.2937,
+      "step": 1965
+    },
+    {
+      "epoch": 2.244152880775813,
+      "grad_norm": 0.21193313579929754,
+      "learning_rate": 1.3995771670190274e-05,
+      "loss": 0.2972,
+      "step": 1966
+    },
+    {
+      "epoch": 2.2452937820878494,
+      "grad_norm": 0.238901730891219,
+      "learning_rate": 1.397463002114165e-05,
+      "loss": 0.2956,
+      "step": 1967
+    },
+    {
+      "epoch": 2.246434683399886,
+      "grad_norm": 0.2189680066373367,
+      "learning_rate": 1.3953488372093024e-05,
+      "loss": 0.2849,
+      "step": 1968
+    },
+    {
+      "epoch": 2.2475755847119223,
+      "grad_norm": 0.2129342166114009,
+      "learning_rate": 1.3932346723044398e-05,
+      "loss": 0.2988,
+      "step": 1969
+    },
+    {
+      "epoch": 2.248716486023959,
+      "grad_norm": 0.260367276734894,
+      "learning_rate": 1.3911205073995774e-05,
+      "loss": 0.3065,
+      "step": 1970
+    },
+    {
+      "epoch": 2.2498573873359953,
+      "grad_norm": 0.23116646083828635,
+      "learning_rate": 1.3890063424947148e-05,
+      "loss": 0.2802,
+      "step": 1971
+    },
+    {
+      "epoch": 2.2509982886480318,
+      "grad_norm": 0.23575182654734847,
+      "learning_rate": 1.3868921775898522e-05,
+      "loss": 0.3006,
+      "step": 1972
+    },
+    {
+      "epoch": 2.2521391899600687,
+      "grad_norm": 0.23514500823576498,
+      "learning_rate": 1.3847780126849896e-05,
+      "loss": 0.286,
+      "step": 1973
+    },
+    {
+      "epoch": 2.253280091272105,
+      "grad_norm": 0.2455896325382261,
+      "learning_rate": 1.3826638477801268e-05,
+      "loss": 0.2873,
+      "step": 1974
+    },
+    {
+      "epoch": 2.2544209925841416,
+      "grad_norm": 0.23967956911885113,
+      "learning_rate": 1.3805496828752642e-05,
+      "loss": 0.3073,
+      "step": 1975
+    },
+    {
+      "epoch": 2.255561893896178,
+      "grad_norm": 0.2586497727165531,
+      "learning_rate": 1.3784355179704016e-05,
+      "loss": 0.2953,
+      "step": 1976
+    },
+    {
+      "epoch": 2.2567027952082146,
+      "grad_norm": 0.22560983749066454,
+      "learning_rate": 1.3763213530655392e-05,
+      "loss": 0.282,
+      "step": 1977
+    },
+    {
+      "epoch": 2.257843696520251,
+      "grad_norm": 0.24888255862010317,
+      "learning_rate": 1.3742071881606766e-05,
+      "loss": 0.3008,
+      "step": 1978
+    },
+    {
+      "epoch": 2.2589845978322876,
+      "grad_norm": 0.24641463696901852,
+      "learning_rate": 1.372093023255814e-05,
+      "loss": 0.2981,
+      "step": 1979
+    },
+    {
+      "epoch": 2.260125499144324,
+      "grad_norm": 0.2161155293814851,
+      "learning_rate": 1.3699788583509516e-05,
+      "loss": 0.2926,
+      "step": 1980
+    },
+    {
+      "epoch": 2.2612664004563605,
+      "grad_norm": 0.24532498098160077,
+      "learning_rate": 1.367864693446089e-05,
+      "loss": 0.2832,
+      "step": 1981
+    },
+    {
+      "epoch": 2.262407301768397,
+      "grad_norm": 0.2015557674816523,
+      "learning_rate": 1.365750528541226e-05,
+      "loss": 0.2889,
+      "step": 1982
+    },
+    {
+      "epoch": 2.2635482030804335,
+      "grad_norm": 0.2763485972223904,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 0.3033,
+      "step": 1983
+    },
+    {
+      "epoch": 2.26468910439247,
+      "grad_norm": 0.23187776287773346,
+      "learning_rate": 1.361522198731501e-05,
+      "loss": 0.2805,
+      "step": 1984
+    },
+    {
+      "epoch": 2.2658300057045064,
+      "grad_norm": 0.21897157764795727,
+      "learning_rate": 1.3594080338266385e-05,
+      "loss": 0.2804,
+      "step": 1985
+    },
+    {
+      "epoch": 2.266970907016543,
+      "grad_norm": 0.2196787744861354,
+      "learning_rate": 1.3572938689217759e-05,
+      "loss": 0.2838,
+      "step": 1986
+    },
+    {
+      "epoch": 2.2681118083285794,
+      "grad_norm": 0.23196784986462152,
+      "learning_rate": 1.3551797040169135e-05,
+      "loss": 0.3006,
+      "step": 1987
+    },
+    {
+      "epoch": 2.2692527096406163,
+      "grad_norm": 0.2521901334329305,
+      "learning_rate": 1.3530655391120509e-05,
+      "loss": 0.2873,
+      "step": 1988
+    },
+    {
+      "epoch": 2.2703936109526524,
+      "grad_norm": 0.23247846103417055,
+      "learning_rate": 1.3509513742071883e-05,
+      "loss": 0.2866,
+      "step": 1989
+    },
+    {
+      "epoch": 2.2715345122646893,
+      "grad_norm": 0.269518441841022,
+      "learning_rate": 1.3488372093023258e-05,
+      "loss": 0.3135,
+      "step": 1990
+    },
+    {
+      "epoch": 2.2726754135767258,
+      "grad_norm": 0.23004727811310058,
+      "learning_rate": 1.3467230443974629e-05,
+      "loss": 0.2914,
+      "step": 1991
+    },
+    {
+      "epoch": 2.2738163148887622,
+      "grad_norm": 0.2483150678244816,
+      "learning_rate": 1.3446088794926003e-05,
+      "loss": 0.3105,
+      "step": 1992
+    },
+    {
+      "epoch": 2.2749572162007987,
+      "grad_norm": 0.24207883143871808,
+      "learning_rate": 1.3424947145877379e-05,
+      "loss": 0.292,
+      "step": 1993
+    },
+    {
+      "epoch": 2.276098117512835,
+      "grad_norm": 0.2292806033794932,
+      "learning_rate": 1.3403805496828753e-05,
+      "loss": 0.2896,
+      "step": 1994
+    },
+    {
+      "epoch": 2.2772390188248717,
+      "grad_norm": 0.21696103815825676,
+      "learning_rate": 1.3382663847780127e-05,
+      "loss": 0.3019,
+      "step": 1995
+    },
+    {
+      "epoch": 2.278379920136908,
+      "grad_norm": 0.24517163113794077,
+      "learning_rate": 1.3361522198731501e-05,
+      "loss": 0.3005,
+      "step": 1996
+    },
+    {
+      "epoch": 2.2795208214489446,
+      "grad_norm": 0.25085833283446013,
+      "learning_rate": 1.3340380549682877e-05,
+      "loss": 0.3081,
+      "step": 1997
+    },
+    {
+      "epoch": 2.280661722760981,
+      "grad_norm": 0.21047117649117664,
+      "learning_rate": 1.3319238900634251e-05,
+      "loss": 0.2961,
+      "step": 1998
+    },
+    {
+      "epoch": 2.2818026240730176,
+      "grad_norm": 0.23178127896840034,
+      "learning_rate": 1.3298097251585625e-05,
+      "loss": 0.2902,
+      "step": 1999
+    },
+    {
+      "epoch": 2.282943525385054,
+      "grad_norm": 0.23749065791689886,
+      "learning_rate": 1.3276955602536997e-05,
+      "loss": 0.2872,
+      "step": 2000
+    },
+    {
+      "epoch": 2.2840844266970906,
+      "grad_norm": 0.2075544823403873,
+      "learning_rate": 1.3255813953488372e-05,
+      "loss": 0.2819,
+      "step": 2001
+    },
+    {
+      "epoch": 2.285225328009127,
+      "grad_norm": 0.20462954299061928,
+      "learning_rate": 1.3234672304439746e-05,
+      "loss": 0.2948,
+      "step": 2002
+    },
+    {
+      "epoch": 2.286366229321164,
+      "grad_norm": 0.2365701652861174,
+      "learning_rate": 1.3213530655391121e-05,
+      "loss": 0.281,
+      "step": 2003
+    },
+    {
+      "epoch": 2.2875071306332,
+      "grad_norm": 0.23216020430008696,
+      "learning_rate": 1.3192389006342495e-05,
+      "loss": 0.2957,
+      "step": 2004
+    },
+    {
+      "epoch": 2.288648031945237,
+      "grad_norm": 0.22488722691116822,
+      "learning_rate": 1.317124735729387e-05,
+      "loss": 0.3038,
+      "step": 2005
+    },
+    {
+      "epoch": 2.2897889332572734,
+      "grad_norm": 0.2247997716859529,
+      "learning_rate": 1.3150105708245244e-05,
+      "loss": 0.2948,
+      "step": 2006
+    },
+    {
+      "epoch": 2.29092983456931,
+      "grad_norm": 0.23588924302825817,
+      "learning_rate": 1.312896405919662e-05,
+      "loss": 0.3015,
+      "step": 2007
+    },
+    {
+      "epoch": 2.2920707358813464,
+      "grad_norm": 0.203969359420377,
+      "learning_rate": 1.3107822410147993e-05,
+      "loss": 0.2833,
+      "step": 2008
+    },
+    {
+      "epoch": 2.293211637193383,
+      "grad_norm": 0.2235108319356959,
+      "learning_rate": 1.3086680761099366e-05,
+      "loss": 0.2984,
+      "step": 2009
+    },
+    {
+      "epoch": 2.2943525385054193,
+      "grad_norm": 0.23896468835409596,
+      "learning_rate": 1.306553911205074e-05,
+      "loss": 0.2864,
+      "step": 2010
+    },
+    {
+      "epoch": 2.295493439817456,
+      "grad_norm": 0.2212794132404757,
+      "learning_rate": 1.3044397463002114e-05,
+      "loss": 0.3006,
+      "step": 2011
+    },
+    {
+      "epoch": 2.2966343411294923,
+      "grad_norm": 0.23225833735132453,
+      "learning_rate": 1.3023255813953488e-05,
+      "loss": 0.2852,
+      "step": 2012
+    },
+    {
+      "epoch": 2.2977752424415288,
+      "grad_norm": 0.23082470670882826,
+      "learning_rate": 1.3002114164904864e-05,
+      "loss": 0.3054,
+      "step": 2013
+    },
+    {
+      "epoch": 2.2989161437535652,
+      "grad_norm": 0.2360677430344232,
+      "learning_rate": 1.2980972515856238e-05,
+      "loss": 0.2877,
+      "step": 2014
+    },
+    {
+      "epoch": 2.3000570450656017,
+      "grad_norm": 0.2404968090073572,
+      "learning_rate": 1.2959830866807612e-05,
+      "loss": 0.296,
+      "step": 2015
+    },
+    {
+      "epoch": 2.301197946377638,
+      "grad_norm": 0.24836315958275604,
+      "learning_rate": 1.2938689217758988e-05,
+      "loss": 0.2932,
+      "step": 2016
+    },
+    {
+      "epoch": 2.3023388476896747,
+      "grad_norm": 0.2337381742291952,
+      "learning_rate": 1.2917547568710358e-05,
+      "loss": 0.287,
+      "step": 2017
+    },
+    {
+      "epoch": 2.3034797490017116,
+      "grad_norm": 0.2232820057505206,
+      "learning_rate": 1.2896405919661732e-05,
+      "loss": 0.2992,
+      "step": 2018
+    },
+    {
+      "epoch": 2.3046206503137476,
+      "grad_norm": 0.21481193248947944,
+      "learning_rate": 1.2875264270613108e-05,
+      "loss": 0.2941,
+      "step": 2019
+    },
+    {
+      "epoch": 2.3057615516257846,
+      "grad_norm": 0.23287659442211797,
+      "learning_rate": 1.2854122621564482e-05,
+      "loss": 0.2888,
+      "step": 2020
+    },
+    {
+      "epoch": 2.306902452937821,
+      "grad_norm": 0.2441450826097555,
+      "learning_rate": 1.2832980972515856e-05,
+      "loss": 0.2881,
+      "step": 2021
+    },
+    {
+      "epoch": 2.3080433542498575,
+      "grad_norm": 0.21631900341301358,
+      "learning_rate": 1.281183932346723e-05,
+      "loss": 0.3031,
+      "step": 2022
+    },
+    {
+      "epoch": 2.309184255561894,
+      "grad_norm": 0.23685841465022145,
+      "learning_rate": 1.2790697674418606e-05,
+      "loss": 0.2889,
+      "step": 2023
+    },
+    {
+      "epoch": 2.3103251568739305,
+      "grad_norm": 0.24752960920348968,
+      "learning_rate": 1.276955602536998e-05,
+      "loss": 0.3048,
+      "step": 2024
+    },
+    {
+      "epoch": 2.311466058185967,
+      "grad_norm": 0.2141925809115063,
+      "learning_rate": 1.2748414376321354e-05,
+      "loss": 0.3126,
+      "step": 2025
+    },
+    {
+      "epoch": 2.3126069594980034,
+      "grad_norm": 0.23469096398654987,
+      "learning_rate": 1.2727272727272727e-05,
+      "loss": 0.3067,
+      "step": 2026
+    },
+    {
+      "epoch": 2.31374786081004,
+      "grad_norm": 0.24021289203570448,
+      "learning_rate": 1.27061310782241e-05,
+      "loss": 0.2849,
+      "step": 2027
+    },
+    {
+      "epoch": 2.3148887621220764,
+      "grad_norm": 0.2224644417413248,
+      "learning_rate": 1.2684989429175475e-05,
+      "loss": 0.2924,
+      "step": 2028
+    },
+    {
+      "epoch": 2.316029663434113,
+      "grad_norm": 0.23049113832054982,
+      "learning_rate": 1.266384778012685e-05,
+      "loss": 0.2971,
+      "step": 2029
+    },
+    {
+      "epoch": 2.3171705647461494,
+      "grad_norm": 0.2340970748955576,
+      "learning_rate": 1.2642706131078225e-05,
+      "loss": 0.2934,
+      "step": 2030
+    },
+    {
+      "epoch": 2.318311466058186,
+      "grad_norm": 0.22710680663083077,
+      "learning_rate": 1.2621564482029599e-05,
+      "loss": 0.3011,
+      "step": 2031
+    },
+    {
+      "epoch": 2.3194523673702223,
+      "grad_norm": 0.2214909866498558,
+      "learning_rate": 1.2600422832980973e-05,
+      "loss": 0.2922,
+      "step": 2032
+    },
+    {
+      "epoch": 2.320593268682259,
+      "grad_norm": 0.2370022891229573,
+      "learning_rate": 1.2579281183932349e-05,
+      "loss": 0.2952,
+      "step": 2033
+    },
+    {
+      "epoch": 2.3217341699942953,
+      "grad_norm": 0.23549458230786854,
+      "learning_rate": 1.2558139534883723e-05,
+      "loss": 0.3035,
+      "step": 2034
+    },
+    {
+      "epoch": 2.322875071306332,
+      "grad_norm": 0.21809602956352517,
+      "learning_rate": 1.2536997885835095e-05,
+      "loss": 0.3024,
+      "step": 2035
+    },
+    {
+      "epoch": 2.3240159726183687,
+      "grad_norm": 0.2166411089709946,
+      "learning_rate": 1.251585623678647e-05,
+      "loss": 0.296,
+      "step": 2036
+    },
+    {
+      "epoch": 2.325156873930405,
+      "grad_norm": 0.26463884719859104,
+      "learning_rate": 1.2494714587737843e-05,
+      "loss": 0.2671,
+      "step": 2037
+    },
+    {
+      "epoch": 2.3262977752424416,
+      "grad_norm": 0.22532754670856348,
+      "learning_rate": 1.2473572938689217e-05,
+      "loss": 0.3002,
+      "step": 2038
+    },
+    {
+      "epoch": 2.327438676554478,
+      "grad_norm": 0.2407040758208473,
+      "learning_rate": 1.2452431289640593e-05,
+      "loss": 0.2984,
+      "step": 2039
+    },
+    {
+      "epoch": 2.3285795778665146,
+      "grad_norm": 0.26048354489435793,
+      "learning_rate": 1.2431289640591967e-05,
+      "loss": 0.2936,
+      "step": 2040
+    },
+    {
+      "epoch": 2.329720479178551,
+      "grad_norm": 0.24546852772138006,
+      "learning_rate": 1.2410147991543341e-05,
+      "loss": 0.2935,
+      "step": 2041
+    },
+    {
+      "epoch": 2.3308613804905876,
+      "grad_norm": 0.2884947521367451,
+      "learning_rate": 1.2389006342494715e-05,
+      "loss": 0.2987,
+      "step": 2042
+    },
+    {
+      "epoch": 2.332002281802624,
+      "grad_norm": 0.2147628745332112,
+      "learning_rate": 1.236786469344609e-05,
+      "loss": 0.2902,
+      "step": 2043
+    },
+    {
+      "epoch": 2.3331431831146605,
+      "grad_norm": 0.25053785188898486,
+      "learning_rate": 1.2346723044397463e-05,
+      "loss": 0.2936,
+      "step": 2044
+    },
+    {
+      "epoch": 2.334284084426697,
+      "grad_norm": 0.2579239729203553,
+      "learning_rate": 1.2325581395348838e-05,
+      "loss": 0.2935,
+      "step": 2045
+    },
+    {
+      "epoch": 2.3354249857387335,
+      "grad_norm": 0.21423908107360926,
+      "learning_rate": 1.2304439746300212e-05,
+      "loss": 0.296,
+      "step": 2046
+    },
+    {
+      "epoch": 2.33656588705077,
+      "grad_norm": 0.25558415641280685,
+      "learning_rate": 1.2283298097251586e-05,
+      "loss": 0.293,
+      "step": 2047
+    },
+    {
+      "epoch": 2.3377067883628064,
+      "grad_norm": 0.243119900240472,
+      "learning_rate": 1.226215644820296e-05,
+      "loss": 0.2909,
