diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,46078 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.99878308487983,
+  "eval_steps": 500,
+  "global_step": 6572,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000608457560085184,
+      "grad_norm": 9.286478996276855,
+      "learning_rate": 1.9011406844106465e-07,
+      "loss": 0.8452,
+      "step": 1
+    },
+    {
+      "epoch": 0.001216915120170368,
+      "grad_norm": 9.28686809539795,
+      "learning_rate": 3.802281368821293e-07,
+      "loss": 0.8357,
+      "step": 2
+    },
+    {
+      "epoch": 0.0018253726802555522,
+      "grad_norm": 9.397164344787598,
+      "learning_rate": 5.70342205323194e-07,
+      "loss": 0.8038,
+      "step": 3
+    },
+    {
+      "epoch": 0.002433830240340736,
+      "grad_norm": 8.887619972229004,
+      "learning_rate": 7.604562737642586e-07,
+      "loss": 0.7907,
+      "step": 4
+    },
+    {
+      "epoch": 0.0030422878004259203,
+      "grad_norm": 10.438508033752441,
+      "learning_rate": 9.505703422053232e-07,
+      "loss": 0.8433,
+      "step": 5
+    },
+    {
+      "epoch": 0.0036507453605111044,
+      "grad_norm": 9.955141067504883,
+      "learning_rate": 1.140684410646388e-06,
+      "loss": 0.8832,
+      "step": 6
+    },
+    {
+      "epoch": 0.0042592029205962886,
+      "grad_norm": 7.124298095703125,
+      "learning_rate": 1.3307984790874525e-06,
+      "loss": 0.7259,
+      "step": 7
+    },
+    {
+      "epoch": 0.004867660480681472,
+      "grad_norm": 6.805534839630127,
+      "learning_rate": 1.5209125475285172e-06,
+      "loss": 0.7877,
+      "step": 8
+    },
+    {
+      "epoch": 0.005476118040766657,
+      "grad_norm": 5.027353286743164,
+      "learning_rate": 1.711026615969582e-06,
+      "loss": 0.6938,
+      "step": 9
+    },
+    {
+      "epoch": 0.0060845756008518406,
+      "grad_norm": 4.865087509155273,
+      "learning_rate": 1.9011406844106463e-06,
+      "loss": 0.704,
+      "step": 10
+    },
+    {
+      "epoch": 0.006693033160937024,
+      "grad_norm": 4.999395847320557,
+      "learning_rate": 2.091254752851711e-06,
+      "loss": 0.6225,
+      "step": 11
+    },
+    {
+      "epoch": 0.007301490721022209,
+      "grad_norm": 4.837018013000488,
+      "learning_rate": 2.281368821292776e-06,
+      "loss": 0.6357,
+      "step": 12
+    },
+    {
+      "epoch": 0.007909948281107393,
+      "grad_norm": 4.025296211242676,
+      "learning_rate": 2.4714828897338406e-06,
+      "loss": 0.6002,
+      "step": 13
+    },
+    {
+      "epoch": 0.008518405841192577,
+      "grad_norm": 3.6638612747192383,
+      "learning_rate": 2.661596958174905e-06,
+      "loss": 0.5499,
+      "step": 14
+    },
+    {
+      "epoch": 0.009126863401277762,
+      "grad_norm": 5.011624336242676,
+      "learning_rate": 2.8517110266159697e-06,
+      "loss": 0.6042,
+      "step": 15
+    },
+    {
+      "epoch": 0.009735320961362945,
+      "grad_norm": 4.899325847625732,
+      "learning_rate": 3.0418250950570345e-06,
+      "loss": 0.6279,
+      "step": 16
+    },
+    {
+      "epoch": 0.010343778521448129,
+      "grad_norm": 4.226297378540039,
+      "learning_rate": 3.2319391634980988e-06,
+      "loss": 0.6134,
+      "step": 17
+    },
+    {
+      "epoch": 0.010952236081533314,
+      "grad_norm": 4.012551784515381,
+      "learning_rate": 3.422053231939164e-06,
+      "loss": 0.6285,
+      "step": 18
+    },
+    {
+      "epoch": 0.011560693641618497,
+      "grad_norm": 3.9116461277008057,
+      "learning_rate": 3.612167300380228e-06,
+      "loss": 0.5381,
+      "step": 19
+    },
+    {
+      "epoch": 0.012169151201703681,
+      "grad_norm": 2.996412515640259,
+      "learning_rate": 3.8022813688212926e-06,
+      "loss": 0.4892,
+      "step": 20
+    },
+    {
+      "epoch": 0.012777608761788866,
+      "grad_norm": 2.7711269855499268,
+      "learning_rate": 3.992395437262358e-06,
+      "loss": 0.4954,
+      "step": 21
+    },
+    {
+      "epoch": 0.013386066321874049,
+      "grad_norm": 3.1407854557037354,
+      "learning_rate": 4.182509505703422e-06,
+      "loss": 0.6063,
+      "step": 22
+    },
+    {
+      "epoch": 0.013994523881959233,
+      "grad_norm": 2.8379178047180176,
+      "learning_rate": 4.3726235741444865e-06,
+      "loss": 0.5944,
+      "step": 23
+    },
+    {
+      "epoch": 0.014602981442044418,
+      "grad_norm": 2.903754234313965,
+      "learning_rate": 4.562737642585552e-06,
+      "loss": 0.5006,
+      "step": 24
+    },
+    {
+      "epoch": 0.015211439002129602,
+      "grad_norm": 3.051934003829956,
+      "learning_rate": 4.752851711026616e-06,
+      "loss": 0.5743,
+      "step": 25
+    },
+    {
+      "epoch": 0.015819896562214785,
+      "grad_norm": 2.664247751235962,
+      "learning_rate": 4.942965779467681e-06,
+      "loss": 0.5162,
+      "step": 26
+    },
+    {
+      "epoch": 0.016428354122299968,
+      "grad_norm": 3.0488321781158447,
+      "learning_rate": 5.1330798479087455e-06,
+      "loss": 0.5195,
+      "step": 27
+    },
+    {
+      "epoch": 0.017036811682385154,
+      "grad_norm": 2.9393467903137207,
+      "learning_rate": 5.32319391634981e-06,
+      "loss": 0.5466,
+      "step": 28
+    },
+    {
+      "epoch": 0.017645269242470337,
+      "grad_norm": 2.603747844696045,
+      "learning_rate": 5.513307984790875e-06,
+      "loss": 0.5294,
+      "step": 29
+    },
+    {
+      "epoch": 0.018253726802555523,
+      "grad_norm": 2.919405698776245,
+      "learning_rate": 5.703422053231939e-06,
+      "loss": 0.5621,
+      "step": 30
+    },
+    {
+      "epoch": 0.018862184362640706,
+      "grad_norm": 2.5989558696746826,
+      "learning_rate": 5.893536121673004e-06,
+      "loss": 0.5719,
+      "step": 31
+    },
+    {
+      "epoch": 0.01947064192272589,
+      "grad_norm": 2.815772294998169,
+      "learning_rate": 6.083650190114069e-06,
+      "loss": 0.5748,
+      "step": 32
+    },
+    {
+      "epoch": 0.020079099482811075,
+      "grad_norm": 2.713179111480713,
+      "learning_rate": 6.273764258555133e-06,
+      "loss": 0.4973,
+      "step": 33
+    },
+    {
+      "epoch": 0.020687557042896258,
+      "grad_norm": 2.5474116802215576,
+      "learning_rate": 6.4638783269961976e-06,
+      "loss": 0.5012,
+      "step": 34
+    },
+    {
+      "epoch": 0.02129601460298144,
+      "grad_norm": 2.737952947616577,
+      "learning_rate": 6.653992395437263e-06,
+      "loss": 0.4963,
+      "step": 35
+    },
+    {
+      "epoch": 0.021904472163066627,
+      "grad_norm": 2.6271746158599854,
+      "learning_rate": 6.844106463878328e-06,
+      "loss": 0.5512,
+      "step": 36
+    },
+    {
+      "epoch": 0.02251292972315181,
+      "grad_norm": 2.547055721282959,
+      "learning_rate": 7.034220532319392e-06,
+      "loss": 0.5489,
+      "step": 37
+    },
+    {
+      "epoch": 0.023121387283236993,
+      "grad_norm": 2.6234169006347656,
+      "learning_rate": 7.224334600760456e-06,
+      "loss": 0.4627,
+      "step": 38
+    },
+    {
+      "epoch": 0.02372984484332218,
+      "grad_norm": 2.4380102157592773,
+      "learning_rate": 7.414448669201521e-06,
+      "loss": 0.5985,
+      "step": 39
+    },
+    {
+      "epoch": 0.024338302403407362,
+      "grad_norm": 2.633112668991089,
+      "learning_rate": 7.604562737642585e-06,
+      "loss": 0.4824,
+      "step": 40
+    },
+    {
+      "epoch": 0.024946759963492545,
+      "grad_norm": 2.357767343521118,
+      "learning_rate": 7.79467680608365e-06,
+      "loss": 0.4725,
+      "step": 41
+    },
+    {
+      "epoch": 0.02555521752357773,
+      "grad_norm": 2.5731077194213867,
+      "learning_rate": 7.984790874524716e-06,
+      "loss": 0.5152,
+      "step": 42
+    },
+    {
+      "epoch": 0.026163675083662914,
+      "grad_norm": 2.6900103092193604,
+      "learning_rate": 8.17490494296578e-06,
+      "loss": 0.4789,
+      "step": 43
+    },
+    {
+      "epoch": 0.026772132643748097,
+      "grad_norm": 2.833528757095337,
+      "learning_rate": 8.365019011406844e-06,
+      "loss": 0.534,
+      "step": 44
+    },
+    {
+      "epoch": 0.027380590203833283,
+      "grad_norm": 2.824359893798828,
+      "learning_rate": 8.55513307984791e-06,
+      "loss": 0.4799,
+      "step": 45
+    },
+    {
+      "epoch": 0.027989047763918466,
+      "grad_norm": 2.2743606567382812,
+      "learning_rate": 8.745247148288973e-06,
+      "loss": 0.4938,
+      "step": 46
+    },
+    {
+      "epoch": 0.02859750532400365,
+      "grad_norm": 2.405674934387207,
+      "learning_rate": 8.935361216730038e-06,
+      "loss": 0.5277,
+      "step": 47
+    },
+    {
+      "epoch": 0.029205962884088835,
+      "grad_norm": 2.473686695098877,
+      "learning_rate": 9.125475285171103e-06,
+      "loss": 0.4787,
+      "step": 48
+    },
+    {
+      "epoch": 0.029814420444174018,
+      "grad_norm": 2.3129618167877197,
+      "learning_rate": 9.315589353612169e-06,
+      "loss": 0.4755,
+      "step": 49
+    },
+    {
+      "epoch": 0.030422878004259205,
+      "grad_norm": 2.2328269481658936,
+      "learning_rate": 9.505703422053232e-06,
+      "loss": 0.5602,
+      "step": 50
+    },
+    {
+      "epoch": 0.031031335564344387,
+      "grad_norm": 2.6735682487487793,
+      "learning_rate": 9.695817490494297e-06,
+      "loss": 0.5262,
+      "step": 51
+    },
+    {
+      "epoch": 0.03163979312442957,
+      "grad_norm": 2.37196946144104,
+      "learning_rate": 9.885931558935362e-06,
+      "loss": 0.5204,
+      "step": 52
+    },
+    {
+      "epoch": 0.032248250684514757,
+      "grad_norm": 2.7100112438201904,
+      "learning_rate": 1.0076045627376426e-05,
+      "loss": 0.5656,
+      "step": 53
+    },
+    {
+      "epoch": 0.032856708244599936,
+      "grad_norm": 2.4364941120147705,
+      "learning_rate": 1.0266159695817491e-05,
+      "loss": 0.4961,
+      "step": 54
+    },
+    {
+      "epoch": 0.03346516580468512,
+      "grad_norm": 2.3422157764434814,
+      "learning_rate": 1.0456273764258556e-05,
+      "loss": 0.492,
+      "step": 55
+    },
+    {
+      "epoch": 0.03407362336477031,
+      "grad_norm": 2.0109498500823975,
+      "learning_rate": 1.064638783269962e-05,
+      "loss": 0.503,
+      "step": 56
+    },
+    {
+      "epoch": 0.03468208092485549,
+      "grad_norm": 2.42881441116333,
+      "learning_rate": 1.0836501901140685e-05,
+      "loss": 0.5093,
+      "step": 57
+    },
+    {
+      "epoch": 0.035290538484940674,
+      "grad_norm": 2.187835693359375,
+      "learning_rate": 1.102661596958175e-05,
+      "loss": 0.4663,
+      "step": 58
+    },
+    {
+      "epoch": 0.03589899604502586,
+      "grad_norm": 2.3914639949798584,
+      "learning_rate": 1.1216730038022814e-05,
+      "loss": 0.4773,
+      "step": 59
+    },
+    {
+      "epoch": 0.03650745360511105,
+      "grad_norm": 2.4356846809387207,
+      "learning_rate": 1.1406844106463879e-05,
+      "loss": 0.5378,
+      "step": 60
+    },
+    {
+      "epoch": 0.037115911165196226,
+      "grad_norm": 2.398271083831787,
+      "learning_rate": 1.1596958174904944e-05,
+      "loss": 0.4864,
+      "step": 61
+    },
+    {
+      "epoch": 0.03772436872528141,
+      "grad_norm": 2.2733852863311768,
+      "learning_rate": 1.1787072243346007e-05,
+      "loss": 0.4751,
+      "step": 62
+    },
+    {
+      "epoch": 0.0383328262853666,
+      "grad_norm": 2.115436315536499,
+      "learning_rate": 1.1977186311787073e-05,
+      "loss": 0.4688,
+      "step": 63
+    },
+    {
+      "epoch": 0.03894128384545178,
+      "grad_norm": 2.285353660583496,
+      "learning_rate": 1.2167300380228138e-05,
+      "loss": 0.5107,
+      "step": 64
+    },
+    {
+      "epoch": 0.039549741405536964,
+      "grad_norm": 2.3437185287475586,
+      "learning_rate": 1.2357414448669203e-05,
+      "loss": 0.5132,
+      "step": 65
+    },
+    {
+      "epoch": 0.04015819896562215,
+      "grad_norm": 2.5334279537200928,
+      "learning_rate": 1.2547528517110266e-05,
+      "loss": 0.4868,
+      "step": 66
+    },
+    {
+      "epoch": 0.04076665652570733,
+      "grad_norm": 2.2732203006744385,
+      "learning_rate": 1.2737642585551332e-05,
+      "loss": 0.4784,
+      "step": 67
+    },
+    {
+      "epoch": 0.041375114085792516,
+      "grad_norm": 2.3344926834106445,
+      "learning_rate": 1.2927756653992395e-05,
+      "loss": 0.5702,
+      "step": 68
+    },
+    {
+      "epoch": 0.0419835716458777,
+      "grad_norm": 2.6183888912200928,
+      "learning_rate": 1.3117870722433462e-05,
+      "loss": 0.4828,
+      "step": 69
+    },
+    {
+      "epoch": 0.04259202920596288,
+      "grad_norm": 3.995020627975464,
+      "learning_rate": 1.3307984790874526e-05,
+      "loss": 0.5182,
+      "step": 70
+    },
+    {
+      "epoch": 0.04320048676604807,
+      "grad_norm": 2.3043413162231445,
+      "learning_rate": 1.3498098859315589e-05,
+      "loss": 0.5346,
+      "step": 71
+    },
+    {
+      "epoch": 0.043808944326133255,
+      "grad_norm": 2.5298848152160645,
+      "learning_rate": 1.3688212927756656e-05,
+      "loss": 0.6079,
+      "step": 72
+    },
+    {
+      "epoch": 0.044417401886218434,
+      "grad_norm": 2.1517183780670166,
+      "learning_rate": 1.387832699619772e-05,
+      "loss": 0.4962,
+      "step": 73
+    },
+    {
+      "epoch": 0.04502585944630362,
+      "grad_norm": 2.3889400959014893,
+      "learning_rate": 1.4068441064638785e-05,
+      "loss": 0.5005,
+      "step": 74
+    },
+    {
+      "epoch": 0.04563431700638881,
+      "grad_norm": 2.265465259552002,
+      "learning_rate": 1.4258555133079848e-05,
+      "loss": 0.5684,
+      "step": 75
+    },
+    {
+      "epoch": 0.046242774566473986,
+      "grad_norm": 2.3578243255615234,
+      "learning_rate": 1.4448669201520912e-05,
+      "loss": 0.5401,
+      "step": 76
+    },
+    {
+      "epoch": 0.04685123212655917,
+      "grad_norm": 2.2916693687438965,
+      "learning_rate": 1.4638783269961978e-05,
+      "loss": 0.482,
+      "step": 77
+    },
+    {
+      "epoch": 0.04745968968664436,
+      "grad_norm": 2.678074359893799,
+      "learning_rate": 1.4828897338403042e-05,
+      "loss": 0.5157,
+      "step": 78
+    },
+    {
+      "epoch": 0.04806814724672954,
+      "grad_norm": 2.3698818683624268,
+      "learning_rate": 1.5019011406844107e-05,
+      "loss": 0.4558,
+      "step": 79
+    },
+    {
+      "epoch": 0.048676604806814724,
+      "grad_norm": 2.1744048595428467,
+      "learning_rate": 1.520912547528517e-05,
+      "loss": 0.4658,
+      "step": 80
+    },
+    {
+      "epoch": 0.04928506236689991,
+      "grad_norm": 2.1933300495147705,
+      "learning_rate": 1.5399239543726237e-05,
+      "loss": 0.518,
+      "step": 81
+    },
+    {
+      "epoch": 0.04989351992698509,
+      "grad_norm": 2.573246479034424,
+      "learning_rate": 1.55893536121673e-05,
+      "loss": 0.5668,
+      "step": 82
+    },
+    {
+      "epoch": 0.050501977487070276,
+      "grad_norm": 2.249875068664551,
+      "learning_rate": 1.5779467680608364e-05,
+      "loss": 0.4757,
+      "step": 83
+    },
+    {
+      "epoch": 0.05111043504715546,
+      "grad_norm": 2.1911234855651855,
+      "learning_rate": 1.596958174904943e-05,
+      "loss": 0.4753,
+      "step": 84
+    },
+    {
+      "epoch": 0.05171889260724064,
+      "grad_norm": 3.867871046066284,
+      "learning_rate": 1.6159695817490495e-05,
+      "loss": 0.4797,
+      "step": 85
+    },
+    {
+      "epoch": 0.05232735016732583,
+      "grad_norm": 2.745781421661377,
+      "learning_rate": 1.634980988593156e-05,
+      "loss": 0.4968,
+      "step": 86
+    },
+    {
+      "epoch": 0.052935807727411015,
+      "grad_norm": 2.0692405700683594,
+      "learning_rate": 1.6539923954372625e-05,
+      "loss": 0.481,
+      "step": 87
+    },
+    {
+      "epoch": 0.053544265287496194,
+      "grad_norm": 2.262392997741699,
+      "learning_rate": 1.673003802281369e-05,
+      "loss": 0.5532,
+      "step": 88
+    },
+    {
+      "epoch": 0.05415272284758138,
+      "grad_norm": 2.1288435459136963,
+      "learning_rate": 1.6920152091254756e-05,
+      "loss": 0.4839,
+      "step": 89
+    },
+    {
+      "epoch": 0.05476118040766657,
+      "grad_norm": 2.0853445529937744,
+      "learning_rate": 1.711026615969582e-05,
+      "loss": 0.5071,
+      "step": 90
+    },
+    {
+      "epoch": 0.055369637967751746,
+      "grad_norm": 2.296994686126709,
+      "learning_rate": 1.7300380228136882e-05,
+      "loss": 0.5614,
+      "step": 91
+    },
+    {
+      "epoch": 0.05597809552783693,
+      "grad_norm": 2.261611223220825,
+      "learning_rate": 1.7490494296577946e-05,
+      "loss": 0.536,
+      "step": 92
+    },
+    {
+      "epoch": 0.05658655308792212,
+      "grad_norm": 2.2473561763763428,
+      "learning_rate": 1.7680608365019013e-05,
+      "loss": 0.4652,
+      "step": 93
+    },
+    {
+      "epoch": 0.0571950106480073,
+      "grad_norm": 2.532961368560791,
+      "learning_rate": 1.7870722433460076e-05,
+      "loss": 0.5331,
+      "step": 94
+    },
+    {
+      "epoch": 0.057803468208092484,
+      "grad_norm": 2.3481945991516113,
+      "learning_rate": 1.806083650190114e-05,
+      "loss": 0.52,
+      "step": 95
+    },
+    {
+      "epoch": 0.05841192576817767,
+      "grad_norm": 2.024817943572998,
+      "learning_rate": 1.8250950570342207e-05,
+      "loss": 0.4745,
+      "step": 96
+    },
+    {
+      "epoch": 0.05902038332826286,
+      "grad_norm": 2.2880470752716064,
+      "learning_rate": 1.844106463878327e-05,
+      "loss": 0.5019,
+      "step": 97
+    },
+    {
+      "epoch": 0.059628840888348036,
+      "grad_norm": 1.9540607929229736,
+      "learning_rate": 1.8631178707224337e-05,
+      "loss": 0.5138,
+      "step": 98
+    },
+    {
+      "epoch": 0.06023729844843322,
+      "grad_norm": 2.5749783515930176,
+      "learning_rate": 1.88212927756654e-05,
+      "loss": 0.5639,
+      "step": 99
+    },
+    {
+      "epoch": 0.06084575600851841,
+      "grad_norm": 2.0296337604522705,
+      "learning_rate": 1.9011406844106464e-05,
+      "loss": 0.4929,
+      "step": 100
+    },
+    {
+      "epoch": 0.06145421356860359,
+      "grad_norm": 2.068000316619873,
+      "learning_rate": 1.920152091254753e-05,
+      "loss": 0.4568,
+      "step": 101
+    },
+    {
+      "epoch": 0.062062671128688775,
+      "grad_norm": 2.4328064918518066,
+      "learning_rate": 1.9391634980988594e-05,
+      "loss": 0.5764,
+      "step": 102
+    },
+    {
+      "epoch": 0.06267112868877396,
+      "grad_norm": 1.961745262145996,
+      "learning_rate": 1.958174904942966e-05,
+      "loss": 0.4756,
+      "step": 103
+    },
+    {
+      "epoch": 0.06327958624885914,
+      "grad_norm": 2.189243793487549,
+      "learning_rate": 1.9771863117870725e-05,
+      "loss": 0.5254,
+      "step": 104
+    },
+    {
+      "epoch": 0.06388804380894432,
+      "grad_norm": 1.9438084363937378,
+      "learning_rate": 1.9961977186311788e-05,
+      "loss": 0.4921,
+      "step": 105
+    },
+    {
+      "epoch": 0.06449650136902951,
+      "grad_norm": 1.9308326244354248,
+      "learning_rate": 2.0152091254752852e-05,
+      "loss": 0.4508,
+      "step": 106
+    },
+    {
+      "epoch": 0.06510495892911469,
+      "grad_norm": 1.9742590188980103,
+      "learning_rate": 2.0342205323193915e-05,
+      "loss": 0.4899,
+      "step": 107
+    },
+    {
+      "epoch": 0.06571341648919987,
+      "grad_norm": 2.1796038150787354,
+      "learning_rate": 2.0532319391634982e-05,
+      "loss": 0.5594,
+      "step": 108
+    },
+    {
+      "epoch": 0.06632187404928507,
+      "grad_norm": 2.2020134925842285,
+      "learning_rate": 2.0722433460076046e-05,
+      "loss": 0.4946,
+      "step": 109
+    },
+    {
+      "epoch": 0.06693033160937024,
+      "grad_norm": 1.9152429103851318,
+      "learning_rate": 2.0912547528517112e-05,
+      "loss": 0.485,
+      "step": 110
+    },
+    {
+      "epoch": 0.06753878916945542,
+      "grad_norm": 2.1203103065490723,
+      "learning_rate": 2.1102661596958176e-05,
+      "loss": 0.5062,
+      "step": 111
+    },
+    {
+      "epoch": 0.06814724672954062,
+      "grad_norm": 1.934693694114685,
+      "learning_rate": 2.129277566539924e-05,
+      "loss": 0.4841,
+      "step": 112
+    },
+    {
+      "epoch": 0.0687557042896258,
+      "grad_norm": 2.1393649578094482,
+      "learning_rate": 2.1482889733840306e-05,
+      "loss": 0.585,
+      "step": 113
+    },
+    {
+      "epoch": 0.06936416184971098,
+      "grad_norm": 2.0923867225646973,
+      "learning_rate": 2.167300380228137e-05,
+      "loss": 0.5467,
+      "step": 114
+    },
+    {
+      "epoch": 0.06997261940979617,
+      "grad_norm": 1.947996735572815,
+      "learning_rate": 2.1863117870722437e-05,
+      "loss": 0.4804,
+      "step": 115
+    },
+    {
+      "epoch": 0.07058107696988135,
+      "grad_norm": 2.1590495109558105,
+      "learning_rate": 2.20532319391635e-05,
+      "loss": 0.5335,
+      "step": 116
+    },
+    {
+      "epoch": 0.07118953452996654,
+      "grad_norm": 2.0173380374908447,
+      "learning_rate": 2.2243346007604564e-05,
+      "loss": 0.5273,
+      "step": 117
+    },
+    {
+      "epoch": 0.07179799209005172,
+      "grad_norm": 2.251988649368286,
+      "learning_rate": 2.2433460076045627e-05,
+      "loss": 0.5975,
+      "step": 118
+    },
+    {
+      "epoch": 0.0724064496501369,
+      "grad_norm": 2.279703378677368,
+      "learning_rate": 2.262357414448669e-05,
+      "loss": 0.5642,
+      "step": 119
+    },
+    {
+      "epoch": 0.0730149072102221,
+      "grad_norm": 2.0928151607513428,
+      "learning_rate": 2.2813688212927758e-05,
+      "loss": 0.5462,
+      "step": 120
+    },
+    {
+      "epoch": 0.07362336477030727,
+      "grad_norm": 9.408754348754883,
+      "learning_rate": 2.300380228136882e-05,
+      "loss": 0.5737,
+      "step": 121
+    },
+    {
+      "epoch": 0.07423182233039245,
+      "grad_norm": 2.0510871410369873,
+      "learning_rate": 2.3193916349809888e-05,
+      "loss": 0.4958,
+      "step": 122
+    },
+    {
+      "epoch": 0.07484027989047765,
+      "grad_norm": 1.8674732446670532,
+      "learning_rate": 2.338403041825095e-05,
+      "loss": 0.4752,
+      "step": 123
+    },
+    {
+      "epoch": 0.07544873745056282,
+      "grad_norm": 1.8963392972946167,
+      "learning_rate": 2.3574144486692015e-05,
+      "loss": 0.5461,
+      "step": 124
+    },
+    {
+      "epoch": 0.076057195010648,
+      "grad_norm": 1.7347912788391113,
+      "learning_rate": 2.3764258555133082e-05,
+      "loss": 0.4789,
+      "step": 125
+    },
+    {
+      "epoch": 0.0766656525707332,
+      "grad_norm": 2.5174317359924316,
+      "learning_rate": 2.3954372623574145e-05,
+      "loss": 0.5642,
+      "step": 126
+    },
+    {
+      "epoch": 0.07727411013081838,
+      "grad_norm": 2.2092862129211426,
+      "learning_rate": 2.4144486692015212e-05,
+      "loss": 0.5812,
+      "step": 127
+    },
+    {
+      "epoch": 0.07788256769090356,
+      "grad_norm": 2.6390881538391113,
+      "learning_rate": 2.4334600760456276e-05,
+      "loss": 0.5021,
+      "step": 128
+    },
+    {
+      "epoch": 0.07849102525098875,
+      "grad_norm": 1.8941545486450195,
+      "learning_rate": 2.452471482889734e-05,
+      "loss": 0.4816,
+      "step": 129
+    },
+    {
+      "epoch": 0.07909948281107393,
+      "grad_norm": 1.7453467845916748,
+      "learning_rate": 2.4714828897338406e-05,
+      "loss": 0.4823,
+      "step": 130
+    },
+    {
+      "epoch": 0.07970794037115911,
+      "grad_norm": 1.9925953149795532,
+      "learning_rate": 2.490494296577947e-05,
+      "loss": 0.5208,
+      "step": 131
+    },
+    {
+      "epoch": 0.0803163979312443,
+      "grad_norm": 2.222625494003296,
+      "learning_rate": 2.5095057034220533e-05,
+      "loss": 0.4962,
+      "step": 132
+    },
+    {
+      "epoch": 0.08092485549132948,
+      "grad_norm": 1.99649178981781,
+      "learning_rate": 2.5285171102661596e-05,
+      "loss": 0.5195,
+      "step": 133
+    },
+    {
+      "epoch": 0.08153331305141466,
+      "grad_norm": 4.0735273361206055,
+      "learning_rate": 2.5475285171102663e-05,
+      "loss": 0.5421,
+      "step": 134
+    },
+    {
+      "epoch": 0.08214177061149985,
+      "grad_norm": 2.117365837097168,
+      "learning_rate": 2.5665399239543723e-05,
+      "loss": 0.5745,
+      "step": 135
+    },
+    {
+      "epoch": 0.08275022817158503,
+      "grad_norm": 2.24652099609375,
+      "learning_rate": 2.585551330798479e-05,
+      "loss": 0.5328,
+      "step": 136
+    },
+    {
+      "epoch": 0.08335868573167021,
+      "grad_norm": 2.288801670074463,
+      "learning_rate": 2.6045627376425857e-05,
+      "loss": 0.5145,
+      "step": 137
+    },
+    {
+      "epoch": 0.0839671432917554,
+      "grad_norm": 1.9129204750061035,
+      "learning_rate": 2.6235741444866924e-05,
+      "loss": 0.5466,
+      "step": 138
+    },
+    {
+      "epoch": 0.08457560085184058,
+      "grad_norm": 2.242582082748413,
+      "learning_rate": 2.6425855513307984e-05,
+      "loss": 0.5592,
+      "step": 139
+    },
+    {
+      "epoch": 0.08518405841192576,
+      "grad_norm": 1.8476616144180298,
+      "learning_rate": 2.661596958174905e-05,
+      "loss": 0.5257,
+      "step": 140
+    },
+    {
+      "epoch": 0.08579251597201096,
+      "grad_norm": 1.9070931673049927,
+      "learning_rate": 2.6806083650190118e-05,
+      "loss": 0.5186,
+      "step": 141
+    },
+    {
+      "epoch": 0.08640097353209614,
+      "grad_norm": 1.9647867679595947,
+      "learning_rate": 2.6996197718631178e-05,
+      "loss": 0.595,
+      "step": 142
+    },
+    {
+      "epoch": 0.08700943109218132,
+      "grad_norm": 2.1235222816467285,
+      "learning_rate": 2.7186311787072245e-05,
+      "loss": 0.4921,
+      "step": 143
+    },
+    {
+      "epoch": 0.08761788865226651,
+      "grad_norm": 2.1548256874084473,
+      "learning_rate": 2.7376425855513312e-05,
+      "loss": 0.5166,
+      "step": 144
+    },
+    {
+      "epoch": 0.08822634621235169,
+      "grad_norm": 3.3486247062683105,
+      "learning_rate": 2.7566539923954375e-05,
+      "loss": 0.5909,
+      "step": 145
+    },
+    {
+      "epoch": 0.08883480377243687,
+      "grad_norm": 2.245957136154175,
+      "learning_rate": 2.775665399239544e-05,
+      "loss": 0.5726,
+      "step": 146
+    },
+    {
+      "epoch": 0.08944326133252206,
+      "grad_norm": 1.7607914209365845,
+      "learning_rate": 2.7946768060836502e-05,
+      "loss": 0.4802,
+      "step": 147
+    },
+    {
+      "epoch": 0.09005171889260724,
+      "grad_norm": 1.840922474861145,
+      "learning_rate": 2.813688212927757e-05,
+      "loss": 0.5054,
+      "step": 148
+    },
+    {
+      "epoch": 0.09066017645269242,
+      "grad_norm": 1.794522762298584,
+      "learning_rate": 2.832699619771863e-05,
+      "loss": 0.5443,
+      "step": 149
+    },
+    {
+      "epoch": 0.09126863401277761,
+      "grad_norm": 1.9434462785720825,
+      "learning_rate": 2.8517110266159696e-05,
+      "loss": 0.5159,
+      "step": 150
+    },
+    {
+      "epoch": 0.09187709157286279,
+      "grad_norm": 2.252189874649048,
+      "learning_rate": 2.8707224334600763e-05,
+      "loss": 0.5802,
+      "step": 151
+    },
+    {
+      "epoch": 0.09248554913294797,
+      "grad_norm": 1.9505751132965088,
+      "learning_rate": 2.8897338403041823e-05,
+      "loss": 0.5033,
+      "step": 152
+    },
+    {
+      "epoch": 0.09309400669303317,
+      "grad_norm": 2.0647778511047363,
+      "learning_rate": 2.908745247148289e-05,
+      "loss": 0.5465,
+      "step": 153
+    },
+    {
+      "epoch": 0.09370246425311834,
+      "grad_norm": 2.034170150756836,
+      "learning_rate": 2.9277566539923957e-05,
+      "loss": 0.5278,
+      "step": 154
+    },
+    {
+      "epoch": 0.09431092181320352,
+      "grad_norm": 2.2246432304382324,
+      "learning_rate": 2.9467680608365024e-05,
+      "loss": 0.5276,
+      "step": 155
+    },
+    {
+      "epoch": 0.09491937937328872,
+      "grad_norm": 1.8863154649734497,
+      "learning_rate": 2.9657794676806084e-05,
+      "loss": 0.4909,
+      "step": 156
+    },
+    {
+      "epoch": 0.0955278369333739,
+      "grad_norm": 2.914320945739746,
+      "learning_rate": 2.984790874524715e-05,
+      "loss": 0.674,
+      "step": 157
+    },
+    {
+      "epoch": 0.09613629449345908,
+      "grad_norm": 4.749186038970947,
+      "learning_rate": 3.0038022813688214e-05,
+      "loss": 0.5386,
+      "step": 158
+    },
+    {
+      "epoch": 0.09674475205354427,
+      "grad_norm": 2.0162932872772217,
+      "learning_rate": 3.0228136882129278e-05,
+      "loss": 0.5441,
+      "step": 159
+    },
+    {
+      "epoch": 0.09735320961362945,
+      "grad_norm": 1.8768587112426758,
+      "learning_rate": 3.041825095057034e-05,
+      "loss": 0.5373,
+      "step": 160
+    },
+    {
+      "epoch": 0.09796166717371463,
+      "grad_norm": 1.9952895641326904,
+      "learning_rate": 3.060836501901141e-05,
+      "loss": 0.5037,
+      "step": 161
+    },
+    {
+      "epoch": 0.09857012473379982,
+      "grad_norm": 1.961302638053894,
+      "learning_rate": 3.0798479087452475e-05,
+      "loss": 0.4827,
+      "step": 162
+    },
+    {
+      "epoch": 0.099178582293885,
+      "grad_norm": 2.2090559005737305,
+      "learning_rate": 3.098859315589354e-05,
+      "loss": 0.5783,
+      "step": 163
+    },
+    {
+      "epoch": 0.09978703985397018,
+      "grad_norm": 1.7766873836517334,
+      "learning_rate": 3.11787072243346e-05,
+      "loss": 0.4749,
+      "step": 164
+    },
+    {
+      "epoch": 0.10039549741405537,
+      "grad_norm": 2.083379030227661,
+      "learning_rate": 3.1368821292775665e-05,
+      "loss": 0.5149,
+      "step": 165
+    },
+    {
+      "epoch": 0.10100395497414055,
+      "grad_norm": 2.4703352451324463,
+      "learning_rate": 3.155893536121673e-05,
+      "loss": 0.5672,
+      "step": 166
+    },
+    {
+      "epoch": 0.10161241253422573,
+      "grad_norm": 2.7795681953430176,
+      "learning_rate": 3.174904942965779e-05,
+      "loss": 0.583,
+      "step": 167
+    },
+    {
+      "epoch": 0.10222087009431093,
+      "grad_norm": 1.8572745323181152,
+      "learning_rate": 3.193916349809886e-05,
+      "loss": 0.4859,
+      "step": 168
+    },
+    {
+      "epoch": 0.1028293276543961,
+      "grad_norm": 2.1731905937194824,
+      "learning_rate": 3.2129277566539926e-05,
+      "loss": 0.4851,
+      "step": 169
+    },
+    {
+      "epoch": 0.10343778521448128,
+      "grad_norm": 2.5642776489257812,
+      "learning_rate": 3.231939163498099e-05,
+      "loss": 0.5155,
+      "step": 170
+    },
+    {
+      "epoch": 0.10404624277456648,
+      "grad_norm": 2.716291904449463,
+      "learning_rate": 3.250950570342205e-05,
+      "loss": 0.5387,
+      "step": 171
+    },
+    {
+      "epoch": 0.10465470033465166,
+      "grad_norm": 2.027435541152954,
+      "learning_rate": 3.269961977186312e-05,
+      "loss": 0.5266,
+      "step": 172
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 2.161743402481079,
+      "learning_rate": 3.288973384030418e-05,
+      "loss": 0.58,
+      "step": 173
+    },
+    {
+      "epoch": 0.10587161545482203,
+      "grad_norm": 2.0740175247192383,
+      "learning_rate": 3.307984790874525e-05,
+      "loss": 0.5476,
+      "step": 174
+    },
+    {
+      "epoch": 0.10648007301490721,
+      "grad_norm": 1.8054178953170776,
+      "learning_rate": 3.3269961977186314e-05,
+      "loss": 0.513,
+      "step": 175
+    },
+    {
+      "epoch": 0.10708853057499239,
+      "grad_norm": 1.872541069984436,
+      "learning_rate": 3.346007604562738e-05,
+      "loss": 0.5266,
+      "step": 176
+    },
+    {
+      "epoch": 0.10769698813507758,
+      "grad_norm": 2.594137191772461,
+      "learning_rate": 3.365019011406844e-05,
+      "loss": 0.56,
+      "step": 177
+    },
+    {
+      "epoch": 0.10830544569516276,
+      "grad_norm": 1.95430588722229,
+      "learning_rate": 3.384030418250951e-05,
+      "loss": 0.5971,
+      "step": 178
+    },
+    {
+      "epoch": 0.10891390325524794,
+      "grad_norm": 2.0000500679016113,
+      "learning_rate": 3.4030418250950574e-05,
+      "loss": 0.4828,
+      "step": 179
+    },
+    {
+      "epoch": 0.10952236081533313,
+      "grad_norm": 1.7855273485183716,
+      "learning_rate": 3.422053231939164e-05,
+      "loss": 0.5368,
+      "step": 180
+    },
+    {
+      "epoch": 0.11013081837541831,
+      "grad_norm": 2.0093138217926025,
+      "learning_rate": 3.44106463878327e-05,
+      "loss": 0.5519,
+      "step": 181
+    },
+    {
+      "epoch": 0.11073927593550349,
+      "grad_norm": 1.8398433923721313,
+      "learning_rate": 3.4600760456273765e-05,
+      "loss": 0.5089,
+      "step": 182
+    },
+    {
+      "epoch": 0.11134773349558869,
+      "grad_norm": 1.9832158088684082,
+      "learning_rate": 3.479087452471483e-05,
+      "loss": 0.578,
+      "step": 183
+    },
+    {
+      "epoch": 0.11195619105567386,
+      "grad_norm": 1.8576923608779907,
+      "learning_rate": 3.498098859315589e-05,
+      "loss": 0.545,
+      "step": 184
+    },
+    {
+      "epoch": 0.11256464861575904,
+      "grad_norm": 1.8127909898757935,
+      "learning_rate": 3.517110266159696e-05,
+      "loss": 0.5694,
+      "step": 185
+    },
+    {
+      "epoch": 0.11317310617584424,
+      "grad_norm": 2.124783515930176,
+      "learning_rate": 3.5361216730038026e-05,
+      "loss": 0.6131,
+      "step": 186
+    },
+    {
+      "epoch": 0.11378156373592942,
+      "grad_norm": 1.8797577619552612,
+      "learning_rate": 3.555133079847909e-05,
+      "loss": 0.5098,
+      "step": 187
+    },
+    {
+      "epoch": 0.1143900212960146,
+      "grad_norm": 1.871032476425171,
+      "learning_rate": 3.574144486692015e-05,
+      "loss": 0.5785,
+      "step": 188
+    },
+    {
+      "epoch": 0.11499847885609979,
+      "grad_norm": 1.6941313743591309,
+      "learning_rate": 3.593155893536122e-05,
+      "loss": 0.5115,
+      "step": 189
+    },
+    {
+      "epoch": 0.11560693641618497,
+      "grad_norm": 2.301297903060913,
+      "learning_rate": 3.612167300380228e-05,
+      "loss": 0.587,
+      "step": 190
+    },
+    {
+      "epoch": 0.11621539397627015,
+      "grad_norm": 1.9337679147720337,
+      "learning_rate": 3.631178707224335e-05,
+      "loss": 0.5393,
+      "step": 191
+    },
+    {
+      "epoch": 0.11682385153635534,
+      "grad_norm": 2.24143648147583,
+      "learning_rate": 3.6501901140684413e-05,
+      "loss": 0.5152,
+      "step": 192
+    },
+    {
+      "epoch": 0.11743230909644052,
+      "grad_norm": 2.0369713306427,
+      "learning_rate": 3.669201520912548e-05,
+      "loss": 0.5691,
+      "step": 193
+    },
+    {
+      "epoch": 0.11804076665652571,
+      "grad_norm": 1.8279515504837036,
+      "learning_rate": 3.688212927756654e-05,
+      "loss": 0.4814,
+      "step": 194
+    },
+    {
+      "epoch": 0.1186492242166109,
+      "grad_norm": 2.0841150283813477,
+      "learning_rate": 3.7072243346007604e-05,
+      "loss": 0.5266,
+      "step": 195
+    },
+    {
+      "epoch": 0.11925768177669607,
+      "grad_norm": 1.830251693725586,
+      "learning_rate": 3.7262357414448674e-05,
+      "loss": 0.5533,
+      "step": 196
+    },
+    {
+      "epoch": 0.11986613933678127,
+      "grad_norm": 1.8735334873199463,
+      "learning_rate": 3.745247148288973e-05,
+      "loss": 0.5265,
+      "step": 197
+    },
+    {
+      "epoch": 0.12047459689686645,
+      "grad_norm": 2.144216299057007,
+      "learning_rate": 3.76425855513308e-05,
+      "loss": 0.5795,
+      "step": 198
+    },
+    {
+      "epoch": 0.12108305445695162,
+      "grad_norm": 1.7879762649536133,
+      "learning_rate": 3.7832699619771865e-05,
+      "loss": 0.5834,
+      "step": 199
+    },
+    {
+      "epoch": 0.12169151201703682,
+      "grad_norm": 2.255518674850464,
+      "learning_rate": 3.802281368821293e-05,
+      "loss": 0.5094,
+      "step": 200
+    },
+    {
+      "epoch": 0.122299969577122,
+      "grad_norm": 1.92243230342865,
+      "learning_rate": 3.821292775665399e-05,
+      "loss": 0.5215,
+      "step": 201
+    },
+    {
+      "epoch": 0.12290842713720718,
+      "grad_norm": 2.395334005355835,
+      "learning_rate": 3.840304182509506e-05,
+      "loss": 0.5715,
+      "step": 202
+    },
+    {
+      "epoch": 0.12351688469729237,
+      "grad_norm": 2.1944241523742676,
+      "learning_rate": 3.8593155893536125e-05,
+      "loss": 0.5486,
+      "step": 203
+    },
+    {
+      "epoch": 0.12412534225737755,
+      "grad_norm": 1.983839750289917,
+      "learning_rate": 3.878326996197719e-05,
+      "loss": 0.5876,
+      "step": 204
+    },
+    {
+      "epoch": 0.12473379981746273,
+      "grad_norm": 2.557204246520996,
+      "learning_rate": 3.897338403041825e-05,
+      "loss": 0.5066,
+      "step": 205
+    },
+    {
+      "epoch": 0.12534225737754792,
+      "grad_norm": 1.9440828561782837,
+      "learning_rate": 3.916349809885932e-05,
+      "loss": 0.567,
+      "step": 206
+    },
+    {
+      "epoch": 0.1259507149376331,
+      "grad_norm": 1.8540195226669312,
+      "learning_rate": 3.935361216730038e-05,
+      "loss": 0.5652,
+      "step": 207
+    },
+    {
+      "epoch": 0.12655917249771828,
+      "grad_norm": 1.9962955713272095,
+      "learning_rate": 3.954372623574145e-05,
+      "loss": 0.5995,
+      "step": 208
+    },
+    {
+      "epoch": 0.12716763005780346,
+      "grad_norm": 2.1377549171447754,
+      "learning_rate": 3.973384030418251e-05,
+      "loss": 0.671,
+      "step": 209
+    },
+    {
+      "epoch": 0.12777608761788864,
+      "grad_norm": 2.1312196254730225,
+      "learning_rate": 3.9923954372623577e-05,
+      "loss": 0.5925,
+      "step": 210
+    },
+    {
+      "epoch": 0.12838454517797385,
+      "grad_norm": 1.659692645072937,
+      "learning_rate": 4.011406844106464e-05,
+      "loss": 0.5313,
+      "step": 211
+    },
+    {
+      "epoch": 0.12899300273805903,
+      "grad_norm": 2.2088122367858887,
+      "learning_rate": 4.0304182509505703e-05,
+      "loss": 0.5368,
+      "step": 212
+    },
+    {
+      "epoch": 0.1296014602981442,
+      "grad_norm": 1.8686944246292114,
+      "learning_rate": 4.0494296577946774e-05,
+      "loss": 0.5576,
+      "step": 213
+    },
+    {
+      "epoch": 0.13020991785822938,
+      "grad_norm": 1.934857726097107,
+      "learning_rate": 4.068441064638783e-05,
+      "loss": 0.5519,
+      "step": 214
+    },
+    {
+      "epoch": 0.13081837541831456,
+      "grad_norm": 1.8353568315505981,
+      "learning_rate": 4.08745247148289e-05,
+      "loss": 0.5312,
+      "step": 215
+    },
+    {
+      "epoch": 0.13142683297839974,
+      "grad_norm": 1.9443809986114502,
+      "learning_rate": 4.1064638783269964e-05,
+      "loss": 0.5108,
+      "step": 216
+    },
+    {
+      "epoch": 0.13203529053848495,
+      "grad_norm": 1.8545550107955933,
+      "learning_rate": 4.125475285171103e-05,
+      "loss": 0.559,
+      "step": 217
+    },
+    {
+      "epoch": 0.13264374809857013,
+      "grad_norm": 1.9417598247528076,
+      "learning_rate": 4.144486692015209e-05,
+      "loss": 0.5631,
+      "step": 218
+    },
+    {
+      "epoch": 0.1332522056586553,
+      "grad_norm": 1.7785390615463257,
+      "learning_rate": 4.163498098859316e-05,
+      "loss": 0.5519,
+      "step": 219
+    },
+    {
+      "epoch": 0.1338606632187405,
+      "grad_norm": 1.9775593280792236,
+      "learning_rate": 4.1825095057034225e-05,
+      "loss": 0.6203,
+      "step": 220
+    },
+    {
+      "epoch": 0.13446912077882567,
+      "grad_norm": 1.7184250354766846,
+      "learning_rate": 4.201520912547529e-05,
+      "loss": 0.5356,
+      "step": 221
+    },
+    {
+      "epoch": 0.13507757833891085,
+      "grad_norm": 1.7691705226898193,
+      "learning_rate": 4.220532319391635e-05,
+      "loss": 0.5185,
+      "step": 222
+    },
+    {
+      "epoch": 0.13568603589899605,
+      "grad_norm": 1.759832739830017,
+      "learning_rate": 4.2395437262357415e-05,
+      "loss": 0.5276,
+      "step": 223
+    },
+    {
+      "epoch": 0.13629449345908123,
+      "grad_norm": 1.8581064939498901,
+      "learning_rate": 4.258555133079848e-05,
+      "loss": 0.5341,
+      "step": 224
+    },
+    {
+      "epoch": 0.1369029510191664,
+      "grad_norm": 1.9540525674819946,
+      "learning_rate": 4.277566539923954e-05,
+      "loss": 0.5544,
+      "step": 225
+    },
+    {
+      "epoch": 0.1375114085792516,
+      "grad_norm": 2.0600829124450684,
+      "learning_rate": 4.296577946768061e-05,
+      "loss": 0.5259,
+      "step": 226
+    },
+    {
+      "epoch": 0.13811986613933677,
+      "grad_norm": 1.8912417888641357,
+      "learning_rate": 4.3155893536121676e-05,
+      "loss": 0.5448,
+      "step": 227
+    },
+    {
+      "epoch": 0.13872832369942195,
+      "grad_norm": 1.7389650344848633,
+      "learning_rate": 4.334600760456274e-05,
+      "loss": 0.5143,
+      "step": 228
+    },
+    {
+      "epoch": 0.13933678125950716,
+      "grad_norm": 1.844450831413269,
+      "learning_rate": 4.35361216730038e-05,
+      "loss": 0.542,
+      "step": 229
+    },
+    {
+      "epoch": 0.13994523881959234,
+      "grad_norm": 1.8148777484893799,
+      "learning_rate": 4.3726235741444873e-05,
+      "loss": 0.5151,
+      "step": 230
+    },
+    {
+      "epoch": 0.14055369637967752,
+      "grad_norm": 1.9488646984100342,
+      "learning_rate": 4.391634980988593e-05,
+      "loss": 0.6014,
+      "step": 231
+    },
+    {
+      "epoch": 0.1411621539397627,
+      "grad_norm": 1.8368064165115356,
+      "learning_rate": 4.4106463878327e-05,
+      "loss": 0.5752,
+      "step": 232
+    },
+    {
+      "epoch": 0.14177061149984788,
+      "grad_norm": 2.520059823989868,
+      "learning_rate": 4.4296577946768064e-05,
+      "loss": 0.5604,
+      "step": 233
+    },
+    {
+      "epoch": 0.14237906905993308,
+      "grad_norm": 1.6602476835250854,
+      "learning_rate": 4.448669201520913e-05,
+      "loss": 0.5702,
+      "step": 234
+    },
+    {
+      "epoch": 0.14298752662001826,
+      "grad_norm": 2.0854058265686035,
+      "learning_rate": 4.467680608365019e-05,
+      "loss": 0.6663,
+      "step": 235
+    },
+    {
+      "epoch": 0.14359598418010344,
+      "grad_norm": 1.8286073207855225,
+      "learning_rate": 4.4866920152091254e-05,
+      "loss": 0.5896,
+      "step": 236
+    },
+    {
+      "epoch": 0.14420444174018862,
+      "grad_norm": 1.8150510787963867,
+      "learning_rate": 4.5057034220532325e-05,
+      "loss": 0.5698,
+      "step": 237
+    },
+    {
+      "epoch": 0.1448128993002738,
+      "grad_norm": 1.7325129508972168,
+      "learning_rate": 4.524714828897338e-05,
+      "loss": 0.5399,
+      "step": 238
+    },
+    {
+      "epoch": 0.14542135686035898,
+      "grad_norm": 1.7553200721740723,
+      "learning_rate": 4.543726235741445e-05,
+      "loss": 0.5335,
+      "step": 239
+    },
+    {
+      "epoch": 0.1460298144204442,
+      "grad_norm": 1.677822470664978,
+      "learning_rate": 4.5627376425855515e-05,
+      "loss": 0.5662,
+      "step": 240
+    },
+    {
+      "epoch": 0.14663827198052937,
+      "grad_norm": 2.0692317485809326,
+      "learning_rate": 4.581749049429658e-05,
+      "loss": 0.7082,
+      "step": 241
+    },
+    {
+      "epoch": 0.14724672954061455,
+      "grad_norm": 1.5937222242355347,
+      "learning_rate": 4.600760456273764e-05,
+      "loss": 0.5736,
+      "step": 242
+    },
+    {
+      "epoch": 0.14785518710069973,
+      "grad_norm": 1.6691651344299316,
+      "learning_rate": 4.619771863117871e-05,
+      "loss": 0.5081,
+      "step": 243
+    },
+    {
+      "epoch": 0.1484636446607849,
+      "grad_norm": 1.916918396949768,
+      "learning_rate": 4.6387832699619776e-05,
+      "loss": 0.5834,
+      "step": 244
+    },
+    {
+      "epoch": 0.14907210222087008,
+      "grad_norm": 1.7552460432052612,
+      "learning_rate": 4.657794676806084e-05,
+      "loss": 0.5644,
+      "step": 245
+    },
+    {
+      "epoch": 0.1496805597809553,
+      "grad_norm": 1.794980764389038,
+      "learning_rate": 4.67680608365019e-05,
+      "loss": 0.5694,
+      "step": 246
+    },
+    {
+      "epoch": 0.15028901734104047,
+      "grad_norm": 1.563038945198059,
+      "learning_rate": 4.695817490494297e-05,
+      "loss": 0.5564,
+      "step": 247
+    },
+    {
+      "epoch": 0.15089747490112565,
+      "grad_norm": 1.6664254665374756,
+      "learning_rate": 4.714828897338403e-05,
+      "loss": 0.5837,
+      "step": 248
+    },
+    {
+      "epoch": 0.15150593246121083,
+      "grad_norm": 1.7116878032684326,
+      "learning_rate": 4.73384030418251e-05,
+      "loss": 0.5491,
+      "step": 249
+    },
+    {
+      "epoch": 0.152114390021296,
+      "grad_norm": 1.987626075744629,
+      "learning_rate": 4.7528517110266163e-05,
+      "loss": 0.6203,
+      "step": 250
+    },
+    {
+      "epoch": 0.1527228475813812,
+      "grad_norm": 1.6196045875549316,
+      "learning_rate": 4.771863117870723e-05,
+      "loss": 0.5393,
+      "step": 251
+    },
+    {
+      "epoch": 0.1533313051414664,
+      "grad_norm": 1.8573944568634033,
+      "learning_rate": 4.790874524714829e-05,
+      "loss": 0.575,
+      "step": 252
+    },
+    {
+      "epoch": 0.15393976270155157,
+      "grad_norm": 1.9413925409317017,
+      "learning_rate": 4.8098859315589354e-05,
+      "loss": 0.5713,
+      "step": 253
+    },
+    {
+      "epoch": 0.15454822026163675,
+      "grad_norm": 1.883753776550293,
+      "learning_rate": 4.8288973384030424e-05,
+      "loss": 0.5714,
+      "step": 254
+    },
+    {
+      "epoch": 0.15515667782172193,
+      "grad_norm": 1.7571375370025635,
+      "learning_rate": 4.847908745247148e-05,
+      "loss": 0.6265,
+      "step": 255
+    },
+    {
+      "epoch": 0.1557651353818071,
+      "grad_norm": 1.780029058456421,
+      "learning_rate": 4.866920152091255e-05,
+      "loss": 0.5677,
+      "step": 256
+    },
+    {
+      "epoch": 0.1563735929418923,
+      "grad_norm": 2.0636849403381348,
+      "learning_rate": 4.8859315589353615e-05,
+      "loss": 0.5038,
+      "step": 257
+    },
+    {
+      "epoch": 0.1569820505019775,
+      "grad_norm": 1.6824311017990112,
+      "learning_rate": 4.904942965779468e-05,
+      "loss": 0.5905,
+      "step": 258
+    },
+    {
+      "epoch": 0.15759050806206268,
+      "grad_norm": 2.048978328704834,
+      "learning_rate": 4.923954372623574e-05,
+      "loss": 0.6073,
+      "step": 259
+    },
+    {
+      "epoch": 0.15819896562214786,
+      "grad_norm": 2.075801134109497,
+      "learning_rate": 4.942965779467681e-05,
+      "loss": 0.5321,
+      "step": 260
+    },
+    {
+      "epoch": 0.15880742318223304,
+      "grad_norm": 1.8397325277328491,
+      "learning_rate": 4.9619771863117875e-05,
+      "loss": 0.6598,
+      "step": 261
+    },
+    {
+      "epoch": 0.15941588074231822,
+      "grad_norm": 2.7808053493499756,
+      "learning_rate": 4.980988593155894e-05,
+      "loss": 0.5265,
+      "step": 262
+    },
+    {
+      "epoch": 0.1600243383024034,
+      "grad_norm": 2.129298686981201,
+      "learning_rate": 5e-05,
+      "loss": 0.5235,
+      "step": 263
+    },
+    {
+      "epoch": 0.1606327958624886,
+      "grad_norm": 6.710923671722412,
+      "learning_rate": 4.999999690052103e-05,
+      "loss": 0.562,
+      "step": 264
+    },
+    {
+      "epoch": 0.16124125342257378,
+      "grad_norm": 4.849277973175049,
+      "learning_rate": 4.9999987602084876e-05,
+      "loss": 0.5677,
+      "step": 265
+    },
+    {
+      "epoch": 0.16184971098265896,
+      "grad_norm": 1.6120680570602417,
+      "learning_rate": 4.999997210469385e-05,
+      "loss": 0.5322,
+      "step": 266
+    },
+    {
+      "epoch": 0.16245816854274414,
+      "grad_norm": 27.193222045898438,
+      "learning_rate": 4.9999950408351784e-05,
+      "loss": 0.7448,
+      "step": 267
+    },
+    {
+      "epoch": 0.16306662610282932,
+      "grad_norm": 2.693523645401001,
+      "learning_rate": 4.999992251306407e-05,
+      "loss": 0.6533,
+      "step": 268
+    },
+    {
+      "epoch": 0.1636750836629145,
+      "grad_norm": 1.7629886865615845,
+      "learning_rate": 4.999988841883763e-05,
+      "loss": 0.5668,
+      "step": 269
+    },
+    {
+      "epoch": 0.1642835412229997,
+      "grad_norm": 1.9537672996520996,
+      "learning_rate": 4.999984812568089e-05,
+      "loss": 0.5254,
+      "step": 270
+    },
+    {
+      "epoch": 0.1648919987830849,
+      "grad_norm": 4.926598072052002,
+      "learning_rate": 4.999980163360388e-05,
+      "loss": 0.5795,
+      "step": 271
+    },
+    {
+      "epoch": 0.16550045634317007,
+      "grad_norm": 2.0768604278564453,
+      "learning_rate": 4.9999748942618094e-05,
+      "loss": 0.6014,
+      "step": 272
+    },
+    {
+      "epoch": 0.16610891390325525,
+      "grad_norm": 2.3579049110412598,
+      "learning_rate": 4.999969005273661e-05,
+      "loss": 0.6518,
+      "step": 273
+    },
+    {
+      "epoch": 0.16671737146334042,
+      "grad_norm": 2.4885222911834717,
+      "learning_rate": 4.9999624963974045e-05,
+      "loss": 0.6194,
+      "step": 274
+    },
+    {
+      "epoch": 0.1673258290234256,
+      "grad_norm": 2.2712252140045166,
+      "learning_rate": 4.999955367634652e-05,
+      "loss": 0.6094,
+      "step": 275
+    },
+    {
+      "epoch": 0.1679342865835108,
+      "grad_norm": 1.9794126749038696,
+      "learning_rate": 4.999947618987171e-05,
+      "loss": 0.5214,
+      "step": 276
+    },
+    {
+      "epoch": 0.168542744143596,
+      "grad_norm": 1.6654919385910034,
+      "learning_rate": 4.999939250456884e-05,
+      "loss": 0.5547,
+      "step": 277
+    },
+    {
+      "epoch": 0.16915120170368117,
+      "grad_norm": 1.932141900062561,
+      "learning_rate": 4.999930262045865e-05,
+      "loss": 0.5427,
+      "step": 278
+    },
+    {
+      "epoch": 0.16975965926376635,
+      "grad_norm": 1.752447247505188,
+      "learning_rate": 4.999920653756344e-05,
+      "loss": 0.5661,
+      "step": 279
+    },
+    {
+      "epoch": 0.17036811682385153,
+      "grad_norm": 2.4955854415893555,
+      "learning_rate": 4.9999104255907015e-05,
+      "loss": 0.6635,
+      "step": 280
+    },
+    {
+      "epoch": 0.1709765743839367,
+      "grad_norm": 1.774271011352539,
+      "learning_rate": 4.999899577551476e-05,
+      "loss": 0.6433,
+      "step": 281
+    },
+    {
+      "epoch": 0.17158503194402192,
+      "grad_norm": 1.7379838228225708,
+      "learning_rate": 4.9998881096413554e-05,
+      "loss": 0.4936,
+      "step": 282
+    },
+    {
+      "epoch": 0.1721934895041071,
+      "grad_norm": 1.5458259582519531,
+      "learning_rate": 4.9998760218631845e-05,
+      "loss": 0.5219,
+      "step": 283
+    },
+    {
+      "epoch": 0.17280194706419227,
+      "grad_norm": 1.8059486150741577,
+      "learning_rate": 4.99986331421996e-05,
+      "loss": 0.5384,
+      "step": 284
+    },
+    {
+      "epoch": 0.17341040462427745,
+      "grad_norm": 1.9652469158172607,
+      "learning_rate": 4.999849986714833e-05,
+      "loss": 0.6101,
+      "step": 285
+    },
+    {
+      "epoch": 0.17401886218436263,
+      "grad_norm": 1.7066302299499512,
+      "learning_rate": 4.999836039351108e-05,
+      "loss": 0.5981,
+      "step": 286
+    },
+    {
+      "epoch": 0.1746273197444478,
+      "grad_norm": 1.5256247520446777,
+      "learning_rate": 4.999821472132244e-05,
+      "loss": 0.5109,
+      "step": 287
+    },
+    {
+      "epoch": 0.17523577730453302,
+      "grad_norm": 1.713955283164978,
+      "learning_rate": 4.999806285061852e-05,
+      "loss": 0.557,
+      "step": 288
+    },
+    {
+      "epoch": 0.1758442348646182,
+      "grad_norm": 1.8485389947891235,
+      "learning_rate": 4.999790478143699e-05,
+      "loss": 0.575,
+      "step": 289
+    },
+    {
+      "epoch": 0.17645269242470338,
+      "grad_norm": 1.7782764434814453,
+      "learning_rate": 4.999774051381704e-05,
+      "loss": 0.5693,
+      "step": 290
+    },
+    {
+      "epoch": 0.17706114998478856,
+      "grad_norm": 2.34918212890625,
+      "learning_rate": 4.999757004779939e-05,
+      "loss": 0.6003,
+      "step": 291
+    },
+    {
+      "epoch": 0.17766960754487374,
+      "grad_norm": 1.9476317167282104,
+      "learning_rate": 4.999739338342633e-05,
+      "loss": 0.6231,
+      "step": 292
+    },
+    {
+      "epoch": 0.17827806510495892,
+      "grad_norm": 1.6623529195785522,
+      "learning_rate": 4.999721052074164e-05,
+      "loss": 0.5573,
+      "step": 293
+    },
+    {
+      "epoch": 0.17888652266504412,
+      "grad_norm": 1.4444576501846313,
+      "learning_rate": 4.999702145979069e-05,
+      "loss": 0.5658,
+      "step": 294
+    },
+    {
+      "epoch": 0.1794949802251293,
+      "grad_norm": 1.9570586681365967,
+      "learning_rate": 4.9996826200620336e-05,
+      "loss": 0.538,
+      "step": 295
+    },
+    {
+      "epoch": 0.18010343778521448,
+      "grad_norm": 1.6489269733428955,
+      "learning_rate": 4.999662474327901e-05,
+      "loss": 0.6005,
+      "step": 296
+    },
+    {
+      "epoch": 0.18071189534529966,
+      "grad_norm": 1.8315165042877197,
+      "learning_rate": 4.999641708781665e-05,
+      "loss": 0.5694,
+      "step": 297
+    },
+    {
+      "epoch": 0.18132035290538484,
+      "grad_norm": 1.9754319190979004,
+      "learning_rate": 4.9996203234284755e-05,
+      "loss": 0.5658,
+      "step": 298
+    },
+    {
+      "epoch": 0.18192881046547002,
+      "grad_norm": 1.5500773191452026,
+      "learning_rate": 4.999598318273636e-05,
+      "loss": 0.5115,
+      "step": 299
+    },
+    {
+      "epoch": 0.18253726802555523,
+      "grad_norm": 2.730600595474243,
+      "learning_rate": 4.999575693322601e-05,
+      "loss": 0.5655,
+      "step": 300
+    },
+    {
+      "epoch": 0.1831457255856404,
+      "grad_norm": 2.1181812286376953,
+      "learning_rate": 4.9995524485809816e-05,
+      "loss": 0.5391,
+      "step": 301
+    },
+    {
+      "epoch": 0.18375418314572559,
+      "grad_norm": 1.870579481124878,
+      "learning_rate": 4.9995285840545425e-05,
+      "loss": 0.5846,
+      "step": 302
+    },
+    {
+      "epoch": 0.18436264070581077,
+      "grad_norm": 1.8520623445510864,
+      "learning_rate": 4.9995040997491993e-05,
+      "loss": 0.5664,
+      "step": 303
+    },
+    {
+      "epoch": 0.18497109826589594,
+      "grad_norm": 1.7064025402069092,
+      "learning_rate": 4.999478995671024e-05,
+      "loss": 0.5567,
+      "step": 304
+    },
+    {
+      "epoch": 0.18557955582598112,
+      "grad_norm": 2.016923427581787,
+      "learning_rate": 4.99945327182624e-05,
+      "loss": 0.5576,
+      "step": 305
+    },
+    {
+      "epoch": 0.18618801338606633,
+      "grad_norm": 1.6377376317977905,
+      "learning_rate": 4.999426928221229e-05,
+      "loss": 0.5462,
+      "step": 306
+    },
+    {
+      "epoch": 0.1867964709461515,
+      "grad_norm": 1.6940040588378906,
+      "learning_rate": 4.9993999648625197e-05,
+      "loss": 0.5148,
+      "step": 307
+    },
+    {
+      "epoch": 0.1874049285062367,
+      "grad_norm": 2.552048921585083,
+      "learning_rate": 4.9993723817567996e-05,
+      "loss": 0.6583,
+      "step": 308
+    },
+    {
+      "epoch": 0.18801338606632187,
+      "grad_norm": 1.4757635593414307,
+      "learning_rate": 4.9993441789109074e-05,
+      "loss": 0.5491,
+      "step": 309
+    },
+    {
+      "epoch": 0.18862184362640705,
+      "grad_norm": 1.6392966508865356,
+      "learning_rate": 4.999315356331837e-05,
+      "loss": 0.5338,
+      "step": 310
+    },
+    {
+      "epoch": 0.18923030118649226,
+      "grad_norm": 1.7045416831970215,
+      "learning_rate": 4.999285914026734e-05,
+      "loss": 0.627,
+      "step": 311
+    },
+    {
+      "epoch": 0.18983875874657744,
+      "grad_norm": 1.4857970476150513,
+      "learning_rate": 4.999255852002901e-05,
+      "loss": 0.6172,
+      "step": 312
+    },
+    {
+      "epoch": 0.19044721630666261,
+      "grad_norm": 3.2289485931396484,
+      "learning_rate": 4.9992251702677904e-05,
+      "loss": 0.5705,
+      "step": 313
+    },
+    {
+      "epoch": 0.1910556738667478,
+      "grad_norm": 1.866363763809204,
+      "learning_rate": 4.9991938688290105e-05,
+      "loss": 0.5102,
+      "step": 314
+    },
+    {
+      "epoch": 0.19166413142683297,
+      "grad_norm": 2.0050652027130127,
+      "learning_rate": 4.999161947694322e-05,
+      "loss": 0.5714,
+      "step": 315
+    },
+    {
+      "epoch": 0.19227258898691815,
+      "grad_norm": 1.7371540069580078,
+      "learning_rate": 4.9991294068716416e-05,
+      "loss": 0.5933,
+      "step": 316
+    },
+    {
+      "epoch": 0.19288104654700336,
+      "grad_norm": 1.7616161108016968,
+      "learning_rate": 4.9990962463690364e-05,
+      "loss": 0.5356,
+      "step": 317
+    },
+    {
+      "epoch": 0.19348950410708854,
+      "grad_norm": 1.7778440713882446,
+      "learning_rate": 4.99906246619473e-05,
+      "loss": 0.5461,
+      "step": 318
+    },
+    {
+      "epoch": 0.19409796166717372,
+      "grad_norm": 1.7068995237350464,
+      "learning_rate": 4.999028066357098e-05,
+      "loss": 0.509,
+      "step": 319
+    },
+    {
+      "epoch": 0.1947064192272589,
+      "grad_norm": 1.5592914819717407,
+      "learning_rate": 4.9989930468646703e-05,
+      "loss": 0.5702,
+      "step": 320
+    },
+    {
+      "epoch": 0.19531487678734408,
+      "grad_norm": 1.4004979133605957,
+      "learning_rate": 4.99895740772613e-05,
+      "loss": 0.5483,
+      "step": 321
+    },
+    {
+      "epoch": 0.19592333434742926,
+      "grad_norm": 1.4833515882492065,
+      "learning_rate": 4.998921148950314e-05,
+      "loss": 0.5739,
+      "step": 322
+    },
+    {
+      "epoch": 0.19653179190751446,
+      "grad_norm": 1.715824007987976,
+      "learning_rate": 4.998884270546214e-05,
+      "loss": 0.5488,
+      "step": 323
+    },
+    {
+      "epoch": 0.19714024946759964,
+      "grad_norm": 1.557228684425354,
+      "learning_rate": 4.998846772522972e-05,
+      "loss": 0.5267,
+      "step": 324
+    },
+    {
+      "epoch": 0.19774870702768482,
+      "grad_norm": 1.7454534769058228,
+      "learning_rate": 4.998808654889888e-05,
+      "loss": 0.533,
+      "step": 325
+    },
+    {
+      "epoch": 0.19835716458777,
+      "grad_norm": 1.6410638093948364,
+      "learning_rate": 4.998769917656414e-05,
+      "loss": 0.5841,
+      "step": 326
+    },
+    {
+      "epoch": 0.19896562214785518,
+      "grad_norm": 1.7179322242736816,
+      "learning_rate": 4.998730560832154e-05,
+      "loss": 0.6051,
+      "step": 327
+    },
+    {
+      "epoch": 0.19957407970794036,
+      "grad_norm": 1.5742508172988892,
+      "learning_rate": 4.9986905844268667e-05,
+      "loss": 0.501,
+      "step": 328
+    },
+    {
+      "epoch": 0.20018253726802557,
+      "grad_norm": 1.944952130317688,
+      "learning_rate": 4.998649988450465e-05,
+      "loss": 0.543,
+      "step": 329
+    },
+    {
+      "epoch": 0.20079099482811075,
+      "grad_norm": 2.1101162433624268,
+      "learning_rate": 4.998608772913015e-05,
+      "loss": 0.7782,
+      "step": 330
+    },
+    {
+      "epoch": 0.20139945238819593,
+      "grad_norm": 1.778678297996521,
+      "learning_rate": 4.998566937824737e-05,
+      "loss": 0.5768,
+      "step": 331
+    },
+    {
+      "epoch": 0.2020079099482811,
+      "grad_norm": 1.5352647304534912,
+      "learning_rate": 4.9985244831960034e-05,
+      "loss": 0.5784,
+      "step": 332
+    },
+    {
+      "epoch": 0.20261636750836629,
+      "grad_norm": 1.6977012157440186,
+      "learning_rate": 4.998481409037342e-05,
+      "loss": 0.537,
+      "step": 333
+    },
+    {
+      "epoch": 0.20322482506845146,
+      "grad_norm": 1.7605286836624146,
+      "learning_rate": 4.9984377153594327e-05,
+      "loss": 0.5381,
+      "step": 334
+    },
+    {
+      "epoch": 0.20383328262853667,
+      "grad_norm": 2.0058248043060303,
+      "learning_rate": 4.99839340217311e-05,
+      "loss": 0.6097,
+      "step": 335
+    },
+    {
+      "epoch": 0.20444174018862185,
+      "grad_norm": 2.2246720790863037,
+      "learning_rate": 4.9983484694893615e-05,
+      "loss": 0.5455,
+      "step": 336
+    },
+    {
+      "epoch": 0.20505019774870703,
+      "grad_norm": 1.8565617799758911,
+      "learning_rate": 4.99830291731933e-05,
+      "loss": 0.6721,
+      "step": 337
+    },
+    {
+      "epoch": 0.2056586553087922,
+      "grad_norm": 1.691021203994751,
+      "learning_rate": 4.998256745674308e-05,
+      "loss": 0.58,
+      "step": 338
+    },
+    {
+      "epoch": 0.2062671128688774,
+      "grad_norm": 2.6199169158935547,
+      "learning_rate": 4.998209954565746e-05,
+      "loss": 0.6639,
+      "step": 339
+    },
+    {
+      "epoch": 0.20687557042896257,
+      "grad_norm": 1.632177710533142,
+      "learning_rate": 4.998162544005246e-05,
+      "loss": 0.5949,
+      "step": 340
+    },
+    {
+      "epoch": 0.20748402798904778,
+      "grad_norm": 2.9773964881896973,
+      "learning_rate": 4.9981145140045634e-05,
+      "loss": 0.547,
+      "step": 341
+    },
+    {
+      "epoch": 0.20809248554913296,
+      "grad_norm": 2.0288548469543457,
+      "learning_rate": 4.998065864575608e-05,
+      "loss": 0.619,
+      "step": 342
+    },
+    {
+      "epoch": 0.20870094310921813,
+      "grad_norm": 1.6673290729522705,
+      "learning_rate": 4.998016595730442e-05,
+      "loss": 0.592,
+      "step": 343
+    },
+    {
+      "epoch": 0.2093094006693033,
+      "grad_norm": 1.5016260147094727,
+      "learning_rate": 4.997966707481284e-05,
+      "loss": 0.6014,
+      "step": 344
+    },
+    {
+      "epoch": 0.2099178582293885,
+      "grad_norm": 2.3136346340179443,
+      "learning_rate": 4.9979161998405024e-05,
+      "loss": 0.5583,
+      "step": 345
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 1.6642321348190308,
+      "learning_rate": 4.997865072820621e-05,
+      "loss": 0.5468,
+      "step": 346
+    },
+    {
+      "epoch": 0.21113477334955888,
+      "grad_norm": 1.8811125755310059,
+      "learning_rate": 4.9978133264343186e-05,
+      "loss": 0.6072,
+      "step": 347
+    },
+    {
+      "epoch": 0.21174323090964406,
+      "grad_norm": 1.5865939855575562,
+      "learning_rate": 4.997760960694424e-05,
+      "loss": 0.509,
+      "step": 348
+    },
+    {
+      "epoch": 0.21235168846972924,
+      "grad_norm": 1.6846966743469238,
+      "learning_rate": 4.9977079756139247e-05,
+      "loss": 0.6028,
+      "step": 349
+    },
+    {
+      "epoch": 0.21296014602981442,
+      "grad_norm": 2.0997796058654785,
+      "learning_rate": 4.997654371205955e-05,
+      "loss": 0.5462,
+      "step": 350
+    },
+    {
+      "epoch": 0.2135686035898996,
+      "grad_norm": 2.4784786701202393,
+      "learning_rate": 4.9976001474838105e-05,
+      "loss": 0.5719,
+      "step": 351
+    },
+    {
+      "epoch": 0.21417706114998478,
+      "grad_norm": 2.554870367050171,
+      "learning_rate": 4.997545304460933e-05,
+      "loss": 0.5194,
+      "step": 352
+    },
+    {
+      "epoch": 0.21478551871006998,
+      "grad_norm": 1.4950354099273682,
+      "learning_rate": 4.997489842150924e-05,
+      "loss": 0.5605,
+      "step": 353
+    },
+    {
+      "epoch": 0.21539397627015516,
+      "grad_norm": 1.4311949014663696,
+      "learning_rate": 4.9974337605675335e-05,
+      "loss": 0.5508,
+      "step": 354
+    },
+    {
+      "epoch": 0.21600243383024034,
+      "grad_norm": 1.7525572776794434,
+      "learning_rate": 4.9973770597246696e-05,
+      "loss": 0.5666,
+      "step": 355
+    },
+    {
+      "epoch": 0.21661089139032552,
+      "grad_norm": 2.4212021827697754,
+      "learning_rate": 4.99731973963639e-05,
+      "loss": 0.5436,
+      "step": 356
+    },
+    {
+      "epoch": 0.2172193489504107,
+      "grad_norm": 6.587791919708252,
+      "learning_rate": 4.997261800316909e-05,
+      "loss": 0.6509,
+      "step": 357
+    },
+    {
+      "epoch": 0.21782780651049588,
+      "grad_norm": 1.5659176111221313,
+      "learning_rate": 4.997203241780592e-05,
+      "loss": 0.5326,
+      "step": 358
+    },
+    {
+      "epoch": 0.2184362640705811,
+      "grad_norm": 1.5538829565048218,
+      "learning_rate": 4.99714406404196e-05,
+      "loss": 0.6136,
+      "step": 359
+    },
+    {
+      "epoch": 0.21904472163066627,
+      "grad_norm": 1.7956405878067017,
+      "learning_rate": 4.997084267115686e-05,
+      "loss": 0.604,
+      "step": 360
+    },
+    {
+      "epoch": 0.21965317919075145,
+      "grad_norm": 1.8146167993545532,
+      "learning_rate": 4.997023851016598e-05,
+      "loss": 0.5673,
+      "step": 361
+    },
+    {
+      "epoch": 0.22026163675083663,
+      "grad_norm": 1.5407328605651855,
+      "learning_rate": 4.996962815759675e-05,
+      "loss": 0.5848,
+      "step": 362
+    },
+    {
+      "epoch": 0.2208700943109218,
+      "grad_norm": 1.560318946838379,
+      "learning_rate": 4.9969011613600525e-05,
+      "loss": 0.4974,
+      "step": 363
+    },
+    {
+      "epoch": 0.22147855187100698,
+      "grad_norm": 1.5309618711471558,
+      "learning_rate": 4.996838887833018e-05,
+      "loss": 0.6155,
+      "step": 364
+    },
+    {
+      "epoch": 0.2220870094310922,
+      "grad_norm": 1.6208105087280273,
+      "learning_rate": 4.9967759951940127e-05,
+      "loss": 0.5615,
+      "step": 365
+    },
+    {
+      "epoch": 0.22269546699117737,
+      "grad_norm": 1.5246007442474365,
+      "learning_rate": 4.996712483458632e-05,
+      "loss": 0.5311,
+      "step": 366
+    },
+    {
+      "epoch": 0.22330392455126255,
+      "grad_norm": 1.9340999126434326,
+      "learning_rate": 4.9966483526426223e-05,
+      "loss": 0.5678,
+      "step": 367
+    },
+    {
+      "epoch": 0.22391238211134773,
+      "grad_norm": 1.4264390468597412,
+      "learning_rate": 4.996583602761887e-05,
+      "loss": 0.5468,
+      "step": 368
+    },
+    {
+      "epoch": 0.2245208396714329,
+      "grad_norm": 1.5261754989624023,
+      "learning_rate": 4.996518233832481e-05,
+      "loss": 0.5593,
+      "step": 369
+    },
+    {
+      "epoch": 0.2251292972315181,
+      "grad_norm": 1.5662906169891357,
+      "learning_rate": 4.996452245870614e-05,
+      "loss": 0.5155,
+      "step": 370
+    },
+    {
+      "epoch": 0.2257377547916033,
+      "grad_norm": 1.6063148975372314,
+      "learning_rate": 4.9963856388926464e-05,
+      "loss": 0.5384,
+      "step": 371
+    },
+    {
+      "epoch": 0.22634621235168847,
+      "grad_norm": 1.5906771421432495,
+      "learning_rate": 4.996318412915095e-05,
+      "loss": 0.5049,
+      "step": 372
+    },
+    {
+      "epoch": 0.22695466991177365,
+      "grad_norm": 1.676358938217163,
+      "learning_rate": 4.9962505679546285e-05,
+      "loss": 0.5344,
+      "step": 373
+    },
+    {
+      "epoch": 0.22756312747185883,
+      "grad_norm": 1.6183301210403442,
+      "learning_rate": 4.9961821040280697e-05,
+      "loss": 0.5611,
+      "step": 374
+    },
+    {
+      "epoch": 0.228171585031944,
+      "grad_norm": 1.670375108718872,
+      "learning_rate": 4.996113021152397e-05,
+      "loss": 0.5533,
+      "step": 375
+    },
+    {
+      "epoch": 0.2287800425920292,
+      "grad_norm": 1.6146153211593628,
+      "learning_rate": 4.996043319344736e-05,
+      "loss": 0.5368,
+      "step": 376
+    },
+    {
+      "epoch": 0.2293885001521144,
+      "grad_norm": 1.5858700275421143,
+      "learning_rate": 4.9959729986223725e-05,
+      "loss": 0.5109,
+      "step": 377
+    },
+    {
+      "epoch": 0.22999695771219958,
+      "grad_norm": 1.696811318397522,
+      "learning_rate": 4.995902059002743e-05,
+      "loss": 0.5879,
+      "step": 378
+    },
+    {
+      "epoch": 0.23060541527228476,
+      "grad_norm": 1.5121724605560303,
+      "learning_rate": 4.995830500503438e-05,
+      "loss": 0.518,
+      "step": 379
+    },
+    {
+      "epoch": 0.23121387283236994,
+      "grad_norm": 1.4365136623382568,
+      "learning_rate": 4.995758323142199e-05,
+      "loss": 0.5384,
+      "step": 380
+    },
+    {
+      "epoch": 0.23182233039245512,
+      "grad_norm": 1.4836268424987793,
+      "learning_rate": 4.995685526936924e-05,
+      "loss": 0.5867,
+      "step": 381
+    },
+    {
+      "epoch": 0.2324307879525403,
+      "grad_norm": 1.8685383796691895,
+      "learning_rate": 4.9956121119056646e-05,
+      "loss": 0.628,
+      "step": 382
+    },
+    {
+      "epoch": 0.2330392455126255,
+      "grad_norm": 1.426859974861145,
+      "learning_rate": 4.9955380780666233e-05,
+      "loss": 0.4974,
+      "step": 383
+    },
+    {
+      "epoch": 0.23364770307271068,
+      "grad_norm": 2.282662868499756,
+      "learning_rate": 4.9954634254381576e-05,
+      "loss": 0.5333,
+      "step": 384
+    },
+    {
+      "epoch": 0.23425616063279586,
+      "grad_norm": 1.881532073020935,
+      "learning_rate": 4.995388154038779e-05,
+      "loss": 0.6067,
+      "step": 385
+    },
+    {
+      "epoch": 0.23486461819288104,
+      "grad_norm": 1.4905718564987183,
+      "learning_rate": 4.9953122638871505e-05,
+      "loss": 0.5767,
+      "step": 386
+    },
+    {
+      "epoch": 0.23547307575296622,
+      "grad_norm": 1.5915948152542114,
+      "learning_rate": 4.99523575500209e-05,
+      "loss": 0.5713,
+      "step": 387
+    },
+    {
+      "epoch": 0.23608153331305143,
+      "grad_norm": 1.8757609128952026,
+      "learning_rate": 4.9951586274025695e-05,
+      "loss": 0.5355,
+      "step": 388
+    },
+    {
+      "epoch": 0.2366899908731366,
+      "grad_norm": 1.7897573709487915,
+      "learning_rate": 4.9950808811077135e-05,
+      "loss": 0.5522,
+      "step": 389
+    },
+    {
+      "epoch": 0.2372984484332218,
+      "grad_norm": 1.5204428434371948,
+      "learning_rate": 4.995002516136797e-05,
+      "loss": 0.5935,
+      "step": 390
+    },
+    {
+      "epoch": 0.23790690599330697,
+      "grad_norm": 1.376928448677063,
+      "learning_rate": 4.994923532509255e-05,
+      "loss": 0.5006,
+      "step": 391
+    },
+    {
+      "epoch": 0.23851536355339215,
+      "grad_norm": 1.6038637161254883,
+      "learning_rate": 4.99484393024467e-05,
+      "loss": 0.4999,
+      "step": 392
+    },
+    {
+      "epoch": 0.23912382111347733,
+      "grad_norm": 1.59404456615448,
+      "learning_rate": 4.99476370936278e-05,
+      "loss": 0.5816,
+      "step": 393
+    },
+    {
+      "epoch": 0.23973227867356253,
+      "grad_norm": 1.748207688331604,
+      "learning_rate": 4.994682869883478e-05,
+      "loss": 0.5596,
+      "step": 394
+    },
+    {
+      "epoch": 0.2403407362336477,
+      "grad_norm": 1.8243966102600098,
+      "learning_rate": 4.994601411826807e-05,
+      "loss": 0.6351,
+      "step": 395
+    },
+    {
+      "epoch": 0.2409491937937329,
+      "grad_norm": 1.70762300491333,
+      "learning_rate": 4.994519335212966e-05,
+      "loss": 0.644,
+      "step": 396
+    },
+    {
+      "epoch": 0.24155765135381807,
+      "grad_norm": 1.5090755224227905,
+      "learning_rate": 4.9944366400623066e-05,
+      "loss": 0.5991,
+      "step": 397
+    },
+    {
+      "epoch": 0.24216610891390325,
+      "grad_norm": 1.3629839420318604,
+      "learning_rate": 4.994353326395334e-05,
+      "loss": 0.4862,
+      "step": 398
+    },
+    {
+      "epoch": 0.24277456647398843,
+      "grad_norm": 1.7295219898223877,
+      "learning_rate": 4.9942693942327054e-05,
+      "loss": 0.5848,
+      "step": 399
+    },
+    {
+      "epoch": 0.24338302403407364,
+      "grad_norm": 1.5481480360031128,
+      "learning_rate": 4.994184843595234e-05,
+      "loss": 0.4668,
+      "step": 400
+    },
+    {
+      "epoch": 0.24399148159415882,
+      "grad_norm": 1.3394376039505005,
+      "learning_rate": 4.994099674503885e-05,
+      "loss": 0.5137,
+      "step": 401
+    },
+    {
+      "epoch": 0.244599939154244,
+      "grad_norm": 1.7005079984664917,
+      "learning_rate": 4.994013886979775e-05,
+      "loss": 0.6468,
+      "step": 402
+    },
+    {
+      "epoch": 0.24520839671432917,
+      "grad_norm": 1.682056188583374,
+      "learning_rate": 4.993927481044176e-05,
+      "loss": 0.5919,
+      "step": 403
+    },
+    {
+      "epoch": 0.24581685427441435,
+      "grad_norm": 1.6984502077102661,
+      "learning_rate": 4.993840456718515e-05,
+      "loss": 0.5864,
+      "step": 404
+    },
+    {
+      "epoch": 0.24642531183449953,
+      "grad_norm": 1.8722606897354126,
+      "learning_rate": 4.993752814024368e-05,
+      "loss": 0.4948,
+      "step": 405
+    },
+    {
+      "epoch": 0.24703376939458474,
+      "grad_norm": 1.4154351949691772,
+      "learning_rate": 4.993664552983469e-05,
+      "loss": 0.5495,
+      "step": 406
+    },
+    {
+      "epoch": 0.24764222695466992,
+      "grad_norm": 1.7346272468566895,
+      "learning_rate": 4.9935756736177006e-05,
+      "loss": 0.6606,
+      "step": 407
+    },
+    {
+      "epoch": 0.2482506845147551,
+      "grad_norm": 1.3692924976348877,
+      "learning_rate": 4.993486175949104e-05,
+      "loss": 0.4961,
+      "step": 408
+    },
+    {
+      "epoch": 0.24885914207484028,
+      "grad_norm": 1.9384909868240356,
+      "learning_rate": 4.993396059999868e-05,
+      "loss": 0.6164,
+      "step": 409
+    },
+    {
+      "epoch": 0.24946759963492546,
+      "grad_norm": 1.8157694339752197,
+      "learning_rate": 4.99330532579234e-05,
+      "loss": 0.5164,
+      "step": 410
+    },
+    {
+      "epoch": 0.25007605719501064,
+      "grad_norm": 1.4756087064743042,
+      "learning_rate": 4.993213973349017e-05,
+      "loss": 0.5389,
+      "step": 411
+    },
+    {
+      "epoch": 0.25068451475509584,
+      "grad_norm": 1.3147797584533691,
+      "learning_rate": 4.9931220026925506e-05,
+      "loss": 0.4488,
+      "step": 412
+    },
+    {
+      "epoch": 0.251292972315181,
+      "grad_norm": 1.9356815814971924,
+      "learning_rate": 4.993029413845746e-05,
+      "loss": 0.5484,
+      "step": 413
+    },
+    {
+      "epoch": 0.2519014298752662,
+      "grad_norm": 1.7992812395095825,
+      "learning_rate": 4.992936206831561e-05,
+      "loss": 0.5857,
+      "step": 414
+    },
+    {
+      "epoch": 0.2525098874353514,
+      "grad_norm": 1.4807167053222656,
+      "learning_rate": 4.9928423816731086e-05,
+      "loss": 0.5957,
+      "step": 415
+    },
+    {
+      "epoch": 0.25311834499543656,
+      "grad_norm": 1.877353549003601,
+      "learning_rate": 4.9927479383936516e-05,
+      "loss": 0.5062,
+      "step": 416
+    },
+    {
+      "epoch": 0.25372680255552177,
+      "grad_norm": 1.6704474687576294,
+      "learning_rate": 4.992652877016608e-05,
+      "loss": 0.6027,
+      "step": 417
+    },
+    {
+      "epoch": 0.2543352601156069,
+      "grad_norm": 1.7480648756027222,
+      "learning_rate": 4.992557197565551e-05,
+      "loss": 0.5166,
+      "step": 418
+    },
+    {
+      "epoch": 0.2549437176756921,
+      "grad_norm": 1.8488813638687134,
+      "learning_rate": 4.992460900064203e-05,
+      "loss": 0.627,
+      "step": 419
+    },
+    {
+      "epoch": 0.2555521752357773,
+      "grad_norm": 1.7881669998168945,
+      "learning_rate": 4.992363984536443e-05,
+      "loss": 0.5289,
+      "step": 420
+    },
+    {
+      "epoch": 0.2561606327958625,
+      "grad_norm": 1.5146019458770752,
+      "learning_rate": 4.9922664510063024e-05,
+      "loss": 0.5334,
+      "step": 421
+    },
+    {
+      "epoch": 0.2567690903559477,
+      "grad_norm": 3.4007725715637207,
+      "learning_rate": 4.992168299497964e-05,
+      "loss": 0.6357,
+      "step": 422
+    },
+    {
+      "epoch": 0.25737754791603284,
+      "grad_norm": 1.778419017791748,
+      "learning_rate": 4.9920695300357664e-05,
+      "loss": 0.5248,
+      "step": 423
+    },
+    {
+      "epoch": 0.25798600547611805,
+      "grad_norm": 1.3506314754486084,
+      "learning_rate": 4.9919701426442e-05,
+      "loss": 0.4993,
+      "step": 424
+    },
+    {
+      "epoch": 0.2585944630362032,
+      "grad_norm": 1.7571451663970947,
+      "learning_rate": 4.991870137347908e-05,
+      "loss": 0.591,
+      "step": 425
+    },
+    {
+      "epoch": 0.2592029205962884,
+      "grad_norm": 1.8246015310287476,
+      "learning_rate": 4.9917695141716884e-05,
+      "loss": 0.5652,
+      "step": 426
+    },
+    {
+      "epoch": 0.2598113781563736,
+      "grad_norm": 1.585374355316162,
+      "learning_rate": 4.991668273140492e-05,
+      "loss": 0.494,
+      "step": 427
+    },
+    {
+      "epoch": 0.26041983571645877,
+      "grad_norm": 1.5021830797195435,
+      "learning_rate": 4.991566414279421e-05,
+      "loss": 0.5471,
+      "step": 428
+    },
+    {
+      "epoch": 0.261028293276544,
+      "grad_norm": 1.493720293045044,
+      "learning_rate": 4.991463937613733e-05,
+      "loss": 0.5409,
+      "step": 429
+    },
+    {
+      "epoch": 0.26163675083662913,
+      "grad_norm": 1.550927758216858,
+      "learning_rate": 4.991360843168838e-05,
+      "loss": 0.5049,
+      "step": 430
+    },
+    {
+      "epoch": 0.26224520839671434,
+      "grad_norm": 1.793480634689331,
+      "learning_rate": 4.991257130970299e-05,
+      "loss": 0.5746,
+      "step": 431
+    },
+    {
+      "epoch": 0.2628536659567995,
+      "grad_norm": 1.3120859861373901,
+      "learning_rate": 4.991152801043832e-05,
+      "loss": 0.48,
+      "step": 432
+    },
+    {
+      "epoch": 0.2634621235168847,
+      "grad_norm": 1.489817500114441,
+      "learning_rate": 4.991047853415307e-05,
+      "loss": 0.5376,
+      "step": 433
+    },
+    {
+      "epoch": 0.2640705810769699,
+      "grad_norm": 1.668395757675171,
+      "learning_rate": 4.990942288110746e-05,
+      "loss": 0.5131,
+      "step": 434
+    },
+    {
+      "epoch": 0.26467903863705505,
+      "grad_norm": 1.8639039993286133,
+      "learning_rate": 4.9908361051563244e-05,
+      "loss": 0.5146,
+      "step": 435
+    },
+    {
+      "epoch": 0.26528749619714026,
+      "grad_norm": 1.944084644317627,
+      "learning_rate": 4.990729304578373e-05,
+      "loss": 0.5916,
+      "step": 436
+    },
+    {
+      "epoch": 0.2658959537572254,
+      "grad_norm": 1.5414292812347412,
+      "learning_rate": 4.990621886403373e-05,
+      "loss": 0.5606,
+      "step": 437
+    },
+    {
+      "epoch": 0.2665044113173106,
+      "grad_norm": 1.4970581531524658,
+      "learning_rate": 4.990513850657958e-05,
+      "loss": 0.5137,
+      "step": 438
+    },
+    {
+      "epoch": 0.2671128688773958,
+      "grad_norm": 1.5545778274536133,
+      "learning_rate": 4.990405197368919e-05,
+      "loss": 0.5935,
+      "step": 439
+    },
+    {
+      "epoch": 0.267721326437481,
+      "grad_norm": 1.836313009262085,
+      "learning_rate": 4.9902959265631966e-05,
+      "loss": 0.4994,
+      "step": 440
+    },
+    {
+      "epoch": 0.2683297839975662,
+      "grad_norm": 2.459785223007202,
+      "learning_rate": 4.990186038267884e-05,
+      "loss": 0.5529,
+      "step": 441
+    },
+    {
+      "epoch": 0.26893824155765134,
+      "grad_norm": 1.6246469020843506,
+      "learning_rate": 4.990075532510231e-05,
+      "loss": 0.5785,
+      "step": 442
+    },
+    {
+      "epoch": 0.26954669911773654,
+      "grad_norm": 1.563841700553894,
+      "learning_rate": 4.989964409317637e-05,
+      "loss": 0.5102,
+      "step": 443
+    },
+    {
+      "epoch": 0.2701551566778217,
+      "grad_norm": 1.729370355606079,
+      "learning_rate": 4.9898526687176554e-05,
+      "loss": 0.6466,
+      "step": 444
+    },
+    {
+      "epoch": 0.2707636142379069,
+      "grad_norm": 1.5674041509628296,
+      "learning_rate": 4.989740310737995e-05,
+      "loss": 0.5578,
+      "step": 445
+    },
+    {
+      "epoch": 0.2713720717979921,
+      "grad_norm": 2.1116135120391846,
+      "learning_rate": 4.989627335406515e-05,
+      "loss": 0.6759,
+      "step": 446
+    },
+    {
+      "epoch": 0.27198052935807726,
+      "grad_norm": 1.3847877979278564,
+      "learning_rate": 4.9895137427512284e-05,
+      "loss": 0.4906,
+      "step": 447
+    },
+    {
+      "epoch": 0.27258898691816247,
+      "grad_norm": 2.1487720012664795,
+      "learning_rate": 4.989399532800302e-05,
+      "loss": 0.4625,
+      "step": 448
+    },
+    {
+      "epoch": 0.2731974444782476,
+      "grad_norm": 2.71392560005188,
+      "learning_rate": 4.989284705582055e-05,
+      "loss": 0.5155,
+      "step": 449
+    },
+    {
+      "epoch": 0.2738059020383328,
+      "grad_norm": 1.6772319078445435,
+      "learning_rate": 4.989169261124958e-05,
+      "loss": 0.5715,
+      "step": 450
+    },
+    {
+      "epoch": 0.27441435959841803,
+      "grad_norm": 1.6435317993164062,
+      "learning_rate": 4.9890531994576394e-05,
+      "loss": 0.5688,
+      "step": 451
+    },
+    {
+      "epoch": 0.2750228171585032,
+      "grad_norm": 1.4484583139419556,
+      "learning_rate": 4.9889365206088755e-05,
+      "loss": 0.5269,
+      "step": 452
+    },
+    {
+      "epoch": 0.2756312747185884,
+      "grad_norm": 1.4935004711151123,
+      "learning_rate": 4.9888192246075986e-05,
+      "loss": 0.5619,
+      "step": 453
+    },
+    {
+      "epoch": 0.27623973227867354,
+      "grad_norm": 2.1516928672790527,
+      "learning_rate": 4.988701311482893e-05,
+      "loss": 0.6126,
+      "step": 454
+    },
+    {
+      "epoch": 0.27684818983875875,
+      "grad_norm": 2.0537219047546387,
+      "learning_rate": 4.988582781263997e-05,
+      "loss": 0.5805,
+      "step": 455
+    },
+    {
+      "epoch": 0.2774566473988439,
+      "grad_norm": 1.384886622428894,
+      "learning_rate": 4.9884636339803e-05,
+      "loss": 0.5707,
+      "step": 456
+    },
+    {
+      "epoch": 0.2780651049589291,
+      "grad_norm": 1.4588273763656616,
+      "learning_rate": 4.988343869661346e-05,
+      "loss": 0.5806,
+      "step": 457
+    },
+    {
+      "epoch": 0.2786735625190143,
+      "grad_norm": 1.3114267587661743,
+      "learning_rate": 4.988223488336832e-05,
+      "loss": 0.5397,
+      "step": 458
+    },
+    {
+      "epoch": 0.27928202007909947,
+      "grad_norm": 1.5628232955932617,
+      "learning_rate": 4.988102490036606e-05,
+      "loss": 0.5486,
+      "step": 459
+    },
+    {
+      "epoch": 0.2798904776391847,
+      "grad_norm": 1.5347833633422852,
+      "learning_rate": 4.987980874790673e-05,
+      "loss": 0.492,
+      "step": 460
+    },
+    {
+      "epoch": 0.2804989351992698,
+      "grad_norm": 1.5386556386947632,
+      "learning_rate": 4.9878586426291864e-05,
+      "loss": 0.5133,
+      "step": 461
+    },
+    {
+      "epoch": 0.28110739275935503,
+      "grad_norm": 1.603733777999878,
+      "learning_rate": 4.987735793582456e-05,
+      "loss": 0.5197,
+      "step": 462
+    },
+    {
+      "epoch": 0.28171585031944024,
+      "grad_norm": 1.7382800579071045,
+      "learning_rate": 4.987612327680943e-05,
+      "loss": 0.6745,
+      "step": 463
+    },
+    {
+      "epoch": 0.2823243078795254,
+      "grad_norm": 1.6408590078353882,
+      "learning_rate": 4.987488244955261e-05,
+      "loss": 0.5331,
+      "step": 464
+    },
+    {
+      "epoch": 0.2829327654396106,
+      "grad_norm": 1.5231281518936157,
+      "learning_rate": 4.987363545436178e-05,
+      "loss": 0.5993,
+      "step": 465
+    },
+    {
+      "epoch": 0.28354122299969575,
+      "grad_norm": 1.5222764015197754,
+      "learning_rate": 4.9872382291546136e-05,
+      "loss": 0.4993,
+      "step": 466
+    },
+    {
+      "epoch": 0.28414968055978096,
+      "grad_norm": 1.3687183856964111,
+      "learning_rate": 4.9871122961416417e-05,
+      "loss": 0.5918,
+      "step": 467
+    },
+    {
+      "epoch": 0.28475813811986617,
+      "grad_norm": 1.5194501876831055,
+      "learning_rate": 4.9869857464284885e-05,
+      "loss": 0.5499,
+      "step": 468
+    },
+    {
+      "epoch": 0.2853665956799513,
+      "grad_norm": 1.4260061979293823,
+      "learning_rate": 4.986858580046534e-05,
+      "loss": 0.5771,
+      "step": 469
+    },
+    {
+      "epoch": 0.2859750532400365,
+      "grad_norm": 1.58305823802948,
+      "learning_rate": 4.986730797027307e-05,
+      "loss": 0.5616,
+      "step": 470
+    },
+    {
+      "epoch": 0.2865835108001217,
+      "grad_norm": 1.5807416439056396,
+      "learning_rate": 4.9866023974024954e-05,
+      "loss": 0.513,
+      "step": 471
+    },
+    {
+      "epoch": 0.2871919683602069,
+      "grad_norm": 1.3939672708511353,
+      "learning_rate": 4.986473381203937e-05,
+      "loss": 0.587,
+      "step": 472
+    },
+    {
+      "epoch": 0.28780042592029204,
+      "grad_norm": 1.4460794925689697,
+      "learning_rate": 4.98634374846362e-05,
+      "loss": 0.5267,
+      "step": 473
+    },
+    {
+      "epoch": 0.28840888348037724,
+      "grad_norm": 1.4511432647705078,
+      "learning_rate": 4.986213499213689e-05,
+      "loss": 0.497,
+      "step": 474
+    },
+    {
+      "epoch": 0.28901734104046245,
+      "grad_norm": 1.877480149269104,
+      "learning_rate": 4.986082633486442e-05,
+      "loss": 0.5801,
+      "step": 475
+    },
+    {
+      "epoch": 0.2896257986005476,
+      "grad_norm": 1.3986396789550781,
+      "learning_rate": 4.985951151314326e-05,
+      "loss": 0.5483,
+      "step": 476
+    },
+    {
+      "epoch": 0.2902342561606328,
+      "grad_norm": 2.344064474105835,
+      "learning_rate": 4.985819052729944e-05,
+      "loss": 0.4837,
+      "step": 477
+    },
+    {
+      "epoch": 0.29084271372071796,
+      "grad_norm": 1.5389190912246704,
+      "learning_rate": 4.9856863377660515e-05,
+      "loss": 0.5699,
+      "step": 478
+    },
+    {
+      "epoch": 0.29145117128080317,
+      "grad_norm": 1.4740633964538574,
+      "learning_rate": 4.985553006455556e-05,
+      "loss": 0.512,
+      "step": 479
+    },
+    {
+      "epoch": 0.2920596288408884,
+      "grad_norm": 3.1011834144592285,
+      "learning_rate": 4.985419058831517e-05,
+      "loss": 0.5975,
+      "step": 480
+    },
+    {
+      "epoch": 0.2926680864009735,
+      "grad_norm": 1.6393465995788574,
+      "learning_rate": 4.9852844949271496e-05,
+      "loss": 0.5588,
+      "step": 481
+    },
+    {
+      "epoch": 0.29327654396105873,
+      "grad_norm": 1.4237196445465088,
+      "learning_rate": 4.985149314775818e-05,
+      "loss": 0.5831,
+      "step": 482
+    },
+    {
+      "epoch": 0.2938850015211439,
+      "grad_norm": 1.3287136554718018,
+      "learning_rate": 4.985013518411044e-05,
+      "loss": 0.4769,
+      "step": 483
+    },
+    {
+      "epoch": 0.2944934590812291,
+      "grad_norm": 1.4517908096313477,
+      "learning_rate": 4.984877105866497e-05,
+      "loss": 0.5007,
+      "step": 484
+    },
+    {
+      "epoch": 0.29510191664131424,
+      "grad_norm": 1.5900462865829468,
+      "learning_rate": 4.984740077176002e-05,
+      "loss": 0.6005,
+      "step": 485
+    },
+    {
+      "epoch": 0.29571037420139945,
+      "grad_norm": 1.5251234769821167,
+      "learning_rate": 4.984602432373537e-05,
+      "loss": 0.5368,
+      "step": 486
+    },
+    {
+      "epoch": 0.29631883176148466,
+      "grad_norm": 1.551039457321167,
+      "learning_rate": 4.984464171493233e-05,
+      "loss": 0.6079,
+      "step": 487
+    },
+    {
+      "epoch": 0.2969272893215698,
+      "grad_norm": 2.1228880882263184,
+      "learning_rate": 4.984325294569372e-05,
+      "loss": 0.6416,
+      "step": 488
+    },
+    {
+      "epoch": 0.297535746881655,
+      "grad_norm": 1.4991825819015503,
+      "learning_rate": 4.98418580163639e-05,
+      "loss": 0.4824,
+      "step": 489
+    },
+    {
+      "epoch": 0.29814420444174017,
+      "grad_norm": 1.746289610862732,
+      "learning_rate": 4.9840456927288734e-05,
+      "loss": 0.5908,
+      "step": 490
+    },
+    {
+      "epoch": 0.2987526620018254,
+      "grad_norm": 1.5192077159881592,
+      "learning_rate": 4.983904967881567e-05,
+      "loss": 0.6048,
+      "step": 491
+    },
+    {
+      "epoch": 0.2993611195619106,
+      "grad_norm": 1.415953278541565,
+      "learning_rate": 4.983763627129362e-05,
+      "loss": 0.5042,
+      "step": 492
+    },
+    {
+      "epoch": 0.29996957712199573,
+      "grad_norm": 1.255936622619629,
+      "learning_rate": 4.983621670507306e-05,
+      "loss": 0.5223,
+      "step": 493
+    },
+    {
+      "epoch": 0.30057803468208094,
+      "grad_norm": 1.3377435207366943,
+      "learning_rate": 4.9834790980505985e-05,
+      "loss": 0.553,
+      "step": 494
+    },
+    {
+      "epoch": 0.3011864922421661,
+      "grad_norm": 1.7899580001831055,
+      "learning_rate": 4.983335909794591e-05,
+      "loss": 0.4885,
+      "step": 495
+    },
+    {
+      "epoch": 0.3017949498022513,
+      "grad_norm": 1.4705740213394165,
+      "learning_rate": 4.983192105774788e-05,
+      "loss": 0.5556,
+      "step": 496
+    },
+    {
+      "epoch": 0.30240340736233645,
+      "grad_norm": 1.9178767204284668,
+      "learning_rate": 4.983047686026847e-05,
+      "loss": 0.4836,
+      "step": 497
+    },
+    {
+      "epoch": 0.30301186492242166,
+      "grad_norm": 1.54046630859375,
+      "learning_rate": 4.9829026505865794e-05,
+      "loss": 0.5707,
+      "step": 498
+    },
+    {
+      "epoch": 0.30362032248250687,
+      "grad_norm": 1.558304786682129,
+      "learning_rate": 4.982756999489947e-05,
+      "loss": 0.5615,
+      "step": 499
+    },
+    {
+      "epoch": 0.304228780042592,
+      "grad_norm": 1.782384991645813,
+      "learning_rate": 4.982610732773064e-05,
+      "loss": 0.4886,
+      "step": 500
+    },
+    {
+      "epoch": 0.3048372376026772,
+      "grad_norm": 1.3223114013671875,
+      "learning_rate": 4.9824638504722005e-05,
+      "loss": 0.517,
+      "step": 501
+    },
+    {
+      "epoch": 0.3054456951627624,
+      "grad_norm": 1.3240535259246826,
+      "learning_rate": 4.982316352623776e-05,
+      "loss": 0.5153,
+      "step": 502
+    },
+    {
+      "epoch": 0.3060541527228476,
+      "grad_norm": 1.610282301902771,
+      "learning_rate": 4.982168239264364e-05,
+      "loss": 0.5231,
+      "step": 503
+    },
+    {
+      "epoch": 0.3066626102829328,
+      "grad_norm": 1.6229352951049805,
+      "learning_rate": 4.982019510430691e-05,
+      "loss": 0.613,
+      "step": 504
+    },
+    {
+      "epoch": 0.30727106784301794,
+      "grad_norm": 1.3226631879806519,
+      "learning_rate": 4.981870166159635e-05,
+      "loss": 0.5142,
+      "step": 505
+    },
+    {
+      "epoch": 0.30787952540310315,
+      "grad_norm": 1.495801568031311,
+      "learning_rate": 4.981720206488226e-05,
+      "loss": 0.5711,
+      "step": 506
+    },
+    {
+      "epoch": 0.3084879829631883,
+      "grad_norm": 1.7586286067962646,
+      "learning_rate": 4.9815696314536504e-05,
+      "loss": 0.5152,
+      "step": 507
+    },
+    {
+      "epoch": 0.3090964405232735,
+      "grad_norm": 1.5324755907058716,
+      "learning_rate": 4.981418441093243e-05,
+      "loss": 0.5501,
+      "step": 508
+    },
+    {
+      "epoch": 0.30970489808335866,
+      "grad_norm": 1.775768756866455,
+      "learning_rate": 4.981266635444492e-05,
+      "loss": 0.6562,
+      "step": 509
+    },
+    {
+      "epoch": 0.31031335564344387,
+      "grad_norm": 1.3444887399673462,
+      "learning_rate": 4.98111421454504e-05,
+      "loss": 0.5188,
+      "step": 510
+    },
+    {
+      "epoch": 0.3109218132035291,
+      "grad_norm": 1.3959381580352783,
+      "learning_rate": 4.9809611784326815e-05,
+      "loss": 0.5093,
+      "step": 511
+    },
+    {
+      "epoch": 0.3115302707636142,
+      "grad_norm": 3.0742530822753906,
+      "learning_rate": 4.9808075271453616e-05,
+      "loss": 0.5801,
+      "step": 512
+    },
+    {
+      "epoch": 0.31213872832369943,
+      "grad_norm": 1.4250247478485107,
+      "learning_rate": 4.9806532607211797e-05,
+      "loss": 0.5579,
+      "step": 513
+    },
+    {
+      "epoch": 0.3127471858837846,
+      "grad_norm": 2.064180374145508,
+      "learning_rate": 4.980498379198389e-05,
+      "loss": 0.5433,
+      "step": 514
+    },
+    {
+      "epoch": 0.3133556434438698,
+      "grad_norm": 1.5153679847717285,
+      "learning_rate": 4.980342882615392e-05,
+      "loss": 0.5676,
+      "step": 515
+    },
+    {
+      "epoch": 0.313964101003955,
+      "grad_norm": 1.5368945598602295,
+      "learning_rate": 4.9801867710107454e-05,
+      "loss": 0.5181,
+      "step": 516
+    },
+    {
+      "epoch": 0.31457255856404015,
+      "grad_norm": 1.2389271259307861,
+      "learning_rate": 4.98003004442316e-05,
+      "loss": 0.4591,
+      "step": 517
+    },
+    {
+      "epoch": 0.31518101612412536,
+      "grad_norm": 1.1532279253005981,
+      "learning_rate": 4.979872702891495e-05,
+      "loss": 0.4422,
+      "step": 518
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 1.6176403760910034,
+      "learning_rate": 4.9797147464547664e-05,
+      "loss": 0.6151,
+      "step": 519
+    },
+    {
+      "epoch": 0.3163979312442957,
+      "grad_norm": 1.5165481567382812,
+      "learning_rate": 4.9795561751521405e-05,
+      "loss": 0.5284,
+      "step": 520
+    },
+    {
+      "epoch": 0.31700638880438087,
+      "grad_norm": 1.6209813356399536,
+      "learning_rate": 4.9793969890229364e-05,
+      "loss": 0.5867,
+      "step": 521
+    },
+    {
+      "epoch": 0.3176148463644661,
+      "grad_norm": 1.3948253393173218,
+      "learning_rate": 4.9792371881066245e-05,
+      "loss": 0.5734,
+      "step": 522
+    },
+    {
+      "epoch": 0.3182233039245513,
+      "grad_norm": 1.2350413799285889,
+      "learning_rate": 4.9790767724428304e-05,
+      "loss": 0.5079,
+      "step": 523
+    },
+    {
+      "epoch": 0.31883176148463643,
+      "grad_norm": 1.3993268013000488,
+      "learning_rate": 4.978915742071329e-05,
+      "loss": 0.5713,
+      "step": 524
+    },
+    {
+      "epoch": 0.31944021904472164,
+      "grad_norm": 1.3843022584915161,
+      "learning_rate": 4.978754097032051e-05,
+      "loss": 0.5258,
+      "step": 525
+    },
+    {
+      "epoch": 0.3200486766048068,
+      "grad_norm": 1.2739397287368774,
+      "learning_rate": 4.978591837365076e-05,
+      "loss": 0.5044,
+      "step": 526
+    },
+    {
+      "epoch": 0.320657134164892,
+      "grad_norm": 2.009359359741211,
+      "learning_rate": 4.978428963110638e-05,
+      "loss": 0.495,
+      "step": 527
+    },
+    {
+      "epoch": 0.3212655917249772,
+      "grad_norm": 1.5859858989715576,
+      "learning_rate": 4.978265474309123e-05,
+      "loss": 0.5793,
+      "step": 528
+    },
+    {
+      "epoch": 0.32187404928506236,
+      "grad_norm": 1.5144940614700317,
+      "learning_rate": 4.9781013710010696e-05,
+      "loss": 0.5305,
+      "step": 529
+    },
+    {
+      "epoch": 0.32248250684514757,
+      "grad_norm": 1.5362976789474487,
+      "learning_rate": 4.977936653227169e-05,
+      "loss": 0.6301,
+      "step": 530
+    },
+    {
+      "epoch": 0.3230909644052327,
+      "grad_norm": 1.516918420791626,
+      "learning_rate": 4.9777713210282636e-05,
+      "loss": 0.581,
+      "step": 531
+    },
+    {
+      "epoch": 0.3236994219653179,
+      "grad_norm": 1.3938144445419312,
+      "learning_rate": 4.977605374445349e-05,
+      "loss": 0.6385,
+      "step": 532
+    },
+    {
+      "epoch": 0.3243078795254031,
+      "grad_norm": 1.1043601036071777,
+      "learning_rate": 4.977438813519574e-05,
+      "loss": 0.4772,
+      "step": 533
+    },
+    {
+      "epoch": 0.3249163370854883,
+      "grad_norm": 1.2884318828582764,
+      "learning_rate": 4.977271638292237e-05,
+      "loss": 0.6186,
+      "step": 534
+    },
+    {
+      "epoch": 0.3255247946455735,
+      "grad_norm": 1.278027057647705,
+      "learning_rate": 4.9771038488047915e-05,
+      "loss": 0.5577,
+      "step": 535
+    },
+    {
+      "epoch": 0.32613325220565864,
+      "grad_norm": 1.1427042484283447,
+      "learning_rate": 4.976935445098843e-05,
+      "loss": 0.4842,
+      "step": 536
+    },
+    {
+      "epoch": 0.32674170976574385,
+      "grad_norm": 1.1449129581451416,
+      "learning_rate": 4.9767664272161474e-05,
+      "loss": 0.4771,
+      "step": 537
+    },
+    {
+      "epoch": 0.327350167325829,
+      "grad_norm": 1.4872303009033203,
+      "learning_rate": 4.976596795198615e-05,
+      "loss": 0.5961,
+      "step": 538
+    },
+    {
+      "epoch": 0.3279586248859142,
+      "grad_norm": 1.3000589609146118,
+      "learning_rate": 4.976426549088307e-05,
+      "loss": 0.5076,
+      "step": 539
+    },
+    {
+      "epoch": 0.3285670824459994,
+      "grad_norm": 1.8529880046844482,
+      "learning_rate": 4.976255688927436e-05,
+      "loss": 0.5203,
+      "step": 540
+    },
+    {
+      "epoch": 0.32917554000608457,
+      "grad_norm": 2.4158706665039062,
+      "learning_rate": 4.976084214758371e-05,
+      "loss": 0.5025,
+      "step": 541
+    },
+    {
+      "epoch": 0.3297839975661698,
+      "grad_norm": 1.3759514093399048,
+      "learning_rate": 4.9759121266236286e-05,
+      "loss": 0.5908,
+      "step": 542
+    },
+    {
+      "epoch": 0.3303924551262549,
+      "grad_norm": 1.3580955266952515,
+      "learning_rate": 4.97573942456588e-05,
+      "loss": 0.4575,
+      "step": 543
+    },
+    {
+      "epoch": 0.33100091268634013,
+      "grad_norm": 3.9890565872192383,
+      "learning_rate": 4.975566108627948e-05,
+      "loss": 0.4681,
+      "step": 544
+    },
+    {
+      "epoch": 0.33160937024642534,
+      "grad_norm": 1.6312041282653809,
+      "learning_rate": 4.975392178852808e-05,
+      "loss": 0.5273,
+      "step": 545
+    },
+    {
+      "epoch": 0.3322178278065105,
+      "grad_norm": 2.074263572692871,
+      "learning_rate": 4.9752176352835866e-05,
+      "loss": 0.5376,
+      "step": 546
+    },
+    {
+      "epoch": 0.3328262853665957,
+      "grad_norm": 2.1829655170440674,
+      "learning_rate": 4.975042477963564e-05,
+      "loss": 0.5093,
+      "step": 547
+    },
+    {
+      "epoch": 0.33343474292668085,
+      "grad_norm": 1.5592143535614014,
+      "learning_rate": 4.9748667069361715e-05,
+      "loss": 0.5319,
+      "step": 548
+    },
+    {
+      "epoch": 0.33404320048676606,
+      "grad_norm": 1.8820182085037231,
+      "learning_rate": 4.974690322244994e-05,
+      "loss": 0.5467,
+      "step": 549
+    },
+    {
+      "epoch": 0.3346516580468512,
+      "grad_norm": 1.3456025123596191,
+      "learning_rate": 4.974513323933766e-05,
+      "loss": 0.5365,
+      "step": 550
+    },
+    {
+      "epoch": 0.3352601156069364,
+      "grad_norm": 1.2476035356521606,
+      "learning_rate": 4.974335712046376e-05,
+      "loss": 0.4927,
+      "step": 551
+    },
+    {
+      "epoch": 0.3358685731670216,
+      "grad_norm": 1.5756597518920898,
+      "learning_rate": 4.974157486626866e-05,
+      "loss": 0.5539,
+      "step": 552
+    },
+    {
+      "epoch": 0.3364770307271068,
+      "grad_norm": 1.4776535034179688,
+      "learning_rate": 4.973978647719426e-05,
+      "loss": 0.5584,
+      "step": 553
+    },
+    {
+      "epoch": 0.337085488287192,
+      "grad_norm": 2.738008975982666,
+      "learning_rate": 4.9737991953684024e-05,
+      "loss": 0.5617,
+      "step": 554
+    },
+    {
+      "epoch": 0.33769394584727713,
+      "grad_norm": 1.5317890644073486,
+      "learning_rate": 4.973619129618292e-05,
+      "loss": 0.5562,
+      "step": 555
+    },
+    {
+      "epoch": 0.33830240340736234,
+      "grad_norm": 1.49234938621521,
+      "learning_rate": 4.973438450513743e-05,
+      "loss": 0.5217,
+      "step": 556
+    },
+    {
+      "epoch": 0.33891086096744755,
+      "grad_norm": 1.301456093788147,
+      "learning_rate": 4.973257158099556e-05,
+      "loss": 0.4828,
+      "step": 557
+    },
+    {
+      "epoch": 0.3395193185275327,
+      "grad_norm": 2.2302095890045166,
+      "learning_rate": 4.9730752524206835e-05,
+      "loss": 0.5493,
+      "step": 558
+    },
+    {
+      "epoch": 0.3401277760876179,
+      "grad_norm": 1.5827584266662598,
+      "learning_rate": 4.972892733522232e-05,
+      "loss": 0.5375,
+      "step": 559
+    },
+    {
+      "epoch": 0.34073623364770306,
+      "grad_norm": 1.3773258924484253,
+      "learning_rate": 4.972709601449458e-05,
+      "loss": 0.5809,
+      "step": 560
+    },
+    {
+      "epoch": 0.34134469120778826,
+      "grad_norm": 1.2351981401443481,
+      "learning_rate": 4.972525856247769e-05,
+      "loss": 0.4553,
+      "step": 561
+    },
+    {
+      "epoch": 0.3419531487678734,
+      "grad_norm": 1.5786877870559692,
+      "learning_rate": 4.972341497962729e-05,
+      "loss": 0.5861,
+      "step": 562
+    },
+    {
+      "epoch": 0.3425616063279586,
+      "grad_norm": 1.3937610387802124,
+      "learning_rate": 4.97215652664005e-05,
+      "loss": 0.5209,
+      "step": 563
+    },
+    {
+      "epoch": 0.34317006388804383,
+      "grad_norm": 1.723417043685913,
+      "learning_rate": 4.971970942325597e-05,
+      "loss": 0.6222,
+      "step": 564
+    },
+    {
+      "epoch": 0.343778521448129,
+      "grad_norm": 1.3207844495773315,
+      "learning_rate": 4.971784745065386e-05,
+      "loss": 0.547,
+      "step": 565
+    },
+    {
+      "epoch": 0.3443869790082142,
+      "grad_norm": 1.4648035764694214,
+      "learning_rate": 4.971597934905587e-05,
+      "loss": 0.4984,
+      "step": 566
+    },
+    {
+      "epoch": 0.34499543656829934,
+      "grad_norm": 1.2435007095336914,
+      "learning_rate": 4.971410511892523e-05,
+      "loss": 0.5244,
+      "step": 567
+    },
+    {
+      "epoch": 0.34560389412838455,
+      "grad_norm": 1.309339165687561,
+      "learning_rate": 4.971222476072665e-05,
+      "loss": 0.5165,
+      "step": 568
+    },
+    {
+      "epoch": 0.34621235168846975,
+      "grad_norm": 1.554057240486145,
+      "learning_rate": 4.9710338274926384e-05,
+      "loss": 0.6061,
+      "step": 569
+    },
+    {
+      "epoch": 0.3468208092485549,
+      "grad_norm": 1.3458071947097778,
+      "learning_rate": 4.97084456619922e-05,
+      "loss": 0.5489,
+      "step": 570
+    },
+    {
+      "epoch": 0.3474292668086401,
+      "grad_norm": 1.2686628103256226,
+      "learning_rate": 4.9706546922393396e-05,
+      "loss": 0.4569,
+      "step": 571
+    },
+    {
+      "epoch": 0.34803772436872527,
+      "grad_norm": 2.3073008060455322,
+      "learning_rate": 4.970464205660077e-05,
+      "loss": 0.4969,
+      "step": 572
+    },
+    {
+      "epoch": 0.3486461819288105,
+      "grad_norm": 1.3341611623764038,
+      "learning_rate": 4.970273106508666e-05,
+      "loss": 0.501,
+      "step": 573
+    },
+    {
+      "epoch": 0.3492546394888956,
+      "grad_norm": 1.5147910118103027,
+      "learning_rate": 4.97008139483249e-05,
+      "loss": 0.5889,
+      "step": 574
+    },
+    {
+      "epoch": 0.34986309704898083,
+      "grad_norm": 2.7254374027252197,
+      "learning_rate": 4.9698890706790866e-05,
+      "loss": 0.598,
+      "step": 575
+    },
+    {
+      "epoch": 0.35047155460906604,
+      "grad_norm": 1.4066073894500732,
+      "learning_rate": 4.969696134096143e-05,
+      "loss": 0.5344,
+      "step": 576
+    },
+    {
+      "epoch": 0.3510800121691512,
+      "grad_norm": 1.7791637182235718,
+      "learning_rate": 4.969502585131502e-05,
+      "loss": 0.4642,
+      "step": 577
+    },
+    {
+      "epoch": 0.3516884697292364,
+      "grad_norm": 1.9573460817337036,
+      "learning_rate": 4.969308423833152e-05,
+      "loss": 0.5322,
+      "step": 578
+    },
+    {
+      "epoch": 0.35229692728932155,
+      "grad_norm": 1.4854366779327393,
+      "learning_rate": 4.96911365024924e-05,
+      "loss": 0.486,
+      "step": 579
+    },
+    {
+      "epoch": 0.35290538484940676,
+      "grad_norm": 1.8296064138412476,
+      "learning_rate": 4.968918264428059e-05,
+      "loss": 0.5649,
+      "step": 580
+    },
+    {
+      "epoch": 0.35351384240949196,
+      "grad_norm": 1.3343514204025269,
+      "learning_rate": 4.9687222664180585e-05,
+      "loss": 0.524,
+      "step": 581
+    },
+    {
+      "epoch": 0.3541222999695771,
+      "grad_norm": 1.521429419517517,
+      "learning_rate": 4.968525656267838e-05,
+      "loss": 0.5018,
+      "step": 582
+    },
+    {
+      "epoch": 0.3547307575296623,
+      "grad_norm": 2.845484972000122,
+      "learning_rate": 4.968328434026148e-05,
+      "loss": 0.6363,
+      "step": 583
+    },
+    {
+      "epoch": 0.3553392150897475,
+      "grad_norm": 1.4245471954345703,
+      "learning_rate": 4.9681305997418906e-05,
+      "loss": 0.5181,
+      "step": 584
+    },
+    {
+      "epoch": 0.3559476726498327,
+      "grad_norm": 1.27480149269104,
+      "learning_rate": 4.9679321534641214e-05,
+      "loss": 0.5343,
+      "step": 585
+    },
+    {
+      "epoch": 0.35655613020991783,
+      "grad_norm": 1.2429792881011963,
+      "learning_rate": 4.967733095242047e-05,
+      "loss": 0.4989,
+      "step": 586
+    },
+    {
+      "epoch": 0.35716458777000304,
+      "grad_norm": 1.9116991758346558,
+      "learning_rate": 4.967533425125025e-05,
+      "loss": 0.611,
+      "step": 587
+    },
+    {
+      "epoch": 0.35777304533008825,
+      "grad_norm": 1.498067855834961,
+      "learning_rate": 4.967333143162565e-05,
+      "loss": 0.5896,
+      "step": 588
+    },
+    {
+      "epoch": 0.3583815028901734,
+      "grad_norm": 1.4009770154953003,
+      "learning_rate": 4.967132249404329e-05,
+      "loss": 0.5002,
+      "step": 589
+    },
+    {
+      "epoch": 0.3589899604502586,
+      "grad_norm": 1.346002459526062,
+      "learning_rate": 4.966930743900131e-05,
+      "loss": 0.5411,
+      "step": 590
+    },
+    {
+      "epoch": 0.35959841801034376,
+      "grad_norm": 1.3442580699920654,
+      "learning_rate": 4.9667286266999354e-05,
+      "loss": 0.5158,
+      "step": 591
+    },
+    {
+      "epoch": 0.36020687557042896,
+      "grad_norm": 1.6261683702468872,
+      "learning_rate": 4.966525897853858e-05,
+      "loss": 0.6546,
+      "step": 592
+    },
+    {
+      "epoch": 0.36081533313051417,
+      "grad_norm": 1.5490068197250366,
+      "learning_rate": 4.966322557412168e-05,
+      "loss": 0.5536,
+      "step": 593
+    },
+    {
+      "epoch": 0.3614237906905993,
+      "grad_norm": 1.3161338567733765,
+      "learning_rate": 4.966118605425285e-05,
+      "loss": 0.5315,
+      "step": 594
+    },
+    {
+      "epoch": 0.36203224825068453,
+      "grad_norm": 1.6688951253890991,
+      "learning_rate": 4.965914041943781e-05,
+      "loss": 0.5392,
+      "step": 595
+    },
+    {
+      "epoch": 0.3626407058107697,
+      "grad_norm": 1.894654393196106,
+      "learning_rate": 4.9657088670183794e-05,
+      "loss": 0.5337,
+      "step": 596
+    },
+    {
+      "epoch": 0.3632491633708549,
+      "grad_norm": 1.5747034549713135,
+      "learning_rate": 4.9655030806999534e-05,
+      "loss": 0.5689,
+      "step": 597
+    },
+    {
+      "epoch": 0.36385762093094004,
+      "grad_norm": 1.315900206565857,
+      "learning_rate": 4.965296683039532e-05,
+      "loss": 0.5897,
+      "step": 598
+    },
+    {
+      "epoch": 0.36446607849102525,
+      "grad_norm": 1.3753433227539062,
+      "learning_rate": 4.9650896740882905e-05,
+      "loss": 0.5475,
+      "step": 599
+    },
+    {
+      "epoch": 0.36507453605111045,
+      "grad_norm": 1.546858787536621,
+      "learning_rate": 4.96488205389756e-05,
+      "loss": 0.5734,
+      "step": 600
+    },
+    {
+      "epoch": 0.3656829936111956,
+      "grad_norm": 1.2179769277572632,
+      "learning_rate": 4.9646738225188226e-05,
+      "loss": 0.4914,
+      "step": 601
+    },
+    {
+      "epoch": 0.3662914511712808,
+      "grad_norm": 1.370540738105774,
+      "learning_rate": 4.964464980003709e-05,
+      "loss": 0.4964,
+      "step": 602
+    },
+    {
+      "epoch": 0.36689990873136596,
+      "grad_norm": 1.3386831283569336,
+      "learning_rate": 4.9642555264040046e-05,
+      "loss": 0.5037,
+      "step": 603
+    },
+    {
+      "epoch": 0.36750836629145117,
+      "grad_norm": 1.3772310018539429,
+      "learning_rate": 4.964045461771645e-05,
+      "loss": 0.4962,
+      "step": 604
+    },
+    {
+      "epoch": 0.3681168238515364,
+      "grad_norm": 1.5445538759231567,
+      "learning_rate": 4.963834786158717e-05,
+      "loss": 0.4728,
+      "step": 605
+    },
+    {
+      "epoch": 0.36872528141162153,
+      "grad_norm": 1.419505000114441,
+      "learning_rate": 4.96362349961746e-05,
+      "loss": 0.502,
+      "step": 606
+    },
+    {
+      "epoch": 0.36933373897170674,
+      "grad_norm": 1.3720256090164185,
+      "learning_rate": 4.963411602200264e-05,
+      "loss": 0.5714,
+      "step": 607
+    },
+    {
+      "epoch": 0.3699421965317919,
+      "grad_norm": 1.3194488286972046,
+      "learning_rate": 4.963199093959671e-05,
+      "loss": 0.5201,
+      "step": 608
+    },
+    {
+      "epoch": 0.3705506540918771,
+      "grad_norm": 1.3677685260772705,
+      "learning_rate": 4.962985974948373e-05,
+      "loss": 0.5827,
+      "step": 609
+    },
+    {
+      "epoch": 0.37115911165196225,
+      "grad_norm": 1.1780776977539062,
+      "learning_rate": 4.9627722452192164e-05,
+      "loss": 0.4458,
+      "step": 610
+    },
+    {
+      "epoch": 0.37176756921204746,
+      "grad_norm": 1.2468417882919312,
+      "learning_rate": 4.962557904825196e-05,
+      "loss": 0.5893,
+      "step": 611
+    },
+    {
+      "epoch": 0.37237602677213266,
+      "grad_norm": 1.4214433431625366,
+      "learning_rate": 4.962342953819459e-05,
+      "loss": 0.5349,
+      "step": 612
+    },
+    {
+      "epoch": 0.3729844843322178,
+      "grad_norm": 1.3830993175506592,
+      "learning_rate": 4.9621273922553055e-05,
+      "loss": 0.4956,
+      "step": 613
+    },
+    {
+      "epoch": 0.373592941892303,
+      "grad_norm": 1.322393774986267,
+      "learning_rate": 4.961911220186186e-05,
+      "loss": 0.528,
+      "step": 614
+    },
+    {
+      "epoch": 0.3742013994523882,
+      "grad_norm": 1.3154321908950806,
+      "learning_rate": 4.961694437665701e-05,
+      "loss": 0.4628,
+      "step": 615
+    },
+    {
+      "epoch": 0.3748098570124734,
+      "grad_norm": 1.4259490966796875,
+      "learning_rate": 4.9614770447476037e-05,
+      "loss": 0.5198,
+      "step": 616
+    },
+    {
+      "epoch": 0.3754183145725586,
+      "grad_norm": 1.643979549407959,
+      "learning_rate": 4.961259041485799e-05,
+      "loss": 0.4984,
+      "step": 617
+    },
+    {
+      "epoch": 0.37602677213264374,
+      "grad_norm": 1.36431884765625,
+      "learning_rate": 4.9610404279343415e-05,
+      "loss": 0.5148,
+      "step": 618
+    },
+    {
+      "epoch": 0.37663522969272895,
+      "grad_norm": 1.2996978759765625,
+      "learning_rate": 4.96082120414744e-05,
+      "loss": 0.5277,
+      "step": 619
+    },
+    {
+      "epoch": 0.3772436872528141,
+      "grad_norm": 1.7564353942871094,
+      "learning_rate": 4.960601370179452e-05,
+      "loss": 0.5884,
+      "step": 620
+    },
+    {
+      "epoch": 0.3778521448128993,
+      "grad_norm": 1.3980865478515625,
+      "learning_rate": 4.9603809260848864e-05,
+      "loss": 0.4982,
+      "step": 621
+    },
+    {
+      "epoch": 0.3784606023729845,
+      "grad_norm": 1.2997080087661743,
+      "learning_rate": 4.960159871918405e-05,
+      "loss": 0.5634,
+      "step": 622
+    },
+    {
+      "epoch": 0.37906905993306966,
+      "grad_norm": 1.2841782569885254,
+      "learning_rate": 4.9599382077348205e-05,
+      "loss": 0.4826,
+      "step": 623
+    },
+    {
+      "epoch": 0.37967751749315487,
+      "grad_norm": 1.347280502319336,
+      "learning_rate": 4.959715933589095e-05,
+      "loss": 0.5076,
+      "step": 624
+    },
+    {
+      "epoch": 0.38028597505324,
+      "grad_norm": 1.6214102506637573,
+      "learning_rate": 4.9594930495363445e-05,
+      "loss": 0.5878,
+      "step": 625
+    },
+    {
+      "epoch": 0.38089443261332523,
+      "grad_norm": 1.3995033502578735,
+      "learning_rate": 4.959269555631835e-05,
+      "loss": 0.476,
+      "step": 626
+    },
+    {
+      "epoch": 0.3815028901734104,
+      "grad_norm": 1.3252878189086914,
+      "learning_rate": 4.959045451930982e-05,
+      "loss": 0.5469,
+      "step": 627
+    },
+    {
+      "epoch": 0.3821113477334956,
+      "grad_norm": 1.2933142185211182,
+      "learning_rate": 4.958820738489355e-05,
+      "loss": 0.535,
+      "step": 628
+    },
+    {
+      "epoch": 0.3827198052935808,
+      "grad_norm": 1.3921632766723633,
+      "learning_rate": 4.958595415362675e-05,
+      "loss": 0.5282,
+      "step": 629
+    },
+    {
+      "epoch": 0.38332826285366595,
+      "grad_norm": 1.2744975090026855,
+      "learning_rate": 4.95836948260681e-05,
+      "loss": 0.5035,
+      "step": 630
+    },
+    {
+      "epoch": 0.38393672041375115,
+      "grad_norm": 2.0861146450042725,
+      "learning_rate": 4.9581429402777826e-05,
+      "loss": 0.5104,
+      "step": 631
+    },
+    {
+      "epoch": 0.3845451779738363,
+      "grad_norm": 1.7195478677749634,
+      "learning_rate": 4.957915788431768e-05,
+      "loss": 0.5618,
+      "step": 632
+    },
+    {
+      "epoch": 0.3851536355339215,
+      "grad_norm": 1.2929902076721191,
+      "learning_rate": 4.957688027125088e-05,
+      "loss": 0.4858,
+      "step": 633
+    },
+    {
+      "epoch": 0.3857620930940067,
+      "grad_norm": 1.324310064315796,
+      "learning_rate": 4.957459656414219e-05,
+      "loss": 0.576,
+      "step": 634
+    },
+    {
+      "epoch": 0.38637055065409187,
+      "grad_norm": 1.3465945720672607,
+      "learning_rate": 4.957230676355787e-05,
+      "loss": 0.5138,
+      "step": 635
+    },
+    {
+      "epoch": 0.3869790082141771,
+      "grad_norm": 1.2938052415847778,
+      "learning_rate": 4.95700108700657e-05,
+      "loss": 0.5101,
+      "step": 636
+    },
+    {
+      "epoch": 0.38758746577426223,
+      "grad_norm": 1.6239326000213623,
+      "learning_rate": 4.956770888423495e-05,
+      "loss": 0.5932,
+      "step": 637
+    },
+    {
+      "epoch": 0.38819592333434744,
+      "grad_norm": 1.3024382591247559,
+      "learning_rate": 4.9565400806636447e-05,
+      "loss": 0.5088,
+      "step": 638
+    },
+    {
+      "epoch": 0.3888043808944326,
+      "grad_norm": 1.557921051979065,
+      "learning_rate": 4.956308663784247e-05,
+      "loss": 0.5358,
+      "step": 639
+    },
+    {
+      "epoch": 0.3894128384545178,
+      "grad_norm": 1.2118357419967651,
+      "learning_rate": 4.956076637842685e-05,
+      "loss": 0.5089,
+      "step": 640
+    },
+    {
+      "epoch": 0.390021296014603,
+      "grad_norm": 1.4801642894744873,
+      "learning_rate": 4.9558440028964914e-05,
+      "loss": 0.4758,
+      "step": 641
+    },
+    {
+      "epoch": 0.39062975357468815,
+      "grad_norm": 1.3965163230895996,
+      "learning_rate": 4.95561075900335e-05,
+      "loss": 0.5498,
+      "step": 642
+    },
+    {
+      "epoch": 0.39123821113477336,
+      "grad_norm": 1.5353859663009644,
+      "learning_rate": 4.955376906221094e-05,
+      "loss": 0.4723,
+      "step": 643
+    },
+    {
+      "epoch": 0.3918466686948585,
+      "grad_norm": 1.248234748840332,
+      "learning_rate": 4.955142444607711e-05,
+      "loss": 0.5461,
+      "step": 644
+    },
+    {
+      "epoch": 0.3924551262549437,
+      "grad_norm": 1.5625355243682861,
+      "learning_rate": 4.9549073742213375e-05,
+      "loss": 0.5561,
+      "step": 645
+    },
+    {
+      "epoch": 0.3930635838150289,
+      "grad_norm": 1.327852487564087,
+      "learning_rate": 4.9546716951202606e-05,
+      "loss": 0.4967,
+      "step": 646
+    },
+    {
+      "epoch": 0.3936720413751141,
+      "grad_norm": 1.3109062910079956,
+      "learning_rate": 4.9544354073629186e-05,
+      "loss": 0.5108,
+      "step": 647
+    },
+    {
+      "epoch": 0.3942804989351993,
+      "grad_norm": 1.6410901546478271,
+      "learning_rate": 4.954198511007902e-05,
+      "loss": 0.5998,
+      "step": 648
+    },
+    {
+      "epoch": 0.39488895649528444,
+      "grad_norm": 1.4157142639160156,
+      "learning_rate": 4.9539610061139505e-05,
+      "loss": 0.54,
+      "step": 649
+    },
+    {
+      "epoch": 0.39549741405536964,
+      "grad_norm": 1.2379719018936157,
+      "learning_rate": 4.953722892739956e-05,
+      "loss": 0.4716,
+      "step": 650
+    },
+    {
+      "epoch": 0.3961058716154548,
+      "grad_norm": 1.3392927646636963,
+      "learning_rate": 4.95348417094496e-05,
+      "loss": 0.4665,
+      "step": 651
+    },
+    {
+      "epoch": 0.39671432917554,
+      "grad_norm": 1.3489748239517212,
+      "learning_rate": 4.953244840788156e-05,
+      "loss": 0.5533,
+      "step": 652
+    },
+    {
+      "epoch": 0.3973227867356252,
+      "grad_norm": 1.4216614961624146,
+      "learning_rate": 4.953004902328887e-05,
+      "loss": 0.5107,
+      "step": 653
+    },
+    {
+      "epoch": 0.39793124429571036,
+      "grad_norm": 1.2721261978149414,
+      "learning_rate": 4.9527643556266493e-05,
+      "loss": 0.5454,
+      "step": 654
+    },
+    {
+      "epoch": 0.39853970185579557,
+      "grad_norm": 1.292842984199524,
+      "learning_rate": 4.952523200741088e-05,
+      "loss": 0.4989,
+      "step": 655
+    },
+    {
+      "epoch": 0.3991481594158807,
+      "grad_norm": 1.2680416107177734,
+      "learning_rate": 4.952281437731998e-05,
+      "loss": 0.4913,
+      "step": 656
+    },
+    {
+      "epoch": 0.39975661697596593,
+      "grad_norm": 1.3643546104431152,
+      "learning_rate": 4.9520390666593286e-05,
+      "loss": 0.5144,
+      "step": 657
+    },
+    {
+      "epoch": 0.40036507453605114,
+      "grad_norm": 1.310077428817749,
+      "learning_rate": 4.951796087583176e-05,
+      "loss": 0.5468,
+      "step": 658
+    },
+    {
+      "epoch": 0.4009735320961363,
+      "grad_norm": 1.3737618923187256,
+      "learning_rate": 4.95155250056379e-05,
+      "loss": 0.5958,
+      "step": 659
+    },
+    {
+      "epoch": 0.4015819896562215,
+      "grad_norm": 1.2556304931640625,
+      "learning_rate": 4.9513083056615695e-05,
+      "loss": 0.5171,
+      "step": 660
+    },
+    {
+      "epoch": 0.40219044721630665,
+      "grad_norm": 1.2657099962234497,
+      "learning_rate": 4.9510635029370646e-05,
+      "loss": 0.5154,
+      "step": 661
+    },
+    {
+      "epoch": 0.40279890477639185,
+      "grad_norm": 1.3634426593780518,
+      "learning_rate": 4.9508180924509763e-05,
+      "loss": 0.4383,
+      "step": 662
+    },
+    {
+      "epoch": 0.403407362336477,
+      "grad_norm": 1.582513689994812,
+      "learning_rate": 4.950572074264156e-05,
+      "loss": 0.4719,
+      "step": 663
+    },
+    {
+      "epoch": 0.4040158198965622,
+      "grad_norm": 2.247115135192871,
+      "learning_rate": 4.9503254484376074e-05,
+      "loss": 0.5479,
+      "step": 664
+    },
+    {
+      "epoch": 0.4046242774566474,
+      "grad_norm": 1.3867719173431396,
+      "learning_rate": 4.950078215032481e-05,
+      "loss": 0.5422,
+      "step": 665
+    },
+    {
+      "epoch": 0.40523273501673257,
+      "grad_norm": 1.2318755388259888,
+      "learning_rate": 4.949830374110081e-05,
+      "loss": 0.5662,
+      "step": 666
+    },
+    {
+      "epoch": 0.4058411925768178,
+      "grad_norm": 1.5015090703964233,
+      "learning_rate": 4.9495819257318635e-05,
+      "loss": 0.5895,
+      "step": 667
+    },
+    {
+      "epoch": 0.40644965013690293,
+      "grad_norm": 1.2691212892532349,
+      "learning_rate": 4.949332869959432e-05,
+      "loss": 0.4953,
+      "step": 668
+    },
+    {
+      "epoch": 0.40705810769698814,
+      "grad_norm": 1.7610334157943726,
+      "learning_rate": 4.9490832068545414e-05,
+      "loss": 0.5327,
+      "step": 669
+    },
+    {
+      "epoch": 0.40766656525707334,
+      "grad_norm": 1.3113877773284912,
+      "learning_rate": 4.9488329364790986e-05,
+      "loss": 0.5369,
+      "step": 670
+    },
+    {
+      "epoch": 0.4082750228171585,
+      "grad_norm": 1.3471949100494385,
+      "learning_rate": 4.948582058895159e-05,
+      "loss": 0.4791,
+      "step": 671
+    },
+    {
+      "epoch": 0.4088834803772437,
+      "grad_norm": 1.1278780698776245,
+      "learning_rate": 4.9483305741649324e-05,
+      "loss": 0.4704,
+      "step": 672
+    },
+    {
+      "epoch": 0.40949193793732885,
+      "grad_norm": 1.2446277141571045,
+      "learning_rate": 4.948078482350774e-05,
+      "loss": 0.4728,
+      "step": 673
+    },
+    {
+      "epoch": 0.41010039549741406,
+      "grad_norm": 1.507495641708374,
+      "learning_rate": 4.947825783515193e-05,
+      "loss": 0.6533,
+      "step": 674
+    },
+    {
+      "epoch": 0.4107088530574992,
+      "grad_norm": 1.4783514738082886,
+      "learning_rate": 4.9475724777208474e-05,
+      "loss": 0.5185,
+      "step": 675
+    },
+    {
+      "epoch": 0.4113173106175844,
+      "grad_norm": 1.186607003211975,
+      "learning_rate": 4.947318565030548e-05,
+      "loss": 0.4684,
+      "step": 676
+    },
+    {
+      "epoch": 0.4119257681776696,
+      "grad_norm": 1.4391772747039795,
+      "learning_rate": 4.947064045507253e-05,
+      "loss": 0.5561,
+      "step": 677
+    },
+    {
+      "epoch": 0.4125342257377548,
+      "grad_norm": 1.3930890560150146,
+      "learning_rate": 4.946808919214074e-05,
+      "loss": 0.5353,
+      "step": 678
+    },
+    {
+      "epoch": 0.41314268329784,
+      "grad_norm": 1.2394658327102661,
+      "learning_rate": 4.946553186214271e-05,
+      "loss": 0.4428,
+      "step": 679
+    },
+    {
+      "epoch": 0.41375114085792514,
+      "grad_norm": 1.2956104278564453,
+      "learning_rate": 4.9462968465712555e-05,
+      "loss": 0.4658,
+      "step": 680
+    },
+    {
+      "epoch": 0.41435959841801034,
+      "grad_norm": 2.5605669021606445,
+      "learning_rate": 4.946039900348588e-05,
+      "loss": 0.5784,
+      "step": 681
+    },
+    {
+      "epoch": 0.41496805597809555,
+      "grad_norm": 1.3224326372146606,
+      "learning_rate": 4.945782347609982e-05,
+      "loss": 0.4152,
+      "step": 682
+    },
+    {
+      "epoch": 0.4155765135381807,
+      "grad_norm": 1.4174445867538452,
+      "learning_rate": 4.945524188419298e-05,
+      "loss": 0.5388,
+      "step": 683
+    },
+    {
+      "epoch": 0.4161849710982659,
+      "grad_norm": 1.2914178371429443,
+      "learning_rate": 4.9452654228405506e-05,
+      "loss": 0.4773,
+      "step": 684
+    },
+    {
+      "epoch": 0.41679342865835106,
+      "grad_norm": 1.3811676502227783,
+      "learning_rate": 4.945006050937902e-05,
+      "loss": 0.551,
+      "step": 685
+    },
+    {
+      "epoch": 0.41740188621843627,
+      "grad_norm": 1.2903940677642822,
+      "learning_rate": 4.944746072775665e-05,
+      "loss": 0.535,
+      "step": 686
+    },
+    {
+      "epoch": 0.4180103437785214,
+      "grad_norm": 1.8577662706375122,
+      "learning_rate": 4.9444854884183046e-05,
+      "loss": 0.533,
+      "step": 687
+    },
+    {
+      "epoch": 0.4186188013386066,
+      "grad_norm": 1.3438717126846313,
+      "learning_rate": 4.944224297930434e-05,
+      "loss": 0.5539,
+      "step": 688
+    },
+    {
+      "epoch": 0.41922725889869183,
+      "grad_norm": 1.184409260749817,
+      "learning_rate": 4.943962501376818e-05,
+      "loss": 0.4907,
+      "step": 689
+    },
+    {
+      "epoch": 0.419835716458777,
+      "grad_norm": 1.1499260663986206,
+      "learning_rate": 4.9437000988223705e-05,
+      "loss": 0.4658,
+      "step": 690
+    },
+    {
+      "epoch": 0.4204441740188622,
+      "grad_norm": 1.16009521484375,
+      "learning_rate": 4.9434370903321566e-05,
+      "loss": 0.5088,
+      "step": 691
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 1.2988529205322266,
+      "learning_rate": 4.943173475971393e-05,
+      "loss": 0.5152,
+      "step": 692
+    },
+    {
+      "epoch": 0.42166108913903255,
+      "grad_norm": 1.246013879776001,
+      "learning_rate": 4.942909255805443e-05,
+      "loss": 0.4863,
+      "step": 693
+    },
+    {
+      "epoch": 0.42226954669911776,
+      "grad_norm": 1.4207082986831665,
+      "learning_rate": 4.942644429899824e-05,
+      "loss": 0.5582,
+      "step": 694
+    },
+    {
+      "epoch": 0.4228780042592029,
+      "grad_norm": 1.1736639738082886,
+      "learning_rate": 4.9423789983201994e-05,
+      "loss": 0.4872,
+      "step": 695
+    },
+    {
+      "epoch": 0.4234864618192881,
+      "grad_norm": 1.3265244960784912,
+      "learning_rate": 4.942112961132388e-05,
+      "loss": 0.4979,
+      "step": 696
+    },
+    {
+      "epoch": 0.42409491937937327,
+      "grad_norm": 1.4081310033798218,
+      "learning_rate": 4.941846318402353e-05,
+      "loss": 0.5683,
+      "step": 697
+    },
+    {
+      "epoch": 0.4247033769394585,
+      "grad_norm": 1.3100318908691406,
+      "learning_rate": 4.941579070196214e-05,
+      "loss": 0.5316,
+      "step": 698
+    },
+    {
+      "epoch": 0.4253118344995437,
+      "grad_norm": 1.369620442390442,
+      "learning_rate": 4.9413112165802345e-05,
+      "loss": 0.5076,
+      "step": 699
+    },
+    {
+      "epoch": 0.42592029205962884,
+      "grad_norm": 1.268389105796814,
+      "learning_rate": 4.9410427576208316e-05,
+      "loss": 0.4792,
+      "step": 700
+    },
+    {
+      "epoch": 0.42652874961971404,
+      "grad_norm": 1.2050520181655884,
+      "learning_rate": 4.940773693384574e-05,
+      "loss": 0.4954,
+      "step": 701
+    },
+    {
+      "epoch": 0.4271372071797992,
+      "grad_norm": 1.5060069561004639,
+      "learning_rate": 4.940504023938176e-05,
+      "loss": 0.5535,
+      "step": 702
+    },
+    {
+      "epoch": 0.4277456647398844,
+      "grad_norm": 1.380094289779663,
+      "learning_rate": 4.940233749348505e-05,
+      "loss": 0.48,
+      "step": 703
+    },
+    {
+      "epoch": 0.42835412229996955,
+      "grad_norm": 1.2237451076507568,
+      "learning_rate": 4.9399628696825786e-05,
+      "loss": 0.4835,
+      "step": 704
+    },
+    {
+      "epoch": 0.42896257986005476,
+      "grad_norm": 1.639406442642212,
+      "learning_rate": 4.9396913850075636e-05,
+      "loss": 0.5267,
+      "step": 705
+    },
+    {
+      "epoch": 0.42957103742013997,
+      "grad_norm": 1.3455315828323364,
+      "learning_rate": 4.9394192953907757e-05,
+      "loss": 0.5586,
+      "step": 706
+    },
+    {
+      "epoch": 0.4301794949802251,
+      "grad_norm": 1.4164340496063232,
+      "learning_rate": 4.939146600899683e-05,
+      "loss": 0.4524,
+      "step": 707
+    },
+    {
+      "epoch": 0.4307879525403103,
+      "grad_norm": 1.3369954824447632,
+      "learning_rate": 4.938873301601902e-05,
+      "loss": 0.5494,
+      "step": 708
+    },
+    {
+      "epoch": 0.4313964101003955,
+      "grad_norm": 1.3146734237670898,
+      "learning_rate": 4.938599397565199e-05,
+      "loss": 0.4943,
+      "step": 709
+    },
+    {
+      "epoch": 0.4320048676604807,
+      "grad_norm": 1.4245113134384155,
+      "learning_rate": 4.9383248888574916e-05,
+      "loss": 0.5151,
+      "step": 710
+    },
+    {
+      "epoch": 0.4326133252205659,
+      "grad_norm": 1.6800333261489868,
+      "learning_rate": 4.938049775546846e-05,
+      "loss": 0.5604,
+      "step": 711
+    },
+    {
+      "epoch": 0.43322178278065104,
+      "grad_norm": 1.3402103185653687,
+      "learning_rate": 4.9377740577014784e-05,
+      "loss": 0.5283,
+      "step": 712
+    },
+    {
+      "epoch": 0.43383024034073625,
+      "grad_norm": 1.3066250085830688,
+      "learning_rate": 4.9374977353897566e-05,
+      "loss": 0.4636,
+      "step": 713
+    },
+    {
+      "epoch": 0.4344386979008214,
+      "grad_norm": 1.3179736137390137,
+      "learning_rate": 4.937220808680196e-05,
+      "loss": 0.471,
+      "step": 714
+    },
+    {
+      "epoch": 0.4350471554609066,
+      "grad_norm": 1.3768038749694824,
+      "learning_rate": 4.9369432776414634e-05,
+      "loss": 0.538,
+      "step": 715
+    },
+    {
+      "epoch": 0.43565561302099176,
+      "grad_norm": 1.1771548986434937,
+      "learning_rate": 4.936665142342375e-05,
+      "loss": 0.4856,
+      "step": 716
+    },
+    {
+      "epoch": 0.43626407058107697,
+      "grad_norm": 2.29483962059021,
+      "learning_rate": 4.936386402851896e-05,
+      "loss": 0.4863,
+      "step": 717
+    },
+    {
+      "epoch": 0.4368725281411622,
+      "grad_norm": 1.554796576499939,
+      "learning_rate": 4.936107059239143e-05,
+      "loss": 0.5158,
+      "step": 718
+    },
+    {
+      "epoch": 0.4374809857012473,
+      "grad_norm": 1.3302370309829712,
+      "learning_rate": 4.935827111573381e-05,
+      "loss": 0.5295,
+      "step": 719
+    },
+    {
+      "epoch": 0.43808944326133253,
+      "grad_norm": 1.2260332107543945,
+      "learning_rate": 4.9355465599240265e-05,
+      "loss": 0.5124,
+      "step": 720
+    },
+    {
+      "epoch": 0.4386979008214177,
+      "grad_norm": 1.272337555885315,
+      "learning_rate": 4.935265404360643e-05,
+      "loss": 0.4962,
+      "step": 721
+    },
+    {
+      "epoch": 0.4393063583815029,
+      "grad_norm": 1.1549131870269775,
+      "learning_rate": 4.9349836449529463e-05,
+      "loss": 0.5208,
+      "step": 722