+      "step": 2048
+    },
+    {
+      "epoch": 2.338847689674843,
+      "grad_norm": 0.2328624664465957,
+      "learning_rate": 1.2241014799154336e-05,
+      "loss": 0.2824,
+      "step": 2049
+    },
+    {
+      "epoch": 2.33998859098688,
+      "grad_norm": 0.2152655139207514,
+      "learning_rate": 1.2219873150105708e-05,
+      "loss": 0.294,
+      "step": 2050
+    },
+    {
+      "epoch": 2.3411294922989163,
+      "grad_norm": 0.2298995263704241,
+      "learning_rate": 1.2198731501057082e-05,
+      "loss": 0.3064,
+      "step": 2051
+    },
+    {
+      "epoch": 2.342270393610953,
+      "grad_norm": 0.2313471598713492,
+      "learning_rate": 1.2177589852008458e-05,
+      "loss": 0.2936,
+      "step": 2052
+    },
+    {
+      "epoch": 2.3434112949229893,
+      "grad_norm": 0.23453565544223756,
+      "learning_rate": 1.2156448202959832e-05,
+      "loss": 0.2889,
+      "step": 2053
+    },
+    {
+      "epoch": 2.3445521962350258,
+      "grad_norm": 0.21356838246444326,
+      "learning_rate": 1.2135306553911206e-05,
+      "loss": 0.2835,
+      "step": 2054
+    },
+    {
+      "epoch": 2.3456930975470622,
+      "grad_norm": 0.20309860095085247,
+      "learning_rate": 1.211416490486258e-05,
+      "loss": 0.2943,
+      "step": 2055
+    },
+    {
+      "epoch": 2.3468339988590987,
+      "grad_norm": 0.21615460920929327,
+      "learning_rate": 1.2093023255813954e-05,
+      "loss": 0.2864,
+      "step": 2056
+    },
+    {
+      "epoch": 2.347974900171135,
+      "grad_norm": 0.23818904219323894,
+      "learning_rate": 1.2071881606765328e-05,
+      "loss": 0.2899,
+      "step": 2057
+    },
+    {
+      "epoch": 2.3491158014831717,
+      "grad_norm": 0.2378427224755335,
+      "learning_rate": 1.2050739957716702e-05,
+      "loss": 0.2674,
+      "step": 2058
+    },
+    {
+      "epoch": 2.350256702795208,
+      "grad_norm": 0.22758941310375355,
+      "learning_rate": 1.2029598308668076e-05,
+      "loss": 0.306,
+      "step": 2059
+    },
+    {
+      "epoch": 2.3513976041072446,
+      "grad_norm": 0.25202783217641567,
+      "learning_rate": 1.200845665961945e-05,
+      "loss": 0.2889,
+      "step": 2060
+    },
+    {
+      "epoch": 2.352538505419281,
+      "grad_norm": 0.21174390435097373,
+      "learning_rate": 1.1987315010570824e-05,
+      "loss": 0.2912,
+      "step": 2061
+    },
+    {
+      "epoch": 2.3536794067313176,
+      "grad_norm": 0.22922473498775675,
+      "learning_rate": 1.19661733615222e-05,
+      "loss": 0.2954,
+      "step": 2062
+    },
+    {
+      "epoch": 2.354820308043354,
+      "grad_norm": 0.24159279897526093,
+      "learning_rate": 1.1945031712473574e-05,
+      "loss": 0.2958,
+      "step": 2063
+    },
+    {
+      "epoch": 2.3559612093553906,
+      "grad_norm": 0.25268059685668554,
+      "learning_rate": 1.1923890063424947e-05,
+      "loss": 0.2868,
+      "step": 2064
+    },
+    {
+      "epoch": 2.3571021106674275,
+      "grad_norm": 0.2263904802772943,
+      "learning_rate": 1.1902748414376322e-05,
+      "loss": 0.299,
+      "step": 2065
+    },
+    {
+      "epoch": 2.358243011979464,
+      "grad_norm": 0.2246350090323415,
+      "learning_rate": 1.1881606765327696e-05,
+      "loss": 0.2851,
+      "step": 2066
+    },
+    {
+      "epoch": 2.3593839132915004,
+      "grad_norm": 0.22816530324187517,
+      "learning_rate": 1.186046511627907e-05,
+      "loss": 0.2899,
+      "step": 2067
+    },
+    {
+      "epoch": 2.360524814603537,
+      "grad_norm": 0.2164548765906234,
+      "learning_rate": 1.1839323467230445e-05,
+      "loss": 0.2766,
+      "step": 2068
+    },
+    {
+      "epoch": 2.3616657159155734,
+      "grad_norm": 0.23513941860792906,
+      "learning_rate": 1.1818181818181819e-05,
+      "loss": 0.2841,
+      "step": 2069
+    },
+    {
+      "epoch": 2.36280661722761,
+      "grad_norm": 0.22235880511179723,
+      "learning_rate": 1.1797040169133193e-05,
+      "loss": 0.2732,
+      "step": 2070
+    },
+    {
+      "epoch": 2.3639475185396464,
+      "grad_norm": 0.2303519935227575,
+      "learning_rate": 1.1775898520084567e-05,
+      "loss": 0.3139,
+      "step": 2071
+    },
+    {
+      "epoch": 2.365088419851683,
+      "grad_norm": 0.20961595927626084,
+      "learning_rate": 1.1754756871035941e-05,
+      "loss": 0.2875,
+      "step": 2072
+    },
+    {
+      "epoch": 2.3662293211637193,
+      "grad_norm": 0.21603133983276732,
+      "learning_rate": 1.1733615221987315e-05,
+      "loss": 0.2884,
+      "step": 2073
+    },
+    {
+      "epoch": 2.367370222475756,
+      "grad_norm": 0.21596790812542116,
+      "learning_rate": 1.1712473572938689e-05,
+      "loss": 0.3083,
+      "step": 2074
+    },
+    {
+      "epoch": 2.3685111237877923,
+      "grad_norm": 0.25087015544580316,
+      "learning_rate": 1.1691331923890065e-05,
+      "loss": 0.2832,
+      "step": 2075
+    },
+    {
+      "epoch": 2.3696520250998288,
+      "grad_norm": 0.2349808227008684,
+      "learning_rate": 1.1670190274841439e-05,
+      "loss": 0.3084,
+      "step": 2076
+    },
+    {
+      "epoch": 2.3707929264118652,
+      "grad_norm": 0.21001164027904437,
+      "learning_rate": 1.1649048625792811e-05,
+      "loss": 0.2802,
+      "step": 2077
+    },
+    {
+      "epoch": 2.3719338277239017,
+      "grad_norm": 0.2474660750534766,
+      "learning_rate": 1.1627906976744187e-05,
+      "loss": 0.2921,
+      "step": 2078
+    },
+    {
+      "epoch": 2.373074729035938,
+      "grad_norm": 0.24463465462042744,
+      "learning_rate": 1.1606765327695561e-05,
+      "loss": 0.2937,
+      "step": 2079
+    },
+    {
+      "epoch": 2.374215630347975,
+      "grad_norm": 0.24342454656237167,
+      "learning_rate": 1.1585623678646935e-05,
+      "loss": 0.292,
+      "step": 2080
+    },
+    {
+      "epoch": 2.3753565316600116,
+      "grad_norm": 0.2463926019467057,
+      "learning_rate": 1.156448202959831e-05,
+      "loss": 0.2972,
+      "step": 2081
+    },
+    {
+      "epoch": 2.376497432972048,
+      "grad_norm": 0.23720071752094882,
+      "learning_rate": 1.1543340380549683e-05,
+      "loss": 0.2917,
+      "step": 2082
+    },
+    {
+      "epoch": 2.3776383342840846,
+      "grad_norm": 0.21546829355069966,
+      "learning_rate": 1.1522198731501057e-05,
+      "loss": 0.2934,
+      "step": 2083
+    },
+    {
+      "epoch": 2.378779235596121,
+      "grad_norm": 0.2518528409099341,
+      "learning_rate": 1.1501057082452431e-05,
+      "loss": 0.2957,
+      "step": 2084
+    },
+    {
+      "epoch": 2.3799201369081575,
+      "grad_norm": 0.23902483436916538,
+      "learning_rate": 1.1479915433403807e-05,
+      "loss": 0.2801,
+      "step": 2085
+    },
+    {
+      "epoch": 2.381061038220194,
+      "grad_norm": 0.22742879481333675,
+      "learning_rate": 1.145877378435518e-05,
+      "loss": 0.3089,
+      "step": 2086
+    },
+    {
+      "epoch": 2.3822019395322305,
+      "grad_norm": 0.2058956665211023,
+      "learning_rate": 1.1437632135306554e-05,
+      "loss": 0.2823,
+      "step": 2087
+    },
+    {
+      "epoch": 2.383342840844267,
+      "grad_norm": 0.22631377036694728,
+      "learning_rate": 1.141649048625793e-05,
+      "loss": 0.2775,
+      "step": 2088
+    },
+    {
+      "epoch": 2.3844837421563034,
+      "grad_norm": 0.2214324306878773,
+      "learning_rate": 1.1395348837209304e-05,
+      "loss": 0.2891,
+      "step": 2089
+    },
+    {
+      "epoch": 2.38562464346834,
+      "grad_norm": 0.23356628029155893,
+      "learning_rate": 1.1374207188160676e-05,
+      "loss": 0.2874,
+      "step": 2090
+    },
+    {
+      "epoch": 2.3867655447803764,
+      "grad_norm": 0.24243095398966297,
+      "learning_rate": 1.1353065539112052e-05,
+      "loss": 0.2934,
+      "step": 2091
+    },
+    {
+      "epoch": 2.387906446092413,
+      "grad_norm": 0.21508838940257163,
+      "learning_rate": 1.1331923890063426e-05,
+      "loss": 0.2913,
+      "step": 2092
+    },
+    {
+      "epoch": 2.3890473474044494,
+      "grad_norm": 0.24449508059101233,
+      "learning_rate": 1.13107822410148e-05,
+      "loss": 0.2932,
+      "step": 2093
+    },
+    {
+      "epoch": 2.390188248716486,
+      "grad_norm": 0.23075538723084565,
+      "learning_rate": 1.1289640591966174e-05,
+      "loss": 0.2946,
+      "step": 2094
+    },
+    {
+      "epoch": 2.3913291500285228,
+      "grad_norm": 0.21406924795031834,
+      "learning_rate": 1.1268498942917548e-05,
+      "loss": 0.2757,
+      "step": 2095
+    },
+    {
+      "epoch": 2.3924700513405592,
+      "grad_norm": 0.2137290364922405,
+      "learning_rate": 1.1247357293868922e-05,
+      "loss": 0.2976,
+      "step": 2096
+    },
+    {
+      "epoch": 2.3936109526525957,
+      "grad_norm": 0.23118716853112048,
+      "learning_rate": 1.1226215644820296e-05,
+      "loss": 0.2792,
+      "step": 2097
+    },
+    {
+      "epoch": 2.394751853964632,
+      "grad_norm": 0.23055275379491955,
+      "learning_rate": 1.1205073995771672e-05,
+      "loss": 0.2877,
+      "step": 2098
+    },
+    {
+      "epoch": 2.3958927552766687,
+      "grad_norm": 0.23762556799317847,
+      "learning_rate": 1.1183932346723044e-05,
+      "loss": 0.2909,
+      "step": 2099
+    },
+    {
+      "epoch": 2.397033656588705,
+      "grad_norm": 0.22760056957997504,
+      "learning_rate": 1.1162790697674418e-05,
+      "loss": 0.305,
+      "step": 2100
+    },
+    {
+      "epoch": 2.3981745579007416,
+      "grad_norm": 0.2164934127449452,
+      "learning_rate": 1.1141649048625794e-05,
+      "loss": 0.2836,
+      "step": 2101
+    },
+    {
+      "epoch": 2.399315459212778,
+      "grad_norm": 0.2369221361495161,
+      "learning_rate": 1.1120507399577168e-05,
+      "loss": 0.2883,
+      "step": 2102
+    },
+    {
+      "epoch": 2.4004563605248146,
+      "grad_norm": 0.21848646801955754,
+      "learning_rate": 1.109936575052854e-05,
+      "loss": 0.2785,
+      "step": 2103
+    },
+    {
+      "epoch": 2.401597261836851,
+      "grad_norm": 0.22794077248598318,
+      "learning_rate": 1.1078224101479916e-05,
+      "loss": 0.2817,
+      "step": 2104
+    },
+    {
+      "epoch": 2.4027381631488876,
+      "grad_norm": 0.23285629065646596,
+      "learning_rate": 1.105708245243129e-05,
+      "loss": 0.2886,
+      "step": 2105
+    },
+    {
+      "epoch": 2.403879064460924,
+      "grad_norm": 0.23290817753384455,
+      "learning_rate": 1.1035940803382664e-05,
+      "loss": 0.2806,
+      "step": 2106
+    },
+    {
+      "epoch": 2.4050199657729605,
+      "grad_norm": 0.23158661692280794,
+      "learning_rate": 1.1014799154334039e-05,
+      "loss": 0.3063,
+      "step": 2107
+    },
+    {
+      "epoch": 2.406160867084997,
+      "grad_norm": 0.20833408878949644,
+      "learning_rate": 1.0993657505285413e-05,
+      "loss": 0.2886,
+      "step": 2108
+    },
+    {
+      "epoch": 2.4073017683970335,
+      "grad_norm": 0.23083635889774368,
+      "learning_rate": 1.0972515856236787e-05,
+      "loss": 0.2869,
+      "step": 2109
+    },
+    {
+      "epoch": 2.4084426697090704,
+      "grad_norm": 0.22576281424305952,
+      "learning_rate": 1.095137420718816e-05,
+      "loss": 0.2854,
+      "step": 2110
+    },
+    {
+      "epoch": 2.4095835710211064,
+      "grad_norm": 0.22884347743104985,
+      "learning_rate": 1.0930232558139537e-05,
+      "loss": 0.2967,
+      "step": 2111
+    },
+    {
+      "epoch": 2.4107244723331434,
+      "grad_norm": 0.22415283034169417,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 0.2913,
+      "step": 2112
+    },
+    {
+      "epoch": 2.41186537364518,
+      "grad_norm": 0.21200894004522153,
+      "learning_rate": 1.0887949260042283e-05,
+      "loss": 0.2888,
+      "step": 2113
+    },
+    {
+      "epoch": 2.4130062749572163,
+      "grad_norm": 0.20964684591494562,
+      "learning_rate": 1.0866807610993659e-05,
+      "loss": 0.2925,
+      "step": 2114
+    },
+    {
+      "epoch": 2.414147176269253,
+      "grad_norm": 0.24476217114724486,
+      "learning_rate": 1.0845665961945033e-05,
+      "loss": 0.2767,
+      "step": 2115
+    },
+    {
+      "epoch": 2.4152880775812893,
+      "grad_norm": 0.2505050259919476,
+      "learning_rate": 1.0824524312896405e-05,
+      "loss": 0.2974,
+      "step": 2116
+    },
+    {
+      "epoch": 2.4164289788933258,
+      "grad_norm": 0.22788028722264833,
+      "learning_rate": 1.0803382663847781e-05,
+      "loss": 0.2748,
+      "step": 2117
+    },
+    {
+      "epoch": 2.4175698802053622,
+      "grad_norm": 0.23594844937596657,
+      "learning_rate": 1.0782241014799155e-05,
+      "loss": 0.2773,
+      "step": 2118
+    },
+    {
+      "epoch": 2.4187107815173987,
+      "grad_norm": 0.22373608681920468,
+      "learning_rate": 1.0761099365750529e-05,
+      "loss": 0.2968,
+      "step": 2119
+    },
+    {
+      "epoch": 2.419851682829435,
+      "grad_norm": 0.20267836607609419,
+      "learning_rate": 1.0739957716701903e-05,
+      "loss": 0.2875,
+      "step": 2120
+    },
+    {
+      "epoch": 2.4209925841414717,
+      "grad_norm": 0.23304066668147824,
+      "learning_rate": 1.0718816067653277e-05,
+      "loss": 0.2784,
+      "step": 2121
+    },
+    {
+      "epoch": 2.422133485453508,
+      "grad_norm": 0.22386813115682758,
+      "learning_rate": 1.0697674418604651e-05,
+      "loss": 0.2895,
+      "step": 2122
+    },
+    {
+      "epoch": 2.4232743867655446,
+      "grad_norm": 0.23791481430326425,
+      "learning_rate": 1.0676532769556025e-05,
+      "loss": 0.2973,
+      "step": 2123
+    },
+    {
+      "epoch": 2.424415288077581,
+      "grad_norm": 0.22929508872335239,
+      "learning_rate": 1.0655391120507401e-05,
+      "loss": 0.2882,
+      "step": 2124
+    },
+    {
+      "epoch": 2.425556189389618,
+      "grad_norm": 0.2548891481570048,
+      "learning_rate": 1.0634249471458774e-05,
+      "loss": 0.2862,
+      "step": 2125
+    },
+    {
+      "epoch": 2.426697090701654,
+      "grad_norm": 0.21746741378702147,
+      "learning_rate": 1.0613107822410148e-05,
+      "loss": 0.2987,
+      "step": 2126
+    },
+    {
+      "epoch": 2.427837992013691,
+      "grad_norm": 0.21251661156113041,
+      "learning_rate": 1.0591966173361523e-05,
+      "loss": 0.2894,
+      "step": 2127
+    },
+    {
+      "epoch": 2.4289788933257275,
+      "grad_norm": 0.25780499036245796,
+      "learning_rate": 1.0570824524312897e-05,
+      "loss": 0.2856,
+      "step": 2128
+    },
+    {
+      "epoch": 2.430119794637764,
+      "grad_norm": 0.23654473246154903,
+      "learning_rate": 1.0549682875264272e-05,
+      "loss": 0.3007,
+      "step": 2129
+    },
+    {
+      "epoch": 2.4312606959498004,
+      "grad_norm": 0.23356461407264056,
+      "learning_rate": 1.0528541226215646e-05,
+      "loss": 0.3256,
+      "step": 2130
+    },
+    {
+      "epoch": 2.432401597261837,
+      "grad_norm": 0.24245974715235033,
+      "learning_rate": 1.050739957716702e-05,