+    },
+    {
+      "epoch": 0.4399148159415881,
+      "grad_norm": 1.3710724115371704,
+      "learning_rate": 4.9347012817708e-05,
+      "loss": 0.5462,
+      "step": 723
+    },
+    {
+      "epoch": 0.44052327350167325,
+      "grad_norm": 1.3693798780441284,
+      "learning_rate": 4.93441831488422e-05,
+      "loss": 0.5133,
+      "step": 724
+    },
+    {
+      "epoch": 0.44113173106175846,
+      "grad_norm": 1.2647064924240112,
+      "learning_rate": 4.934134744363369e-05,
+      "loss": 0.5089,
+      "step": 725
+    },
+    {
+      "epoch": 0.4417401886218436,
+      "grad_norm": 1.302121877670288,
+      "learning_rate": 4.933850570278562e-05,
+      "loss": 0.4566,
+      "step": 726
+    },
+    {
+      "epoch": 0.4423486461819288,
+      "grad_norm": 1.633015751838684,
+      "learning_rate": 4.933565792700261e-05,
+      "loss": 0.5596,
+      "step": 727
+    },
+    {
+      "epoch": 0.44295710374201397,
+      "grad_norm": 1.376413345336914,
+      "learning_rate": 4.9332804116990795e-05,
+      "loss": 0.5597,
+      "step": 728
+    },
+    {
+      "epoch": 0.4435655613020992,
+      "grad_norm": 1.3377238512039185,
+      "learning_rate": 4.9329944273457794e-05,
+      "loss": 0.4782,
+      "step": 729
+    },
+    {
+      "epoch": 0.4441740188621844,
+      "grad_norm": 1.3485767841339111,
+      "learning_rate": 4.932707839711273e-05,
+      "loss": 0.504,
+      "step": 730
+    },
+    {
+      "epoch": 0.44478247642226953,
+      "grad_norm": 1.3925731182098389,
+      "learning_rate": 4.9324206488666244e-05,
+      "loss": 0.5548,
+      "step": 731
+    },
+    {
+      "epoch": 0.44539093398235474,
+      "grad_norm": 1.1115320920944214,
+      "learning_rate": 4.9321328548830426e-05,
+      "loss": 0.5035,
+      "step": 732
+    },
+    {
+      "epoch": 0.4459993915424399,
+      "grad_norm": 1.2374329566955566,
+      "learning_rate": 4.9318444578318886e-05,
+      "loss": 0.47,
+      "step": 733
+    },
+    {
+      "epoch": 0.4466078491025251,
+      "grad_norm": 1.378677487373352,
+      "learning_rate": 4.931555457784674e-05,
+      "loss": 0.5206,
+      "step": 734
+    },
+    {
+      "epoch": 0.4472163066626103,
+      "grad_norm": 1.3036342859268188,
+      "learning_rate": 4.931265854813057e-05,
+      "loss": 0.4869,
+      "step": 735
+    },
+    {
+      "epoch": 0.44782476422269546,
+      "grad_norm": 1.1578119993209839,
+      "learning_rate": 4.930975648988849e-05,
+      "loss": 0.495,
+      "step": 736
+    },
+    {
+      "epoch": 0.44843322178278067,
+      "grad_norm": 1.815414309501648,
+      "learning_rate": 4.930684840384008e-05,
+      "loss": 0.6301,
+      "step": 737
+    },
+    {
+      "epoch": 0.4490416793428658,
+      "grad_norm": 1.4656637907028198,
+      "learning_rate": 4.9303934290706424e-05,
+      "loss": 0.4809,
+      "step": 738
+    },
+    {
+      "epoch": 0.449650136902951,
+      "grad_norm": 1.5585298538208008,
+      "learning_rate": 4.93010141512101e-05,
+      "loss": 0.5067,
+      "step": 739
+    },
+    {
+      "epoch": 0.4502585944630362,
+      "grad_norm": 1.466367483139038,
+      "learning_rate": 4.929808798607518e-05,
+      "loss": 0.5188,
+      "step": 740
+    },
+    {
+      "epoch": 0.4508670520231214,
+      "grad_norm": 1.1880073547363281,
+      "learning_rate": 4.9295155796027244e-05,
+      "loss": 0.5133,
+      "step": 741
+    },
+    {
+      "epoch": 0.4514755095832066,
+      "grad_norm": 1.425815224647522,
+      "learning_rate": 4.929221758179333e-05,
+      "loss": 0.5826,
+      "step": 742
+    },
+    {
+      "epoch": 0.45208396714329174,
+      "grad_norm": 1.4366766214370728,
+      "learning_rate": 4.9289273344102014e-05,
+      "loss": 0.5136,
+      "step": 743
+    },
+    {
+      "epoch": 0.45269242470337695,
+      "grad_norm": 1.2157377004623413,
+      "learning_rate": 4.928632308368334e-05,
+      "loss": 0.4507,
+      "step": 744
+    },
+    {
+      "epoch": 0.4533008822634621,
+      "grad_norm": 1.1545383930206299,
+      "learning_rate": 4.928336680126884e-05,
+      "loss": 0.4523,
+      "step": 745
+    },
+    {
+      "epoch": 0.4539093398235473,
+      "grad_norm": 1.2865034341812134,
+      "learning_rate": 4.9280404497591545e-05,
+      "loss": 0.4843,
+      "step": 746
+    },
+    {
+      "epoch": 0.4545177973836325,
+      "grad_norm": 1.1949740648269653,
+      "learning_rate": 4.9277436173386006e-05,
+      "loss": 0.4669,
+      "step": 747
+    },
+    {
+      "epoch": 0.45512625494371767,
+      "grad_norm": 1.2178860902786255,
+      "learning_rate": 4.927446182938822e-05,
+      "loss": 0.5262,
+      "step": 748
+    },
+    {
+      "epoch": 0.4557347125038029,
+      "grad_norm": 1.3532716035842896,
+      "learning_rate": 4.927148146633571e-05,
+      "loss": 0.4994,
+      "step": 749
+    },
+    {
+      "epoch": 0.456343170063888,
+      "grad_norm": 1.2420387268066406,
+      "learning_rate": 4.9268495084967485e-05,
+      "loss": 0.4912,
+      "step": 750
+    },
+    {
+      "epoch": 0.45695162762397323,
+      "grad_norm": 1.242753267288208,
+      "learning_rate": 4.926550268602404e-05,
+      "loss": 0.5301,
+      "step": 751
+    },
+    {
+      "epoch": 0.4575600851840584,
+      "grad_norm": 1.2774479389190674,
+      "learning_rate": 4.926250427024736e-05,
+      "loss": 0.4994,
+      "step": 752
+    },
+    {
+      "epoch": 0.4581685427441436,
+      "grad_norm": 1.3734140396118164,
+      "learning_rate": 4.925949983838094e-05,
+      "loss": 0.5812,
+      "step": 753
+    },
+    {
+      "epoch": 0.4587770003042288,
+      "grad_norm": 1.4266109466552734,
+      "learning_rate": 4.925648939116974e-05,
+      "loss": 0.5418,
+      "step": 754
+    },
+    {
+      "epoch": 0.45938545786431395,
+      "grad_norm": 1.2357332706451416,
+      "learning_rate": 4.9253472929360235e-05,
+      "loss": 0.5441,
+      "step": 755
+    },
+    {
+      "epoch": 0.45999391542439916,
+      "grad_norm": 1.4429142475128174,
+      "learning_rate": 4.925045045370037e-05,
+      "loss": 0.5116,
+      "step": 756
+    },
+    {
+      "epoch": 0.4606023729844843,
+      "grad_norm": 1.2204004526138306,
+      "learning_rate": 4.924742196493961e-05,
+      "loss": 0.4969,
+      "step": 757
+    },
+    {
+      "epoch": 0.4612108305445695,
+      "grad_norm": 1.2525185346603394,
+      "learning_rate": 4.9244387463828876e-05,
+      "loss": 0.4672,
+      "step": 758
+    },
+    {
+      "epoch": 0.4618192881046547,
+      "grad_norm": 1.238554835319519,
+      "learning_rate": 4.9241346951120616e-05,
+      "loss": 0.5296,
+      "step": 759
+    },
+    {
+      "epoch": 0.4624277456647399,
+      "grad_norm": 1.2626805305480957,
+      "learning_rate": 4.923830042756874e-05,
+      "loss": 0.5025,
+      "step": 760
+    },
+    {
+      "epoch": 0.4630362032248251,
+      "grad_norm": 1.3802168369293213,
+      "learning_rate": 4.923524789392866e-05,
+      "loss": 0.5273,
+      "step": 761
+    },
+    {
+      "epoch": 0.46364466078491023,
+      "grad_norm": 1.498776912689209,
+      "learning_rate": 4.923218935095727e-05,
+      "loss": 0.4597,
+      "step": 762
+    },
+    {
+      "epoch": 0.46425311834499544,
+      "grad_norm": 1.449130892753601,
+      "learning_rate": 4.922912479941297e-05,
+      "loss": 0.533,
+      "step": 763
+    },
+    {
+      "epoch": 0.4648615759050806,
+      "grad_norm": 2.3711140155792236,
+      "learning_rate": 4.922605424005565e-05,
+      "loss": 0.5551,
+      "step": 764
+    },
+    {
+      "epoch": 0.4654700334651658,
+      "grad_norm": 1.1763001680374146,
+      "learning_rate": 4.922297767364666e-05,
+      "loss": 0.4951,
+      "step": 765
+    },
+    {
+      "epoch": 0.466078491025251,
+      "grad_norm": 1.5831891298294067,
+      "learning_rate": 4.921989510094888e-05,
+      "loss": 0.6159,
+      "step": 766
+    },
+    {
+      "epoch": 0.46668694858533616,
+      "grad_norm": 1.3769197463989258,
+      "learning_rate": 4.921680652272665e-05,
+      "loss": 0.5048,
+      "step": 767
+    },
+    {
+      "epoch": 0.46729540614542137,
+      "grad_norm": 1.137963056564331,
+      "learning_rate": 4.9213711939745795e-05,
+      "loss": 0.4931,
+      "step": 768
+    },
+    {
+      "epoch": 0.4679038637055065,
+      "grad_norm": 1.7989009618759155,
+      "learning_rate": 4.921061135277366e-05,
+      "loss": 0.5065,
+      "step": 769
+    },
+    {
+      "epoch": 0.4685123212655917,
+      "grad_norm": 1.294255256652832,
+      "learning_rate": 4.920750476257906e-05,
+      "loss": 0.4863,
+      "step": 770
+    },
+    {
+      "epoch": 0.46912077882567693,
+      "grad_norm": 1.5027679204940796,
+      "learning_rate": 4.92043921699323e-05,
+      "loss": 0.507,
+      "step": 771
+    },
+    {
+      "epoch": 0.4697292363857621,
+      "grad_norm": 1.1163376569747925,
+      "learning_rate": 4.920127357560517e-05,
+      "loss": 0.4507,
+      "step": 772
+    },
+    {
+      "epoch": 0.4703376939458473,
+      "grad_norm": 1.1562211513519287,
+      "learning_rate": 4.919814898037095e-05,
+      "loss": 0.4933,
+      "step": 773
+    },
+    {
+      "epoch": 0.47094615150593244,
+      "grad_norm": 1.191632866859436,
+      "learning_rate": 4.919501838500441e-05,
+      "loss": 0.4549,
+      "step": 774
+    },
+    {
+      "epoch": 0.47155460906601765,
+      "grad_norm": 1.2718918323516846,
+      "learning_rate": 4.9191881790281815e-05,
+      "loss": 0.5133,
+      "step": 775
+    },
+    {
+      "epoch": 0.47216306662610286,
+      "grad_norm": 1.677453875541687,
+      "learning_rate": 4.91887391969809e-05,
+      "loss": 0.4517,
+      "step": 776
+    },
+    {
+      "epoch": 0.472771524186188,
+      "grad_norm": 1.3520299196243286,
+      "learning_rate": 4.91855906058809e-05,
+      "loss": 0.4373,
+      "step": 777
+    },
+    {
+      "epoch": 0.4733799817462732,
+      "grad_norm": 1.4283875226974487,
+      "learning_rate": 4.9182436017762535e-05,
+      "loss": 0.517,
+      "step": 778
+    },
+    {
+      "epoch": 0.47398843930635837,
+      "grad_norm": 1.2896888256072998,
+      "learning_rate": 4.917927543340801e-05,
+      "loss": 0.5267,
+      "step": 779
+    },
+    {
+      "epoch": 0.4745968968664436,
+      "grad_norm": 1.275476098060608,
+      "learning_rate": 4.9176108853601024e-05,
+      "loss": 0.4542,
+      "step": 780
+    },
+    {
+      "epoch": 0.4752053544265287,
+      "grad_norm": 1.1607316732406616,
+      "learning_rate": 4.917293627912675e-05,
+      "loss": 0.4794,
+      "step": 781
+    },
+    {
+      "epoch": 0.47581381198661393,
+      "grad_norm": 1.6451760530471802,
+      "learning_rate": 4.916975771077185e-05,
+      "loss": 0.5549,
+      "step": 782
+    },
+    {
+      "epoch": 0.47642226954669914,
+      "grad_norm": 1.3104053735733032,
+      "learning_rate": 4.9166573149324486e-05,
+      "loss": 0.4835,
+      "step": 783
+    },
+    {
+      "epoch": 0.4770307271067843,
+      "grad_norm": 1.5024620294570923,
+      "learning_rate": 4.916338259557429e-05,
+      "loss": 0.5095,
+      "step": 784
+    },
+    {
+      "epoch": 0.4776391846668695,
+      "grad_norm": 1.4474741220474243,
+      "learning_rate": 4.91601860503124e-05,
+      "loss": 0.5574,
+      "step": 785
+    },
+    {
+      "epoch": 0.47824764222695465,
+      "grad_norm": 1.329298734664917,
+      "learning_rate": 4.915698351433141e-05,
+      "loss": 0.5496,
+      "step": 786
+    },
+    {
+      "epoch": 0.47885609978703986,
+      "grad_norm": 1.1795927286148071,
+      "learning_rate": 4.915377498842542e-05,
+      "loss": 0.4553,
+      "step": 787
+    },
+    {
+      "epoch": 0.47946455734712506,
+      "grad_norm": 1.329775094985962,
+      "learning_rate": 4.915056047339002e-05,
+      "loss": 0.5445,
+      "step": 788
+    },
+    {
+      "epoch": 0.4800730149072102,
+      "grad_norm": 1.1533706188201904,
+      "learning_rate": 4.9147339970022256e-05,
+      "loss": 0.507,
+      "step": 789
+    },
+    {
+      "epoch": 0.4806814724672954,
+      "grad_norm": 1.1243101358413696,
+      "learning_rate": 4.9144113479120695e-05,
+      "loss": 0.5018,
+      "step": 790
+    },
+    {
+      "epoch": 0.4812899300273806,
+      "grad_norm": 1.1606624126434326,
+      "learning_rate": 4.9140881001485374e-05,
+      "loss": 0.4971,
+      "step": 791
+    },
+    {
+      "epoch": 0.4818983875874658,
+      "grad_norm": 1.342514991760254,
+      "learning_rate": 4.91376425379178e-05,
+      "loss": 0.541,
+      "step": 792
+    },
+    {
+      "epoch": 0.48250684514755093,
+      "grad_norm": 1.347623586654663,
+      "learning_rate": 4.913439808922098e-05,
+      "loss": 0.509,
+      "step": 793
+    },
+    {
+      "epoch": 0.48311530270763614,
+      "grad_norm": 1.340950608253479,
+      "learning_rate": 4.91311476561994e-05,
+      "loss": 0.4773,
+      "step": 794
+    },
+    {
+      "epoch": 0.48372376026772135,
+      "grad_norm": 1.0880382061004639,
+      "learning_rate": 4.912789123965905e-05,
+      "loss": 0.4569,
+      "step": 795
+    },
+    {
+      "epoch": 0.4843322178278065,
+      "grad_norm": 1.922781229019165,
+      "learning_rate": 4.912462884040737e-05,
+      "loss": 0.4752,
+      "step": 796
+    },
+    {
+      "epoch": 0.4849406753878917,
+      "grad_norm": 1.0886273384094238,
+      "learning_rate": 4.91213604592533e-05,
+      "loss": 0.4649,
+      "step": 797
+    },
+    {
+      "epoch": 0.48554913294797686,
+      "grad_norm": 1.411138892173767,
+      "learning_rate": 4.911808609700726e-05,
+      "loss": 0.5271,
+      "step": 798
+    },
+    {
+      "epoch": 0.48615759050806207,
+      "grad_norm": 0.9868936538696289,
+      "learning_rate": 4.911480575448116e-05,
+      "loss": 0.435,
+      "step": 799
+    },
+    {
+      "epoch": 0.4867660480681473,
+      "grad_norm": 1.06026291847229,
+      "learning_rate": 4.911151943248839e-05,
+      "loss": 0.4693,
+      "step": 800
+    },
+    {
+      "epoch": 0.4873745056282324,
+      "grad_norm": 1.5773180723190308,
+      "learning_rate": 4.910822713184382e-05,
+      "loss": 0.5369,
+      "step": 801
+    },
+    {
+      "epoch": 0.48798296318831763,
+      "grad_norm": 1.2286581993103027,
+      "learning_rate": 4.910492885336381e-05,
+      "loss": 0.4814,
+      "step": 802
+    },
+    {
+      "epoch": 0.4885914207484028,
+      "grad_norm": 1.3670451641082764,
+      "learning_rate": 4.910162459786617e-05,
+      "loss": 0.5145,
+      "step": 803
+    },
+    {
+      "epoch": 0.489199878308488,
+      "grad_norm": 1.6404004096984863,
+      "learning_rate": 4.9098314366170245e-05,
+      "loss": 0.5741,
+      "step": 804
+    },
+    {
+      "epoch": 0.48980833586857314,
+      "grad_norm": 1.1956790685653687,
+      "learning_rate": 4.909499815909682e-05,
+      "loss": 0.4935,
+      "step": 805
+    },
+    {
+      "epoch": 0.49041679342865835,
+      "grad_norm": 1.4000447988510132,
+      "learning_rate": 4.909167597746819e-05,
+      "loss": 0.4661,
+      "step": 806
+    },
+    {
+      "epoch": 0.49102525098874356,
+      "grad_norm": 1.4137040376663208,
+      "learning_rate": 4.908834782210809e-05,
+      "loss": 0.5235,
+      "step": 807
+    },
+    {
+      "epoch": 0.4916337085488287,
+      "grad_norm": 1.3332107067108154,
+      "learning_rate": 4.90850136938418e-05,
+      "loss": 0.5003,
+      "step": 808
+    },
+    {
+      "epoch": 0.4922421661089139,
+      "grad_norm": 1.3718630075454712,
+      "learning_rate": 4.908167359349601e-05,
+      "loss": 0.4756,
+      "step": 809
+    },
+    {
+      "epoch": 0.49285062366899907,
+      "grad_norm": 1.3051707744598389,
+      "learning_rate": 4.907832752189896e-05,
+      "loss": 0.5024,
+      "step": 810
+    },
+    {
+      "epoch": 0.4934590812290843,
+      "grad_norm": 1.2241960763931274,
+      "learning_rate": 4.907497547988031e-05,
+      "loss": 0.4989,
+      "step": 811
+    },
+    {
+      "epoch": 0.4940675387891695,
+      "grad_norm": 1.3939323425292969,
+      "learning_rate": 4.9071617468271234e-05,
+      "loss": 0.5231,
+      "step": 812
+    },
+    {
+      "epoch": 0.49467599634925463,
+      "grad_norm": 1.4326186180114746,
+      "learning_rate": 4.906825348790438e-05,
+      "loss": 0.4494,
+      "step": 813
+    },
+    {
+      "epoch": 0.49528445390933984,
+      "grad_norm": 1.1022182703018188,
+      "learning_rate": 4.9064883539613884e-05,
+      "loss": 0.5195,
+      "step": 814
+    },
+    {
+      "epoch": 0.495892911469425,
+      "grad_norm": 1.2197952270507812,
+      "learning_rate": 4.9061507624235334e-05,
+      "loss": 0.4793,
+      "step": 815
+    },
+    {
+      "epoch": 0.4965013690295102,
+      "grad_norm": 1.329502820968628,
+      "learning_rate": 4.905812574260583e-05,
+      "loss": 0.5823,
+      "step": 816
+    },
+    {
+      "epoch": 0.49710982658959535,
+      "grad_norm": 1.4672378301620483,
+      "learning_rate": 4.9054737895563935e-05,
+      "loss": 0.5231,
+      "step": 817
+    },
+    {
+      "epoch": 0.49771828414968056,
+      "grad_norm": 1.1329609155654907,
+      "learning_rate": 4.905134408394969e-05,
+      "loss": 0.501,
+      "step": 818
+    },
+    {
+      "epoch": 0.49832674170976576,
+      "grad_norm": 1.1195259094238281,
+      "learning_rate": 4.904794430860462e-05,
+      "loss": 0.482,
+      "step": 819
+    },
+    {
+      "epoch": 0.4989351992698509,
+      "grad_norm": 1.258161187171936,
+      "learning_rate": 4.904453857037173e-05,
+      "loss": 0.4575,
+      "step": 820
+    },
+    {
+      "epoch": 0.4995436568299361,
+      "grad_norm": 1.2955543994903564,
+      "learning_rate": 4.904112687009551e-05,
+      "loss": 0.563,
+      "step": 821
+    },
+    {
+      "epoch": 0.5001521143900213,
+      "grad_norm": 1.3455455303192139,
+      "learning_rate": 4.90377092086219e-05,
+      "loss": 0.5372,
+      "step": 822
+    },
+    {
+      "epoch": 0.5007605719501065,
+      "grad_norm": 1.2363775968551636,
+      "learning_rate": 4.903428558679835e-05,
+      "loss": 0.4754,
+      "step": 823
+    },
+    {
+      "epoch": 0.5013690295101917,
+      "grad_norm": 1.1832222938537598,
+      "learning_rate": 4.9030856005473776e-05,
+      "loss": 0.4748,
+      "step": 824
+    },
+    {
+      "epoch": 0.5019774870702769,
+      "grad_norm": 1.2365323305130005,
+      "learning_rate": 4.902742046549856e-05,
+      "loss": 0.5218,
+      "step": 825
+    },
+    {
+      "epoch": 0.502585944630362,
+      "grad_norm": 1.120848298072815,
+      "learning_rate": 4.902397896772459e-05,
+      "loss": 0.4514,
+      "step": 826
+    },
+    {
+      "epoch": 0.5031944021904472,
+      "grad_norm": 1.7753840684890747,
+      "learning_rate": 4.9020531513005194e-05,
+      "loss": 0.5859,
+      "step": 827
+    },
+    {
+      "epoch": 0.5038028597505324,
+      "grad_norm": 1.2528876066207886,
+      "learning_rate": 4.901707810219522e-05,
+      "loss": 0.4974,
+      "step": 828
+    },
+    {
+      "epoch": 0.5044113173106176,
+      "grad_norm": 1.1275379657745361,
+      "learning_rate": 4.901361873615095e-05,
+      "loss": 0.4611,
+      "step": 829
+    },
+    {
+      "epoch": 0.5050197748707028,
+      "grad_norm": 1.2789126634597778,
+      "learning_rate": 4.901015341573017e-05,
+      "loss": 0.4913,
+      "step": 830
+    },
+    {
+      "epoch": 0.5056282324307879,
+      "grad_norm": 1.2548869848251343,
+      "learning_rate": 4.900668214179214e-05,
+      "loss": 0.5161,
+      "step": 831
+    },
+    {
+      "epoch": 0.5062366899908731,
+      "grad_norm": 1.1596348285675049,
+      "learning_rate": 4.900320491519759e-05,
+      "loss": 0.5323,
+      "step": 832
+    },
+    {
+      "epoch": 0.5068451475509583,
+      "grad_norm": 1.0443363189697266,
+      "learning_rate": 4.8999721736808714e-05,
+      "loss": 0.4074,
+      "step": 833
+    },
+    {
+      "epoch": 0.5074536051110435,
+      "grad_norm": 1.1528189182281494,
+      "learning_rate": 4.899623260748921e-05,
+      "loss": 0.4656,
+      "step": 834
+    },
+    {
+      "epoch": 0.5080620626711286,
+      "grad_norm": 1.4290034770965576,
+      "learning_rate": 4.899273752810423e-05,
+      "loss": 0.5818,
+      "step": 835
+    },
+    {
+      "epoch": 0.5086705202312138,
+      "grad_norm": 1.298559546470642,
+      "learning_rate": 4.898923649952041e-05,
+      "loss": 0.5283,
+      "step": 836
+    },
+    {
+      "epoch": 0.509278977791299,
+      "grad_norm": 1.4868043661117554,
+      "learning_rate": 4.8985729522605864e-05,
+      "loss": 0.4575,
+      "step": 837
+    },
+    {
+      "epoch": 0.5098874353513843,
+      "grad_norm": 1.40908944606781,
+      "learning_rate": 4.898221659823016e-05,
+      "loss": 0.5026,
+      "step": 838
+    },
+    {
+      "epoch": 0.5104958929114695,
+      "grad_norm": 1.5016429424285889,
+      "learning_rate": 4.897869772726438e-05,
+      "loss": 0.4975,
+      "step": 839
+    },
+    {
+      "epoch": 0.5111043504715546,
+      "grad_norm": 1.2684623003005981,
+      "learning_rate": 4.8975172910581033e-05,
+      "loss": 0.5549,
+      "step": 840
+    },
+    {
+      "epoch": 0.5117128080316398,
+      "grad_norm": 1.4201849699020386,
+      "learning_rate": 4.897164214905414e-05,
+      "loss": 0.5264,
+      "step": 841
+    },
+    {
+      "epoch": 0.512321265591725,
+      "grad_norm": 1.300970435142517,
+      "learning_rate": 4.8968105443559194e-05,
+      "loss": 0.509,
+      "step": 842
+    },
+    {
+      "epoch": 0.5129297231518102,
+      "grad_norm": 1.3174622058868408,
+      "learning_rate": 4.896456279497312e-05,
+      "loss": 0.4918,
+      "step": 843
+    },
+    {
+      "epoch": 0.5135381807118954,
+      "grad_norm": 1.2107031345367432,
+      "learning_rate": 4.8961014204174384e-05,
+      "loss": 0.5297,
+      "step": 844
+    },
+    {
+      "epoch": 0.5141466382719805,
+      "grad_norm": 1.4777075052261353,
+      "learning_rate": 4.895745967204286e-05,
+      "loss": 0.4719,
+      "step": 845
+    },
+    {
+      "epoch": 0.5147550958320657,
+      "grad_norm": 1.252469539642334,
+      "learning_rate": 4.895389919945993e-05,
+      "loss": 0.4674,
+      "step": 846
+    },
+    {
+      "epoch": 0.5153635533921509,
+      "grad_norm": 1.2273253202438354,
+      "learning_rate": 4.895033278730845e-05,
+      "loss": 0.5269,
+      "step": 847
+    },
+    {
+      "epoch": 0.5159720109522361,
+      "grad_norm": 1.6084344387054443,
+      "learning_rate": 4.894676043647274e-05,
+      "loss": 0.4133,
+      "step": 848
+    },
+    {
+      "epoch": 0.5165804685123213,
+      "grad_norm": 0.9914146065711975,
+      "learning_rate": 4.894318214783859e-05,
+      "loss": 0.4391,
+      "step": 849
+    },
+    {
+      "epoch": 0.5171889260724064,
+      "grad_norm": 1.3645412921905518,
+      "learning_rate": 4.893959792229327e-05,
+      "loss": 0.5509,
+      "step": 850
+    },
+    {
+      "epoch": 0.5177973836324916,
+      "grad_norm": 1.2271888256072998,
+      "learning_rate": 4.8936007760725514e-05,
+      "loss": 0.5111,
+      "step": 851
+    },
+    {
+      "epoch": 0.5184058411925768,
+      "grad_norm": 1.262268304824829,
+      "learning_rate": 4.893241166402553e-05,
+      "loss": 0.4809,
+      "step": 852
+    },
+    {
+      "epoch": 0.519014298752662,
+      "grad_norm": 1.3610913753509521,
+      "learning_rate": 4.892880963308502e-05,
+      "loss": 0.5159,
+      "step": 853
+    },
+    {
+      "epoch": 0.5196227563127472,
+      "grad_norm": 1.3238227367401123,
+      "learning_rate": 4.8925201668797117e-05,
+      "loss": 0.5162,
+      "step": 854
+    },
+    {
+      "epoch": 0.5202312138728323,
+      "grad_norm": 1.2725099325180054,
+      "learning_rate": 4.8921587772056444e-05,
+      "loss": 0.4616,
+      "step": 855
+    },
+    {
+      "epoch": 0.5208396714329175,
+      "grad_norm": 1.4523481130599976,
+      "learning_rate": 4.8917967943759114e-05,
+      "loss": 0.6087,
+      "step": 856
+    },
+    {
+      "epoch": 0.5214481289930027,
+      "grad_norm": 1.192795753479004,
+      "learning_rate": 4.8914342184802675e-05,
+      "loss": 0.4805,
+      "step": 857
+    },
+    {
+      "epoch": 0.522056586553088,
+      "grad_norm": 1.2161285877227783,
+      "learning_rate": 4.891071049608618e-05,
+      "loss": 0.481,
+      "step": 858
+    },
+    {
+      "epoch": 0.5226650441131732,
+      "grad_norm": 1.391660213470459,
+      "learning_rate": 4.890707287851013e-05,
+      "loss": 0.4796,
+      "step": 859
+    },
+    {
+      "epoch": 0.5232735016732583,
+      "grad_norm": 1.189489722251892,
+      "learning_rate": 4.8903429332976494e-05,
+      "loss": 0.5197,
+      "step": 860
+    },
+    {
+      "epoch": 0.5238819592333435,
+      "grad_norm": 1.2670400142669678,
+      "learning_rate": 4.889977986038874e-05,
+      "loss": 0.4648,
+      "step": 861
+    },
+    {
+      "epoch": 0.5244904167934287,
+      "grad_norm": 1.570553183555603,
+      "learning_rate": 4.889612446165176e-05,
+      "loss": 0.5102,
+      "step": 862
+    },
+    {
+      "epoch": 0.5250988743535139,
+      "grad_norm": 1.1414783000946045,
+      "learning_rate": 4.8892463137671963e-05,
+      "loss": 0.4489,
+      "step": 863
+    },
+    {
+      "epoch": 0.525707331913599,
+      "grad_norm": 1.4709701538085938,
+      "learning_rate": 4.888879588935719e-05,
+      "loss": 0.5205,
+      "step": 864
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 1.218973159790039,
+      "learning_rate": 4.888512271761677e-05,
+      "loss": 0.466,
+      "step": 865
+    },
+    {
+      "epoch": 0.5269242470337694,
+      "grad_norm": 1.2933622598648071,
+      "learning_rate": 4.88814436233615e-05,
+      "loss": 0.5473,
+      "step": 866
+    },
+    {
+      "epoch": 0.5275327045938546,
+      "grad_norm": 1.085365891456604,
+      "learning_rate": 4.887775860750363e-05,
+      "loss": 0.4572,
+      "step": 867
+    },
+    {
+      "epoch": 0.5281411621539398,
+      "grad_norm": 1.5779204368591309,
+      "learning_rate": 4.8874067670956905e-05,
+      "loss": 0.4428,
+      "step": 868
+    },
+    {
+      "epoch": 0.5287496197140249,
+      "grad_norm": 1.36058509349823,
+      "learning_rate": 4.887037081463652e-05,
+      "loss": 0.5667,
+      "step": 869
+    },
+    {
+      "epoch": 0.5293580772741101,
+      "grad_norm": 1.1790504455566406,
+      "learning_rate": 4.886666803945914e-05,
+      "loss": 0.5067,
+      "step": 870
+    },
+    {
+      "epoch": 0.5299665348341953,
+      "grad_norm": 1.2920281887054443,
+      "learning_rate": 4.886295934634289e-05,
+      "loss": 0.5173,
+      "step": 871
+    },
+    {
+      "epoch": 0.5305749923942805,
+      "grad_norm": 1.085190773010254,
+      "learning_rate": 4.8859244736207395e-05,
+      "loss": 0.4904,
+      "step": 872
+    },
+    {
+      "epoch": 0.5311834499543657,
+      "grad_norm": 1.1711935997009277,
+      "learning_rate": 4.885552420997369e-05,
+      "loss": 0.4416,
+      "step": 873
+    },
+    {
+      "epoch": 0.5317919075144508,
+      "grad_norm": 1.4009342193603516,
+      "learning_rate": 4.885179776856435e-05,
+      "loss": 0.529,
+      "step": 874
+    },
+    {
+      "epoch": 0.532400365074536,
+      "grad_norm": 1.261432409286499,
+      "learning_rate": 4.8848065412903335e-05,
+      "loss": 0.5338,
+      "step": 875
+    },
+    {
+      "epoch": 0.5330088226346212,
+      "grad_norm": 1.0821967124938965,
+      "learning_rate": 4.884432714391613e-05,
+      "loss": 0.4251,
+      "step": 876
+    },
+    {
+      "epoch": 0.5336172801947064,
+      "grad_norm": 1.4930713176727295,
+      "learning_rate": 4.884058296252969e-05,
+      "loss": 0.5063,
+      "step": 877
+    },
+    {
+      "epoch": 0.5342257377547917,
+      "grad_norm": 1.2942625284194946,
+      "learning_rate": 4.88368328696724e-05,
+      "loss": 0.4825,
+      "step": 878
+    },
+    {
+      "epoch": 0.5348341953148767,
+      "grad_norm": 1.1449693441390991,
+      "learning_rate": 4.883307686627412e-05,
+      "loss": 0.4833,
+      "step": 879
+    },
+    {
+      "epoch": 0.535442652874962,
+      "grad_norm": 1.2848173379898071,
+      "learning_rate": 4.882931495326619e-05,
+      "loss": 0.4981,
+      "step": 880
+    },
+    {
+      "epoch": 0.5360511104350472,
+      "grad_norm": 1.0941145420074463,
+      "learning_rate": 4.882554713158141e-05,
+      "loss": 0.4457,
+      "step": 881
+    },
+    {
+      "epoch": 0.5366595679951324,
+      "grad_norm": 1.2851606607437134,
+      "learning_rate": 4.8821773402154025e-05,
+      "loss": 0.4238,
+      "step": 882
+    },
+    {
+      "epoch": 0.5372680255552176,
+      "grad_norm": 1.5043134689331055,
+      "learning_rate": 4.881799376591979e-05,
+      "loss": 0.53,
+      "step": 883
+    },
+    {
+      "epoch": 0.5378764831153027,
+      "grad_norm": 1.2732433080673218,
+      "learning_rate": 4.8814208223815886e-05,
+      "loss": 0.5126,
+      "step": 884
+    },
+    {
+      "epoch": 0.5384849406753879,
+      "grad_norm": 1.2558897733688354,
+      "learning_rate": 4.8810416776780956e-05,
+      "loss": 0.4542,
+      "step": 885
+    },
+    {
+      "epoch": 0.5390933982354731,
+      "grad_norm": 1.1462931632995605,
+      "learning_rate": 4.880661942575514e-05,
+      "loss": 0.4536,
+      "step": 886
+    },
+    {
+      "epoch": 0.5397018557955583,
+      "grad_norm": 1.2264448404312134,
+      "learning_rate": 4.880281617168001e-05,
+      "loss": 0.4497,
+      "step": 887
+    },
+    {
+      "epoch": 0.5403103133556434,
+      "grad_norm": 1.3015364408493042,
+      "learning_rate": 4.879900701549863e-05,
+      "loss": 0.4697,
+      "step": 888
+    },
+    {
+      "epoch": 0.5409187709157286,
+      "grad_norm": 1.5858356952667236,
+      "learning_rate": 4.879519195815549e-05,
+      "loss": 0.4573,
+      "step": 889
+    },
+    {
+      "epoch": 0.5415272284758138,
+      "grad_norm": 1.2215701341629028,
+      "learning_rate": 4.8791371000596585e-05,
+      "loss": 0.5027,
+      "step": 890
+    },
+    {
+      "epoch": 0.542135686035899,
+      "grad_norm": 1.3186434507369995,
+      "learning_rate": 4.8787544143769335e-05,
+      "loss": 0.5387,
+      "step": 891
+    },
+    {
+      "epoch": 0.5427441435959842,
+      "grad_norm": 1.487970232963562,
+      "learning_rate": 4.878371138862267e-05,
+      "loss": 0.5038,
+      "step": 892
+    },
+    {
+      "epoch": 0.5433526011560693,
+      "grad_norm": 1.4170433282852173,
+      "learning_rate": 4.8779872736106916e-05,
+      "loss": 0.4948,
+      "step": 893
+    },
+    {
+      "epoch": 0.5439610587161545,
+      "grad_norm": 1.167901635169983,
+      "learning_rate": 4.877602818717393e-05,
+      "loss": 0.503,
+      "step": 894
+    },
+    {
+      "epoch": 0.5445695162762397,
+      "grad_norm": 1.3846609592437744,
+      "learning_rate": 4.877217774277698e-05,
+      "loss": 0.6564,
+      "step": 895
+    },
+    {
+      "epoch": 0.5451779738363249,
+      "grad_norm": 1.4311436414718628,
+      "learning_rate": 4.876832140387082e-05,
+      "loss": 0.5094,
+      "step": 896
+    },
+    {
+      "epoch": 0.5457864313964101,
+      "grad_norm": 1.1872178316116333,
+      "learning_rate": 4.876445917141167e-05,
+      "loss": 0.4632,
+      "step": 897
+    },
+    {
+      "epoch": 0.5463948889564952,
+      "grad_norm": 1.4088075160980225,
+      "learning_rate": 4.8760591046357196e-05,
+      "loss": 0.4989,
+      "step": 898
+    },
+    {
+      "epoch": 0.5470033465165804,
+      "grad_norm": 1.2164874076843262,
+      "learning_rate": 4.875671702966653e-05,
+      "loss": 0.4258,
+      "step": 899
+    },
+    {
+      "epoch": 0.5476118040766657,
+      "grad_norm": 1.2215385437011719,
+      "learning_rate": 4.875283712230027e-05,
+      "loss": 0.4936,
+      "step": 900
+    },
+    {
+      "epoch": 0.5482202616367509,
+      "grad_norm": 1.4215302467346191,
+      "learning_rate": 4.874895132522047e-05,
+      "loss": 0.425,
+      "step": 901
+    },
+    {
+      "epoch": 0.5488287191968361,
+      "grad_norm": 1.227898359298706,
+      "learning_rate": 4.874505963939066e-05,
+      "loss": 0.4306,
+      "step": 902
+    },
+    {
+      "epoch": 0.5494371767569212,
+      "grad_norm": 1.2220474481582642,
+      "learning_rate": 4.874116206577578e-05,
+      "loss": 0.5173,
+      "step": 903
+    },
+    {
+      "epoch": 0.5500456343170064,
+      "grad_norm": 1.1359342336654663,
+      "learning_rate": 4.8737258605342304e-05,
+      "loss": 0.4496,
+      "step": 904
+    },
+    {
+      "epoch": 0.5506540918770916,
+      "grad_norm": 1.4966158866882324,
+      "learning_rate": 4.8733349259058105e-05,
+      "loss": 0.4165,
+      "step": 905
+    },
+    {
+      "epoch": 0.5512625494371768,
+      "grad_norm": 1.28777277469635,
+      "learning_rate": 4.872943402789255e-05,
+      "loss": 0.5104,
+      "step": 906
+    },
+    {
+      "epoch": 0.551871006997262,
+      "grad_norm": 1.30458664894104,
+      "learning_rate": 4.872551291281644e-05,
+      "loss": 0.4773,
+      "step": 907
+    },
+    {
+      "epoch": 0.5524794645573471,
+      "grad_norm": 1.116233229637146,
+      "learning_rate": 4.872158591480206e-05,
+      "loss": 0.4654,
+      "step": 908
+    },
+    {
+      "epoch": 0.5530879221174323,
+      "grad_norm": 1.341030240058899,
+      "learning_rate": 4.871765303482314e-05,
+      "loss": 0.4994,
+      "step": 909
+    },
+    {
+      "epoch": 0.5536963796775175,
+      "grad_norm": 1.6074213981628418,
+      "learning_rate": 4.871371427385486e-05,
+      "loss": 0.4937,
+      "step": 910
+    },
+    {
+      "epoch": 0.5543048372376027,
+      "grad_norm": 1.4934123754501343,
+      "learning_rate": 4.870976963287389e-05,
+      "loss": 0.4713,
+      "step": 911
+    },
+    {
+      "epoch": 0.5549132947976878,
+      "grad_norm": 1.2505860328674316,
+      "learning_rate": 4.8705819112858306e-05,
+      "loss": 0.5438,
+      "step": 912
+    },
+    {
+      "epoch": 0.555521752357773,
+      "grad_norm": 1.4453731775283813,
+      "learning_rate": 4.8701862714787704e-05,
+      "loss": 0.5272,
+      "step": 913
+    },
+    {
+      "epoch": 0.5561302099178582,
+      "grad_norm": 1.0299216508865356,
+      "learning_rate": 4.8697900439643087e-05,
+      "loss": 0.4406,
+      "step": 914
+    },
+    {
+      "epoch": 0.5567386674779434,
+      "grad_norm": 1.381816029548645,
+      "learning_rate": 4.869393228840693e-05,
+      "loss": 0.4839,
+      "step": 915
+    },
+    {
+      "epoch": 0.5573471250380286,
+      "grad_norm": 1.1093368530273438,
+      "learning_rate": 4.8689958262063186e-05,
+      "loss": 0.4567,
+      "step": 916
+    },
+    {
+      "epoch": 0.5579555825981137,
+      "grad_norm": 1.302557349205017,
+      "learning_rate": 4.8685978361597234e-05,
+      "loss": 0.4838,
+      "step": 917
+    },
+    {
+      "epoch": 0.5585640401581989,
+      "grad_norm": 1.2300167083740234,
+      "learning_rate": 4.868199258799593e-05,
+      "loss": 0.4985,
+      "step": 918
+    },
+    {
+      "epoch": 0.5591724977182841,
+      "grad_norm": 1.171228051185608,
+      "learning_rate": 4.867800094224758e-05,
+      "loss": 0.4577,
+      "step": 919
+    },
+    {
+      "epoch": 0.5597809552783694,
+      "grad_norm": 1.348331093788147,
+      "learning_rate": 4.867400342534194e-05,
+      "loss": 0.4892,
+      "step": 920
+    },
+    {
+      "epoch": 0.5603894128384546,
+      "grad_norm": 1.2311930656433105,
+      "learning_rate": 4.8670000038270236e-05,
+      "loss": 0.4981,
+      "step": 921
+    },
+    {
+      "epoch": 0.5609978703985397,
+      "grad_norm": 1.3021697998046875,
+      "learning_rate": 4.866599078202514e-05,
+      "loss": 0.5145,
+      "step": 922
+    },
+    {
+      "epoch": 0.5616063279586249,
+      "grad_norm": 1.426529049873352,
+      "learning_rate": 4.8661975657600765e-05,
+      "loss": 0.502,
+      "step": 923
+    },
+    {
+      "epoch": 0.5622147855187101,
+      "grad_norm": 1.096643090248108,
+      "learning_rate": 4.865795466599272e-05,
+      "loss": 0.4567,
+      "step": 924
+    },
+    {
+      "epoch": 0.5628232430787953,
+      "grad_norm": 1.1198285818099976,
+      "learning_rate": 4.8653927808198024e-05,
+      "loss": 0.4742,
+      "step": 925
+    },
+    {
+      "epoch": 0.5634317006388805,
+      "grad_norm": 1.3766636848449707,
+      "learning_rate": 4.8649895085215177e-05,
+      "loss": 0.5223,
+      "step": 926
+    },
+    {
+      "epoch": 0.5640401581989656,
+      "grad_norm": 1.1397812366485596,
+      "learning_rate": 4.8645856498044125e-05,
+      "loss": 0.4926,
+      "step": 927
+    },
+    {
+      "epoch": 0.5646486157590508,
+      "grad_norm": 1.2435989379882812,
+      "learning_rate": 4.8641812047686266e-05,
+      "loss": 0.44,
+      "step": 928
+    },
+    {
+      "epoch": 0.565257073319136,
+      "grad_norm": 1.5014331340789795,
+      "learning_rate": 4.863776173514446e-05,
+      "loss": 0.4282,
+      "step": 929
+    },
+    {
+      "epoch": 0.5658655308792212,
+      "grad_norm": 1.1607805490493774,
+      "learning_rate": 4.8633705561423014e-05,
+      "loss": 0.4856,
+      "step": 930
+    },
+    {
+      "epoch": 0.5664739884393064,
+      "grad_norm": 1.08901846408844,
+      "learning_rate": 4.862964352752769e-05,
+      "loss": 0.5092,
+      "step": 931
+    },
+    {
+      "epoch": 0.5670824459993915,
+      "grad_norm": 1.114229440689087,
+      "learning_rate": 4.862557563446571e-05,
+      "loss": 0.4652,
+      "step": 932
+    },
+    {
+      "epoch": 0.5676909035594767,
+      "grad_norm": 1.2063711881637573,
+      "learning_rate": 4.862150188324573e-05,
+      "loss": 0.4485,
+      "step": 933
+    },
+    {
+      "epoch": 0.5682993611195619,
+      "grad_norm": 1.2078471183776855,
+      "learning_rate": 4.861742227487788e-05,
+      "loss": 0.5443,
+      "step": 934
+    },
+    {
+      "epoch": 0.5689078186796471,
+      "grad_norm": 1.0555641651153564,
+      "learning_rate": 4.861333681037372e-05,
+      "loss": 0.4346,
+      "step": 935
+    },
+    {
+      "epoch": 0.5695162762397323,
+      "grad_norm": 0.9378579258918762,
+      "learning_rate": 4.8609245490746283e-05,
+      "loss": 0.4181,
+      "step": 936
+    },
+    {
+      "epoch": 0.5701247337998174,
+      "grad_norm": 1.4141374826431274,
+      "learning_rate": 4.8605148317010054e-05,
+      "loss": 0.4459,
+      "step": 937
+    },
+    {
+      "epoch": 0.5707331913599026,
+      "grad_norm": 1.106641173362732,
+      "learning_rate": 4.8601045290180946e-05,
+      "loss": 0.4694,
+      "step": 938
+    },
+    {
+      "epoch": 0.5713416489199878,
+      "grad_norm": 1.272030234336853,
+      "learning_rate": 4.8596936411276354e-05,
+      "loss": 0.5127,
+      "step": 939
+    },
+    {
+      "epoch": 0.571950106480073,
+      "grad_norm": 1.3205722570419312,
+      "learning_rate": 4.8592821681315096e-05,
+      "loss": 0.4937,
+      "step": 940
+    },
+    {
+      "epoch": 0.5725585640401581,
+      "grad_norm": 1.1618589162826538,
+      "learning_rate": 4.858870110131746e-05,
+      "loss": 0.5017,
+      "step": 941
+    },
+    {
+      "epoch": 0.5731670216002434,
+      "grad_norm": 1.2044800519943237,
+      "learning_rate": 4.858457467230517e-05,
+      "loss": 0.4831,
+      "step": 942
+    },
+    {
+      "epoch": 0.5737754791603286,
+      "grad_norm": 1.5071766376495361,
+      "learning_rate": 4.858044239530143e-05,
+      "loss": 0.5215,
+      "step": 943
+    },
+    {
+      "epoch": 0.5743839367204138,
+      "grad_norm": 1.1561444997787476,
+      "learning_rate": 4.857630427133084e-05,
+      "loss": 0.4897,
+      "step": 944
+    },
+    {
+      "epoch": 0.574992394280499,
+      "grad_norm": 1.1617779731750488,
+      "learning_rate": 4.857216030141951e-05,
+      "loss": 0.4417,
+      "step": 945
+    },
+    {
+      "epoch": 0.5756008518405841,
+      "grad_norm": 1.083703637123108,
+      "learning_rate": 4.8568010486594964e-05,
+      "loss": 0.4986,
+      "step": 946
+    },
+    {
+      "epoch": 0.5762093094006693,
+      "grad_norm": 1.3023382425308228,
+      "learning_rate": 4.8563854827886166e-05,
+      "loss": 0.5696,
+      "step": 947
+    },
+    {
+      "epoch": 0.5768177669607545,
+      "grad_norm": 1.446385383605957,
+      "learning_rate": 4.855969332632357e-05,
+      "loss": 0.49,
+      "step": 948
+    },
+    {
+      "epoch": 0.5774262245208397,
+      "grad_norm": 1.1349619626998901,
+      "learning_rate": 4.8555525982939046e-05,
+      "loss": 0.5179,
+      "step": 949
+    },
+    {
+      "epoch": 0.5780346820809249,
+      "grad_norm": 1.1926076412200928,
+      "learning_rate": 4.855135279876592e-05,
+      "loss": 0.4852,
+      "step": 950
+    },
+    {
+      "epoch": 0.57864313964101,
+      "grad_norm": 1.0169259309768677,
+      "learning_rate": 4.8547173774838975e-05,
+      "loss": 0.4002,
+      "step": 951
+    },
+    {
+      "epoch": 0.5792515972010952,
+      "grad_norm": 1.2183793783187866,
+      "learning_rate": 4.854298891219441e-05,
+      "loss": 0.5109,
+      "step": 952
+    },
+    {
+      "epoch": 0.5798600547611804,
+      "grad_norm": 1.2902944087982178,
+      "learning_rate": 4.853879821186993e-05,
+      "loss": 0.5235,
+      "step": 953
+    },
+    {
+      "epoch": 0.5804685123212656,
+      "grad_norm": 1.0406612157821655,
+      "learning_rate": 4.8534601674904635e-05,
+      "loss": 0.4147,
+      "step": 954
+    },
+    {
+      "epoch": 0.5810769698813508,
+      "grad_norm": 1.2372276782989502,
+      "learning_rate": 4.85303993023391e-05,
+      "loss": 0.4658,
+      "step": 955
+    },
+    {
+      "epoch": 0.5816854274414359,
+      "grad_norm": 1.1386033296585083,
+      "learning_rate": 4.852619109521533e-05,
+      "loss": 0.4555,
+      "step": 956
+    },
+    {
+      "epoch": 0.5822938850015211,
+      "grad_norm": 1.4579136371612549,
+      "learning_rate": 4.8521977054576783e-05,
+      "loss": 0.4944,
+      "step": 957
+    },
+    {
+      "epoch": 0.5829023425616063,
+      "grad_norm": 1.3394607305526733,
+      "learning_rate": 4.851775718146838e-05,
+      "loss": 0.5533,
+      "step": 958
+    },
+    {
+      "epoch": 0.5835108001216915,
+      "grad_norm": 1.263627290725708,
+      "learning_rate": 4.851353147693646e-05,
+      "loss": 0.5234,
+      "step": 959
+    },
+    {
+      "epoch": 0.5841192576817767,
+      "grad_norm": 1.30045747756958,
+      "learning_rate": 4.850929994202882e-05,
+      "loss": 0.5035,
+      "step": 960
+    },
+    {
+      "epoch": 0.5847277152418618,
+      "grad_norm": 1.3387147188186646,
+      "learning_rate": 4.8505062577794716e-05,
+      "loss": 0.4725,
+      "step": 961
+    },
+    {
+      "epoch": 0.585336172801947,
+      "grad_norm": 1.1547223329544067,
+      "learning_rate": 4.8500819385284835e-05,
+      "loss": 0.4985,
+      "step": 962
+    },
+    {
+      "epoch": 0.5859446303620323,
+      "grad_norm": 1.1327009201049805,
+      "learning_rate": 4.849657036555131e-05,
+      "loss": 0.4579,
+      "step": 963
+    },
+    {
+      "epoch": 0.5865530879221175,
+      "grad_norm": 0.9808380007743835,
+      "learning_rate": 4.849231551964771e-05,
+      "loss": 0.4366,
+      "step": 964
+    },
+    {
+      "epoch": 0.5871615454822026,
+      "grad_norm": 1.2232143878936768,
+      "learning_rate": 4.848805484862908e-05,
+      "loss": 0.4611,
+      "step": 965
+    },
+    {
+      "epoch": 0.5877700030422878,
+      "grad_norm": 1.2399910688400269,
+      "learning_rate": 4.8483788353551876e-05,
+      "loss": 0.5725,
+      "step": 966
+    },
+    {
+      "epoch": 0.588378460602373,
+      "grad_norm": 1.1158905029296875,
+      "learning_rate": 4.8479516035474003e-05,
+      "loss": 0.4809,
+      "step": 967
+    },
+    {
+      "epoch": 0.5889869181624582,
+      "grad_norm": 1.1464722156524658,
+      "learning_rate": 4.8475237895454833e-05,
+      "loss": 0.4469,
+      "step": 968
+    },
+    {
+      "epoch": 0.5895953757225434,
+      "grad_norm": 1.3894585371017456,
+      "learning_rate": 4.847095393455516e-05,
+      "loss": 0.4433,
+      "step": 969
+    },
+    {
+      "epoch": 0.5902038332826285,
+      "grad_norm": 1.2021290063858032,
+      "learning_rate": 4.846666415383724e-05,
+      "loss": 0.4745,
+      "step": 970
+    },
+    {
+      "epoch": 0.5908122908427137,
+      "grad_norm": 1.128505825996399,
+      "learning_rate": 4.846236855436473e-05,
+      "loss": 0.3727,
+      "step": 971
+    },
+    {
+      "epoch": 0.5914207484027989,
+      "grad_norm": 1.2462071180343628,
+      "learning_rate": 4.845806713720279e-05,
+      "loss": 0.5052,
+      "step": 972
+    },
+    {
+      "epoch": 0.5920292059628841,
+      "grad_norm": 1.1397188901901245,
+      "learning_rate": 4.845375990341798e-05,
+      "loss": 0.447,
+      "step": 973
+    },
+    {
+      "epoch": 0.5926376635229693,
+      "grad_norm": 1.1206566095352173,
+      "learning_rate": 4.844944685407831e-05,
+      "loss": 0.4872,
+      "step": 974
+    },
+    {
+      "epoch": 0.5932461210830544,
+      "grad_norm": 1.167952299118042,
+      "learning_rate": 4.8445127990253244e-05,
+      "loss": 0.489,
+      "step": 975
+    },
+    {
+      "epoch": 0.5938545786431396,
+      "grad_norm": 1.4353526830673218,
+      "learning_rate": 4.844080331301368e-05,
+      "loss": 0.5701,
+      "step": 976
+    },
+    {
+      "epoch": 0.5944630362032248,
+      "grad_norm": 1.1396512985229492,
+      "learning_rate": 4.843647282343195e-05,
+      "loss": 0.4253,
+      "step": 977
+    },
+    {
+      "epoch": 0.59507149376331,
+      "grad_norm": 1.1176103353500366,
+      "learning_rate": 4.843213652258185e-05,
+      "loss": 0.4273,
+      "step": 978
+    },
+    {
+      "epoch": 0.5956799513233952,
+      "grad_norm": 0.9949272274971008,
+      "learning_rate": 4.842779441153858e-05,
+      "loss": 0.4386,
+      "step": 979
+    },
+    {
+      "epoch": 0.5962884088834803,
+      "grad_norm": 1.0839688777923584,
+      "learning_rate": 4.842344649137882e-05,
+      "loss": 0.4394,
+      "step": 980
+    },
+    {
+      "epoch": 0.5968968664435655,
+      "grad_norm": 1.2163331508636475,
+      "learning_rate": 4.8419092763180673e-05,
+      "loss": 0.4894,
+      "step": 981
+    },
+    {
+      "epoch": 0.5975053240036508,
+      "grad_norm": 1.1652277708053589,
+      "learning_rate": 4.841473322802367e-05,
+      "loss": 0.4745,
+      "step": 982
+    },
+    {
+      "epoch": 0.598113781563736,
+      "grad_norm": 1.3939238786697388,
+      "learning_rate": 4.84103678869888e-05,
+      "loss": 0.5641,
+      "step": 983
+    },
+    {
+      "epoch": 0.5987222391238212,
+      "grad_norm": 1.0871104001998901,
+      "learning_rate": 4.840599674115849e-05,
+      "loss": 0.457,
+      "step": 984
+    },
+    {
+      "epoch": 0.5993306966839063,
+      "grad_norm": 1.193232536315918,
+      "learning_rate": 4.8401619791616595e-05,
+      "loss": 0.5823,
+      "step": 985
+    },
+    {
+      "epoch": 0.5999391542439915,
+      "grad_norm": 1.1120039224624634,
+      "learning_rate": 4.839723703944842e-05,
+      "loss": 0.464,
+      "step": 986
+    },
+    {
+      "epoch": 0.6005476118040767,
+      "grad_norm": 1.0750576257705688,
+      "learning_rate": 4.8392848485740706e-05,
+      "loss": 0.5022,
+      "step": 987
+    },
+    {
+      "epoch": 0.6011560693641619,
+      "grad_norm": 1.062077522277832,
+      "learning_rate": 4.838845413158162e-05,
+      "loss": 0.4077,
+      "step": 988
+    },
+    {
+      "epoch": 0.601764526924247,
+      "grad_norm": 1.1485848426818848,
+      "learning_rate": 4.83840539780608e-05,
+      "loss": 0.4517,
+      "step": 989
+    },
+    {
+      "epoch": 0.6023729844843322,
+      "grad_norm": 1.139553427696228,
+      "learning_rate": 4.837964802626929e-05,
+      "loss": 0.5155,
+      "step": 990
+    },
+    {
+      "epoch": 0.6029814420444174,
+      "grad_norm": 1.2427302598953247,
+      "learning_rate": 4.8375236277299575e-05,
+      "loss": 0.4747,
+      "step": 991
+    },
+    {
+      "epoch": 0.6035898996045026,
+      "grad_norm": 1.172774076461792,
+      "learning_rate": 4.837081873224559e-05,
+      "loss": 0.5702,
+      "step": 992
+    },
+    {
+      "epoch": 0.6041983571645878,
+      "grad_norm": 1.059998869895935,
+      "learning_rate": 4.83663953922027e-05,
+      "loss": 0.4608,
+      "step": 993
+    },
+    {
+      "epoch": 0.6048068147246729,
+      "grad_norm": 1.0970420837402344,
+      "learning_rate": 4.836196625826772e-05,
+      "loss": 0.4902,
+      "step": 994
+    },
+    {
+      "epoch": 0.6054152722847581,
+      "grad_norm": 1.263503909111023,
+      "learning_rate": 4.835753133153888e-05,
+      "loss": 0.4825,
+      "step": 995
+    },
+    {
+      "epoch": 0.6060237298448433,
+      "grad_norm": 1.0827324390411377,
+      "learning_rate": 4.8353090613115856e-05,
+      "loss": 0.4678,
+      "step": 996
+    },
+    {
+      "epoch": 0.6066321874049285,
+      "grad_norm": 1.2301872968673706,
+      "learning_rate": 4.8348644104099773e-05,
+      "loss": 0.4315,
+      "step": 997
+    },
+    {
+      "epoch": 0.6072406449650137,
+      "grad_norm": 1.1217997074127197,
+      "learning_rate": 4.834419180559317e-05,
+      "loss": 0.4979,
+      "step": 998
+    },
+    {
+      "epoch": 0.6078491025250988,
+      "grad_norm": 0.9536124467849731,
+      "learning_rate": 4.833973371870003e-05,
+      "loss": 0.4011,
+      "step": 999
+    },
+    {
+      "epoch": 0.608457560085184,
+      "grad_norm": 1.1605530977249146,
+      "learning_rate": 4.833526984452578e-05,
+      "loss": 0.5327,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6090660176452692,
+      "grad_norm": 1.1049379110336304,
+      "learning_rate": 4.833080018417726e-05,
+      "loss": 0.4454,
+      "step": 1001
+    },
+    {
+      "epoch": 0.6096744752053544,
+      "grad_norm": 1.0455185174942017,
+      "learning_rate": 4.8326324738762774e-05,
+      "loss": 0.4685,
+      "step": 1002
+    },
+    {
+      "epoch": 0.6102829327654397,
+      "grad_norm": 1.3146376609802246,
+      "learning_rate": 4.832184350939205e-05,
+      "loss": 0.465,
+      "step": 1003
+    },
+    {
+      "epoch": 0.6108913903255248,
+      "grad_norm": 1.5373934507369995,
+      "learning_rate": 4.831735649717623e-05,
+      "loss": 0.5181,
+      "step": 1004
+    },
+    {
+      "epoch": 0.61149984788561,
+      "grad_norm": 1.0241584777832031,
+      "learning_rate": 4.831286370322792e-05,
+      "loss": 0.4116,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6121083054456952,
+      "grad_norm": 1.1920700073242188,
+      "learning_rate": 4.830836512866113e-05,
+      "loss": 0.5441,
+      "step": 1006
+    },
+    {
+      "epoch": 0.6127167630057804,
+      "grad_norm": 1.1300125122070312,
+      "learning_rate": 4.8303860774591336e-05,
+      "loss": 0.4439,
+      "step": 1007
+    },
+    {
+      "epoch": 0.6133252205658656,
+      "grad_norm": 1.3509349822998047,
+      "learning_rate": 4.8299350642135424e-05,
+      "loss": 0.5142,
+      "step": 1008
+    },
+    {
+      "epoch": 0.6139336781259507,
+      "grad_norm": 1.2040926218032837,
+      "learning_rate": 4.8294834732411714e-05,
+      "loss": 0.5396,
+      "step": 1009
+    },
+    {
+      "epoch": 0.6145421356860359,
+      "grad_norm": 1.0937200784683228,
+      "learning_rate": 4.829031304653997e-05,
+      "loss": 0.4634,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6151505932461211,
+      "grad_norm": 1.4639860391616821,
+      "learning_rate": 4.8285785585641375e-05,
+      "loss": 0.4316,
+      "step": 1011
+    },
+    {
+      "epoch": 0.6157590508062063,
+      "grad_norm": 1.1813201904296875,
+      "learning_rate": 4.8281252350838557e-05,
+      "loss": 0.4838,
+      "step": 1012
+    },
+    {
+      "epoch": 0.6163675083662915,
+      "grad_norm": 1.0957978963851929,
+      "learning_rate": 4.827671334325556e-05,
+      "loss": 0.4231,
+      "step": 1013
+    },
+    {
+      "epoch": 0.6169759659263766,
+      "grad_norm": 1.061135172843933,
+      "learning_rate": 4.827216856401788e-05,
+      "loss": 0.4508,
+      "step": 1014
+    },
+    {
+      "epoch": 0.6175844234864618,
+      "grad_norm": 1.2017053365707397,
+      "learning_rate": 4.826761801425243e-05,
+      "loss": 0.4579,
+      "step": 1015
+    },
+    {
+      "epoch": 0.618192881046547,
+      "grad_norm": 1.263437271118164,
+      "learning_rate": 4.826306169508755e-05,
+      "loss": 0.511,
+      "step": 1016
+    },
+    {
+      "epoch": 0.6188013386066322,
+      "grad_norm": 1.6724522113800049,
+      "learning_rate": 4.825849960765303e-05,
+      "loss": 0.4669,
+      "step": 1017
+    },
+    {
+      "epoch": 0.6194097961667173,
+      "grad_norm": 1.3626164197921753,
+      "learning_rate": 4.825393175308006e-05,
+      "loss": 0.5212,
+      "step": 1018
+    },
+    {
+      "epoch": 0.6200182537268025,
+      "grad_norm": 1.1306309700012207,
+      "learning_rate": 4.824935813250129e-05,
+      "loss": 0.4786,
+      "step": 1019
+    },
+    {
+      "epoch": 0.6206267112868877,
+      "grad_norm": 1.1889963150024414,
+      "learning_rate": 4.824477874705079e-05,
+      "loss": 0.4653,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6212351688469729,
+      "grad_norm": 1.1174501180648804,
+      "learning_rate": 4.8240193597864044e-05,
+      "loss": 0.4382,
+      "step": 1021
+    },
+    {
+      "epoch": 0.6218436264070581,
+      "grad_norm": 1.1703723669052124,
+      "learning_rate": 4.8235602686077986e-05,
+      "loss": 0.4405,
+      "step": 1022
+    },
+    {
+      "epoch": 0.6224520839671432,
+      "grad_norm": 1.100953221321106,
+      "learning_rate": 4.823100601283097e-05,
+      "loss": 0.417,
+      "step": 1023
+    },
+    {
+      "epoch": 0.6230605415272285,
+      "grad_norm": 0.9949542880058289,
+      "learning_rate": 4.822640357926278e-05,
+      "loss": 0.4109,
+      "step": 1024
+    },
+    {
+      "epoch": 0.6236689990873137,
+      "grad_norm": 1.0636481046676636,
+      "learning_rate": 4.8221795386514625e-05,
+      "loss": 0.4349,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6242774566473989,
+      "grad_norm": 1.30801522731781,
+      "learning_rate": 4.821718143572914e-05,
+      "loss": 0.4903,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6248859142074841,
+      "grad_norm": 1.0893471240997314,
+      "learning_rate": 4.821256172805041e-05,
+      "loss": 0.4154,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6254943717675692,
+      "grad_norm": 1.0990405082702637,
+      "learning_rate": 4.820793626462391e-05,
+      "loss": 0.3986,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6261028293276544,
+      "grad_norm": 1.1979132890701294,
+      "learning_rate": 4.8203305046596584e-05,
+      "loss": 0.5231,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6267112868877396,
+      "grad_norm": 1.0785170793533325,
+      "learning_rate": 4.8198668075116754e-05,
+      "loss": 0.495,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6273197444478248,
+      "grad_norm": 1.0589004755020142,
+      "learning_rate": 4.819402535133422e-05,
+      "loss": 0.4546,
+      "step": 1031
+    },
+    {
+      "epoch": 0.62792820200791,
+      "grad_norm": 1.189632534980774,
+      "learning_rate": 4.818937687640016e-05,
+      "loss": 0.4908,
+      "step": 1032
+    },
+    {
+      "epoch": 0.6285366595679951,
+      "grad_norm": 1.1523140668869019,
+      "learning_rate": 4.818472265146722e-05,
+      "loss": 0.4496,
+      "step": 1033
+    },
+    {
+      "epoch": 0.6291451171280803,
+      "grad_norm": 1.1302261352539062,
+      "learning_rate": 4.818006267768945e-05,
+      "loss": 0.4783,
+      "step": 1034
+    },
+    {
+      "epoch": 0.6297535746881655,
+      "grad_norm": 1.0538748502731323,
+      "learning_rate": 4.817539695622234e-05,
+      "loss": 0.4308,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6303620322482507,
+      "grad_norm": 1.1539735794067383,
+      "learning_rate": 4.817072548822277e-05,
+      "loss": 0.4797,
+      "step": 1036
+    },
+    {
+      "epoch": 0.6309704898083359,
+      "grad_norm": 1.2327778339385986,
+      "learning_rate": 4.816604827484908e-05,
+      "loss": 0.4814,
+      "step": 1037
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 1.306960105895996,
+      "learning_rate": 4.816136531726104e-05,
+      "loss": 0.5386,
+      "step": 1038
+    },
+    {
+      "epoch": 0.6321874049285062,
+      "grad_norm": 1.0629810094833374,
+      "learning_rate": 4.815667661661981e-05,
+      "loss": 0.4891,
+      "step": 1039
+    },
+    {
+      "epoch": 0.6327958624885914,
+      "grad_norm": 1.113374948501587,
+      "learning_rate": 4.8151982174088e-05,
+      "loss": 0.4601,
+      "step": 1040
+    },
+    {
+      "epoch": 0.6334043200486766,
+      "grad_norm": 1.1330082416534424,
+      "learning_rate": 4.814728199082962e-05,
+      "loss": 0.4374,
+      "step": 1041
+    },
+    {
+      "epoch": 0.6340127776087617,
+      "grad_norm": 1.10567307472229,
+      "learning_rate": 4.8142576068010135e-05,
+      "loss": 0.4555,
+      "step": 1042
+    },
+    {
+      "epoch": 0.6346212351688469,
+      "grad_norm": 1.0396449565887451,
+      "learning_rate": 4.813786440679642e-05,
+      "loss": 0.4372,
+      "step": 1043
+    },
+    {
+      "epoch": 0.6352296927289321,
+      "grad_norm": 1.0404661893844604,
+      "learning_rate": 4.813314700835677e-05,
+      "loss": 0.4632,
+      "step": 1044
+    },
+    {
+      "epoch": 0.6358381502890174,
+      "grad_norm": 1.0946401357650757,
+      "learning_rate": 4.8128423873860894e-05,
+      "loss": 0.4613,
+      "step": 1045
+    },
+    {
+      "epoch": 0.6364466078491026,
+      "grad_norm": 1.223552942276001,
+      "learning_rate": 4.812369500447994e-05,
+      "loss": 0.475,
+      "step": 1046
+    },
+    {
+      "epoch": 0.6370550654091877,
+      "grad_norm": 1.0299303531646729,
+      "learning_rate": 4.8118960401386466e-05,
+      "loss": 0.4937,
+      "step": 1047
+    },
+    {
+      "epoch": 0.6376635229692729,
+      "grad_norm": 1.256249189376831,
+      "learning_rate": 4.811422006575446e-05,
+      "loss": 0.4967,
+      "step": 1048
+    },
+    {
+      "epoch": 0.6382719805293581,
+      "grad_norm": 1.0132943391799927,
+      "learning_rate": 4.810947399875933e-05,
+      "loss": 0.4349,
+      "step": 1049
+    },
+    {
+      "epoch": 0.6388804380894433,
+      "grad_norm": 1.097622036933899,
+      "learning_rate": 4.810472220157789e-05,
+      "loss": 0.4606,
+      "step": 1050
+    },
+    {
+      "epoch": 0.6394888956495285,
+      "grad_norm": 1.0819436311721802,
+      "learning_rate": 4.80999646753884e-05,
+      "loss": 0.4715,
+      "step": 1051
+    },
+    {
+      "epoch": 0.6400973532096136,
+      "grad_norm": 1.1682971715927124,
+      "learning_rate": 4.8095201421370515e-05,
+      "loss": 0.477,
+      "step": 1052
+    },
+    {
+      "epoch": 0.6407058107696988,
+      "grad_norm": 0.9915129542350769,
+      "learning_rate": 4.8090432440705344e-05,
+      "loss": 0.4397,
+      "step": 1053
+    },
+    {
+      "epoch": 0.641314268329784,
+      "grad_norm": 1.028359293937683,
+      "learning_rate": 4.8085657734575387e-05,
+      "loss": 0.4533,
+      "step": 1054
+    },
+    {
+      "epoch": 0.6419227258898692,
+      "grad_norm": 1.2199668884277344,
+      "learning_rate": 4.8080877304164564e-05,
+      "loss": 0.4632,
+      "step": 1055
+    },
+    {
+      "epoch": 0.6425311834499544,
+      "grad_norm": 1.3181172609329224,
+      "learning_rate": 4.807609115065823e-05,
+      "loss": 0.5162,
+      "step": 1056
+    },
+    {
+      "epoch": 0.6431396410100395,
+      "grad_norm": 1.1091372966766357,
+      "learning_rate": 4.8071299275243145e-05,
+      "loss": 0.4368,
+      "step": 1057
+    },
+    {
+      "epoch": 0.6437480985701247,
+      "grad_norm": 1.1425464153289795,
+      "learning_rate": 4.80665016791075e-05,
+      "loss": 0.4589,
+      "step": 1058
+    },
+    {
+      "epoch": 0.6443565561302099,
+      "grad_norm": 1.0934760570526123,
+      "learning_rate": 4.80616983634409e-05,
+      "loss": 0.4216,
+      "step": 1059
+    },
+    {
+      "epoch": 0.6449650136902951,
+      "grad_norm": 1.2468703985214233,
+      "learning_rate": 4.805688932943436e-05,
+      "loss": 0.4859,
+      "step": 1060
+    },
+    {
+      "epoch": 0.6455734712503803,
+      "grad_norm": 1.0578359365463257,
+      "learning_rate": 4.805207457828034e-05,
+      "loss": 0.4568,
+      "step": 1061
+    },
+    {
+      "epoch": 0.6461819288104654,
+      "grad_norm": 1.2633030414581299,
+      "learning_rate": 4.8047254111172665e-05,
+      "loss": 0.4284,
+      "step": 1062
+    },
+    {
+      "epoch": 0.6467903863705506,
+      "grad_norm": 1.122072458267212,
+      "learning_rate": 4.804242792930663e-05,
+      "loss": 0.4531,
+      "step": 1063
+    },
+    {
+      "epoch": 0.6473988439306358,
+      "grad_norm": 1.2295277118682861,
+      "learning_rate": 4.803759603387894e-05,
+      "loss": 0.5087,
+      "step": 1064
+    },
+    {
+      "epoch": 0.648007301490721,
+      "grad_norm": 1.3162295818328857,
+      "learning_rate": 4.803275842608767e-05,
+      "loss": 0.5562,
+      "step": 1065
+    },
+    {
+      "epoch": 0.6486157590508062,
+      "grad_norm": 1.1886372566223145,
+      "learning_rate": 4.802791510713237e-05,
+      "loss": 0.4722,
+      "step": 1066
+    },
+    {
+      "epoch": 0.6492242166108914,
+      "grad_norm": 1.1018664836883545,
+      "learning_rate": 4.802306607821398e-05,
+      "loss": 0.4603,
+      "step": 1067
+    },
+    {
+      "epoch": 0.6498326741709766,
+      "grad_norm": 1.2671383619308472,
+      "learning_rate": 4.8018211340534835e-05,
+      "loss": 0.506,
+      "step": 1068
+    },
+    {
+      "epoch": 0.6504411317310618,
+      "grad_norm": 1.1520271301269531,
+      "learning_rate": 4.8013350895298735e-05,
+      "loss": 0.465,
+      "step": 1069
+    },
+    {
+      "epoch": 0.651049589291147,
+      "grad_norm": 1.2675180435180664,
+      "learning_rate": 4.8008484743710854e-05,
+      "loss": 0.4874,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6516580468512321,
+      "grad_norm": 1.073280692100525,
+      "learning_rate": 4.80036128869778e-05,
+      "loss": 0.439,
+      "step": 1071
+    },
+    {
+      "epoch": 0.6522665044113173,
+      "grad_norm": 1.365029215812683,
+      "learning_rate": 4.7998735326307585e-05,
+      "loss": 0.3928,
+      "step": 1072
+    },
+    {
+      "epoch": 0.6528749619714025,
+      "grad_norm": 1.06315016746521,
+      "learning_rate": 4.799385206290965e-05,
+      "loss": 0.4725,
+      "step": 1073
+    },
+    {
+      "epoch": 0.6534834195314877,
+      "grad_norm": 1.0730098485946655,
+      "learning_rate": 4.798896309799483e-05,
+      "loss": 0.447,
+      "step": 1074
+    },
+    {
+      "epoch": 0.6540918770915729,
+      "grad_norm": 1.4145798683166504,
+      "learning_rate": 4.798406843277538e-05,
+      "loss": 0.5263,
+      "step": 1075
+    },
+    {
+      "epoch": 0.654700334651658,
+      "grad_norm": 1.3397396802902222,
+      "learning_rate": 4.7979168068465e-05,
+      "loss": 0.4561,
+      "step": 1076
+    },
+    {
+      "epoch": 0.6553087922117432,
+      "grad_norm": 1.0366169214248657,
+      "learning_rate": 4.7974262006278745e-05,
+      "loss": 0.3994,
+      "step": 1077
+    },
+    {
+      "epoch": 0.6559172497718284,
+      "grad_norm": 1.1666359901428223,
+      "learning_rate": 4.796935024743313e-05,
+      "loss": 0.4478,
+      "step": 1078
+    },
+    {
+      "epoch": 0.6565257073319136,
+      "grad_norm": 1.1900277137756348,
+      "learning_rate": 4.7964432793146065e-05,
+      "loss": 0.4719,
+      "step": 1079
+    },
+    {
+      "epoch": 0.6571341648919988,
+      "grad_norm": 1.5251827239990234,
+      "learning_rate": 4.795950964463687e-05,
+      "loss": 0.475,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6577426224520839,
+      "grad_norm": 1.298077940940857,
+      "learning_rate": 4.795458080312628e-05,
+      "loss": 0.4526,
+      "step": 1081
+    },
+    {
+      "epoch": 0.6583510800121691,
+      "grad_norm": 1.0009061098098755,
+      "learning_rate": 4.794964626983646e-05,
+      "loss": 0.3941,
+      "step": 1082
+    },
+    {
+      "epoch": 0.6589595375722543,
+      "grad_norm": 1.320255994796753,
+      "learning_rate": 4.794470604599093e-05,
+      "loss": 0.54,
+      "step": 1083
+    },
+    {
+      "epoch": 0.6595679951323395,
+      "grad_norm": 1.177507758140564,
+      "learning_rate": 4.79397601328147e-05,
+      "loss": 0.4851,
+      "step": 1084
+    },
+    {
+      "epoch": 0.6601764526924248,
+      "grad_norm": 1.2261627912521362,
+      "learning_rate": 4.793480853153412e-05,
+      "loss": 0.4907,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6607849102525098,
+      "grad_norm": 1.0971055030822754,
+      "learning_rate": 4.792985124337701e-05,
+      "loss": 0.4562,
+      "step": 1086
+    },
+    {
+      "epoch": 0.6613933678125951,
+      "grad_norm": 1.3122934103012085,
+      "learning_rate": 4.7924888269572545e-05,
+      "loss": 0.5057,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6620018253726803,
+      "grad_norm": 1.1012256145477295,
+      "learning_rate": 4.791991961135135e-05,
+      "loss": 0.4912,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6626102829327655,
+      "grad_norm": 1.0547481775283813,
+      "learning_rate": 4.791494526994544e-05,
+      "loss": 0.4898,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6632187404928507,
+      "grad_norm": 1.085845708847046,
+      "learning_rate": 4.790996524658824e-05,
+      "loss": 0.4328,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6638271980529358,
+      "grad_norm": 1.1783792972564697,
+      "learning_rate": 4.790497954251459e-05,
+      "loss": 0.4246,
+      "step": 1091
+    },
+    {
+      "epoch": 0.664435655613021,
+      "grad_norm": 1.1917533874511719,
+      "learning_rate": 4.789998815896075e-05,
+      "loss": 0.494,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6650441131731062,
+      "grad_norm": 1.0818932056427002,
+      "learning_rate": 4.7894991097164366e-05,
+      "loss": 0.5001,
+      "step": 1093
+    },
+    {
+      "epoch": 0.6656525707331914,
+      "grad_norm": 1.1833261251449585,
+      "learning_rate": 4.788998835836449e-05,
+      "loss": 0.5394,
+      "step": 1094
+    },
+    {
+      "epoch": 0.6662610282932765,
+      "grad_norm": 1.4886353015899658,
+      "learning_rate": 4.788497994380162e-05,
+      "loss": 0.3904,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6668694858533617,
+      "grad_norm": 1.6799591779708862,
+      "learning_rate": 4.78799658547176e-05,
+      "loss": 0.478,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6674779434134469,
+      "grad_norm": 1.1299816370010376,
+      "learning_rate": 4.787494609235575e-05,
+      "loss": 0.5295,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6680864009735321,
+      "grad_norm": 1.024692177772522,
+      "learning_rate": 4.786992065796072e-05,
+      "loss": 0.4684,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6686948585336173,
+      "grad_norm": 1.1255176067352295,
+      "learning_rate": 4.786488955277865e-05,
+      "loss": 0.455,
+      "step": 1099
+    },
+    {
+      "epoch": 0.6693033160937024,
+      "grad_norm": 1.1396479606628418,
+      "learning_rate": 4.7859852778057016e-05,
+      "loss": 0.5006,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6699117736537876,
+      "grad_norm": 1.0285197496414185,
+      "learning_rate": 4.7854810335044745e-05,
+      "loss": 0.4194,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6705202312138728,
+      "grad_norm": 0.9936107993125916,
+      "learning_rate": 4.7849762224992144e-05,
+      "loss": 0.4164,
+      "step": 1102
+    },
+    {
+      "epoch": 0.671128688773958,
+      "grad_norm": 1.3650137186050415,
+      "learning_rate": 4.784470844915093e-05,
+      "loss": 0.4367,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6717371463340432,
+      "grad_norm": 1.2089672088623047,
+      "learning_rate": 4.783964900877425e-05,
+      "loss": 0.491,
+      "step": 1104
+    },
+    {
+      "epoch": 0.6723456038941283,
+      "grad_norm": 1.2130366563796997,
+      "learning_rate": 4.78345839051166e-05,
+      "loss": 0.4589,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6729540614542135,
+      "grad_norm": 1.3927088975906372,
+      "learning_rate": 4.782951313943395e-05,
+      "loss": 0.5312,
+      "step": 1106
+    },
+    {
+      "epoch": 0.6735625190142988,
+      "grad_norm": 1.0126924514770508,
+      "learning_rate": 4.782443671298362e-05,
+      "loss": 0.4162,
+      "step": 1107
+    },
+    {
+      "epoch": 0.674170976574384,
+      "grad_norm": 1.0403350591659546,
+      "learning_rate": 4.781935462702435e-05,
+      "loss": 0.407,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6747794341344692,
+      "grad_norm": 1.1724456548690796,
+      "learning_rate": 4.7814266882816296e-05,
+      "loss": 0.4615,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6753878916945543,
+      "grad_norm": 1.178318977355957,
+      "learning_rate": 4.780917348162099e-05,
+      "loss": 0.5085,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6759963492546395,
+      "grad_norm": 1.088005542755127,
+      "learning_rate": 4.7804074424701406e-05,
+      "loss": 0.4744,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6766048068147247,
+      "grad_norm": 1.1180678606033325,
+      "learning_rate": 4.7798969713321874e-05,
+      "loss": 0.4814,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6772132643748099,
+      "grad_norm": 1.1669161319732666,
+      "learning_rate": 4.779385934874817e-05,
+      "loss": 0.4187,
+      "step": 1113
+    },
+    {
+      "epoch": 0.6778217219348951,
+      "grad_norm": 1.0521961450576782,
+      "learning_rate": 4.7788743332247437e-05,
+      "loss": 0.459,
+      "step": 1114
+    },
+    {
+      "epoch": 0.6784301794949802,
+      "grad_norm": 1.1008024215698242,
+      "learning_rate": 4.778362166508824e-05,
+      "loss": 0.5048,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6790386370550654,
+      "grad_norm": 1.0038844347000122,
+      "learning_rate": 4.777849434854054e-05,
+      "loss": 0.3917,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6796470946151506,
+      "grad_norm": 1.1488128900527954,
+      "learning_rate": 4.7773361383875697e-05,
+      "loss": 0.4351,
+      "step": 1117
+    },
+    {
+      "epoch": 0.6802555521752358,
+      "grad_norm": 1.1399554014205933,
+      "learning_rate": 4.7768222772366466e-05,
+      "loss": 0.4365,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6808640097353209,
+      "grad_norm": 1.3681772947311401,
+      "learning_rate": 4.776307851528702e-05,
+      "loss": 0.5406,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6814724672954061,
+      "grad_norm": 1.2079730033874512,
+      "learning_rate": 4.7757928613912914e-05,
+      "loss": 0.4459,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6820809248554913,
+      "grad_norm": 1.1010768413543701,
+      "learning_rate": 4.7752773069521104e-05,
+      "loss": 0.42,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6826893824155765,
+      "grad_norm": 1.095807433128357,
+      "learning_rate": 4.774761188338995e-05,
+      "loss": 0.4267,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6832978399756617,
+      "grad_norm": 1.1266393661499023,
+      "learning_rate": 4.774244505679923e-05,
+      "loss": 0.4663,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6839062975357468,
+      "grad_norm": 1.1465984582901,
+      "learning_rate": 4.773727259103008e-05,
+      "loss": 0.4362,
+      "step": 1124
+    },
+    {
+      "epoch": 0.684514755095832,
+      "grad_norm": 1.1885544061660767,
+      "learning_rate": 4.7732094487365065e-05,
+      "loss": 0.4638,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6851232126559172,
+      "grad_norm": 1.164554476737976,
+      "learning_rate": 4.772691074708814e-05,
+      "loss": 0.409,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6857316702160025,
+      "grad_norm": 1.0508933067321777,
+      "learning_rate": 4.7721721371484654e-05,
+      "loss": 0.4334,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6863401277760877,
+      "grad_norm": 1.0762606859207153,
+      "learning_rate": 4.771652636184135e-05,
+      "loss": 0.4735,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6869485853361728,
+      "grad_norm": 1.0883415937423706,
+      "learning_rate": 4.771132571944639e-05,
+      "loss": 0.4749,
+      "step": 1129
+    },
+    {
+      "epoch": 0.687557042896258,
+      "grad_norm": 1.061272144317627,
+      "learning_rate": 4.770611944558929e-05,
+      "loss": 0.4797,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6881655004563432,
+      "grad_norm": 1.0837888717651367,
+      "learning_rate": 4.770090754156102e-05,
+      "loss": 0.516,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6887739580164284,
+      "grad_norm": 1.197726845741272,
+      "learning_rate": 4.7695690008653896e-05,
+      "loss": 0.4487,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6893824155765136,
+      "grad_norm": 0.9252657294273376,
+      "learning_rate": 4.769046684816165e-05,
+      "loss": 0.3675,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6899908731365987,
+      "grad_norm": 1.079782247543335,
+      "learning_rate": 4.768523806137941e-05,
+      "loss": 0.4489,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6905993306966839,
+      "grad_norm": 1.2456952333450317,
+      "learning_rate": 4.76800036496037e-05,
+      "loss": 0.5507,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6912077882567691,
+      "grad_norm": 1.1496522426605225,
+      "learning_rate": 4.7674763614132434e-05,
+      "loss": 0.496,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6918162458168543,
+      "grad_norm": 1.1047946214675903,
+      "learning_rate": 4.766951795626493e-05,
+      "loss": 0.4656,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6924247033769395,
+      "grad_norm": 1.0433160066604614,
+      "learning_rate": 4.7664266677301874e-05,
+      "loss": 0.4197,
+      "step": 1138
+    },
+    {
+      "epoch": 0.6930331609370246,
+      "grad_norm": 1.2909495830535889,
+      "learning_rate": 4.7659009778545384e-05,
+      "loss": 0.4733,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6936416184971098,
+      "grad_norm": 1.110093355178833,
+      "learning_rate": 4.765374726129893e-05,
+      "loss": 0.4824,
+      "step": 1140
+    },
+    {
+      "epoch": 0.694250076057195,
+      "grad_norm": 1.285891056060791,
+      "learning_rate": 4.764847912686742e-05,
+      "loss": 0.485,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6948585336172802,
+      "grad_norm": 1.0111443996429443,
+      "learning_rate": 4.764320537655712e-05,
+      "loss": 0.4081,
+      "step": 1142
+    },
+    {
+      "epoch": 0.6954669911773653,
+      "grad_norm": 1.2149741649627686,
+      "learning_rate": 4.76379260116757e-05,
+      "loss": 0.5088,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6960754487374505,
+      "grad_norm": 1.2215123176574707,
+      "learning_rate": 4.7632641033532226e-05,
+      "loss": 0.4491,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6966839062975357,
+      "grad_norm": 1.1258493661880493,
+      "learning_rate": 4.762735044343715e-05,
+      "loss": 0.4473,
+      "step": 1145
+    },
+    {
+      "epoch": 0.697292363857621,
+      "grad_norm": 1.2326325178146362,
+      "learning_rate": 4.7622054242702316e-05,
+      "loss": 0.4819,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6979008214177062,
+      "grad_norm": 1.06281316280365,
+      "learning_rate": 4.761675243264097e-05,
+      "loss": 0.4228,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6985092789777912,
+      "grad_norm": 1.1261435747146606,
+      "learning_rate": 4.761144501456773e-05,
+      "loss": 0.4255,
+      "step": 1148
+    },
+    {
+      "epoch": 0.6991177365378765,
+      "grad_norm": 1.1780974864959717,
+      "learning_rate": 4.760613198979862e-05,
+      "loss": 0.4302,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6997261940979617,
+      "grad_norm": 1.119859218597412,
+      "learning_rate": 4.760081335965104e-05,
+      "loss": 0.4293,
+      "step": 1150
+    },
+    {
+      "epoch": 0.7003346516580469,
+      "grad_norm": 1.2447001934051514,
+      "learning_rate": 4.75954891254438e-05,
+      "loss": 0.4851,
+      "step": 1151
+    },
+    {
+      "epoch": 0.7009431092181321,
+      "grad_norm": 1.1865140199661255,
+      "learning_rate": 4.759015928849709e-05,
+      "loss": 0.4618,
+      "step": 1152
+    },
+    {
+      "epoch": 0.7015515667782172,
+      "grad_norm": 1.0534874200820923,
+      "learning_rate": 4.758482385013247e-05,
+      "loss": 0.4203,
+      "step": 1153
+    },
+    {
+      "epoch": 0.7021600243383024,
+      "grad_norm": 0.9607533812522888,
+      "learning_rate": 4.757948281167292e-05,
+      "loss": 0.368,
+      "step": 1154
+    },
+    {
+      "epoch": 0.7027684818983876,
+      "grad_norm": 1.1073296070098877,
+      "learning_rate": 4.75741361744428e-05,
+      "loss": 0.4375,
+      "step": 1155
+    },
+    {
+      "epoch": 0.7033769394584728,
+      "grad_norm": 1.2616397142410278,
+      "learning_rate": 4.756878393976783e-05,
+      "loss": 0.5214,
+      "step": 1156
+    },
+    {
+      "epoch": 0.703985397018558,
+      "grad_norm": 1.152910590171814,
+      "learning_rate": 4.756342610897517e-05,
+      "loss": 0.5083,
+      "step": 1157
+    },
+    {
+      "epoch": 0.7045938545786431,
+      "grad_norm": 1.2775417566299438,
+      "learning_rate": 4.7558062683393314e-05,
+      "loss": 0.5371,
+      "step": 1158
+    },
+    {
+      "epoch": 0.7052023121387283,
+      "grad_norm": 1.0952919721603394,
+      "learning_rate": 4.755269366435219e-05,
+      "loss": 0.4826,
+      "step": 1159
+    },
+    {
+      "epoch": 0.7058107696988135,
+      "grad_norm": 1.0538053512573242,
+      "learning_rate": 4.754731905318307e-05,
+      "loss": 0.4789,
+      "step": 1160
+    },
+    {
+      "epoch": 0.7064192272588987,
+      "grad_norm": 1.293050765991211,
+      "learning_rate": 4.754193885121865e-05,
+      "loss": 0.5337,
+      "step": 1161
+    },
+    {
+      "epoch": 0.7070276848189839,
+      "grad_norm": 1.0085028409957886,
+      "learning_rate": 4.7536553059792984e-05,
+      "loss": 0.4581,
+      "step": 1162
+    },
+    {
+      "epoch": 0.707636142379069,
+      "grad_norm": 1.063040852546692,
+      "learning_rate": 4.753116168024153e-05,
+      "loss": 0.5102,
+      "step": 1163
+    },
+    {
+      "epoch": 0.7082445999391542,
+      "grad_norm": 0.9961603283882141,
+      "learning_rate": 4.752576471390112e-05,
+      "loss": 0.4818,
+      "step": 1164
+    },
+    {
+      "epoch": 0.7088530574992394,
+      "grad_norm": 1.0528628826141357,
+      "learning_rate": 4.7520362162109986e-05,
+      "loss": 0.4396,
+      "step": 1165
+    },
+    {
+      "epoch": 0.7094615150593246,
+      "grad_norm": 1.2712112665176392,
+      "learning_rate": 4.751495402620774e-05,
+      "loss": 0.496,
+      "step": 1166
+    },
+    {
+      "epoch": 0.7100699726194099,
+      "grad_norm": 1.0716404914855957,
+      "learning_rate": 4.750954030753535e-05,
+      "loss": 0.4065,
+      "step": 1167
+    },
+    {
+      "epoch": 0.710678430179495,
+      "grad_norm": 1.2594822645187378,
+      "learning_rate": 4.7504121007435224e-05,
+      "loss": 0.5146,
+      "step": 1168
+    },
+    {
+      "epoch": 0.7112868877395802,
+      "grad_norm": 1.2636467218399048,
+      "learning_rate": 4.749869612725108e-05,
+      "loss": 0.5225,
+      "step": 1169
+    },
+    {
+      "epoch": 0.7118953452996654,
+      "grad_norm": 1.1434320211410522,
+      "learning_rate": 4.749326566832811e-05,
+      "loss": 0.4127,
+      "step": 1170
+    },
+    {
+      "epoch": 0.7125038028597506,
+      "grad_norm": 1.0728880167007446,
+      "learning_rate": 4.7487829632012816e-05,
+      "loss": 0.4583,
+      "step": 1171
+    },
+    {
+      "epoch": 0.7131122604198357,
+      "grad_norm": 1.2107247114181519,
+      "learning_rate": 4.7482388019653114e-05,
+      "loss": 0.4255,
+      "step": 1172
+    },
+    {
+      "epoch": 0.7137207179799209,
+      "grad_norm": 1.1021230220794678,
+      "learning_rate": 4.7476940832598295e-05,
+      "loss": 0.4829,
+      "step": 1173
+    },
+    {
+      "epoch": 0.7143291755400061,
+      "grad_norm": 1.2605928182601929,
+      "learning_rate": 4.747148807219902e-05,
+      "loss": 0.4636,
+      "step": 1174
+    },
+    {
+      "epoch": 0.7149376331000913,
+      "grad_norm": 1.0456511974334717,
+      "learning_rate": 4.746602973980738e-05,
+      "loss": 0.4248,
+      "step": 1175
+    },
+    {
+      "epoch": 0.7155460906601765,
+      "grad_norm": 1.1637083292007446,
+      "learning_rate": 4.746056583677678e-05,
+      "loss": 0.4544,
+      "step": 1176
+    },
+    {
+      "epoch": 0.7161545482202616,
+      "grad_norm": 1.1425434350967407,
+      "learning_rate": 4.745509636446207e-05,
+      "loss": 0.4557,
+      "step": 1177
+    },
+    {
+      "epoch": 0.7167630057803468,
+      "grad_norm": 1.1743497848510742,
+      "learning_rate": 4.744962132421943e-05,
+      "loss": 0.425,
+      "step": 1178
+    },
+    {
+      "epoch": 0.717371463340432,
+      "grad_norm": 1.1483440399169922,
+      "learning_rate": 4.744414071740644e-05,
+      "loss": 0.4819,
+      "step": 1179
+    },
+    {
+      "epoch": 0.7179799209005172,
+      "grad_norm": 1.0176396369934082,
+      "learning_rate": 4.7438654545382076e-05,
+      "loss": 0.4357,
+      "step": 1180
+    },
+    {
+      "epoch": 0.7185883784606024,
+      "grad_norm": 1.065658450126648,
+      "learning_rate": 4.743316280950667e-05,
+      "loss": 0.4108,
+      "step": 1181
+    },
+    {
+      "epoch": 0.7191968360206875,
+      "grad_norm": 0.9383304119110107,
+      "learning_rate": 4.7427665511141955e-05,
+      "loss": 0.3916,
+      "step": 1182
+    },
+    {
+      "epoch": 0.7198052935807727,
+      "grad_norm": 1.0549168586730957,
+      "learning_rate": 4.7422162651651026e-05,
+      "loss": 0.4165,
+      "step": 1183
+    },
+    {
+      "epoch": 0.7204137511408579,
+      "grad_norm": 1.241140365600586,
+      "learning_rate": 4.741665423239835e-05,
+      "loss": 0.4203,
+      "step": 1184
+    },
+    {
+      "epoch": 0.7210222087009431,
+      "grad_norm": 1.204376220703125,
+      "learning_rate": 4.741114025474981e-05,
+      "loss": 0.4497,
+      "step": 1185
+    },
+    {
+      "epoch": 0.7216306662610283,
+      "grad_norm": 1.0280909538269043,
+      "learning_rate": 4.7405620720072616e-05,
+      "loss": 0.4068,
+      "step": 1186
+    },
+    {
+      "epoch": 0.7222391238211134,
+      "grad_norm": 1.3876770734786987,
+      "learning_rate": 4.74000956297354e-05,
+      "loss": 0.4713,
+      "step": 1187
+    },
+    {
+      "epoch": 0.7228475813811986,
+      "grad_norm": 1.0588303804397583,
+      "learning_rate": 4.739456498510815e-05,
+      "loss": 0.4042,
+      "step": 1188
+    },
+    {
+      "epoch": 0.7234560389412839,
+      "grad_norm": 1.0324196815490723,
+      "learning_rate": 4.738902878756224e-05,
+      "loss": 0.469,
+      "step": 1189
+    },
+    {
+      "epoch": 0.7240644965013691,
+      "grad_norm": 0.9513891935348511,
+      "learning_rate": 4.738348703847041e-05,
+      "loss": 0.4032,
+      "step": 1190
+    },
+    {
+      "epoch": 0.7246729540614543,
+      "grad_norm": 1.0217005014419556,
+      "learning_rate": 4.737793973920678e-05,
+      "loss": 0.4112,
+      "step": 1191
+    },
+    {
+      "epoch": 0.7252814116215394,
+      "grad_norm": 1.1146938800811768,
+      "learning_rate": 4.737238689114686e-05,
+      "loss": 0.4397,
+      "step": 1192
+    },
+    {
+      "epoch": 0.7258898691816246,
+      "grad_norm": 1.017473816871643,
+      "learning_rate": 4.736682849566751e-05,
+      "loss": 0.4661,
+      "step": 1193
+    },
+    {
+      "epoch": 0.7264983267417098,
+      "grad_norm": 1.3287296295166016,
+      "learning_rate": 4.736126455414699e-05,
+      "loss": 0.4159,
+      "step": 1194
+    },
+    {
+      "epoch": 0.727106784301795,
+      "grad_norm": 1.1023236513137817,
+      "learning_rate": 4.7355695067964925e-05,
+      "loss": 0.425,
+      "step": 1195
+    },
+    {
+      "epoch": 0.7277152418618801,
+      "grad_norm": 1.1281861066818237,
+      "learning_rate": 4.735012003850232e-05,
+      "loss": 0.4161,
+      "step": 1196
+    },
+    {
+      "epoch": 0.7283236994219653,
+      "grad_norm": 1.031778335571289,
+      "learning_rate": 4.734453946714154e-05,
+      "loss": 0.4355,
+      "step": 1197
+    },
+    {
+      "epoch": 0.7289321569820505,
+      "grad_norm": 1.101798176765442,
+      "learning_rate": 4.733895335526633e-05,
+      "loss": 0.4299,
+      "step": 1198
+    },
+    {
+      "epoch": 0.7295406145421357,
+      "grad_norm": 1.27028489112854,
+      "learning_rate": 4.7333361704261834e-05,
+      "loss": 0.4291,
+      "step": 1199
+    },
+    {
+      "epoch": 0.7301490721022209,
+      "grad_norm": 1.2124416828155518,
+      "learning_rate": 4.732776451551453e-05,
+      "loss": 0.4873,
+      "step": 1200
+    },
+    {
+      "epoch": 0.730757529662306,
+      "grad_norm": 1.1454896926879883,
+      "learning_rate": 4.73221617904123e-05,
+      "loss": 0.4799,
+      "step": 1201
+    },
+    {
+      "epoch": 0.7313659872223912,
+      "grad_norm": 1.1555596590042114,
+      "learning_rate": 4.731655353034437e-05,
+      "loss": 0.465,
+      "step": 1202
+    },
+    {
+      "epoch": 0.7319744447824764,
+      "grad_norm": 1.1674630641937256,
+      "learning_rate": 4.7310939736701364e-05,
+      "loss": 0.4619,
+      "step": 1203
+    },
+    {
+      "epoch": 0.7325829023425616,
+      "grad_norm": 1.320896863937378,
+      "learning_rate": 4.7305320410875277e-05,
+      "loss": 0.4513,
+      "step": 1204
+    },
+    {
+      "epoch": 0.7331913599026468,
+      "grad_norm": 3.161492347717285,
+      "learning_rate": 4.7299695554259455e-05,
+      "loss": 0.512,
+      "step": 1205
+    },
+    {
+      "epoch": 0.7337998174627319,
+      "grad_norm": 1.1867998838424683,
+      "learning_rate": 4.729406516824864e-05,
+      "loss": 0.4125,
+      "step": 1206
+    },
+    {
+      "epoch": 0.7344082750228171,
+      "grad_norm": 1.1488641500473022,
+      "learning_rate": 4.7288429254238917e-05,
+      "loss": 0.4309,
+      "step": 1207
+    },
+    {
+      "epoch": 0.7350167325829023,
+      "grad_norm": 1.1211802959442139,
+      "learning_rate": 4.728278781362777e-05,
+      "loss": 0.4359,
+      "step": 1208
+    },
+    {
+      "epoch": 0.7356251901429876,
+      "grad_norm": 0.9918986558914185,
+      "learning_rate": 4.7277140847814025e-05,
+      "loss": 0.4088,
+      "step": 1209
+    },
+    {
+      "epoch": 0.7362336477030728,
+      "grad_norm": 1.1625863313674927,
+      "learning_rate": 4.7271488358197924e-05,
+      "loss": 0.424,
+      "step": 1210
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 1.432852029800415,
+      "learning_rate": 4.726583034618103e-05,
+      "loss": 0.4789,
+      "step": 1211
+    },
+    {
+      "epoch": 0.7374505628232431,
+      "grad_norm": 1.154141902923584,
+      "learning_rate": 4.7260166813166285e-05,
+      "loss": 0.4522,
+      "step": 1212
+    },
+    {
+      "epoch": 0.7380590203833283,
+      "grad_norm": 1.013265609741211,
+      "learning_rate": 4.7254497760558024e-05,
+      "loss": 0.3628,
+      "step": 1213
+    },
+    {
+      "epoch": 0.7386674779434135,
+      "grad_norm": 1.169954538345337,
+      "learning_rate": 4.724882318976194e-05,
+      "loss": 0.4394,
+      "step": 1214
+    },
+    {
+      "epoch": 0.7392759355034987,
+      "grad_norm": 1.1213183403015137,
+      "learning_rate": 4.724314310218507e-05,
+      "loss": 0.4598,
+      "step": 1215
+    },
+    {
+      "epoch": 0.7398843930635838,
+      "grad_norm": 1.1155545711517334,
+      "learning_rate": 4.723745749923586e-05,
+      "loss": 0.4315,
+      "step": 1216
+    },
+    {
+      "epoch": 0.740492850623669,
+      "grad_norm": 1.197703242301941,
+      "learning_rate": 4.723176638232408e-05,
+      "loss": 0.4351,
+      "step": 1217
+    },
+    {
+      "epoch": 0.7411013081837542,
+      "grad_norm": 1.4502240419387817,
+      "learning_rate": 4.7226069752860915e-05,
+      "loss": 0.47,
+      "step": 1218
+    },
+    {
+      "epoch": 0.7417097657438394,
+      "grad_norm": 1.4130961894989014,
+      "learning_rate": 4.722036761225888e-05,
+      "loss": 0.5292,
+      "step": 1219
+    },
+    {
+      "epoch": 0.7423182233039245,
+      "grad_norm": 1.2829434871673584,
+      "learning_rate": 4.7214659961931864e-05,
+      "loss": 0.4497,
+      "step": 1220
+    },
+    {
+      "epoch": 0.7429266808640097,
+      "grad_norm": 1.1895910501480103,
+      "learning_rate": 4.720894680329513e-05,
+      "loss": 0.3774,
+      "step": 1221
+    },
+    {
+      "epoch": 0.7435351384240949,
+      "grad_norm": 1.1315902471542358,
+      "learning_rate": 4.720322813776531e-05,
+      "loss": 0.4519,
+      "step": 1222
+    },
+    {
+      "epoch": 0.7441435959841801,
+      "grad_norm": 1.225305199623108,
+      "learning_rate": 4.7197503966760375e-05,
+      "loss": 0.5347,
+      "step": 1223
+    },
+    {
+      "epoch": 0.7447520535442653,
+      "grad_norm": 1.2546653747558594,
+      "learning_rate": 4.7191774291699695e-05,
+      "loss": 0.4597,
+      "step": 1224
+    },
+    {
+      "epoch": 0.7453605111043504,
+      "grad_norm": 0.880062997341156,
+      "learning_rate": 4.7186039114004e-05,
+      "loss": 0.3922,
+      "step": 1225
+    },
+    {
+      "epoch": 0.7459689686644356,
+      "grad_norm": 1.0395137071609497,
+      "learning_rate": 4.718029843509536e-05,
+      "loss": 0.4635,
+      "step": 1226
+    },
+    {
+      "epoch": 0.7465774262245208,
+      "grad_norm": 1.2031601667404175,
+      "learning_rate": 4.717455225639723e-05,
+      "loss": 0.3788,
+      "step": 1227
+    },
+    {
+      "epoch": 0.747185883784606,
+      "grad_norm": 0.9248659014701843,
+      "learning_rate": 4.716880057933441e-05,
+      "loss": 0.4171,
+      "step": 1228
+    },
+    {
+      "epoch": 0.7477943413446912,
+      "grad_norm": 0.9494261145591736,
+      "learning_rate": 4.71630434053331e-05,
+      "loss": 0.3695,
+      "step": 1229
+    },
+    {
+      "epoch": 0.7484027989047763,
+      "grad_norm": 1.4120838642120361,
+      "learning_rate": 4.715728073582082e-05,
+      "loss": 0.4758,
+      "step": 1230
+    },
+    {
+      "epoch": 0.7490112564648616,
+      "grad_norm": 1.4075794219970703,
+      "learning_rate": 4.715151257222649e-05,
+      "loss": 0.4857,
+      "step": 1231
+    },
+    {
+      "epoch": 0.7496197140249468,
+      "grad_norm": 1.2198188304901123,
+      "learning_rate": 4.7145738915980354e-05,
+      "loss": 0.4428,
+      "step": 1232
+    },
+    {
+      "epoch": 0.750228171585032,
+      "grad_norm": 1.3420535326004028,
+      "learning_rate": 4.7139959768514044e-05,
+      "loss": 0.4027,
+      "step": 1233
+    },
+    {
+      "epoch": 0.7508366291451172,
+      "grad_norm": 1.2325292825698853,
+      "learning_rate": 4.713417513126055e-05,
+      "loss": 0.4469,
+      "step": 1234
+    },
+    {
+      "epoch": 0.7514450867052023,
+      "grad_norm": 1.3523552417755127,
+      "learning_rate": 4.712838500565423e-05,
+      "loss": 0.4269,
+      "step": 1235
+    },
+    {
+      "epoch": 0.7520535442652875,
+      "grad_norm": 1.099088191986084,
+      "learning_rate": 4.712258939313078e-05,
+      "loss": 0.4336,
+      "step": 1236
+    },
+    {
+      "epoch": 0.7526620018253727,
+      "grad_norm": 1.248347520828247,
+      "learning_rate": 4.7116788295127275e-05,
+      "loss": 0.3627,
+      "step": 1237
+    },
+    {
+      "epoch": 0.7532704593854579,
+      "grad_norm": 1.3930177688598633,
+      "learning_rate": 4.711098171308214e-05,
+      "loss": 0.4765,
+      "step": 1238
+    },
+    {
+      "epoch": 0.7538789169455431,
+      "grad_norm": 0.9983875155448914,
+      "learning_rate": 4.7105169648435176e-05,
+      "loss": 0.4179,
+      "step": 1239
+    },
+    {
+      "epoch": 0.7544873745056282,
+      "grad_norm": 1.0693212747573853,
+      "learning_rate": 4.7099352102627536e-05,
+      "loss": 0.4764,
+      "step": 1240
+    },
+    {
+      "epoch": 0.7550958320657134,
+      "grad_norm": 1.1990723609924316,
+      "learning_rate": 4.7093529077101714e-05,
+      "loss": 0.4924,
+      "step": 1241
+    },
+    {
+      "epoch": 0.7557042896257986,
+      "grad_norm": 1.0372933149337769,
+      "learning_rate": 4.7087700573301585e-05,
+      "loss": 0.4129,
+      "step": 1242
+    },
+    {
+      "epoch": 0.7563127471858838,
+      "grad_norm": 1.0057023763656616,
+      "learning_rate": 4.7081866592672376e-05,
+      "loss": 0.4319,
+      "step": 1243
+    },
+    {
+      "epoch": 0.756921204745969,
+      "grad_norm": 1.0604325532913208,
+      "learning_rate": 4.7076027136660663e-05,
+      "loss": 0.4533,
+      "step": 1244
+    },
+    {
+      "epoch": 0.7575296623060541,
+      "grad_norm": 1.1443201303482056,
+      "learning_rate": 4.70701822067144e-05,
+      "loss": 0.4355,
+      "step": 1245
+    },
+    {
+      "epoch": 0.7581381198661393,
+      "grad_norm": 1.1489462852478027,
+      "learning_rate": 4.706433180428288e-05,
+      "loss": 0.4098,
+      "step": 1246
+    },
+    {
+      "epoch": 0.7587465774262245,
+      "grad_norm": 1.0587549209594727,
+      "learning_rate": 4.705847593081676e-05,
+      "loss": 0.3956,
+      "step": 1247
+    },
+    {
+      "epoch": 0.7593550349863097,
+      "grad_norm": 1.0858579874038696,
+      "learning_rate": 4.705261458776805e-05,
+      "loss": 0.4166,
+      "step": 1248
+    },
+    {
+      "epoch": 0.7599634925463948,
+      "grad_norm": 1.7913252115249634,
+      "learning_rate": 4.704674777659012e-05,
+      "loss": 0.4482,
+      "step": 1249
+    },
+    {
+      "epoch": 0.76057195010648,
+      "grad_norm": 1.3128273487091064,
+      "learning_rate": 4.70408754987377e-05,
+      "loss": 0.4596,
+      "step": 1250
+    },
+    {
+      "epoch": 0.7611804076665653,
+      "grad_norm": 1.0580440759658813,
+      "learning_rate": 4.703499775566686e-05,
+      "loss": 0.4517,
+      "step": 1251
+    },
+    {
+      "epoch": 0.7617888652266505,
+      "grad_norm": 1.076398491859436,
+      "learning_rate": 4.702911454883504e-05,
+      "loss": 0.3933,
+      "step": 1252
+    },
+    {
+      "epoch": 0.7623973227867357,
+      "grad_norm": 1.267801284790039,
+      "learning_rate": 4.702322587970104e-05,
+      "loss": 0.4953,
+      "step": 1253
+    },
+    {
+      "epoch": 0.7630057803468208,
+      "grad_norm": 0.9255053400993347,
+      "learning_rate": 4.701733174972498e-05,
+      "loss": 0.3944,
+      "step": 1254
+    },
+    {
+      "epoch": 0.763614237906906,
+      "grad_norm": 0.9950581192970276,
+      "learning_rate": 4.7011432160368385e-05,
+      "loss": 0.4811,
+      "step": 1255
+    },
+    {
+      "epoch": 0.7642226954669912,
+      "grad_norm": 1.2969601154327393,
+      "learning_rate": 4.7005527113094094e-05,
+      "loss": 0.4454,
+      "step": 1256
+    },
+    {
+      "epoch": 0.7648311530270764,
+      "grad_norm": 1.045771837234497,
+      "learning_rate": 4.699961660936631e-05,
+      "loss": 0.4248,
+      "step": 1257
+    },
+    {
+      "epoch": 0.7654396105871616,
+      "grad_norm": 1.1103192567825317,
+      "learning_rate": 4.69937006506506e-05,
+      "loss": 0.402,
+      "step": 1258
+    },
+    {
+      "epoch": 0.7660480681472467,
+      "grad_norm": 1.0572874546051025,
+      "learning_rate": 4.698777923841386e-05,
+      "loss": 0.4319,
+      "step": 1259
+    },
+    {
+      "epoch": 0.7666565257073319,
+      "grad_norm": 1.1772733926773071,
+      "learning_rate": 4.6981852374124384e-05,
+      "loss": 0.4311,
+      "step": 1260
+    },
+    {
+      "epoch": 0.7672649832674171,
+      "grad_norm": 1.0701271295547485,
+      "learning_rate": 4.697592005925176e-05,
+      "loss": 0.4412,
+      "step": 1261
+    },
+    {
+      "epoch": 0.7678734408275023,
+      "grad_norm": 1.0132157802581787,
+      "learning_rate": 4.696998229526696e-05,
+      "loss": 0.3985,
+      "step": 1262
+    },
+    {
+      "epoch": 0.7684818983875875,
+      "grad_norm": 1.0832394361495972,
+      "learning_rate": 4.696403908364231e-05,
+      "loss": 0.4161,
+      "step": 1263
+    },
+    {
+      "epoch": 0.7690903559476726,
+      "grad_norm": 1.0529204607009888,
+      "learning_rate": 4.6958090425851465e-05,
+      "loss": 0.5481,
+      "step": 1264
+    },
+    {
+      "epoch": 0.7696988135077578,
+      "grad_norm": 1.1204801797866821,
+      "learning_rate": 4.6952136323369463e-05,
+      "loss": 0.4214,
+      "step": 1265
+    },
+    {
+      "epoch": 0.770307271067843,
+      "grad_norm": 1.0171352624893188,
+      "learning_rate": 4.6946176777672654e-05,
+      "loss": 0.4146,
+      "step": 1266
+    },
+    {
+      "epoch": 0.7709157286279282,
+      "grad_norm": 0.9407910704612732,
+      "learning_rate": 4.694021179023877e-05,
+      "loss": 0.4224,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7715241861880134,
+      "grad_norm": 1.1776046752929688,
+      "learning_rate": 4.6934241362546874e-05,
+      "loss": 0.5192,
+      "step": 1268
+    },
+    {
+      "epoch": 0.7721326437480985,
+      "grad_norm": 1.0098623037338257,
+      "learning_rate": 4.692826549607738e-05,
+      "loss": 0.4352,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7727411013081837,
+      "grad_norm": 1.0229098796844482,
+      "learning_rate": 4.6922284192312074e-05,
+      "loss": 0.4095,
+      "step": 1270
+    },
+    {
+      "epoch": 0.773349558868269,
+      "grad_norm": 1.0688673257827759,
+      "learning_rate": 4.691629745273404e-05,
+      "loss": 0.4093,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7739580164283542,
+      "grad_norm": 1.7956994771957397,
+      "learning_rate": 4.691030527882776e-05,
+      "loss": 0.4594,
+      "step": 1272
+    },
+    {
+      "epoch": 0.7745664739884393,
+      "grad_norm": 1.1307945251464844,
+      "learning_rate": 4.690430767207903e-05,
+      "loss": 0.4544,
+      "step": 1273
+    },
+    {
+      "epoch": 0.7751749315485245,
+      "grad_norm": 0.9435398578643799,
+      "learning_rate": 4.689830463397502e-05,
+      "loss": 0.3981,
+      "step": 1274
+    },
+    {
+      "epoch": 0.7757833891086097,
+      "grad_norm": 1.1967155933380127,
+      "learning_rate": 4.689229616600422e-05,
+      "loss": 0.4067,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7763918466686949,
+      "grad_norm": 1.2077351808547974,
+      "learning_rate": 4.68862822696565e-05,
+      "loss": 0.4763,
+      "step": 1276
+    },
+    {
+      "epoch": 0.7770003042287801,
+      "grad_norm": 1.1517752408981323,
+      "learning_rate": 4.688026294642303e-05,
+      "loss": 0.4822,
+      "step": 1277
+    },
+    {
+      "epoch": 0.7776087617888652,
+      "grad_norm": 1.3446024656295776,
+      "learning_rate": 4.687423819779637e-05,
+      "loss": 0.3934,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7782172193489504,
+      "grad_norm": 0.9664615988731384,
+      "learning_rate": 4.6868208025270396e-05,
+      "loss": 0.3821,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7788256769090356,
+      "grad_norm": 1.1611120700836182,
+      "learning_rate": 4.6862172430340344e-05,
+      "loss": 0.4274,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7794341344691208,
+      "grad_norm": 1.4211469888687134,
+      "learning_rate": 4.6856131414502795e-05,
+      "loss": 0.5036,
+      "step": 1281
+    },
+    {
+      "epoch": 0.780042592029206,
+      "grad_norm": 1.0818086862564087,
+      "learning_rate": 4.685008497925566e-05,
+      "loss": 0.4481,
+      "step": 1282
+    },
+    {
+      "epoch": 0.7806510495892911,
+      "grad_norm": 1.3467516899108887,
+      "learning_rate": 4.6844033126098206e-05,
+      "loss": 0.4211,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7812595071493763,
+      "grad_norm": 1.0704116821289062,
+      "learning_rate": 4.683797585653104e-05,
+      "loss": 0.4344,
+      "step": 1284
+    },
+    {
+      "epoch": 0.7818679647094615,
+      "grad_norm": 1.2070298194885254,
+      "learning_rate": 4.683191317205612e-05,
+      "loss": 0.4463,
+      "step": 1285
+    },
+    {
+      "epoch": 0.7824764222695467,
+      "grad_norm": 1.0552793741226196,
+      "learning_rate": 4.682584507417672e-05,
+      "loss": 0.4259,
+      "step": 1286
+    },
+    {
+      "epoch": 0.7830848798296319,
+      "grad_norm": 1.0915594100952148,
+      "learning_rate": 4.6819771564397496e-05,
+      "loss": 0.4085,
+      "step": 1287
+    },
+    {
+      "epoch": 0.783693337389717,
+      "grad_norm": 1.2251918315887451,
+      "learning_rate": 4.681369264422441e-05,
+      "loss": 0.4623,
+      "step": 1288
+    },
+    {
+      "epoch": 0.7843017949498022,
+      "grad_norm": 1.0790926218032837,
+      "learning_rate": 4.68076083151648e-05,
+      "loss": 0.454,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7849102525098874,
+      "grad_norm": 1.0661593675613403,
+      "learning_rate": 4.680151857872731e-05,
+      "loss": 0.406,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7855187100699726,
+      "grad_norm": 0.9886572957038879,
+      "learning_rate": 4.6795423436421934e-05,
+      "loss": 0.4058,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7861271676300579,
+      "grad_norm": 1.0567951202392578,
+      "learning_rate": 4.678932288976004e-05,
+      "loss": 0.4134,
+      "step": 1292
+    },
+    {
+      "epoch": 0.786735625190143,
+      "grad_norm": 1.0652602910995483,
+      "learning_rate": 4.678321694025428e-05,
+      "loss": 0.4439,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7873440827502282,
+      "grad_norm": 1.048807144165039,
+      "learning_rate": 4.6777105589418695e-05,
+      "loss": 0.3748,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7879525403103134,
+      "grad_norm": 1.044026494026184,
+      "learning_rate": 4.6770988838768634e-05,
+      "loss": 0.3903,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7885609978703986,
+      "grad_norm": 0.9846064448356628,
+      "learning_rate": 4.676486668982081e-05,
+      "loss": 0.4294,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7891694554304837,
+      "grad_norm": 1.2005590200424194,
+      "learning_rate": 4.675873914409324e-05,
+      "loss": 0.4848,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7897779129905689,
+      "grad_norm": 1.192406177520752,
+      "learning_rate": 4.6752606203105314e-05,
+      "loss": 0.3836,
+      "step": 1298
+    },
+    {
+      "epoch": 0.7903863705506541,
+      "grad_norm": 0.9216708540916443,
+      "learning_rate": 4.6746467868377744e-05,
+      "loss": 0.419,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7909948281107393,
+      "grad_norm": 1.1610586643218994,
+      "learning_rate": 4.674032414143258e-05,
+      "loss": 0.4027,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7916032856708245,
+      "grad_norm": 1.142449975013733,
+      "learning_rate": 4.673417502379321e-05,
+      "loss": 0.4675,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7922117432309096,
+      "grad_norm": 1.1200268268585205,
+      "learning_rate": 4.672802051698436e-05,
+      "loss": 0.4913,
+      "step": 1302
+    },
+    {
+      "epoch": 0.7928202007909948,
+      "grad_norm": 1.5124874114990234,
+      "learning_rate": 4.672186062253209e-05,
+      "loss": 0.5245,
+      "step": 1303
+    },
+    {
+      "epoch": 0.79342865835108,
+      "grad_norm": 1.1234627962112427,
+      "learning_rate": 4.671569534196379e-05,
+      "loss": 0.4452,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7940371159111652,
+      "grad_norm": 1.069880723953247,
+      "learning_rate": 4.6709524676808215e-05,
+      "loss": 0.4368,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7946455734712504,
+      "grad_norm": 1.0936565399169922,
+      "learning_rate": 4.670334862859541e-05,
+      "loss": 0.4527,
+      "step": 1306
+    },
+    {
+      "epoch": 0.7952540310313355,
+      "grad_norm": 1.0277516841888428,
+      "learning_rate": 4.669716719885679e-05,
+      "loss": 0.4186,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7958624885914207,
+      "grad_norm": 0.9716640114784241,
+      "learning_rate": 4.6690980389125075e-05,
+      "loss": 0.3765,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7964709461515059,
+      "grad_norm": 1.0334478616714478,
+      "learning_rate": 4.668478820093436e-05,
+      "loss": 0.416,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7970794037115911,
+      "grad_norm": 1.0650469064712524,
+      "learning_rate": 4.667859063582003e-05,
+      "loss": 0.428,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7976878612716763,
+      "grad_norm": 1.0338752269744873,
+      "learning_rate": 4.667238769531883e-05,
+      "loss": 0.4597,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7982963188317614,
+      "grad_norm": 1.0277113914489746,
+      "learning_rate": 4.666617938096884e-05,
+      "loss": 0.4062,
+      "step": 1312
+    },
+    {
+      "epoch": 0.7989047763918466,
+      "grad_norm": 1.0234774351119995,
+      "learning_rate": 4.6659965694309446e-05,
+      "loss": 0.3839,
+      "step": 1313
+    },
+    {
+      "epoch": 0.7995132339519319,
+      "grad_norm": 1.0939502716064453,
+      "learning_rate": 4.66537466368814e-05,
+      "loss": 0.4286,
+      "step": 1314
+    },
+    {
+      "epoch": 0.8001216915120171,
+      "grad_norm": 1.0294753313064575,
+      "learning_rate": 4.664752221022676e-05,
+      "loss": 0.3898,
+      "step": 1315
+    },
+    {
+      "epoch": 0.8007301490721023,
+      "grad_norm": 1.0266523361206055,
+      "learning_rate": 4.6641292415888916e-05,
+      "loss": 0.3917,
+      "step": 1316
+    },
+    {
+      "epoch": 0.8013386066321874,
+      "grad_norm": 1.0287213325500488,
+      "learning_rate": 4.6635057255412606e-05,
+      "loss": 0.4362,
+      "step": 1317
+    },
+    {
+      "epoch": 0.8019470641922726,
+      "grad_norm": 1.0150967836380005,
+      "learning_rate": 4.662881673034389e-05,
+      "loss": 0.3872,
+      "step": 1318
+    },
+    {
+      "epoch": 0.8025555217523578,
+      "grad_norm": 1.106278896331787,
+      "learning_rate": 4.662257084223017e-05,
+      "loss": 0.4082,
+      "step": 1319
+    },
+    {
+      "epoch": 0.803163979312443,
+      "grad_norm": 0.9770355224609375,
+      "learning_rate": 4.661631959262015e-05,
+      "loss": 0.4392,
+      "step": 1320
+    },
+    {
+      "epoch": 0.8037724368725282,
+      "grad_norm": 1.2682089805603027,
+      "learning_rate": 4.661006298306388e-05,
+      "loss": 0.4548,
+      "step": 1321
+    },
+    {
+      "epoch": 0.8043808944326133,
+      "grad_norm": 1.000850796699524,
+      "learning_rate": 4.660380101511275e-05,
+      "loss": 0.374,
+      "step": 1322
+    },
+    {
+      "epoch": 0.8049893519926985,
+      "grad_norm": 0.9677361249923706,
+      "learning_rate": 4.659753369031945e-05,
+      "loss": 0.3699,
+      "step": 1323
+    },
+    {
+      "epoch": 0.8055978095527837,
+      "grad_norm": 1.096622109413147,
+      "learning_rate": 4.659126101023802e-05,
+      "loss": 0.4508,
+      "step": 1324
+    },
+    {
+      "epoch": 0.8062062671128689,
+      "grad_norm": 0.9467645883560181,
+      "learning_rate": 4.658498297642384e-05,
+      "loss": 0.3777,
+      "step": 1325
+    },
+    {
+      "epoch": 0.806814724672954,
+      "grad_norm": 0.9743712544441223,
+      "learning_rate": 4.6578699590433585e-05,
+      "loss": 0.3962,
+      "step": 1326
+    },
+    {
+      "epoch": 0.8074231822330392,
+      "grad_norm": 1.0897773504257202,
+      "learning_rate": 4.657241085382527e-05,
+      "loss": 0.4304,
+      "step": 1327
+    },
+    {
+      "epoch": 0.8080316397931244,
+      "grad_norm": 1.0162287950515747,
+      "learning_rate": 4.6566116768158254e-05,
+      "loss": 0.4047,
+      "step": 1328
+    },
+    {
+      "epoch": 0.8086400973532096,
+      "grad_norm": 0.9446073770523071,
+      "learning_rate": 4.65598173349932e-05,
+      "loss": 0.3908,
+      "step": 1329
+    },
+    {
+      "epoch": 0.8092485549132948,
+      "grad_norm": 1.0082128047943115,
+      "learning_rate": 4.655351255589209e-05,
+      "loss": 0.3977,
+      "step": 1330
+    },
+    {
+      "epoch": 0.8098570124733799,
+      "grad_norm": 1.08930242061615,
+      "learning_rate": 4.6547202432418274e-05,
+      "loss": 0.4132,
+      "step": 1331
+    },
+    {
+      "epoch": 0.8104654700334651,
+      "grad_norm": 1.1194206476211548,
+      "learning_rate": 4.654088696613638e-05,
+      "loss": 0.3912,
+      "step": 1332
+    },
+    {
+      "epoch": 0.8110739275935503,
+      "grad_norm": 1.0496968030929565,
+      "learning_rate": 4.6534566158612395e-05,
+      "loss": 0.3896,
+      "step": 1333
+    },
+    {
+      "epoch": 0.8116823851536356,
+      "grad_norm": 1.0310301780700684,
+      "learning_rate": 4.65282400114136e-05,
+      "loss": 0.4467,
+      "step": 1334
+    },
+    {
+      "epoch": 0.8122908427137208,
+      "grad_norm": 1.234102487564087,
+      "learning_rate": 4.6521908526108624e-05,
+      "loss": 0.403,
+      "step": 1335
+    },
+    {
+      "epoch": 0.8128993002738059,
+      "grad_norm": 1.0917807817459106,
+      "learning_rate": 4.6515571704267414e-05,
+      "loss": 0.4464,
+      "step": 1336
+    },
+    {
+      "epoch": 0.8135077578338911,
+      "grad_norm": 1.1464637517929077,
+      "learning_rate": 4.650922954746123e-05,
+      "loss": 0.4617,
+      "step": 1337
+    },
+    {
+      "epoch": 0.8141162153939763,
+      "grad_norm": 1.1391267776489258,
+      "learning_rate": 4.6502882057262675e-05,
+      "loss": 0.4209,
+      "step": 1338
+    },
+    {
+      "epoch": 0.8147246729540615,
+      "grad_norm": 1.1518268585205078,
+      "learning_rate": 4.6496529235245644e-05,
+      "loss": 0.4591,
+      "step": 1339
+    },
+    {
+      "epoch": 0.8153331305141467,
+      "grad_norm": 1.207227349281311,
+      "learning_rate": 4.649017108298539e-05,
+      "loss": 0.4252,
+      "step": 1340
+    },
+    {
+      "epoch": 0.8159415880742318,
+      "grad_norm": 1.172620415687561,
+      "learning_rate": 4.648380760205846e-05,
+      "loss": 0.4601,
+      "step": 1341
+    },
+    {
+      "epoch": 0.816550045634317,
+      "grad_norm": 1.1918513774871826,
+      "learning_rate": 4.647743879404273e-05,
+      "loss": 0.4373,
+      "step": 1342
+    },
+    {
+      "epoch": 0.8171585031944022,
+      "grad_norm": 1.1502752304077148,
+      "learning_rate": 4.647106466051741e-05,
+      "loss": 0.4047,
+      "step": 1343
+    },
+    {
+      "epoch": 0.8177669607544874,
+      "grad_norm": 1.0501556396484375,
+      "learning_rate": 4.6464685203063005e-05,
+      "loss": 0.422,
+      "step": 1344
+    },
+    {
+      "epoch": 0.8183754183145726,
+      "grad_norm": 0.8644611835479736,
+      "learning_rate": 4.645830042326137e-05,
+      "loss": 0.3404,
+      "step": 1345
+    },
+    {
+      "epoch": 0.8189838758746577,
+      "grad_norm": 0.9725202322006226,
+      "learning_rate": 4.645191032269565e-05,
+      "loss": 0.4371,
+      "step": 1346
+    },
+    {
+      "epoch": 0.8195923334347429,
+      "grad_norm": 1.0984727144241333,
+      "learning_rate": 4.644551490295033e-05,
+      "loss": 0.396,
+      "step": 1347
+    },
+    {
+      "epoch": 0.8202007909948281,
+      "grad_norm": 1.0615260601043701,
+      "learning_rate": 4.643911416561121e-05,
+      "loss": 0.4147,
+      "step": 1348
+    },
+    {
+      "epoch": 0.8208092485549133,
+      "grad_norm": 1.1247777938842773,
+      "learning_rate": 4.6432708112265397e-05,
+      "loss": 0.3993,
+      "step": 1349
+    },
+    {
+      "epoch": 0.8214177061149984,
+      "grad_norm": 1.0866373777389526,
+      "learning_rate": 4.642629674450134e-05,
+      "loss": 0.439,
+      "step": 1350
+    },
+    {
+      "epoch": 0.8220261636750836,
+      "grad_norm": 1.415543556213379,
+      "learning_rate": 4.641988006390877e-05,
+      "loss": 0.4355,
+      "step": 1351
+    },
+    {
+      "epoch": 0.8226346212351688,
+      "grad_norm": 3.2268388271331787,
+      "learning_rate": 4.641345807207879e-05,
+      "loss": 0.4287,
+      "step": 1352
+    },
+    {
+      "epoch": 0.823243078795254,
+      "grad_norm": 1.167698860168457,
+      "learning_rate": 4.640703077060374e-05,
+      "loss": 0.4728,
+      "step": 1353
+    },
+    {
+      "epoch": 0.8238515363553393,
+      "grad_norm": 1.2120864391326904,
+      "learning_rate": 4.640059816107737e-05,
+      "loss": 0.4648,
+      "step": 1354
+    },
+    {
+      "epoch": 0.8244599939154243,
+      "grad_norm": 1.0012540817260742,
+      "learning_rate": 4.639416024509466e-05,
+      "loss": 0.4621,
+      "step": 1355
+    },
+    {
+      "epoch": 0.8250684514755096,
+      "grad_norm": 2.5287537574768066,
+      "learning_rate": 4.638771702425197e-05,
+      "loss": 0.4767,
+      "step": 1356
+    },
+    {
+      "epoch": 0.8256769090355948,
+      "grad_norm": 1.1467103958129883,
+      "learning_rate": 4.638126850014694e-05,
+      "loss": 0.4715,
+      "step": 1357
+    },
+    {
+      "epoch": 0.82628536659568,
+      "grad_norm": 1.286732792854309,
+      "learning_rate": 4.637481467437854e-05,
+      "loss": 0.404,
+      "step": 1358
+    },
+    {
+      "epoch": 0.8268938241557652,
+      "grad_norm": 1.0348389148712158,
+      "learning_rate": 4.6368355548547046e-05,
+      "loss": 0.4483,
+      "step": 1359
+    },
+    {
+      "epoch": 0.8275022817158503,
+      "grad_norm": 1.1077744960784912,
+      "learning_rate": 4.636189112425405e-05,
+      "loss": 0.4157,
+      "step": 1360
+    },
+    {
+      "epoch": 0.8281107392759355,
+      "grad_norm": 0.9787223935127258,
+      "learning_rate": 4.635542140310246e-05,
+      "loss": 0.3761,
+      "step": 1361
+    },
+    {
+      "epoch": 0.8287191968360207,
+      "grad_norm": 1.1058927774429321,
+      "learning_rate": 4.6348946386696506e-05,
+      "loss": 0.3918,
+      "step": 1362
+    },
+    {
+      "epoch": 0.8293276543961059,
+      "grad_norm": 1.0943537950515747,
+      "learning_rate": 4.6342466076641715e-05,
+      "loss": 0.4403,
+      "step": 1363
+    },
+    {
+      "epoch": 0.8299361119561911,
+      "grad_norm": 1.1616437435150146,
+      "learning_rate": 4.633598047454494e-05,
+      "loss": 0.4496,
+      "step": 1364
+    },
+    {
+      "epoch": 0.8305445695162762,
+      "grad_norm": 1.026498794555664,
+      "learning_rate": 4.632948958201432e-05,
+      "loss": 0.4414,
+      "step": 1365
+    },
+    {
+      "epoch": 0.8311530270763614,
+      "grad_norm": 1.1366041898727417,
+      "learning_rate": 4.6322993400659355e-05,
+      "loss": 0.4226,
+      "step": 1366
+    },
+    {
+      "epoch": 0.8317614846364466,
+      "grad_norm": 0.9318983554840088,
+      "learning_rate": 4.631649193209081e-05,
+      "loss": 0.3743,
+      "step": 1367
+    },
+    {
+      "epoch": 0.8323699421965318,
+      "grad_norm": 1.0723328590393066,
+      "learning_rate": 4.6309985177920776e-05,
+      "loss": 0.3779,
+      "step": 1368
+    },
+    {
+      "epoch": 0.832978399756617,
+      "grad_norm": 0.9842046499252319,
+      "learning_rate": 4.630347313976266e-05,
+      "loss": 0.3871,
+      "step": 1369
+    },
+    {
+      "epoch": 0.8335868573167021,
+      "grad_norm": 1.232901930809021,
+      "learning_rate": 4.629695581923118e-05,
+      "loss": 0.3983,
+      "step": 1370
+    },
+    {
+      "epoch": 0.8341953148767873,
+      "grad_norm": 1.4811533689498901,
+      "learning_rate": 4.629043321794237e-05,
+      "loss": 0.4282,
+      "step": 1371
+    },
+    {
+      "epoch": 0.8348037724368725,
+      "grad_norm": 0.9871484637260437,
+      "learning_rate": 4.628390533751353e-05,
+      "loss": 0.3485,
+      "step": 1372
+    },
+    {
+      "epoch": 0.8354122299969577,
+      "grad_norm": 1.0894109010696411,
+      "learning_rate": 4.6277372179563336e-05,
+      "loss": 0.4226,
+      "step": 1373
+    },
+    {
+      "epoch": 0.8360206875570428,
+      "grad_norm": 0.9860946536064148,
+      "learning_rate": 4.627083374571173e-05,
+      "loss": 0.3852,
+      "step": 1374
+    },
+    {
+      "epoch": 0.836629145117128,
+      "grad_norm": 0.9105382561683655,
+      "learning_rate": 4.6264290037579955e-05,
+      "loss": 0.3875,
+      "step": 1375
+    },
+    {
+      "epoch": 0.8372376026772133,
+      "grad_norm": 0.9577265977859497,
+      "learning_rate": 4.625774105679059e-05,
+      "loss": 0.4193,
+      "step": 1376
+    },
+    {
+      "epoch": 0.8378460602372985,
+      "grad_norm": 1.2966721057891846,
+      "learning_rate": 4.625118680496752e-05,
+      "loss": 0.4104,
+      "step": 1377
+    },
+    {
+      "epoch": 0.8384545177973837,
+      "grad_norm": 1.9938510656356812,
+      "learning_rate": 4.624462728373591e-05,
+      "loss": 0.4041,
+      "step": 1378
+    },
+    {
+      "epoch": 0.8390629753574688,
+      "grad_norm": 0.9969017505645752,
+      "learning_rate": 4.6238062494722254e-05,
+      "loss": 0.4448,
+      "step": 1379
+    },
+    {
+      "epoch": 0.839671432917554,
+      "grad_norm": 1.0461878776550293,
+      "learning_rate": 4.623149243955435e-05,
+      "loss": 0.4411,
+      "step": 1380
+    },
+    {
+      "epoch": 0.8402798904776392,
+      "grad_norm": 1.1554114818572998,
+      "learning_rate": 4.6224917119861286e-05,
+      "loss": 0.4828,
+      "step": 1381
+    },
+    {
+      "epoch": 0.8408883480377244,
+      "grad_norm": 1.0953601598739624,
+      "learning_rate": 4.6218336537273476e-05,
+      "loss": 0.4883,
+      "step": 1382
+    },
+    {
+      "epoch": 0.8414968055978096,
+      "grad_norm": 1.0780771970748901,
+      "learning_rate": 4.621175069342263e-05,
+      "loss": 0.4425,
+      "step": 1383
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.8909410834312439,
+      "learning_rate": 4.620515958994176e-05,
+      "loss": 0.4176,
+      "step": 1384
+    },
+    {
+      "epoch": 0.8427137207179799,
+      "grad_norm": 1.1288505792617798,
+      "learning_rate": 4.619856322846518e-05,
+      "loss": 0.4576,
+      "step": 1385
+    },
+    {
+      "epoch": 0.8433221782780651,
+      "grad_norm": 1.2266353368759155,
+      "learning_rate": 4.619196161062854e-05,
+      "loss": 0.3678,
+      "step": 1386
+    },
+    {
+      "epoch": 0.8439306358381503,
+      "grad_norm": 1.1094207763671875,
+      "learning_rate": 4.6185354738068726e-05,
+      "loss": 0.3636,
+      "step": 1387
+    },
+    {
+      "epoch": 0.8445390933982355,
+      "grad_norm": 1.0035345554351807,
+      "learning_rate": 4.617874261242399e-05,
+      "loss": 0.3561,
+      "step": 1388
+    },
+    {
+      "epoch": 0.8451475509583206,
+      "grad_norm": 1.1671876907348633,
+      "learning_rate": 4.617212523533386e-05,
+      "loss": 0.395,
+      "step": 1389
+    },
+    {
+      "epoch": 0.8457560085184058,
+      "grad_norm": 1.2926161289215088,
+      "learning_rate": 4.616550260843917e-05,
+      "loss": 0.4506,
+      "step": 1390
+    },
+    {
+      "epoch": 0.846364466078491,
+      "grad_norm": 1.220173954963684,
+      "learning_rate": 4.6158874733382056e-05,
+      "loss": 0.4585,
+      "step": 1391
+    },
+    {
+      "epoch": 0.8469729236385762,
+      "grad_norm": 0.9969469308853149,
+      "learning_rate": 4.6152241611805956e-05,
+      "loss": 0.3849,
+      "step": 1392
+    },
+    {
+      "epoch": 0.8475813811986614,
+      "grad_norm": 0.9798814058303833,
+      "learning_rate": 4.61456032453556e-05,
+      "loss": 0.4175,
+      "step": 1393
+    },
+    {
+      "epoch": 0.8481898387587465,
+      "grad_norm": 1.095361351966858,
+      "learning_rate": 4.613895963567704e-05,
+      "loss": 0.4047,
+      "step": 1394
+    },
+    {
+      "epoch": 0.8487982963188317,
+      "grad_norm": 0.9531521201133728,
+      "learning_rate": 4.6132310784417595e-05,
+      "loss": 0.348,
+      "step": 1395
+    },
+    {
+      "epoch": 0.849406753878917,
+      "grad_norm": 1.015824556350708,
+      "learning_rate": 4.612565669322592e-05,
+      "loss": 0.4401,
+      "step": 1396
+    },
+    {
+      "epoch": 0.8500152114390022,
+      "grad_norm": 1.1718368530273438,
+      "learning_rate": 4.611899736375194e-05,
+      "loss": 0.3683,
+      "step": 1397
+    },
+    {
+      "epoch": 0.8506236689990874,
+      "grad_norm": 1.0078071355819702,
+      "learning_rate": 4.61123327976469e-05,
+      "loss": 0.3512,
+      "step": 1398
+    },
+    {
+      "epoch": 0.8512321265591725,
+      "grad_norm": 1.269882321357727,
+      "learning_rate": 4.610566299656332e-05,
+      "loss": 0.425,
+      "step": 1399
+    },
+    {
+      "epoch": 0.8518405841192577,
+      "grad_norm": 1.0912398099899292,
+      "learning_rate": 4.609898796215506e-05,
+      "loss": 0.421,
+      "step": 1400
+    },
+    {
+      "epoch": 0.8524490416793429,
+      "grad_norm": 1.116542100906372,
+      "learning_rate": 4.609230769607723e-05,
+      "loss": 0.4179,
+      "step": 1401
+    },
+    {
+      "epoch": 0.8530574992394281,
+      "grad_norm": 1.4198042154312134,
+      "learning_rate": 4.6085622199986266e-05,
+      "loss": 0.4061,
+      "step": 1402
+    },
+    {
+      "epoch": 0.8536659567995132,
+      "grad_norm": 1.0006221532821655,
+      "learning_rate": 4.607893147553989e-05,
+      "loss": 0.3707,
+      "step": 1403
+    },
+    {
+      "epoch": 0.8542744143595984,
+      "grad_norm": 1.0920796394348145,
+      "learning_rate": 4.607223552439711e-05,
+      "loss": 0.3692,
+      "step": 1404
+    },
+    {
+      "epoch": 0.8548828719196836,
+      "grad_norm": 0.9862841367721558,
+      "learning_rate": 4.606553434821826e-05,
+      "loss": 0.3852,
+      "step": 1405
+    },
+    {
+      "epoch": 0.8554913294797688,
+      "grad_norm": 1.6820799112319946,
+      "learning_rate": 4.605882794866495e-05,
+      "loss": 0.4196,
+      "step": 1406
+    },
+    {
+      "epoch": 0.856099787039854,
+      "grad_norm": 1.1753647327423096,
+      "learning_rate": 4.605211632740008e-05,
+      "loss": 0.4056,
+      "step": 1407
+    },
+    {
+      "epoch": 0.8567082445999391,
+      "grad_norm": 1.041169285774231,
+      "learning_rate": 4.6045399486087856e-05,
+      "loss": 0.4012,
+      "step": 1408
+    },
+    {
+      "epoch": 0.8573167021600243,
+      "grad_norm": 1.1424866914749146,
+      "learning_rate": 4.603867742639377e-05,
+      "loss": 0.4843,
+      "step": 1409
+    },
+    {
+      "epoch": 0.8579251597201095,
+      "grad_norm": 1.1785975694656372,
+      "learning_rate": 4.6031950149984624e-05,
+      "loss": 0.3528,
+      "step": 1410
+    },
+    {
+      "epoch": 0.8585336172801947,
+      "grad_norm": 1.092089056968689,
+      "learning_rate": 4.6025217658528497e-05,
+      "loss": 0.4248,
+      "step": 1411
+    },
+    {
+      "epoch": 0.8591420748402799,
+      "grad_norm": 1.6886132955551147,
+      "learning_rate": 4.601847995369477e-05,
+      "loss": 0.3746,
+      "step": 1412
+    },
+    {
+      "epoch": 0.859750532400365,
+      "grad_norm": 0.9200694561004639,
+      "learning_rate": 4.60117370371541e-05,
+      "loss": 0.3175,
+      "step": 1413
+    },
+    {
+      "epoch": 0.8603589899604502,
+      "grad_norm": 0.9181839823722839,
+      "learning_rate": 4.600498891057845e-05,
+      "loss": 0.3799,
+      "step": 1414
+    },
+    {
+      "epoch": 0.8609674475205354,
+      "grad_norm": 0.9873498678207397,
+      "learning_rate": 4.599823557564109e-05,
+      "loss": 0.3687,
+      "step": 1415
+    },
+    {
+      "epoch": 0.8615759050806207,
+      "grad_norm": 1.0958657264709473,
+      "learning_rate": 4.5991477034016564e-05,
+      "loss": 0.4309,
+      "step": 1416
+    },
+    {
+      "epoch": 0.8621843626407059,
+      "grad_norm": 0.935565710067749,
+      "learning_rate": 4.598471328738069e-05,
+      "loss": 0.3494,
+      "step": 1417
+    },
+    {
+      "epoch": 0.862792820200791,
+      "grad_norm": 1.0181663036346436,
+      "learning_rate": 4.597794433741061e-05,
+      "loss": 0.3567,
+      "step": 1418
+    },
+    {
+      "epoch": 0.8634012777608762,
+      "grad_norm": 1.1385202407836914,
+      "learning_rate": 4.597117018578473e-05,
+      "loss": 0.4259,
+      "step": 1419
+    },
+    {
+      "epoch": 0.8640097353209614,
+      "grad_norm": 1.041556715965271,
+      "learning_rate": 4.596439083418278e-05,
+      "loss": 0.3825,
+      "step": 1420
+    },
+    {
+      "epoch": 0.8646181928810466,
+      "grad_norm": 0.9178049564361572,
+      "learning_rate": 4.5957606284285736e-05,
+      "loss": 0.3587,
+      "step": 1421
+    },
+    {
+      "epoch": 0.8652266504411318,
+      "grad_norm": 1.0861730575561523,
+      "learning_rate": 4.595081653777589e-05,
+      "loss": 0.4345,
+      "step": 1422
+    },
+    {
+      "epoch": 0.8658351080012169,
+      "grad_norm": 1.1010128259658813,
+      "learning_rate": 4.594402159633681e-05,
+      "loss": 0.4007,
+      "step": 1423
+    },
+    {
+      "epoch": 0.8664435655613021,
+      "grad_norm": 1.000855803489685,
+      "learning_rate": 4.593722146165337e-05,
+      "loss": 0.4319,
+      "step": 1424
+    },
+    {
+      "epoch": 0.8670520231213873,
+      "grad_norm": 1.2340881824493408,
+      "learning_rate": 4.5930416135411715e-05,
+      "loss": 0.4397,
+      "step": 1425
+    },
+    {
+      "epoch": 0.8676604806814725,
+      "grad_norm": 1.198055386543274,
+      "learning_rate": 4.592360561929928e-05,
+      "loss": 0.4217,
+      "step": 1426
+    },
+    {
+      "epoch": 0.8682689382415576,
+      "grad_norm": 1.011428952217102,
+      "learning_rate": 4.591678991500479e-05,
+      "loss": 0.3898,
+      "step": 1427
+    },
+    {
+      "epoch": 0.8688773958016428,
+      "grad_norm": 1.2541236877441406,
+      "learning_rate": 4.590996902421825e-05,
+      "loss": 0.485,
+      "step": 1428
+    },
+    {
+      "epoch": 0.869485853361728,
+      "grad_norm": 1.0267783403396606,
+      "learning_rate": 4.590314294863097e-05,
+      "loss": 0.4073,
+      "step": 1429
+    },
+    {
+      "epoch": 0.8700943109218132,
+      "grad_norm": 1.1068730354309082,
+      "learning_rate": 4.589631168993552e-05,
+      "loss": 0.3804,
+      "step": 1430
+    },
+    {
+      "epoch": 0.8707027684818984,
+      "grad_norm": 0.9509475827217102,
+      "learning_rate": 4.5889475249825774e-05,
+      "loss": 0.3519,
+      "step": 1431
+    },
+    {
+      "epoch": 0.8713112260419835,
+      "grad_norm": 1.0428766012191772,
+      "learning_rate": 4.5882633629996886e-05,
+      "loss": 0.3894,
+      "step": 1432
+    },
+    {
+      "epoch": 0.8719196836020687,
+      "grad_norm": 1.0796087980270386,
+      "learning_rate": 4.5875786832145287e-05,
+      "loss": 0.3919,
+      "step": 1433
+    },
+    {
+      "epoch": 0.8725281411621539,
+      "grad_norm": 1.0344951152801514,
+      "learning_rate": 4.5868934857968695e-05,
+      "loss": 0.4074,
+      "step": 1434
+    },
+    {
+      "epoch": 0.8731365987222391,
+      "grad_norm": 1.127869963645935,
+      "learning_rate": 4.586207770916612e-05,
+      "loss": 0.3942,
+      "step": 1435
+    },
+    {
+      "epoch": 0.8737450562823244,
+      "grad_norm": 1.5959341526031494,
+      "learning_rate": 4.585521538743785e-05,
+      "loss": 0.3944,
+      "step": 1436
+    },
+    {
+      "epoch": 0.8743535138424094,
+      "grad_norm": 0.9701662659645081,
+      "learning_rate": 4.584834789448544e-05,
+      "loss": 0.3523,
+      "step": 1437
+    },
+    {
+      "epoch": 0.8749619714024947,
+      "grad_norm": 1.1680253744125366,
+      "learning_rate": 4.5841475232011773e-05,
+      "loss": 0.4448,
+      "step": 1438
+    },
+    {
+      "epoch": 0.8755704289625799,
+      "grad_norm": 1.0042173862457275,
+      "learning_rate": 4.5834597401720956e-05,
+      "loss": 0.3796,
+      "step": 1439
+    },
+    {
+      "epoch": 0.8761788865226651,
+      "grad_norm": 1.100633978843689,
+      "learning_rate": 4.582771440531841e-05,
+      "loss": 0.3943,
+      "step": 1440
+    },
+    {
+      "epoch": 0.8767873440827503,
+      "grad_norm": 1.1571612358093262,
+      "learning_rate": 4.582082624451084e-05,
+      "loss": 0.3959,
+      "step": 1441
+    },
+    {
+      "epoch": 0.8773958016428354,
+      "grad_norm": 1.4444702863693237,
+      "learning_rate": 4.581393292100621e-05,
+      "loss": 0.4962,
+      "step": 1442
+    },
+    {
+      "epoch": 0.8780042592029206,
+      "grad_norm": 0.991072952747345,
+      "learning_rate": 4.5807034436513784e-05,
+      "loss": 0.384,
+      "step": 1443
+    },
+    {
+      "epoch": 0.8786127167630058,
+      "grad_norm": 1.3088997602462769,
+      "learning_rate": 4.5800130792744096e-05,
+      "loss": 0.3756,
+      "step": 1444
+    },
+    {
+      "epoch": 0.879221174323091,
+      "grad_norm": 1.1207791566848755,
+      "learning_rate": 4.5793221991408966e-05,
+      "loss": 0.4245,
+      "step": 1445
+    },
+    {
+      "epoch": 0.8798296318831762,