+      "loss": 0.3106,
+      "step": 2131
+    },
+    {
+      "epoch": 2.4335424985738734,
+      "grad_norm": 0.24303489817253957,
+      "learning_rate": 1.0486257928118394e-05,
+      "loss": 0.2955,
+      "step": 2132
+    },
+    {
+      "epoch": 2.43468339988591,
+      "grad_norm": 0.24005732028686433,
+      "learning_rate": 1.0465116279069768e-05,
+      "loss": 0.2893,
+      "step": 2133
+    },
+    {
+      "epoch": 2.4358243011979464,
+      "grad_norm": 0.2121910038291667,
+      "learning_rate": 1.0443974630021142e-05,
+      "loss": 0.2945,
+      "step": 2134
+    },
+    {
+      "epoch": 2.436965202509983,
+      "grad_norm": 0.2387084030486226,
+      "learning_rate": 1.0422832980972516e-05,
+      "loss": 0.2903,
+      "step": 2135
+    },
+    {
+      "epoch": 2.4381061038220193,
+      "grad_norm": 0.23633958382284798,
+      "learning_rate": 1.040169133192389e-05,
+      "loss": 0.2996,
+      "step": 2136
+    },
+    {
+      "epoch": 2.439247005134056,
+      "grad_norm": 0.22037384346987707,
+      "learning_rate": 1.0380549682875266e-05,
+      "loss": 0.2839,
+      "step": 2137
+    },
+    {
+      "epoch": 2.4403879064460923,
+      "grad_norm": 0.21706680631361816,
+      "learning_rate": 1.0359408033826638e-05,
+      "loss": 0.2975,
+      "step": 2138
+    },
+    {
+      "epoch": 2.4415288077581287,
+      "grad_norm": 0.2111778175282013,
+      "learning_rate": 1.0338266384778012e-05,
+      "loss": 0.2729,
+      "step": 2139
+    },
+    {
+      "epoch": 2.4426697090701657,
+      "grad_norm": 0.2370242617846862,
+      "learning_rate": 1.0317124735729388e-05,
+      "loss": 0.2909,
+      "step": 2140
+    },
+    {
+      "epoch": 2.4438106103822017,
+      "grad_norm": 0.23853498865235603,
+      "learning_rate": 1.0295983086680762e-05,
+      "loss": 0.2932,
+      "step": 2141
+    },
+    {
+      "epoch": 2.4449515116942386,
+      "grad_norm": 0.2521182998898568,
+      "learning_rate": 1.0274841437632136e-05,
+      "loss": 0.3059,
+      "step": 2142
+    },
+    {
+      "epoch": 2.446092413006275,
+      "grad_norm": 0.23322546814866815,
+      "learning_rate": 1.025369978858351e-05,
+      "loss": 0.295,
+      "step": 2143
+    },
+    {
+      "epoch": 2.4472333143183116,
+      "grad_norm": 0.2321246083656867,
+      "learning_rate": 1.0232558139534884e-05,
+      "loss": 0.2916,
+      "step": 2144
+    },
+    {
+      "epoch": 2.448374215630348,
+      "grad_norm": 0.21203888608803423,
+      "learning_rate": 1.0211416490486258e-05,
+      "loss": 0.3057,
+      "step": 2145
+    },
+    {
+      "epoch": 2.4495151169423846,
+      "grad_norm": 0.24675500559447644,
+      "learning_rate": 1.0190274841437632e-05,
+      "loss": 0.3005,
+      "step": 2146
+    },
+    {
+      "epoch": 2.450656018254421,
+      "grad_norm": 0.2149188388562516,
+      "learning_rate": 1.0169133192389007e-05,
+      "loss": 0.2913,
+      "step": 2147
+    },
+    {
+      "epoch": 2.4517969195664575,
+      "grad_norm": 0.22455843922740962,
+      "learning_rate": 1.014799154334038e-05,
+      "loss": 0.2911,
+      "step": 2148
+    },
+    {
+      "epoch": 2.452937820878494,
+      "grad_norm": 0.2365688821621369,
+      "learning_rate": 1.0126849894291755e-05,
+      "loss": 0.3009,
+      "step": 2149
+    },
+    {
+      "epoch": 2.4540787221905305,
+      "grad_norm": 0.236174371463899,
+      "learning_rate": 1.010570824524313e-05,
+      "loss": 0.2963,
+      "step": 2150
+    },
+    {
+      "epoch": 2.455219623502567,
+      "grad_norm": 0.22166605922121535,
+      "learning_rate": 1.0084566596194505e-05,
+      "loss": 0.3,
+      "step": 2151
+    },
+    {
+      "epoch": 2.4563605248146034,
+      "grad_norm": 0.21578175432032534,
+      "learning_rate": 1.0063424947145877e-05,
+      "loss": 0.3036,
+      "step": 2152
+    },
+    {
+      "epoch": 2.45750142612664,
+      "grad_norm": 0.20735663165711912,
+      "learning_rate": 1.0042283298097253e-05,
+      "loss": 0.2903,
+      "step": 2153
+    },
+    {
+      "epoch": 2.4586423274386764,
+      "grad_norm": 0.21996192089426445,
+      "learning_rate": 1.0021141649048627e-05,
+      "loss": 0.2907,
+      "step": 2154
+    },
+    {
+      "epoch": 2.4597832287507133,
+      "grad_norm": 0.2458406652272878,
+      "learning_rate": 1e-05,
+      "loss": 0.2969,
+      "step": 2155
+    },
+    {
+      "epoch": 2.4609241300627493,
+      "grad_norm": 0.22647445428101715,
+      "learning_rate": 9.978858350951375e-06,
+      "loss": 0.3052,
+      "step": 2156
+    },
+    {
+      "epoch": 2.4620650313747863,
+      "grad_norm": 0.2219779270023204,
+      "learning_rate": 9.957716701902749e-06,
+      "loss": 0.2824,
+      "step": 2157
+    },
+    {
+      "epoch": 2.4632059326868228,
+      "grad_norm": 0.23756600468667605,
+      "learning_rate": 9.936575052854123e-06,
+      "loss": 0.3011,
+      "step": 2158
+    },
+    {
+      "epoch": 2.4643468339988592,
+      "grad_norm": 0.2272830607275422,
+      "learning_rate": 9.915433403805497e-06,
+      "loss": 0.2718,
+      "step": 2159
+    },
+    {
+      "epoch": 2.4654877353108957,
+      "grad_norm": 0.2145277871400877,
+      "learning_rate": 9.894291754756871e-06,
+      "loss": 0.2941,
+      "step": 2160
+    },
+    {
+      "epoch": 2.466628636622932,
+      "grad_norm": 0.2376090963043766,
+      "learning_rate": 9.873150105708245e-06,
+      "loss": 0.3067,
+      "step": 2161
+    },
+    {
+      "epoch": 2.4677695379349687,
+      "grad_norm": 0.24356987631727217,
+      "learning_rate": 9.85200845665962e-06,
+      "loss": 0.3022,
+      "step": 2162
+    },
+    {
+      "epoch": 2.468910439247005,
+      "grad_norm": 0.22299783197978315,
+      "learning_rate": 9.830866807610995e-06,
+      "loss": 0.2869,
+      "step": 2163
+    },
+    {
+      "epoch": 2.4700513405590416,
+      "grad_norm": 0.20939756640471135,
+      "learning_rate": 9.80972515856237e-06,
+      "loss": 0.3098,
+      "step": 2164
+    },
+    {
+      "epoch": 2.471192241871078,
+      "grad_norm": 0.21609710238107147,
+      "learning_rate": 9.788583509513742e-06,
+      "loss": 0.2952,
+      "step": 2165
+    },
+    {
+      "epoch": 2.4723331431831146,
+      "grad_norm": 0.2372720663818419,
+      "learning_rate": 9.767441860465117e-06,
+      "loss": 0.2898,
+      "step": 2166
+    },
+    {
+      "epoch": 2.473474044495151,
+      "grad_norm": 0.21121175197196843,
+      "learning_rate": 9.746300211416491e-06,
+      "loss": 0.3025,
+      "step": 2167
+    },
+    {
+      "epoch": 2.4746149458071875,
+      "grad_norm": 0.24198203548706435,
+      "learning_rate": 9.725158562367865e-06,
+      "loss": 0.296,
+      "step": 2168
+    },
+    {
+      "epoch": 2.475755847119224,
+      "grad_norm": 0.24194232553304587,
+      "learning_rate": 9.70401691331924e-06,
+      "loss": 0.2853,
+      "step": 2169
+    },
+    {
+      "epoch": 2.4768967484312605,
+      "grad_norm": 0.22799957616792252,
+      "learning_rate": 9.682875264270614e-06,
+      "loss": 0.2862,
+      "step": 2170
+    },
+    {
+      "epoch": 2.478037649743297,
+      "grad_norm": 0.227565930275027,
+      "learning_rate": 9.661733615221988e-06,
+      "loss": 0.2922,
+      "step": 2171
+    },
+    {
+      "epoch": 2.479178551055334,
+      "grad_norm": 0.21324291845071908,
+      "learning_rate": 9.640591966173362e-06,
+      "loss": 0.2957,
+      "step": 2172
+    },
+    {
+      "epoch": 2.4803194523673704,
+      "grad_norm": 0.24298054311441278,
+      "learning_rate": 9.619450317124738e-06,
+      "loss": 0.2781,
+      "step": 2173
+    },
+    {
+      "epoch": 2.481460353679407,
+      "grad_norm": 0.19336392182500792,
+      "learning_rate": 9.59830866807611e-06,
+      "loss": 0.2779,
+      "step": 2174
+    },
+    {
+      "epoch": 2.4826012549914434,
+      "grad_norm": 0.20345662355740218,
+      "learning_rate": 9.577167019027484e-06,
+      "loss": 0.2823,
+      "step": 2175
+    },
+    {
+      "epoch": 2.48374215630348,
+      "grad_norm": 0.20790668803942755,
+      "learning_rate": 9.55602536997886e-06,
+      "loss": 0.2947,
+      "step": 2176
+    },
+    {
+      "epoch": 2.4848830576155163,
+      "grad_norm": 0.21470421388061137,
+      "learning_rate": 9.534883720930234e-06,
+      "loss": 0.288,
+      "step": 2177
+    },
+    {
+      "epoch": 2.486023958927553,
+      "grad_norm": 0.2373711556108657,
+      "learning_rate": 9.513742071881606e-06,
+      "loss": 0.2871,
+      "step": 2178
+    },
+    {
+      "epoch": 2.4871648602395893,
+      "grad_norm": 0.2137356437142869,
+      "learning_rate": 9.492600422832982e-06,
+      "loss": 0.281,
+      "step": 2179
+    },
+    {
+      "epoch": 2.4883057615516257,
+      "grad_norm": 0.2160640542317364,
+      "learning_rate": 9.471458773784356e-06,
+      "loss": 0.2909,
+      "step": 2180
+    },
+    {
+      "epoch": 2.4894466628636622,
+      "grad_norm": 0.23681297605390203,
+      "learning_rate": 9.45031712473573e-06,
+      "loss": 0.2976,
+      "step": 2181
+    },
+    {
+      "epoch": 2.4905875641756987,
+      "grad_norm": 0.21619935440988564,
+      "learning_rate": 9.429175475687104e-06,
+      "loss": 0.2965,
+      "step": 2182
+    },
+    {
+      "epoch": 2.491728465487735,
+      "grad_norm": 0.22582877145805355,
+      "learning_rate": 9.408033826638478e-06,
+      "loss": 0.2884,
+      "step": 2183
+    },
+    {
+      "epoch": 2.4928693667997717,
+      "grad_norm": 0.24291200698676815,
+      "learning_rate": 9.386892177589852e-06,
+      "loss": 0.2833,
+      "step": 2184
+    },
+    {
+      "epoch": 2.494010268111808,
+      "grad_norm": 0.22462516903821608,
+      "learning_rate": 9.365750528541226e-06,
+      "loss": 0.2814,
+      "step": 2185
+    },
+    {
+      "epoch": 2.4951511694238446,
+      "grad_norm": 0.2150912847250357,
+      "learning_rate": 9.344608879492602e-06,
+      "loss": 0.2865,
+      "step": 2186
+    },
+    {
+      "epoch": 2.4962920707358816,
+      "grad_norm": 0.22046725631639622,
+      "learning_rate": 9.323467230443975e-06,
+      "loss": 0.3042,
+      "step": 2187
+    },
+    {
+      "epoch": 2.497432972047918,
+      "grad_norm": 0.21876472045649098,
+      "learning_rate": 9.302325581395349e-06,
+      "loss": 0.3086,
+      "step": 2188
+    },
+    {
+      "epoch": 2.4985738733599545,
+      "grad_norm": 0.22155910910209872,
+      "learning_rate": 9.281183932346724e-06,
+      "loss": 0.2913,
+      "step": 2189
+    },
+    {
+      "epoch": 2.499714774671991,
+      "grad_norm": 0.21640048073756032,
+      "learning_rate": 9.260042283298098e-06,
+      "loss": 0.2945,
+      "step": 2190
+    },
+    {
+      "epoch": 2.5008556759840275,
+      "grad_norm": 0.22250078526965852,
+      "learning_rate": 9.23890063424947e-06,
+      "loss": 0.3035,
+      "step": 2191
+    },
+    {
+      "epoch": 2.501996577296064,
+      "grad_norm": 0.24474339500328393,
+      "learning_rate": 9.217758985200847e-06,
+      "loss": 0.3127,
+      "step": 2192
+    },
+    {
+      "epoch": 2.5031374786081004,
+      "grad_norm": 0.251308976662145,
+      "learning_rate": 9.19661733615222e-06,
+      "loss": 0.2875,
+      "step": 2193
+    },
+    {
+      "epoch": 2.504278379920137,
+      "grad_norm": 0.22794637405036033,
+      "learning_rate": 9.175475687103595e-06,
+      "loss": 0.2913,
+      "step": 2194
+    },
+    {
+      "epoch": 2.5054192812321734,
+      "grad_norm": 0.259791985825867,
+      "learning_rate": 9.154334038054969e-06,
+      "loss": 0.3014,
+      "step": 2195
+    },
+    {
+      "epoch": 2.50656018254421,
+      "grad_norm": 0.22447975268191678,
+      "learning_rate": 9.133192389006343e-06,
+      "loss": 0.2852,
+      "step": 2196
+    },
+    {
+      "epoch": 2.5077010838562463,
+      "grad_norm": 0.2339651033453112,
+      "learning_rate": 9.112050739957717e-06,
+      "loss": 0.2981,
+      "step": 2197
+    },
+    {
+      "epoch": 2.508841985168283,
+      "grad_norm": 0.2413662095243327,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 0.297,
+      "step": 2198
+    },
+    {
+      "epoch": 2.5099828864803193,
+      "grad_norm": 0.25405342039518297,
+      "learning_rate": 9.069767441860467e-06,
+      "loss": 0.2871,
+      "step": 2199
+    },
+    {
+      "epoch": 2.5111237877923562,
+      "grad_norm": 0.23920425875866733,
+      "learning_rate": 9.04862579281184e-06,
+      "loss": 0.2875,
+      "step": 2200
+    },
+    {
+      "epoch": 2.5122646891043923,
+      "grad_norm": 0.2051445130760935,
+      "learning_rate": 9.027484143763213e-06,
+      "loss": 0.2758,
+      "step": 2201
+    },
+    {
+      "epoch": 2.513405590416429,
+      "grad_norm": 0.20124858870683618,
+      "learning_rate": 9.006342494714589e-06,
+      "loss": 0.2972,
+      "step": 2202
+    },
+    {
+      "epoch": 2.5145464917284652,
+      "grad_norm": 0.23302207610442713,
+      "learning_rate": 8.985200845665963e-06,
+      "loss": 0.2978,
+      "step": 2203
+    },
+    {
+      "epoch": 2.515687393040502,
+      "grad_norm": 0.21344843871744304,
+      "learning_rate": 8.964059196617335e-06,
+      "loss": 0.301,
+      "step": 2204
+    },
+    {
+      "epoch": 2.5168282943525386,
+      "grad_norm": 0.2407781938188346,
+      "learning_rate": 8.942917547568711e-06,
+      "loss": 0.281,
+      "step": 2205
+    },
+    {
+      "epoch": 2.517969195664575,
+      "grad_norm": 0.20516201626203046,
+      "learning_rate": 8.921775898520085e-06,
+      "loss": 0.3002,
+      "step": 2206
+    },
+    {
+      "epoch": 2.5191100969766116,
+      "grad_norm": 0.2104381049294445,
+      "learning_rate": 8.90063424947146e-06,
+      "loss": 0.2979,
+      "step": 2207
+    },
+    {
+      "epoch": 2.520250998288648,
+      "grad_norm": 0.22424201493799636,
+      "learning_rate": 8.879492600422833e-06,
+      "loss": 0.2845,
+      "step": 2208
+    },
+    {
+      "epoch": 2.5213918996006845,
+      "grad_norm": 0.23906885464829122,
+      "learning_rate": 8.858350951374208e-06,
+      "loss": 0.3064,
+      "step": 2209
+    },
+    {
+      "epoch": 2.522532800912721,
+      "grad_norm": 0.21700945798058027,
+      "learning_rate": 8.837209302325582e-06,
+      "loss": 0.2946,
+      "step": 2210
+    },
+    {
+      "epoch": 2.5236737022247575,
+      "grad_norm": 0.20269201104999035,
+      "learning_rate": 8.816067653276956e-06,
+      "loss": 0.2918,
+      "step": 2211
+    },
+    {
+      "epoch": 2.524814603536794,
+      "grad_norm": 0.1983271277482634,
+      "learning_rate": 8.794926004228331e-06,
+      "loss": 0.2823,
+      "step": 2212
+    },
+    {
+      "epoch": 2.5259555048488305,
+      "grad_norm": 0.2071653143990948,
+      "learning_rate": 8.773784355179704e-06,
+      "loss": 0.2987,
+      "step": 2213
+    },
+    {
+      "epoch": 2.527096406160867,
+      "grad_norm": 0.21535784337570055,
+      "learning_rate": 8.752642706131078e-06,
+      "loss": 0.2962,
+      "step": 2214
+    },
+    {
+      "epoch": 2.5282373074729034,