+      "grad_norm": 0.9720820188522339,
+      "learning_rate": 4.578630803422148e-05,
+      "loss": 0.4084,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8804380894432613,
+      "grad_norm": 0.9813503623008728,
+      "learning_rate": 4.577938892289603e-05,
+      "loss": 0.3907,
+      "step": 1447
+    },
+    {
+      "epoch": 0.8810465470033465,
+      "grad_norm": 0.9682145714759827,
+      "learning_rate": 4.577246465914825e-05,
+      "loss": 0.4057,
+      "step": 1448
+    },
+    {
+      "epoch": 0.8816550045634317,
+      "grad_norm": 1.0260658264160156,
+      "learning_rate": 4.576553524469507e-05,
+      "loss": 0.351,
+      "step": 1449
+    },
+    {
+      "epoch": 0.8822634621235169,
+      "grad_norm": 1.1361267566680908,
+      "learning_rate": 4.575860068125471e-05,
+      "loss": 0.4011,
+      "step": 1450
+    },
+    {
+      "epoch": 0.882871919683602,
+      "grad_norm": 1.084456443786621,
+      "learning_rate": 4.575166097054662e-05,
+      "loss": 0.3984,
+      "step": 1451
+    },
+    {
+      "epoch": 0.8834803772436872,
+      "grad_norm": 1.2074880599975586,
+      "learning_rate": 4.57447161142916e-05,
+      "loss": 0.3984,
+      "step": 1452
+    },
+    {
+      "epoch": 0.8840888348037724,
+      "grad_norm": 1.044838786125183,
+      "learning_rate": 4.5737766114211654e-05,
+      "loss": 0.3779,
+      "step": 1453
+    },
+    {
+      "epoch": 0.8846972923638576,
+      "grad_norm": 1.0638879537582397,
+      "learning_rate": 4.5730810972030114e-05,
+      "loss": 0.3623,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8853057499239428,
+      "grad_norm": 1.0590424537658691,
+      "learning_rate": 4.572385068947155e-05,
+      "loss": 0.3891,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8859142074840279,
+      "grad_norm": 1.0808945894241333,
+      "learning_rate": 4.5716885268261834e-05,
+      "loss": 0.374,
+      "step": 1456
+    },
+    {
+      "epoch": 0.8865226650441131,
+      "grad_norm": 0.9356719851493835,
+      "learning_rate": 4.570991471012809e-05,
+      "loss": 0.3635,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8871311226041984,
+      "grad_norm": 1.1011765003204346,
+      "learning_rate": 4.570293901679873e-05,
+      "loss": 0.3804,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8877395801642836,
+      "grad_norm": 1.0141364336013794,
+      "learning_rate": 4.569595819000344e-05,
+      "loss": 0.4072,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8883480377243688,
+      "grad_norm": 1.7948359251022339,
+      "learning_rate": 4.568897223147316e-05,
+      "loss": 0.3803,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8889564952844539,
+      "grad_norm": 1.2225985527038574,
+      "learning_rate": 4.5681981142940126e-05,
+      "loss": 0.3975,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8895649528445391,
+      "grad_norm": 0.9955945611000061,
+      "learning_rate": 4.5674984926137844e-05,
+      "loss": 0.3601,
+      "step": 1462
+    },
+    {
+      "epoch": 0.8901734104046243,
+      "grad_norm": 1.1640013456344604,
+      "learning_rate": 4.5667983582801064e-05,
+      "loss": 0.3927,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8907818679647095,
+      "grad_norm": 1.3846951723098755,
+      "learning_rate": 4.566097711466585e-05,
+      "loss": 0.3688,
+      "step": 1464
+    },
+    {
+      "epoch": 0.8913903255247947,
+      "grad_norm": 1.1909358501434326,
+      "learning_rate": 4.56539655234695e-05,
+      "loss": 0.4304,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8919987830848798,
+      "grad_norm": 1.1379756927490234,
+      "learning_rate": 4.56469488109506e-05,
+      "loss": 0.4431,
+      "step": 1466
+    },
+    {
+      "epoch": 0.892607240644965,
+      "grad_norm": 0.9373139142990112,
+      "learning_rate": 4.5639926978849e-05,
+      "loss": 0.4213,
+      "step": 1467
+    },
+    {
+      "epoch": 0.8932156982050502,
+      "grad_norm": 1.0285266637802124,
+      "learning_rate": 4.563290002890583e-05,
+      "loss": 0.4201,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8938241557651354,
+      "grad_norm": 1.1803969144821167,
+      "learning_rate": 4.5625867962863466e-05,
+      "loss": 0.4099,
+      "step": 1469
+    },
+    {
+      "epoch": 0.8944326133252206,
+      "grad_norm": 1.115708589553833,
+      "learning_rate": 4.5618830782465584e-05,
+      "loss": 0.4085,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8950410708853057,
+      "grad_norm": 1.1141424179077148,
+      "learning_rate": 4.56117884894571e-05,
+      "loss": 0.4274,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8956495284453909,
+      "grad_norm": 1.2159873247146606,
+      "learning_rate": 4.5604741085584215e-05,
+      "loss": 0.4164,
+      "step": 1472
+    },
+    {
+      "epoch": 0.8962579860054761,
+      "grad_norm": 1.0636794567108154,
+      "learning_rate": 4.559768857259438e-05,
+      "loss": 0.4036,
+      "step": 1473
+    },
+    {
+      "epoch": 0.8968664435655613,
+      "grad_norm": 1.2936556339263916,
+      "learning_rate": 4.5590630952236336e-05,
+      "loss": 0.4509,
+      "step": 1474
+    },
+    {
+      "epoch": 0.8974749011256465,
+      "grad_norm": 1.1456127166748047,
+      "learning_rate": 4.558356822626008e-05,
+      "loss": 0.4016,
+      "step": 1475
+    },
+    {
+      "epoch": 0.8980833586857316,
+      "grad_norm": 1.0449507236480713,
+      "learning_rate": 4.557650039641687e-05,
+      "loss": 0.4636,
+      "step": 1476
+    },
+    {
+      "epoch": 0.8986918162458168,
+      "grad_norm": 1.048593282699585,
+      "learning_rate": 4.5569427464459226e-05,
+      "loss": 0.4334,
+      "step": 1477
+    },
+    {
+      "epoch": 0.899300273805902,
+      "grad_norm": 0.8729017972946167,
+      "learning_rate": 4.556234943214095e-05,
+      "loss": 0.3823,
+      "step": 1478
+    },
+    {
+      "epoch": 0.8999087313659873,
+      "grad_norm": 1.1806243658065796,
+      "learning_rate": 4.55552663012171e-05,
+      "loss": 0.4189,
+      "step": 1479
+    },
+    {
+      "epoch": 0.9005171889260724,
+      "grad_norm": 1.0129870176315308,
+      "learning_rate": 4.554817807344399e-05,
+      "loss": 0.3643,
+      "step": 1480
+    },
+    {
+      "epoch": 0.9011256464861576,
+      "grad_norm": 1.1537796258926392,
+      "learning_rate": 4.5541084750579205e-05,
+      "loss": 0.4925,
+      "step": 1481
+    },
+    {
+      "epoch": 0.9017341040462428,
+      "grad_norm": 0.9418867230415344,
+      "learning_rate": 4.55339863343816e-05,
+      "loss": 0.417,
+      "step": 1482
+    },
+    {
+      "epoch": 0.902342561606328,
+      "grad_norm": 1.1266993284225464,
+      "learning_rate": 4.5526882826611285e-05,
+      "loss": 0.4141,
+      "step": 1483
+    },
+    {
+      "epoch": 0.9029510191664132,
+      "grad_norm": 1.0497965812683105,
+      "learning_rate": 4.5519774229029625e-05,
+      "loss": 0.3798,
+      "step": 1484
+    },
+    {
+      "epoch": 0.9035594767264983,
+      "grad_norm": 1.159199595451355,
+      "learning_rate": 4.551266054339927e-05,
+      "loss": 0.3659,
+      "step": 1485
+    },
+    {
+      "epoch": 0.9041679342865835,
+      "grad_norm": 0.9190648794174194,
+      "learning_rate": 4.55055417714841e-05,
+      "loss": 0.3857,
+      "step": 1486
+    },
+    {
+      "epoch": 0.9047763918466687,
+      "grad_norm": 1.1719613075256348,
+      "learning_rate": 4.549841791504929e-05,
+      "loss": 0.3847,
+      "step": 1487
+    },
+    {
+      "epoch": 0.9053848494067539,
+      "grad_norm": 0.9929993748664856,
+      "learning_rate": 4.5491288975861254e-05,
+      "loss": 0.4189,
+      "step": 1488
+    },
+    {
+      "epoch": 0.9059933069668391,
+      "grad_norm": 1.0294368267059326,
+      "learning_rate": 4.548415495568767e-05,
+      "loss": 0.4341,
+      "step": 1489
+    },
+    {
+      "epoch": 0.9066017645269242,
+      "grad_norm": 1.082219123840332,
+      "learning_rate": 4.5477015856297475e-05,
+      "loss": 0.418,
+      "step": 1490
+    },
+    {
+      "epoch": 0.9072102220870094,
+      "grad_norm": 1.0178451538085938,
+      "learning_rate": 4.546987167946088e-05,
+      "loss": 0.4209,
+      "step": 1491
+    },
+    {
+      "epoch": 0.9078186796470946,
+      "grad_norm": 0.9833378791809082,
+      "learning_rate": 4.546272242694933e-05,
+      "loss": 0.4089,
+      "step": 1492
+    },
+    {
+      "epoch": 0.9084271372071798,
+      "grad_norm": 1.0812345743179321,
+      "learning_rate": 4.5455568100535545e-05,
+      "loss": 0.39,
+      "step": 1493
+    },
+    {
+      "epoch": 0.909035594767265,
+      "grad_norm": 1.0747591257095337,
+      "learning_rate": 4.544840870199351e-05,
+      "loss": 0.4182,
+      "step": 1494
+    },
+    {
+      "epoch": 0.9096440523273501,
+      "grad_norm": 0.9758956432342529,
+      "learning_rate": 4.5441244233098434e-05,
+      "loss": 0.32,
+      "step": 1495
+    },
+    {
+      "epoch": 0.9102525098874353,
+      "grad_norm": 1.013469934463501,
+      "learning_rate": 4.5434074695626826e-05,
+      "loss": 0.4348,
+      "step": 1496
+    },
+    {
+      "epoch": 0.9108609674475205,
+      "grad_norm": 0.9901841878890991,
+      "learning_rate": 4.542690009135643e-05,
+      "loss": 0.3635,
+      "step": 1497
+    },
+    {
+      "epoch": 0.9114694250076057,
+      "grad_norm": 1.0742686986923218,
+      "learning_rate": 4.541972042206625e-05,
+      "loss": 0.4546,
+      "step": 1498
+    },
+    {
+      "epoch": 0.912077882567691,
+      "grad_norm": 1.0467991828918457,
+      "learning_rate": 4.541253568953654e-05,
+      "loss": 0.4002,
+      "step": 1499
+    },
+    {
+      "epoch": 0.912686340127776,
+      "grad_norm": 1.4334415197372437,
+      "learning_rate": 4.540534589554881e-05,
+      "loss": 0.3806,
+      "step": 1500
+    },
+    {
+      "epoch": 0.9132947976878613,
+      "grad_norm": 1.107809066772461,
+      "learning_rate": 4.539815104188584e-05,
+      "loss": 0.4402,
+      "step": 1501
+    },
+    {
+      "epoch": 0.9139032552479465,
+      "grad_norm": 1.0222909450531006,
+      "learning_rate": 4.539095113033165e-05,
+      "loss": 0.4404,
+      "step": 1502
+    },
+    {
+      "epoch": 0.9145117128080317,
+      "grad_norm": 1.0362690687179565,
+      "learning_rate": 4.538374616267151e-05,
+      "loss": 0.385,
+      "step": 1503
+    },
+    {
+      "epoch": 0.9151201703681168,
+      "grad_norm": 1.0879029035568237,
+      "learning_rate": 4.537653614069196e-05,
+      "loss": 0.4255,
+      "step": 1504
+    },
+    {
+      "epoch": 0.915728627928202,
+      "grad_norm": 1.040208339691162,
+      "learning_rate": 4.536932106618078e-05,
+      "loss": 0.3637,
+      "step": 1505
+    },
+    {
+      "epoch": 0.9163370854882872,
+      "grad_norm": 1.043503999710083,
+      "learning_rate": 4.536210094092702e-05,
+      "loss": 0.3691,
+      "step": 1506
+    },
+    {
+      "epoch": 0.9169455430483724,
+      "grad_norm": 1.0326288938522339,
+      "learning_rate": 4.535487576672095e-05,
+      "loss": 0.3499,
+      "step": 1507
+    },
+    {
+      "epoch": 0.9175540006084576,
+      "grad_norm": 1.1494892835617065,
+      "learning_rate": 4.5347645545354136e-05,
+      "loss": 0.3684,
+      "step": 1508
+    },
+    {
+      "epoch": 0.9181624581685427,
+      "grad_norm": 1.2776802778244019,
+      "learning_rate": 4.534041027861935e-05,
+      "loss": 0.4245,
+      "step": 1509
+    },
+    {
+      "epoch": 0.9187709157286279,
+      "grad_norm": 1.056604266166687,
+      "learning_rate": 4.533316996831064e-05,
+      "loss": 0.3585,
+      "step": 1510
+    },
+    {
+      "epoch": 0.9193793732887131,
+      "grad_norm": 1.12501859664917,
+      "learning_rate": 4.532592461622331e-05,
+      "loss": 0.3689,
+      "step": 1511
+    },
+    {
+      "epoch": 0.9199878308487983,
+      "grad_norm": 1.0672175884246826,
+      "learning_rate": 4.531867422415391e-05,
+      "loss": 0.393,
+      "step": 1512
+    },
+    {
+      "epoch": 0.9205962884088835,
+      "grad_norm": 1.0935568809509277,
+      "learning_rate": 4.531141879390022e-05,
+      "loss": 0.3826,
+      "step": 1513
+    },
+    {
+      "epoch": 0.9212047459689686,
+      "grad_norm": 1.1044657230377197,
+      "learning_rate": 4.5304158327261294e-05,
+      "loss": 0.4487,
+      "step": 1514
+    },
+    {
+      "epoch": 0.9218132035290538,
+      "grad_norm": 1.1453882455825806,
+      "learning_rate": 4.5296892826037414e-05,
+      "loss": 0.3947,
+      "step": 1515
+    },
+    {
+      "epoch": 0.922421661089139,
+      "grad_norm": 0.9416497349739075,
+      "learning_rate": 4.5289622292030134e-05,
+      "loss": 0.3573,
+      "step": 1516
+    },
+    {
+      "epoch": 0.9230301186492242,
+      "grad_norm": 1.0041788816452026,
+      "learning_rate": 4.528234672704224e-05,
+      "loss": 0.3347,
+      "step": 1517
+    },
+    {
+      "epoch": 0.9236385762093094,
+      "grad_norm": 0.9780565500259399,
+      "learning_rate": 4.527506613287776e-05,
+      "loss": 0.3597,
+      "step": 1518
+    },
+    {
+      "epoch": 0.9242470337693945,
+      "grad_norm": 1.2946865558624268,
+      "learning_rate": 4.526778051134199e-05,
+      "loss": 0.4323,
+      "step": 1519
+    },
+    {
+      "epoch": 0.9248554913294798,
+      "grad_norm": 1.030163049697876,
+      "learning_rate": 4.526048986424146e-05,
+      "loss": 0.3541,
+      "step": 1520
+    },
+    {
+      "epoch": 0.925463948889565,
+      "grad_norm": 1.0969247817993164,
+      "learning_rate": 4.525319419338394e-05,
+      "loss": 0.415,
+      "step": 1521
+    },
+    {
+      "epoch": 0.9260724064496502,
+      "grad_norm": 1.0787875652313232,
+      "learning_rate": 4.5245893500578455e-05,
+      "loss": 0.4565,
+      "step": 1522
+    },
+    {
+      "epoch": 0.9266808640097354,
+      "grad_norm": 1.2321113348007202,
+      "learning_rate": 4.523858778763528e-05,
+      "loss": 0.4511,
+      "step": 1523
+    },
+    {
+      "epoch": 0.9272893215698205,
+      "grad_norm": 1.0210340023040771,
+      "learning_rate": 4.523127705636591e-05,
+      "loss": 0.4325,
+      "step": 1524
+    },
+    {
+      "epoch": 0.9278977791299057,
+      "grad_norm": 0.9291285872459412,
+      "learning_rate": 4.522396130858311e-05,
+      "loss": 0.3934,
+      "step": 1525
+    },
+    {
+      "epoch": 0.9285062366899909,
+      "grad_norm": 1.1080670356750488,
+      "learning_rate": 4.5216640546100884e-05,
+      "loss": 0.381,
+      "step": 1526
+    },
+    {
+      "epoch": 0.9291146942500761,
+      "grad_norm": 0.9290838837623596,
+      "learning_rate": 4.5209314770734475e-05,
+      "loss": 0.3477,
+      "step": 1527
+    },
+    {
+      "epoch": 0.9297231518101612,
+      "grad_norm": 1.5479507446289062,
+      "learning_rate": 4.520198398430037e-05,
+      "loss": 0.3941,
+      "step": 1528
+    },
+    {
+      "epoch": 0.9303316093702464,
+      "grad_norm": 1.0913853645324707,
+      "learning_rate": 4.5194648188616294e-05,
+      "loss": 0.3909,
+      "step": 1529
+    },
+    {
+      "epoch": 0.9309400669303316,
+      "grad_norm": 1.1218714714050293,
+      "learning_rate": 4.518730738550122e-05,
+      "loss": 0.4052,
+      "step": 1530
+    },
+    {
+      "epoch": 0.9315485244904168,
+      "grad_norm": 1.0080488920211792,
+      "learning_rate": 4.517996157677537e-05,
+      "loss": 0.3412,
+      "step": 1531
+    },
+    {
+      "epoch": 0.932156982050502,
+      "grad_norm": 1.1207358837127686,
+      "learning_rate": 4.517261076426018e-05,
+      "loss": 0.3878,
+      "step": 1532
+    },
+    {
+      "epoch": 0.9327654396105871,
+      "grad_norm": 1.1212834119796753,
+      "learning_rate": 4.516525494977837e-05,
+      "loss": 0.407,
+      "step": 1533
+    },
+    {
+      "epoch": 0.9333738971706723,
+      "grad_norm": 1.1488605737686157,
+      "learning_rate": 4.5157894135153845e-05,
+      "loss": 0.4392,
+      "step": 1534
+    },
+    {
+      "epoch": 0.9339823547307575,
+      "grad_norm": 1.036495327949524,
+      "learning_rate": 4.515052832221181e-05,
+      "loss": 0.3508,
+      "step": 1535
+    },
+    {
+      "epoch": 0.9345908122908427,
+      "grad_norm": 1.168699860572815,
+      "learning_rate": 4.514315751277867e-05,
+      "loss": 0.4334,
+      "step": 1536
+    },
+    {
+      "epoch": 0.9351992698509279,
+      "grad_norm": 1.0106819868087769,
+      "learning_rate": 4.513578170868206e-05,
+      "loss": 0.3479,
+      "step": 1537
+    },
+    {
+      "epoch": 0.935807727411013,
+      "grad_norm": 0.9819125533103943,
+      "learning_rate": 4.512840091175089e-05,
+      "loss": 0.3646,
+      "step": 1538
+    },
+    {
+      "epoch": 0.9364161849710982,
+      "grad_norm": 0.8794593811035156,
+      "learning_rate": 4.51210151238153e-05,
+      "loss": 0.3568,
+      "step": 1539
+    },
+    {
+      "epoch": 0.9370246425311834,
+      "grad_norm": 0.9659105539321899,
+      "learning_rate": 4.511362434670663e-05,
+      "loss": 0.3865,
+      "step": 1540
+    },
+    {
+      "epoch": 0.9376331000912687,
+      "grad_norm": 1.012877345085144,
+      "learning_rate": 4.510622858225752e-05,
+      "loss": 0.4071,
+      "step": 1541
+    },
+    {
+      "epoch": 0.9382415576513539,
+      "grad_norm": 0.9462942481040955,
+      "learning_rate": 4.509882783230177e-05,
+      "loss": 0.345,
+      "step": 1542
+    },
+    {
+      "epoch": 0.938850015211439,
+      "grad_norm": 1.1898541450500488,
+      "learning_rate": 4.509142209867448e-05,
+      "loss": 0.3663,
+      "step": 1543
+    },
+    {
+      "epoch": 0.9394584727715242,
+      "grad_norm": 1.2533682584762573,
+      "learning_rate": 4.508401138321196e-05,
+      "loss": 0.3441,
+      "step": 1544
+    },
+    {
+      "epoch": 0.9400669303316094,
+      "grad_norm": 1.0154099464416504,
+      "learning_rate": 4.507659568775177e-05,
+      "loss": 0.3578,
+      "step": 1545
+    },
+    {
+      "epoch": 0.9406753878916946,
+      "grad_norm": 1.03611159324646,
+      "learning_rate": 4.506917501413268e-05,
+      "loss": 0.3639,
+      "step": 1546
+    },
+    {
+      "epoch": 0.9412838454517798,
+      "grad_norm": 1.214494228363037,
+      "learning_rate": 4.506174936419471e-05,
+      "loss": 0.3985,
+      "step": 1547
+    },
+    {
+      "epoch": 0.9418923030118649,
+      "grad_norm": 1.0812007188796997,
+      "learning_rate": 4.505431873977911e-05,
+      "loss": 0.3469,
+      "step": 1548
+    },
+    {
+      "epoch": 0.9425007605719501,
+      "grad_norm": 1.1645228862762451,
+      "learning_rate": 4.504688314272837e-05,
+      "loss": 0.4381,
+      "step": 1549
+    },
+    {
+      "epoch": 0.9431092181320353,
+      "grad_norm": 1.1173677444458008,
+      "learning_rate": 4.5039442574886204e-05,
+      "loss": 0.3876,
+      "step": 1550
+    },
+    {
+      "epoch": 0.9437176756921205,
+      "grad_norm": 1.0114907026290894,
+      "learning_rate": 4.503199703809757e-05,
+      "loss": 0.4089,
+      "step": 1551
+    },
+    {
+      "epoch": 0.9443261332522057,
+      "grad_norm": 0.9891692996025085,
+      "learning_rate": 4.5024546534208645e-05,
+      "loss": 0.3515,
+      "step": 1552
+    },
+    {
+      "epoch": 0.9449345908122908,
+      "grad_norm": 1.005385398864746,
+      "learning_rate": 4.5017091065066837e-05,
+      "loss": 0.3328,
+      "step": 1553
+    },
+    {
+      "epoch": 0.945543048372376,
+      "grad_norm": 1.1059056520462036,
+      "learning_rate": 4.50096306325208e-05,
+      "loss": 0.4516,
+      "step": 1554
+    },
+    {
+      "epoch": 0.9461515059324612,
+      "grad_norm": 0.9695335626602173,
+      "learning_rate": 4.500216523842041e-05,
+      "loss": 0.3779,
+      "step": 1555
+    },
+    {
+      "epoch": 0.9467599634925464,
+      "grad_norm": 1.2391901016235352,
+      "learning_rate": 4.499469488461677e-05,
+      "loss": 0.3915,
+      "step": 1556
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.9961259961128235,
+      "learning_rate": 4.4987219572962224e-05,
+      "loss": 0.3433,
+      "step": 1557
+    },
+    {
+      "epoch": 0.9479768786127167,
+      "grad_norm": 1.959693193435669,
+      "learning_rate": 4.497973930531033e-05,
+      "loss": 0.3842,
+      "step": 1558
+    },
+    {
+      "epoch": 0.9485853361728019,
+      "grad_norm": 1.5181488990783691,
+      "learning_rate": 4.497225408351589e-05,
+      "loss": 0.4688,
+      "step": 1559
+    },
+    {
+      "epoch": 0.9491937937328871,
+      "grad_norm": 1.1111069917678833,
+      "learning_rate": 4.4964763909434914e-05,
+      "loss": 0.3564,
+      "step": 1560
+    },
+    {
+      "epoch": 0.9498022512929724,
+      "grad_norm": 1.0508601665496826,
+      "learning_rate": 4.495726878492465e-05,
+      "loss": 0.3788,
+      "step": 1561
+    },
+    {
+      "epoch": 0.9504107088530575,
+      "grad_norm": 1.1294502019882202,
+      "learning_rate": 4.494976871184361e-05,
+      "loss": 0.3692,
+      "step": 1562
+    },
+    {
+      "epoch": 0.9510191664131427,
+      "grad_norm": 1.0184043645858765,
+      "learning_rate": 4.494226369205147e-05,
+      "loss": 0.4318,
+      "step": 1563
+    },
+    {
+      "epoch": 0.9516276239732279,
+      "grad_norm": 1.053466796875,
+      "learning_rate": 4.493475372740916e-05,
+      "loss": 0.4011,
+      "step": 1564
+    },
+    {
+      "epoch": 0.9522360815333131,
+      "grad_norm": 1.0696157217025757,
+      "learning_rate": 4.492723881977885e-05,
+      "loss": 0.4152,
+      "step": 1565
+    },
+    {
+      "epoch": 0.9528445390933983,
+      "grad_norm": 1.0308218002319336,
+      "learning_rate": 4.4919718971023926e-05,
+      "loss": 0.3829,
+      "step": 1566
+    },
+    {
+      "epoch": 0.9534529966534834,
+      "grad_norm": 1.0396000146865845,
+      "learning_rate": 4.4912194183008994e-05,
+      "loss": 0.3648,
+      "step": 1567
+    },
+    {
+      "epoch": 0.9540614542135686,
+      "grad_norm": 1.12269926071167,
+      "learning_rate": 4.490466445759988e-05,
+      "loss": 0.4426,
+      "step": 1568
+    },
+    {
+      "epoch": 0.9546699117736538,
+      "grad_norm": 1.1220544576644897,
+      "learning_rate": 4.489712979666365e-05,
+      "loss": 0.3841,
+      "step": 1569
+    },
+    {
+      "epoch": 0.955278369333739,
+      "grad_norm": 0.9514595866203308,
+      "learning_rate": 4.4889590202068584e-05,
+      "loss": 0.384,
+      "step": 1570
+    },
+    {
+      "epoch": 0.9558868268938242,
+      "grad_norm": 0.9942927956581116,
+      "learning_rate": 4.4882045675684184e-05,
+      "loss": 0.3854,
+      "step": 1571
+    },
+    {
+      "epoch": 0.9564952844539093,
+      "grad_norm": 1.0529226064682007,
+      "learning_rate": 4.487449621938118e-05,
+      "loss": 0.3861,
+      "step": 1572
+    },
+    {
+      "epoch": 0.9571037420139945,
+      "grad_norm": 1.1254863739013672,
+      "learning_rate": 4.486694183503153e-05,
+      "loss": 0.4056,
+      "step": 1573
+    },
+    {
+      "epoch": 0.9577121995740797,
+      "grad_norm": 0.9550037980079651,
+      "learning_rate": 4.48593825245084e-05,
+      "loss": 0.3541,
+      "step": 1574
+    },
+    {
+      "epoch": 0.9583206571341649,
+      "grad_norm": 1.0226643085479736,
+      "learning_rate": 4.4851818289686175e-05,
+      "loss": 0.4194,
+      "step": 1575
+    },
+    {
+      "epoch": 0.9589291146942501,
+      "grad_norm": 1.5648815631866455,
+      "learning_rate": 4.484424913244049e-05,
+      "loss": 0.3899,
+      "step": 1576
+    },
+    {
+      "epoch": 0.9595375722543352,
+      "grad_norm": 1.0767285823822021,
+      "learning_rate": 4.4836675054648156e-05,
+      "loss": 0.386,
+      "step": 1577
+    },
+    {
+      "epoch": 0.9601460298144204,
+      "grad_norm": 1.153304100036621,
+      "learning_rate": 4.482909605818725e-05,
+      "loss": 0.4087,
+      "step": 1578
+    },
+    {
+      "epoch": 0.9607544873745056,
+      "grad_norm": 0.8764035105705261,
+      "learning_rate": 4.482151214493704e-05,
+      "loss": 0.3476,
+      "step": 1579
+    },
+    {
+      "epoch": 0.9613629449345908,
+      "grad_norm": 2.1990628242492676,
+      "learning_rate": 4.4813923316778014e-05,
+      "loss": 0.43,
+      "step": 1580
+    },
+    {
+      "epoch": 0.9619714024946759,
+      "grad_norm": 1.129573106765747,
+      "learning_rate": 4.48063295755919e-05,
+      "loss": 0.4173,
+      "step": 1581
+    },
+    {
+      "epoch": 0.9625798600547611,
+      "grad_norm": 1.4812309741973877,
+      "learning_rate": 4.4798730923261614e-05,
+      "loss": 0.4175,
+      "step": 1582
+    },
+    {
+      "epoch": 0.9631883176148464,
+      "grad_norm": 0.9128961563110352,
+      "learning_rate": 4.4791127361671304e-05,
+      "loss": 0.3943,
+      "step": 1583
+    },
+    {
+      "epoch": 0.9637967751749316,
+      "grad_norm": 1.1068278551101685,
+      "learning_rate": 4.478351889270635e-05,
+      "loss": 0.4152,
+      "step": 1584
+    },
+    {
+      "epoch": 0.9644052327350168,
+      "grad_norm": 0.88494873046875,
+      "learning_rate": 4.477590551825333e-05,
+      "loss": 0.3745,
+      "step": 1585
+    },
+    {
+      "epoch": 0.9650136902951019,
+      "grad_norm": 1.0295628309249878,
+      "learning_rate": 4.476828724020004e-05,
+      "loss": 0.4369,
+      "step": 1586
+    },
+    {
+      "epoch": 0.9656221478551871,
+      "grad_norm": 0.9472355246543884,
+      "learning_rate": 4.47606640604355e-05,
+      "loss": 0.3364,
+      "step": 1587
+    },
+    {
+      "epoch": 0.9662306054152723,
+      "grad_norm": 0.9680747389793396,
+      "learning_rate": 4.4753035980849935e-05,
+      "loss": 0.3878,
+      "step": 1588
+    },
+    {
+      "epoch": 0.9668390629753575,
+      "grad_norm": 0.9909029603004456,
+      "learning_rate": 4.4745403003334784e-05,
+      "loss": 0.3365,
+      "step": 1589
+    },
+    {
+      "epoch": 0.9674475205354427,
+      "grad_norm": 1.3338682651519775,
+      "learning_rate": 4.4737765129782735e-05,
+      "loss": 0.3644,
+      "step": 1590
+    },
+    {
+      "epoch": 0.9680559780955278,
+      "grad_norm": 0.9651134014129639,
+      "learning_rate": 4.473012236208763e-05,
+      "loss": 0.3977,
+      "step": 1591
+    },
+    {
+      "epoch": 0.968664435655613,
+      "grad_norm": 1.1207659244537354,
+      "learning_rate": 4.472247470214458e-05,
+      "loss": 0.395,
+      "step": 1592
+    },
+    {
+      "epoch": 0.9692728932156982,
+      "grad_norm": 1.0144009590148926,
+      "learning_rate": 4.471482215184988e-05,
+      "loss": 0.3607,
+      "step": 1593
+    },
+    {
+      "epoch": 0.9698813507757834,
+      "grad_norm": 0.911418616771698,
+      "learning_rate": 4.470716471310103e-05,
+      "loss": 0.3452,
+      "step": 1594
+    },
+    {
+      "epoch": 0.9704898083358686,
+      "grad_norm": 0.9752874374389648,
+      "learning_rate": 4.469950238779677e-05,
+      "loss": 0.3708,
+      "step": 1595
+    },
+    {
+      "epoch": 0.9710982658959537,
+      "grad_norm": 0.8316762447357178,
+      "learning_rate": 4.469183517783704e-05,
+      "loss": 0.3127,
+      "step": 1596
+    },
+    {
+      "epoch": 0.9717067234560389,
+      "grad_norm": 0.9781742095947266,
+      "learning_rate": 4.4684163085122976e-05,
+      "loss": 0.3197,
+      "step": 1597
+    },
+    {
+      "epoch": 0.9723151810161241,
+      "grad_norm": 1.1461424827575684,
+      "learning_rate": 4.4676486111556936e-05,
+      "loss": 0.3754,
+      "step": 1598
+    },
+    {
+      "epoch": 0.9729236385762093,
+      "grad_norm": 1.0407254695892334,
+      "learning_rate": 4.466880425904251e-05,
+      "loss": 0.4198,
+      "step": 1599
+    },
+    {
+      "epoch": 0.9735320961362945,
+      "grad_norm": 0.9534010887145996,
+      "learning_rate": 4.466111752948446e-05,
+      "loss": 0.3313,
+      "step": 1600
+    },
+    {
+      "epoch": 0.9741405536963796,
+      "grad_norm": 1.008195400238037,
+      "learning_rate": 4.465342592478878e-05,
+      "loss": 0.366,
+      "step": 1601
+    },
+    {
+      "epoch": 0.9747490112564648,
+      "grad_norm": 0.928765058517456,
+      "learning_rate": 4.464572944686266e-05,
+      "loss": 0.3674,
+      "step": 1602
+    },
+    {
+      "epoch": 0.97535746881655,
+      "grad_norm": 1.0775413513183594,
+      "learning_rate": 4.4638028097614515e-05,
+      "loss": 0.448,
+      "step": 1603
+    },
+    {
+      "epoch": 0.9759659263766353,
+      "grad_norm": 0.9224221706390381,
+      "learning_rate": 4.463032187895395e-05,
+      "loss": 0.3466,
+      "step": 1604
+    },
+    {
+      "epoch": 0.9765743839367204,
+      "grad_norm": 1.0811947584152222,
+      "learning_rate": 4.46226107927918e-05,
+      "loss": 0.4303,
+      "step": 1605
+    },
+    {
+      "epoch": 0.9771828414968056,
+      "grad_norm": 1.023656964302063,
+      "learning_rate": 4.4614894841040076e-05,
+      "loss": 0.3192,
+      "step": 1606
+    },
+    {
+      "epoch": 0.9777912990568908,
+      "grad_norm": 1.0087846517562866,
+      "learning_rate": 4.460717402561203e-05,
+      "loss": 0.3878,
+      "step": 1607
+    },
+    {
+      "epoch": 0.978399756616976,
+      "grad_norm": 1.254363775253296,
+      "learning_rate": 4.4599448348422087e-05,
+      "loss": 0.4048,
+      "step": 1608
+    },
+    {
+      "epoch": 0.9790082141770612,
+      "grad_norm": 1.086838722229004,
+      "learning_rate": 4.45917178113859e-05,
+      "loss": 0.3949,
+      "step": 1609
+    },
+    {
+      "epoch": 0.9796166717371463,
+      "grad_norm": 1.0660868883132935,
+      "learning_rate": 4.458398241642032e-05,
+      "loss": 0.3275,
+      "step": 1610
+    },
+    {
+      "epoch": 0.9802251292972315,
+      "grad_norm": 1.0936857461929321,
+      "learning_rate": 4.4576242165443394e-05,
+      "loss": 0.4235,
+      "step": 1611
+    },
+    {
+      "epoch": 0.9808335868573167,
+      "grad_norm": 1.1144509315490723,
+      "learning_rate": 4.456849706037439e-05,
+      "loss": 0.4732,
+      "step": 1612
+    },
+    {
+      "epoch": 0.9814420444174019,
+      "grad_norm": 1.0750304460525513,
+      "learning_rate": 4.456074710313378e-05,
+      "loss": 0.3902,
+      "step": 1613
+    },
+    {
+      "epoch": 0.9820505019774871,
+      "grad_norm": 1.7120141983032227,
+      "learning_rate": 4.455299229564321e-05,
+      "loss": 0.5309,
+      "step": 1614
+    },
+    {
+      "epoch": 0.9826589595375722,
+      "grad_norm": 1.0807467699050903,
+      "learning_rate": 4.454523263982557e-05,
+      "loss": 0.4076,
+      "step": 1615
+    },
+    {
+      "epoch": 0.9832674170976574,
+      "grad_norm": 1.0277988910675049,
+      "learning_rate": 4.453746813760492e-05,
+      "loss": 0.3838,
+      "step": 1616
+    },
+    {
+      "epoch": 0.9838758746577426,
+      "grad_norm": 1.0415362119674683,
+      "learning_rate": 4.452969879090653e-05,
+      "loss": 0.3689,
+      "step": 1617
+    },
+    {
+      "epoch": 0.9844843322178278,
+      "grad_norm": 1.0815787315368652,
+      "learning_rate": 4.452192460165687e-05,
+      "loss": 0.337,
+      "step": 1618
+    },
+    {
+      "epoch": 0.985092789777913,
+      "grad_norm": 1.2625021934509277,
+      "learning_rate": 4.451414557178363e-05,
+      "loss": 0.3757,
+      "step": 1619
+    },
+    {
+      "epoch": 0.9857012473379981,
+      "grad_norm": 1.092826247215271,
+      "learning_rate": 4.450636170321568e-05,
+      "loss": 0.3752,
+      "step": 1620
+    },
+    {
+      "epoch": 0.9863097048980833,
+      "grad_norm": 1.0274910926818848,
+      "learning_rate": 4.449857299788309e-05,
+      "loss": 0.3388,
+      "step": 1621
+    },
+    {
+      "epoch": 0.9869181624581685,
+      "grad_norm": 0.9931478500366211,
+      "learning_rate": 4.449077945771714e-05,
+      "loss": 0.366,
+      "step": 1622
+    },
+    {
+      "epoch": 0.9875266200182538,
+      "grad_norm": 1.035408616065979,
+      "learning_rate": 4.44829810846503e-05,
+      "loss": 0.3855,
+      "step": 1623
+    },
+    {
+      "epoch": 0.988135077578339,
+      "grad_norm": 1.1031928062438965,
+      "learning_rate": 4.447517788061624e-05,
+      "loss": 0.4198,
+      "step": 1624
+    },
+    {
+      "epoch": 0.9887435351384241,
+      "grad_norm": 1.122717261314392,
+      "learning_rate": 4.446736984754982e-05,
+      "loss": 0.4499,
+      "step": 1625
+    },
+    {
+      "epoch": 0.9893519926985093,
+      "grad_norm": 0.972223699092865,
+      "learning_rate": 4.445955698738714e-05,
+      "loss": 0.4174,
+      "step": 1626
+    },
+    {
+      "epoch": 0.9899604502585945,
+      "grad_norm": 0.8995710611343384,
+      "learning_rate": 4.445173930206543e-05,
+      "loss": 0.3257,
+      "step": 1627
+    },
+    {
+      "epoch": 0.9905689078186797,
+      "grad_norm": 0.9216899871826172,
+      "learning_rate": 4.444391679352315e-05,
+      "loss": 0.3732,
+      "step": 1628
+    },
+    {
+      "epoch": 0.9911773653787649,
+      "grad_norm": 1.138875961303711,
+      "learning_rate": 4.4436089463699984e-05,
+      "loss": 0.4355,
+      "step": 1629
+    },
+    {
+      "epoch": 0.99178582293885,
+      "grad_norm": 0.9231753945350647,
+      "learning_rate": 4.442825731453676e-05,
+      "loss": 0.3415,
+      "step": 1630
+    },
+    {
+      "epoch": 0.9923942804989352,
+      "grad_norm": 1.0919394493103027,
+      "learning_rate": 4.4420420347975535e-05,
+      "loss": 0.3939,
+      "step": 1631
+    },
+    {
+      "epoch": 0.9930027380590204,
+      "grad_norm": 1.25301194190979,
+      "learning_rate": 4.4412578565959554e-05,
+      "loss": 0.3438,
+      "step": 1632
+    },
+    {
+      "epoch": 0.9936111956191056,
+      "grad_norm": 1.0430591106414795,
+      "learning_rate": 4.440473197043323e-05,
+      "loss": 0.4462,
+      "step": 1633
+    },
+    {
+      "epoch": 0.9942196531791907,
+      "grad_norm": 0.9053996205329895,
+      "learning_rate": 4.439688056334221e-05,
+      "loss": 0.3515,
+      "step": 1634
+    },
+    {
+      "epoch": 0.9948281107392759,
+      "grad_norm": 1.0166853666305542,
+      "learning_rate": 4.438902434663331e-05,
+      "loss": 0.3846,
+      "step": 1635
+    },
+    {
+      "epoch": 0.9954365682993611,
+      "grad_norm": 1.1809371709823608,
+      "learning_rate": 4.438116332225456e-05,
+      "loss": 0.4166,
+      "step": 1636
+    },
+    {
+      "epoch": 0.9960450258594463,
+      "grad_norm": 0.8665146231651306,
+      "learning_rate": 4.437329749215514e-05,
+      "loss": 0.3229,
+      "step": 1637
+    },
+    {
+      "epoch": 0.9966534834195315,
+      "grad_norm": 0.9558749198913574,
+      "learning_rate": 4.4365426858285466e-05,
+      "loss": 0.3278,
+      "step": 1638
+    },
+    {
+      "epoch": 0.9972619409796166,
+      "grad_norm": 0.9763336777687073,
+      "learning_rate": 4.435755142259712e-05,
+      "loss": 0.3675,
+      "step": 1639
+    },
+    {
+      "epoch": 0.9978703985397018,
+      "grad_norm": 1.078508734703064,
+      "learning_rate": 4.434967118704289e-05,
+      "loss": 0.3435,
+      "step": 1640
+    },
+    {
+      "epoch": 0.998478856099787,
+      "grad_norm": 1.0996962785720825,
+      "learning_rate": 4.434178615357673e-05,
+      "loss": 0.4125,
+      "step": 1641
+    },
+    {
+      "epoch": 0.9990873136598722,
+      "grad_norm": 1.0043129920959473,
+      "learning_rate": 4.433389632415381e-05,
+      "loss": 0.4054,
+      "step": 1642
+    },
+    {
+      "epoch": 0.9996957712199575,
+      "grad_norm": 1.0766171216964722,
+      "learning_rate": 4.432600170073048e-05,
+      "loss": 0.3377,
+      "step": 1643
+    },
+    {
+      "epoch": 0.9996957712199575,
+      "eval_loss": 1.0254058837890625,
+      "eval_runtime": 108.4457,
+      "eval_samples_per_second": 7.027,
+      "eval_steps_per_second": 0.443,
+      "step": 1643
+    },
+    {
+      "epoch": 1.0003042287800425,
+      "grad_norm": 0.9912607073783875,
+      "learning_rate": 4.431810228526428e-05,
+      "loss": 0.3113,
+      "step": 1644
+    },
+    {
+      "epoch": 1.0009126863401279,
+      "grad_norm": 0.9680461287498474,
+      "learning_rate": 4.431019807971393e-05,
+      "loss": 0.3006,
+      "step": 1645
+    },
+    {
+      "epoch": 1.001521143900213,
+      "grad_norm": 0.9305565357208252,
+      "learning_rate": 4.430228908603934e-05,
+      "loss": 0.2665,
+      "step": 1646
+    },
+    {
+      "epoch": 1.002129601460298,
+      "grad_norm": 0.9174848198890686,
+      "learning_rate": 4.4294375306201617e-05,
+      "loss": 0.2765,
+      "step": 1647
+    },
+    {
+      "epoch": 1.0027380590203834,
+      "grad_norm": 0.8088688254356384,
+      "learning_rate": 4.4286456742163055e-05,
+      "loss": 0.2378,
+      "step": 1648
+    },
+    {
+      "epoch": 1.0033465165804685,
+      "grad_norm": 1.254451870918274,
+      "learning_rate": 4.427853339588711e-05,
+      "loss": 0.2796,
+      "step": 1649
+    },
+    {
+      "epoch": 1.0039549741405538,
+      "grad_norm": 0.9241443872451782,
+      "learning_rate": 4.427060526933846e-05,
+      "loss": 0.2814,
+      "step": 1650
+    },
+    {
+      "epoch": 1.0045634317006389,
+      "grad_norm": 1.0514885187149048,
+      "learning_rate": 4.4262672364482935e-05,
+      "loss": 0.2402,
+      "step": 1651
+    },
+    {
+      "epoch": 1.005171889260724,
+      "grad_norm": 0.931946337223053,
+      "learning_rate": 4.4254734683287575e-05,
+      "loss": 0.2502,
+      "step": 1652
+    },
+    {
+      "epoch": 1.0057803468208093,
+      "grad_norm": 2.174992561340332,
+      "learning_rate": 4.424679222772059e-05,
+      "loss": 0.2845,
+      "step": 1653
+    },
+    {
+      "epoch": 1.0063888043808944,
+      "grad_norm": 0.9909196496009827,
+      "learning_rate": 4.4238844999751376e-05,
+      "loss": 0.2355,
+      "step": 1654
+    },
+    {
+      "epoch": 1.0069972619409797,
+      "grad_norm": 0.9932165741920471,
+      "learning_rate": 4.423089300135052e-05,
+      "loss": 0.2375,
+      "step": 1655
+    },
+    {
+      "epoch": 1.0076057195010648,
+      "grad_norm": 1.1847784519195557,
+      "learning_rate": 4.422293623448978e-05,
+      "loss": 0.2185,
+      "step": 1656
+    },
+    {
+      "epoch": 1.00821417706115,
+      "grad_norm": 0.9104436039924622,
+      "learning_rate": 4.421497470114211e-05,
+      "loss": 0.2347,
+      "step": 1657
+    },
+    {
+      "epoch": 1.0088226346212352,
+      "grad_norm": 1.0744693279266357,
+      "learning_rate": 4.420700840328162e-05,
+      "loss": 0.2824,
+      "step": 1658
+    },
+    {
+      "epoch": 1.0094310921813203,
+      "grad_norm": 1.0241225957870483,
+      "learning_rate": 4.4199037342883656e-05,
+      "loss": 0.2483,
+      "step": 1659
+    },
+    {
+      "epoch": 1.0100395497414056,
+      "grad_norm": 1.3999607563018799,
+      "learning_rate": 4.419106152192467e-05,
+      "loss": 0.2848,
+      "step": 1660
+    },
+    {
+      "epoch": 1.0106480073014907,
+      "grad_norm": 1.127521276473999,
+      "learning_rate": 4.418308094238235e-05,
+      "loss": 0.2263,
+      "step": 1661
+    },
+    {
+      "epoch": 1.0112564648615758,
+      "grad_norm": 1.0944982767105103,
+      "learning_rate": 4.417509560623555e-05,
+      "loss": 0.2932,
+      "step": 1662
+    },
+    {
+      "epoch": 1.0118649224216612,
+      "grad_norm": 0.9151603579521179,
+      "learning_rate": 4.4167105515464305e-05,
+      "loss": 0.2578,
+      "step": 1663
+    },
+    {
+      "epoch": 1.0124733799817462,
+      "grad_norm": 1.0785083770751953,
+      "learning_rate": 4.415911067204981e-05,
+      "loss": 0.2092,
+      "step": 1664
+    },
+    {
+      "epoch": 1.0130818375418316,
+      "grad_norm": 1.1762471199035645,
+      "learning_rate": 4.415111107797445e-05,
+      "loss": 0.3176,
+      "step": 1665
+    },
+    {
+      "epoch": 1.0136902951019167,
+      "grad_norm": 0.9114542007446289,
+      "learning_rate": 4.414310673522181e-05,
+      "loss": 0.1926,
+      "step": 1666
+    },
+    {
+      "epoch": 1.0142987526620018,
+      "grad_norm": 1.0898659229278564,
+      "learning_rate": 4.413509764577663e-05,
+      "loss": 0.2444,
+      "step": 1667
+    },
+    {
+      "epoch": 1.014907210222087,
+      "grad_norm": 1.3012317419052124,
+      "learning_rate": 4.412708381162481e-05,
+      "loss": 0.2119,
+      "step": 1668
+    },
+    {
+      "epoch": 1.0155156677821722,
+      "grad_norm": 0.9848833680152893,
+      "learning_rate": 4.411906523475347e-05,
+      "loss": 0.2507,
+      "step": 1669
+    },
+    {
+      "epoch": 1.0161241253422575,
+      "grad_norm": 1.0226378440856934,
+      "learning_rate": 4.411104191715087e-05,
+      "loss": 0.2876,
+      "step": 1670
+    },
+    {
+      "epoch": 1.0167325829023426,
+      "grad_norm": 0.9476386308670044,
+      "learning_rate": 4.410301386080646e-05,
+      "loss": 0.209,
+      "step": 1671
+    },
+    {
+      "epoch": 1.0173410404624277,
+      "grad_norm": 1.0117278099060059,
+      "learning_rate": 4.4094981067710864e-05,
+      "loss": 0.2523,
+      "step": 1672
+    },
+    {
+      "epoch": 1.017949498022513,
+      "grad_norm": 1.3588906526565552,
+      "learning_rate": 4.408694353985589e-05,
+      "loss": 0.2786,
+      "step": 1673
+    },
+    {
+      "epoch": 1.018557955582598,
+      "grad_norm": 1.4190876483917236,
+      "learning_rate": 4.40789012792345e-05,
+      "loss": 0.2771,
+      "step": 1674
+    },
+    {
+      "epoch": 1.0191664131426832,
+      "grad_norm": 1.1058770418167114,
+      "learning_rate": 4.4070854287840836e-05,
+      "loss": 0.2673,
+      "step": 1675
+    },
+    {
+      "epoch": 1.0197748707027685,
+      "grad_norm": 0.8771906495094299,
+      "learning_rate": 4.406280256767022e-05,
+      "loss": 0.2254,
+      "step": 1676
+    },
+    {
+      "epoch": 1.0203833282628536,
+      "grad_norm": 0.8889464735984802,
+      "learning_rate": 4.4054746120719145e-05,
+      "loss": 0.2608,
+      "step": 1677
+    },
+    {
+      "epoch": 1.020991785822939,
+      "grad_norm": 0.8301272392272949,
+      "learning_rate": 4.404668494898527e-05,
+      "loss": 0.2089,
+      "step": 1678
+    },
+    {
+      "epoch": 1.021600243383024,
+      "grad_norm": 0.9632772207260132,
+      "learning_rate": 4.403861905446744e-05,
+      "loss": 0.2448,
+      "step": 1679
+    },
+    {
+      "epoch": 1.0222087009431091,
+      "grad_norm": 0.8604543209075928,
+      "learning_rate": 4.4030548439165654e-05,
+      "loss": 0.254,
+      "step": 1680
+    },
+    {
+      "epoch": 1.0228171585031944,
+      "grad_norm": 0.9880070090293884,
+      "learning_rate": 4.402247310508108e-05,
+      "loss": 0.2649,
+      "step": 1681
+    },
+    {
+      "epoch": 1.0234256160632795,
+      "grad_norm": 1.0415153503417969,
+      "learning_rate": 4.4014393054216076e-05,
+      "loss": 0.243,
+      "step": 1682
+    },
+    {
+      "epoch": 1.0240340736233648,
+      "grad_norm": 1.1902730464935303,
+      "learning_rate": 4.400630828857415e-05,
+      "loss": 0.2617,
+      "step": 1683
+    },
+    {
+      "epoch": 1.02464253118345,
+      "grad_norm": 0.9809486269950867,
+      "learning_rate": 4.3998218810159995e-05,
+      "loss": 0.2415,
+      "step": 1684
+    },
+    {
+      "epoch": 1.025250988743535,
+      "grad_norm": 1.0021169185638428,
+      "learning_rate": 4.3990124620979454e-05,
+      "loss": 0.2443,
+      "step": 1685
+    },
+    {
+      "epoch": 1.0258594463036204,
+      "grad_norm": 0.8601137399673462,
+      "learning_rate": 4.3982025723039564e-05,
+      "loss": 0.2291,
+      "step": 1686
+    },
+    {
+      "epoch": 1.0264679038637055,
+      "grad_norm": 1.0233848094940186,
+      "learning_rate": 4.3973922118348497e-05,
+      "loss": 0.2482,
+      "step": 1687
+    },
+    {
+      "epoch": 1.0270763614237908,
+      "grad_norm": 1.0119305849075317,
+      "learning_rate": 4.396581380891562e-05,
+      "loss": 0.2517,
+      "step": 1688
+    },
+    {
+      "epoch": 1.0276848189838759,
+      "grad_norm": 0.9090347290039062,
+      "learning_rate": 4.3957700796751446e-05,
+      "loss": 0.2357,
+      "step": 1689
+    },
+    {
+      "epoch": 1.028293276543961,
+      "grad_norm": 0.8543658256530762,
+      "learning_rate": 4.394958308386768e-05,
+      "loss": 0.2061,
+      "step": 1690
+    },
+    {
+      "epoch": 1.0289017341040463,
+      "grad_norm": 1.0883513689041138,
+      "learning_rate": 4.394146067227716e-05,
+      "loss": 0.2704,
+      "step": 1691
+    },
+    {
+      "epoch": 1.0295101916641314,
+      "grad_norm": 0.9570848345756531,
+      "learning_rate": 4.393333356399391e-05,
+      "loss": 0.2509,
+      "step": 1692
+    },
+    {
+      "epoch": 1.0301186492242167,
+      "grad_norm": 0.9206425547599792,
+      "learning_rate": 4.3925201761033116e-05,
+      "loss": 0.2453,
+      "step": 1693
+    },
+    {
+      "epoch": 1.0307271067843018,
+      "grad_norm": 0.9678740501403809,
+      "learning_rate": 4.391706526541114e-05,
+      "loss": 0.2438,
+      "step": 1694
+    },
+    {
+      "epoch": 1.031335564344387,
+      "grad_norm": 0.9064133763313293,
+      "learning_rate": 4.390892407914547e-05,
+      "loss": 0.2221,
+      "step": 1695
+    },
+    {
+      "epoch": 1.0319440219044722,
+      "grad_norm": 1.0082942247390747,
+      "learning_rate": 4.390077820425479e-05,
+      "loss": 0.2859,
+      "step": 1696
+    },
+    {
+      "epoch": 1.0325524794645573,
+      "grad_norm": 0.9536372423171997,
+      "learning_rate": 4.3892627642758946e-05,
+      "loss": 0.2616,
+      "step": 1697
+    },
+    {
+      "epoch": 1.0331609370246426,
+      "grad_norm": 0.9097919464111328,
+      "learning_rate": 4.388447239667892e-05,
+      "loss": 0.2647,
+      "step": 1698
+    },
+    {
+      "epoch": 1.0337693945847277,
+      "grad_norm": 0.9455602765083313,
+      "learning_rate": 4.387631246803689e-05,
+      "loss": 0.2271,
+      "step": 1699
+    },
+    {
+      "epoch": 1.0343778521448128,
+      "grad_norm": 1.0754998922348022,
+      "learning_rate": 4.386814785885617e-05,
+      "loss": 0.2671,
+      "step": 1700
+    },
+    {
+      "epoch": 1.0349863097048981,
+      "grad_norm": 0.8816566467285156,
+      "learning_rate": 4.3859978571161245e-05,
+      "loss": 0.2149,
+      "step": 1701
+    },
+    {
+      "epoch": 1.0355947672649832,
+      "grad_norm": 0.8995217084884644,
+      "learning_rate": 4.3851804606977756e-05,
+      "loss": 0.2361,
+      "step": 1702
+    },
+    {
+      "epoch": 1.0362032248250685,
+      "grad_norm": 1.0205188989639282,
+      "learning_rate": 4.38436259683325e-05,
+      "loss": 0.2671,
+      "step": 1703
+    },
+    {
+      "epoch": 1.0368116823851536,
+      "grad_norm": 0.9358905553817749,
+      "learning_rate": 4.383544265725346e-05,
+      "loss": 0.2259,
+      "step": 1704
+    },
+    {
+      "epoch": 1.0374201399452387,
+      "grad_norm": 1.6457610130310059,
+      "learning_rate": 4.382725467576973e-05,
+      "loss": 0.3296,
+      "step": 1705
+    },
+    {
+      "epoch": 1.038028597505324,
+      "grad_norm": 0.9523285627365112,
+      "learning_rate": 4.38190620259116e-05,
+      "loss": 0.2466,
+      "step": 1706
+    },
+    {
+      "epoch": 1.0386370550654092,
+      "grad_norm": 2.4622488021850586,
+      "learning_rate": 4.381086470971051e-05,
+      "loss": 0.2464,
+      "step": 1707
+    },
+    {
+      "epoch": 1.0392455126254945,
+      "grad_norm": 0.9736045002937317,
+      "learning_rate": 4.380266272919904e-05,
+      "loss": 0.226,
+      "step": 1708
+    },
+    {
+      "epoch": 1.0398539701855796,
+      "grad_norm": 0.9621555209159851,
+      "learning_rate": 4.379445608641095e-05,
+      "loss": 0.247,
+      "step": 1709
+    },
+    {
+      "epoch": 1.0404624277456647,
+      "grad_norm": 0.9200358390808105,
+      "learning_rate": 4.378624478338115e-05,
+      "loss": 0.2385,
+      "step": 1710
+    },
+    {
+      "epoch": 1.04107088530575,
+      "grad_norm": 0.921256422996521,
+      "learning_rate": 4.377802882214568e-05,
+      "loss": 0.2536,
+      "step": 1711
+    },
+    {
+      "epoch": 1.041679342865835,
+      "grad_norm": 1.0362918376922607,
+      "learning_rate": 4.3769808204741766e-05,
+      "loss": 0.2505,
+      "step": 1712
+    },
+    {
+      "epoch": 1.0422878004259204,
+      "grad_norm": 0.9670946002006531,
+      "learning_rate": 4.37615829332078e-05,
+      "loss": 0.2311,
+      "step": 1713
+    },
+    {
+      "epoch": 1.0428962579860055,
+      "grad_norm": 0.9415709376335144,
+      "learning_rate": 4.3753353009583275e-05,
+      "loss": 0.2073,
+      "step": 1714
+    },
+    {
+      "epoch": 1.0435047155460906,
+      "grad_norm": 0.9122322797775269,
+      "learning_rate": 4.374511843590888e-05,
+      "loss": 0.2956,
+      "step": 1715
+    },
+    {
+      "epoch": 1.044113173106176,
+      "grad_norm": 0.8133691549301147,
+      "learning_rate": 4.373687921422646e-05,
+      "loss": 0.2302,
+      "step": 1716
+    },
+    {
+      "epoch": 1.044721630666261,
+      "grad_norm": 0.9611904621124268,
+      "learning_rate": 4.372863534657897e-05,
+      "loss": 0.242,
+      "step": 1717
+    },
+    {
+      "epoch": 1.045330088226346,
+      "grad_norm": 1.105997085571289,
+      "learning_rate": 4.372038683501057e-05,
+      "loss": 0.2738,
+      "step": 1718
+    },
+    {
+      "epoch": 1.0459385457864314,
+      "grad_norm": 0.9072719216346741,
+      "learning_rate": 4.3712133681566546e-05,
+      "loss": 0.3021,
+      "step": 1719
+    },
+    {
+      "epoch": 1.0465470033465165,
+      "grad_norm": 0.9673424959182739,
+      "learning_rate": 4.370387588829332e-05,
+      "loss": 0.2756,
+      "step": 1720
+    },
+    {
+      "epoch": 1.0471554609066018,
+      "grad_norm": 0.8735299706459045,
+      "learning_rate": 4.369561345723849e-05,
+      "loss": 0.244,
+      "step": 1721
+    },
+    {
+      "epoch": 1.047763918466687,
+      "grad_norm": 0.8752166628837585,
+      "learning_rate": 4.36873463904508e-05,
+      "loss": 0.2251,
+      "step": 1722
+    },
+    {
+      "epoch": 1.048372376026772,
+      "grad_norm": 0.845640242099762,
+      "learning_rate": 4.367907468998013e-05,
+      "loss": 0.2487,
+      "step": 1723
+    },
+    {
+      "epoch": 1.0489808335868573,
+      "grad_norm": 0.846122682094574,
+      "learning_rate": 4.3670798357877515e-05,
+      "loss": 0.2061,
+      "step": 1724
+    },
+    {
+      "epoch": 1.0495892911469424,
+      "grad_norm": 0.9644899368286133,
+      "learning_rate": 4.366251739619515e-05,
+      "loss": 0.2853,
+      "step": 1725
+    },
+    {
+      "epoch": 1.0501977487070278,
+      "grad_norm": 1.052946925163269,
+      "learning_rate": 4.365423180698636e-05,
+      "loss": 0.2502,
+      "step": 1726
+    },
+    {
+      "epoch": 1.0508062062671129,
+      "grad_norm": 1.0697331428527832,
+      "learning_rate": 4.364594159230563e-05,
+      "loss": 0.2259,
+      "step": 1727
+    },
+    {
+      "epoch": 1.051414663827198,
+      "grad_norm": 0.9684371948242188,
+      "learning_rate": 4.363764675420858e-05,
+      "loss": 0.2298,
+      "step": 1728
+    },
+    {
+      "epoch": 1.0520231213872833,
+      "grad_norm": 0.9283509254455566,
+      "learning_rate": 4.3629347294752e-05,
+      "loss": 0.2221,
+      "step": 1729
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.9486536383628845,
+      "learning_rate": 4.3621043215993793e-05,
+      "loss": 0.2321,
+      "step": 1730
+    },
+    {
+      "epoch": 1.0532400365074537,
+      "grad_norm": 0.9327476024627686,
+      "learning_rate": 4.3612734519993035e-05,
+      "loss": 0.2353,
+      "step": 1731
+    },
+    {
+      "epoch": 1.0538484940675388,
+      "grad_norm": 0.9721671342849731,
+      "learning_rate": 4.360442120880994e-05,
+      "loss": 0.2255,
+      "step": 1732
+    },
+    {
+      "epoch": 1.0544569516276239,
+      "grad_norm": 1.0690611600875854,
+      "learning_rate": 4.3596103284505854e-05,
+      "loss": 0.2511,
+      "step": 1733
+    },
+    {
+      "epoch": 1.0550654091877092,
+      "grad_norm": 0.9945971965789795,
+      "learning_rate": 4.358778074914326e-05,
+      "loss": 0.2729,
+      "step": 1734
+    },
+    {
+      "epoch": 1.0556738667477943,
+      "grad_norm": 1.0413203239440918,
+      "learning_rate": 4.357945360478584e-05,
+      "loss": 0.2918,
+      "step": 1735
+    },
+    {
+      "epoch": 1.0562823243078796,
+      "grad_norm": 0.8384091854095459,
+      "learning_rate": 4.357112185349834e-05,
+      "loss": 0.2324,
+      "step": 1736
+    },
+    {
+      "epoch": 1.0568907818679647,
+      "grad_norm": 1.018020510673523,
+      "learning_rate": 4.3562785497346706e-05,
+      "loss": 0.262,
+      "step": 1737
+    },
+    {
+      "epoch": 1.0574992394280498,
+      "grad_norm": 0.9685660004615784,
+      "learning_rate": 4.355444453839801e-05,
+      "loss": 0.2684,
+      "step": 1738
+    },
+    {
+      "epoch": 1.0581076969881351,
+      "grad_norm": 1.006584882736206,
+      "learning_rate": 4.354609897872044e-05,
+      "loss": 0.2912,
+      "step": 1739
+    },
+    {
+      "epoch": 1.0587161545482202,
+      "grad_norm": 0.9230665564537048,
+      "learning_rate": 4.3537748820383386e-05,
+      "loss": 0.2542,
+      "step": 1740
+    },
+    {
+      "epoch": 1.0593246121083055,
+      "grad_norm": 0.9952296018600464,
+      "learning_rate": 4.35293940654573e-05,
+      "loss": 0.2588,
+      "step": 1741
+    },
+    {
+      "epoch": 1.0599330696683906,
+      "grad_norm": 0.8689709305763245,
+      "learning_rate": 4.352103471601383e-05,
+      "loss": 0.2159,
+      "step": 1742
+    },
+    {
+      "epoch": 1.0605415272284757,
+      "grad_norm": 0.9874710440635681,
+      "learning_rate": 4.351267077412575e-05,
+      "loss": 0.2766,
+      "step": 1743
+    },
+    {
+      "epoch": 1.061149984788561,
+      "grad_norm": 0.9209951758384705,
+      "learning_rate": 4.350430224186696e-05,
+      "loss": 0.2918,
+      "step": 1744
+    },
+    {
+      "epoch": 1.0617584423486461,
+      "grad_norm": 0.9789804220199585,
+      "learning_rate": 4.349592912131252e-05,
+      "loss": 0.26,
+      "step": 1745
+    },
+    {
+      "epoch": 1.0623668999087315,
+      "grad_norm": 0.963188111782074,
+      "learning_rate": 4.3487551414538595e-05,
+      "loss": 0.2295,
+      "step": 1746
+    },
+    {
+      "epoch": 1.0629753574688166,
+      "grad_norm": 0.8524497151374817,
+      "learning_rate": 4.347916912362252e-05,
+      "loss": 0.2094,
+      "step": 1747
+    },
+    {
+      "epoch": 1.0635838150289016,
+      "grad_norm": 1.0623359680175781,
+      "learning_rate": 4.347078225064276e-05,
+      "loss": 0.263,
+      "step": 1748
+    },
+    {
+      "epoch": 1.064192272588987,
+      "grad_norm": 1.013371467590332,
+      "learning_rate": 4.34623907976789e-05,
+      "loss": 0.2404,
+      "step": 1749
+    },
+    {
+      "epoch": 1.064800730149072,
+      "grad_norm": 0.8995924592018127,
+      "learning_rate": 4.345399476681167e-05,
+      "loss": 0.2167,
+      "step": 1750
+    },
+    {
+      "epoch": 1.0654091877091574,
+      "grad_norm": 1.011319875717163,
+      "learning_rate": 4.344559416012293e-05,
+      "loss": 0.2701,
+      "step": 1751
+    },
+    {
+      "epoch": 1.0660176452692425,
+      "grad_norm": 0.9735434055328369,
+      "learning_rate": 4.34371889796957e-05,
+      "loss": 0.2159,
+      "step": 1752
+    },
+    {
+      "epoch": 1.0666261028293276,
+      "grad_norm": 0.8756011128425598,
+      "learning_rate": 4.34287792276141e-05,
+      "loss": 0.2148,
+      "step": 1753
+    },
+    {
+      "epoch": 1.067234560389413,
+      "grad_norm": 0.8667939305305481,
+      "learning_rate": 4.34203649059634e-05,
+      "loss": 0.1809,
+      "step": 1754
+    },
+    {
+      "epoch": 1.067843017949498,
+      "grad_norm": 0.9129221439361572,
+      "learning_rate": 4.341194601683e-05,
+      "loss": 0.2344,
+      "step": 1755
+    },
+    {
+      "epoch": 1.0684514755095833,
+      "grad_norm": 0.952587902545929,
+      "learning_rate": 4.340352256230144e-05,
+      "loss": 0.2158,
+      "step": 1756
+    },
+    {
+      "epoch": 1.0690599330696684,
+      "grad_norm": 0.932098388671875,
+      "learning_rate": 4.339509454446637e-05,
+      "loss": 0.2488,
+      "step": 1757
+    },
+    {
+      "epoch": 1.0696683906297535,
+      "grad_norm": 1.0526845455169678,
+      "learning_rate": 4.338666196541461e-05,
+      "loss": 0.276,
+      "step": 1758
+    },
+    {
+      "epoch": 1.0702768481898388,
+      "grad_norm": 1.0784953832626343,
+      "learning_rate": 4.337822482723708e-05,
+      "loss": 0.2574,
+      "step": 1759
+    },
+    {
+      "epoch": 1.070885305749924,
+      "grad_norm": 1.0775099992752075,
+      "learning_rate": 4.336978313202583e-05,
+      "loss": 0.2513,
+      "step": 1760
+    },
+    {
+      "epoch": 1.0714937633100092,
+      "grad_norm": 0.990006685256958,
+      "learning_rate": 4.336133688187405e-05,
+      "loss": 0.237,
+      "step": 1761
+    },
+    {
+      "epoch": 1.0721022208700943,
+      "grad_norm": 0.9512050151824951,
+      "learning_rate": 4.3352886078876065e-05,
+      "loss": 0.2594,
+      "step": 1762
+    },
+    {
+      "epoch": 1.0727106784301794,
+      "grad_norm": 1.0135704278945923,
+      "learning_rate": 4.3344430725127315e-05,
+      "loss": 0.2292,
+      "step": 1763
+    },
+    {
+      "epoch": 1.0733191359902647,
+      "grad_norm": 1.0516897439956665,
+      "learning_rate": 4.333597082272438e-05,
+      "loss": 0.2289,
+      "step": 1764
+    },
+    {
+      "epoch": 1.0739275935503498,
+      "grad_norm": 0.9647794365882874,
+      "learning_rate": 4.332750637376496e-05,
+      "loss": 0.2488,
+      "step": 1765
+    },
+    {
+      "epoch": 1.0745360511104352,
+      "grad_norm": 0.84974604845047,
+      "learning_rate": 4.331903738034789e-05,
+      "loss": 0.2196,
+      "step": 1766
+    },
+    {
+      "epoch": 1.0751445086705202,
+      "grad_norm": 0.9360343813896179,
+      "learning_rate": 4.331056384457313e-05,
+      "loss": 0.2452,
+      "step": 1767
+    },
+    {
+      "epoch": 1.0757529662306053,
+      "grad_norm": 0.920875608921051,
+      "learning_rate": 4.330208576854176e-05,
+      "loss": 0.2618,
+      "step": 1768
+    },
+    {
+      "epoch": 1.0763614237906907,
+      "grad_norm": 0.9963610172271729,
+      "learning_rate": 4.3293603154355976e-05,
+      "loss": 0.2536,
+      "step": 1769
+    },
+    {
+      "epoch": 1.0769698813507758,
+      "grad_norm": 0.914340615272522,
+      "learning_rate": 4.328511600411913e-05,
+      "loss": 0.2487,
+      "step": 1770
+    },
+    {
+      "epoch": 1.077578338910861,
+      "grad_norm": 0.8099347949028015,
+      "learning_rate": 4.327662431993568e-05,
+      "loss": 0.2186,
+      "step": 1771
+    },
+    {
+      "epoch": 1.0781867964709462,
+      "grad_norm": 0.8634665608406067,
+      "learning_rate": 4.32681281039112e-05,
+      "loss": 0.2381,
+      "step": 1772
+    },
+    {
+      "epoch": 1.0787952540310313,
+      "grad_norm": 1.6494232416152954,
+      "learning_rate": 4.325962735815241e-05,
+      "loss": 0.3326,
+      "step": 1773
+    },
+    {
+      "epoch": 1.0794037115911166,
+      "grad_norm": 0.9181280732154846,
+      "learning_rate": 4.3251122084767124e-05,
+      "loss": 0.2321,
+      "step": 1774
+    },
+    {
+      "epoch": 1.0800121691512017,
+      "grad_norm": 1.1291999816894531,
+      "learning_rate": 4.324261228586431e-05,
+      "loss": 0.2762,
+      "step": 1775
+    },
+    {
+      "epoch": 1.080620626711287,
+      "grad_norm": 0.8536227941513062,
+      "learning_rate": 4.323409796355404e-05,
+      "loss": 0.1981,
+      "step": 1776
+    },
+    {
+      "epoch": 1.081229084271372,
+      "grad_norm": 0.9988884329795837,
+      "learning_rate": 4.322557911994751e-05,
+      "loss": 0.2943,
+      "step": 1777
+    },
+    {
+      "epoch": 1.0818375418314572,
+      "grad_norm": 0.9455403089523315,
+      "learning_rate": 4.321705575715703e-05,
+      "loss": 0.2459,
+      "step": 1778
+    },
+    {
+      "epoch": 1.0824459993915425,
+      "grad_norm": 0.9276781678199768,
+      "learning_rate": 4.320852787729606e-05,
+      "loss": 0.2779,
+      "step": 1779
+    },
+    {
+      "epoch": 1.0830544569516276,
+      "grad_norm": 1.0638751983642578,
+      "learning_rate": 4.319999548247914e-05,
+      "loss": 0.2595,
+      "step": 1780
+    },
+    {
+      "epoch": 1.0836629145117127,
+      "grad_norm": 0.9427651762962341,
+      "learning_rate": 4.3191458574821955e-05,
+      "loss": 0.2628,
+      "step": 1781
+    },
+    {
+      "epoch": 1.084271372071798,
+      "grad_norm": 0.9555470943450928,
+      "learning_rate": 4.3182917156441296e-05,
+      "loss": 0.245,
+      "step": 1782
+    },
+    {
+      "epoch": 1.0848798296318831,
+      "grad_norm": 0.9582526683807373,
+      "learning_rate": 4.317437122945508e-05,
+      "loss": 0.2842,
+      "step": 1783
+    },
+    {
+      "epoch": 1.0854882871919684,
+      "grad_norm": 0.9005835652351379,
+      "learning_rate": 4.316582079598235e-05,
+      "loss": 0.2646,
+      "step": 1784
+    },
+    {
+      "epoch": 1.0860967447520535,
+      "grad_norm": 0.7882102727890015,
+      "learning_rate": 4.3157265858143247e-05,
+      "loss": 0.2415,
+      "step": 1785
+    },
+    {
+      "epoch": 1.0867052023121386,
+      "grad_norm": 0.8828130960464478,
+      "learning_rate": 4.3148706418059046e-05,
+      "loss": 0.2327,
+      "step": 1786
+    },
+    {
+      "epoch": 1.087313659872224,
+      "grad_norm": 0.823904812335968,
+      "learning_rate": 4.314014247785214e-05,
+      "loss": 0.2377,
+      "step": 1787
+    },
+    {
+      "epoch": 1.087922117432309,
+      "grad_norm": 0.9788227677345276,
+      "learning_rate": 4.313157403964601e-05,
+      "loss": 0.2552,
+      "step": 1788
+    },
+    {
+      "epoch": 1.0885305749923944,
+      "grad_norm": 0.935871422290802,
+      "learning_rate": 4.312300110556527e-05,
+      "loss": 0.2474,
+      "step": 1789
+    },
+    {
+      "epoch": 1.0891390325524795,
+      "grad_norm": 17.13036346435547,
+      "learning_rate": 4.311442367773567e-05,
+      "loss": 0.2759,
+      "step": 1790
+    },
+    {
+      "epoch": 1.0897474901125646,
+      "grad_norm": 1.1542432308197021,
+      "learning_rate": 4.3105841758284035e-05,
+      "loss": 0.2905,
+      "step": 1791
+    },
+    {
+      "epoch": 1.0903559476726499,
+      "grad_norm": 1.1316038370132446,
+      "learning_rate": 4.3097255349338344e-05,
+      "loss": 0.2549,
+      "step": 1792
+    },
+    {
+      "epoch": 1.090964405232735,
+      "grad_norm": 1.0348681211471558,
+      "learning_rate": 4.308866445302766e-05,
+      "loss": 0.2649,
+      "step": 1793
+    },
+    {
+      "epoch": 1.0915728627928203,
+      "grad_norm": 1.0835968255996704,
+      "learning_rate": 4.308006907148215e-05,
+      "loss": 0.2424,
+      "step": 1794
+    },
+    {
+      "epoch": 1.0921813203529054,
+      "grad_norm": 0.9840952754020691,
+      "learning_rate": 4.307146920683313e-05,
+      "loss": 0.27,
+      "step": 1795
+    },
+    {
+      "epoch": 1.0927897779129905,
+      "grad_norm": 0.8699171543121338,
+      "learning_rate": 4.3062864861213e-05,
+      "loss": 0.2061,
+      "step": 1796
+    },
+    {
+      "epoch": 1.0933982354730758,
+      "grad_norm": 0.9339224100112915,
+      "learning_rate": 4.305425603675529e-05,
+      "loss": 0.2612,
+      "step": 1797
+    },
+    {
+      "epoch": 1.094006693033161,
+      "grad_norm": 1.0255519151687622,
+      "learning_rate": 4.304564273559462e-05,
+      "loss": 0.2652,
+      "step": 1798
+    },
+    {
+      "epoch": 1.0946151505932462,
+      "grad_norm": 1.1433568000793457,
+      "learning_rate": 4.303702495986672e-05,
+      "loss": 0.2684,
+      "step": 1799
+    },
+    {
+      "epoch": 1.0952236081533313,
+      "grad_norm": 1.2076388597488403,
+      "learning_rate": 4.302840271170846e-05,
+      "loss": 0.2964,
+      "step": 1800
+    },
+    {
+      "epoch": 1.0958320657134164,
+      "grad_norm": 0.9302932620048523,
+      "learning_rate": 4.301977599325779e-05,
+      "loss": 0.2291,
+      "step": 1801
+    },
+    {
+      "epoch": 1.0964405232735017,
+      "grad_norm": 0.9104456901550293,
+      "learning_rate": 4.301114480665377e-05,
+      "loss": 0.235,
+      "step": 1802
+    },
+    {
+      "epoch": 1.0970489808335868,
+      "grad_norm": 1.0696483850479126,
+      "learning_rate": 4.3002509154036585e-05,
+      "loss": 0.3616,
+      "step": 1803
+    },
+    {
+      "epoch": 1.0976574383936721,
+      "grad_norm": 0.8358986377716064,
+      "learning_rate": 4.299386903754751e-05,
+      "loss": 0.197,
+      "step": 1804
+    },
+    {
+      "epoch": 1.0982658959537572,
+      "grad_norm": 0.949712872505188,
+      "learning_rate": 4.2985224459328934e-05,
+      "loss": 0.2621,
+      "step": 1805
+    },
+    {
+      "epoch": 1.0988743535138423,
+      "grad_norm": 1.1388565301895142,
+      "learning_rate": 4.297657542152434e-05,
+      "loss": 0.3019,
+      "step": 1806
+    },
+    {
+      "epoch": 1.0994828110739276,
+      "grad_norm": 0.9658118486404419,
+      "learning_rate": 4.296792192627836e-05,
+      "loss": 0.2406,
+      "step": 1807
+    },
+    {
+      "epoch": 1.1000912686340127,
+      "grad_norm": 0.9098191261291504,
+      "learning_rate": 4.2959263975736676e-05,
+      "loss": 0.2166,
+      "step": 1808
+    },
+    {
+      "epoch": 1.100699726194098,
+      "grad_norm": 1.0185233354568481,
+      "learning_rate": 4.29506015720461e-05,
+      "loss": 0.2498,
+      "step": 1809
+    },
+    {
+      "epoch": 1.1013081837541832,
+      "grad_norm": 1.152592658996582,
+      "learning_rate": 4.294193471735456e-05,
+      "loss": 0.2417,
+      "step": 1810
+    },
+    {
+      "epoch": 1.1019166413142683,
+      "grad_norm": 0.8422519564628601,
+      "learning_rate": 4.2933263413811065e-05,
+      "loss": 0.2361,
+      "step": 1811
+    },
+    {
+      "epoch": 1.1025250988743536,
+      "grad_norm": 0.9988545179367065,
+      "learning_rate": 4.292458766356574e-05,
+      "loss": 0.273,
+      "step": 1812
+    },
+    {
+      "epoch": 1.1031335564344387,
+      "grad_norm": 1.068097472190857,
+      "learning_rate": 4.29159074687698e-05,
+      "loss": 0.2403,
+      "step": 1813
+    },
+    {
+      "epoch": 1.103742013994524,
+      "grad_norm": 1.0569621324539185,
+      "learning_rate": 4.290722283157559e-05,
+      "loss": 0.2446,
+      "step": 1814
+    },
+    {
+      "epoch": 1.104350471554609,
+      "grad_norm": 0.9275895357131958,
+      "learning_rate": 4.289853375413652e-05,
+      "loss": 0.2369,
+      "step": 1815
+    },
+    {
+      "epoch": 1.1049589291146942,
+      "grad_norm": 1.1227593421936035,
+      "learning_rate": 4.2889840238607135e-05,
+      "loss": 0.262,
+      "step": 1816
+    },
+    {
+      "epoch": 1.1055673866747795,
+      "grad_norm": 0.9638474583625793,
+      "learning_rate": 4.2881142287143044e-05,
+      "loss": 0.232,
+      "step": 1817
+    },
+    {
+      "epoch": 1.1061758442348646,
+      "grad_norm": 0.9380021691322327,
+      "learning_rate": 4.2872439901901e-05,
+      "loss": 0.244,
+      "step": 1818
+    },
+    {
+      "epoch": 1.1067843017949497,
+      "grad_norm": 0.9702509641647339,
+      "learning_rate": 4.286373308503881e-05,
+      "loss": 0.2451,
+      "step": 1819
+    },
+    {
+      "epoch": 1.107392759355035,
+      "grad_norm": 0.9465189576148987,
+      "learning_rate": 4.285502183871542e-05,
+      "loss": 0.2479,
+      "step": 1820
+    },
+    {
+      "epoch": 1.10800121691512,
+      "grad_norm": 0.9717510342597961,
+      "learning_rate": 4.284630616509084e-05,
+      "loss": 0.2178,
+      "step": 1821
+    },
+    {
+      "epoch": 1.1086096744752054,
+      "grad_norm": 0.9717493653297424,
+      "learning_rate": 4.283758606632621e-05,
+      "loss": 0.2601,
+      "step": 1822
+    },
+    {
+      "epoch": 1.1092181320352905,
+      "grad_norm": 1.4151792526245117,
+      "learning_rate": 4.2828861544583746e-05,
+      "loss": 0.29,
+      "step": 1823
+    },
+    {
+      "epoch": 1.1098265895953756,
+      "grad_norm": 1.0413273572921753,
+      "learning_rate": 4.282013260202675e-05,
+      "loss": 0.262,
+      "step": 1824
+    },
+    {
+      "epoch": 1.110435047155461,
+      "grad_norm": 1.031923532485962,
+      "learning_rate": 4.281139924081966e-05,
+      "loss": 0.2929,
+      "step": 1825
+    },
+    {
+      "epoch": 1.111043504715546,
+      "grad_norm": 0.9092714786529541,
+      "learning_rate": 4.280266146312797e-05,
+      "loss": 0.2703,
+      "step": 1826
+    },
+    {
+      "epoch": 1.1116519622756313,
+      "grad_norm": 1.0851198434829712,
+      "learning_rate": 4.279391927111828e-05,
+      "loss": 0.3064,
+      "step": 1827
+    },
+    {
+      "epoch": 1.1122604198357164,
+      "grad_norm": 0.839637279510498,
+      "learning_rate": 4.2785172666958305e-05,
+      "loss": 0.2491,
+      "step": 1828
+    },
+    {
+      "epoch": 1.1128688773958015,
+      "grad_norm": 0.8800391554832458,
+      "learning_rate": 4.2776421652816834e-05,
+      "loss": 0.2591,
+      "step": 1829
+    },
+    {
+      "epoch": 1.1134773349558869,
+      "grad_norm": 0.9919121861457825,
+      "learning_rate": 4.2767666230863743e-05,
+      "loss": 0.2876,
+      "step": 1830
+    },
+    {
+      "epoch": 1.114085792515972,
+      "grad_norm": 1.0813971757888794,
+      "learning_rate": 4.2758906403270026e-05,
+      "loss": 0.2622,
+      "step": 1831
+    },
+    {
+      "epoch": 1.1146942500760573,
+      "grad_norm": 0.9650895595550537,
+      "learning_rate": 4.275014217220775e-05,
+      "loss": 0.2185,
+      "step": 1832
+    },
+    {
+      "epoch": 1.1153027076361424,
+      "grad_norm": 0.9158915281295776,
+      "learning_rate": 4.2741373539850076e-05,
+      "loss": 0.2519,
+      "step": 1833
+    },
+    {
+      "epoch": 1.1159111651962275,
+      "grad_norm": 1.001535415649414,
+      "learning_rate": 4.273260050837126e-05,
+      "loss": 0.268,
+      "step": 1834
+    },
+    {
+      "epoch": 1.1165196227563128,
+      "grad_norm": 0.9505628347396851,
+      "learning_rate": 4.272382307994665e-05,
+      "loss": 0.2334,
+      "step": 1835
+    },
+    {
+      "epoch": 1.1171280803163979,
+      "grad_norm": 1.0092861652374268,
+      "learning_rate": 4.271504125675269e-05,
+      "loss": 0.2992,
+      "step": 1836
+    },
+    {
+      "epoch": 1.1177365378764832,
+      "grad_norm": 0.9263057708740234,
+      "learning_rate": 4.270625504096688e-05,
+      "loss": 0.2283,
+      "step": 1837
+    },
+    {
+      "epoch": 1.1183449954365683,
+      "grad_norm": 1.5665264129638672,
+      "learning_rate": 4.269746443476787e-05,
+      "loss": 0.235,
+      "step": 1838
+    },
+    {
+      "epoch": 1.1189534529966534,
+      "grad_norm": 1.0436450242996216,
+      "learning_rate": 4.268866944033533e-05,
+      "loss": 0.26,
+      "step": 1839
+    },
+    {
+      "epoch": 1.1195619105567387,
+      "grad_norm": 0.8918569684028625,
+      "learning_rate": 4.267987005985008e-05,
+      "loss": 0.2506,
+      "step": 1840
+    },
+    {
+      "epoch": 1.1201703681168238,
+      "grad_norm": 0.9875538349151611,
+      "learning_rate": 4.267106629549398e-05,
+      "loss": 0.2407,
+      "step": 1841
+    },
+    {
+      "epoch": 1.1207788256769091,
+      "grad_norm": 1.0079050064086914,
+      "learning_rate": 4.266225814945001e-05,
+      "loss": 0.3127,
+      "step": 1842
+    },
+    {
+      "epoch": 1.1213872832369942,
+      "grad_norm": 0.8947494029998779,
+      "learning_rate": 4.265344562390222e-05,
+      "loss": 0.2271,
+      "step": 1843
+    },
+    {
+      "epoch": 1.1219957407970793,
+      "grad_norm": 0.9922159314155579,
+      "learning_rate": 4.264462872103575e-05,
+      "loss": 0.2515,
+      "step": 1844
+    },
+    {
+      "epoch": 1.1226041983571646,
+      "grad_norm": 0.8498148918151855,
+      "learning_rate": 4.263580744303681e-05,
+      "loss": 0.2378,
+      "step": 1845
+    },
+    {
+      "epoch": 1.1232126559172497,
+      "grad_norm": 1.0088249444961548,
+      "learning_rate": 4.2626981792092735e-05,
+      "loss": 0.2551,
+      "step": 1846
+    },
+    {
+      "epoch": 1.123821113477335,
+      "grad_norm": 0.899652898311615,
+      "learning_rate": 4.261815177039189e-05,
+      "loss": 0.2272,
+      "step": 1847
+    },
+    {
+      "epoch": 1.1244295710374201,
+      "grad_norm": 1.3068557977676392,
+      "learning_rate": 4.260931738012378e-05,
+      "loss": 0.2594,
+      "step": 1848
+    },
+    {
+      "epoch": 1.1250380285975052,
+      "grad_norm": 1.037555456161499,
+      "learning_rate": 4.260047862347894e-05,
+      "loss": 0.2607,
+      "step": 1849
+    },
+    {
+      "epoch": 1.1256464861575906,
+      "grad_norm": 1.0918004512786865,
+      "learning_rate": 4.259163550264904e-05,
+      "loss": 0.2699,
+      "step": 1850
+    },
+    {
+      "epoch": 1.1262549437176756,
+      "grad_norm": 0.9314645528793335,
+      "learning_rate": 4.258278801982678e-05,
+      "loss": 0.2482,
+      "step": 1851
+    },
+    {
+      "epoch": 1.126863401277761,
+      "grad_norm": 0.8757783770561218,
+      "learning_rate": 4.257393617720599e-05,
+      "loss": 0.2579,
+      "step": 1852
+    },
+    {
+      "epoch": 1.127471858837846,
+      "grad_norm": 1.0207921266555786,
+      "learning_rate": 4.256507997698152e-05,
+      "loss": 0.2311,
+      "step": 1853
+    },
+    {
+      "epoch": 1.1280803163979312,
+      "grad_norm": 0.9114212989807129,
+      "learning_rate": 4.2556219421349394e-05,
+      "loss": 0.2494,
+      "step": 1854
+    },
+    {
+      "epoch": 1.1286887739580165,
+      "grad_norm": 1.038779616355896,
+      "learning_rate": 4.254735451250661e-05,
+      "loss": 0.2665,
+      "step": 1855
+    },
+    {
+      "epoch": 1.1292972315181016,
+      "grad_norm": 0.8652352094650269,
+      "learning_rate": 4.2538485252651326e-05,
+      "loss": 0.2155,
+      "step": 1856
+    },
+    {
+      "epoch": 1.129905689078187,
+      "grad_norm": 0.9267082810401917,
+      "learning_rate": 4.2529611643982735e-05,
+      "loss": 0.2499,
+      "step": 1857
+    },
+    {
+      "epoch": 1.130514146638272,
+      "grad_norm": 1.116803526878357,
+      "learning_rate": 4.252073368870113e-05,
+      "loss": 0.2744,
+      "step": 1858
+    },
+    {
+      "epoch": 1.131122604198357,
+      "grad_norm": 0.9374389052391052,
+      "learning_rate": 4.251185138900787e-05,
+      "loss": 0.2698,
+      "step": 1859
+    },
+    {
+      "epoch": 1.1317310617584424,
+      "grad_norm": 0.8052027821540833,
+      "learning_rate": 4.25029647471054e-05,
+      "loss": 0.2098,
+      "step": 1860
+    },
+    {
+      "epoch": 1.1323395193185275,
+      "grad_norm": 0.9955977201461792,
+      "learning_rate": 4.249407376519722e-05,
+      "loss": 0.2417,
+      "step": 1861
+    },
+    {
+      "epoch": 1.1329479768786128,
+      "grad_norm": 0.8962797522544861,
+      "learning_rate": 4.248517844548795e-05,
+      "loss": 0.2428,
+      "step": 1862
+    },
+    {
+      "epoch": 1.133556434438698,
+      "grad_norm": 1.001209020614624,
+      "learning_rate": 4.247627879018323e-05,
+      "loss": 0.2477,
+      "step": 1863
+    },
+    {
+      "epoch": 1.134164891998783,
+      "grad_norm": 1.0366243124008179,
+      "learning_rate": 4.246737480148983e-05,
+      "loss": 0.2382,
+      "step": 1864
+    },
+    {
+      "epoch": 1.1347733495588683,
+      "grad_norm": 0.9120834469795227,
+      "learning_rate": 4.245846648161554e-05,
+      "loss": 0.2555,
+      "step": 1865
+    },
+    {
+      "epoch": 1.1353818071189534,
+      "grad_norm": 0.9888159036636353,
+      "learning_rate": 4.2449553832769284e-05,
+      "loss": 0.2803,
+      "step": 1866
+    },
+    {
+      "epoch": 1.1359902646790387,
+      "grad_norm": 0.9909419417381287,
+      "learning_rate": 4.2440636857161e-05,
+      "loss": 0.27,
+      "step": 1867
+    },
+    {
+      "epoch": 1.1365987222391238,
+      "grad_norm": 1.0697113275527954,
+      "learning_rate": 4.243171555700174e-05,
+      "loss": 0.2664,
+      "step": 1868
+    },
+    {
+      "epoch": 1.137207179799209,
+      "grad_norm": 0.9328745603561401,
+      "learning_rate": 4.242278993450361e-05,
+      "loss": 0.2762,
+      "step": 1869
+    },
+    {
+      "epoch": 1.1378156373592943,
+      "grad_norm": 0.8406645059585571,
+      "learning_rate": 4.24138599918798e-05,
+      "loss": 0.2151,
+      "step": 1870
+    },
+    {
+      "epoch": 1.1384240949193793,
+      "grad_norm": 0.9748327136039734,
+      "learning_rate": 4.240492573134455e-05,
+      "loss": 0.2286,
+      "step": 1871
+    },
+    {
+      "epoch": 1.1390325524794647,
+      "grad_norm": 0.8779383301734924,
+      "learning_rate": 4.239598715511319e-05,
+      "loss": 0.2185,
+      "step": 1872
+    },
+    {
+      "epoch": 1.1396410100395498,
+      "grad_norm": 0.96523517370224,
+      "learning_rate": 4.238704426540213e-05,
+      "loss": 0.2499,
+      "step": 1873
+    },
+    {
+      "epoch": 1.1402494675996349,
+      "grad_norm": 1.156562328338623,
+      "learning_rate": 4.23780970644288e-05,
+      "loss": 0.272,
+      "step": 1874
+    },
+    {
+      "epoch": 1.1408579251597202,
+      "grad_norm": 1.0613001585006714,
+      "learning_rate": 4.236914555441177e-05,
+      "loss": 0.268,
+      "step": 1875
+    },
+    {
+      "epoch": 1.1414663827198053,
+      "grad_norm": 0.803700864315033,
+      "learning_rate": 4.236018973757061e-05,
+      "loss": 0.2026,
+      "step": 1876
+    },
+    {
+      "epoch": 1.1420748402798906,
+      "grad_norm": 1.021272897720337,
+      "learning_rate": 4.235122961612602e-05,
+      "loss": 0.2904,
+      "step": 1877
+    },
+    {
+      "epoch": 1.1426832978399757,
+      "grad_norm": 0.7920538187026978,
+      "learning_rate": 4.234226519229971e-05,
+      "loss": 0.2118,
+      "step": 1878
+    },
+    {
+      "epoch": 1.1432917554000608,
+      "grad_norm": 1.0034376382827759,
+      "learning_rate": 4.233329646831449e-05,
+      "loss": 0.2355,
+      "step": 1879
+    },
+    {
+      "epoch": 1.143900212960146,
+      "grad_norm": 1.0037422180175781,
+      "learning_rate": 4.2324323446394244e-05,
+      "loss": 0.2941,
+      "step": 1880
+    },
+    {
+      "epoch": 1.1445086705202312,
+      "grad_norm": 0.7395780086517334,
+      "learning_rate": 4.2315346128763886e-05,
+      "loss": 0.1974,
+      "step": 1881
+    },
+    {
+      "epoch": 1.1451171280803165,
+      "grad_norm": 0.9243549108505249,
+      "learning_rate": 4.2306364517649434e-05,
+      "loss": 0.2657,
+      "step": 1882
+    },
+    {
+      "epoch": 1.1457255856404016,
+      "grad_norm": 0.8601450324058533,
+      "learning_rate": 4.2297378615277935e-05,
+      "loss": 0.2241,
+      "step": 1883
+    },
+    {
+      "epoch": 1.1463340432004867,
+      "grad_norm": 0.9598115682601929,
+      "learning_rate": 4.228838842387755e-05,
+      "loss": 0.2484,
+      "step": 1884
+    },
+    {
+      "epoch": 1.146942500760572,
+      "grad_norm": 1.2198432683944702,
+      "learning_rate": 4.2279393945677437e-05,
+      "loss": 0.217,
+      "step": 1885
+    },
+    {
+      "epoch": 1.1475509583206571,
+      "grad_norm": 0.9545729756355286,
+      "learning_rate": 4.227039518290786e-05,
+      "loss": 0.2229,
+      "step": 1886
+    },
+    {
+      "epoch": 1.1481594158807424,
+      "grad_norm": 1.0584890842437744,
+      "learning_rate": 4.226139213780016e-05,
+      "loss": 0.2805,
+      "step": 1887
+    },
+    {
+      "epoch": 1.1487678734408275,
+      "grad_norm": 0.9882286787033081,
+      "learning_rate": 4.225238481258669e-05,
+      "loss": 0.2621,
+      "step": 1888
+    },
+    {
+      "epoch": 1.1493763310009126,
+      "grad_norm": 0.944648265838623,
+      "learning_rate": 4.22433732095009e-05,
+      "loss": 0.2358,
+      "step": 1889
+    },
+    {
+      "epoch": 1.149984788560998,
+      "grad_norm": 0.7821002006530762,
+      "learning_rate": 4.223435733077731e-05,
+      "loss": 0.1971,
+      "step": 1890
+    },
+    {
+      "epoch": 1.150593246121083,
+      "grad_norm": 1.0753979682922363,
+      "learning_rate": 4.2225337178651444e-05,
+      "loss": 0.2761,
+      "step": 1891
+    },
+    {
+      "epoch": 1.1512017036811681,
+      "grad_norm": 0.9069803953170776,
+      "learning_rate": 4.221631275535996e-05,
+      "loss": 0.258,
+      "step": 1892
+    },
+    {
+      "epoch": 1.1518101612412535,
+      "grad_norm": 1.0057759284973145,
+      "learning_rate": 4.2207284063140514e-05,
+      "loss": 0.2679,
+      "step": 1893
+    },
+    {
+      "epoch": 1.1524186188013386,
+      "grad_norm": 1.0301549434661865,
+      "learning_rate": 4.2198251104231854e-05,
+      "loss": 0.2995,
+      "step": 1894
+    },
+    {
+      "epoch": 1.1530270763614239,
+      "grad_norm": 0.9355301260948181,
+      "learning_rate": 4.218921388087379e-05,
+      "loss": 0.2179,
+      "step": 1895
+    },
+    {
+      "epoch": 1.153635533921509,
+      "grad_norm": 1.027159571647644,
+      "learning_rate": 4.2180172395307156e-05,
+      "loss": 0.2577,
+      "step": 1896
+    },
+    {
+      "epoch": 1.154243991481594,
+      "grad_norm": 1.046653389930725,
+      "learning_rate": 4.217112664977387e-05,
+      "loss": 0.2326,
+      "step": 1897
+    },
+    {
+      "epoch": 1.1548524490416794,
+      "grad_norm": 0.9612645506858826,
+      "learning_rate": 4.216207664651691e-05,
+      "loss": 0.2137,
+      "step": 1898
+    },
+    {
+      "epoch": 1.1554609066017645,
+      "grad_norm": 0.9418598413467407,
+      "learning_rate": 4.21530223877803e-05,
+      "loss": 0.2772,
+      "step": 1899
+    },
+    {
+      "epoch": 1.1560693641618498,
+      "grad_norm": 1.0486358404159546,
+      "learning_rate": 4.2143963875809096e-05,
+      "loss": 0.2663,
+      "step": 1900
+    },
+    {
+      "epoch": 1.156677821721935,
+      "grad_norm": 1.084235668182373,
+      "learning_rate": 4.213490111284945e-05,
+      "loss": 0.2499,
+      "step": 1901
+    },
+    {
+      "epoch": 1.15728627928202,
+      "grad_norm": 0.9962263107299805,
+      "learning_rate": 4.212583410114855e-05,
+      "loss": 0.2512,
+      "step": 1902
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.9712923765182495,
+      "learning_rate": 4.2116762842954625e-05,
+      "loss": 0.2396,
+      "step": 1903
+    },
+    {
+      "epoch": 1.1585031944021904,
+      "grad_norm": 0.9264492988586426,
+      "learning_rate": 4.210768734051699e-05,
+      "loss": 0.2584,
+      "step": 1904
+    },
+    {
+      "epoch": 1.1591116519622757,
+      "grad_norm": 1.201596975326538,
+      "learning_rate": 4.209860759608597e-05,
+      "loss": 0.287,
+      "step": 1905
+    },
+    {
+      "epoch": 1.1597201095223608,
+      "grad_norm": 0.9659140110015869,
+      "learning_rate": 4.2089523611912966e-05,
+      "loss": 0.2453,
+      "step": 1906
+    },
+    {
+      "epoch": 1.160328567082446,
+      "grad_norm": 0.8334513306617737,
+      "learning_rate": 4.208043539025044e-05,
+      "loss": 0.2573,
+      "step": 1907
+    },
+    {
+      "epoch": 1.1609370246425312,
+      "grad_norm": 0.9269493222236633,
+      "learning_rate": 4.2071342933351886e-05,
+      "loss": 0.2988,
+      "step": 1908
+    },
+    {
+      "epoch": 1.1615454822026163,
+      "grad_norm": 0.8946632146835327,
+      "learning_rate": 4.2062246243471846e-05,
+      "loss": 0.2378,
+      "step": 1909
+    },
+    {
+      "epoch": 1.1621539397627014,
+      "grad_norm": 0.9380949139595032,
+      "learning_rate": 4.2053145322865936e-05,
+      "loss": 0.2316,
+      "step": 1910
+    },
+    {
+      "epoch": 1.1627623973227867,
+      "grad_norm": 0.8564540147781372,
+      "learning_rate": 4.204404017379079e-05,
+      "loss": 0.2255,
+      "step": 1911
+    },
+    {
+      "epoch": 1.1633708548828718,
+      "grad_norm": 1.0175832509994507,
+      "learning_rate": 4.2034930798504114e-05,
+      "loss": 0.2384,
+      "step": 1912
+    },
+    {
+      "epoch": 1.1639793124429572,
+      "grad_norm": 0.9850523471832275,
+      "learning_rate": 4.202581719926465e-05,
+      "loss": 0.2682,
+      "step": 1913
+    },
+    {
+      "epoch": 1.1645877700030423,
+      "grad_norm": 1.0594969987869263,
+      "learning_rate": 4.201669937833219e-05,
+      "loss": 0.2491,
+      "step": 1914
+    },
+    {
+      "epoch": 1.1651962275631274,
+      "grad_norm": 0.9938765168190002,
+      "learning_rate": 4.2007577337967574e-05,
+      "loss": 0.2847,
+      "step": 1915
+    },
+    {
+      "epoch": 1.1658046851232127,
+      "grad_norm": 0.8707656860351562,
+      "learning_rate": 4.19984510804327e-05,
+      "loss": 0.2301,
+      "step": 1916
+    },
+    {
+      "epoch": 1.1664131426832978,
+      "grad_norm": 0.9624968767166138,
+      "learning_rate": 4.1989320607990474e-05,
+      "loss": 0.2459,
+      "step": 1917
+    },
+    {
+      "epoch": 1.167021600243383,
+      "grad_norm": 1.0143096446990967,
+      "learning_rate": 4.1980185922904894e-05,
+      "loss": 0.2169,
+      "step": 1918
+    },
+    {
+      "epoch": 1.1676300578034682,
+      "grad_norm": 0.8604801893234253,
+      "learning_rate": 4.197104702744097e-05,
+      "loss": 0.2162,
+      "step": 1919
+    },
+    {
+      "epoch": 1.1682385153635533,
+      "grad_norm": 1.1591583490371704,
+      "learning_rate": 4.1961903923864775e-05,
+      "loss": 0.2425,
+      "step": 1920
+    },
+    {
+      "epoch": 1.1688469729236386,
+      "grad_norm": 1.020859718322754,
+      "learning_rate": 4.1952756614443415e-05,
+      "loss": 0.2657,
+      "step": 1921
+    },
+    {
+      "epoch": 1.1694554304837237,
+      "grad_norm": 0.9535967111587524,
+      "learning_rate": 4.194360510144504e-05,
+      "loss": 0.2221,
+      "step": 1922
+    },
+    {
+      "epoch": 1.170063888043809,
+      "grad_norm": 1.1001362800598145,
+      "learning_rate": 4.1934449387138845e-05,
+      "loss": 0.3064,
+      "step": 1923
+    },
+    {
+      "epoch": 1.170672345603894,
+      "grad_norm": 0.851999819278717,
+      "learning_rate": 4.192528947379506e-05,
+      "loss": 0.2202,
+      "step": 1924
+    },
+    {
+      "epoch": 1.1712808031639792,
+      "grad_norm": 0.888411283493042,
+      "learning_rate": 4.1916125363684965e-05,
+      "loss": 0.2617,
+      "step": 1925
+    },
+    {
+      "epoch": 1.1718892607240645,
+      "grad_norm": 0.9610258340835571,
+      "learning_rate": 4.1906957059080886e-05,
+      "loss": 0.2411,
+      "step": 1926
+    },
+    {
+      "epoch": 1.1724977182841496,
+      "grad_norm": 0.9737393856048584,
+      "learning_rate": 4.189778456225617e-05,
+      "loss": 0.2612,
+      "step": 1927
+    },
+    {
+      "epoch": 1.173106175844235,
+      "grad_norm": 0.9638609886169434,
+      "learning_rate": 4.188860787548522e-05,
+      "loss": 0.2328,
+      "step": 1928
+    },
+    {
+      "epoch": 1.17371463340432,
+      "grad_norm": 1.1524028778076172,
+      "learning_rate": 4.187942700104346e-05,
+      "loss": 0.2513,
+      "step": 1929
+    },
+    {
+      "epoch": 1.1743230909644051,
+      "grad_norm": 0.9276042580604553,
+      "learning_rate": 4.1870241941207375e-05,
+      "loss": 0.2542,
+      "step": 1930
+    },
+    {
+      "epoch": 1.1749315485244904,
+      "grad_norm": 0.9600099325180054,
+      "learning_rate": 4.186105269825449e-05,
+      "loss": 0.2704,
+      "step": 1931
+    },
+    {
+      "epoch": 1.1755400060845755,
+      "grad_norm": 0.940571129322052,
+      "learning_rate": 4.1851859274463326e-05,
+      "loss": 0.2159,
+      "step": 1932
+    },
+    {
+      "epoch": 1.1761484636446609,
+      "grad_norm": 0.9095624685287476,
+      "learning_rate": 4.184266167211348e-05,
+      "loss": 0.2321,
+      "step": 1933
+    },
+    {
+      "epoch": 1.176756921204746,
+      "grad_norm": 0.9772641658782959,
+      "learning_rate": 4.183345989348558e-05,
+      "loss": 0.2242,
+      "step": 1934
+    },
+    {
+      "epoch": 1.177365378764831,
+      "grad_norm": 0.8371894359588623,
+      "learning_rate": 4.182425394086128e-05,
+      "loss": 0.1883,
+      "step": 1935
+    },
+    {
+      "epoch": 1.1779738363249164,
+      "grad_norm": 0.9441771507263184,
+      "learning_rate": 4.181504381652327e-05,
+      "loss": 0.2521,
+      "step": 1936
+    },
+    {
+      "epoch": 1.1785822938850015,
+      "grad_norm": 1.1066073179244995,
+      "learning_rate": 4.180582952275528e-05,
+      "loss": 0.275,
+      "step": 1937
+    },
+    {
+      "epoch": 1.1791907514450868,
+      "grad_norm": 0.9698868989944458,
+      "learning_rate": 4.179661106184207e-05,
+      "loss": 0.2583,
+      "step": 1938
+    },
+    {
+      "epoch": 1.1797992090051719,
+      "grad_norm": 0.9682748317718506,
+      "learning_rate": 4.178738843606943e-05,
+      "loss": 0.2726,
+      "step": 1939
+    },
+    {
+      "epoch": 1.180407666565257,
+      "grad_norm": 0.9256848692893982,
+      "learning_rate": 4.1778161647724203e-05,
+      "loss": 0.225,
+      "step": 1940
+    },
+    {
+      "epoch": 1.1810161241253423,
+      "grad_norm": 0.8582398891448975,
+      "learning_rate": 4.176893069909422e-05,
+      "loss": 0.2164,
+      "step": 1941
+    },
+    {
+      "epoch": 1.1816245816854274,
+      "grad_norm": 0.9054145812988281,
+      "learning_rate": 4.1759695592468395e-05,
+      "loss": 0.2086,
+      "step": 1942
+    },
+    {
+      "epoch": 1.1822330392455127,
+      "grad_norm": 0.9626482725143433,
+      "learning_rate": 4.175045633013665e-05,
+      "loss": 0.2939,
+      "step": 1943
+    },
+    {
+      "epoch": 1.1828414968055978,
+      "grad_norm": 0.8653888702392578,
+      "learning_rate": 4.1741212914389914e-05,
+      "loss": 0.2459,
+      "step": 1944
+    },
+    {
+      "epoch": 1.183449954365683,
+      "grad_norm": 0.94068843126297,
+      "learning_rate": 4.173196534752019e-05,
+      "loss": 0.2704,
+      "step": 1945
+    },
+    {
+      "epoch": 1.1840584119257682,
+      "grad_norm": 0.8980154991149902,
+      "learning_rate": 4.1722713631820485e-05,
+      "loss": 0.2332,
+      "step": 1946
+    },
+    {
+      "epoch": 1.1846668694858533,
+      "grad_norm": 0.9154375791549683,
+      "learning_rate": 4.171345776958483e-05,
+      "loss": 0.2293,
+      "step": 1947
+    },
+    {
+      "epoch": 1.1852753270459386,
+      "grad_norm": 0.9358067512512207,
+      "learning_rate": 4.17041977631083e-05,
+      "loss": 0.2325,
+      "step": 1948
+    },
+    {
+      "epoch": 1.1858837846060237,
+      "grad_norm": 0.8783141374588013,
+      "learning_rate": 4.1694933614686995e-05,
+      "loss": 0.2125,
+      "step": 1949
+    },
+    {
+      "epoch": 1.1864922421661088,
+      "grad_norm": 1.0046310424804688,
+      "learning_rate": 4.168566532661803e-05,
+      "loss": 0.2796,
+      "step": 1950
+    },
+    {
+      "epoch": 1.1871006997261941,
+      "grad_norm": 0.9390839338302612,
+      "learning_rate": 4.167639290119956e-05,
+      "loss": 0.2875,
+      "step": 1951
+    },
+    {
+      "epoch": 1.1877091572862792,
+      "grad_norm": 0.9401413202285767,
+      "learning_rate": 4.166711634073075e-05,
+      "loss": 0.2614,
+      "step": 1952
+    },
+    {
+      "epoch": 1.1883176148463646,
+      "grad_norm": 1.0055378675460815,
+      "learning_rate": 4.1657835647511804e-05,
+      "loss": 0.2489,
+      "step": 1953
+    },
+    {
+      "epoch": 1.1889260724064497,
+      "grad_norm": 0.8895730376243591,
+      "learning_rate": 4.164855082384396e-05,
+      "loss": 0.2491,
+      "step": 1954
+    },
+    {
+      "epoch": 1.1895345299665347,
+      "grad_norm": 0.8464339375495911,
+      "learning_rate": 4.163926187202946e-05,
+      "loss": 0.199,
+      "step": 1955
+    },
+    {
+      "epoch": 1.19014298752662,
+      "grad_norm": 0.9686809778213501,
+      "learning_rate": 4.162996879437156e-05,
+      "loss": 0.2742,
+      "step": 1956
+    },
+    {
+      "epoch": 1.1907514450867052,
+      "grad_norm": 1.0053505897521973,
+      "learning_rate": 4.1620671593174585e-05,
+      "loss": 0.2875,
+      "step": 1957
+    },
+    {
+      "epoch": 1.1913599026467905,
+      "grad_norm": 1.0522572994232178,
+      "learning_rate": 4.1611370270743826e-05,
+      "loss": 0.2174,
+      "step": 1958
+    },
+    {
+      "epoch": 1.1919683602068756,
+      "grad_norm": 0.8975366353988647,
+      "learning_rate": 4.160206482938565e-05,
+      "loss": 0.2535,
+      "step": 1959
+    },
+    {
+      "epoch": 1.1925768177669607,
+      "grad_norm": 0.8640589714050293,
+      "learning_rate": 4.159275527140739e-05,
+      "loss": 0.2178,
+      "step": 1960
+    },
+    {
+      "epoch": 1.193185275327046,
+      "grad_norm": 0.9383702874183655,
+      "learning_rate": 4.1583441599117453e-05,
+      "loss": 0.2912,
+      "step": 1961
+    },
+    {
+      "epoch": 1.193793732887131,
+      "grad_norm": 0.8808874487876892,
+      "learning_rate": 4.1574123814825226e-05,
+      "loss": 0.2189,
+      "step": 1962
+    },
+    {
+      "epoch": 1.1944021904472164,
+      "grad_norm": 0.9913263320922852,
+      "learning_rate": 4.156480192084114e-05,
+      "loss": 0.2931,
+      "step": 1963
+    },
+    {
+      "epoch": 1.1950106480073015,
+      "grad_norm": 1.0202916860580444,
+      "learning_rate": 4.155547591947663e-05,
+      "loss": 0.2677,
+      "step": 1964
+    },
+    {
+      "epoch": 1.1956191055673866,
+      "grad_norm": 1.405519962310791,
+      "learning_rate": 4.154614581304416e-05,
+      "loss": 0.3883,
+      "step": 1965
+    },
+    {
+      "epoch": 1.196227563127472,
+      "grad_norm": 0.9594994783401489,
+      "learning_rate": 4.15368116038572e-05,
+      "loss": 0.2595,
+      "step": 1966
+    },
+    {
+      "epoch": 1.196836020687557,
+      "grad_norm": 1.043186902999878,
+      "learning_rate": 4.1527473294230255e-05,
+      "loss": 0.2648,
+      "step": 1967
+    },
+    {
+      "epoch": 1.1974444782476423,
+      "grad_norm": 0.9707902073860168,
+      "learning_rate": 4.151813088647883e-05,
+      "loss": 0.2546,
+      "step": 1968
+    },
+    {
+      "epoch": 1.1980529358077274,
+      "grad_norm": 0.9176664352416992,
+      "learning_rate": 4.1508784382919466e-05,
+      "loss": 0.2161,
+      "step": 1969
+    },
+    {
+      "epoch": 1.1986613933678125,
+      "grad_norm": 0.9123236536979675,
+      "learning_rate": 4.149943378586968e-05,
+      "loss": 0.242,
+      "step": 1970
+    },
+    {
+      "epoch": 1.1992698509278978,
+      "grad_norm": 0.9186065793037415,
+      "learning_rate": 4.149007909764805e-05,
+      "loss": 0.2001,
+      "step": 1971
+    },
+    {
+      "epoch": 1.199878308487983,
+      "grad_norm": 1.0215867757797241,
+      "learning_rate": 4.148072032057415e-05,
+      "loss": 0.2529,
+      "step": 1972
+    },
+    {
+      "epoch": 1.2004867660480683,
+      "grad_norm": 0.9522402882575989,
+      "learning_rate": 4.1471357456968556e-05,
+      "loss": 0.2096,
+      "step": 1973
+    },
+    {
+      "epoch": 1.2010952236081534,
+      "grad_norm": 0.9875971674919128,
+      "learning_rate": 4.146199050915288e-05,
+      "loss": 0.2659,
+      "step": 1974
+    },
+    {
+      "epoch": 1.2017036811682384,
+      "grad_norm": 0.892595112323761,
+      "learning_rate": 4.1452619479449714e-05,
+      "loss": 0.2422,
+      "step": 1975
+    },
+    {
+      "epoch": 1.2023121387283238,
+      "grad_norm": 0.9332039952278137,
+      "learning_rate": 4.14432443701827e-05,
+      "loss": 0.1921,
+      "step": 1976
+    },
+    {
+      "epoch": 1.2029205962884089,
+      "grad_norm": 0.8022421598434448,
+      "learning_rate": 4.143386518367648e-05,
+      "loss": 0.1945,
+      "step": 1977
+    },
+    {
+      "epoch": 1.2035290538484942,
+      "grad_norm": 0.9475464224815369,
+      "learning_rate": 4.142448192225669e-05,
+      "loss": 0.2119,
+      "step": 1978
+    },
+    {
+      "epoch": 1.2041375114085793,
+      "grad_norm": 0.8126545548439026,
+      "learning_rate": 4.1415094588249975e-05,
+      "loss": 0.2195,
+      "step": 1979
+    },
+    {
+      "epoch": 1.2047459689686644,
+      "grad_norm": 1.1211334466934204,
+      "learning_rate": 4.140570318398403e-05,
+      "loss": 0.2904,
+      "step": 1980
+    },
+    {
+      "epoch": 1.2053544265287497,
+      "grad_norm": 1.054579734802246,
+      "learning_rate": 4.1396307711787516e-05,
+      "loss": 0.2495,
+      "step": 1981
+    },
+    {
+      "epoch": 1.2059628840888348,
+      "grad_norm": 0.9004554152488708,
+      "learning_rate": 4.1386908173990126e-05,
+      "loss": 0.2597,
+      "step": 1982
+    },
+    {
+      "epoch": 1.20657134164892,
+      "grad_norm": 0.9053953886032104,
+      "learning_rate": 4.1377504572922534e-05,
+      "loss": 0.2532,
+      "step": 1983
+    },
+    {
+      "epoch": 1.2071797992090052,
+      "grad_norm": 0.8492591381072998,
+      "learning_rate": 4.136809691091647e-05,
+      "loss": 0.2266,
+      "step": 1984
+    },
+    {
+      "epoch": 1.2077882567690903,
+      "grad_norm": 0.9747501611709595,
+      "learning_rate": 4.135868519030463e-05,
+      "loss": 0.2336,
+      "step": 1985
+    },
+    {
+      "epoch": 1.2083967143291756,
+      "grad_norm": 0.907906711101532,
+      "learning_rate": 4.134926941342071e-05,
+      "loss": 0.2306,
+      "step": 1986
+    },
+    {
+      "epoch": 1.2090051718892607,
+      "grad_norm": 1.0222368240356445,
+      "learning_rate": 4.1339849582599454e-05,
+      "loss": 0.2196,
+      "step": 1987
+    },
+    {
+      "epoch": 1.209613629449346,
+      "grad_norm": 1.05088210105896,
+      "learning_rate": 4.1330425700176586e-05,
+      "loss": 0.2503,
+      "step": 1988
+    },
+    {
+      "epoch": 1.2102220870094311,
+      "grad_norm": 1.1211512088775635,
+      "learning_rate": 4.132099776848882e-05,
+      "loss": 0.2371,
+      "step": 1989
+    },
+    {
+      "epoch": 1.2108305445695162,
+      "grad_norm": 0.9060412645339966,
+      "learning_rate": 4.1311565789873914e-05,
+      "loss": 0.2377,
+      "step": 1990
+    },
+    {
+      "epoch": 1.2114390021296015,
+      "grad_norm": 0.8177694082260132,
+      "learning_rate": 4.1302129766670586e-05,
+      "loss": 0.2452,
+      "step": 1991
+    },
+    {
+      "epoch": 1.2120474596896866,
+      "grad_norm": 0.9358994960784912,
+      "learning_rate": 4.129268970121858e-05,
+      "loss": 0.2222,
+      "step": 1992
+    },
+    {
+      "epoch": 1.212655917249772,
+      "grad_norm": 0.8588563203811646,
+      "learning_rate": 4.128324559585863e-05,
+      "loss": 0.219,
+      "step": 1993
+    },
+    {
+      "epoch": 1.213264374809857,
+      "grad_norm": 0.9672694206237793,
+      "learning_rate": 4.127379745293251e-05,
+      "loss": 0.2587,
+      "step": 1994
+    },
+    {
+      "epoch": 1.2138728323699421,
+      "grad_norm": 0.8904327154159546,
+      "learning_rate": 4.1264345274782937e-05,
+      "loss": 0.2371,
+      "step": 1995
+    },
+    {
+      "epoch": 1.2144812899300275,
+      "grad_norm": 1.017808198928833,
+      "learning_rate": 4.125488906375367e-05,
+      "loss": 0.2382,
+      "step": 1996
+    },
+    {
+      "epoch": 1.2150897474901126,
+      "grad_norm": 0.9038395881652832,
+      "learning_rate": 4.1245428822189444e-05,
+      "loss": 0.2444,
+      "step": 1997
+    },
+    {
+      "epoch": 1.2156982050501977,
+      "grad_norm": 1.1696877479553223,
+      "learning_rate": 4.123596455243603e-05,
+      "loss": 0.2993,
+      "step": 1998
+    },
+    {
+      "epoch": 1.216306662610283,
+      "grad_norm": 0.8989875316619873,
+      "learning_rate": 4.122649625684014e-05,
+      "loss": 0.1964,
+      "step": 1999
+    },
+    {
+      "epoch": 1.216915120170368,
+      "grad_norm": 1.0550363063812256,
+      "learning_rate": 4.1217023937749544e-05,
+      "loss": 0.2987,
+      "step": 2000
+    },
+    {
+      "epoch": 1.2175235777304534,
+      "grad_norm": 0.9538779258728027,
+      "learning_rate": 4.120754759751296e-05,
+      "loss": 0.2522,
+      "step": 2001
+    },
+    {
+      "epoch": 1.2181320352905385,
+      "grad_norm": 1.279292106628418,
+      "learning_rate": 4.1198067238480145e-05,
+      "loss": 0.2853,
+      "step": 2002
+    },
+    {
+      "epoch": 1.2187404928506236,
+      "grad_norm": 1.0189933776855469,
+      "learning_rate": 4.118858286300182e-05,
+      "loss": 0.2265,
+      "step": 2003
+    },
+    {
+      "epoch": 1.219348950410709,
+      "grad_norm": 0.8648117184638977,
+      "learning_rate": 4.117909447342972e-05,
+      "loss": 0.2193,
+      "step": 2004
+    },
+    {
+      "epoch": 1.219957407970794,
+      "grad_norm": 0.8088192939758301,
+      "learning_rate": 4.1169602072116567e-05,
+      "loss": 0.2087,
+      "step": 2005
+    },
+    {
+      "epoch": 1.2205658655308793,
+      "grad_norm": 1.034005045890808,
+      "learning_rate": 4.116010566141608e-05,
+      "loss": 0.2648,
+      "step": 2006
+    },
+    {
+      "epoch": 1.2211743230909644,
+      "grad_norm": 1.020052194595337,
+      "learning_rate": 4.115060524368297e-05,
+      "loss": 0.2715,
+      "step": 2007
+    },
+    {
+      "epoch": 1.2217827806510495,
+      "grad_norm": 0.8594374656677246,
+      "learning_rate": 4.114110082127296e-05,
+      "loss": 0.2395,
+      "step": 2008
+    },
+    {
+      "epoch": 1.2223912382111348,
+      "grad_norm": 0.9257277250289917,
+      "learning_rate": 4.113159239654273e-05,
+      "loss": 0.2356,
+      "step": 2009
+    },
+    {
+      "epoch": 1.22299969577122,
+      "grad_norm": 0.9078371524810791,
+      "learning_rate": 4.1122079971849994e-05,
+      "loss": 0.2536,
+      "step": 2010
+    },
+    {
+      "epoch": 1.2236081533313052,
+      "grad_norm": 0.9304254055023193,
+      "learning_rate": 4.1112563549553427e-05,
+      "loss": 0.2504,
+      "step": 2011
+    },
+    {
+      "epoch": 1.2242166108913903,
+      "grad_norm": 0.9377618432044983,
+      "learning_rate": 4.11030431320127e-05,
+      "loss": 0.2245,
+      "step": 2012
+    },
+    {
+      "epoch": 1.2248250684514754,
+      "grad_norm": 0.8383059501647949,
+      "learning_rate": 4.1093518721588484e-05,
+      "loss": 0.1871,
+      "step": 2013
+    },
+    {
+      "epoch": 1.2254335260115607,
+      "grad_norm": 0.853079617023468,
+      "learning_rate": 4.108399032064243e-05,
+      "loss": 0.2169,
+      "step": 2014
+    },
+    {
+      "epoch": 1.2260419835716458,
+      "grad_norm": 0.9059860110282898,
+      "learning_rate": 4.10744579315372e-05,
+      "loss": 0.2204,
+      "step": 2015
+    },
+    {
+      "epoch": 1.226650441131731,
+      "grad_norm": 0.9250550866127014,
+      "learning_rate": 4.1064921556636413e-05,
+      "loss": 0.2556,
+      "step": 2016
+    },
+    {
+      "epoch": 1.2272588986918163,
+      "grad_norm": 0.9633865356445312,
+      "learning_rate": 4.1055381198304705e-05,
+      "loss": 0.2426,
+      "step": 2017
+    },
+    {
+      "epoch": 1.2278673562519014,
+      "grad_norm": 0.865479588508606,
+      "learning_rate": 4.1045836858907676e-05,
+      "loss": 0.2093,
+      "step": 2018
+    },
+    {
+      "epoch": 1.2284758138119867,
+      "grad_norm": 0.8409626483917236,
+      "learning_rate": 4.1036288540811935e-05,
+      "loss": 0.1839,
+      "step": 2019
+    },
+    {
+      "epoch": 1.2290842713720718,
+      "grad_norm": 0.8517163991928101,
+      "learning_rate": 4.1026736246385055e-05,
+      "loss": 0.1791,
+      "step": 2020
+    },
+    {
+      "epoch": 1.2296927289321569,
+      "grad_norm": 1.0229874849319458,
+      "learning_rate": 4.101717997799562e-05,
+      "loss": 0.2466,
+      "step": 2021
+    },
+    {
+      "epoch": 1.2303011864922422,
+      "grad_norm": 1.0059734582901,
+      "learning_rate": 4.100761973801317e-05,
+      "loss": 0.2218,
+      "step": 2022
+    },
+    {
+      "epoch": 1.2309096440523273,
+      "grad_norm": 1.6977488994598389,
+      "learning_rate": 4.0998055528808266e-05,
+      "loss": 0.2822,
+      "step": 2023
+    },
+    {
+      "epoch": 1.2315181016124126,
+      "grad_norm": 1.067182183265686,
+      "learning_rate": 4.0988487352752414e-05,
+      "loss": 0.2576,
+      "step": 2024
+    },
+    {
+      "epoch": 1.2321265591724977,
+      "grad_norm": 0.9440335631370544,
+      "learning_rate": 4.097891521221814e-05,
+      "loss": 0.2397,
+      "step": 2025
+    },
+    {
+      "epoch": 1.2327350167325828,
+      "grad_norm": 0.9076058268547058,
+      "learning_rate": 4.096933910957892e-05,
+      "loss": 0.2329,
+      "step": 2026
+    },
+    {
+      "epoch": 1.233343474292668,
+      "grad_norm": 0.8461357951164246,
+      "learning_rate": 4.0959759047209234e-05,
+      "loss": 0.2133,
+      "step": 2027
+    },
+    {
+      "epoch": 1.2339519318527532,
+      "grad_norm": 0.8329628109931946,
+      "learning_rate": 4.095017502748455e-05,
+      "loss": 0.2251,
+      "step": 2028
+    },
+    {
+      "epoch": 1.2345603894128385,
+      "grad_norm": 0.8861196041107178,
+      "learning_rate": 4.094058705278129e-05,
+      "loss": 0.2232,
+      "step": 2029
+    },
+    {
+      "epoch": 1.2351688469729236,
+      "grad_norm": 0.9130899906158447,
+      "learning_rate": 4.093099512547687e-05,
+      "loss": 0.2517,
+      "step": 2030
+    },
+    {
+      "epoch": 1.2357773045330087,
+      "grad_norm": 0.8920985460281372,
+      "learning_rate": 4.09213992479497e-05,
+      "loss": 0.2403,
+      "step": 2031
+    },
+    {
+      "epoch": 1.236385762093094,
+      "grad_norm": 0.9351683259010315,
+      "learning_rate": 4.0911799422579155e-05,
+      "loss": 0.2416,
+      "step": 2032
+    },
+    {
+      "epoch": 1.2369942196531791,
+      "grad_norm": 0.8944339156150818,
+      "learning_rate": 4.090219565174559e-05,
+      "loss": 0.2441,
+      "step": 2033
+    },
+    {
+      "epoch": 1.2376026772132644,
+      "grad_norm": 0.9203975200653076,
+      "learning_rate": 4.089258793783034e-05,
+      "loss": 0.2541,
+      "step": 2034
+    },
+    {
+      "epoch": 1.2382111347733495,
+      "grad_norm": 0.880241334438324,
+      "learning_rate": 4.0882976283215714e-05,
+      "loss": 0.2492,
+      "step": 2035
+    },
+    {
+      "epoch": 1.2388195923334346,
+      "grad_norm": 0.9325846433639526,
+      "learning_rate": 4.087336069028501e-05,
+      "loss": 0.2663,
+      "step": 2036
+    },
+    {
+      "epoch": 1.23942804989352,
+      "grad_norm": 1.0154072046279907,
+      "learning_rate": 4.086374116142249e-05,
+      "loss": 0.2699,
+      "step": 2037
+    },
+    {
+      "epoch": 1.240036507453605,
+      "grad_norm": 0.9841558933258057,
+      "learning_rate": 4.0854117699013396e-05,
+      "loss": 0.2496,
+      "step": 2038
+    },
+    {
+      "epoch": 1.2406449650136904,
+      "grad_norm": 0.8984540700912476,
+      "learning_rate": 4.0844490305443934e-05,
+      "loss": 0.2008,
+      "step": 2039
+    },
+    {
+      "epoch": 1.2412534225737755,
+      "grad_norm": 0.8929226994514465,
+      "learning_rate": 4.083485898310131e-05,
+      "loss": 0.2588,
+      "step": 2040
+    },
+    {
+      "epoch": 1.2418618801338606,
+      "grad_norm": 0.7902721762657166,
+      "learning_rate": 4.0825223734373696e-05,
+      "loss": 0.1924,
+      "step": 2041
+    },
+    {
+      "epoch": 1.2424703376939459,
+      "grad_norm": 0.8901758790016174,
+      "learning_rate": 4.081558456165022e-05,
+      "loss": 0.246,
+      "step": 2042
+    },
+    {
+      "epoch": 1.243078795254031,
+      "grad_norm": 1.1308690309524536,
+      "learning_rate": 4.080594146732099e-05,
+      "loss": 0.286,
+      "step": 2043
+    },
+    {
+      "epoch": 1.2436872528141163,
+      "grad_norm": 1.1202785968780518,
+      "learning_rate": 4.079629445377712e-05,
+      "loss": 0.3143,
+      "step": 2044
+    },
+    {
+      "epoch": 1.2442957103742014,
+      "grad_norm": 0.8539735078811646,
+      "learning_rate": 4.078664352341063e-05,
+      "loss": 0.2171,
+      "step": 2045
+    },
+    {
+      "epoch": 1.2449041679342865,
+      "grad_norm": 1.0043787956237793,
+      "learning_rate": 4.077698867861457e-05,
+      "loss": 0.2746,
+      "step": 2046
+    },
+    {
+      "epoch": 1.2455126254943718,
+      "grad_norm": 1.0371819734573364,
+      "learning_rate": 4.076732992178294e-05,
+      "loss": 0.269,
+      "step": 2047
+    },
+    {
+      "epoch": 1.246121083054457,
+      "grad_norm": 0.9970436096191406,
+      "learning_rate": 4.075766725531069e-05,
+      "loss": 0.2185,
+      "step": 2048
+    },
+    {
+      "epoch": 1.2467295406145422,
+      "grad_norm": 1.058952808380127,
+      "learning_rate": 4.074800068159379e-05,
+      "loss": 0.2503,
+      "step": 2049
+    },
+    {
+      "epoch": 1.2473379981746273,
+      "grad_norm": 0.8718734383583069,
+      "learning_rate": 4.073833020302912e-05,
+      "loss": 0.2422,
+      "step": 2050
+    },
+    {
+      "epoch": 1.2479464557347124,
+      "grad_norm": 0.8423463106155396,
+      "learning_rate": 4.0728655822014574e-05,
+      "loss": 0.2396,
+      "step": 2051
+    },
+    {
+      "epoch": 1.2485549132947977,
+      "grad_norm": 0.9200191497802734,
+      "learning_rate": 4.071897754094898e-05,
+      "loss": 0.201,
+      "step": 2052
+    },
+    {
+      "epoch": 1.2491633708548828,
+      "grad_norm": 1.0252035856246948,
+      "learning_rate": 4.0709295362232156e-05,
+      "loss": 0.2548,
+      "step": 2053
+    },
+    {
+      "epoch": 1.2497718284149681,
+      "grad_norm": 0.875744640827179,
+      "learning_rate": 4.069960928826488e-05,
+      "loss": 0.2319,
+      "step": 2054
+    },
+    {
+      "epoch": 1.2503802859750532,
+      "grad_norm": 1.0111494064331055,
+      "learning_rate": 4.0689919321448885e-05,
+      "loss": 0.2188,
+      "step": 2055
+    },
+    {
+      "epoch": 1.2509887435351383,
+      "grad_norm": 1.0929096937179565,
+      "learning_rate": 4.0680225464186895e-05,
+      "loss": 0.2029,
+      "step": 2056
+    },
+    {
+      "epoch": 1.2515972010952237,
+      "grad_norm": 0.9133079051971436,
+      "learning_rate": 4.067052771888257e-05,
+      "loss": 0.2084,
+      "step": 2057
+    },
+    {
+      "epoch": 1.2522056586553088,
+      "grad_norm": 0.8972569108009338,
+      "learning_rate": 4.066082608794055e-05,
+      "loss": 0.238,
+      "step": 2058
+    },
+    {
+      "epoch": 1.252814116215394,
+      "grad_norm": 0.9103267788887024,
+      "learning_rate": 4.0651120573766447e-05,
+      "loss": 0.2144,
+      "step": 2059
+    },
+    {
+      "epoch": 1.2534225737754792,
+      "grad_norm": 0.8961523771286011,
+      "learning_rate": 4.0641411178766795e-05,
+      "loss": 0.2639,
+      "step": 2060
+    },
+    {
+      "epoch": 1.2540310313355643,
+      "grad_norm": 0.907996654510498,
+      "learning_rate": 4.0631697905349144e-05,
+      "loss": 0.2035,
+      "step": 2061
+    },
+    {
+      "epoch": 1.2546394888956496,
+      "grad_norm": 1.2141492366790771,
+      "learning_rate": 4.0621980755921974e-05,
+      "loss": 0.2882,
+      "step": 2062
+    },
+    {
+      "epoch": 1.2552479464557347,
+      "grad_norm": 0.889756441116333,
+      "learning_rate": 4.061225973289473e-05,
+      "loss": 0.2281,
+      "step": 2063
+    },
+    {
+      "epoch": 1.25585640401582,
+      "grad_norm": 0.9822306632995605,
+      "learning_rate": 4.060253483867783e-05,
+      "loss": 0.2966,
+      "step": 2064
+    },
+    {
+      "epoch": 1.256464861575905,
+      "grad_norm": 0.9463213682174683,
+      "learning_rate": 4.059280607568263e-05,
+      "loss": 0.2358,
+      "step": 2065
+    },
+    {
+      "epoch": 1.2570733191359902,
+      "grad_norm": 0.9215368032455444,
+      "learning_rate": 4.058307344632147e-05,
+      "loss": 0.2188,
+      "step": 2066
+    },
+    {
+      "epoch": 1.2576817766960755,
+      "grad_norm": 0.9397695064544678,
+      "learning_rate": 4.057333695300762e-05,
+      "loss": 0.2423,
+      "step": 2067
+    },
+    {
+      "epoch": 1.2582902342561606,
+      "grad_norm": 0.9168745875358582,
+      "learning_rate": 4.056359659815534e-05,
+      "loss": 0.1766,
+      "step": 2068
+    },
+    {
+      "epoch": 1.258898691816246,
+      "grad_norm": 0.8033076524734497,
+      "learning_rate": 4.055385238417984e-05,
+      "loss": 0.2049,
+      "step": 2069
+    },
+    {
+      "epoch": 1.259507149376331,
+      "grad_norm": 0.8447003960609436,
+      "learning_rate": 4.054410431349724e-05,
+      "loss": 0.2532,
+      "step": 2070
+    },
+    {
+      "epoch": 1.260115606936416,
+      "grad_norm": 0.9117369651794434,
+      "learning_rate": 4.053435238852469e-05,
+      "loss": 0.2495,
+      "step": 2071
+    },
+    {
+      "epoch": 1.2607240644965014,
+      "grad_norm": 0.838207483291626,
+      "learning_rate": 4.052459661168025e-05,
+      "loss": 0.2442,
+      "step": 2072
+    },
+    {
+      "epoch": 1.2613325220565865,
+      "grad_norm": 0.9510546922683716,
+      "learning_rate": 4.051483698538295e-05,
+      "loss": 0.2255,
+      "step": 2073
+    },
+    {
+      "epoch": 1.2619409796166718,
+      "grad_norm": 1.0252444744110107,
+      "learning_rate": 4.0505073512052774e-05,
+      "loss": 0.2537,
+      "step": 2074
+    },
+    {
+      "epoch": 1.262549437176757,
+      "grad_norm": 0.8936956524848938,
+      "learning_rate": 4.049530619411065e-05,
+      "loss": 0.1971,
+      "step": 2075
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.9542527198791504,
+      "learning_rate": 4.0485535033978455e-05,
+      "loss": 0.2105,
+      "step": 2076
+    },
+    {
+      "epoch": 1.2637663522969274,
+      "grad_norm": 0.9937785863876343,
+      "learning_rate": 4.047576003407905e-05,
+      "loss": 0.2452,
+      "step": 2077
+    },
+    {
+      "epoch": 1.2643748098570124,
+      "grad_norm": 0.9868922233581543,
+      "learning_rate": 4.046598119683621e-05,
+      "loss": 0.2215,
+      "step": 2078
+    },
+    {
+      "epoch": 1.2649832674170978,
+      "grad_norm": 0.9198853373527527,
+      "learning_rate": 4.045619852467469e-05,
+      "loss": 0.2079,
+      "step": 2079
+    },
+    {
+      "epoch": 1.2655917249771829,
+      "grad_norm": 0.9050553441047668,
+      "learning_rate": 4.0446412020020185e-05,
+      "loss": 0.2373,
+      "step": 2080
+    },
+    {
+      "epoch": 1.266200182537268,
+      "grad_norm": 1.0860339403152466,
+      "learning_rate": 4.0436621685299334e-05,
+      "loss": 0.2303,
+      "step": 2081
+    },
+    {
+      "epoch": 1.2668086400973533,
+      "grad_norm": 0.9267223477363586,
+      "learning_rate": 4.0426827522939735e-05,
+      "loss": 0.2241,
+      "step": 2082
+    },
+    {
+      "epoch": 1.2674170976574384,
+      "grad_norm": 0.8770591616630554,
+      "learning_rate": 4.041702953536994e-05,
+      "loss": 0.2209,
+      "step": 2083
+    },
+    {
+      "epoch": 1.2680255552175237,
+      "grad_norm": 0.7461957335472107,
+      "learning_rate": 4.0407227725019426e-05,
+      "loss": 0.2294,
+      "step": 2084
+    },
+    {
+      "epoch": 1.2686340127776088,
+      "grad_norm": 1.0627920627593994,
+      "learning_rate": 4.039742209431864e-05,
+      "loss": 0.2419,
+      "step": 2085
+    },
+    {
+      "epoch": 1.2692424703376939,
+      "grad_norm": 0.9372715353965759,
+      "learning_rate": 4.0387612645698974e-05,
+      "loss": 0.2496,
+      "step": 2086
+    },
+    {
+      "epoch": 1.2698509278977792,
+      "grad_norm": 0.9207741022109985,
+      "learning_rate": 4.037779938159276e-05,
+      "loss": 0.2682,
+      "step": 2087
+    },
+    {
+      "epoch": 1.2704593854578643,
+      "grad_norm": 0.9947006702423096,
+      "learning_rate": 4.036798230443328e-05,
+      "loss": 0.241,
+      "step": 2088
+    },
+    {
+      "epoch": 1.2710678430179496,
+      "grad_norm": 0.9042183756828308,
+      "learning_rate": 4.035816141665475e-05,
+      "loss": 0.2357,
+      "step": 2089
+    },
+    {
+      "epoch": 1.2716763005780347,
+      "grad_norm": 0.9522048830986023,
+      "learning_rate": 4.0348336720692345e-05,
+      "loss": 0.1831,
+      "step": 2090
+    },
+    {
+      "epoch": 1.2722847581381198,
+      "grad_norm": 0.8158754706382751,
+      "learning_rate": 4.0338508218982197e-05,
+      "loss": 0.2126,
+      "step": 2091
+    },
+    {
+      "epoch": 1.2728932156982051,
+      "grad_norm": 1.012651801109314,
+      "learning_rate": 4.032867591396135e-05,
+      "loss": 0.2472,
+      "step": 2092
+    },
+    {
+      "epoch": 1.2735016732582902,
+      "grad_norm": 0.9281013011932373,
+      "learning_rate": 4.0318839808067796e-05,
+      "loss": 0.2052,
+      "step": 2093
+    },
+    {
+      "epoch": 1.2741101308183755,
+      "grad_norm": 1.0591697692871094,
+      "learning_rate": 4.0308999903740496e-05,
+      "loss": 0.2503,
+      "step": 2094
+    },
+    {
+      "epoch": 1.2747185883784606,
+      "grad_norm": 0.852708637714386,
+      "learning_rate": 4.029915620341933e-05,
+      "loss": 0.2026,
+      "step": 2095
+    },
+    {
+      "epoch": 1.2753270459385457,
+      "grad_norm": 0.9270033836364746,
+      "learning_rate": 4.028930870954512e-05,
+      "loss": 0.223,
+      "step": 2096
+    },
+    {
+      "epoch": 1.275935503498631,
+      "grad_norm": 0.8873688578605652,
+      "learning_rate": 4.0279457424559654e-05,
+      "loss": 0.2448,
+      "step": 2097
+    },
+    {
+      "epoch": 1.2765439610587161,
+      "grad_norm": 0.8902617692947388,
+      "learning_rate": 4.0269602350905615e-05,
+      "loss": 0.2214,
+      "step": 2098
+    },
+    {
+      "epoch": 1.2771524186188015,
+      "grad_norm": 0.8026160597801208,
+      "learning_rate": 4.025974349102667e-05,
+      "loss": 0.201,
+      "step": 2099
+    },
+    {
+      "epoch": 1.2777608761788866,
+      "grad_norm": 0.8452394008636475,
+      "learning_rate": 4.024988084736739e-05,
+      "loss": 0.2145,
+      "step": 2100
+    },
+    {
+      "epoch": 1.2783693337389717,
+      "grad_norm": 0.859439492225647,
+      "learning_rate": 4.0240014422373304e-05,
+      "loss": 0.2221,
+      "step": 2101
+    },
+    {
+      "epoch": 1.2789777912990568,
+      "grad_norm": 0.929378867149353,
+      "learning_rate": 4.023014421849088e-05,
+      "loss": 0.2458,
+      "step": 2102
+    },
+    {
+      "epoch": 1.279586248859142,
+      "grad_norm": 0.9007444381713867,
+      "learning_rate": 4.0220270238167514e-05,
+      "loss": 0.2225,
+      "step": 2103
+    },
+    {
+      "epoch": 1.2801947064192274,
+      "grad_norm": 0.821709394454956,
+      "learning_rate": 4.021039248385154e-05,
+      "loss": 0.1803,
+      "step": 2104
+    },
+    {
+      "epoch": 1.2808031639793125,
+      "grad_norm": 1.293845534324646,
+      "learning_rate": 4.0200510957992234e-05,
+      "loss": 0.397,
+      "step": 2105
+    },
+    {
+      "epoch": 1.2814116215393976,
+      "grad_norm": 0.8068326711654663,
+      "learning_rate": 4.0190625663039796e-05,
+      "loss": 0.2132,
+      "step": 2106
+    },
+    {
+      "epoch": 1.2820200790994827,
+      "grad_norm": 0.9025530219078064,
+      "learning_rate": 4.0180736601445365e-05,
+      "loss": 0.2425,
+      "step": 2107
+    },
+    {
+      "epoch": 1.282628536659568,
+      "grad_norm": 0.9707067012786865,
+      "learning_rate": 4.0170843775661025e-05,
+      "loss": 0.2608,
+      "step": 2108
+    },
+    {
+      "epoch": 1.2832369942196533,
+      "grad_norm": 0.9354795217514038,
+      "learning_rate": 4.0160947188139786e-05,
+      "loss": 0.2323,
+      "step": 2109
+    },
+    {
+      "epoch": 1.2838454517797384,
+      "grad_norm": 1.0434750318527222,
+      "learning_rate": 4.015104684133558e-05,
+      "loss": 0.2411,
+      "step": 2110
+    },
+    {
+      "epoch": 1.2844539093398235,
+      "grad_norm": 0.9146111011505127,
+      "learning_rate": 4.014114273770328e-05,
+      "loss": 0.247,
+      "step": 2111
+    },
+    {
+      "epoch": 1.2850623668999086,
+      "grad_norm": 0.9847277402877808,
+      "learning_rate": 4.01312348796987e-05,
+      "loss": 0.2576,
+      "step": 2112
+    },
+    {
+      "epoch": 1.285670824459994,
+      "grad_norm": 0.9638257026672363,
+      "learning_rate": 4.0121323269778565e-05,
+      "loss": 0.2398,
+      "step": 2113
+    },
+    {
+      "epoch": 1.286279282020079,
+      "grad_norm": 0.8207821249961853,
+      "learning_rate": 4.0111407910400555e-05,
+      "loss": 0.2134,
+      "step": 2114
+    },
+    {
+      "epoch": 1.2868877395801643,
+      "grad_norm": 0.9123344421386719,
+      "learning_rate": 4.010148880402326e-05,
+      "loss": 0.1845,
+      "step": 2115
+    },
+    {
+      "epoch": 1.2874961971402494,
+      "grad_norm": 0.9210152626037598,
+      "learning_rate": 4.009156595310619e-05,
+      "loss": 0.2483,
+      "step": 2116
+    },
+    {
+      "epoch": 1.2881046547003345,
+      "grad_norm": 0.9474868178367615,
+      "learning_rate": 4.0081639360109816e-05,
+      "loss": 0.2452,
+      "step": 2117
+    },
+    {
+      "epoch": 1.2887131122604198,
+      "grad_norm": 0.931167721748352,
+      "learning_rate": 4.007170902749552e-05,
+      "loss": 0.25,
+      "step": 2118
+    },
+    {
+      "epoch": 1.289321569820505,
+      "grad_norm": 0.9042659401893616,
+      "learning_rate": 4.00617749577256e-05,
+      "loss": 0.2474,
+      "step": 2119
+    },
+    {
+      "epoch": 1.2899300273805903,
+      "grad_norm": 0.9150558114051819,
+      "learning_rate": 4.0051837153263296e-05,
+      "loss": 0.2376,
+      "step": 2120
+    },
+    {
+      "epoch": 1.2905384849406754,
+      "grad_norm": 0.7283451557159424,
+      "learning_rate": 4.0041895616572765e-05,
+      "loss": 0.1869,
+      "step": 2121
+    },
+    {
+      "epoch": 1.2911469425007605,
+      "grad_norm": 0.8587160110473633,
+      "learning_rate": 4.0031950350119106e-05,
+      "loss": 0.2099,
+      "step": 2122
+    },
+    {
+      "epoch": 1.2917554000608458,
+      "grad_norm": 0.7844583988189697,
+      "learning_rate": 4.002200135636832e-05,
+      "loss": 0.1998,
+      "step": 2123
+    },
+    {
+      "epoch": 1.2923638576209309,
+      "grad_norm": 0.8835726380348206,
+      "learning_rate": 4.001204863778735e-05,
+      "loss": 0.2201,
+      "step": 2124
+    },
+    {
+      "epoch": 1.2929723151810162,
+      "grad_norm": 0.8643476366996765,
+      "learning_rate": 4.0002092196844046e-05,
+      "loss": 0.2101,
+      "step": 2125
+    },
+    {
+      "epoch": 1.2935807727411013,
+      "grad_norm": 0.8908064961433411,
+      "learning_rate": 3.999213203600719e-05,
+      "loss": 0.2303,
+      "step": 2126
+    },
+    {
+      "epoch": 1.2941892303011864,
+      "grad_norm": 1.0389541387557983,
+      "learning_rate": 3.99821681577465e-05,
+      "loss": 0.233,
+      "step": 2127
+    },
+    {
+      "epoch": 1.2947976878612717,
+      "grad_norm": 1.1947957277297974,
+      "learning_rate": 3.997220056453259e-05,
+      "loss": 0.2872,
+      "step": 2128
+    },
+    {
+      "epoch": 1.2954061454213568,
+      "grad_norm": 1.0398141145706177,
+      "learning_rate": 3.9962229258837013e-05,
+      "loss": 0.2141,
+      "step": 2129
+    },
+    {
+      "epoch": 1.296014602981442,
+      "grad_norm": 0.8543111085891724,
+      "learning_rate": 3.9952254243132243e-05,
+      "loss": 0.2153,
+      "step": 2130
+    },
+    {
+      "epoch": 1.2966230605415272,
+      "grad_norm": 1.0683271884918213,
+      "learning_rate": 3.9942275519891656e-05,
+      "loss": 0.2487,
+      "step": 2131
+    },
+    {
+      "epoch": 1.2972315181016123,
+      "grad_norm": 0.8721834421157837,
+      "learning_rate": 3.993229309158957e-05,
+      "loss": 0.2295,
+      "step": 2132
+    },
+    {
+      "epoch": 1.2978399756616976,
+      "grad_norm": 0.8389768004417419,
+      "learning_rate": 3.9922306960701196e-05,
+      "loss": 0.2181,
+      "step": 2133
+    },
+    {
+      "epoch": 1.2984484332217827,
+      "grad_norm": 0.9460014700889587,
+      "learning_rate": 3.991231712970269e-05,
+      "loss": 0.2342,
+      "step": 2134
+    },
+    {
+      "epoch": 1.299056890781868,
+      "grad_norm": 0.9399822950363159,
+      "learning_rate": 3.990232360107111e-05,
+      "loss": 0.2097,
+      "step": 2135
+    },
+    {
+      "epoch": 1.2996653483419531,
+      "grad_norm": 0.8793272376060486,
+      "learning_rate": 3.989232637728445e-05,
+      "loss": 0.2349,
+      "step": 2136
+    },
+    {
+      "epoch": 1.3002738059020382,
+      "grad_norm": 0.9901358485221863,
+      "learning_rate": 3.988232546082158e-05,
+      "loss": 0.2422,
+      "step": 2137
+    },
+    {
+      "epoch": 1.3008822634621235,
+      "grad_norm": 0.8974143862724304,
+      "learning_rate": 3.9872320854162324e-05,
+      "loss": 0.2433,
+      "step": 2138
+    },
+    {
+      "epoch": 1.3014907210222086,
+      "grad_norm": 0.9541013836860657,
+      "learning_rate": 3.9862312559787404e-05,
+      "loss": 0.2531,
+      "step": 2139
+    },
+    {
+      "epoch": 1.302099178582294,
+      "grad_norm": 1.2113054990768433,
+      "learning_rate": 3.985230058017846e-05,
+      "loss": 0.2654,
+      "step": 2140
+    },
+    {
+      "epoch": 1.302707636142379,
+      "grad_norm": 0.8780795931816101,
+      "learning_rate": 3.984228491781805e-05,
+      "loss": 0.2094,
+      "step": 2141
+    },
+    {
+      "epoch": 1.3033160937024642,
+      "grad_norm": 0.9655569791793823,
+      "learning_rate": 3.9832265575189635e-05,
+      "loss": 0.2571,
+      "step": 2142
+    },
+    {
+      "epoch": 1.3039245512625495,
+      "grad_norm": 0.9380378723144531,
+      "learning_rate": 3.98222425547776e-05,
+      "loss": 0.2191,
+      "step": 2143
+    },
+    {
+      "epoch": 1.3045330088226346,
+      "grad_norm": 0.887397825717926,
+      "learning_rate": 3.981221585906723e-05,
+      "loss": 0.2585,
+      "step": 2144
+    },
+    {
+      "epoch": 1.3051414663827199,
+      "grad_norm": 0.8565236926078796,
+      "learning_rate": 3.980218549054473e-05,
+      "loss": 0.1939,
+      "step": 2145
+    },
+    {
+      "epoch": 1.305749923942805,
+      "grad_norm": 0.8174847364425659,
+      "learning_rate": 3.979215145169721e-05,
+      "loss": 0.2119,
+      "step": 2146
+    },
+    {
+      "epoch": 1.30635838150289,
+      "grad_norm": 0.8997043371200562,
+      "learning_rate": 3.97821137450127e-05,
+      "loss": 0.2329,
+      "step": 2147
+    },
+    {
+      "epoch": 1.3069668390629754,
+      "grad_norm": 0.9833238124847412,
+      "learning_rate": 3.977207237298014e-05,
+      "loss": 0.2813,
+      "step": 2148
+    },
+    {
+      "epoch": 1.3075752966230605,
+      "grad_norm": 0.8356313705444336,
+      "learning_rate": 3.9762027338089356e-05,
+      "loss": 0.2254,
+      "step": 2149
+    },
+    {
+      "epoch": 1.3081837541831458,
+      "grad_norm": 1.0737109184265137,
+      "learning_rate": 3.97519786428311e-05,
+      "loss": 0.2502,
+      "step": 2150
+    },
+    {
+      "epoch": 1.308792211743231,
+      "grad_norm": 0.8580512404441833,
+      "learning_rate": 3.9741926289697036e-05,
+      "loss": 0.2491,
+      "step": 2151
+    },
+    {
+      "epoch": 1.309400669303316,
+      "grad_norm": 0.9692168831825256,
+      "learning_rate": 3.973187028117972e-05,
+      "loss": 0.2251,
+      "step": 2152
+    },
+    {
+      "epoch": 1.3100091268634013,
+      "grad_norm": 0.8899025917053223,
+      "learning_rate": 3.9721810619772636e-05,
+      "loss": 0.2117,
+      "step": 2153
+    },
+    {
+      "epoch": 1.3106175844234864,
+      "grad_norm": 0.784062385559082,
+      "learning_rate": 3.971174730797015e-05,
+      "loss": 0.2105,
+      "step": 2154
+    },
+    {
+      "epoch": 1.3112260419835717,
+      "grad_norm": 0.7936700582504272,
+      "learning_rate": 3.970168034826755e-05,
+      "loss": 0.1936,
+      "step": 2155
+    },
+    {
+      "epoch": 1.3118344995436568,
+      "grad_norm": 0.844158947467804,
+      "learning_rate": 3.9691609743161015e-05,
+      "loss": 0.2318,
+      "step": 2156
+    },
+    {
+      "epoch": 1.312442957103742,
+      "grad_norm": 0.9515892267227173,
+      "learning_rate": 3.968153549514765e-05,
+      "loss": 0.2188,
+      "step": 2157
+    },
+    {
+      "epoch": 1.3130514146638272,
+      "grad_norm": 0.8544421792030334,
+      "learning_rate": 3.967145760672543e-05,
+      "loss": 0.225,
+      "step": 2158
+    },
+    {
+      "epoch": 1.3136598722239123,
+      "grad_norm": 0.9000852704048157,
+      "learning_rate": 3.9661376080393266e-05,
+      "loss": 0.2036,
+      "step": 2159
+    },
+    {
+      "epoch": 1.3142683297839977,
+      "grad_norm": 0.9248766899108887,
+      "learning_rate": 3.965129091865094e-05,
+      "loss": 0.2235,
+      "step": 2160
+    },
+    {
+      "epoch": 1.3148767873440828,
+      "grad_norm": 1.002901315689087,
+      "learning_rate": 3.9641202123999174e-05,
+      "loss": 0.2248,
+      "step": 2161
+    },
+    {
+      "epoch": 1.3154852449041678,
+      "grad_norm": 0.8213080763816833,
+      "learning_rate": 3.963110969893955e-05,
+      "loss": 0.2088,
+      "step": 2162
+    },
+    {
+      "epoch": 1.3160937024642532,
+      "grad_norm": 1.0072423219680786,
+      "learning_rate": 3.9621013645974574e-05,
+      "loss": 0.2406,
+      "step": 2163
+    },
+    {
+      "epoch": 1.3167021600243383,
+      "grad_norm": 1.0771087408065796,
+      "learning_rate": 3.961091396760765e-05,
+      "loss": 0.2093,
+      "step": 2164
+    },
+    {
+      "epoch": 1.3173106175844236,
+      "grad_norm": 0.9404158592224121,
+      "learning_rate": 3.960081066634308e-05,
+      "loss": 0.2502,
+      "step": 2165
+    },
+    {
+      "epoch": 1.3179190751445087,
+      "grad_norm": 0.910447359085083,
+      "learning_rate": 3.959070374468605e-05,