+      "grad_norm": 0.2296330117605354,
+      "learning_rate": 8.731501057082454e-06,
+      "loss": 0.3056,
+      "step": 2215
+    },
+    {
+      "epoch": 2.52937820878494,
+      "grad_norm": 0.2271558756128669,
+      "learning_rate": 8.710359408033828e-06,
+      "loss": 0.2892,
+      "step": 2216
+    },
+    {
+      "epoch": 2.530519110096977,
+      "grad_norm": 0.2105137344073212,
+      "learning_rate": 8.6892177589852e-06,
+      "loss": 0.2947,
+      "step": 2217
+    },
+    {
+      "epoch": 2.531660011409013,
+      "grad_norm": 0.2184000564503314,
+      "learning_rate": 8.668076109936576e-06,
+      "loss": 0.2809,
+      "step": 2218
+    },
+    {
+      "epoch": 2.53280091272105,
+      "grad_norm": 0.2263050553315689,
+      "learning_rate": 8.64693446088795e-06,
+      "loss": 0.2788,
+      "step": 2219
+    },
+    {
+      "epoch": 2.5339418140330863,
+      "grad_norm": 0.21604440983648726,
+      "learning_rate": 8.625792811839324e-06,
+      "loss": 0.2936,
+      "step": 2220
+    },
+    {
+      "epoch": 2.5350827153451228,
+      "grad_norm": 0.21737368520487446,
+      "learning_rate": 8.604651162790698e-06,
+      "loss": 0.2961,
+      "step": 2221
+    },
+    {
+      "epoch": 2.5362236166571592,
+      "grad_norm": 0.21773148010014565,
+      "learning_rate": 8.583509513742072e-06,
+      "loss": 0.2884,
+      "step": 2222
+    },
+    {
+      "epoch": 2.5373645179691957,
+      "grad_norm": 0.22892926444409067,
+      "learning_rate": 8.562367864693446e-06,
+      "loss": 0.3067,
+      "step": 2223
+    },
+    {
+      "epoch": 2.538505419281232,
+      "grad_norm": 0.21226778252034528,
+      "learning_rate": 8.54122621564482e-06,
+      "loss": 0.2936,
+      "step": 2224
+    },
+    {
+      "epoch": 2.5396463205932687,
+      "grad_norm": 0.2152635025793325,
+      "learning_rate": 8.520084566596196e-06,
+      "loss": 0.3102,
+      "step": 2225
+    },
+    {
+      "epoch": 2.540787221905305,
+      "grad_norm": 0.21770436762764223,
+      "learning_rate": 8.498942917547568e-06,
+      "loss": 0.2983,
+      "step": 2226
+    },
+    {
+      "epoch": 2.5419281232173416,
+      "grad_norm": 0.22410803091422948,
+      "learning_rate": 8.477801268498943e-06,
+      "loss": 0.2982,
+      "step": 2227
+    },
+    {
+      "epoch": 2.543069024529378,
+      "grad_norm": 0.20713974718886613,
+      "learning_rate": 8.456659619450318e-06,
+      "loss": 0.2785,
+      "step": 2228
+    },
+    {
+      "epoch": 2.5442099258414146,
+      "grad_norm": 0.21352924445956112,
+      "learning_rate": 8.435517970401692e-06,
+      "loss": 0.2773,
+      "step": 2229
+    },
+    {
+      "epoch": 2.545350827153451,
+      "grad_norm": 0.20187171844616342,
+      "learning_rate": 8.414376321353066e-06,
+      "loss": 0.2995,
+      "step": 2230
+    },
+    {
+      "epoch": 2.5464917284654875,
+      "grad_norm": 0.20941049925190608,
+      "learning_rate": 8.39323467230444e-06,
+      "loss": 0.2784,
+      "step": 2231
+    },
+    {
+      "epoch": 2.5476326297775245,
+      "grad_norm": 0.20180411877439458,
+      "learning_rate": 8.372093023255815e-06,
+      "loss": 0.2773,
+      "step": 2232
+    },
+    {
+      "epoch": 2.5487735310895605,
+      "grad_norm": 0.22093030719312548,
+      "learning_rate": 8.350951374207189e-06,
+      "loss": 0.3012,
+      "step": 2233
+    },
+    {
+      "epoch": 2.5499144324015974,
+      "grad_norm": 0.22523096523898684,
+      "learning_rate": 8.329809725158563e-06,
+      "loss": 0.2922,
+      "step": 2234
+    },
+    {
+      "epoch": 2.551055333713634,
+      "grad_norm": 0.22343591900578977,
+      "learning_rate": 8.308668076109937e-06,
+      "loss": 0.3093,
+      "step": 2235
+    },
+    {
+      "epoch": 2.5521962350256704,
+      "grad_norm": 0.2161498369058084,
+      "learning_rate": 8.287526427061311e-06,
+      "loss": 0.2969,
+      "step": 2236
+    },
+    {
+      "epoch": 2.553337136337707,
+      "grad_norm": 0.22562631086956914,
+      "learning_rate": 8.266384778012685e-06,
+      "loss": 0.3016,
+      "step": 2237
+    },
+    {
+      "epoch": 2.5544780376497433,
+      "grad_norm": 0.2369459126707438,
+      "learning_rate": 8.24524312896406e-06,
+      "loss": 0.2904,
+      "step": 2238
+    },
+    {
+      "epoch": 2.55561893896178,
+      "grad_norm": 0.22448499213847747,
+      "learning_rate": 8.224101479915433e-06,
+      "loss": 0.2853,
+      "step": 2239
+    },
+    {
+      "epoch": 2.5567598402738163,
+      "grad_norm": 0.21029356493483137,
+      "learning_rate": 8.202959830866807e-06,
+      "loss": 0.2843,
+      "step": 2240
+    },
+    {
+      "epoch": 2.557900741585853,
+      "grad_norm": 0.22200835961714357,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 0.3057,
+      "step": 2241
+    },
+    {
+      "epoch": 2.5590416428978893,
+      "grad_norm": 0.23044204049657901,
+      "learning_rate": 8.160676532769557e-06,
+      "loss": 0.2962,
+      "step": 2242
+    },
+    {
+      "epoch": 2.5601825442099257,
+      "grad_norm": 0.20553292626259104,
+      "learning_rate": 8.139534883720931e-06,
+      "loss": 0.2745,
+      "step": 2243
+    },
+    {
+      "epoch": 2.5613234455219622,
+      "grad_norm": 0.20915386111446832,
+      "learning_rate": 8.118393234672305e-06,
+      "loss": 0.2958,
+      "step": 2244
+    },
+    {
+      "epoch": 2.5624643468339987,
+      "grad_norm": 0.22334896184814618,
+      "learning_rate": 8.09725158562368e-06,
+      "loss": 0.2972,
+      "step": 2245
+    },
+    {
+      "epoch": 2.563605248146035,
+      "grad_norm": 0.23809497758629466,
+      "learning_rate": 8.076109936575053e-06,
+      "loss": 0.2942,
+      "step": 2246
+    },
+    {
+      "epoch": 2.564746149458072,
+      "grad_norm": 0.24069339582590796,
+      "learning_rate": 8.054968287526427e-06,
+      "loss": 0.2918,
+      "step": 2247
+    },
+    {
+      "epoch": 2.565887050770108,
+      "grad_norm": 0.23602440927239052,
+      "learning_rate": 8.033826638477801e-06,
+      "loss": 0.3028,
+      "step": 2248
+    },
+    {
+      "epoch": 2.567027952082145,
+      "grad_norm": 0.2171746495174681,
+      "learning_rate": 8.012684989429176e-06,
+      "loss": 0.2916,
+      "step": 2249
+    },
+    {
+      "epoch": 2.5681688533941815,
+      "grad_norm": 0.2068130866348627,
+      "learning_rate": 7.99154334038055e-06,
+      "loss": 0.2745,
+      "step": 2250
+    },
+    {
+      "epoch": 2.569309754706218,
+      "grad_norm": 0.2177101252393696,
+      "learning_rate": 7.970401691331925e-06,
+      "loss": 0.2975,
+      "step": 2251
+    },
+    {
+      "epoch": 2.5704506560182545,
+      "grad_norm": 0.2735012912904912,
+      "learning_rate": 7.9492600422833e-06,
+      "loss": 0.2926,
+      "step": 2252
+    },
+    {
+      "epoch": 2.571591557330291,
+      "grad_norm": 0.20464785110092396,
+      "learning_rate": 7.928118393234672e-06,
+      "loss": 0.2927,
+      "step": 2253
+    },
+    {
+      "epoch": 2.5727324586423275,
+      "grad_norm": 0.21455342420906462,
+      "learning_rate": 7.906976744186048e-06,
+      "loss": 0.3056,
+      "step": 2254
+    },
+    {
+      "epoch": 2.573873359954364,
+      "grad_norm": 0.22559564092009435,
+      "learning_rate": 7.885835095137422e-06,
+      "loss": 0.2865,
+      "step": 2255
+    },
+    {
+      "epoch": 2.5750142612664004,
+      "grad_norm": 0.2294901793229017,
+      "learning_rate": 7.864693446088796e-06,
+      "loss": 0.2853,
+      "step": 2256
+    },
+    {
+      "epoch": 2.576155162578437,
+      "grad_norm": 0.2147093578647149,
+      "learning_rate": 7.843551797040168e-06,
+      "loss": 0.2936,
+      "step": 2257
+    },
+    {
+      "epoch": 2.5772960638904734,
+      "grad_norm": 0.2031963357013273,
+      "learning_rate": 7.822410147991544e-06,
+      "loss": 0.2852,
+      "step": 2258
+    },
+    {
+      "epoch": 2.57843696520251,
+      "grad_norm": 0.23046080341522127,
+      "learning_rate": 7.801268498942918e-06,
+      "loss": 0.2934,
+      "step": 2259
+    },
+    {
+      "epoch": 2.5795778665145463,
+      "grad_norm": 0.2273970672894729,
+      "learning_rate": 7.780126849894292e-06,
+      "loss": 0.3049,
+      "step": 2260
+    },
+    {
+      "epoch": 2.580718767826583,
+      "grad_norm": 0.20061748470118457,
+      "learning_rate": 7.758985200845666e-06,
+      "loss": 0.2856,
+      "step": 2261
+    },
+    {
+      "epoch": 2.5818596691386198,
+      "grad_norm": 0.22065013702452976,
+      "learning_rate": 7.73784355179704e-06,
+      "loss": 0.3123,
+      "step": 2262
+    },
+    {
+      "epoch": 2.583000570450656,
+      "grad_norm": 0.22822659923283384,
+      "learning_rate": 7.716701902748414e-06,
+      "loss": 0.2822,
+      "step": 2263
+    },
+    {
+      "epoch": 2.5841414717626927,
+      "grad_norm": 0.21413158459854104,
+      "learning_rate": 7.69556025369979e-06,
+      "loss": 0.2976,
+      "step": 2264
+    },
+    {
+      "epoch": 2.585282373074729,
+      "grad_norm": 0.21690708574915987,
+      "learning_rate": 7.674418604651164e-06,
+      "loss": 0.3013,
+      "step": 2265
+    },
+    {
+      "epoch": 2.5864232743867657,
+      "grad_norm": 0.20274340989060452,
+      "learning_rate": 7.653276955602536e-06,
+      "loss": 0.2986,
+      "step": 2266
+    },
+    {
+      "epoch": 2.587564175698802,
+      "grad_norm": 0.22857150975728216,
+      "learning_rate": 7.632135306553912e-06,
+      "loss": 0.2988,
+      "step": 2267
+    },
+    {
+      "epoch": 2.5887050770108386,
+      "grad_norm": 0.24680265809712104,
+      "learning_rate": 7.610993657505286e-06,
+      "loss": 0.3058,
+      "step": 2268
+    },
+    {
+      "epoch": 2.589845978322875,
+      "grad_norm": 0.21717337738812287,
+      "learning_rate": 7.58985200845666e-06,
+      "loss": 0.2759,
+      "step": 2269
+    },
+    {
+      "epoch": 2.5909868796349116,
+      "grad_norm": 0.1956386453417279,
+      "learning_rate": 7.568710359408034e-06,
+      "loss": 0.2997,
+      "step": 2270
+    },
+    {
+      "epoch": 2.592127780946948,
+      "grad_norm": 0.21352206012373648,
+      "learning_rate": 7.547568710359408e-06,
+      "loss": 0.2947,
+      "step": 2271
+    },
+    {
+      "epoch": 2.5932686822589845,
+      "grad_norm": 0.21869565704668345,
+      "learning_rate": 7.526427061310783e-06,
+      "loss": 0.2867,
+      "step": 2272
+    },
+    {
+      "epoch": 2.594409583571021,
+      "grad_norm": 0.2199935791732718,
+      "learning_rate": 7.5052854122621575e-06,
+      "loss": 0.286,
+      "step": 2273
+    },
+    {
+      "epoch": 2.5955504848830575,
+      "grad_norm": 0.2077071830689117,
+      "learning_rate": 7.484143763213532e-06,
+      "loss": 0.2862,
+      "step": 2274
+    },
+    {
+      "epoch": 2.596691386195094,
+      "grad_norm": 0.19250504102605728,
+      "learning_rate": 7.463002114164905e-06,
+      "loss": 0.272,
+      "step": 2275
+    },
+    {
+      "epoch": 2.5978322875071305,
+      "grad_norm": 0.2543282781990224,
+      "learning_rate": 7.44186046511628e-06,
+      "loss": 0.2899,
+      "step": 2276
+    },
+    {
+      "epoch": 2.5989731888191674,
+      "grad_norm": 0.2227953001922955,
+      "learning_rate": 7.420718816067654e-06,
+      "loss": 0.3072,
+      "step": 2277
+    },
+    {
+      "epoch": 2.6001140901312034,
+      "grad_norm": 0.2119324966397027,
+      "learning_rate": 7.399577167019029e-06,
+      "loss": 0.2818,
+      "step": 2278
+    },
+    {
+      "epoch": 2.6012549914432403,
+      "grad_norm": 0.21825598326524892,
+      "learning_rate": 7.378435517970401e-06,
+      "loss": 0.307,
+      "step": 2279
+    },
+    {
+      "epoch": 2.602395892755277,
+      "grad_norm": 0.19780618819467255,
+      "learning_rate": 7.357293868921776e-06,
+      "loss": 0.2719,
+      "step": 2280
+    },
+    {
+      "epoch": 2.6035367940673133,
+      "grad_norm": 0.20187862019079622,
+      "learning_rate": 7.336152219873151e-06,
+      "loss": 0.3001,
+      "step": 2281
+    },
+    {
+      "epoch": 2.60467769537935,
+      "grad_norm": 0.1957602418547089,
+      "learning_rate": 7.315010570824525e-06,
+      "loss": 0.2893,
+      "step": 2282
+    },
+    {
+      "epoch": 2.6058185966913863,
+      "grad_norm": 0.32010591246894465,
+      "learning_rate": 7.293868921775898e-06,
+      "loss": 0.3003,
+      "step": 2283
+    },
+    {
+      "epoch": 2.6069594980034227,
+      "grad_norm": 0.21826237399726542,
+      "learning_rate": 7.272727272727272e-06,
+      "loss": 0.2889,
+      "step": 2284
+    },
+    {
+      "epoch": 2.6081003993154592,
+      "grad_norm": 0.2106352951066305,
+      "learning_rate": 7.251585623678647e-06,
+      "loss": 0.2827,
+      "step": 2285
+    },
+    {
+      "epoch": 2.6092413006274957,
+      "grad_norm": 0.2078174915090593,
+      "learning_rate": 7.230443974630022e-06,
+      "loss": 0.2914,
+      "step": 2286
+    },
+    {
+      "epoch": 2.610382201939532,
+      "grad_norm": 0.22396488141909118,
+      "learning_rate": 7.209302325581396e-06,
+      "loss": 0.3042,
+      "step": 2287
+    },
+    {
+      "epoch": 2.6115231032515687,
+      "grad_norm": 0.2290191223253906,
+      "learning_rate": 7.1881606765327695e-06,
+      "loss": 0.2893,
+      "step": 2288
+    },
+    {
+      "epoch": 2.612664004563605,
+      "grad_norm": 0.21481339577130223,
+      "learning_rate": 7.1670190274841435e-06,
+      "loss": 0.278,
+      "step": 2289
+    },
+    {
+      "epoch": 2.6138049058756416,
+      "grad_norm": 0.22863107764420157,
+      "learning_rate": 7.1458773784355185e-06,
+      "loss": 0.2937,
+      "step": 2290
+    },
+    {
+      "epoch": 2.614945807187678,
+      "grad_norm": 0.20446333945789158,
+      "learning_rate": 7.124735729386893e-06,
+      "loss": 0.2803,
+      "step": 2291
+    },
+    {
+      "epoch": 2.616086708499715,
+      "grad_norm": 0.2139924313184036,
+      "learning_rate": 7.103594080338266e-06,
+      "loss": 0.2973,
+      "step": 2292
+    },
+    {
+      "epoch": 2.617227609811751,
+      "grad_norm": 0.2272213739485047,
+      "learning_rate": 7.082452431289641e-06,
+      "loss": 0.2915,
+      "step": 2293
+    },
+    {
+      "epoch": 2.618368511123788,
+      "grad_norm": 0.23274850175302836,
+      "learning_rate": 7.061310782241015e-06,
+      "loss": 0.2871,
+      "step": 2294
+    },
+    {
+      "epoch": 2.619509412435824,
+      "grad_norm": 0.2200811712932776,
+      "learning_rate": 7.04016913319239e-06,
+      "loss": 0.2939,
+      "step": 2295
+    },
+    {
+      "epoch": 2.620650313747861,
+      "grad_norm": 0.21688950824573724,
+      "learning_rate": 7.019027484143765e-06,
+      "loss": 0.2981,
+      "step": 2296
+    },
+    {
+      "epoch": 2.6217912150598974,
+      "grad_norm": 0.21223187386382966,
+      "learning_rate": 6.997885835095137e-06,
+      "loss": 0.2898,
+      "step": 2297
+    },
+    {
+      "epoch": 2.622932116371934,
+      "grad_norm": 0.21679812155701778,
+      "learning_rate": 6.976744186046512e-06,
+      "loss": 0.303,
+      "step": 2298
+    },
+    {
+      "epoch": 2.6240730176839704,
+      "grad_norm": 0.213840792985224,
+      "learning_rate": 6.955602536997887e-06,
+      "loss": 0.2905,
+      "step": 2299
+    },
+    {
+      "epoch": 2.625213918996007,
+      "grad_norm": 0.2324092568365301,
+      "learning_rate": 6.934460887949261e-06,
+      "loss": 0.3023,
+      "step": 2300
+    },
+    {
+      "epoch": 2.6263548203080433,
+      "grad_norm": 0.21224593855037074,
+      "learning_rate": 6.913319238900634e-06,
+      "loss": 0.2786,
+      "step": 2301
+    },
+    {
+      "epoch": 2.62749572162008,
+      "grad_norm": 0.2191741025335595,
+      "learning_rate": 6.892177589852008e-06,
+      "loss": 0.2915,
+      "step": 2302
+    },
+    {
+      "epoch": 2.6286366229321163,
+      "grad_norm": 0.2050190605945917,
+      "learning_rate": 6.871035940803383e-06,
+      "loss": 0.3005,
+      "step": 2303
+    },
+    {
+      "epoch": 2.629777524244153,
+      "grad_norm": 0.19422534480342057,
+      "learning_rate": 6.849894291754758e-06,
+      "loss": 0.2813,
+      "step": 2304
+    },
+    {
+      "epoch": 2.6309184255561893,
+      "grad_norm": 0.2024986617966428,
+      "learning_rate": 6.82875264270613e-06,
+      "loss": 0.2931,
+      "step": 2305
+    },
+    {
+      "epoch": 2.6320593268682257,
+      "grad_norm": 0.2016325798838157,
+      "learning_rate": 6.807610993657505e-06,
+      "loss": 0.2757,
+      "step": 2306
+    },
+    {
+      "epoch": 2.6332002281802627,
+      "grad_norm": 0.19900248841956134,
+      "learning_rate": 6.786469344608879e-06,
+      "loss": 0.2731,
+      "step": 2307
+    },
+    {
+      "epoch": 2.6343411294922987,
+      "grad_norm": 0.22345370720066865,
+      "learning_rate": 6.765327695560254e-06,
+      "loss": 0.2915,
+      "step": 2308
+    },
+    {
+      "epoch": 2.6354820308043356,
+      "grad_norm": 0.22172722286619925,
+      "learning_rate": 6.744186046511629e-06,
+      "loss": 0.2763,
+      "step": 2309
+    },
+    {
+      "epoch": 2.6366229321163717,
+      "grad_norm": 0.20927209846966097,
+      "learning_rate": 6.723044397463002e-06,
+      "loss": 0.3024,
+      "step": 2310
+    },
+    {
+      "epoch": 2.6377638334284086,
+      "grad_norm": 0.1995585604700486,
+      "learning_rate": 6.7019027484143765e-06,
+      "loss": 0.2986,
+      "step": 2311
+    },
+    {
+      "epoch": 2.638904734740445,
+      "grad_norm": 0.24437397232747674,
+      "learning_rate": 6.680761099365751e-06,
+      "loss": 0.2906,
+      "step": 2312
+    },
+    {
+      "epoch": 2.6400456360524815,
+      "grad_norm": 0.23672311901146992,
+      "learning_rate": 6.6596194503171255e-06,
+      "loss": 0.3059,
+      "step": 2313
+    },
+    {
+      "epoch": 2.641186537364518,
+      "grad_norm": 0.20833196620847985,
+      "learning_rate": 6.638477801268499e-06,
+      "loss": 0.2814,
+      "step": 2314
+    },
+    {
+      "epoch": 2.6423274386765545,
+      "grad_norm": 0.21171378116649958,
+      "learning_rate": 6.617336152219873e-06,
+      "loss": 0.2923,
+      "step": 2315
+    },
+    {
+      "epoch": 2.643468339988591,
+      "grad_norm": 0.21992329243569908,
+      "learning_rate": 6.596194503171248e-06,
+      "loss": 0.293,
+      "step": 2316
+    },
+    {
+      "epoch": 2.6446092413006275,
+      "grad_norm": 0.19236579441830645,
+      "learning_rate": 6.575052854122622e-06,
+      "loss": 0.2882,
+      "step": 2317
+    },
+    {
+      "epoch": 2.645750142612664,
+      "grad_norm": 0.21732952448802234,
+      "learning_rate": 6.553911205073997e-06,
+      "loss": 0.3002,
+      "step": 2318
+    },
+    {
+      "epoch": 2.6468910439247004,
+      "grad_norm": 0.21930948530716543,
+      "learning_rate": 6.53276955602537e-06,
+      "loss": 0.2837,
+      "step": 2319
+    },
+    {
+      "epoch": 2.648031945236737,
+      "grad_norm": 0.20276319027259337,
+      "learning_rate": 6.511627906976744e-06,
+      "loss": 0.2941,
+      "step": 2320
+    },
+    {
+      "epoch": 2.6491728465487734,
+      "grad_norm": 0.20664191060677842,
+      "learning_rate": 6.490486257928119e-06,
+      "loss": 0.2973,
+      "step": 2321
+    },
+    {
+      "epoch": 2.6503137478608103,
+      "grad_norm": 0.20686816856995033,
+      "learning_rate": 6.469344608879494e-06,
+      "loss": 0.2876,
+      "step": 2322
+    },
+    {
+      "epoch": 2.6514546491728463,
+      "grad_norm": 0.21791289664974559,
+      "learning_rate": 6.448202959830866e-06,
+      "loss": 0.2938,
+      "step": 2323
+    },
+    {
+      "epoch": 2.6525955504848833,
+      "grad_norm": 0.20835972672843145,
+      "learning_rate": 6.427061310782241e-06,
+      "loss": 0.2965,
+      "step": 2324
+    },
+    {
+      "epoch": 2.6537364517969193,
+      "grad_norm": 0.19895481748874236,
+      "learning_rate": 6.405919661733615e-06,
+      "loss": 0.2878,
+      "step": 2325
+    },
+    {
+      "epoch": 2.6548773531089562,
+      "grad_norm": 0.2061114756476564,
+      "learning_rate": 6.38477801268499e-06,
+      "loss": 0.2883,
+      "step": 2326
+    },
+    {
+      "epoch": 2.6560182544209927,
+      "grad_norm": 0.19673018088924696,
+      "learning_rate": 6.363636363636363e-06,
+      "loss": 0.2982,
+      "step": 2327
+    },
+    {
+      "epoch": 2.657159155733029,
+      "grad_norm": 0.20789548693413587,
+      "learning_rate": 6.3424947145877375e-06,
+      "loss": 0.283,
+      "step": 2328
+    },
+    {
+      "epoch": 2.6583000570450657,
+      "grad_norm": 0.20423004517268986,
+      "learning_rate": 6.321353065539112e-06,
+      "loss": 0.2985,
+      "step": 2329
+    },
+    {
+      "epoch": 2.659440958357102,
+      "grad_norm": 0.21369437414101625,
+      "learning_rate": 6.3002114164904865e-06,
+      "loss": 0.2951,
+      "step": 2330
+    },
+    {
+      "epoch": 2.6605818596691386,
+      "grad_norm": 0.2024038323173226,
+      "learning_rate": 6.279069767441861e-06,
+      "loss": 0.2983,
+      "step": 2331
+    },
+    {
+      "epoch": 2.661722760981175,
+      "grad_norm": 0.22749689909229562,
+      "learning_rate": 6.257928118393235e-06,
+      "loss": 0.29,
+      "step": 2332
+    },
+    {
+      "epoch": 2.6628636622932116,
+      "grad_norm": 0.2054459051238387,
+      "learning_rate": 6.236786469344609e-06,
+      "loss": 0.2975,
+      "step": 2333
+    },
+    {
+      "epoch": 2.664004563605248,
+      "grad_norm": 0.22028377224426043,
+      "learning_rate": 6.215644820295984e-06,
+      "loss": 0.2935,
+      "step": 2334
+    },
+    {
+      "epoch": 2.6651454649172845,
+      "grad_norm": 0.21504971194559508,
+      "learning_rate": 6.194503171247358e-06,
+      "loss": 0.3184,
+      "step": 2335
+    },
+    {
+      "epoch": 2.666286366229321,
+      "grad_norm": 0.2453904967082021,
+      "learning_rate": 6.173361522198732e-06,
+      "loss": 0.286,
+      "step": 2336
+    },
+    {
+      "epoch": 2.667427267541358,
+      "grad_norm": 0.21061911886303802,
+      "learning_rate": 6.152219873150106e-06,
+      "loss": 0.3064,
+      "step": 2337
+    },
+    {
+      "epoch": 2.668568168853394,
+      "grad_norm": 0.2133958177486806,
+      "learning_rate": 6.13107822410148e-06,
+      "loss": 0.2897,
+      "step": 2338
+    },
+    {
+      "epoch": 2.669709070165431,
+      "grad_norm": 0.22350309093921972,
+      "learning_rate": 6.109936575052854e-06,
+      "loss": 0.2908,
+      "step": 2339
+    },
+    {
+      "epoch": 2.670849971477467,
+      "grad_norm": 0.1994310994630131,
+      "learning_rate": 6.088794926004229e-06,
+      "loss": 0.3016,
+      "step": 2340
+    },
+    {
+      "epoch": 2.671990872789504,
+      "grad_norm": 0.242456861299583,
+      "learning_rate": 6.067653276955603e-06,
+      "loss": 0.2782,
+      "step": 2341
+    },
+    {
+      "epoch": 2.6731317741015403,
+      "grad_norm": 0.20070378596638844,
+      "learning_rate": 6.046511627906977e-06,
+      "loss": 0.275,
+      "step": 2342
+    },
+    {
+      "epoch": 2.674272675413577,
+      "grad_norm": 0.20533693117962545,
+      "learning_rate": 6.025369978858351e-06,
+      "loss": 0.2886,
+      "step": 2343
+    },
+    {
+      "epoch": 2.6754135767256133,
+      "grad_norm": 0.24120919621556927,
+      "learning_rate": 6.004228329809725e-06,
+      "loss": 0.3033,
+      "step": 2344
+    },
+    {
+      "epoch": 2.67655447803765,
+      "grad_norm": 0.2154183445042602,
+      "learning_rate": 5.9830866807611e-06,
+      "loss": 0.3066,
+      "step": 2345
+    },
+    {
+      "epoch": 2.6776953793496863,
+      "grad_norm": 0.20613740694158517,
+      "learning_rate": 5.961945031712473e-06,
+      "loss": 0.3162,
+      "step": 2346
+    },
+    {
+      "epoch": 2.6788362806617227,
+      "grad_norm": 0.22079946197964784,
+      "learning_rate": 5.940803382663848e-06,
+      "loss": 0.2847,
+      "step": 2347
+    },
+    {
+      "epoch": 2.6799771819737592,
+      "grad_norm": 0.21631086212021008,
+      "learning_rate": 5.919661733615222e-06,
+      "loss": 0.2809,
+      "step": 2348
+    },
+    {
+      "epoch": 2.6811180832857957,
+      "grad_norm": 0.22122238445645395,
+      "learning_rate": 5.898520084566596e-06,
+      "loss": 0.3071,
+      "step": 2349
+    },
+    {
+      "epoch": 2.682258984597832,
+      "grad_norm": 0.2209980769754549,
+      "learning_rate": 5.8773784355179705e-06,
+      "loss": 0.2933,
+      "step": 2350
+    },
+    {
+      "epoch": 2.6833998859098687,
+      "grad_norm": 0.20106379921531103,
+      "learning_rate": 5.8562367864693445e-06,
+      "loss": 0.2692,
+      "step": 2351
+    },
+    {
+      "epoch": 2.684540787221905,
+      "grad_norm": 0.20397362922563297,
+      "learning_rate": 5.8350951374207194e-06,
+      "loss": 0.2815,
+      "step": 2352
+    },
+    {
+      "epoch": 2.6856816885339416,
+      "grad_norm": 0.20098081864470793,
+      "learning_rate": 5.8139534883720935e-06,
+      "loss": 0.2811,
+      "step": 2353
+    },
+    {
+      "epoch": 2.6868225898459785,
+      "grad_norm": 0.19634429911239146,
+      "learning_rate": 5.792811839323468e-06,
+      "loss": 0.2754,
+      "step": 2354
+    },
+    {
+      "epoch": 2.6879634911580146,
+      "grad_norm": 0.22762609745240756,
+      "learning_rate": 5.771670190274842e-06,
+      "loss": 0.2772,
+      "step": 2355
+    },
+    {
+      "epoch": 2.6891043924700515,
+      "grad_norm": 0.19275931276843722,
+      "learning_rate": 5.750528541226216e-06,
+      "loss": 0.3034,
+      "step": 2356
+    },
+    {
+      "epoch": 2.690245293782088,
+      "grad_norm": 0.20721106140373188,
+      "learning_rate": 5.72938689217759e-06,
+      "loss": 0.2817,
+      "step": 2357
+    },
+    {
+      "epoch": 2.6913861950941245,
+      "grad_norm": 0.21585429405699674,
+      "learning_rate": 5.708245243128965e-06,
+      "loss": 0.3001,
+      "step": 2358
+    },
+    {
+      "epoch": 2.692527096406161,
+      "grad_norm": 0.22737139061335532,
+      "learning_rate": 5.687103594080338e-06,
+      "loss": 0.2791,
+      "step": 2359
+    },
+    {
+      "epoch": 2.6936679977181974,
+      "grad_norm": 0.2052614228941748,
+      "learning_rate": 5.665961945031713e-06,
+      "loss": 0.2986,
+      "step": 2360
+    },
+    {
+      "epoch": 2.694808899030234,
+      "grad_norm": 0.21375194844104548,
+      "learning_rate": 5.644820295983087e-06,
+      "loss": 0.2929,
+      "step": 2361
+    },
+    {
+      "epoch": 2.6959498003422704,
+      "grad_norm": 0.21440238679198403,
+      "learning_rate": 5.623678646934461e-06,
+      "loss": 0.2774,
+      "step": 2362
+    },
+    {
+      "epoch": 2.697090701654307,
+      "grad_norm": 0.20394690527332443,
+      "learning_rate": 5.602536997885836e-06,
+      "loss": 0.2999,
+      "step": 2363
+    },
+    {
+      "epoch": 2.6982316029663433,
+      "grad_norm": 0.22082160108841933,
+      "learning_rate": 5.581395348837209e-06,
+      "loss": 0.2891,
+      "step": 2364
+    },
+    {
+      "epoch": 2.69937250427838,
+      "grad_norm": 0.2154910615773614,
+      "learning_rate": 5.560253699788584e-06,
+      "loss": 0.2907,
+      "step": 2365
+    },
+    {
+      "epoch": 2.7005134055904163,
+      "grad_norm": 0.19966209756699427,
+      "learning_rate": 5.539112050739958e-06,
+      "loss": 0.2792,
+      "step": 2366
+    },
+    {
+      "epoch": 2.701654306902453,
+      "grad_norm": 0.20758371417711657,
+      "learning_rate": 5.517970401691332e-06,
+      "loss": 0.2772,
+      "step": 2367
+    },
+    {
+      "epoch": 2.7027952082144893,
+      "grad_norm": 0.19655111008329076,
+      "learning_rate": 5.496828752642706e-06,
+      "loss": 0.2859,
+      "step": 2368
+    },
+    {
+      "epoch": 2.703936109526526,
+      "grad_norm": 0.21344377891772437,
+      "learning_rate": 5.47568710359408e-06,
+      "loss": 0.273,
+      "step": 2369
+    },
+    {
+      "epoch": 2.7050770108385622,
+      "grad_norm": 0.19954667152573125,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 0.27,
+      "step": 2370
+    },
+    {
+      "epoch": 2.706217912150599,
+      "grad_norm": 0.20696421237106963,
+      "learning_rate": 5.433403805496829e-06,
+      "loss": 0.2741,
+      "step": 2371
+    },
+    {
+      "epoch": 2.7073588134626356,
+      "grad_norm": 0.19888310424958752,
+      "learning_rate": 5.412262156448203e-06,
+      "loss": 0.2895,
+      "step": 2372
+    },
+    {
+      "epoch": 2.708499714774672,
+      "grad_norm": 0.22935801181778648,
+      "learning_rate": 5.3911205073995775e-06,
+      "loss": 0.2893,
+      "step": 2373
+    },
+    {
+      "epoch": 2.7096406160867086,
+      "grad_norm": 0.21186470683679917,
+      "learning_rate": 5.369978858350952e-06,
+      "loss": 0.287,
+      "step": 2374
+    },
+    {
+      "epoch": 2.710781517398745,
+      "grad_norm": 0.18801028249751817,
+      "learning_rate": 5.348837209302326e-06,
+      "loss": 0.2861,
+      "step": 2375
+    },
+    {
+      "epoch": 2.7119224187107815,
+      "grad_norm": 0.2116535183061984,
+      "learning_rate": 5.327695560253701e-06,
+      "loss": 0.3056,
+      "step": 2376
+    },
+    {
+      "epoch": 2.713063320022818,
+      "grad_norm": 0.21346624281502724,
+      "learning_rate": 5.306553911205074e-06,
+      "loss": 0.2971,
+      "step": 2377
+    },
+    {
+      "epoch": 2.7142042213348545,
+      "grad_norm": 0.20321563705952317,
+      "learning_rate": 5.285412262156449e-06,
+      "loss": 0.2808,
+      "step": 2378
+    },
+    {
+      "epoch": 2.715345122646891,
+      "grad_norm": 0.2115089124503202,
+      "learning_rate": 5.264270613107823e-06,
+      "loss": 0.2885,
+      "step": 2379
+    },
+    {
+      "epoch": 2.7164860239589275,
+      "grad_norm": 0.22902298132952414,
+      "learning_rate": 5.243128964059197e-06,
+      "loss": 0.2921,
+      "step": 2380
+    },
+    {
+      "epoch": 2.717626925270964,
+      "grad_norm": 0.20349307589386711,
+      "learning_rate": 5.221987315010571e-06,
+      "loss": 0.2846,