+      "loss": 0.2086,
+      "step": 2166
+    },
+    {
+      "epoch": 1.3185275327045938,
+      "grad_norm": 0.7833483219146729,
+      "learning_rate": 3.958059320514267e-05,
+      "loss": 0.2102,
+      "step": 2167
+    },
+    {
+      "epoch": 1.319135990264679,
+      "grad_norm": 1.0158355236053467,
+      "learning_rate": 3.957047905021991e-05,
+      "loss": 0.2798,
+      "step": 2168
+    },
+    {
+      "epoch": 1.3197444478247642,
+      "grad_norm": 0.9133449196815491,
+      "learning_rate": 3.956036128242568e-05,
+      "loss": 0.2381,
+      "step": 2169
+    },
+    {
+      "epoch": 1.3203529053848495,
+      "grad_norm": 0.9535892009735107,
+      "learning_rate": 3.955023990426876e-05,
+      "loss": 0.2382,
+      "step": 2170
+    },
+    {
+      "epoch": 1.3209613629449346,
+      "grad_norm": 0.919695258140564,
+      "learning_rate": 3.954011491825883e-05,
+      "loss": 0.2257,
+      "step": 2171
+    },
+    {
+      "epoch": 1.3215698205050197,
+      "grad_norm": 0.8631134033203125,
+      "learning_rate": 3.952998632690646e-05,
+      "loss": 0.1773,
+      "step": 2172
+    },
+    {
+      "epoch": 1.322178278065105,
+      "grad_norm": 1.0971744060516357,
+      "learning_rate": 3.951985413272312e-05,
+      "loss": 0.251,
+      "step": 2173
+    },
+    {
+      "epoch": 1.3227867356251901,
+      "grad_norm": 0.8867610692977905,
+      "learning_rate": 3.950971833822117e-05,
+      "loss": 0.1989,
+      "step": 2174
+    },
+    {
+      "epoch": 1.3233951931852754,
+      "grad_norm": 0.8583669066429138,
+      "learning_rate": 3.949957894591387e-05,
+      "loss": 0.2238,
+      "step": 2175
+    },
+    {
+      "epoch": 1.3240036507453605,
+      "grad_norm": 0.8695077300071716,
+      "learning_rate": 3.9489435958315354e-05,
+      "loss": 0.2755,
+      "step": 2176
+    },
+    {
+      "epoch": 1.3246121083054456,
+      "grad_norm": 0.9294130206108093,
+      "learning_rate": 3.947928937794069e-05,
+      "loss": 0.2271,
+      "step": 2177
+    },
+    {
+      "epoch": 1.325220565865531,
+      "grad_norm": 0.871168851852417,
+      "learning_rate": 3.946913920730577e-05,
+      "loss": 0.215,
+      "step": 2178
+    },
+    {
+      "epoch": 1.325829023425616,
+      "grad_norm": 0.9917259812355042,
+      "learning_rate": 3.945898544892744e-05,
+      "loss": 0.2764,
+      "step": 2179
+    },
+    {
+      "epoch": 1.3264374809857014,
+      "grad_norm": 0.689329206943512,
+      "learning_rate": 3.944882810532339e-05,
+      "loss": 0.1973,
+      "step": 2180
+    },
+    {
+      "epoch": 1.3270459385457865,
+      "grad_norm": 0.8238869905471802,
+      "learning_rate": 3.943866717901223e-05,
+      "loss": 0.2145,
+      "step": 2181
+    },
+    {
+      "epoch": 1.3276543961058715,
+      "grad_norm": 0.8981241583824158,
+      "learning_rate": 3.9428502672513446e-05,
+      "loss": 0.2292,
+      "step": 2182
+    },
+    {
+      "epoch": 1.3282628536659569,
+      "grad_norm": 0.902793824672699,
+      "learning_rate": 3.9418334588347406e-05,
+      "loss": 0.2068,
+      "step": 2183
+    },
+    {
+      "epoch": 1.328871311226042,
+      "grad_norm": 1.0001530647277832,
+      "learning_rate": 3.9408162929035375e-05,
+      "loss": 0.282,
+      "step": 2184
+    },
+    {
+      "epoch": 1.3294797687861273,
+      "grad_norm": 0.8582153916358948,
+      "learning_rate": 3.939798769709949e-05,
+      "loss": 0.2009,
+      "step": 2185
+    },
+    {
+      "epoch": 1.3300882263462124,
+      "grad_norm": 1.967231035232544,
+      "learning_rate": 3.93878088950628e-05,
+      "loss": 0.2382,
+      "step": 2186
+    },
+    {
+      "epoch": 1.3306966839062975,
+      "grad_norm": 1.2576441764831543,
+      "learning_rate": 3.937762652544923e-05,
+      "loss": 0.3125,
+      "step": 2187
+    },
+    {
+      "epoch": 1.3313051414663828,
+      "grad_norm": 1.0156564712524414,
+      "learning_rate": 3.9367440590783554e-05,
+      "loss": 0.2311,
+      "step": 2188
+    },
+    {
+      "epoch": 1.3319135990264679,
+      "grad_norm": 0.9007154703140259,
+      "learning_rate": 3.9357251093591485e-05,
+      "loss": 0.2335,
+      "step": 2189
+    },
+    {
+      "epoch": 1.3325220565865532,
+      "grad_norm": 0.9337432980537415,
+      "learning_rate": 3.934705803639959e-05,
+      "loss": 0.2229,
+      "step": 2190
+    },
+    {
+      "epoch": 1.3331305141466383,
+      "grad_norm": 0.875648021697998,
+      "learning_rate": 3.9336861421735305e-05,
+      "loss": 0.2179,
+      "step": 2191
+    },
+    {
+      "epoch": 1.3337389717067234,
+      "grad_norm": 0.7988814115524292,
+      "learning_rate": 3.9326661252126984e-05,
+      "loss": 0.2233,
+      "step": 2192
+    },
+    {
+      "epoch": 1.3343474292668087,
+      "grad_norm": 0.8480890989303589,
+      "learning_rate": 3.931645753010383e-05,
+      "loss": 0.2495,
+      "step": 2193
+    },
+    {
+      "epoch": 1.3349558868268938,
+      "grad_norm": 0.9835476279258728,
+      "learning_rate": 3.9306250258195954e-05,
+      "loss": 0.2171,
+      "step": 2194
+    },
+    {
+      "epoch": 1.3355643443869791,
+      "grad_norm": 1.1172250509262085,
+      "learning_rate": 3.929603943893432e-05,
+      "loss": 0.2434,
+      "step": 2195
+    },
+    {
+      "epoch": 1.3361728019470642,
+      "grad_norm": 0.8083821535110474,
+      "learning_rate": 3.92858250748508e-05,
+      "loss": 0.2134,
+      "step": 2196
+    },
+    {
+      "epoch": 1.3367812595071493,
+      "grad_norm": 0.8030064702033997,
+      "learning_rate": 3.9275607168478126e-05,
+      "loss": 0.2281,
+      "step": 2197
+    },
+    {
+      "epoch": 1.3373897170672346,
+      "grad_norm": 0.8596853017807007,
+      "learning_rate": 3.926538572234991e-05,
+      "loss": 0.1805,
+      "step": 2198
+    },
+    {
+      "epoch": 1.3379981746273197,
+      "grad_norm": 1.0698977708816528,
+      "learning_rate": 3.925516073900064e-05,
+      "loss": 0.2718,
+      "step": 2199
+    },
+    {
+      "epoch": 1.338606632187405,
+      "grad_norm": 1.0942227840423584,
+      "learning_rate": 3.9244932220965704e-05,
+      "loss": 0.2411,
+      "step": 2200
+    },
+    {
+      "epoch": 1.3392150897474902,
+      "grad_norm": 0.9584268927574158,
+      "learning_rate": 3.9234700170781316e-05,
+      "loss": 0.277,
+      "step": 2201
+    },
+    {
+      "epoch": 1.3398235473075752,
+      "grad_norm": 1.1233789920806885,
+      "learning_rate": 3.922446459098463e-05,
+      "loss": 0.2515,
+      "step": 2202
+    },
+    {
+      "epoch": 1.3404320048676603,
+      "grad_norm": 1.0286003351211548,
+      "learning_rate": 3.9214225484113634e-05,
+      "loss": 0.199,
+      "step": 2203
+    },
+    {
+      "epoch": 1.3410404624277457,
+      "grad_norm": 0.920686662197113,
+      "learning_rate": 3.9203982852707184e-05,
+      "loss": 0.2214,
+      "step": 2204
+    },
+    {
+      "epoch": 1.341648919987831,
+      "grad_norm": 0.9834322333335876,
+      "learning_rate": 3.9193736699305044e-05,
+      "loss": 0.2745,
+      "step": 2205
+    },
+    {
+      "epoch": 1.342257377547916,
+      "grad_norm": 0.8372621536254883,
+      "learning_rate": 3.9183487026447824e-05,
+      "loss": 0.2355,
+      "step": 2206
+    },
+    {
+      "epoch": 1.3428658351080012,
+      "grad_norm": 0.7708995938301086,
+      "learning_rate": 3.9173233836677024e-05,
+      "loss": 0.1949,
+      "step": 2207
+    },
+    {
+      "epoch": 1.3434742926680863,
+      "grad_norm": 0.9579456448554993,
+      "learning_rate": 3.9162977132534996e-05,
+      "loss": 0.2197,
+      "step": 2208
+    },
+    {
+      "epoch": 1.3440827502281716,
+      "grad_norm": 0.8540910482406616,
+      "learning_rate": 3.915271691656498e-05,
+      "loss": 0.2385,
+      "step": 2209
+    },
+    {
+      "epoch": 1.344691207788257,
+      "grad_norm": 0.8921311497688293,
+      "learning_rate": 3.914245319131109e-05,
+      "loss": 0.2211,
+      "step": 2210
+    },
+    {
+      "epoch": 1.345299665348342,
+      "grad_norm": 0.8396940231323242,
+      "learning_rate": 3.913218595931829e-05,
+      "loss": 0.2176,
+      "step": 2211
+    },
+    {
+      "epoch": 1.345908122908427,
+      "grad_norm": 0.8367871642112732,
+      "learning_rate": 3.9121915223132436e-05,
+      "loss": 0.2457,
+      "step": 2212
+    },
+    {
+      "epoch": 1.3465165804685122,
+      "grad_norm": 0.9021058082580566,
+      "learning_rate": 3.911164098530023e-05,
+      "loss": 0.2251,
+      "step": 2213
+    },
+    {
+      "epoch": 1.3471250380285975,
+      "grad_norm": 0.911104679107666,
+      "learning_rate": 3.910136324836927e-05,
+      "loss": 0.2683,
+      "step": 2214
+    },
+    {
+      "epoch": 1.3477334955886826,
+      "grad_norm": 0.906808614730835,
+      "learning_rate": 3.909108201488799e-05,
+      "loss": 0.2472,
+      "step": 2215
+    },
+    {
+      "epoch": 1.348341953148768,
+      "grad_norm": 0.8513998985290527,
+      "learning_rate": 3.908079728740571e-05,
+      "loss": 0.205,
+      "step": 2216
+    },
+    {
+      "epoch": 1.348950410708853,
+      "grad_norm": 0.9133200645446777,
+      "learning_rate": 3.9070509068472635e-05,
+      "loss": 0.2453,
+      "step": 2217
+    },
+    {
+      "epoch": 1.3495588682689381,
+      "grad_norm": 0.8982611298561096,
+      "learning_rate": 3.90602173606398e-05,
+      "loss": 0.1938,
+      "step": 2218
+    },
+    {
+      "epoch": 1.3501673258290234,
+      "grad_norm": 1.0791575908660889,
+      "learning_rate": 3.90499221664591e-05,
+      "loss": 0.2885,
+      "step": 2219
+    },
+    {
+      "epoch": 1.3507757833891085,
+      "grad_norm": 0.8639183044433594,
+      "learning_rate": 3.9039623488483346e-05,
+      "loss": 0.2382,
+      "step": 2220
+    },
+    {
+      "epoch": 1.3513842409491938,
+      "grad_norm": 1.2875553369522095,
+      "learning_rate": 3.902932132926616e-05,
+      "loss": 0.2692,
+      "step": 2221
+    },
+    {
+      "epoch": 1.351992698509279,
+      "grad_norm": 0.9130270481109619,
+      "learning_rate": 3.901901569136206e-05,
+      "loss": 0.2361,
+      "step": 2222
+    },
+    {
+      "epoch": 1.352601156069364,
+      "grad_norm": 0.832845151424408,
+      "learning_rate": 3.900870657732641e-05,
+      "loss": 0.2147,
+      "step": 2223
+    },
+    {
+      "epoch": 1.3532096136294494,
+      "grad_norm": 1.053515911102295,
+      "learning_rate": 3.8998393989715434e-05,
+      "loss": 0.2453,
+      "step": 2224
+    },
+    {
+      "epoch": 1.3538180711895345,
+      "grad_norm": 0.9339423179626465,
+      "learning_rate": 3.898807793108624e-05,
+      "loss": 0.1898,
+      "step": 2225
+    },
+    {
+      "epoch": 1.3544265287496198,
+      "grad_norm": 0.9041353464126587,
+      "learning_rate": 3.8977758403996765e-05,
+      "loss": 0.2065,
+      "step": 2226
+    },
+    {
+      "epoch": 1.3550349863097049,
+      "grad_norm": 0.8681005239486694,
+      "learning_rate": 3.896743541100583e-05,
+      "loss": 0.2186,
+      "step": 2227
+    },
+    {
+      "epoch": 1.35564344386979,
+      "grad_norm": 0.8641691207885742,
+      "learning_rate": 3.89571089546731e-05,
+      "loss": 0.2016,
+      "step": 2228
+    },
+    {
+      "epoch": 1.3562519014298753,
+      "grad_norm": 0.8914021253585815,
+      "learning_rate": 3.8946779037559115e-05,
+      "loss": 0.2332,
+      "step": 2229
+    },
+    {
+      "epoch": 1.3568603589899604,
+      "grad_norm": 0.9768247604370117,
+      "learning_rate": 3.8936445662225264e-05,
+      "loss": 0.1901,
+      "step": 2230
+    },
+    {
+      "epoch": 1.3574688165500457,
+      "grad_norm": 0.982510507106781,
+      "learning_rate": 3.892610883123378e-05,
+      "loss": 0.2356,
+      "step": 2231
+    },
+    {
+      "epoch": 1.3580772741101308,
+      "grad_norm": 0.9750663638114929,
+      "learning_rate": 3.891576854714777e-05,
+      "loss": 0.2241,
+      "step": 2232
+    },
+    {
+      "epoch": 1.358685731670216,
+      "grad_norm": 0.8455461263656616,
+      "learning_rate": 3.890542481253121e-05,
+      "loss": 0.2119,
+      "step": 2233
+    },
+    {
+      "epoch": 1.3592941892303012,
+      "grad_norm": 1.088861107826233,
+      "learning_rate": 3.88950776299489e-05,
+      "loss": 0.2778,
+      "step": 2234
+    },
+    {
+      "epoch": 1.3599026467903863,
+      "grad_norm": 1.2681231498718262,
+      "learning_rate": 3.888472700196651e-05,
+      "loss": 0.2204,
+      "step": 2235
+    },
+    {
+      "epoch": 1.3605111043504716,
+      "grad_norm": 0.9089130163192749,
+      "learning_rate": 3.887437293115057e-05,
+      "loss": 0.2306,
+      "step": 2236
+    },
+    {
+      "epoch": 1.3611195619105567,
+      "grad_norm": 1.256752610206604,
+      "learning_rate": 3.8864015420068454e-05,
+      "loss": 0.2447,
+      "step": 2237
+    },
+    {
+      "epoch": 1.3617280194706418,
+      "grad_norm": 0.9000508189201355,
+      "learning_rate": 3.88536544712884e-05,
+      "loss": 0.2213,
+      "step": 2238
+    },
+    {
+      "epoch": 1.3623364770307271,
+      "grad_norm": 1.016454815864563,
+      "learning_rate": 3.884329008737947e-05,
+      "loss": 0.2309,
+      "step": 2239
+    },
+    {
+      "epoch": 1.3629449345908122,
+      "grad_norm": 0.9450438618659973,
+      "learning_rate": 3.883292227091163e-05,
+      "loss": 0.2443,
+      "step": 2240
+    },
+    {
+      "epoch": 1.3635533921508975,
+      "grad_norm": 1.0585530996322632,
+      "learning_rate": 3.882255102445565e-05,
+      "loss": 0.2413,
+      "step": 2241
+    },
+    {
+      "epoch": 1.3641618497109826,
+      "grad_norm": 0.9803087115287781,
+      "learning_rate": 3.8812176350583164e-05,
+      "loss": 0.219,
+      "step": 2242
+    },
+    {
+      "epoch": 1.3647703072710677,
+      "grad_norm": 0.93431156873703,
+      "learning_rate": 3.880179825186667e-05,
+      "loss": 0.2136,
+      "step": 2243
+    },
+    {
+      "epoch": 1.365378764831153,
+      "grad_norm": 0.9018591642379761,
+      "learning_rate": 3.879141673087949e-05,
+      "loss": 0.2249,
+      "step": 2244
+    },
+    {
+      "epoch": 1.3659872223912382,
+      "grad_norm": 1.3140918016433716,
+      "learning_rate": 3.878103179019581e-05,
+      "loss": 0.2312,
+      "step": 2245
+    },
+    {
+      "epoch": 1.3665956799513235,
+      "grad_norm": 0.9108927249908447,
+      "learning_rate": 3.877064343239068e-05,
+      "loss": 0.2274,
+      "step": 2246
+    },
+    {
+      "epoch": 1.3672041375114086,
+      "grad_norm": 1.067284107208252,
+      "learning_rate": 3.8760251660039956e-05,
+      "loss": 0.2968,
+      "step": 2247
+    },
+    {
+      "epoch": 1.3678125950714937,
+      "grad_norm": 0.8846755623817444,
+      "learning_rate": 3.874985647572039e-05,
+      "loss": 0.2381,
+      "step": 2248
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.9099476933479309,
+      "learning_rate": 3.8739457882009526e-05,
+      "loss": 0.1729,
+      "step": 2249
+    },
+    {
+      "epoch": 1.369029510191664,
+      "grad_norm": 0.866028904914856,
+      "learning_rate": 3.87290558814858e-05,
+      "loss": 0.2237,
+      "step": 2250
+    },
+    {
+      "epoch": 1.3696379677517494,
+      "grad_norm": 0.8966884016990662,
+      "learning_rate": 3.871865047672848e-05,
+      "loss": 0.2003,
+      "step": 2251
+    },
+    {
+      "epoch": 1.3702464253118345,
+      "grad_norm": 0.9942765831947327,
+      "learning_rate": 3.8708241670317645e-05,
+      "loss": 0.2374,
+      "step": 2252
+    },
+    {
+      "epoch": 1.3708548828719196,
+      "grad_norm": 0.9010903239250183,
+      "learning_rate": 3.869782946483428e-05,
+      "loss": 0.1948,
+      "step": 2253
+    },
+    {
+      "epoch": 1.371463340432005,
+      "grad_norm": 0.9701775908470154,
+      "learning_rate": 3.868741386286016e-05,
+      "loss": 0.2412,
+      "step": 2254
+    },
+    {
+      "epoch": 1.37207179799209,
+      "grad_norm": 0.7762331366539001,
+      "learning_rate": 3.867699486697791e-05,
+      "loss": 0.1873,
+      "step": 2255
+    },
+    {
+      "epoch": 1.3726802555521753,
+      "grad_norm": 0.9885874390602112,
+      "learning_rate": 3.866657247977103e-05,
+      "loss": 0.2072,
+      "step": 2256
+    },
+    {
+      "epoch": 1.3732887131122604,
+      "grad_norm": 0.9853124618530273,
+      "learning_rate": 3.865614670382382e-05,
+      "loss": 0.2562,
+      "step": 2257
+    },
+    {
+      "epoch": 1.3738971706723455,
+      "grad_norm": 0.9936604499816895,
+      "learning_rate": 3.864571754172144e-05,
+      "loss": 0.2386,
+      "step": 2258
+    },
+    {
+      "epoch": 1.3745056282324308,
+      "grad_norm": 0.9598672389984131,
+      "learning_rate": 3.8635284996049904e-05,
+      "loss": 0.2379,
+      "step": 2259
+    },
+    {
+      "epoch": 1.375114085792516,
+      "grad_norm": 0.9736387133598328,
+      "learning_rate": 3.8624849069396024e-05,
+      "loss": 0.2614,
+      "step": 2260
+    },
+    {
+      "epoch": 1.3757225433526012,
+      "grad_norm": 0.8407406210899353,
+      "learning_rate": 3.8614409764347494e-05,
+      "loss": 0.2115,
+      "step": 2261
+    },
+    {
+      "epoch": 1.3763310009126863,
+      "grad_norm": 0.8771750330924988,
+      "learning_rate": 3.860396708349281e-05,
+      "loss": 0.2091,
+      "step": 2262
+    },
+    {
+      "epoch": 1.3769394584727714,
+      "grad_norm": 0.9052261114120483,
+      "learning_rate": 3.859352102942134e-05,
+      "loss": 0.2031,
+      "step": 2263
+    },
+    {
+      "epoch": 1.3775479160328568,
+      "grad_norm": 0.8526843190193176,
+      "learning_rate": 3.8583071604723256e-05,
+      "loss": 0.1998,
+      "step": 2264
+    },
+    {
+      "epoch": 1.3781563735929419,
+      "grad_norm": 1.0175981521606445,
+      "learning_rate": 3.857261881198958e-05,
+      "loss": 0.209,
+      "step": 2265
+    },
+    {
+      "epoch": 1.3787648311530272,
+      "grad_norm": 1.1324195861816406,
+      "learning_rate": 3.856216265381219e-05,
+      "loss": 0.2031,
+      "step": 2266
+    },
+    {
+      "epoch": 1.3793732887131123,
+      "grad_norm": 1.1673684120178223,
+      "learning_rate": 3.8551703132783745e-05,
+      "loss": 0.2526,
+      "step": 2267
+    },
+    {
+      "epoch": 1.3799817462731974,
+      "grad_norm": 0.929223895072937,
+      "learning_rate": 3.854124025149778e-05,
+      "loss": 0.2443,
+      "step": 2268
+    },
+    {
+      "epoch": 1.3805902038332827,
+      "grad_norm": 0.9426333904266357,
+      "learning_rate": 3.8530774012548674e-05,
+      "loss": 0.2002,
+      "step": 2269
+    },
+    {
+      "epoch": 1.3811986613933678,
+      "grad_norm": 0.7749611139297485,
+      "learning_rate": 3.85203044185316e-05,
+      "loss": 0.2044,
+      "step": 2270
+    },
+    {
+      "epoch": 1.381807118953453,
+      "grad_norm": 0.8301417231559753,
+      "learning_rate": 3.8509831472042585e-05,
+      "loss": 0.2198,
+      "step": 2271
+    },
+    {
+      "epoch": 1.3824155765135382,
+      "grad_norm": 1.038061499595642,
+      "learning_rate": 3.849935517567848e-05,
+      "loss": 0.2578,
+      "step": 2272
+    },
+    {
+      "epoch": 1.3830240340736233,
+      "grad_norm": 0.9880487322807312,
+      "learning_rate": 3.8488875532036975e-05,
+      "loss": 0.2688,
+      "step": 2273
+    },
+    {
+      "epoch": 1.3836324916337086,
+      "grad_norm": 0.9047033190727234,
+      "learning_rate": 3.847839254371658e-05,
+      "loss": 0.2179,
+      "step": 2274
+    },
+    {
+      "epoch": 1.3842409491937937,
+      "grad_norm": 0.9710778594017029,
+      "learning_rate": 3.8467906213316636e-05,
+      "loss": 0.2569,
+      "step": 2275
+    },
+    {
+      "epoch": 1.384849406753879,
+      "grad_norm": 1.0498100519180298,
+      "learning_rate": 3.845741654343733e-05,
+      "loss": 0.2389,
+      "step": 2276
+    },
+    {
+      "epoch": 1.3854578643139641,
+      "grad_norm": 0.9659872055053711,
+      "learning_rate": 3.8446923536679644e-05,
+      "loss": 0.2312,
+      "step": 2277
+    },
+    {
+      "epoch": 1.3860663218740492,
+      "grad_norm": 0.956907331943512,
+      "learning_rate": 3.843642719564542e-05,
+      "loss": 0.2527,
+      "step": 2278
+    },
+    {
+      "epoch": 1.3866747794341345,
+      "grad_norm": 0.9140104651451111,
+      "learning_rate": 3.842592752293731e-05,
+      "loss": 0.2255,
+      "step": 2279
+    },
+    {
+      "epoch": 1.3872832369942196,
+      "grad_norm": 0.8140145540237427,
+      "learning_rate": 3.8415424521158804e-05,
+      "loss": 0.2145,
+      "step": 2280
+    },
+    {
+      "epoch": 1.387891694554305,
+      "grad_norm": 0.9981208443641663,
+      "learning_rate": 3.8404918192914184e-05,
+      "loss": 0.2338,
+      "step": 2281
+    },
+    {
+      "epoch": 1.38850015211439,
+      "grad_norm": 1.1599704027175903,
+      "learning_rate": 3.839440854080861e-05,
+      "loss": 0.3089,
+      "step": 2282
+    },
+    {
+      "epoch": 1.3891086096744751,
+      "grad_norm": 0.9346492886543274,
+      "learning_rate": 3.8383895567448015e-05,
+      "loss": 0.1997,
+      "step": 2283
+    },
+    {
+      "epoch": 1.3897170672345605,
+      "grad_norm": 0.9026308655738831,
+      "learning_rate": 3.8373379275439194e-05,
+      "loss": 0.198,
+      "step": 2284
+    },
+    {
+      "epoch": 1.3903255247946456,
+      "grad_norm": 0.8270521759986877,
+      "learning_rate": 3.836285966738974e-05,
+      "loss": 0.2308,
+      "step": 2285
+    },
+    {
+      "epoch": 1.3909339823547309,
+      "grad_norm": 0.8808834552764893,
+      "learning_rate": 3.8352336745908076e-05,
+      "loss": 0.2238,
+      "step": 2286
+    },
+    {
+      "epoch": 1.391542439914816,
+      "grad_norm": 0.8379645347595215,
+      "learning_rate": 3.834181051360346e-05,
+      "loss": 0.1938,
+      "step": 2287
+    },
+    {
+      "epoch": 1.392150897474901,
+      "grad_norm": 0.894270658493042,
+      "learning_rate": 3.833128097308594e-05,
+      "loss": 0.2172,
+      "step": 2288
+    },
+    {
+      "epoch": 1.3927593550349864,
+      "grad_norm": 0.8662185668945312,
+      "learning_rate": 3.832074812696642e-05,
+      "loss": 0.2332,
+      "step": 2289
+    },
+    {
+      "epoch": 1.3933678125950715,
+      "grad_norm": 1.1685889959335327,
+      "learning_rate": 3.8310211977856605e-05,
+      "loss": 0.254,
+      "step": 2290
+    },
+    {
+      "epoch": 1.3939762701551568,
+      "grad_norm": 0.9299534559249878,
+      "learning_rate": 3.8299672528369014e-05,
+      "loss": 0.2024,
+      "step": 2291
+    },
+    {
+      "epoch": 1.394584727715242,
+      "grad_norm": 0.7964719533920288,
+      "learning_rate": 3.828912978111699e-05,
+      "loss": 0.1907,
+      "step": 2292
+    },
+    {
+      "epoch": 1.395193185275327,
+      "grad_norm": 0.9740342497825623,
+      "learning_rate": 3.8278583738714696e-05,
+      "loss": 0.2479,
+      "step": 2293
+    },
+    {
+      "epoch": 1.3958016428354123,
+      "grad_norm": 1.016251564025879,
+      "learning_rate": 3.826803440377712e-05,
+      "loss": 0.2517,
+      "step": 2294
+    },
+    {
+      "epoch": 1.3964101003954974,
+      "grad_norm": 1.0677728652954102,
+      "learning_rate": 3.8257481778920045e-05,
+      "loss": 0.23,
+      "step": 2295
+    },
+    {
+      "epoch": 1.3970185579555827,
+      "grad_norm": 0.9394170045852661,
+      "learning_rate": 3.824692586676009e-05,
+      "loss": 0.2383,
+      "step": 2296
+    },
+    {
+      "epoch": 1.3976270155156678,
+      "grad_norm": 0.9719864130020142,
+      "learning_rate": 3.823636666991468e-05,
+      "loss": 0.2135,
+      "step": 2297
+    },
+    {
+      "epoch": 1.398235473075753,
+      "grad_norm": 0.9637436270713806,
+      "learning_rate": 3.8225804191002054e-05,
+      "loss": 0.2471,
+      "step": 2298
+    },
+    {
+      "epoch": 1.3988439306358382,
+      "grad_norm": 1.0839718580245972,
+      "learning_rate": 3.821523843264127e-05,
+      "loss": 0.2424,
+      "step": 2299
+    },
+    {
+      "epoch": 1.3994523881959233,
+      "grad_norm": 0.9015427231788635,
+      "learning_rate": 3.82046693974522e-05,
+      "loss": 0.2568,
+      "step": 2300
+    },
+    {
+      "epoch": 1.4000608457560086,
+      "grad_norm": 0.9065006971359253,
+      "learning_rate": 3.8194097088055505e-05,
+      "loss": 0.1932,
+      "step": 2301
+    },
+    {
+      "epoch": 1.4006693033160937,
+      "grad_norm": 0.876019299030304,
+      "learning_rate": 3.818352150707269e-05,
+      "loss": 0.2278,
+      "step": 2302
+    },
+    {
+      "epoch": 1.4012777608761788,
+      "grad_norm": 0.8508959412574768,
+      "learning_rate": 3.817294265712606e-05,
+      "loss": 0.2087,
+      "step": 2303
+    },
+    {
+      "epoch": 1.4018862184362642,
+      "grad_norm": 0.8045951128005981,
+      "learning_rate": 3.8162360540838726e-05,
+      "loss": 0.1957,
+      "step": 2304
+    },
+    {
+      "epoch": 1.4024946759963492,
+      "grad_norm": 0.893156111240387,
+      "learning_rate": 3.815177516083461e-05,
+      "loss": 0.1938,
+      "step": 2305
+    },
+    {
+      "epoch": 1.4031031335564346,
+      "grad_norm": 0.9133938550949097,
+      "learning_rate": 3.8141186519738456e-05,
+      "loss": 0.2445,
+      "step": 2306
+    },
+    {
+      "epoch": 1.4037115911165197,
+      "grad_norm": 0.9459069967269897,
+      "learning_rate": 3.8130594620175786e-05,
+      "loss": 0.2532,
+      "step": 2307
+    },
+    {
+      "epoch": 1.4043200486766048,
+      "grad_norm": 0.9361640810966492,
+      "learning_rate": 3.811999946477296e-05,
+      "loss": 0.2845,
+      "step": 2308
+    },
+    {
+      "epoch": 1.4049285062366899,
+      "grad_norm": 0.7425569295883179,
+      "learning_rate": 3.810940105615715e-05,
+      "loss": 0.1709,
+      "step": 2309
+    },
+    {
+      "epoch": 1.4055369637967752,
+      "grad_norm": 0.9451609253883362,
+      "learning_rate": 3.8098799396956284e-05,
+      "loss": 0.2082,
+      "step": 2310
+    },
+    {
+      "epoch": 1.4061454213568605,
+      "grad_norm": 0.9120255708694458,
+      "learning_rate": 3.808819448979917e-05,
+      "loss": 0.2128,
+      "step": 2311
+    },
+    {
+      "epoch": 1.4067538789169456,
+      "grad_norm": 0.9225917458534241,
+      "learning_rate": 3.8077586337315365e-05,
+      "loss": 0.2095,
+      "step": 2312
+    },
+    {
+      "epoch": 1.4073623364770307,
+      "grad_norm": 0.9341639876365662,
+      "learning_rate": 3.8066974942135234e-05,
+      "loss": 0.2372,
+      "step": 2313
+    },
+    {
+      "epoch": 1.4079707940371158,
+      "grad_norm": 0.9186847805976868,
+      "learning_rate": 3.8056360306889985e-05,
+      "loss": 0.2227,
+      "step": 2314
+    },
+    {
+      "epoch": 1.408579251597201,
+      "grad_norm": 0.8264140486717224,
+      "learning_rate": 3.8045742434211595e-05,
+      "loss": 0.1934,
+      "step": 2315
+    },
+    {
+      "epoch": 1.4091877091572864,
+      "grad_norm": 0.8795403242111206,
+      "learning_rate": 3.803512132673286e-05,
+      "loss": 0.2349,
+      "step": 2316
+    },
+    {
+      "epoch": 1.4097961667173715,
+      "grad_norm": 0.9693627953529358,
+      "learning_rate": 3.802449698708736e-05,
+      "loss": 0.2146,
+      "step": 2317
+    },
+    {
+      "epoch": 1.4104046242774566,
+      "grad_norm": 0.9687555432319641,
+      "learning_rate": 3.8013869417909496e-05,
+      "loss": 0.2455,
+      "step": 2318
+    },
+    {
+      "epoch": 1.4110130818375417,
+      "grad_norm": 4.701800346374512,
+      "learning_rate": 3.800323862183446e-05,
+      "loss": 0.2436,
+      "step": 2319
+    },
+    {
+      "epoch": 1.411621539397627,
+      "grad_norm": 0.9624002575874329,
+      "learning_rate": 3.799260460149825e-05,
+      "loss": 0.2143,
+      "step": 2320
+    },
+    {
+      "epoch": 1.4122299969577121,
+      "grad_norm": 1.0840182304382324,
+      "learning_rate": 3.7981967359537656e-05,
+      "loss": 0.219,
+      "step": 2321
+    },
+    {
+      "epoch": 1.4128384545177974,
+      "grad_norm": 0.9719953536987305,
+      "learning_rate": 3.797132689859027e-05,
+      "loss": 0.2347,
+      "step": 2322
+    },
+    {
+      "epoch": 1.4134469120778825,
+      "grad_norm": 0.9404840469360352,
+      "learning_rate": 3.796068322129449e-05,
+      "loss": 0.2089,
+      "step": 2323
+    },
+    {
+      "epoch": 1.4140553696379676,
+      "grad_norm": 0.7975339889526367,
+      "learning_rate": 3.795003633028949e-05,
+      "loss": 0.2036,
+      "step": 2324
+    },
+    {
+      "epoch": 1.414663827198053,
+      "grad_norm": 0.9089730978012085,
+      "learning_rate": 3.793938622821528e-05,
+      "loss": 0.2158,
+      "step": 2325
+    },
+    {
+      "epoch": 1.415272284758138,
+      "grad_norm": 1.023104190826416,
+      "learning_rate": 3.792873291771261e-05,
+      "loss": 0.2297,
+      "step": 2326
+    },
+    {
+      "epoch": 1.4158807423182234,
+      "grad_norm": 0.9837586283683777,
+      "learning_rate": 3.7918076401423076e-05,
+      "loss": 0.2383,
+      "step": 2327
+    },
+    {
+      "epoch": 1.4164891998783085,
+      "grad_norm": 0.9490445852279663,
+      "learning_rate": 3.790741668198906e-05,
+      "loss": 0.2691,
+      "step": 2328
+    },
+    {
+      "epoch": 1.4170976574383936,
+      "grad_norm": 1.8430839776992798,
+      "learning_rate": 3.7896753762053693e-05,
+      "loss": 0.2033,
+      "step": 2329
+    },
+    {
+      "epoch": 1.4177061149984789,
+      "grad_norm": 0.8883879780769348,
+      "learning_rate": 3.788608764426097e-05,
+      "loss": 0.2133,
+      "step": 2330
+    },
+    {
+      "epoch": 1.418314572558564,
+      "grad_norm": 0.9225703477859497,
+      "learning_rate": 3.787541833125563e-05,
+      "loss": 0.1979,
+      "step": 2331
+    },
+    {
+      "epoch": 1.4189230301186493,
+      "grad_norm": 0.8261762261390686,
+      "learning_rate": 3.786474582568321e-05,
+      "loss": 0.2257,
+      "step": 2332
+    },
+    {
+      "epoch": 1.4195314876787344,
+      "grad_norm": 0.8989948630332947,
+      "learning_rate": 3.785407013019006e-05,
+      "loss": 0.2048,
+      "step": 2333
+    },
+    {
+      "epoch": 1.4201399452388195,
+      "grad_norm": 0.8330014944076538,
+      "learning_rate": 3.78433912474233e-05,
+      "loss": 0.2363,
+      "step": 2334
+    },
+    {
+      "epoch": 1.4207484027989048,
+      "grad_norm": 0.8741576075553894,
+      "learning_rate": 3.783270918003085e-05,
+      "loss": 0.2149,
+      "step": 2335
+    },
+    {
+      "epoch": 1.42135686035899,
+      "grad_norm": 0.9070691466331482,
+      "learning_rate": 3.782202393066141e-05,
+      "loss": 0.2316,
+      "step": 2336
+    },
+    {
+      "epoch": 1.4219653179190752,
+      "grad_norm": 0.9362895488739014,
+      "learning_rate": 3.7811335501964495e-05,
+      "loss": 0.2179,
+      "step": 2337
+    },
+    {
+      "epoch": 1.4225737754791603,
+      "grad_norm": 0.8994218111038208,
+      "learning_rate": 3.7800643896590375e-05,
+      "loss": 0.1861,
+      "step": 2338
+    },
+    {
+      "epoch": 1.4231822330392454,
+      "grad_norm": 0.9640007019042969,
+      "learning_rate": 3.778994911719013e-05,
+      "loss": 0.2578,
+      "step": 2339
+    },
+    {
+      "epoch": 1.4237906905993307,
+      "grad_norm": 0.8841598629951477,
+      "learning_rate": 3.7779251166415606e-05,
+      "loss": 0.2208,
+      "step": 2340
+    },
+    {
+      "epoch": 1.4243991481594158,
+      "grad_norm": 0.7951266765594482,
+      "learning_rate": 3.776855004691946e-05,
+      "loss": 0.1759,
+      "step": 2341
+    },
+    {
+      "epoch": 1.4250076057195011,
+      "grad_norm": 0.9464953541755676,
+      "learning_rate": 3.775784576135513e-05,
+      "loss": 0.2117,
+      "step": 2342
+    },
+    {
+      "epoch": 1.4256160632795862,
+      "grad_norm": 0.8788123726844788,
+      "learning_rate": 3.774713831237682e-05,
+      "loss": 0.2078,
+      "step": 2343
+    },
+    {
+      "epoch": 1.4262245208396713,
+      "grad_norm": 0.9868049621582031,
+      "learning_rate": 3.7736427702639526e-05,
+      "loss": 0.2355,
+      "step": 2344
+    },
+    {
+      "epoch": 1.4268329783997566,
+      "grad_norm": 1.0244841575622559,
+      "learning_rate": 3.7725713934799045e-05,
+      "loss": 0.2093,
+      "step": 2345
+    },
+    {
+      "epoch": 1.4274414359598417,
+      "grad_norm": 0.9810612797737122,
+      "learning_rate": 3.7714997011511956e-05,
+      "loss": 0.1909,
+      "step": 2346
+    },
+    {
+      "epoch": 1.428049893519927,
+      "grad_norm": 0.8638327717781067,
+      "learning_rate": 3.770427693543558e-05,
+      "loss": 0.209,
+      "step": 2347
+    },
+    {
+      "epoch": 1.4286583510800122,
+      "grad_norm": 0.9180063009262085,
+      "learning_rate": 3.769355370922807e-05,
+      "loss": 0.1966,
+      "step": 2348
+    },
+    {
+      "epoch": 1.4292668086400973,
+      "grad_norm": 2.471264600753784,
+      "learning_rate": 3.768282733554833e-05,
+      "loss": 0.189,
+      "step": 2349
+    },
+    {
+      "epoch": 1.4298752662001826,
+      "grad_norm": 0.8565406799316406,
+      "learning_rate": 3.7672097817056065e-05,
+      "loss": 0.2314,
+      "step": 2350
+    },
+    {
+      "epoch": 1.4304837237602677,
+      "grad_norm": 0.9157159924507141,
+      "learning_rate": 3.766136515641174e-05,
+      "loss": 0.2094,
+      "step": 2351
+    },
+    {
+      "epoch": 1.431092181320353,
+      "grad_norm": 0.7455861568450928,
+      "learning_rate": 3.76506293562766e-05,
+      "loss": 0.1656,
+      "step": 2352
+    },
+    {
+      "epoch": 1.431700638880438,
+      "grad_norm": 0.8483655452728271,
+      "learning_rate": 3.7639890419312694e-05,
+      "loss": 0.2195,
+      "step": 2353
+    },
+    {
+      "epoch": 1.4323090964405232,
+      "grad_norm": 0.9521898627281189,
+      "learning_rate": 3.7629148348182807e-05,
+      "loss": 0.1907,
+      "step": 2354
+    },
+    {
+      "epoch": 1.4329175540006085,
+      "grad_norm": 0.9046710133552551,
+      "learning_rate": 3.761840314555055e-05,
+      "loss": 0.1728,
+      "step": 2355
+    },
+    {
+      "epoch": 1.4335260115606936,
+      "grad_norm": 0.9136427044868469,
+      "learning_rate": 3.760765481408027e-05,
+      "loss": 0.2079,
+      "step": 2356
+    },
+    {
+      "epoch": 1.434134469120779,
+      "grad_norm": 0.7689423561096191,
+      "learning_rate": 3.759690335643711e-05,
+      "loss": 0.1898,
+      "step": 2357
+    },
+    {
+      "epoch": 1.434742926680864,
+      "grad_norm": 0.868561327457428,
+      "learning_rate": 3.758614877528698e-05,
+      "loss": 0.1768,
+      "step": 2358
+    },
+    {
+      "epoch": 1.435351384240949,
+      "grad_norm": 0.889870285987854,
+      "learning_rate": 3.7575391073296575e-05,
+      "loss": 0.2147,
+      "step": 2359
+    },
+    {
+      "epoch": 1.4359598418010344,
+      "grad_norm": 1.013046145439148,
+      "learning_rate": 3.756463025313335e-05,
+      "loss": 0.2365,
+      "step": 2360
+    },
+    {
+      "epoch": 1.4365682993611195,
+      "grad_norm": 1.0037925243377686,
+      "learning_rate": 3.7553866317465536e-05,
+      "loss": 0.2229,
+      "step": 2361
+    },
+    {
+      "epoch": 1.4371767569212048,
+      "grad_norm": 0.910197913646698,
+      "learning_rate": 3.754309926896215e-05,
+      "loss": 0.2133,
+      "step": 2362
+    },
+    {
+      "epoch": 1.43778521448129,
+      "grad_norm": 0.8792757987976074,
+      "learning_rate": 3.7532329110292966e-05,
+      "loss": 0.2481,
+      "step": 2363
+    },
+    {
+      "epoch": 1.438393672041375,
+      "grad_norm": 0.8445896506309509,
+      "learning_rate": 3.752155584412854e-05,
+      "loss": 0.2402,
+      "step": 2364
+    },
+    {
+      "epoch": 1.4390021296014603,
+      "grad_norm": 0.8201121687889099,
+      "learning_rate": 3.751077947314019e-05,
+      "loss": 0.1696,
+      "step": 2365
+    },
+    {
+      "epoch": 1.4396105871615454,
+      "grad_norm": 0.9546027183532715,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.2487,
+      "step": 2366
+    },
+    {
+      "epoch": 1.4402190447216308,
+      "grad_norm": 0.9001442193984985,
+      "learning_rate": 3.748921742738084e-05,
+      "loss": 0.2163,
+      "step": 2367
+    },
+    {
+      "epoch": 1.4408275022817159,
+      "grad_norm": 0.7833166122436523,
+      "learning_rate": 3.7478431757956335e-05,
+      "loss": 0.2015,
+      "step": 2368
+    },
+    {
+      "epoch": 1.441435959841801,
+      "grad_norm": 0.7708922028541565,
+      "learning_rate": 3.746764299440087e-05,
+      "loss": 0.2106,
+      "step": 2369
+    },
+    {
+      "epoch": 1.4420444174018863,
+      "grad_norm": 0.7704883217811584,
+      "learning_rate": 3.745685113938963e-05,
+      "loss": 0.2011,
+      "step": 2370
+    },
+    {
+      "epoch": 1.4426528749619714,
+      "grad_norm": 0.7907694578170776,
+      "learning_rate": 3.7446056195598536e-05,
+      "loss": 0.2023,
+      "step": 2371
+    },
+    {
+      "epoch": 1.4432613325220567,
+      "grad_norm": 1.1114675998687744,
+      "learning_rate": 3.743525816570428e-05,
+      "loss": 0.1887,
+      "step": 2372
+    },
+    {
+      "epoch": 1.4438697900821418,
+      "grad_norm": 0.7749055624008179,
+      "learning_rate": 3.742445705238432e-05,
+      "loss": 0.1806,
+      "step": 2373
+    },
+    {
+      "epoch": 1.4444782476422269,
+      "grad_norm": 0.8483683466911316,
+      "learning_rate": 3.74136528583169e-05,
+      "loss": 0.1959,
+      "step": 2374
+    },
+    {
+      "epoch": 1.4450867052023122,
+      "grad_norm": 0.9588725566864014,
+      "learning_rate": 3.740284558618099e-05,
+      "loss": 0.2355,
+      "step": 2375
+    },
+    {
+      "epoch": 1.4456951627623973,
+      "grad_norm": 0.9711139798164368,
+      "learning_rate": 3.739203523865635e-05,
+      "loss": 0.2419,
+      "step": 2376
+    },
+    {
+      "epoch": 1.4463036203224826,
+      "grad_norm": 0.8849513530731201,
+      "learning_rate": 3.738122181842349e-05,
+      "loss": 0.2023,
+      "step": 2377
+    },
+    {
+      "epoch": 1.4469120778825677,
+      "grad_norm": 0.8123235702514648,
+      "learning_rate": 3.73704053281637e-05,
+      "loss": 0.2084,
+      "step": 2378
+    },
+    {
+      "epoch": 1.4475205354426528,
+      "grad_norm": 0.9523190259933472,
+      "learning_rate": 3.7359585770559024e-05,
+      "loss": 0.2318,
+      "step": 2379
+    },
+    {
+      "epoch": 1.4481289930027381,
+      "grad_norm": 0.7958305478096008,
+      "learning_rate": 3.7348763148292236e-05,
+      "loss": 0.2133,
+      "step": 2380
+    },
+    {
+      "epoch": 1.4487374505628232,
+      "grad_norm": 0.9427193403244019,
+      "learning_rate": 3.733793746404692e-05,
+      "loss": 0.1828,
+      "step": 2381
+    },
+    {
+      "epoch": 1.4493459081229085,
+      "grad_norm": 0.9071337580680847,
+      "learning_rate": 3.732710872050737e-05,
+      "loss": 0.1915,
+      "step": 2382
+    },
+    {
+      "epoch": 1.4499543656829936,
+      "grad_norm": 0.9505900740623474,
+      "learning_rate": 3.731627692035869e-05,
+      "loss": 0.2522,
+      "step": 2383
+    },
+    {
+      "epoch": 1.4505628232430787,
+      "grad_norm": 0.9712157845497131,
+      "learning_rate": 3.730544206628669e-05,
+      "loss": 0.2114,
+      "step": 2384
+    },
+    {
+      "epoch": 1.451171280803164,
+      "grad_norm": 0.8542957901954651,
+      "learning_rate": 3.729460416097797e-05,
+      "loss": 0.1831,
+      "step": 2385
+    },
+    {
+      "epoch": 1.4517797383632491,
+      "grad_norm": 0.9015089869499207,
+      "learning_rate": 3.7283763207119894e-05,
+      "loss": 0.1942,
+      "step": 2386
+    },
+    {
+      "epoch": 1.4523881959233345,
+      "grad_norm": 0.8781471848487854,
+      "learning_rate": 3.7272919207400556e-05,
+      "loss": 0.2114,
+      "step": 2387
+    },
+    {
+      "epoch": 1.4529966534834196,
+      "grad_norm": 0.9101613163948059,
+      "learning_rate": 3.72620721645088e-05,
+      "loss": 0.1945,
+      "step": 2388
+    },
+    {
+      "epoch": 1.4536051110435046,
+      "grad_norm": 0.9514543414115906,
+      "learning_rate": 3.725122208113427e-05,
+      "loss": 0.2435,
+      "step": 2389
+    },
+    {
+      "epoch": 1.45421356860359,
+      "grad_norm": 0.9269886612892151,
+      "learning_rate": 3.724036895996732e-05,
+      "loss": 0.2401,
+      "step": 2390
+    },
+    {
+      "epoch": 1.454822026163675,
+      "grad_norm": 0.9546543955802917,
+      "learning_rate": 3.722951280369906e-05,
+      "loss": 0.2285,
+      "step": 2391
+    },
+    {
+      "epoch": 1.4554304837237604,
+      "grad_norm": 0.8362632393836975,
+      "learning_rate": 3.7218653615021395e-05,
+      "loss": 0.2115,
+      "step": 2392
+    },
+    {
+      "epoch": 1.4560389412838455,
+      "grad_norm": 0.9519213438034058,
+      "learning_rate": 3.720779139662691e-05,
+      "loss": 0.2214,
+      "step": 2393
+    },
+    {
+      "epoch": 1.4566473988439306,
+      "grad_norm": 0.7873554825782776,
+      "learning_rate": 3.719692615120902e-05,
+      "loss": 0.2028,
+      "step": 2394
+    },
+    {
+      "epoch": 1.457255856404016,
+      "grad_norm": 0.8509390354156494,
+      "learning_rate": 3.718605788146183e-05,
+      "loss": 0.2084,
+      "step": 2395
+    },
+    {
+      "epoch": 1.457864313964101,
+      "grad_norm": 0.9013881087303162,
+      "learning_rate": 3.717518659008023e-05,
+      "loss": 0.2139,
+      "step": 2396
+    },
+    {
+      "epoch": 1.4584727715241863,
+      "grad_norm": 0.8233511447906494,
+      "learning_rate": 3.7164312279759836e-05,
+      "loss": 0.1946,
+      "step": 2397
+    },
+    {
+      "epoch": 1.4590812290842714,
+      "grad_norm": 0.8857746124267578,
+      "learning_rate": 3.7153434953197044e-05,
+      "loss": 0.1972,
+      "step": 2398
+    },
+    {
+      "epoch": 1.4596896866443565,
+      "grad_norm": 0.793010950088501,
+      "learning_rate": 3.714255461308895e-05,
+      "loss": 0.184,
+      "step": 2399
+    },
+    {
+      "epoch": 1.4602981442044418,
+      "grad_norm": 0.9649643301963806,
+      "learning_rate": 3.7131671262133444e-05,
+      "loss": 0.2153,
+      "step": 2400
+    },
+    {
+      "epoch": 1.460906601764527,
+      "grad_norm": 0.9244347810745239,
+      "learning_rate": 3.712078490302913e-05,
+      "loss": 0.1921,
+      "step": 2401
+    },
+    {
+      "epoch": 1.4615150593246122,
+      "grad_norm": 1.1217639446258545,
+      "learning_rate": 3.7109895538475394e-05,
+      "loss": 0.2125,
+      "step": 2402
+    },
+    {
+      "epoch": 1.4621235168846973,
+      "grad_norm": 0.8992077112197876,
+      "learning_rate": 3.709900317117232e-05,
+      "loss": 0.1979,
+      "step": 2403
+    },
+    {
+      "epoch": 1.4627319744447824,
+      "grad_norm": 0.9109466075897217,
+      "learning_rate": 3.708810780382077e-05,
+      "loss": 0.1959,
+      "step": 2404
+    },
+    {
+      "epoch": 1.4633404320048677,
+      "grad_norm": 0.8444501161575317,
+      "learning_rate": 3.707720943912235e-05,
+      "loss": 0.1709,
+      "step": 2405
+    },
+    {
+      "epoch": 1.4639488895649528,
+      "grad_norm": 0.9722615480422974,
+      "learning_rate": 3.706630807977938e-05,
+      "loss": 0.2304,
+      "step": 2406
+    },
+    {
+      "epoch": 1.4645573471250382,
+      "grad_norm": 0.9228814244270325,
+      "learning_rate": 3.705540372849496e-05,
+      "loss": 0.2209,
+      "step": 2407
+    },
+    {
+      "epoch": 1.4651658046851233,
+      "grad_norm": 0.9532406330108643,
+      "learning_rate": 3.7044496387972914e-05,
+      "loss": 0.2162,
+      "step": 2408
+    },
+    {
+      "epoch": 1.4657742622452083,
+      "grad_norm": 0.7384613156318665,
+      "learning_rate": 3.7033586060917795e-05,
+      "loss": 0.1676,
+      "step": 2409
+    },
+    {
+      "epoch": 1.4663827198052934,
+      "grad_norm": 0.804396390914917,
+      "learning_rate": 3.7022672750034926e-05,
+      "loss": 0.2018,
+      "step": 2410
+    },
+    {
+      "epoch": 1.4669911773653788,
+      "grad_norm": 0.8668960332870483,
+      "learning_rate": 3.701175645803034e-05,
+      "loss": 0.238,
+      "step": 2411
+    },
+    {
+      "epoch": 1.467599634925464,
+      "grad_norm": 0.8890330195426941,
+      "learning_rate": 3.7000837187610826e-05,
+      "loss": 0.1914,
+      "step": 2412
+    },
+    {
+      "epoch": 1.4682080924855492,
+      "grad_norm": 1.1121948957443237,
+      "learning_rate": 3.698991494148391e-05,
+      "loss": 0.2169,
+      "step": 2413
+    },
+    {
+      "epoch": 1.4688165500456343,
+      "grad_norm": 1.0296958684921265,
+      "learning_rate": 3.697898972235785e-05,
+      "loss": 0.2164,
+      "step": 2414
+    },
+    {
+      "epoch": 1.4694250076057194,
+      "grad_norm": 0.9091324210166931,
+      "learning_rate": 3.6968061532941654e-05,
+      "loss": 0.207,
+      "step": 2415
+    },
+    {
+      "epoch": 1.4700334651658047,
+      "grad_norm": 0.8409296274185181,
+      "learning_rate": 3.695713037594505e-05,
+      "loss": 0.2317,
+      "step": 2416
+    },
+    {
+      "epoch": 1.47064192272589,
+      "grad_norm": 0.8644183278083801,
+      "learning_rate": 3.6946196254078515e-05,
+      "loss": 0.2301,
+      "step": 2417
+    },
+    {
+      "epoch": 1.471250380285975,
+      "grad_norm": 0.9687103033065796,
+      "learning_rate": 3.693525917005324e-05,
+      "loss": 0.2214,
+      "step": 2418
+    },
+    {
+      "epoch": 1.4718588378460602,
+      "grad_norm": 0.8268604278564453,
+      "learning_rate": 3.692431912658118e-05,
+      "loss": 0.2614,
+      "step": 2419
+    },
+    {
+      "epoch": 1.4724672954061453,
+      "grad_norm": 0.8227600455284119,
+      "learning_rate": 3.691337612637501e-05,
+      "loss": 0.1931,
+      "step": 2420
+    },
+    {
+      "epoch": 1.4730757529662306,
+      "grad_norm": 0.8828481435775757,
+      "learning_rate": 3.690243017214813e-05,
+      "loss": 0.1842,
+      "step": 2421
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.7842615246772766,
+      "learning_rate": 3.689148126661469e-05,
+      "loss": 0.2161,
+      "step": 2422
+    },
+    {
+      "epoch": 1.474292668086401,
+      "grad_norm": 0.9619740843772888,
+      "learning_rate": 3.688052941248956e-05,
+      "loss": 0.2421,
+      "step": 2423
+    },
+    {
+      "epoch": 1.4749011256464861,
+      "grad_norm": 0.8373282551765442,
+      "learning_rate": 3.686957461248833e-05,
+      "loss": 0.2045,
+      "step": 2424
+    },
+    {
+      "epoch": 1.4755095832065712,
+      "grad_norm": 0.9472412467002869,
+      "learning_rate": 3.685861686932735e-05,
+      "loss": 0.2319,
+      "step": 2425
+    },
+    {
+      "epoch": 1.4761180407666565,
+      "grad_norm": 0.9255704283714294,
+      "learning_rate": 3.6847656185723686e-05,
+      "loss": 0.2266,
+      "step": 2426
+    },
+    {
+      "epoch": 1.4767264983267416,
+      "grad_norm": 0.851377010345459,
+      "learning_rate": 3.683669256439511e-05,
+      "loss": 0.2215,
+      "step": 2427
+    },
+    {
+      "epoch": 1.477334955886827,
+      "grad_norm": 0.9094513654708862,
+      "learning_rate": 3.6825726008060155e-05,
+      "loss": 0.2713,
+      "step": 2428
+    },
+    {
+      "epoch": 1.477943413446912,
+      "grad_norm": 0.9373210668563843,
+      "learning_rate": 3.681475651943809e-05,
+      "loss": 0.2021,
+      "step": 2429
+    },
+    {
+      "epoch": 1.4785518710069971,
+      "grad_norm": 0.8862281441688538,
+      "learning_rate": 3.680378410124885e-05,
+      "loss": 0.2172,
+      "step": 2430
+    },
+    {
+      "epoch": 1.4791603285670825,
+      "grad_norm": 0.9415962100028992,
+      "learning_rate": 3.6792808756213166e-05,
+      "loss": 0.2334,
+      "step": 2431
+    },
+    {
+      "epoch": 1.4797687861271676,
+      "grad_norm": 0.8250126838684082,
+      "learning_rate": 3.678183048705246e-05,
+      "loss": 0.2086,
+      "step": 2432
+    },
+    {
+      "epoch": 1.4803772436872529,
+      "grad_norm": 0.8701198697090149,
+      "learning_rate": 3.6770849296488885e-05,
+      "loss": 0.218,
+      "step": 2433
+    },
+    {
+      "epoch": 1.480985701247338,
+      "grad_norm": 0.7581288814544678,
+      "learning_rate": 3.675986518724532e-05,
+      "loss": 0.1843,
+      "step": 2434
+    },
+    {
+      "epoch": 1.481594158807423,
+      "grad_norm": 0.8139165043830872,
+      "learning_rate": 3.674887816204536e-05,
+      "loss": 0.1995,
+      "step": 2435
+    },
+    {
+      "epoch": 1.4822026163675084,
+      "grad_norm": 0.9862119555473328,
+      "learning_rate": 3.6737888223613323e-05,
+      "loss": 0.173,
+      "step": 2436
+    },
+    {
+      "epoch": 1.4828110739275935,
+      "grad_norm": 0.9770458936691284,
+      "learning_rate": 3.6726895374674286e-05,
+      "loss": 0.2326,
+      "step": 2437
+    },
+    {
+      "epoch": 1.4834195314876788,
+      "grad_norm": 0.9238155484199524,
+      "learning_rate": 3.671589961795399e-05,
+      "loss": 0.1974,
+      "step": 2438
+    },
+    {
+      "epoch": 1.484027989047764,
+      "grad_norm": 0.7940313816070557,
+      "learning_rate": 3.6704900956178924e-05,
+      "loss": 0.1888,
+      "step": 2439
+    },
+    {
+      "epoch": 1.484636446607849,
+      "grad_norm": 0.9048517942428589,
+      "learning_rate": 3.6693899392076306e-05,
+      "loss": 0.2122,
+      "step": 2440
+    },
+    {
+      "epoch": 1.4852449041679343,
+      "grad_norm": 0.8240786194801331,
+      "learning_rate": 3.6682894928374074e-05,
+      "loss": 0.169,
+      "step": 2441
+    },
+    {
+      "epoch": 1.4858533617280194,
+      "grad_norm": 0.9464930295944214,
+      "learning_rate": 3.6671887567800853e-05,
+      "loss": 0.1947,
+      "step": 2442
+    },
+    {
+      "epoch": 1.4864618192881047,
+      "grad_norm": 0.8999909162521362,
+      "learning_rate": 3.666087731308604e-05,
+      "loss": 0.2237,
+      "step": 2443
+    },
+    {
+      "epoch": 1.4870702768481898,
+      "grad_norm": 0.8082036972045898,
+      "learning_rate": 3.664986416695969e-05,
+      "loss": 0.1476,
+      "step": 2444
+    },
+    {
+      "epoch": 1.487678734408275,
+      "grad_norm": 0.8836786150932312,
+      "learning_rate": 3.663884813215263e-05,
+      "loss": 0.1995,
+      "step": 2445
+    },
+    {
+      "epoch": 1.4882871919683602,
+      "grad_norm": 0.8656254410743713,
+      "learning_rate": 3.6627829211396345e-05,
+      "loss": 0.2678,
+      "step": 2446
+    },
+    {
+      "epoch": 1.4888956495284453,
+      "grad_norm": 0.7731005549430847,
+      "learning_rate": 3.66168074074231e-05,
+      "loss": 0.1776,
+      "step": 2447
+    },
+    {
+      "epoch": 1.4895041070885306,
+      "grad_norm": 0.8301362991333008,
+      "learning_rate": 3.6605782722965834e-05,
+      "loss": 0.1878,
+      "step": 2448
+    },
+    {
+      "epoch": 1.4901125646486157,
+      "grad_norm": 0.8362447023391724,
+      "learning_rate": 3.65947551607582e-05,
+      "loss": 0.2027,
+      "step": 2449
+    },
+    {
+      "epoch": 1.4907210222087008,
+      "grad_norm": 0.9739522933959961,
+      "learning_rate": 3.6583724723534574e-05,
+      "loss": 0.2361,
+      "step": 2450
+    },
+    {
+      "epoch": 1.4913294797687862,
+      "grad_norm": 0.9524062275886536,
+      "learning_rate": 3.6572691414030055e-05,
+      "loss": 0.2252,
+      "step": 2451
+    },
+    {
+      "epoch": 1.4919379373288713,
+      "grad_norm": 1.0326991081237793,
+      "learning_rate": 3.656165523498044e-05,
+      "loss": 0.2235,
+      "step": 2452
+    },
+    {
+      "epoch": 1.4925463948889566,
+      "grad_norm": 0.9849575161933899,
+      "learning_rate": 3.655061618912224e-05,
+      "loss": 0.2279,
+      "step": 2453
+    },
+    {
+      "epoch": 1.4931548524490417,
+      "grad_norm": 0.8630450367927551,
+      "learning_rate": 3.653957427919268e-05,
+      "loss": 0.1915,
+      "step": 2454
+    },
+    {
+      "epoch": 1.4937633100091268,
+      "grad_norm": 0.845496654510498,
+      "learning_rate": 3.652852950792969e-05,
+      "loss": 0.1721,
+      "step": 2455
+    },
+    {
+      "epoch": 1.494371767569212,
+      "grad_norm": 1.2583898305892944,
+      "learning_rate": 3.651748187807191e-05,
+      "loss": 0.2039,
+      "step": 2456
+    },
+    {
+      "epoch": 1.4949802251292972,
+      "grad_norm": 0.7946431040763855,
+      "learning_rate": 3.650643139235871e-05,
+      "loss": 0.1687,
+      "step": 2457
+    },
+    {
+      "epoch": 1.4955886826893825,
+      "grad_norm": 1.014686107635498,
+      "learning_rate": 3.649537805353013e-05,
+      "loss": 0.2319,
+      "step": 2458
+    },
+    {
+      "epoch": 1.4961971402494676,
+      "grad_norm": 1.0813982486724854,
+      "learning_rate": 3.648432186432694e-05,
+      "loss": 0.2508,
+      "step": 2459
+    },
+    {
+      "epoch": 1.4968055978095527,
+      "grad_norm": 0.8511247038841248,
+      "learning_rate": 3.6473262827490614e-05,
+      "loss": 0.1737,
+      "step": 2460
+    },
+    {
+      "epoch": 1.497414055369638,
+      "grad_norm": 0.963362455368042,
+      "learning_rate": 3.646220094576334e-05,
+      "loss": 0.2431,
+      "step": 2461
+    },
+    {
+      "epoch": 1.498022512929723,
+      "grad_norm": 0.8051499724388123,
+      "learning_rate": 3.6451136221888005e-05,
+      "loss": 0.1954,
+      "step": 2462
+    },
+    {
+      "epoch": 1.4986309704898084,
+      "grad_norm": 0.8649519085884094,
+      "learning_rate": 3.644006865860819e-05,
+      "loss": 0.1716,
+      "step": 2463
+    },
+    {
+      "epoch": 1.4992394280498935,
+      "grad_norm": 0.9137884378433228,
+      "learning_rate": 3.642899825866819e-05,
+      "loss": 0.1964,
+      "step": 2464
+    },
+    {
+      "epoch": 1.4998478856099786,
+      "grad_norm": 0.7866896986961365,
+      "learning_rate": 3.641792502481301e-05,
+      "loss": 0.1809,
+      "step": 2465
+    },
+    {
+      "epoch": 1.500456343170064,
+      "grad_norm": 0.9844456911087036,
+      "learning_rate": 3.640684895978834e-05,
+      "loss": 0.2182,
+      "step": 2466
+    },
+    {
+      "epoch": 1.501064800730149,
+      "grad_norm": 0.9635726809501648,
+      "learning_rate": 3.6395770066340596e-05,
+      "loss": 0.2182,
+      "step": 2467
+    },
+    {
+      "epoch": 1.5016732582902343,
+      "grad_norm": 0.9193302392959595,
+      "learning_rate": 3.6384688347216875e-05,
+      "loss": 0.2455,
+      "step": 2468
+    },
+    {
+      "epoch": 1.5022817158503194,
+      "grad_norm": 1.0144778490066528,
+      "learning_rate": 3.637360380516498e-05,
+      "loss": 0.2717,
+      "step": 2469
+    },
+    {
+      "epoch": 1.5028901734104045,
+      "grad_norm": 0.8575990796089172,
+      "learning_rate": 3.6362516442933416e-05,
+      "loss": 0.2333,
+      "step": 2470
+    },
+    {
+      "epoch": 1.5034986309704899,
+      "grad_norm": 1.1676721572875977,
+      "learning_rate": 3.635142626327139e-05,
+      "loss": 0.2948,
+      "step": 2471
+    },
+    {
+      "epoch": 1.504107088530575,
+      "grad_norm": 0.7690470814704895,
+      "learning_rate": 3.634033326892879e-05,
+      "loss": 0.2268,
+      "step": 2472
+    },
+    {
+      "epoch": 1.5047155460906603,
+      "grad_norm": 0.8740969300270081,
+      "learning_rate": 3.632923746265623e-05,
+      "loss": 0.227,
+      "step": 2473
+    },
+    {
+      "epoch": 1.5053240036507454,
+      "grad_norm": 0.8315359354019165,
+      "learning_rate": 3.631813884720502e-05,
+      "loss": 0.2036,
+      "step": 2474
+    },
+    {
+      "epoch": 1.5059324612108305,
+      "grad_norm": 0.8344742059707642,
+      "learning_rate": 3.630703742532713e-05,
+      "loss": 0.2397,
+      "step": 2475
+    },
+    {
+      "epoch": 1.5065409187709158,
+      "grad_norm": 1.0899635553359985,
+      "learning_rate": 3.6295933199775265e-05,
+      "loss": 0.1978,
+      "step": 2476
+    },
+    {
+      "epoch": 1.5071493763310009,
+      "grad_norm": 0.9321961998939514,
+      "learning_rate": 3.628482617330279e-05,
+      "loss": 0.2296,
+      "step": 2477
+    },
+    {
+      "epoch": 1.5077578338910862,
+      "grad_norm": 0.8498799204826355,
+      "learning_rate": 3.627371634866381e-05,
+      "loss": 0.1986,
+      "step": 2478
+    },
+    {
+      "epoch": 1.5083662914511713,
+      "grad_norm": 0.8041394948959351,
+      "learning_rate": 3.626260372861308e-05,
+      "loss": 0.1844,
+      "step": 2479
+    },
+    {
+      "epoch": 1.5089747490112564,
+      "grad_norm": 0.8486339449882507,
+      "learning_rate": 3.625148831590608e-05,
+      "loss": 0.2032,
+      "step": 2480
+    },
+    {
+      "epoch": 1.5095832065713415,
+      "grad_norm": 0.817248523235321,
+      "learning_rate": 3.624037011329896e-05,
+      "loss": 0.1985,
+      "step": 2481
+    },
+    {
+      "epoch": 1.5101916641314268,
+      "grad_norm": 0.8181208372116089,
+      "learning_rate": 3.6229249123548574e-05,
+      "loss": 0.1856,
+      "step": 2482
+    },
+    {
+      "epoch": 1.5108001216915121,
+      "grad_norm": 0.7917095422744751,
+      "learning_rate": 3.621812534941246e-05,
+      "loss": 0.1694,
+      "step": 2483
+    },
+    {
+      "epoch": 1.5114085792515972,
+      "grad_norm": 0.8646035194396973,
+      "learning_rate": 3.620699879364886e-05,
+      "loss": 0.2158,
+      "step": 2484
+    },
+    {
+      "epoch": 1.5120170368116823,
+      "grad_norm": 0.9511632323265076,
+      "learning_rate": 3.619586945901669e-05,
+      "loss": 0.1938,
+      "step": 2485
+    },
+    {
+      "epoch": 1.5126254943717674,
+      "grad_norm": 0.7660108208656311,
+      "learning_rate": 3.618473734827556e-05,
+      "loss": 0.1687,
+      "step": 2486
+    },
+    {
+      "epoch": 1.5132339519318527,
+      "grad_norm": 0.772885799407959,
+      "learning_rate": 3.6173602464185765e-05,
+      "loss": 0.2055,
+      "step": 2487
+    },
+    {
+      "epoch": 1.513842409491938,
+      "grad_norm": 0.8760683536529541,
+      "learning_rate": 3.61624648095083e-05,
+      "loss": 0.1855,
+      "step": 2488
+    },
+    {
+      "epoch": 1.5144508670520231,
+      "grad_norm": 1.2097331285476685,
+      "learning_rate": 3.615132438700484e-05,
+      "loss": 0.2581,
+      "step": 2489
+    },
+    {
+      "epoch": 1.5150593246121082,
+      "grad_norm": 0.7943812012672424,
+      "learning_rate": 3.614018119943774e-05,
+      "loss": 0.2076,
+      "step": 2490
+    },
+    {
+      "epoch": 1.5156677821721933,
+      "grad_norm": 0.8257327079772949,
+      "learning_rate": 3.6129035249570053e-05,
+      "loss": 0.202,
+      "step": 2491
+    },
+    {
+      "epoch": 1.5162762397322787,
+      "grad_norm": 0.768048107624054,
+      "learning_rate": 3.6117886540165504e-05,
+      "loss": 0.1859,
+      "step": 2492
+    },
+    {
+      "epoch": 1.516884697292364,
+      "grad_norm": 0.8744530081748962,
+      "learning_rate": 3.6106735073988504e-05,
+      "loss": 0.1751,
+      "step": 2493
+    },
+    {
+      "epoch": 1.517493154852449,
+      "grad_norm": 0.9588749408721924,
+      "learning_rate": 3.6095580853804155e-05,
+      "loss": 0.2156,
+      "step": 2494
+    },
+    {
+      "epoch": 1.5181016124125342,
+      "grad_norm": 0.8880223035812378,
+      "learning_rate": 3.608442388237825e-05,
+      "loss": 0.2116,
+      "step": 2495
+    },
+    {
+      "epoch": 1.5187100699726193,
+      "grad_norm": 0.9411167502403259,
+      "learning_rate": 3.607326416247723e-05,
+      "loss": 0.217,
+      "step": 2496
+    },
+    {
+      "epoch": 1.5193185275327046,
+      "grad_norm": 0.9615315198898315,
+      "learning_rate": 3.606210169686827e-05,
+      "loss": 0.2129,
+      "step": 2497
+    },
+    {
+      "epoch": 1.51992698509279,
+      "grad_norm": 0.9808461666107178,
+      "learning_rate": 3.605093648831917e-05,
+      "loss": 0.2242,
+      "step": 2498
+    },
+    {
+      "epoch": 1.520535442652875,
+      "grad_norm": 1.2025115489959717,
+      "learning_rate": 3.603976853959845e-05,
+      "loss": 0.2028,
+      "step": 2499
+    },
+    {
+      "epoch": 1.52114390021296,
+      "grad_norm": 0.8565399646759033,
+      "learning_rate": 3.602859785347529e-05,
+      "loss": 0.228,
+      "step": 2500
+    },
+    {
+      "epoch": 1.5217523577730452,
+      "grad_norm": 0.769661545753479,
+      "learning_rate": 3.601742443271956e-05,
+      "loss": 0.1927,
+      "step": 2501
+    },
+    {
+      "epoch": 1.5223608153331305,
+      "grad_norm": 0.8793697953224182,
+      "learning_rate": 3.600624828010181e-05,
+      "loss": 0.199,
+      "step": 2502
+    },
+    {
+      "epoch": 1.5229692728932158,
+      "grad_norm": 0.9120194315910339,
+      "learning_rate": 3.599506939839323e-05,
+      "loss": 0.2185,
+      "step": 2503
+    },
+    {
+      "epoch": 1.523577730453301,
+      "grad_norm": 0.9638735055923462,
+      "learning_rate": 3.598388779036575e-05,
+      "loss": 0.2264,
+      "step": 2504
+    },
+    {
+      "epoch": 1.524186188013386,
+      "grad_norm": 0.9277864694595337,
+      "learning_rate": 3.597270345879192e-05,
+      "loss": 0.2542,
+      "step": 2505
+    },
+    {
+      "epoch": 1.524794645573471,
+      "grad_norm": 0.8778446316719055,
+      "learning_rate": 3.5961516406445e-05,
+      "loss": 0.2076,
+      "step": 2506
+    },
+    {
+      "epoch": 1.5254031031335564,
+      "grad_norm": 0.8511151671409607,
+      "learning_rate": 3.595032663609891e-05,
+      "loss": 0.1923,
+      "step": 2507
+    },
+    {
+      "epoch": 1.5260115606936417,
+      "grad_norm": 0.7811472415924072,
+      "learning_rate": 3.593913415052825e-05,
+      "loss": 0.1589,
+      "step": 2508
+    },
+    {
+      "epoch": 1.5266200182537268,
+      "grad_norm": 0.8555339574813843,
+      "learning_rate": 3.5927938952508284e-05,
+      "loss": 0.1826,
+      "step": 2509
+    },
+    {
+      "epoch": 1.527228475813812,
+      "grad_norm": 0.8637688159942627,
+      "learning_rate": 3.591674104481495e-05,
+      "loss": 0.193,
+      "step": 2510
+    },
+    {
+      "epoch": 1.527836933373897,
+      "grad_norm": 0.8080950379371643,
+      "learning_rate": 3.590554043022488e-05,
+      "loss": 0.1814,
+      "step": 2511
+    },
+    {
+      "epoch": 1.5284453909339824,
+      "grad_norm": 0.9069305658340454,
+      "learning_rate": 3.5894337111515344e-05,
+      "loss": 0.1547,
+      "step": 2512
+    },
+    {
+      "epoch": 1.5290538484940677,
+      "grad_norm": 0.9532658457756042,
+      "learning_rate": 3.58831310914643e-05,
+      "loss": 0.1953,
+      "step": 2513
+    },
+    {
+      "epoch": 1.5296623060541528,
+      "grad_norm": 1.2767683267593384,
+      "learning_rate": 3.5871922372850376e-05,
+      "loss": 0.2019,
+      "step": 2514
+    },
+    {
+      "epoch": 1.5302707636142379,
+      "grad_norm": 0.8507498502731323,
+      "learning_rate": 3.586071095845287e-05,
+      "loss": 0.2115,
+      "step": 2515
+    },
+    {
+      "epoch": 1.530879221174323,
+      "grad_norm": 1.0665339231491089,
+      "learning_rate": 3.5849496851051744e-05,
+      "loss": 0.2373,
+      "step": 2516
+    },
+    {
+      "epoch": 1.5314876787344083,
+      "grad_norm": 0.9984211325645447,
+      "learning_rate": 3.583828005342763e-05,
+      "loss": 0.245,
+      "step": 2517
+    },
+    {
+      "epoch": 1.5320961362944936,
+      "grad_norm": 0.850425124168396,
+      "learning_rate": 3.5827060568361817e-05,
+      "loss": 0.1987,
+      "step": 2518
+    },
+    {
+      "epoch": 1.5327045938545787,
+      "grad_norm": 1.2310839891433716,
+      "learning_rate": 3.5815838398636284e-05,
+      "loss": 0.2151,
+      "step": 2519
+    },
+    {
+      "epoch": 1.5333130514146638,
+      "grad_norm": 0.8494765162467957,
+      "learning_rate": 3.580461354703365e-05,
+      "loss": 0.245,
+      "step": 2520
+    },
+    {
+      "epoch": 1.5339215089747489,
+      "grad_norm": 0.929420530796051,
+      "learning_rate": 3.579338601633722e-05,
+      "loss": 0.24,
+      "step": 2521
+    },
+    {
+      "epoch": 1.5345299665348342,
+      "grad_norm": 0.7413504719734192,
+      "learning_rate": 3.578215580933095e-05,
+      "loss": 0.1886,
+      "step": 2522
+    },
+    {
+      "epoch": 1.5351384240949195,
+      "grad_norm": 0.9758164286613464,
+      "learning_rate": 3.577092292879946e-05,
+      "loss": 0.2134,
+      "step": 2523
+    },
+    {
+      "epoch": 1.5357468816550046,
+      "grad_norm": 0.8202041983604431,
+      "learning_rate": 3.5759687377528026e-05,
+      "loss": 0.1697,
+      "step": 2524
+    },
+    {
+      "epoch": 1.5363553392150897,
+      "grad_norm": 0.816893994808197,
+      "learning_rate": 3.574844915830263e-05,
+      "loss": 0.1809,
+      "step": 2525
+    },
+    {
+      "epoch": 1.5369637967751748,
+      "grad_norm": 0.8673279285430908,
+      "learning_rate": 3.573720827390984e-05,
+      "loss": 0.2501,
+      "step": 2526
+    },
+    {
+      "epoch": 1.5375722543352601,
+      "grad_norm": 0.8785646557807922,