+      "step": 2381
+    },
+    {
+      "epoch": 2.7187678265830004,
+      "grad_norm": 0.18793703684238383,
+      "learning_rate": 5.200845665961945e-06,
+      "loss": 0.2935,
+      "step": 2382
+    },
+    {
+      "epoch": 2.719908727895037,
+      "grad_norm": 0.2519734868072542,
+      "learning_rate": 5.179704016913319e-06,
+      "loss": 0.2838,
+      "step": 2383
+    },
+    {
+      "epoch": 2.721049629207074,
+      "grad_norm": 0.197819968737116,
+      "learning_rate": 5.158562367864694e-06,
+      "loss": 0.3027,
+      "step": 2384
+    },
+    {
+      "epoch": 2.72219053051911,
+      "grad_norm": 0.2086766984476139,
+      "learning_rate": 5.137420718816068e-06,
+      "loss": 0.2865,
+      "step": 2385
+    },
+    {
+      "epoch": 2.723331431831147,
+      "grad_norm": 0.19345032801304457,
+      "learning_rate": 5.116279069767442e-06,
+      "loss": 0.2973,
+      "step": 2386
+    },
+    {
+      "epoch": 2.7244723331431833,
+      "grad_norm": 0.22577042718511026,
+      "learning_rate": 5.095137420718816e-06,
+      "loss": 0.2906,
+      "step": 2387
+    },
+    {
+      "epoch": 2.7256132344552197,
+      "grad_norm": 0.21603357817376984,
+      "learning_rate": 5.07399577167019e-06,
+      "loss": 0.2807,
+      "step": 2388
+    },
+    {
+      "epoch": 2.7267541357672562,
+      "grad_norm": 0.21219120304091335,
+      "learning_rate": 5.052854122621565e-06,
+      "loss": 0.291,
+      "step": 2389
+    },
+    {
+      "epoch": 2.7278950370792927,
+      "grad_norm": 0.1942871235301068,
+      "learning_rate": 5.0317124735729385e-06,
+      "loss": 0.2863,
+      "step": 2390
+    },
+    {
+      "epoch": 2.729035938391329,
+      "grad_norm": 0.2047396916851407,
+      "learning_rate": 5.010570824524313e-06,
+      "loss": 0.2834,
+      "step": 2391
+    },
+    {
+      "epoch": 2.7301768397033657,
+      "grad_norm": 0.19544577571703675,
+      "learning_rate": 4.9894291754756874e-06,
+      "loss": 0.2985,
+      "step": 2392
+    },
+    {
+      "epoch": 2.731317741015402,
+      "grad_norm": 0.20598080382300346,
+      "learning_rate": 4.9682875264270615e-06,
+      "loss": 0.28,
+      "step": 2393
+    },
+    {
+      "epoch": 2.7324586423274386,
+      "grad_norm": 0.19446571137719146,
+      "learning_rate": 4.947145877378436e-06,
+      "loss": 0.2749,
+      "step": 2394
+    },
+    {
+      "epoch": 2.733599543639475,
+      "grad_norm": 0.19509428149486063,
+      "learning_rate": 4.92600422832981e-06,
+      "loss": 0.309,
+      "step": 2395
+    },
+    {
+      "epoch": 2.7347404449515116,
+      "grad_norm": 0.2091106077639519,
+      "learning_rate": 4.904862579281185e-06,
+      "loss": 0.268,
+      "step": 2396
+    },
+    {
+      "epoch": 2.735881346263548,
+      "grad_norm": 0.19395912384493028,
+      "learning_rate": 4.883720930232559e-06,
+      "loss": 0.2749,
+      "step": 2397
+    },
+    {
+      "epoch": 2.7370222475755845,
+      "grad_norm": 0.2128904563883501,
+      "learning_rate": 4.862579281183933e-06,
+      "loss": 0.3169,
+      "step": 2398
+    },
+    {
+      "epoch": 2.7381631488876215,
+      "grad_norm": 0.2120146570263807,
+      "learning_rate": 4.841437632135307e-06,
+      "loss": 0.2999,
+      "step": 2399
+    },
+    {
+      "epoch": 2.7393040501996575,
+      "grad_norm": 0.21299550942100914,
+      "learning_rate": 4.820295983086681e-06,
+      "loss": 0.2803,
+      "step": 2400
+    },
+    {
+      "epoch": 2.7404449515116944,
+      "grad_norm": 0.20165472101507262,
+      "learning_rate": 4.799154334038055e-06,
+      "loss": 0.2741,
+      "step": 2401
+    },
+    {
+      "epoch": 2.741585852823731,
+      "grad_norm": 0.20587824908237448,
+      "learning_rate": 4.77801268498943e-06,
+      "loss": 0.3079,
+      "step": 2402
+    },
+    {
+      "epoch": 2.7427267541357674,
+      "grad_norm": 0.20745487685308733,
+      "learning_rate": 4.756871035940803e-06,
+      "loss": 0.2988,
+      "step": 2403
+    },
+    {
+      "epoch": 2.743867655447804,
+      "grad_norm": 0.19662301496766177,
+      "learning_rate": 4.735729386892178e-06,
+      "loss": 0.2895,
+      "step": 2404
+    },
+    {
+      "epoch": 2.7450085567598403,
+      "grad_norm": 0.2047896485844238,
+      "learning_rate": 4.714587737843552e-06,
+      "loss": 0.2908,
+      "step": 2405
+    },
+    {
+      "epoch": 2.746149458071877,
+      "grad_norm": 0.22075407121101182,
+      "learning_rate": 4.693446088794926e-06,
+      "loss": 0.2924,
+      "step": 2406
+    },
+    {
+      "epoch": 2.7472903593839133,
+      "grad_norm": 0.20989441506128534,
+      "learning_rate": 4.672304439746301e-06,
+      "loss": 0.2752,
+      "step": 2407
+    },
+    {
+      "epoch": 2.74843126069595,
+      "grad_norm": 0.19330236930243655,
+      "learning_rate": 4.651162790697674e-06,
+      "loss": 0.2873,
+      "step": 2408
+    },
+    {
+      "epoch": 2.7495721620079863,
+      "grad_norm": 0.21471432313236352,
+      "learning_rate": 4.630021141649049e-06,
+      "loss": 0.2984,
+      "step": 2409
+    },
+    {
+      "epoch": 2.7507130633200227,
+      "grad_norm": 0.1990241754745363,
+      "learning_rate": 4.608879492600423e-06,
+      "loss": 0.3024,
+      "step": 2410
+    },
+    {
+      "epoch": 2.7518539646320592,
+      "grad_norm": 0.2175123910831802,
+      "learning_rate": 4.587737843551797e-06,
+      "loss": 0.2862,
+      "step": 2411
+    },
+    {
+      "epoch": 2.7529948659440957,
+      "grad_norm": 0.2022442697895259,
+      "learning_rate": 4.5665961945031714e-06,
+      "loss": 0.3061,
+      "step": 2412
+    },
+    {
+      "epoch": 2.754135767256132,
+      "grad_norm": 0.202210410401036,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.3099,
+      "step": 2413
+    },
+    {
+      "epoch": 2.755276668568169,
+      "grad_norm": 0.20841630948380577,
+      "learning_rate": 4.52431289640592e-06,
+      "loss": 0.2872,
+      "step": 2414
+    },
+    {
+      "epoch": 2.756417569880205,
+      "grad_norm": 0.20962608268730534,
+      "learning_rate": 4.5031712473572945e-06,
+      "loss": 0.2799,
+      "step": 2415
+    },
+    {
+      "epoch": 2.757558471192242,
+      "grad_norm": 0.20684129654773203,
+      "learning_rate": 4.482029598308668e-06,
+      "loss": 0.2755,
+      "step": 2416
+    },
+    {
+      "epoch": 2.7586993725042785,
+      "grad_norm": 0.20412423424441156,
+      "learning_rate": 4.460887949260043e-06,
+      "loss": 0.2872,
+      "step": 2417
+    },
+    {
+      "epoch": 2.759840273816315,
+      "grad_norm": 0.21520346828361941,
+      "learning_rate": 4.439746300211417e-06,
+      "loss": 0.2998,
+      "step": 2418
+    },
+    {
+      "epoch": 2.7609811751283515,
+      "grad_norm": 0.21069047843020636,
+      "learning_rate": 4.418604651162791e-06,
+      "loss": 0.2838,
+      "step": 2419
+    },
+    {
+      "epoch": 2.762122076440388,
+      "grad_norm": 0.2097851230483762,
+      "learning_rate": 4.397463002114166e-06,
+      "loss": 0.3005,
+      "step": 2420
+    },
+    {
+      "epoch": 2.7632629777524245,
+      "grad_norm": 0.1990108557800723,
+      "learning_rate": 4.376321353065539e-06,
+      "loss": 0.2814,
+      "step": 2421
+    },
+    {
+      "epoch": 2.764403879064461,
+      "grad_norm": 0.19641216914165438,
+      "learning_rate": 4.355179704016914e-06,
+      "loss": 0.2813,
+      "step": 2422
+    },
+    {
+      "epoch": 2.7655447803764974,
+      "grad_norm": 0.21405575621485598,
+      "learning_rate": 4.334038054968288e-06,
+      "loss": 0.2878,
+      "step": 2423
+    },
+    {
+      "epoch": 2.766685681688534,
+      "grad_norm": 0.21401069015463817,
+      "learning_rate": 4.312896405919662e-06,
+      "loss": 0.2894,
+      "step": 2424
+    },
+    {
+      "epoch": 2.7678265830005704,
+      "grad_norm": 0.22891323803035088,
+      "learning_rate": 4.291754756871036e-06,
+      "loss": 0.3081,
+      "step": 2425
+    },
+    {
+      "epoch": 2.768967484312607,
+      "grad_norm": 0.2109649751414716,
+      "learning_rate": 4.27061310782241e-06,
+      "loss": 0.2908,
+      "step": 2426
+    },
+    {
+      "epoch": 2.7701083856246433,
+      "grad_norm": 0.21252536464308053,
+      "learning_rate": 4.249471458773784e-06,
+      "loss": 0.2936,
+      "step": 2427
+    },
+    {
+      "epoch": 2.77124928693668,
+      "grad_norm": 0.19865252632831235,
+      "learning_rate": 4.228329809725159e-06,
+      "loss": 0.2954,
+      "step": 2428
+    },
+    {
+      "epoch": 2.7723901882487167,
+      "grad_norm": 0.2153522039069563,
+      "learning_rate": 4.207188160676533e-06,
+      "loss": 0.2974,
+      "step": 2429
+    },
+    {
+      "epoch": 2.773531089560753,
+      "grad_norm": 0.21887123860802177,
+      "learning_rate": 4.186046511627907e-06,
+      "loss": 0.2843,
+      "step": 2430
+    },
+    {
+      "epoch": 2.7746719908727897,
+      "grad_norm": 0.21186766551947683,
+      "learning_rate": 4.164904862579281e-06,
+      "loss": 0.2828,
+      "step": 2431
+    },
+    {
+      "epoch": 2.7758128921848257,
+      "grad_norm": 0.19989097612840076,
+      "learning_rate": 4.1437632135306554e-06,
+      "loss": 0.2701,
+      "step": 2432
+    },
+    {
+      "epoch": 2.7769537934968627,
+      "grad_norm": 0.18958199905520784,
+      "learning_rate": 4.12262156448203e-06,
+      "loss": 0.2759,
+      "step": 2433
+    },
+    {
+      "epoch": 2.778094694808899,
+      "grad_norm": 0.20469102511481307,
+      "learning_rate": 4.101479915433404e-06,
+      "loss": 0.2927,
+      "step": 2434
+    },
+    {
+      "epoch": 2.7792355961209356,
+      "grad_norm": 0.2045040607399336,
+      "learning_rate": 4.0803382663847785e-06,
+      "loss": 0.2808,
+      "step": 2435
+    },
+    {
+      "epoch": 2.780376497432972,
+      "grad_norm": 0.20416663418050893,
+      "learning_rate": 4.059196617336153e-06,
+      "loss": 0.3044,
+      "step": 2436
+    },
+    {
+      "epoch": 2.7815173987450086,
+      "grad_norm": 0.21126468810168858,
+      "learning_rate": 4.038054968287527e-06,
+      "loss": 0.2864,
+      "step": 2437
+    },
+    {
+      "epoch": 2.782658300057045,
+      "grad_norm": 0.21333692375010985,
+      "learning_rate": 4.016913319238901e-06,
+      "loss": 0.3066,
+      "step": 2438
+    },
+    {
+      "epoch": 2.7837992013690815,
+      "grad_norm": 0.2063409848079501,
+      "learning_rate": 3.995771670190275e-06,
+      "loss": 0.2901,
+      "step": 2439
+    },
+    {
+      "epoch": 2.784940102681118,
+      "grad_norm": 0.21945699246555966,
+      "learning_rate": 3.97463002114165e-06,
+      "loss": 0.292,
+      "step": 2440
+    },
+    {
+      "epoch": 2.7860810039931545,
+      "grad_norm": 0.21272225883141066,
+      "learning_rate": 3.953488372093024e-06,
+      "loss": 0.289,
+      "step": 2441
+    },
+    {
+      "epoch": 2.787221905305191,
+      "grad_norm": 0.1998651488460456,
+      "learning_rate": 3.932346723044398e-06,
+      "loss": 0.3048,
+      "step": 2442
+    },
+    {
+      "epoch": 2.7883628066172275,
+      "grad_norm": 0.21027926042200606,
+      "learning_rate": 3.911205073995772e-06,
+      "loss": 0.2788,
+      "step": 2443
+    },
+    {
+      "epoch": 2.7895037079292644,
+      "grad_norm": 0.23160367083032785,
+      "learning_rate": 3.890063424947146e-06,
+      "loss": 0.2791,
+      "step": 2444
+    },
+    {
+      "epoch": 2.7906446092413004,
+      "grad_norm": 0.21688813469461285,
+      "learning_rate": 3.86892177589852e-06,
+      "loss": 0.2808,
+      "step": 2445
+    },
+    {
+      "epoch": 2.7917855105533373,
+      "grad_norm": 0.20893812255772268,
+      "learning_rate": 3.847780126849895e-06,
+      "loss": 0.2939,
+      "step": 2446
+    },
+    {
+      "epoch": 2.7929264118653734,
+      "grad_norm": 0.21161195500726673,
+      "learning_rate": 3.826638477801268e-06,
+      "loss": 0.2982,
+      "step": 2447
+    },
+    {
+      "epoch": 2.7940673131774103,
+      "grad_norm": 0.22159348068245183,
+      "learning_rate": 3.805496828752643e-06,
+      "loss": 0.2909,
+      "step": 2448
+    },
+    {
+      "epoch": 2.795208214489447,
+      "grad_norm": 0.19693523939694024,
+      "learning_rate": 3.784355179704017e-06,
+      "loss": 0.3072,
+      "step": 2449
+    },
+    {
+      "epoch": 2.7963491158014833,
+      "grad_norm": 0.20541335027371957,
+      "learning_rate": 3.7632135306553913e-06,
+      "loss": 0.29,
+      "step": 2450
+    },
+    {
+      "epoch": 2.7974900171135197,
+      "grad_norm": 0.2021310692826138,
+      "learning_rate": 3.742071881606766e-06,
+      "loss": 0.2932,
+      "step": 2451
+    },
+    {
+      "epoch": 2.7986309184255562,
+      "grad_norm": 0.20551180712915548,
+      "learning_rate": 3.72093023255814e-06,
+      "loss": 0.2958,
+      "step": 2452
+    },
+    {
+      "epoch": 2.7997718197375927,
+      "grad_norm": 0.2171587485859702,
+      "learning_rate": 3.6997885835095144e-06,
+      "loss": 0.2901,
+      "step": 2453
+    },
+    {
+      "epoch": 2.800912721049629,
+      "grad_norm": 0.19791283797515585,
+      "learning_rate": 3.678646934460888e-06,
+      "loss": 0.292,
+      "step": 2454
+    },
+    {
+      "epoch": 2.8020536223616657,
+      "grad_norm": 0.1989943243503315,
+      "learning_rate": 3.6575052854122625e-06,
+      "loss": 0.2921,
+      "step": 2455
+    },
+    {
+      "epoch": 2.803194523673702,
+      "grad_norm": 0.20434592117335035,
+      "learning_rate": 3.636363636363636e-06,
+      "loss": 0.3064,
+      "step": 2456
+    },
+    {
+      "epoch": 2.8043354249857386,
+      "grad_norm": 0.2084756469685634,
+      "learning_rate": 3.615221987315011e-06,
+      "loss": 0.2957,
+      "step": 2457
+    },
+    {
+      "epoch": 2.805476326297775,
+      "grad_norm": 0.20296767278473282,
+      "learning_rate": 3.5940803382663847e-06,
+      "loss": 0.2849,
+      "step": 2458
+    },
+    {
+      "epoch": 2.806617227609812,
+      "grad_norm": 0.19379089522361004,
+      "learning_rate": 3.5729386892177592e-06,
+      "loss": 0.306,
+      "step": 2459
+    },
+    {
+      "epoch": 2.807758128921848,
+      "grad_norm": 0.2114373361742132,
+      "learning_rate": 3.551797040169133e-06,
+      "loss": 0.2811,
+      "step": 2460
+    },
+    {
+      "epoch": 2.808899030233885,
+      "grad_norm": 0.19648544797014308,
+      "learning_rate": 3.5306553911205074e-06,
+      "loss": 0.2903,
+      "step": 2461
+    },
+    {
+      "epoch": 2.810039931545921,
+      "grad_norm": 0.19136569333549805,
+      "learning_rate": 3.5095137420718823e-06,
+      "loss": 0.2896,
+      "step": 2462
+    },
+    {
+      "epoch": 2.811180832857958,
+      "grad_norm": 3.255184569034401,
+      "learning_rate": 3.488372093023256e-06,
+      "loss": 0.3308,
+      "step": 2463
+    },
+    {
+      "epoch": 2.8123217341699944,
+      "grad_norm": 0.23148723562030504,
+      "learning_rate": 3.4672304439746304e-06,
+      "loss": 0.2929,
+      "step": 2464
+    },
+    {
+      "epoch": 2.813462635482031,
+      "grad_norm": 0.19251575021641812,
+      "learning_rate": 3.446088794926004e-06,
+      "loss": 0.2816,
+      "step": 2465
+    },
+    {
+      "epoch": 2.8146035367940674,
+      "grad_norm": 0.20358606861072642,
+      "learning_rate": 3.424947145877379e-06,
+      "loss": 0.2809,
+      "step": 2466
+    },
+    {
+      "epoch": 2.815744438106104,
+      "grad_norm": 0.2525881711594538,
+      "learning_rate": 3.4038054968287527e-06,
+      "loss": 0.2799,
+      "step": 2467
+    },
+    {
+      "epoch": 2.8168853394181403,
+      "grad_norm": 0.19051484999821816,
+      "learning_rate": 3.382663847780127e-06,
+      "loss": 0.2996,
+      "step": 2468
+    },
+    {
+      "epoch": 2.818026240730177,
+      "grad_norm": 0.21665936441766315,
+      "learning_rate": 3.361522198731501e-06,
+      "loss": 0.3008,
+      "step": 2469
+    },
+    {
+      "epoch": 2.8191671420422133,
+      "grad_norm": 0.21141505886515194,
+      "learning_rate": 3.3403805496828753e-06,
+      "loss": 0.2803,
+      "step": 2470
+    },
+    {
+      "epoch": 2.82030804335425,
+      "grad_norm": 0.2013970938079833,
+      "learning_rate": 3.3192389006342494e-06,
+      "loss": 0.2894,
+      "step": 2471
+    },
+    {
+      "epoch": 2.8214489446662863,
+      "grad_norm": 0.19323798161487787,
+      "learning_rate": 3.298097251585624e-06,
+      "loss": 0.2864,
+      "step": 2472
+    },
+    {
+      "epoch": 2.8225898459783227,
+      "grad_norm": 0.201226274752575,
+      "learning_rate": 3.2769556025369984e-06,
+      "loss": 0.3071,
+      "step": 2473
+    },
+    {
+      "epoch": 2.823730747290359,
+      "grad_norm": 0.2074222646039902,
+      "learning_rate": 3.255813953488372e-06,
+      "loss": 0.3135,
+      "step": 2474
+    },
+    {
+      "epoch": 2.8248716486023957,
+      "grad_norm": 0.20696506174071955,
+      "learning_rate": 3.234672304439747e-06,
+      "loss": 0.2734,
+      "step": 2475
+    },
+    {
+      "epoch": 2.8260125499144326,
+      "grad_norm": 0.19955249590119115,
+      "learning_rate": 3.2135306553911206e-06,
+      "loss": 0.2826,
+      "step": 2476
+    },
+    {
+      "epoch": 2.8271534512264687,
+      "grad_norm": 0.2156852402716923,
+      "learning_rate": 3.192389006342495e-06,
+      "loss": 0.3096,
+      "step": 2477
+    },
+    {
+      "epoch": 2.8282943525385056,
+      "grad_norm": 0.20841607177541718,
+      "learning_rate": 3.1712473572938687e-06,
+      "loss": 0.2849,
+      "step": 2478
+    },
+    {
+      "epoch": 2.829435253850542,
+      "grad_norm": 0.1941983020977351,
+      "learning_rate": 3.1501057082452432e-06,
+      "loss": 0.278,
+      "step": 2479
+    },
+    {
+      "epoch": 2.8305761551625785,
+      "grad_norm": 0.20869712403007026,
+      "learning_rate": 3.1289640591966173e-06,
+      "loss": 0.2873,
+      "step": 2480
+    },
+    {
+      "epoch": 2.831717056474615,
+      "grad_norm": 0.20706120097683225,
+      "learning_rate": 3.107822410147992e-06,
+      "loss": 0.2833,
+      "step": 2481
+    },
+    {
+      "epoch": 2.8328579577866515,
+      "grad_norm": 0.1982144805085283,
+      "learning_rate": 3.086680761099366e-06,
+      "loss": 0.2942,
+      "step": 2482
+    },
+    {
+      "epoch": 2.833998859098688,
+      "grad_norm": 0.20975348933509538,
+      "learning_rate": 3.06553911205074e-06,
+      "loss": 0.2797,
+      "step": 2483
+    },
+    {
+      "epoch": 2.8351397604107245,
+      "grad_norm": 0.21647387880347907,
+      "learning_rate": 3.0443974630021144e-06,
+      "loss": 0.2802,
+      "step": 2484
+    },
+    {
+      "epoch": 2.836280661722761,
+      "grad_norm": 0.20027142878455498,
+      "learning_rate": 3.0232558139534885e-06,
+      "loss": 0.2901,
+      "step": 2485
+    },
+    {
+      "epoch": 2.8374215630347974,
+      "grad_norm": 0.21680907668024518,
+      "learning_rate": 3.0021141649048626e-06,
+      "loss": 0.2915,
+      "step": 2486
+    },
+    {
+      "epoch": 2.838562464346834,
+      "grad_norm": 0.21538879323186988,
+      "learning_rate": 2.9809725158562367e-06,
+      "loss": 0.3016,
+      "step": 2487
+    },
+    {
+      "epoch": 2.8397033656588704,
+      "grad_norm": 0.21981713377113177,
+      "learning_rate": 2.959830866807611e-06,
+      "loss": 0.2818,
+      "step": 2488
+    },
+    {
+      "epoch": 2.840844266970907,
+      "grad_norm": 0.19697892124577174,
+      "learning_rate": 2.9386892177589852e-06,
+      "loss": 0.2892,
+      "step": 2489
+    },
+    {
+      "epoch": 2.8419851682829433,
+      "grad_norm": 0.18909864372788007,
+      "learning_rate": 2.9175475687103597e-06,
+      "loss": 0.2976,
+      "step": 2490
+    },
+    {
+      "epoch": 2.8431260695949803,
+      "grad_norm": 0.18454775144527993,
+      "learning_rate": 2.896405919661734e-06,
+      "loss": 0.3028,
+      "step": 2491
+    },
+    {
+      "epoch": 2.8442669709070163,
+      "grad_norm": 0.2310587393078742,
+      "learning_rate": 2.875264270613108e-06,
+      "loss": 0.298,
+      "step": 2492
+    },
+    {
+      "epoch": 2.8454078722190532,
+      "grad_norm": 0.21248087744098798,
+      "learning_rate": 2.8541226215644824e-06,
+      "loss": 0.29,
+      "step": 2493
+    },
+    {
+      "epoch": 2.8465487735310897,
+      "grad_norm": 0.19507301929030554,
+      "learning_rate": 2.8329809725158564e-06,
+      "loss": 0.2799,
+      "step": 2494
+    },
+    {
+      "epoch": 2.847689674843126,
+      "grad_norm": 0.19860147238212592,
+      "learning_rate": 2.8118393234672305e-06,
+      "loss": 0.2867,
+      "step": 2495
+    },
+    {
+      "epoch": 2.8488305761551627,
+      "grad_norm": 0.19602116174876164,
+      "learning_rate": 2.7906976744186046e-06,
+      "loss": 0.2984,
+      "step": 2496
+    },
+    {
+      "epoch": 2.849971477467199,
+      "grad_norm": 0.19553454209538595,
+      "learning_rate": 2.769556025369979e-06,
+      "loss": 0.2972,
+      "step": 2497
+    },
+    {
+      "epoch": 2.8511123787792356,
+      "grad_norm": 0.2040292670050931,
+      "learning_rate": 2.748414376321353e-06,
+      "loss": 0.3003,
+      "step": 2498
+    },
+    {
+      "epoch": 2.852253280091272,
+      "grad_norm": 0.20204674645438503,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 0.3008,
+      "step": 2499
+    },
+    {
+      "epoch": 2.8533941814033086,
+      "grad_norm": 0.19934424552625263,
+      "learning_rate": 2.7061310782241013e-06,
+      "loss": 0.2881,
+      "step": 2500
+    },
+    {
+      "epoch": 2.854535082715345,
+      "grad_norm": 0.18785474732210933,
+      "learning_rate": 2.684989429175476e-06,
+      "loss": 0.2794,
+      "step": 2501
+    },
+    {
+      "epoch": 2.8556759840273815,
+      "grad_norm": 0.20006434112569327,
+      "learning_rate": 2.6638477801268503e-06,
+      "loss": 0.2964,
+      "step": 2502
+    },
+    {
+      "epoch": 2.856816885339418,
+      "grad_norm": 0.1977342860009862,
+      "learning_rate": 2.6427061310782244e-06,
+      "loss": 0.2845,
+      "step": 2503
+    },
+    {
+      "epoch": 2.8579577866514545,
+      "grad_norm": 0.20632919899960922,
+      "learning_rate": 2.6215644820295984e-06,
+      "loss": 0.2765,
+      "step": 2504
+    },
+    {
+      "epoch": 2.859098687963491,
+      "grad_norm": 0.21051236583497268,
+      "learning_rate": 2.6004228329809725e-06,
+      "loss": 0.305,
+      "step": 2505
+    },
+    {
+      "epoch": 2.860239589275528,
+      "grad_norm": 0.20023237671477812,
+      "learning_rate": 2.579281183932347e-06,
+      "loss": 0.274,
+      "step": 2506
+    },
+    {
+      "epoch": 2.861380490587564,
+      "grad_norm": 0.1828118884191061,
+      "learning_rate": 2.558139534883721e-06,
+      "loss": 0.3021,
+      "step": 2507
+    },
+    {
+      "epoch": 2.862521391899601,
+      "grad_norm": 0.20392801822251036,
+      "learning_rate": 2.536997885835095e-06,
+      "loss": 0.2843,
+      "step": 2508
+    },
+    {
+      "epoch": 2.8636622932116373,
+      "grad_norm": 0.20830765542926766,
+      "learning_rate": 2.5158562367864692e-06,
+      "loss": 0.2953,
+      "step": 2509
+    },
+    {
+      "epoch": 2.864803194523674,
+      "grad_norm": 0.2120174031270726,
+      "learning_rate": 2.4947145877378437e-06,
+      "loss": 0.2805,
+      "step": 2510
+    },
+    {
+      "epoch": 2.8659440958357103,
+      "grad_norm": 0.1945197850552631,
+      "learning_rate": 2.473572938689218e-06,
+      "loss": 0.289,
+      "step": 2511
+    },
+    {
+      "epoch": 2.867084997147747,
+      "grad_norm": 0.19141709513826796,
+      "learning_rate": 2.4524312896405923e-06,
+      "loss": 0.3101,
+      "step": 2512
+    },
+    {
+      "epoch": 2.8682258984597833,
+      "grad_norm": 0.21186283438035136,
+      "learning_rate": 2.4312896405919664e-06,
+      "loss": 0.2898,
+      "step": 2513
+    },
+    {
+      "epoch": 2.8693667997718197,
+      "grad_norm": 0.1945697871372187,
+      "learning_rate": 2.4101479915433404e-06,
+      "loss": 0.2907,
+      "step": 2514
+    },
+    {
+      "epoch": 2.8705077010838562,
+      "grad_norm": 0.19475854706319662,
+      "learning_rate": 2.389006342494715e-06,
+      "loss": 0.2913,
+      "step": 2515
+    },
+    {
+      "epoch": 2.8716486023958927,
+      "grad_norm": 0.19694923949612048,
+      "learning_rate": 2.367864693446089e-06,
+      "loss": 0.2989,
+      "step": 2516
+    },
+    {
+      "epoch": 2.872789503707929,
+      "grad_norm": 0.20020422907016433,
+      "learning_rate": 2.346723044397463e-06,
+      "loss": 0.2843,
+      "step": 2517
+    },
+    {
+      "epoch": 2.8739304050199657,
+      "grad_norm": 0.20878342093258945,
+      "learning_rate": 2.325581395348837e-06,
+      "loss": 0.283,
+      "step": 2518
+    },
+    {
+      "epoch": 2.875071306332002,
+      "grad_norm": 0.19905491086930102,
+      "learning_rate": 2.3044397463002116e-06,
+      "loss": 0.3028,
+      "step": 2519
+    },
+    {
+      "epoch": 2.8762122076440386,
+      "grad_norm": 0.19423780902965673,
+      "learning_rate": 2.2832980972515857e-06,
+      "loss": 0.2761,
+      "step": 2520
+    },
+    {
+      "epoch": 2.8773531089560755,
+      "grad_norm": 0.2002744524607323,
+      "learning_rate": 2.26215644820296e-06,
+      "loss": 0.2836,
+      "step": 2521
+    },
+    {
+      "epoch": 2.8784940102681116,
+      "grad_norm": 0.1989239278844251,
+      "learning_rate": 2.241014799154334e-06,
+      "loss": 0.3083,
+      "step": 2522
+    },
+    {
+      "epoch": 2.8796349115801485,
+      "grad_norm": 0.21159960733423624,
+      "learning_rate": 2.2198731501057084e-06,
+      "loss": 0.2943,
+      "step": 2523
+    },
+    {
+      "epoch": 2.880775812892185,
+      "grad_norm": 0.18973753627567133,
+      "learning_rate": 2.198731501057083e-06,
+      "loss": 0.2857,
+      "step": 2524
+    },
+    {
+      "epoch": 2.8819167142042215,
+      "grad_norm": 0.20692276429512232,
+      "learning_rate": 2.177589852008457e-06,
+      "loss": 0.3059,
+      "step": 2525
+    },
+    {
+      "epoch": 2.883057615516258,
+      "grad_norm": 0.20607858347875038,
+      "learning_rate": 2.156448202959831e-06,
+      "loss": 0.2748,
+      "step": 2526
+    },
+    {
+      "epoch": 2.8841985168282944,
+      "grad_norm": 0.19766789002244206,
+      "learning_rate": 2.135306553911205e-06,
+      "loss": 0.2864,
+      "step": 2527
+    },
+    {
+      "epoch": 2.885339418140331,
+      "grad_norm": 0.20508355580996102,
+      "learning_rate": 2.1141649048625796e-06,
+      "loss": 0.2825,
+      "step": 2528
+    },
+    {
+      "epoch": 2.8864803194523674,
+      "grad_norm": 0.21621004836804458,
+      "learning_rate": 2.0930232558139536e-06,
+      "loss": 0.2939,
+      "step": 2529
+    },
+    {
+      "epoch": 2.887621220764404,
+      "grad_norm": 0.20185327841415296,
+      "learning_rate": 2.0718816067653277e-06,
+      "loss": 0.2839,
+      "step": 2530
+    },
+    {
+      "epoch": 2.8887621220764403,
+      "grad_norm": 0.1933977567661312,
+      "learning_rate": 2.050739957716702e-06,
+      "loss": 0.3008,
+      "step": 2531
+    },
+    {
+      "epoch": 2.889903023388477,
+      "grad_norm": 0.2089992242477195,
+      "learning_rate": 2.0295983086680763e-06,
+      "loss": 0.2959,
+      "step": 2532
+    },
+    {
+      "epoch": 2.8910439247005133,
+      "grad_norm": 0.2030088255511792,
+      "learning_rate": 2.0084566596194504e-06,
+      "loss": 0.2766,
+      "step": 2533
+    },
+    {
+      "epoch": 2.89218482601255,
+      "grad_norm": 0.20334283930171593,
+      "learning_rate": 1.987315010570825e-06,
+      "loss": 0.2758,
+      "step": 2534
+    },
+    {
+      "epoch": 2.8933257273245863,
+      "grad_norm": 0.19141292320551842,
+      "learning_rate": 1.966173361522199e-06,
+      "loss": 0.2863,
+      "step": 2535
+    },
+    {
+      "epoch": 2.894466628636623,
+      "grad_norm": 0.20218732073147072,
+      "learning_rate": 1.945031712473573e-06,
+      "loss": 0.28,
+      "step": 2536
+    },
+    {
+      "epoch": 2.895607529948659,
+      "grad_norm": 0.1918627608925985,
+      "learning_rate": 1.9238900634249475e-06,
+      "loss": 0.2948,
+      "step": 2537
+    },
+    {
+      "epoch": 2.896748431260696,
+      "grad_norm": 0.19060330399001432,
+      "learning_rate": 1.9027484143763216e-06,
+      "loss": 0.2737,
+      "step": 2538
+    },
+    {
+      "epoch": 2.8978893325727326,
+      "grad_norm": 0.18229171529313887,
+      "learning_rate": 1.8816067653276956e-06,
+      "loss": 0.2916,
+      "step": 2539
+    },
+    {
+      "epoch": 2.899030233884769,
+      "grad_norm": 0.20382857407887608,
+      "learning_rate": 1.86046511627907e-06,
+      "loss": 0.2862,
+      "step": 2540
+    },
+    {
+      "epoch": 2.9001711351968056,
+      "grad_norm": 0.1979576770452718,
+      "learning_rate": 1.839323467230444e-06,
+      "loss": 0.2855,
+      "step": 2541
+    },
+    {
+      "epoch": 2.901312036508842,
+      "grad_norm": 0.23856764466276376,
+      "learning_rate": 1.818181818181818e-06,
+      "loss": 0.2861,
+      "step": 2542
+    },
+    {
+      "epoch": 2.9024529378208785,
+      "grad_norm": 0.18177042300595342,
+      "learning_rate": 1.7970401691331924e-06,
+      "loss": 0.2875,
+      "step": 2543
+    },
+    {
+      "epoch": 2.903593839132915,
+      "grad_norm": 0.20699386822550767,
+      "learning_rate": 1.7758985200845664e-06,
+      "loss": 0.292,
+      "step": 2544
+    },
+    {
+      "epoch": 2.9047347404449515,
+      "grad_norm": 0.1919349320022479,
+      "learning_rate": 1.7547568710359411e-06,
+      "loss": 0.294,
+      "step": 2545
+    },
+    {
+      "epoch": 2.905875641756988,
+      "grad_norm": 0.21020672149500216,
+      "learning_rate": 1.7336152219873152e-06,
+      "loss": 0.2909,
+      "step": 2546
+    },
+    {
+      "epoch": 2.9070165430690245,
+      "grad_norm": 0.19606757753191717,
+      "learning_rate": 1.7124735729386895e-06,
+      "loss": 0.2933,
+      "step": 2547
+    },
+    {
+      "epoch": 2.908157444381061,
+      "grad_norm": 0.21670306350149213,
+      "learning_rate": 1.6913319238900636e-06,
+      "loss": 0.2977,
+      "step": 2548
+    },
+    {
+      "epoch": 2.9092983456930974,
+      "grad_norm": 0.20976365996925933,
+      "learning_rate": 1.6701902748414376e-06,
+      "loss": 0.2873,
+      "step": 2549
+    },
+    {
+      "epoch": 2.910439247005134,
+      "grad_norm": 0.20449092038401195,
+      "learning_rate": 1.649048625792812e-06,
+      "loss": 0.3063,
+      "step": 2550
+    },
+    {
+      "epoch": 2.911580148317171,
+      "grad_norm": 0.19231918973044662,
+      "learning_rate": 1.627906976744186e-06,
+      "loss": 0.2914,
+      "step": 2551
+    },
+    {
+      "epoch": 2.912721049629207,
+      "grad_norm": 0.1936538427333248,
+      "learning_rate": 1.6067653276955603e-06,
+      "loss": 0.2913,
+      "step": 2552
+    },
+    {
+      "epoch": 2.913861950941244,
+      "grad_norm": 0.19368708996765574,
+      "learning_rate": 1.5856236786469344e-06,
+      "loss": 0.2917,
+      "step": 2553
+    },
+    {
+      "epoch": 2.91500285225328,
+      "grad_norm": 0.20383004181874495,
+      "learning_rate": 1.5644820295983086e-06,
+      "loss": 0.3022,
+      "step": 2554
+    },
+    {
+      "epoch": 2.9161437535653167,
+      "grad_norm": 0.19381726160739035,
+      "learning_rate": 1.543340380549683e-06,
+      "loss": 0.2995,
+      "step": 2555
+    },
+    {
+      "epoch": 2.9172846548773532,
+      "grad_norm": 0.20273695423735769,
+      "learning_rate": 1.5221987315010572e-06,
+      "loss": 0.3121,
+      "step": 2556
+    },
+    {
+      "epoch": 2.9184255561893897,
+      "grad_norm": 0.2036090775938918,
+      "learning_rate": 1.5010570824524313e-06,
+      "loss": 0.2825,
+      "step": 2557
+    },
+    {
+      "epoch": 2.919566457501426,
+      "grad_norm": 0.1862428612768711,
+      "learning_rate": 1.4799154334038056e-06,
+      "loss": 0.2778,
+      "step": 2558
+    },
+    {
+      "epoch": 2.9207073588134627,
+      "grad_norm": 0.19043762405083364,
+      "learning_rate": 1.4587737843551799e-06,
+      "loss": 0.2841,
+      "step": 2559
+    },
+    {
+      "epoch": 2.921848260125499,
+      "grad_norm": 0.20187309787287008,
+      "learning_rate": 1.437632135306554e-06,
+      "loss": 0.3121,
+      "step": 2560
+    },
+    {
+      "epoch": 2.9229891614375356,
+      "grad_norm": 0.1946742127017689,
+      "learning_rate": 1.4164904862579282e-06,
+      "loss": 0.293,
+      "step": 2561
+    },
+    {
+      "epoch": 2.924130062749572,
+      "grad_norm": 0.208031129118613,
+      "learning_rate": 1.3953488372093023e-06,
+      "loss": 0.2792,
+      "step": 2562
+    },
+    {
+      "epoch": 2.9252709640616086,
+      "grad_norm": 0.20002498323975684,
+      "learning_rate": 1.3742071881606766e-06,
+      "loss": 0.2943,
+      "step": 2563
+    },
+    {
+      "epoch": 2.926411865373645,
+      "grad_norm": 0.19726715468112263,
+      "learning_rate": 1.3530655391120506e-06,
+      "loss": 0.2843,
+      "step": 2564
+    },
+    {
+      "epoch": 2.9275527666856815,
+      "grad_norm": 0.19063146152359087,
+      "learning_rate": 1.3319238900634251e-06,
+      "loss": 0.2652,
+      "step": 2565
+    },
+    {
+      "epoch": 2.9286936679977185,
+      "grad_norm": 0.19186174328788505,
+      "learning_rate": 1.3107822410147992e-06,
+      "loss": 0.2789,
+      "step": 2566
+    },
+    {
+      "epoch": 2.9298345693097545,
+      "grad_norm": 0.18565312435433468,
+      "learning_rate": 1.2896405919661735e-06,
+      "loss": 0.2711,
+      "step": 2567
+    },
+    {
+      "epoch": 2.9309754706217914,
+      "grad_norm": 0.20681081424437364,
+      "learning_rate": 1.2684989429175476e-06,
+      "loss": 0.295,
+      "step": 2568
+    },
+    {
+      "epoch": 2.9321163719338275,
+      "grad_norm": 0.18971818353896588,
+      "learning_rate": 1.2473572938689219e-06,
+      "loss": 0.2926,
+      "step": 2569
+    },
+    {
+      "epoch": 2.9332572732458644,
+      "grad_norm": 0.19957945564892926,
+      "learning_rate": 1.2262156448202961e-06,
+      "loss": 0.2859,
+      "step": 2570
+    },
+    {
+      "epoch": 2.934398174557901,
+      "grad_norm": 0.19095388008305295,
+      "learning_rate": 1.2050739957716702e-06,
+      "loss": 0.287,
+      "step": 2571
+    },
+    {
+      "epoch": 2.9355390758699373,
+      "grad_norm": 0.1954767979123498,
+      "learning_rate": 1.1839323467230445e-06,
+      "loss": 0.2918,
+      "step": 2572
+    },
+    {
+      "epoch": 2.936679977181974,
+      "grad_norm": 0.18193127967812484,
+      "learning_rate": 1.1627906976744186e-06,
+      "loss": 0.2991,
+      "step": 2573
+    },
+    {
+      "epoch": 2.9378208784940103,
+      "grad_norm": 0.19683880921857755,
+      "learning_rate": 1.1416490486257929e-06,
+      "loss": 0.2959,
+      "step": 2574
+    },
+    {
+      "epoch": 2.938961779806047,
+      "grad_norm": 0.1938599942036892,
+      "learning_rate": 1.120507399577167e-06,
+      "loss": 0.2921,
+      "step": 2575
+    },
+    {
+      "epoch": 2.9401026811180833,
+      "grad_norm": 0.19585647152155478,
+      "learning_rate": 1.0993657505285414e-06,
+      "loss": 0.2876,
+      "step": 2576
+    },
+    {
+      "epoch": 2.9412435824301197,
+      "grad_norm": 0.18568095514600264,
+      "learning_rate": 1.0782241014799155e-06,
+      "loss": 0.2827,
+      "step": 2577
+    },
+    {
+      "epoch": 2.942384483742156,
+      "grad_norm": 0.18521242123000087,
+      "learning_rate": 1.0570824524312898e-06,
+      "loss": 0.2776,
+      "step": 2578
+    },
+    {
+      "epoch": 2.9435253850541927,
+      "grad_norm": 0.19199229150066516,
+      "learning_rate": 1.0359408033826639e-06,
+      "loss": 0.2952,
+      "step": 2579
+    },
+    {
+      "epoch": 2.944666286366229,
+      "grad_norm": 0.21405090207622357,
+      "learning_rate": 1.0147991543340381e-06,
+      "loss": 0.3068,
+      "step": 2580
+    },
+    {
+      "epoch": 2.945807187678266,
+      "grad_norm": 0.19626351454485538,
+      "learning_rate": 9.936575052854124e-07,
+      "loss": 0.2755,
+      "step": 2581
+    },
+    {
+      "epoch": 2.946948088990302,
+      "grad_norm": 0.18502284402400002,
+      "learning_rate": 9.725158562367865e-07,
+      "loss": 0.2841,
+      "step": 2582
+    },
+    {
+      "epoch": 2.948088990302339,
+      "grad_norm": 0.1982315842625302,
+      "learning_rate": 9.513742071881608e-07,
+      "loss": 0.2918,
+      "step": 2583
+    },
+    {
+      "epoch": 2.949229891614375,
+      "grad_norm": 0.18425481134910698,
+      "learning_rate": 9.30232558139535e-07,
+      "loss": 0.2942,
+      "step": 2584
+    },
+    {
+      "epoch": 2.950370792926412,
+      "grad_norm": 0.19674722242444975,
+      "learning_rate": 9.09090909090909e-07,
+      "loss": 0.2962,
+      "step": 2585
+    },
+    {
+      "epoch": 2.9515116942384485,
+      "grad_norm": 0.18753143414095313,
+      "learning_rate": 8.879492600422832e-07,
+      "loss": 0.2852,
+      "step": 2586
+    },
+    {
+      "epoch": 2.952652595550485,
+      "grad_norm": 0.21810746998579797,
+      "learning_rate": 8.668076109936576e-07,
+      "loss": 0.2932,
+      "step": 2587
+    },
+    {
+      "epoch": 2.9537934968625215,
+      "grad_norm": 0.19022816490059893,
+      "learning_rate": 8.456659619450318e-07,
+      "loss": 0.2977,
+      "step": 2588
+    },
+    {
+      "epoch": 2.954934398174558,
+      "grad_norm": 0.18911069763423244,
+      "learning_rate": 8.24524312896406e-07,
+      "loss": 0.2994,
+      "step": 2589
+    },
+    {
+      "epoch": 2.9560752994865944,
+      "grad_norm": 0.19384914458501556,
+      "learning_rate": 8.033826638477801e-07,
+      "loss": 0.298,
+      "step": 2590
+    },
+    {
+      "epoch": 2.957216200798631,
+      "grad_norm": 0.18640080999088518,
+      "learning_rate": 7.822410147991543e-07,
+      "loss": 0.2921,
+      "step": 2591
+    },
+    {
+      "epoch": 2.9583571021106674,
+      "grad_norm": 0.19120986595410241,
+      "learning_rate": 7.610993657505286e-07,
+      "loss": 0.2807,
+      "step": 2592
+    },
+    {
+      "epoch": 2.959498003422704,
+      "grad_norm": 0.17738248639668097,
+      "learning_rate": 7.399577167019028e-07,
+      "loss": 0.2794,
+      "step": 2593
+    },
+    {
+      "epoch": 2.9606389047347403,
+      "grad_norm": 0.20010347777133208,
+      "learning_rate": 7.18816067653277e-07,
+      "loss": 0.2788,
+      "step": 2594
+    },
+    {
+      "epoch": 2.961779806046777,
+      "grad_norm": 0.19032191300940796,
+      "learning_rate": 6.976744186046511e-07,
+      "loss": 0.2844,
+      "step": 2595
+    },
+    {
+      "epoch": 2.9629207073588133,
+      "grad_norm": 0.18550298786618058,
+      "learning_rate": 6.765327695560253e-07,
+      "loss": 0.2864,
+      "step": 2596
+    },
+    {
+      "epoch": 2.96406160867085,
+      "grad_norm": 0.18691192560025538,
+      "learning_rate": 6.553911205073996e-07,
+      "loss": 0.296,
+      "step": 2597
+    },
+    {
+      "epoch": 2.9652025099828867,
+      "grad_norm": 0.18180543168740507,
+      "learning_rate": 6.342494714587738e-07,
+      "loss": 0.2853,
+      "step": 2598
+    },
+    {
+      "epoch": 2.9663434112949227,
+      "grad_norm": 0.18467998613229547,
+      "learning_rate": 6.131078224101481e-07,
+      "loss": 0.2927,
+      "step": 2599
+    },
+    {
+      "epoch": 2.9674843126069597,
+      "grad_norm": 0.19071773172373238,
+      "learning_rate": 5.919661733615223e-07,
+      "loss": 0.288,
+      "step": 2600
+    },
+    {
+      "epoch": 2.968625213918996,
+      "grad_norm": 0.2040159766721208,
+      "learning_rate": 5.708245243128964e-07,
+      "loss": 0.2977,
+      "step": 2601
+    },
+    {
+      "epoch": 2.9697661152310326,
+      "grad_norm": 0.18910715703892397,
+      "learning_rate": 5.496828752642707e-07,
+      "loss": 0.2943,
+      "step": 2602
+    },
+    {
+      "epoch": 2.970907016543069,
+      "grad_norm": 0.20003695025712895,
+      "learning_rate": 5.285412262156449e-07,
+      "loss": 0.2789,
+      "step": 2603
+    },
+    {
+      "epoch": 2.9720479178551056,
+      "grad_norm": 0.18261240038936224,
+      "learning_rate": 5.073995771670191e-07,
+      "loss": 0.3079,
+      "step": 2604
+    },
+    {
+      "epoch": 2.973188819167142,
+      "grad_norm": 0.19686254457945906,
+      "learning_rate": 4.862579281183933e-07,
+      "loss": 0.2868,
+      "step": 2605
+    },
+    {
+      "epoch": 2.9743297204791785,
+      "grad_norm": 0.19786667723149334,
+      "learning_rate": 4.651162790697675e-07,
+      "loss": 0.2968,
+      "step": 2606
+    },
+    {
+      "epoch": 2.975470621791215,
+      "grad_norm": 0.20652336498177548,
+      "learning_rate": 4.439746300211416e-07,
+      "loss": 0.3022,
+      "step": 2607
+    },
+    {
+      "epoch": 2.9766115231032515,
+      "grad_norm": 0.1967118773874935,
+      "learning_rate": 4.228329809725159e-07,
+      "loss": 0.2832,
+      "step": 2608
+    },
+    {
+      "epoch": 2.977752424415288,
+      "grad_norm": 0.1945048452350098,
+      "learning_rate": 4.0169133192389007e-07,
+      "loss": 0.2974,
+      "step": 2609
+    },
+    {
+      "epoch": 2.9788933257273245,
+      "grad_norm": 0.1889319361030166,
+      "learning_rate": 3.805496828752643e-07,
+      "loss": 0.2861,
+      "step": 2610
+    },
+    {
+      "epoch": 2.980034227039361,
+      "grad_norm": 0.20296137411834944,
+      "learning_rate": 3.594080338266385e-07,
+      "loss": 0.2938,
+      "step": 2611
+    },
+    {
+      "epoch": 2.9811751283513974,
+      "grad_norm": 0.18722476513226705,
+      "learning_rate": 3.3826638477801266e-07,
+      "loss": 0.2909,
+      "step": 2612
+    },
+    {
+      "epoch": 2.9823160296634343,
+      "grad_norm": 0.19444064275525808,
+      "learning_rate": 3.171247357293869e-07,
+      "loss": 0.2815,
+      "step": 2613
+    },
+    {
+      "epoch": 2.9834569309754704,
+      "grad_norm": 0.1908410516017102,
+      "learning_rate": 2.959830866807611e-07,
+      "loss": 0.2784,
+      "step": 2614
+    },
+    {
+      "epoch": 2.9845978322875073,
+      "grad_norm": 0.19556815957043194,
+      "learning_rate": 2.7484143763213536e-07,
+      "loss": 0.3052,
+      "step": 2615
+    },
+    {
+      "epoch": 2.985738733599544,
+      "grad_norm": 0.20031795787914,
+      "learning_rate": 2.5369978858350954e-07,
+      "loss": 0.3062,
+      "step": 2616
+    },
+    {
+      "epoch": 2.9868796349115803,
+      "grad_norm": 0.2120042672146075,
+      "learning_rate": 2.3255813953488374e-07,
+      "loss": 0.2793,
+      "step": 2617
+    },
+    {
+      "epoch": 2.9880205362236167,
+      "grad_norm": 0.2135770614047869,
+      "learning_rate": 2.1141649048625795e-07,
+      "loss": 0.278,
+      "step": 2618
+    },
+    {
+      "epoch": 2.989161437535653,
+      "grad_norm": 0.1865270923940558,
+      "learning_rate": 1.9027484143763215e-07,
+      "loss": 0.2858,
+      "step": 2619
+    },
+    {
+      "epoch": 2.9903023388476897,
+      "grad_norm": 0.1917240005764539,
+      "learning_rate": 1.6913319238900633e-07,
+      "loss": 0.2951,
+      "step": 2620
+    },
+    {
+      "epoch": 2.991443240159726,
+      "grad_norm": 0.18904300839758295,
+      "learning_rate": 1.4799154334038056e-07,
+      "loss": 0.2856,
+      "step": 2621
+    },
+    {
+      "epoch": 2.9925841414717627,
+      "grad_norm": 0.1836428754871979,
+      "learning_rate": 1.2684989429175477e-07,
+      "loss": 0.2911,
+      "step": 2622
+    },
+    {
+      "epoch": 2.993725042783799,
+      "grad_norm": 0.2057010148568336,
+      "learning_rate": 1.0570824524312897e-07,
+      "loss": 0.2841,
+      "step": 2623
+    },
+    {
+      "epoch": 2.9948659440958356,
+      "grad_norm": 0.19150844068855372,
+      "learning_rate": 8.456659619450317e-08,
+      "loss": 0.2928,
+      "step": 2624
+    },
+    {
+      "epoch": 2.996006845407872,
+      "grad_norm": 0.18710502056330086,
+      "learning_rate": 6.342494714587738e-08,
+      "loss": 0.2823,
+      "step": 2625
+    },
+    {
+      "epoch": 2.9971477467199086,
+      "grad_norm": 0.18850823231663,
+      "learning_rate": 4.228329809725158e-08,
+      "loss": 0.2842,
+      "step": 2626
+    },
+    {
+      "epoch": 2.998288648031945,
+      "grad_norm": 0.1861174706937187,
+      "learning_rate": 2.114164904862579e-08,
+      "loss": 0.2976,
+      "step": 2627
+    },
+    {
+      "epoch": 2.999429549343982,
+      "grad_norm": 0.18555388787832147,
+      "learning_rate": 0.0,
+      "loss": 0.2851,
+      "step": 2628
+    },
+    {
+      "epoch": 2.999429549343982,
+      "step": 2628,
+      "total_flos": 2.2479308431902638e+18,
+      "train_loss": 0.20299065702763502,
+      "train_runtime": 141958.5173,
+      "train_samples_per_second": 0.296,
+      "train_steps_per_second": 0.019
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2628,
+  "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": 2.2479308431902638e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}