+      "learning_rate": 3.572596472713696e-05,
+      "loss": 0.1991,
+      "step": 2527
+    },
+    {
+      "epoch": 1.5381807118953454,
+      "grad_norm": 0.8835013508796692,
+      "learning_rate": 3.5714718520771904e-05,
+      "loss": 0.1982,
+      "step": 2528
+    },
+    {
+      "epoch": 1.5387891694554305,
+      "grad_norm": 0.8082259297370911,
+      "learning_rate": 3.570346965760326e-05,
+      "loss": 0.174,
+      "step": 2529
+    },
+    {
+      "epoch": 1.5393976270155156,
+      "grad_norm": 0.869325578212738,
+      "learning_rate": 3.5692218140420295e-05,
+      "loss": 0.1844,
+      "step": 2530
+    },
+    {
+      "epoch": 1.5400060845756007,
+      "grad_norm": 0.9018534421920776,
+      "learning_rate": 3.5680963972012894e-05,
+      "loss": 0.1921,
+      "step": 2531
+    },
+    {
+      "epoch": 1.540614542135686,
+      "grad_norm": 0.8284593820571899,
+      "learning_rate": 3.566970715517164e-05,
+      "loss": 0.1818,
+      "step": 2532
+    },
+    {
+      "epoch": 1.5412229996957714,
+      "grad_norm": 0.7905526161193848,
+      "learning_rate": 3.565844769268774e-05,
+      "loss": 0.1917,
+      "step": 2533
+    },
+    {
+      "epoch": 1.5418314572558565,
+      "grad_norm": 0.7893586158752441,
+      "learning_rate": 3.564718558735308e-05,
+      "loss": 0.177,
+      "step": 2534
+    },
+    {
+      "epoch": 1.5424399148159416,
+      "grad_norm": 0.7778828144073486,
+      "learning_rate": 3.56359208419602e-05,
+      "loss": 0.2133,
+      "step": 2535
+    },
+    {
+      "epoch": 1.5430483723760267,
+      "grad_norm": 0.8154110312461853,
+      "learning_rate": 3.562465345930227e-05,
+      "loss": 0.2001,
+      "step": 2536
+    },
+    {
+      "epoch": 1.543656829936112,
+      "grad_norm": 0.8788914084434509,
+      "learning_rate": 3.561338344217314e-05,
+      "loss": 0.1756,
+      "step": 2537
+    },
+    {
+      "epoch": 1.5442652874961973,
+      "grad_norm": 0.7563880085945129,
+      "learning_rate": 3.560211079336731e-05,
+      "loss": 0.1944,
+      "step": 2538
+    },
+    {
+      "epoch": 1.5448737450562824,
+      "grad_norm": 0.7627280354499817,
+      "learning_rate": 3.559083551567991e-05,
+      "loss": 0.1919,
+      "step": 2539
+    },
+    {
+      "epoch": 1.5454822026163675,
+      "grad_norm": 1.1467633247375488,
+      "learning_rate": 3.5579557611906755e-05,
+      "loss": 0.1835,
+      "step": 2540
+    },
+    {
+      "epoch": 1.5460906601764526,
+      "grad_norm": 0.9292544722557068,
+      "learning_rate": 3.5568277084844295e-05,
+      "loss": 0.2053,
+      "step": 2541
+    },
+    {
+      "epoch": 1.546699117736538,
+      "grad_norm": 0.7936065196990967,
+      "learning_rate": 3.555699393728962e-05,
+      "loss": 0.1753,
+      "step": 2542
+    },
+    {
+      "epoch": 1.5473075752966232,
+      "grad_norm": 0.8932594060897827,
+      "learning_rate": 3.554570817204048e-05,
+      "loss": 0.1695,
+      "step": 2543
+    },
+    {
+      "epoch": 1.5479160328567083,
+      "grad_norm": 1.080686330795288,
+      "learning_rate": 3.553441979189529e-05,
+      "loss": 0.259,
+      "step": 2544
+    },
+    {
+      "epoch": 1.5485244904167934,
+      "grad_norm": 0.7938811182975769,
+      "learning_rate": 3.5523128799653084e-05,
+      "loss": 0.1673,
+      "step": 2545
+    },
+    {
+      "epoch": 1.5491329479768785,
+      "grad_norm": 0.8280820846557617,
+      "learning_rate": 3.551183519811356e-05,
+      "loss": 0.1884,
+      "step": 2546
+    },
+    {
+      "epoch": 1.5497414055369638,
+      "grad_norm": 0.8619610071182251,
+      "learning_rate": 3.550053899007707e-05,
+      "loss": 0.1851,
+      "step": 2547
+    },
+    {
+      "epoch": 1.5503498630970491,
+      "grad_norm": 0.824386477470398,
+      "learning_rate": 3.548924017834458e-05,
+      "loss": 0.1824,
+      "step": 2548
+    },
+    {
+      "epoch": 1.5509583206571342,
+      "grad_norm": 0.8854561448097229,
+      "learning_rate": 3.547793876571775e-05,
+      "loss": 0.1681,
+      "step": 2549
+    },
+    {
+      "epoch": 1.5515667782172193,
+      "grad_norm": 0.7834053635597229,
+      "learning_rate": 3.546663475499884e-05,
+      "loss": 0.1574,
+      "step": 2550
+    },
+    {
+      "epoch": 1.5521752357773044,
+      "grad_norm": 1.0175268650054932,
+      "learning_rate": 3.5455328148990794e-05,
+      "loss": 0.2224,
+      "step": 2551
+    },
+    {
+      "epoch": 1.5527836933373897,
+      "grad_norm": 0.7926445007324219,
+      "learning_rate": 3.544401895049716e-05,
+      "loss": 0.1664,
+      "step": 2552
+    },
+    {
+      "epoch": 1.553392150897475,
+      "grad_norm": 0.8977791666984558,
+      "learning_rate": 3.543270716232215e-05,
+      "loss": 0.2121,
+      "step": 2553
+    },
+    {
+      "epoch": 1.5540006084575602,
+      "grad_norm": 0.7897874116897583,
+      "learning_rate": 3.542139278727062e-05,
+      "loss": 0.2033,
+      "step": 2554
+    },
+    {
+      "epoch": 1.5546090660176453,
+      "grad_norm": 0.8928864598274231,
+      "learning_rate": 3.541007582814807e-05,
+      "loss": 0.1999,
+      "step": 2555
+    },
+    {
+      "epoch": 1.5552175235777304,
+      "grad_norm": 0.8382681608200073,
+      "learning_rate": 3.539875628776062e-05,
+      "loss": 0.1996,
+      "step": 2556
+    },
+    {
+      "epoch": 1.5558259811378157,
+      "grad_norm": 0.7771824598312378,
+      "learning_rate": 3.5387434168915065e-05,
+      "loss": 0.1788,
+      "step": 2557
+    },
+    {
+      "epoch": 1.5564344386979008,
+      "grad_norm": 0.9810817241668701,
+      "learning_rate": 3.53761094744188e-05,
+      "loss": 0.1917,
+      "step": 2558
+    },
+    {
+      "epoch": 1.557042896257986,
+      "grad_norm": 0.8481950163841248,
+      "learning_rate": 3.5364782207079886e-05,
+      "loss": 0.1748,
+      "step": 2559
+    },
+    {
+      "epoch": 1.5576513538180712,
+      "grad_norm": 0.8863506317138672,
+      "learning_rate": 3.5353452369707e-05,
+      "loss": 0.1898,
+      "step": 2560
+    },
+    {
+      "epoch": 1.5582598113781563,
+      "grad_norm": 0.9136548638343811,
+      "learning_rate": 3.534211996510949e-05,
+      "loss": 0.2652,
+      "step": 2561
+    },
+    {
+      "epoch": 1.5588682689382416,
+      "grad_norm": 0.814624011516571,
+      "learning_rate": 3.53307849960973e-05,
+      "loss": 0.1837,
+      "step": 2562
+    },
+    {
+      "epoch": 1.5594767264983267,
+      "grad_norm": 0.9348104000091553,
+      "learning_rate": 3.531944746548105e-05,
+      "loss": 0.1614,
+      "step": 2563
+    },
+    {
+      "epoch": 1.560085184058412,
+      "grad_norm": 0.817783534526825,
+      "learning_rate": 3.530810737607195e-05,
+      "loss": 0.1853,
+      "step": 2564
+    },
+    {
+      "epoch": 1.560693641618497,
+      "grad_norm": 0.8149775266647339,
+      "learning_rate": 3.529676473068189e-05,
+      "loss": 0.1745,
+      "step": 2565
+    },
+    {
+      "epoch": 1.5613020991785822,
+      "grad_norm": 0.8247946500778198,
+      "learning_rate": 3.5285419532123375e-05,
+      "loss": 0.1779,
+      "step": 2566
+    },
+    {
+      "epoch": 1.5619105567386675,
+      "grad_norm": 0.8637406826019287,
+      "learning_rate": 3.5274071783209525e-05,
+      "loss": 0.1981,
+      "step": 2567
+    },
+    {
+      "epoch": 1.5625190142987526,
+      "grad_norm": 0.7747231721878052,
+      "learning_rate": 3.5262721486754125e-05,
+      "loss": 0.1827,
+      "step": 2568
+    },
+    {
+      "epoch": 1.563127471858838,
+      "grad_norm": 0.9045551419258118,
+      "learning_rate": 3.525136864557156e-05,
+      "loss": 0.2214,
+      "step": 2569
+    },
+    {
+      "epoch": 1.563735929418923,
+      "grad_norm": 0.9226981997489929,
+      "learning_rate": 3.5240013262476866e-05,
+      "loss": 0.2045,
+      "step": 2570
+    },
+    {
+      "epoch": 1.5643443869790081,
+      "grad_norm": 0.793237566947937,
+      "learning_rate": 3.522865534028572e-05,
+      "loss": 0.1727,
+      "step": 2571
+    },
+    {
+      "epoch": 1.5649528445390934,
+      "grad_norm": 0.9380844235420227,
+      "learning_rate": 3.5217294881814386e-05,
+      "loss": 0.2171,
+      "step": 2572
+    },
+    {
+      "epoch": 1.5655613020991785,
+      "grad_norm": 0.964013397693634,
+      "learning_rate": 3.520593188987982e-05,
+      "loss": 0.2032,
+      "step": 2573
+    },
+    {
+      "epoch": 1.5661697596592639,
+      "grad_norm": 0.9411283731460571,
+      "learning_rate": 3.519456636729953e-05,
+      "loss": 0.1901,
+      "step": 2574
+    },
+    {
+      "epoch": 1.566778217219349,
+      "grad_norm": 0.8983286023139954,
+      "learning_rate": 3.518319831689172e-05,
+      "loss": 0.1837,
+      "step": 2575
+    },
+    {
+      "epoch": 1.567386674779434,
+      "grad_norm": 0.8221237659454346,
+      "learning_rate": 3.517182774147518e-05,
+      "loss": 0.1748,
+      "step": 2576
+    },
+    {
+      "epoch": 1.5679951323395194,
+      "grad_norm": 0.9628398418426514,
+      "learning_rate": 3.516045464386935e-05,
+      "loss": 0.1888,
+      "step": 2577
+    },
+    {
+      "epoch": 1.5686035898996045,
+      "grad_norm": 0.8376243710517883,
+      "learning_rate": 3.5149079026894266e-05,
+      "loss": 0.2096,
+      "step": 2578
+    },
+    {
+      "epoch": 1.5692120474596898,
+      "grad_norm": 0.8941128253936768,
+      "learning_rate": 3.513770089337063e-05,
+      "loss": 0.2027,
+      "step": 2579
+    },
+    {
+      "epoch": 1.5698205050197749,
+      "grad_norm": 0.9198672771453857,
+      "learning_rate": 3.512632024611972e-05,
+      "loss": 0.218,
+      "step": 2580
+    },
+    {
+      "epoch": 1.57042896257986,
+      "grad_norm": 0.9642953872680664,
+      "learning_rate": 3.511493708796348e-05,
+      "loss": 0.2402,
+      "step": 2581
+    },
+    {
+      "epoch": 1.571037420139945,
+      "grad_norm": 0.8804176449775696,
+      "learning_rate": 3.5103551421724457e-05,
+      "loss": 0.2181,
+      "step": 2582
+    },
+    {
+      "epoch": 1.5716458777000304,
+      "grad_norm": 0.7416863441467285,
+      "learning_rate": 3.509216325022582e-05,
+      "loss": 0.192,
+      "step": 2583
+    },
+    {
+      "epoch": 1.5722543352601157,
+      "grad_norm": 0.8433103561401367,
+      "learning_rate": 3.5080772576291356e-05,
+      "loss": 0.2138,
+      "step": 2584
+    },
+    {
+      "epoch": 1.5728627928202008,
+      "grad_norm": 0.8431623578071594,
+      "learning_rate": 3.506937940274547e-05,
+      "loss": 0.1981,
+      "step": 2585
+    },
+    {
+      "epoch": 1.573471250380286,
+      "grad_norm": 0.8424465656280518,
+      "learning_rate": 3.5057983732413224e-05,
+      "loss": 0.1745,
+      "step": 2586
+    },
+    {
+      "epoch": 1.574079707940371,
+      "grad_norm": 0.8414326310157776,
+      "learning_rate": 3.504658556812024e-05,
+      "loss": 0.1687,
+      "step": 2587
+    },
+    {
+      "epoch": 1.5746881655004563,
+      "grad_norm": 0.8884822130203247,
+      "learning_rate": 3.503518491269279e-05,
+      "loss": 0.1771,
+      "step": 2588
+    },
+    {
+      "epoch": 1.5752966230605416,
+      "grad_norm": 0.7992542386054993,
+      "learning_rate": 3.502378176895778e-05,
+      "loss": 0.1782,
+      "step": 2589
+    },
+    {
+      "epoch": 1.5759050806206267,
+      "grad_norm": 0.9021846055984497,
+      "learning_rate": 3.50123761397427e-05,
+      "loss": 0.2075,
+      "step": 2590
+    },
+    {
+      "epoch": 1.5765135381807118,
+      "grad_norm": 1.1060278415679932,
+      "learning_rate": 3.500096802787567e-05,
+      "loss": 0.2263,
+      "step": 2591
+    },
+    {
+      "epoch": 1.577121995740797,
+      "grad_norm": 1.0591530799865723,
+      "learning_rate": 3.4989557436185434e-05,
+      "loss": 0.1848,
+      "step": 2592
+    },
+    {
+      "epoch": 1.5777304533008822,
+      "grad_norm": 0.945749044418335,
+      "learning_rate": 3.4978144367501335e-05,
+      "loss": 0.2026,
+      "step": 2593
+    },
+    {
+      "epoch": 1.5783389108609676,
+      "grad_norm": 0.8668262958526611,
+      "learning_rate": 3.496672882465335e-05,
+      "loss": 0.1701,
+      "step": 2594
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.8319501280784607,
+      "learning_rate": 3.495531081047204e-05,
+      "loss": 0.1532,
+      "step": 2595
+    },
+    {
+      "epoch": 1.5795558259811378,
+      "grad_norm": 0.9786222577095032,
+      "learning_rate": 3.494389032778862e-05,
+      "loss": 0.2242,
+      "step": 2596
+    },
+    {
+      "epoch": 1.5801642835412228,
+      "grad_norm": 0.8528378009796143,
+      "learning_rate": 3.493246737943487e-05,
+      "loss": 0.2037,
+      "step": 2597
+    },
+    {
+      "epoch": 1.5807727411013082,
+      "grad_norm": 1.3592191934585571,
+      "learning_rate": 3.492104196824322e-05,
+      "loss": 0.2073,
+      "step": 2598
+    },
+    {
+      "epoch": 1.5813811986613935,
+      "grad_norm": 0.7618198394775391,
+      "learning_rate": 3.4909614097046686e-05,
+      "loss": 0.1418,
+      "step": 2599
+    },
+    {
+      "epoch": 1.5819896562214786,
+      "grad_norm": 0.8264166712760925,
+      "learning_rate": 3.489818376867891e-05,
+      "loss": 0.1669,
+      "step": 2600
+    },
+    {
+      "epoch": 1.5825981137815637,
+      "grad_norm": 0.8527823090553284,
+      "learning_rate": 3.4886750985974136e-05,
+      "loss": 0.1732,
+      "step": 2601
+    },
+    {
+      "epoch": 1.5832065713416488,
+      "grad_norm": 0.8907186985015869,
+      "learning_rate": 3.487531575176722e-05,
+      "loss": 0.2095,
+      "step": 2602
+    },
+    {
+      "epoch": 1.583815028901734,
+      "grad_norm": 0.8088558912277222,
+      "learning_rate": 3.4863878068893625e-05,
+      "loss": 0.1638,
+      "step": 2603
+    },
+    {
+      "epoch": 1.5844234864618194,
+      "grad_norm": 1.0112056732177734,
+      "learning_rate": 3.4852437940189414e-05,
+      "loss": 0.1842,
+      "step": 2604
+    },
+    {
+      "epoch": 1.5850319440219045,
+      "grad_norm": 0.6868709325790405,
+      "learning_rate": 3.4840995368491255e-05,
+      "loss": 0.1572,
+      "step": 2605
+    },
+    {
+      "epoch": 1.5856404015819896,
+      "grad_norm": 0.8623439073562622,
+      "learning_rate": 3.4829550356636445e-05,
+      "loss": 0.1846,
+      "step": 2606
+    },
+    {
+      "epoch": 1.5862488591420747,
+      "grad_norm": 0.8471298813819885,
+      "learning_rate": 3.481810290746287e-05,
+      "loss": 0.1654,
+      "step": 2607
+    },
+    {
+      "epoch": 1.58685731670216,
+      "grad_norm": 0.8028208017349243,
+      "learning_rate": 3.4806653023808996e-05,
+      "loss": 0.1955,
+      "step": 2608
+    },
+    {
+      "epoch": 1.5874657742622453,
+      "grad_norm": 0.8367049694061279,
+      "learning_rate": 3.4795200708513945e-05,
+      "loss": 0.1996,
+      "step": 2609
+    },
+    {
+      "epoch": 1.5880742318223304,
+      "grad_norm": 0.8771877884864807,
+      "learning_rate": 3.47837459644174e-05,
+      "loss": 0.2201,
+      "step": 2610
+    },
+    {
+      "epoch": 1.5886826893824155,
+      "grad_norm": 0.876565158367157,
+      "learning_rate": 3.477228879435966e-05,
+      "loss": 0.1817,
+      "step": 2611
+    },
+    {
+      "epoch": 1.5892911469425006,
+      "grad_norm": 0.9059187173843384,
+      "learning_rate": 3.4760829201181635e-05,
+      "loss": 0.2115,
+      "step": 2612
+    },
+    {
+      "epoch": 1.589899604502586,
+      "grad_norm": 0.9595020413398743,
+      "learning_rate": 3.474936718772481e-05,
+      "loss": 0.212,
+      "step": 2613
+    },
+    {
+      "epoch": 1.5905080620626713,
+      "grad_norm": 0.8010035753250122,
+      "learning_rate": 3.47379027568313e-05,
+      "loss": 0.1616,
+      "step": 2614
+    },
+    {
+      "epoch": 1.5911165196227564,
+      "grad_norm": 0.9189199805259705,
+      "learning_rate": 3.4726435911343804e-05,
+      "loss": 0.1859,
+      "step": 2615
+    },
+    {
+      "epoch": 1.5917249771828414,
+      "grad_norm": 0.8791665434837341,
+      "learning_rate": 3.471496665410561e-05,
+      "loss": 0.2185,
+      "step": 2616
+    },
+    {
+      "epoch": 1.5923334347429265,
+      "grad_norm": 0.9701746106147766,
+      "learning_rate": 3.4703494987960637e-05,
+      "loss": 0.2412,
+      "step": 2617
+    },
+    {
+      "epoch": 1.5929418923030119,
+      "grad_norm": 0.7525075674057007,
+      "learning_rate": 3.469202091575337e-05,
+      "loss": 0.1916,
+      "step": 2618
+    },
+    {
+      "epoch": 1.5935503498630972,
+      "grad_norm": 0.9566659331321716,
+      "learning_rate": 3.468054444032889e-05,
+      "loss": 0.1985,
+      "step": 2619
+    },
+    {
+      "epoch": 1.5941588074231823,
+      "grad_norm": 0.7912515997886658,
+      "learning_rate": 3.46690655645329e-05,
+      "loss": 0.2016,
+      "step": 2620
+    },
+    {
+      "epoch": 1.5947672649832674,
+      "grad_norm": 0.6893926858901978,
+      "learning_rate": 3.465758429121168e-05,
+      "loss": 0.1687,
+      "step": 2621
+    },
+    {
+      "epoch": 1.5953757225433525,
+      "grad_norm": 0.8185518980026245,
+      "learning_rate": 3.4646100623212094e-05,
+      "loss": 0.1726,
+      "step": 2622
+    },
+    {
+      "epoch": 1.5959841801034378,
+      "grad_norm": 1.019525170326233,
+      "learning_rate": 3.4634614563381624e-05,
+      "loss": 0.1899,
+      "step": 2623
+    },
+    {
+      "epoch": 1.596592637663523,
+      "grad_norm": 0.8567332625389099,
+      "learning_rate": 3.4623126114568336e-05,
+      "loss": 0.1901,
+      "step": 2624
+    },
+    {
+      "epoch": 1.5972010952236082,
+      "grad_norm": 0.7891209125518799,
+      "learning_rate": 3.4611635279620885e-05,
+      "loss": 0.1894,
+      "step": 2625
+    },
+    {
+      "epoch": 1.5978095527836933,
+      "grad_norm": 0.782440721988678,
+      "learning_rate": 3.460014206138851e-05,
+      "loss": 0.1761,
+      "step": 2626
+    },
+    {
+      "epoch": 1.5984180103437784,
+      "grad_norm": 1.1387567520141602,
+      "learning_rate": 3.458864646272107e-05,
+      "loss": 0.1906,
+      "step": 2627
+    },
+    {
+      "epoch": 1.5990264679038637,
+      "grad_norm": 0.766904354095459,
+      "learning_rate": 3.4577148486468975e-05,
+      "loss": 0.1905,
+      "step": 2628
+    },
+    {
+      "epoch": 1.599634925463949,
+      "grad_norm": 0.8891963362693787,
+      "learning_rate": 3.456564813548325e-05,
+      "loss": 0.1965,
+      "step": 2629
+    },
+    {
+      "epoch": 1.6002433830240341,
+      "grad_norm": 0.8071811199188232,
+      "learning_rate": 3.45541454126155e-05,
+      "loss": 0.185,
+      "step": 2630
+    },
+    {
+      "epoch": 1.6008518405841192,
+      "grad_norm": 0.9761980175971985,
+      "learning_rate": 3.4542640320717926e-05,
+      "loss": 0.2078,
+      "step": 2631
+    },
+    {
+      "epoch": 1.6014602981442043,
+      "grad_norm": 1.070326805114746,
+      "learning_rate": 3.453113286264332e-05,
+      "loss": 0.2261,
+      "step": 2632
+    },
+    {
+      "epoch": 1.6020687557042896,
+      "grad_norm": 0.7715917825698853,
+      "learning_rate": 3.4519623041245026e-05,
+      "loss": 0.1596,
+      "step": 2633
+    },
+    {
+      "epoch": 1.602677213264375,
+      "grad_norm": 0.7550219893455505,
+      "learning_rate": 3.450811085937702e-05,
+      "loss": 0.1756,
+      "step": 2634
+    },
+    {
+      "epoch": 1.60328567082446,
+      "grad_norm": 1.0285836458206177,
+      "learning_rate": 3.4496596319893844e-05,
+      "loss": 0.1657,
+      "step": 2635
+    },
+    {
+      "epoch": 1.6038941283845451,
+      "grad_norm": 0.9173053503036499,
+      "learning_rate": 3.448507942565061e-05,
+      "loss": 0.1729,
+      "step": 2636
+    },
+    {
+      "epoch": 1.6045025859446302,
+      "grad_norm": 3.4212629795074463,
+      "learning_rate": 3.4473560179503045e-05,
+      "loss": 0.2194,
+      "step": 2637
+    },
+    {
+      "epoch": 1.6051110435047156,
+      "grad_norm": 0.9600544571876526,
+      "learning_rate": 3.4462038584307424e-05,
+      "loss": 0.16,
+      "step": 2638
+    },
+    {
+      "epoch": 1.6057195010648009,
+      "grad_norm": 0.8442495465278625,
+      "learning_rate": 3.4450514642920636e-05,
+      "loss": 0.1872,
+      "step": 2639
+    },
+    {
+      "epoch": 1.606327958624886,
+      "grad_norm": 0.9204790592193604,
+      "learning_rate": 3.443898835820014e-05,
+      "loss": 0.1905,
+      "step": 2640
+    },
+    {
+      "epoch": 1.606936416184971,
+      "grad_norm": 0.9242935180664062,
+      "learning_rate": 3.442745973300395e-05,
+      "loss": 0.1767,
+      "step": 2641
+    },
+    {
+      "epoch": 1.6075448737450562,
+      "grad_norm": 0.7664668560028076,
+      "learning_rate": 3.441592877019072e-05,
+      "loss": 0.1682,
+      "step": 2642
+    },
+    {
+      "epoch": 1.6081533313051415,
+      "grad_norm": 0.7136707901954651,
+      "learning_rate": 3.440439547261962e-05,
+      "loss": 0.1695,
+      "step": 2643
+    },
+    {
+      "epoch": 1.6087617888652268,
+      "grad_norm": 0.8739480972290039,
+      "learning_rate": 3.4392859843150435e-05,
+      "loss": 0.1854,
+      "step": 2644
+    },
+    {
+      "epoch": 1.609370246425312,
+      "grad_norm": 0.9654523134231567,
+      "learning_rate": 3.4381321884643534e-05,
+      "loss": 0.1728,
+      "step": 2645
+    },
+    {
+      "epoch": 1.609978703985397,
+      "grad_norm": 0.9120699167251587,
+      "learning_rate": 3.4369781599959825e-05,
+      "loss": 0.1915,
+      "step": 2646
+    },
+    {
+      "epoch": 1.610587161545482,
+      "grad_norm": 0.8703420758247375,
+      "learning_rate": 3.4358238991960837e-05,
+      "loss": 0.1929,
+      "step": 2647
+    },
+    {
+      "epoch": 1.6111956191055674,
+      "grad_norm": 1.062511920928955,
+      "learning_rate": 3.434669406350866e-05,
+      "loss": 0.2282,
+      "step": 2648
+    },
+    {
+      "epoch": 1.6118040766656527,
+      "grad_norm": 0.9723806381225586,
+      "learning_rate": 3.433514681746593e-05,
+      "loss": 0.2546,
+      "step": 2649
+    },
+    {
+      "epoch": 1.6124125342257378,
+      "grad_norm": 0.8156774044036865,
+      "learning_rate": 3.4323597256695906e-05,
+      "loss": 0.1962,
+      "step": 2650
+    },
+    {
+      "epoch": 1.613020991785823,
+      "grad_norm": 0.6880114674568176,
+      "learning_rate": 3.4312045384062386e-05,
+      "loss": 0.1365,
+      "step": 2651
+    },
+    {
+      "epoch": 1.613629449345908,
+      "grad_norm": 0.9442213773727417,
+      "learning_rate": 3.430049120242975e-05,
+      "loss": 0.2038,
+      "step": 2652
+    },
+    {
+      "epoch": 1.6142379069059933,
+      "grad_norm": 0.8928872346878052,
+      "learning_rate": 3.428893471466297e-05,
+      "loss": 0.2134,
+      "step": 2653
+    },
+    {
+      "epoch": 1.6148463644660787,
+      "grad_norm": 0.7231988310813904,
+      "learning_rate": 3.427737592362755e-05,
+      "loss": 0.1674,
+      "step": 2654
+    },
+    {
+      "epoch": 1.6154548220261638,
+      "grad_norm": 0.8021813631057739,
+      "learning_rate": 3.4265814832189614e-05,
+      "loss": 0.1836,
+      "step": 2655
+    },
+    {
+      "epoch": 1.6160632795862488,
+      "grad_norm": 0.8998550772666931,
+      "learning_rate": 3.425425144321581e-05,
+      "loss": 0.1854,
+      "step": 2656
+    },
+    {
+      "epoch": 1.616671737146334,
+      "grad_norm": 0.8088403344154358,
+      "learning_rate": 3.424268575957339e-05,
+      "loss": 0.1573,
+      "step": 2657
+    },
+    {
+      "epoch": 1.6172801947064193,
+      "grad_norm": 0.9545161724090576,
+      "learning_rate": 3.423111778413015e-05,
+      "loss": 0.229,
+      "step": 2658
+    },
+    {
+      "epoch": 1.6178886522665044,
+      "grad_norm": 1.0301522016525269,
+      "learning_rate": 3.4219547519754475e-05,
+      "loss": 0.2269,
+      "step": 2659
+    },
+    {
+      "epoch": 1.6184971098265897,
+      "grad_norm": 1.287305474281311,
+      "learning_rate": 3.42079749693153e-05,
+      "loss": 0.2085,
+      "step": 2660
+    },
+    {
+      "epoch": 1.6191055673866748,
+      "grad_norm": 0.7046839594841003,
+      "learning_rate": 3.419640013568215e-05,
+      "loss": 0.1573,
+      "step": 2661
+    },
+    {
+      "epoch": 1.6197140249467599,
+      "grad_norm": 0.8109933137893677,
+      "learning_rate": 3.418482302172508e-05,
+      "loss": 0.173,
+      "step": 2662
+    },
+    {
+      "epoch": 1.6203224825068452,
+      "grad_norm": 0.792901873588562,
+      "learning_rate": 3.4173243630314754e-05,
+      "loss": 0.1872,
+      "step": 2663
+    },
+    {
+      "epoch": 1.6209309400669303,
+      "grad_norm": 0.8485409021377563,
+      "learning_rate": 3.4161661964322354e-05,
+      "loss": 0.1863,
+      "step": 2664
+    },
+    {
+      "epoch": 1.6215393976270156,
+      "grad_norm": 0.7961007356643677,
+      "learning_rate": 3.415007802661967e-05,
+      "loss": 0.1448,
+      "step": 2665
+    },
+    {
+      "epoch": 1.6221478551871007,
+      "grad_norm": 0.7520683407783508,
+      "learning_rate": 3.413849182007903e-05,
+      "loss": 0.1682,
+      "step": 2666
+    },
+    {
+      "epoch": 1.6227563127471858,
+      "grad_norm": 0.8546669483184814,
+      "learning_rate": 3.412690334757334e-05,
+      "loss": 0.1731,
+      "step": 2667
+    },
+    {
+      "epoch": 1.623364770307271,
+      "grad_norm": 0.9142362475395203,
+      "learning_rate": 3.411531261197603e-05,
+      "loss": 0.2352,
+      "step": 2668
+    },
+    {
+      "epoch": 1.6239732278673562,
+      "grad_norm": 0.7282983064651489,
+      "learning_rate": 3.410371961616114e-05,
+      "loss": 0.1176,
+      "step": 2669
+    },
+    {
+      "epoch": 1.6245816854274415,
+      "grad_norm": 0.8929833769798279,
+      "learning_rate": 3.409212436300326e-05,
+      "loss": 0.1912,
+      "step": 2670
+    },
+    {
+      "epoch": 1.6251901429875266,
+      "grad_norm": 0.7896724939346313,
+      "learning_rate": 3.40805268553775e-05,
+      "loss": 0.1885,
+      "step": 2671
+    },
+    {
+      "epoch": 1.6257986005476117,
+      "grad_norm": 0.9245496392250061,
+      "learning_rate": 3.4068927096159594e-05,
+      "loss": 0.1935,
+      "step": 2672
+    },
+    {
+      "epoch": 1.626407058107697,
+      "grad_norm": 0.8307329416275024,
+      "learning_rate": 3.4057325088225764e-05,
+      "loss": 0.1696,
+      "step": 2673
+    },
+    {
+      "epoch": 1.6270155156677821,
+      "grad_norm": 0.9103578925132751,
+      "learning_rate": 3.404572083445285e-05,
+      "loss": 0.1489,
+      "step": 2674
+    },
+    {
+      "epoch": 1.6276239732278674,
+      "grad_norm": 0.868897557258606,
+      "learning_rate": 3.40341143377182e-05,
+      "loss": 0.2073,
+      "step": 2675
+    },
+    {
+      "epoch": 1.6282324307879525,
+      "grad_norm": 0.79030442237854,
+      "learning_rate": 3.402250560089977e-05,
+      "loss": 0.195,
+      "step": 2676
+    },
+    {
+      "epoch": 1.6288408883480376,
+      "grad_norm": 0.7882320284843445,
+      "learning_rate": 3.401089462687602e-05,
+      "loss": 0.1669,
+      "step": 2677
+    },
+    {
+      "epoch": 1.629449345908123,
+      "grad_norm": 0.920220673084259,
+      "learning_rate": 3.399928141852599e-05,
+      "loss": 0.2067,
+      "step": 2678
+    },
+    {
+      "epoch": 1.630057803468208,
+      "grad_norm": 1.1648184061050415,
+      "learning_rate": 3.398766597872928e-05,
+      "loss": 0.1998,
+      "step": 2679
+    },
+    {
+      "epoch": 1.6306662610282934,
+      "grad_norm": 1.0704290866851807,
+      "learning_rate": 3.397604831036604e-05,
+      "loss": 0.2699,
+      "step": 2680
+    },
+    {
+      "epoch": 1.6312747185883785,
+      "grad_norm": 1.071129560470581,
+      "learning_rate": 3.396442841631695e-05,
+      "loss": 0.224,
+      "step": 2681
+    },
+    {
+      "epoch": 1.6318831761484636,
+      "grad_norm": 0.7212443947792053,
+      "learning_rate": 3.395280629946327e-05,
+      "loss": 0.1551,
+      "step": 2682
+    },
+    {
+      "epoch": 1.6324916337085489,
+      "grad_norm": 0.8159400224685669,
+      "learning_rate": 3.39411819626868e-05,
+      "loss": 0.1954,
+      "step": 2683
+    },
+    {
+      "epoch": 1.633100091268634,
+      "grad_norm": 0.8605476021766663,
+      "learning_rate": 3.3929555408869896e-05,
+      "loss": 0.2296,
+      "step": 2684
+    },
+    {
+      "epoch": 1.6337085488287193,
+      "grad_norm": 0.8847449421882629,
+      "learning_rate": 3.3917926640895445e-05,
+      "loss": 0.2023,
+      "step": 2685
+    },
+    {
+      "epoch": 1.6343170063888044,
+      "grad_norm": 0.8677266240119934,
+      "learning_rate": 3.390629566164691e-05,
+      "loss": 0.2,
+      "step": 2686
+    },
+    {
+      "epoch": 1.6349254639488895,
+      "grad_norm": 0.8311256170272827,
+      "learning_rate": 3.389466247400828e-05,
+      "loss": 0.2025,
+      "step": 2687
+    },
+    {
+      "epoch": 1.6355339215089746,
+      "grad_norm": 3.200028896331787,
+      "learning_rate": 3.38830270808641e-05,
+      "loss": 0.2398,
+      "step": 2688
+    },
+    {
+      "epoch": 1.63614237906906,
+      "grad_norm": 0.7760987877845764,
+      "learning_rate": 3.3871389485099474e-05,
+      "loss": 0.1651,
+      "step": 2689
+    },
+    {
+      "epoch": 1.6367508366291452,
+      "grad_norm": 1.227342128753662,
+      "learning_rate": 3.385974968960003e-05,
+      "loss": 0.1905,
+      "step": 2690
+    },
+    {
+      "epoch": 1.6373592941892303,
+      "grad_norm": 0.832114577293396,
+      "learning_rate": 3.384810769725196e-05,
+      "loss": 0.1901,
+      "step": 2691
+    },
+    {
+      "epoch": 1.6379677517493154,
+      "grad_norm": 0.7426611185073853,
+      "learning_rate": 3.383646351094198e-05,
+      "loss": 0.1425,
+      "step": 2692
+    },
+    {
+      "epoch": 1.6385762093094005,
+      "grad_norm": 0.7370221018791199,
+      "learning_rate": 3.382481713355738e-05,
+      "loss": 0.1638,
+      "step": 2693
+    },
+    {
+      "epoch": 1.6391846668694858,
+      "grad_norm": 0.9661339521408081,
+      "learning_rate": 3.381316856798596e-05,
+      "loss": 0.1937,
+      "step": 2694
+    },
+    {
+      "epoch": 1.6397931244295711,
+      "grad_norm": 0.8891693353652954,
+      "learning_rate": 3.3801517817116094e-05,
+      "loss": 0.1547,
+      "step": 2695
+    },
+    {
+      "epoch": 1.6404015819896562,
+      "grad_norm": 1.042301058769226,
+      "learning_rate": 3.378986488383667e-05,
+      "loss": 0.2168,
+      "step": 2696
+    },
+    {
+      "epoch": 1.6410100395497413,
+      "grad_norm": 0.7799018025398254,
+      "learning_rate": 3.377820977103714e-05,
+      "loss": 0.1444,
+      "step": 2697
+    },
+    {
+      "epoch": 1.6416184971098264,
+      "grad_norm": 0.9158018231391907,
+      "learning_rate": 3.376655248160747e-05,
+      "loss": 0.2048,
+      "step": 2698
+    },
+    {
+      "epoch": 1.6422269546699118,
+      "grad_norm": 0.9366563558578491,
+      "learning_rate": 3.37548930184382e-05,
+      "loss": 0.2095,
+      "step": 2699
+    },
+    {
+      "epoch": 1.642835412229997,
+      "grad_norm": 0.7867175340652466,
+      "learning_rate": 3.3743231384420384e-05,
+      "loss": 0.1721,
+      "step": 2700
+    },
+    {
+      "epoch": 1.6434438697900822,
+      "grad_norm": 0.9113629460334778,
+      "learning_rate": 3.3731567582445615e-05,
+      "loss": 0.1875,
+      "step": 2701
+    },
+    {
+      "epoch": 1.6440523273501673,
+      "grad_norm": 0.9016094207763672,
+      "learning_rate": 3.371990161540603e-05,
+      "loss": 0.2127,
+      "step": 2702
+    },
+    {
+      "epoch": 1.6446607849102524,
+      "grad_norm": 0.8337920904159546,
+      "learning_rate": 3.3708233486194324e-05,
+      "loss": 0.1788,
+      "step": 2703
+    },
+    {
+      "epoch": 1.6452692424703377,
+      "grad_norm": 0.8652783036231995,
+      "learning_rate": 3.369656319770369e-05,
+      "loss": 0.1558,
+      "step": 2704
+    },
+    {
+      "epoch": 1.645877700030423,
+      "grad_norm": 0.8067019581794739,
+      "learning_rate": 3.368489075282786e-05,
+      "loss": 0.174,
+      "step": 2705
+    },
+    {
+      "epoch": 1.646486157590508,
+      "grad_norm": 0.7883349657058716,
+      "learning_rate": 3.367321615446113e-05,
+      "loss": 0.1485,
+      "step": 2706
+    },
+    {
+      "epoch": 1.6470946151505932,
+      "grad_norm": 1.0703591108322144,
+      "learning_rate": 3.366153940549832e-05,
+      "loss": 0.1884,
+      "step": 2707
+    },
+    {
+      "epoch": 1.6477030727106783,
+      "grad_norm": 0.6376674175262451,
+      "learning_rate": 3.364986050883476e-05,
+      "loss": 0.1306,
+      "step": 2708
+    },
+    {
+      "epoch": 1.6483115302707636,
+      "grad_norm": 0.9308871030807495,
+      "learning_rate": 3.363817946736634e-05,
+      "loss": 0.1994,
+      "step": 2709
+    },
+    {
+      "epoch": 1.648919987830849,
+      "grad_norm": 0.8675797581672668,
+      "learning_rate": 3.3626496283989476e-05,
+      "loss": 0.197,
+      "step": 2710
+    },
+    {
+      "epoch": 1.649528445390934,
+      "grad_norm": 0.9659071564674377,
+      "learning_rate": 3.361481096160109e-05,
+      "loss": 0.2214,
+      "step": 2711
+    },
+    {
+      "epoch": 1.6501369029510191,
+      "grad_norm": 0.8575699925422668,
+      "learning_rate": 3.3603123503098675e-05,
+      "loss": 0.1685,
+      "step": 2712
+    },
+    {
+      "epoch": 1.6507453605111042,
+      "grad_norm": 0.8685420155525208,
+      "learning_rate": 3.359143391138023e-05,
+      "loss": 0.1846,
+      "step": 2713
+    },
+    {
+      "epoch": 1.6513538180711895,
+      "grad_norm": 0.862989604473114,
+      "learning_rate": 3.357974218934429e-05,
+      "loss": 0.2076,
+      "step": 2714
+    },
+    {
+      "epoch": 1.6519622756312748,
+      "grad_norm": 0.9207594990730286,
+      "learning_rate": 3.356804833988989e-05,
+      "loss": 0.1636,
+      "step": 2715
+    },
+    {
+      "epoch": 1.65257073319136,
+      "grad_norm": 0.9252761602401733,
+      "learning_rate": 3.3556352365916646e-05,
+      "loss": 0.2106,
+      "step": 2716
+    },
+    {
+      "epoch": 1.653179190751445,
+      "grad_norm": 1.0307517051696777,
+      "learning_rate": 3.354465427032467e-05,
+      "loss": 0.255,
+      "step": 2717
+    },
+    {
+      "epoch": 1.6537876483115301,
+      "grad_norm": 0.7410596013069153,
+      "learning_rate": 3.353295405601457e-05,
+      "loss": 0.1325,
+      "step": 2718
+    },
+    {
+      "epoch": 1.6543961058716155,
+      "grad_norm": 0.8445053696632385,
+      "learning_rate": 3.352125172588756e-05,
+      "loss": 0.1847,
+      "step": 2719
+    },
+    {
+      "epoch": 1.6550045634317008,
+      "grad_norm": 0.7926309704780579,
+      "learning_rate": 3.350954728284529e-05,
+      "loss": 0.178,
+      "step": 2720
+    },
+    {
+      "epoch": 1.6556130209917859,
+      "grad_norm": 0.8683445453643799,
+      "learning_rate": 3.349784072979e-05,
+      "loss": 0.1496,
+      "step": 2721
+    },
+    {
+      "epoch": 1.656221478551871,
+      "grad_norm": 0.9647382497787476,
+      "learning_rate": 3.348613206962441e-05,
+      "loss": 0.1771,
+      "step": 2722
+    },
+    {
+      "epoch": 1.656829936111956,
+      "grad_norm": 1.0857597589492798,
+      "learning_rate": 3.3474421305251785e-05,
+      "loss": 0.1893,
+      "step": 2723
+    },
+    {
+      "epoch": 1.6574383936720414,
+      "grad_norm": 0.904272198677063,
+      "learning_rate": 3.3462708439575916e-05,
+      "loss": 0.1812,
+      "step": 2724
+    },
+    {
+      "epoch": 1.6580468512321267,
+      "grad_norm": 0.9223242402076721,
+      "learning_rate": 3.3450993475501084e-05,
+      "loss": 0.2006,
+      "step": 2725
+    },
+    {
+      "epoch": 1.6586553087922118,
+      "grad_norm": 0.8843355178833008,
+      "learning_rate": 3.343927641593213e-05,
+      "loss": 0.1732,
+      "step": 2726
+    },
+    {
+      "epoch": 1.6592637663522969,
+      "grad_norm": 0.8139360547065735,
+      "learning_rate": 3.3427557263774395e-05,
+      "loss": 0.1904,
+      "step": 2727
+    },
+    {
+      "epoch": 1.659872223912382,
+      "grad_norm": 0.8607851266860962,
+      "learning_rate": 3.341583602193373e-05,
+      "loss": 0.1773,
+      "step": 2728
+    },
+    {
+      "epoch": 1.6604806814724673,
+      "grad_norm": 0.7943541407585144,
+      "learning_rate": 3.3404112693316525e-05,
+      "loss": 0.1659,
+      "step": 2729
+    },
+    {
+      "epoch": 1.6610891390325526,
+      "grad_norm": 0.8024083971977234,
+      "learning_rate": 3.339238728082968e-05,
+      "loss": 0.1562,
+      "step": 2730
+    },
+    {
+      "epoch": 1.6616975965926377,
+      "grad_norm": 0.7931082248687744,
+      "learning_rate": 3.338065978738059e-05,
+      "loss": 0.1367,
+      "step": 2731
+    },
+    {
+      "epoch": 1.6623060541527228,
+      "grad_norm": 0.8233804106712341,
+      "learning_rate": 3.33689302158772e-05,
+      "loss": 0.1757,
+      "step": 2732
+    },
+    {
+      "epoch": 1.662914511712808,
+      "grad_norm": 0.9862247705459595,
+      "learning_rate": 3.3357198569227954e-05,
+      "loss": 0.1799,
+      "step": 2733
+    },
+    {
+      "epoch": 1.6635229692728932,
+      "grad_norm": 0.7572802901268005,
+      "learning_rate": 3.33454648503418e-05,
+      "loss": 0.1937,
+      "step": 2734
+    },
+    {
+      "epoch": 1.6641314268329785,
+      "grad_norm": 0.764121949672699,
+      "learning_rate": 3.333372906212823e-05,
+      "loss": 0.1609,
+      "step": 2735
+    },
+    {
+      "epoch": 1.6647398843930636,
+      "grad_norm": 0.9266926646232605,
+      "learning_rate": 3.332199120749721e-05,
+      "loss": 0.1643,
+      "step": 2736
+    },
+    {
+      "epoch": 1.6653483419531487,
+      "grad_norm": 0.7189835906028748,
+      "learning_rate": 3.331025128935926e-05,
+      "loss": 0.1632,
+      "step": 2737
+    },
+    {
+      "epoch": 1.6659567995132338,
+      "grad_norm": 0.9593905806541443,
+      "learning_rate": 3.3298509310625363e-05,
+      "loss": 0.1825,
+      "step": 2738
+    },
+    {
+      "epoch": 1.6665652570733192,
+      "grad_norm": 0.8221538066864014,
+      "learning_rate": 3.328676527420706e-05,
+      "loss": 0.184,
+      "step": 2739
+    },
+    {
+      "epoch": 1.6671737146334045,
+      "grad_norm": 0.8926746249198914,
+      "learning_rate": 3.3275019183016384e-05,
+      "loss": 0.1851,
+      "step": 2740
+    },
+    {
+      "epoch": 1.6677821721934896,
+      "grad_norm": 0.8557993769645691,
+      "learning_rate": 3.326327103996587e-05,
+      "loss": 0.1588,
+      "step": 2741
+    },
+    {
+      "epoch": 1.6683906297535747,
+      "grad_norm": 0.701745867729187,
+      "learning_rate": 3.3251520847968566e-05,
+      "loss": 0.1516,
+      "step": 2742
+    },
+    {
+      "epoch": 1.6689990873136598,
+      "grad_norm": 0.9786536693572998,
+      "learning_rate": 3.323976860993803e-05,
+      "loss": 0.1958,
+      "step": 2743
+    },
+    {
+      "epoch": 1.669607544873745,
+      "grad_norm": 0.7492821216583252,
+      "learning_rate": 3.322801432878833e-05,
+      "loss": 0.1302,
+      "step": 2744
+    },
+    {
+      "epoch": 1.6702160024338304,
+      "grad_norm": 0.8823762536048889,
+      "learning_rate": 3.321625800743404e-05,
+      "loss": 0.1697,
+      "step": 2745
+    },
+    {
+      "epoch": 1.6708244599939155,
+      "grad_norm": 1.0150517225265503,
+      "learning_rate": 3.320449964879023e-05,
+      "loss": 0.1984,
+      "step": 2746
+    },
+    {
+      "epoch": 1.6714329175540006,
+      "grad_norm": 0.9592936635017395,
+      "learning_rate": 3.31927392557725e-05,
+      "loss": 0.2018,
+      "step": 2747
+    },
+    {
+      "epoch": 1.6720413751140857,
+      "grad_norm": 0.9560689926147461,
+      "learning_rate": 3.318097683129691e-05,
+      "loss": 0.1886,
+      "step": 2748
+    },
+    {
+      "epoch": 1.672649832674171,
+      "grad_norm": 1.1188546419143677,
+      "learning_rate": 3.316921237828007e-05,
+      "loss": 0.2193,
+      "step": 2749
+    },
+    {
+      "epoch": 1.6732582902342563,
+      "grad_norm": 0.8860459923744202,
+      "learning_rate": 3.315744589963907e-05,
+      "loss": 0.1493,
+      "step": 2750
+    },
+    {
+      "epoch": 1.6738667477943414,
+      "grad_norm": 0.9111357927322388,
+      "learning_rate": 3.314567739829151e-05,
+      "loss": 0.2217,
+      "step": 2751
+    },
+    {
+      "epoch": 1.6744752053544265,
+      "grad_norm": 0.9053143262863159,
+      "learning_rate": 3.313390687715548e-05,
+      "loss": 0.2102,
+      "step": 2752
+    },
+    {
+      "epoch": 1.6750836629145116,
+      "grad_norm": 0.8748982548713684,
+      "learning_rate": 3.3122134339149585e-05,
+      "loss": 0.1728,
+      "step": 2753
+    },
+    {
+      "epoch": 1.675692120474597,
+      "grad_norm": 0.8217591643333435,
+      "learning_rate": 3.311035978719292e-05,
+      "loss": 0.1708,
+      "step": 2754
+    },
+    {
+      "epoch": 1.6763005780346822,
+      "grad_norm": 0.9044924378395081,
+      "learning_rate": 3.309858322420508e-05,
+      "loss": 0.2065,
+      "step": 2755
+    },
+    {
+      "epoch": 1.6769090355947673,
+      "grad_norm": 0.7668700814247131,
+      "learning_rate": 3.308680465310617e-05,
+      "loss": 0.2101,
+      "step": 2756
+    },
+    {
+      "epoch": 1.6775174931548524,
+      "grad_norm": 0.9722397327423096,
+      "learning_rate": 3.307502407681678e-05,
+      "loss": 0.1881,
+      "step": 2757
+    },
+    {
+      "epoch": 1.6781259507149375,
+      "grad_norm": 0.842012882232666,
+      "learning_rate": 3.3063241498258e-05,
+      "loss": 0.1471,
+      "step": 2758
+    },
+    {
+      "epoch": 1.6787344082750228,
+      "grad_norm": 0.8087152242660522,
+      "learning_rate": 3.305145692035143e-05,
+      "loss": 0.1465,
+      "step": 2759
+    },
+    {
+      "epoch": 1.6793428658351082,
+      "grad_norm": 0.8884347081184387,
+      "learning_rate": 3.303967034601914e-05,
+      "loss": 0.1811,
+      "step": 2760
+    },
+    {
+      "epoch": 1.6799513233951933,
+      "grad_norm": 1.1114667654037476,
+      "learning_rate": 3.3027881778183715e-05,
+      "loss": 0.1998,
+      "step": 2761
+    },
+    {
+      "epoch": 1.6805597809552784,
+      "grad_norm": 0.8144546151161194,
+      "learning_rate": 3.301609121976822e-05,
+      "loss": 0.1806,
+      "step": 2762
+    },
+    {
+      "epoch": 1.6811682385153635,
+      "grad_norm": 0.858739972114563,
+      "learning_rate": 3.300429867369623e-05,
+      "loss": 0.2098,
+      "step": 2763
+    },
+    {
+      "epoch": 1.6817766960754488,
+      "grad_norm": 0.9147648811340332,
+      "learning_rate": 3.299250414289181e-05,
+      "loss": 0.1956,
+      "step": 2764
+    },
+    {
+      "epoch": 1.6823851536355339,
+      "grad_norm": 0.8137851357460022,
+      "learning_rate": 3.298070763027951e-05,
+      "loss": 0.2142,
+      "step": 2765
+    },
+    {
+      "epoch": 1.6829936111956192,
+      "grad_norm": 0.8987473845481873,
+      "learning_rate": 3.296890913878436e-05,
+      "loss": 0.1742,
+      "step": 2766
+    },
+    {
+      "epoch": 1.6836020687557043,
+      "grad_norm": 0.6859608888626099,
+      "learning_rate": 3.295710867133191e-05,
+      "loss": 0.1589,
+      "step": 2767
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.7220624685287476,
+      "learning_rate": 3.2945306230848185e-05,
+      "loss": 0.138,
+      "step": 2768
+    },
+    {
+      "epoch": 1.6848189838758747,
+      "grad_norm": 0.6847586631774902,
+      "learning_rate": 3.293350182025968e-05,
+      "loss": 0.1614,
+      "step": 2769
+    },
+    {
+      "epoch": 1.6854274414359598,
+      "grad_norm": 0.846610426902771,
+      "learning_rate": 3.292169544249341e-05,
+      "loss": 0.1999,
+      "step": 2770
+    },
+    {
+      "epoch": 1.6860358989960451,
+      "grad_norm": 0.7773505449295044,
+      "learning_rate": 3.290988710047687e-05,
+      "loss": 0.1724,
+      "step": 2771
+    },
+    {
+      "epoch": 1.6866443565561302,
+      "grad_norm": 0.7428063750267029,
+      "learning_rate": 3.289807679713803e-05,
+      "loss": 0.1307,
+      "step": 2772
+    },
+    {
+      "epoch": 1.6872528141162153,
+      "grad_norm": 0.8252652883529663,
+      "learning_rate": 3.288626453540535e-05,
+      "loss": 0.2018,
+      "step": 2773
+    },
+    {
+      "epoch": 1.6878612716763006,
+      "grad_norm": 0.8059362173080444,
+      "learning_rate": 3.287445031820777e-05,
+      "loss": 0.1656,
+      "step": 2774
+    },
+    {
+      "epoch": 1.6884697292363857,
+      "grad_norm": 0.9173924326896667,
+      "learning_rate": 3.2862634148474744e-05,
+      "loss": 0.1938,
+      "step": 2775
+    },
+    {
+      "epoch": 1.689078186796471,
+      "grad_norm": 0.8214101791381836,
+      "learning_rate": 3.285081602913618e-05,
+      "loss": 0.1551,
+      "step": 2776
+    },
+    {
+      "epoch": 1.6896866443565561,
+      "grad_norm": 0.9609821438789368,
+      "learning_rate": 3.283899596312247e-05,
+      "loss": 0.1968,
+      "step": 2777
+    },
+    {
+      "epoch": 1.6902951019166412,
+      "grad_norm": 0.8379231095314026,
+      "learning_rate": 3.2827173953364526e-05,
+      "loss": 0.1466,
+      "step": 2778
+    },
+    {
+      "epoch": 1.6909035594767265,
+      "grad_norm": 0.7881366014480591,
+      "learning_rate": 3.281535000279368e-05,
+      "loss": 0.132,
+      "step": 2779
+    },
+    {
+      "epoch": 1.6915120170368116,
+      "grad_norm": 0.8496389389038086,
+      "learning_rate": 3.2803524114341795e-05,
+      "loss": 0.2176,
+      "step": 2780
+    },
+    {
+      "epoch": 1.692120474596897,
+      "grad_norm": 0.7638592720031738,
+      "learning_rate": 3.2791696290941196e-05,
+      "loss": 0.1763,
+      "step": 2781
+    },
+    {
+      "epoch": 1.692728932156982,
+      "grad_norm": 0.9417856335639954,
+      "learning_rate": 3.27798665355247e-05,
+      "loss": 0.1721,
+      "step": 2782
+    },
+    {
+      "epoch": 1.6933373897170672,
+      "grad_norm": 0.788231611251831,
+      "learning_rate": 3.276803485102557e-05,
+      "loss": 0.1493,
+      "step": 2783
+    },
+    {
+      "epoch": 1.6939458472771525,
+      "grad_norm": 0.8815106153488159,
+      "learning_rate": 3.2756201240377596e-05,
+      "loss": 0.1762,
+      "step": 2784
+    },
+    {
+      "epoch": 1.6945543048372376,
+      "grad_norm": 0.8450623750686646,
+      "learning_rate": 3.2744365706514995e-05,
+      "loss": 0.2006,
+      "step": 2785
+    },
+    {
+      "epoch": 1.6951627623973229,
+      "grad_norm": 1.0778264999389648,
+      "learning_rate": 3.273252825237251e-05,
+      "loss": 0.2408,
+      "step": 2786
+    },
+    {
+      "epoch": 1.695771219957408,
+      "grad_norm": 0.7945128083229065,
+      "learning_rate": 3.2720688880885324e-05,
+      "loss": 0.1696,
+      "step": 2787
+    },
+    {
+      "epoch": 1.696379677517493,
+      "grad_norm": 0.7426414489746094,
+      "learning_rate": 3.270884759498911e-05,
+      "loss": 0.157,
+      "step": 2788
+    },
+    {
+      "epoch": 1.6969881350775782,
+      "grad_norm": 0.8385169506072998,
+      "learning_rate": 3.269700439762001e-05,
+      "loss": 0.1832,
+      "step": 2789
+    },
+    {
+      "epoch": 1.6975965926376635,
+      "grad_norm": 1.0522770881652832,
+      "learning_rate": 3.268515929171465e-05,
+      "loss": 0.2037,
+      "step": 2790
+    },
+    {
+      "epoch": 1.6982050501977488,
+      "grad_norm": 0.7473107576370239,
+      "learning_rate": 3.2673312280210124e-05,
+      "loss": 0.1878,
+      "step": 2791
+    },
+    {
+      "epoch": 1.698813507757834,
+      "grad_norm": 0.7824332118034363,
+      "learning_rate": 3.2661463366043985e-05,
+      "loss": 0.1556,
+      "step": 2792
+    },
+    {
+      "epoch": 1.699421965317919,
+      "grad_norm": 0.8999038338661194,
+      "learning_rate": 3.2649612552154276e-05,
+      "loss": 0.1974,
+      "step": 2793
+    },
+    {
+      "epoch": 1.700030422878004,
+      "grad_norm": 0.7992719411849976,
+      "learning_rate": 3.263775984147951e-05,
+      "loss": 0.1679,
+      "step": 2794
+    },
+    {
+      "epoch": 1.7006388804380894,
+      "grad_norm": 0.7823406457901001,
+      "learning_rate": 3.2625905236958655e-05,
+      "loss": 0.1676,
+      "step": 2795
+    },
+    {
+      "epoch": 1.7012473379981747,
+      "grad_norm": 0.8602148294448853,
+      "learning_rate": 3.2614048741531166e-05,
+      "loss": 0.1666,
+      "step": 2796
+    },
+    {
+      "epoch": 1.7018557955582598,
+      "grad_norm": 0.8012514114379883,
+      "learning_rate": 3.2602190358136965e-05,
+      "loss": 0.1704,
+      "step": 2797
+    },
+    {
+      "epoch": 1.702464253118345,
+      "grad_norm": 0.9197515249252319,
+      "learning_rate": 3.259033008971642e-05,
+      "loss": 0.1809,
+      "step": 2798
+    },
+    {
+      "epoch": 1.70307271067843,
+      "grad_norm": 1.046052098274231,
+      "learning_rate": 3.25784679392104e-05,
+      "loss": 0.2223,
+      "step": 2799
+    },
+    {
+      "epoch": 1.7036811682385153,
+      "grad_norm": 0.7376840710639954,
+      "learning_rate": 3.256660390956022e-05,
+      "loss": 0.1596,
+      "step": 2800
+    },
+    {
+      "epoch": 1.7042896257986007,
+      "grad_norm": 0.8339237570762634,
+      "learning_rate": 3.255473800370765e-05,
+      "loss": 0.1818,
+      "step": 2801
+    },
+    {
+      "epoch": 1.7048980833586858,
+      "grad_norm": 0.7850296497344971,
+      "learning_rate": 3.254287022459496e-05,
+      "loss": 0.1752,
+      "step": 2802
+    },
+    {
+      "epoch": 1.7055065409187709,
+      "grad_norm": 0.8697124719619751,
+      "learning_rate": 3.253100057516486e-05,
+      "loss": 0.1768,
+      "step": 2803
+    },
+    {
+      "epoch": 1.706114998478856,
+      "grad_norm": 1.1126950979232788,
+      "learning_rate": 3.251912905836052e-05,
+      "loss": 0.2258,
+      "step": 2804
+    },
+    {
+      "epoch": 1.7067234560389413,
+      "grad_norm": 0.7312096953392029,
+      "learning_rate": 3.250725567712559e-05,
+      "loss": 0.1779,
+      "step": 2805
+    },
+    {
+      "epoch": 1.7073319135990266,
+      "grad_norm": 0.7978745102882385,
+      "learning_rate": 3.2495380434404167e-05,
+      "loss": 0.1554,
+      "step": 2806
+    },
+    {
+      "epoch": 1.7079403711591117,
+      "grad_norm": 0.9084555506706238,
+      "learning_rate": 3.248350333314082e-05,
+      "loss": 0.1468,
+      "step": 2807
+    },
+    {
+      "epoch": 1.7085488287191968,
+      "grad_norm": 0.7953585386276245,
+      "learning_rate": 3.247162437628057e-05,
+      "loss": 0.1308,
+      "step": 2808
+    },
+    {
+      "epoch": 1.7091572862792819,
+      "grad_norm": 0.7833218574523926,
+      "learning_rate": 3.2459743566768916e-05,
+      "loss": 0.1697,
+      "step": 2809
+    },
+    {
+      "epoch": 1.7097657438393672,
+      "grad_norm": 0.7225537896156311,
+      "learning_rate": 3.2447860907551786e-05,
+      "loss": 0.1569,
+      "step": 2810
+    },
+    {
+      "epoch": 1.7103742013994525,
+      "grad_norm": 0.7659318447113037,
+      "learning_rate": 3.24359764015756e-05,
+      "loss": 0.1846,
+      "step": 2811
+    },
+    {
+      "epoch": 1.7109826589595376,
+      "grad_norm": 0.8912885189056396,
+      "learning_rate": 3.242409005178721e-05,
+      "loss": 0.1805,
+      "step": 2812
+    },
+    {
+      "epoch": 1.7115911165196227,
+      "grad_norm": 0.7760403156280518,
+      "learning_rate": 3.241220186113394e-05,
+      "loss": 0.1389,
+      "step": 2813
+    },
+    {
+      "epoch": 1.7121995740797078,
+      "grad_norm": 1.0232988595962524,
+      "learning_rate": 3.2400311832563563e-05,
+      "loss": 0.2085,
+      "step": 2814
+    },
+    {
+      "epoch": 1.7128080316397931,
+      "grad_norm": 0.8213723301887512,
+      "learning_rate": 3.238841996902431e-05,
+      "loss": 0.1491,
+      "step": 2815
+    },
+    {
+      "epoch": 1.7134164891998784,
+      "grad_norm": 0.8206520676612854,
+      "learning_rate": 3.237652627346487e-05,
+      "loss": 0.1543,
+      "step": 2816
+    },
+    {
+      "epoch": 1.7140249467599635,
+      "grad_norm": 0.885313093662262,
+      "learning_rate": 3.2364630748834385e-05,
+      "loss": 0.1985,
+      "step": 2817
+    },
+    {
+      "epoch": 1.7146334043200486,
+      "grad_norm": 0.7833919525146484,
+      "learning_rate": 3.235273339808245e-05,
+      "loss": 0.1493,
+      "step": 2818
+    },
+    {
+      "epoch": 1.7152418618801337,
+      "grad_norm": 0.8376486301422119,
+      "learning_rate": 3.2340834224159104e-05,
+      "loss": 0.164,
+      "step": 2819
+    },
+    {
+      "epoch": 1.715850319440219,
+      "grad_norm": 0.8096402287483215,
+      "learning_rate": 3.232893323001485e-05,
+      "loss": 0.1883,
+      "step": 2820
+    },
+    {
+      "epoch": 1.7164587770003044,
+      "grad_norm": 0.881629467010498,
+      "learning_rate": 3.2317030418600645e-05,
+      "loss": 0.1901,
+      "step": 2821
+    },
+    {
+      "epoch": 1.7170672345603895,
+      "grad_norm": 0.910102128982544,
+      "learning_rate": 3.2305125792867886e-05,
+      "loss": 0.1814,
+      "step": 2822
+    },
+    {
+      "epoch": 1.7176756921204746,
+      "grad_norm": 0.7410189509391785,
+      "learning_rate": 3.229321935576842e-05,
+      "loss": 0.1637,
+      "step": 2823
+    },
+    {
+      "epoch": 1.7182841496805596,
+      "grad_norm": 0.8830031752586365,
+      "learning_rate": 3.228131111025455e-05,
+      "loss": 0.1492,
+      "step": 2824
+    },
+    {
+      "epoch": 1.718892607240645,
+      "grad_norm": 0.9548397660255432,
+      "learning_rate": 3.226940105927903e-05,
+      "loss": 0.1643,
+      "step": 2825
+    },
+    {
+      "epoch": 1.7195010648007303,
+      "grad_norm": 0.829814612865448,
+      "learning_rate": 3.2257489205795034e-05,
+      "loss": 0.1411,
+      "step": 2826
+    },
+    {
+      "epoch": 1.7201095223608154,
+      "grad_norm": 0.9172588586807251,
+      "learning_rate": 3.224557555275623e-05,
+      "loss": 0.1536,
+      "step": 2827
+    },
+    {
+      "epoch": 1.7207179799209005,
+      "grad_norm": 0.8592941164970398,
+      "learning_rate": 3.223366010311671e-05,
+      "loss": 0.1777,
+      "step": 2828
+    },
+    {
+      "epoch": 1.7213264374809856,
+      "grad_norm": 0.8923897743225098,
+      "learning_rate": 3.2221742859831e-05,
+      "loss": 0.1696,
+      "step": 2829
+    },
+    {
+      "epoch": 1.721934895041071,
+      "grad_norm": 1.0709412097930908,
+      "learning_rate": 3.220982382585406e-05,
+      "loss": 0.2554,
+      "step": 2830
+    },
+    {
+      "epoch": 1.7225433526011562,
+      "grad_norm": 0.8038744330406189,
+      "learning_rate": 3.219790300414134e-05,
+      "loss": 0.1586,
+      "step": 2831
+    },
+    {
+      "epoch": 1.7231518101612413,
+      "grad_norm": 0.8866575956344604,
+      "learning_rate": 3.21859803976487e-05,
+      "loss": 0.1691,
+      "step": 2832
+    },
+    {
+      "epoch": 1.7237602677213264,
+      "grad_norm": 0.8635597825050354,
+      "learning_rate": 3.217405600933245e-05,
+      "loss": 0.1865,
+      "step": 2833
+    },
+    {
+      "epoch": 1.7243687252814115,
+      "grad_norm": 0.8194993138313293,
+      "learning_rate": 3.2162129842149336e-05,
+      "loss": 0.1614,
+      "step": 2834
+    },
+    {
+      "epoch": 1.7249771828414968,
+      "grad_norm": 0.8445285558700562,
+      "learning_rate": 3.215020189905655e-05,
+      "loss": 0.2077,
+      "step": 2835
+    },
+    {
+      "epoch": 1.7255856404015821,
+      "grad_norm": 0.9541175365447998,
+      "learning_rate": 3.213827218301173e-05,
+      "loss": 0.2071,
+      "step": 2836
+    },
+    {
+      "epoch": 1.7261940979616672,
+      "grad_norm": 0.7591641545295715,
+      "learning_rate": 3.212634069697295e-05,
+      "loss": 0.1616,
+      "step": 2837
+    },
+    {
+      "epoch": 1.7268025555217523,
+      "grad_norm": 0.7483083605766296,
+      "learning_rate": 3.211440744389871e-05,
+      "loss": 0.1756,
+      "step": 2838
+    },
+    {
+      "epoch": 1.7274110130818374,
+      "grad_norm": 0.7312026619911194,
+      "learning_rate": 3.2102472426747975e-05,
+      "loss": 0.1834,
+      "step": 2839
+    },
+    {
+      "epoch": 1.7280194706419227,
+      "grad_norm": 0.8581993579864502,
+      "learning_rate": 3.2090535648480126e-05,
+      "loss": 0.1491,
+      "step": 2840
+    },
+    {
+      "epoch": 1.728627928202008,
+      "grad_norm": 0.7880068421363831,
+      "learning_rate": 3.207859711205498e-05,
+      "loss": 0.1904,
+      "step": 2841
+    },
+    {
+      "epoch": 1.7292363857620932,
+      "grad_norm": 0.9084392786026001,
+      "learning_rate": 3.20666568204328e-05,
+      "loss": 0.1815,
+      "step": 2842
+    },
+    {
+      "epoch": 1.7298448433221782,
+      "grad_norm": 0.8204681873321533,
+      "learning_rate": 3.205471477657428e-05,
+      "loss": 0.1863,
+      "step": 2843
+    },
+    {
+      "epoch": 1.7304533008822633,
+      "grad_norm": 0.728208065032959,
+      "learning_rate": 3.204277098344055e-05,
+      "loss": 0.1384,
+      "step": 2844
+    },
+    {
+      "epoch": 1.7310617584423487,
+      "grad_norm": 0.9103240966796875,
+      "learning_rate": 3.203082544399318e-05,
+      "loss": 0.1875,
+      "step": 2845
+    },
+    {
+      "epoch": 1.731670216002434,
+      "grad_norm": 0.8469733595848083,
+      "learning_rate": 3.201887816119416e-05,
+      "loss": 0.1546,
+      "step": 2846
+    },
+    {
+      "epoch": 1.732278673562519,
+      "grad_norm": 1.0149333477020264,
+      "learning_rate": 3.2006929138005905e-05,
+      "loss": 0.1469,
+      "step": 2847
+    },
+    {
+      "epoch": 1.7328871311226042,
+      "grad_norm": 0.8808649778366089,
+      "learning_rate": 3.1994978377391295e-05,
+      "loss": 0.1811,
+      "step": 2848
+    },
+    {
+      "epoch": 1.7334955886826893,
+      "grad_norm": 0.77308589220047,
+      "learning_rate": 3.198302588231361e-05,
+      "loss": 0.1662,
+      "step": 2849
+    },
+    {
+      "epoch": 1.7341040462427746,
+      "grad_norm": 0.7700105309486389,
+      "learning_rate": 3.197107165573657e-05,
+      "loss": 0.1371,
+      "step": 2850
+    },
+    {
+      "epoch": 1.73471250380286,
+      "grad_norm": 0.8109223246574402,
+      "learning_rate": 3.195911570062434e-05,
+      "loss": 0.138,
+      "step": 2851
+    },
+    {
+      "epoch": 1.735320961362945,
+      "grad_norm": 0.8431537747383118,
+      "learning_rate": 3.1947158019941476e-05,
+      "loss": 0.1787,
+      "step": 2852
+    },
+    {
+      "epoch": 1.73592941892303,
+      "grad_norm": 0.8387619256973267,
+      "learning_rate": 3.1935198616652996e-05,
+      "loss": 0.1893,
+      "step": 2853
+    },
+    {
+      "epoch": 1.7365378764831152,
+      "grad_norm": 0.8504119515419006,
+      "learning_rate": 3.192323749372433e-05,
+      "loss": 0.1783,
+      "step": 2854
+    },
+    {
+      "epoch": 1.7371463340432005,
+      "grad_norm": 0.8203498125076294,
+      "learning_rate": 3.1911274654121345e-05,
+      "loss": 0.1651,
+      "step": 2855
+    },
+    {
+      "epoch": 1.7377547916032858,
+      "grad_norm": 0.941662073135376,
+      "learning_rate": 3.1899310100810326e-05,
+      "loss": 0.193,
+      "step": 2856
+    },
+    {
+      "epoch": 1.738363249163371,
+      "grad_norm": 0.8915706276893616,
+      "learning_rate": 3.1887343836757977e-05,
+      "loss": 0.2059,
+      "step": 2857
+    },
+    {
+      "epoch": 1.738971706723456,
+      "grad_norm": 0.7377119064331055,
+      "learning_rate": 3.1875375864931426e-05,
+      "loss": 0.1607,
+      "step": 2858
+    },
+    {
+      "epoch": 1.7395801642835411,
+      "grad_norm": 0.8279535174369812,
+      "learning_rate": 3.186340618829825e-05,
+      "loss": 0.1356,
+      "step": 2859
+    },
+    {
+      "epoch": 1.7401886218436264,
+      "grad_norm": 0.9348523020744324,
+      "learning_rate": 3.185143480982642e-05,
+      "loss": 0.2069,
+      "step": 2860
+    },
+    {
+      "epoch": 1.7407970794037118,
+      "grad_norm": 0.7679805159568787,
+      "learning_rate": 3.183946173248433e-05,
+      "loss": 0.1833,
+      "step": 2861
+    },
+    {
+      "epoch": 1.7414055369637969,
+      "grad_norm": 0.8387174010276794,
+      "learning_rate": 3.182748695924082e-05,
+      "loss": 0.2116,
+      "step": 2862
+    },
+    {
+      "epoch": 1.742013994523882,
+      "grad_norm": 0.8004626035690308,
+      "learning_rate": 3.181551049306513e-05,
+      "loss": 0.1897,
+      "step": 2863
+    },
+    {
+      "epoch": 1.742622452083967,
+      "grad_norm": 0.8716647624969482,
+      "learning_rate": 3.180353233692691e-05,
+      "loss": 0.2086,
+      "step": 2864
+    },
+    {
+      "epoch": 1.7432309096440524,
+      "grad_norm": 0.901866614818573,
+      "learning_rate": 3.179155249379628e-05,
+      "loss": 0.2046,
+      "step": 2865
+    },
+    {
+      "epoch": 1.7438393672041375,
+      "grad_norm": 0.7688228487968445,
+      "learning_rate": 3.17795709666437e-05,
+      "loss": 0.1335,
+      "step": 2866
+    },
+    {
+      "epoch": 1.7444478247642228,
+      "grad_norm": 0.7995738983154297,
+      "learning_rate": 3.1767587758440106e-05,
+      "loss": 0.1418,
+      "step": 2867
+    },
+    {
+      "epoch": 1.7450562823243079,
+      "grad_norm": 0.8365558981895447,
+      "learning_rate": 3.175560287215684e-05,
+      "loss": 0.1897,
+      "step": 2868
+    },
+    {
+      "epoch": 1.745664739884393,
+      "grad_norm": 0.8732959032058716,
+      "learning_rate": 3.1743616310765644e-05,
+      "loss": 0.1946,
+      "step": 2869
+    },
+    {
+      "epoch": 1.7462731974444783,
+      "grad_norm": 0.972126841545105,
+      "learning_rate": 3.1731628077238694e-05,
+      "loss": 0.1725,
+      "step": 2870
+    },
+    {
+      "epoch": 1.7468816550045634,
+      "grad_norm": 0.811004102230072,
+      "learning_rate": 3.171963817454857e-05,
+      "loss": 0.1305,
+      "step": 2871
+    },
+    {
+      "epoch": 1.7474901125646487,
+      "grad_norm": 0.9573332667350769,
+      "learning_rate": 3.170764660566826e-05,
+      "loss": 0.1681,
+      "step": 2872
+    },
+    {
+      "epoch": 1.7480985701247338,
+      "grad_norm": 0.7610151767730713,
+      "learning_rate": 3.1695653373571196e-05,
+      "loss": 0.154,
+      "step": 2873
+    },
+    {
+      "epoch": 1.748707027684819,
+      "grad_norm": 1.0020116567611694,
+      "learning_rate": 3.168365848123117e-05,
+      "loss": 0.171,
+      "step": 2874
+    },
+    {
+      "epoch": 1.7493154852449042,
+      "grad_norm": 0.8454048037528992,
+      "learning_rate": 3.167166193162244e-05,
+      "loss": 0.1776,
+      "step": 2875
+    },
+    {
+      "epoch": 1.7499239428049893,
+      "grad_norm": 0.8977103233337402,
+      "learning_rate": 3.1659663727719625e-05,
+      "loss": 0.1786,
+      "step": 2876
+    },
+    {
+      "epoch": 1.7505324003650746,
+      "grad_norm": 0.7743524312973022,
+      "learning_rate": 3.1647663872497804e-05,
+      "loss": 0.1425,
+      "step": 2877
+    },
+    {
+      "epoch": 1.7511408579251597,
+      "grad_norm": 0.9040097594261169,
+      "learning_rate": 3.1635662368932426e-05,
+      "loss": 0.2031,
+      "step": 2878
+    },
+    {
+      "epoch": 1.7517493154852448,
+      "grad_norm": 0.6998404264450073,
+      "learning_rate": 3.1623659219999374e-05,
+      "loss": 0.1344,
+      "step": 2879
+    },
+    {
+      "epoch": 1.7523577730453301,
+      "grad_norm": 0.8501226902008057,
+      "learning_rate": 3.161165442867492e-05,
+      "loss": 0.1767,
+      "step": 2880
+    },
+    {
+      "epoch": 1.7529662306054152,
+      "grad_norm": 0.9239878058433533,
+      "learning_rate": 3.159964799793575e-05,
+      "loss": 0.2296,
+      "step": 2881
+    },
+    {
+      "epoch": 1.7535746881655006,
+      "grad_norm": 0.9915949106216431,
+      "learning_rate": 3.158763993075897e-05,
+      "loss": 0.2277,
+      "step": 2882
+    },
+    {
+      "epoch": 1.7541831457255856,
+      "grad_norm": 0.746919572353363,
+      "learning_rate": 3.157563023012208e-05,
+      "loss": 0.1781,
+      "step": 2883
+    },
+    {
+      "epoch": 1.7547916032856707,
+      "grad_norm": 0.6546300649642944,
+      "learning_rate": 3.1563618899002965e-05,
+      "loss": 0.1176,
+      "step": 2884
+    },
+    {
+      "epoch": 1.755400060845756,
+      "grad_norm": 0.824098527431488,
+      "learning_rate": 3.1551605940379954e-05,
+      "loss": 0.1666,
+      "step": 2885
+    },
+    {
+      "epoch": 1.7560085184058412,
+      "grad_norm": 0.7335473299026489,
+      "learning_rate": 3.1539591357231755e-05,
+      "loss": 0.1489,
+      "step": 2886
+    },
+    {
+      "epoch": 1.7566169759659265,
+      "grad_norm": 0.8762836456298828,
+      "learning_rate": 3.152757515253748e-05,
+      "loss": 0.1802,
+      "step": 2887
+    },
+    {
+      "epoch": 1.7572254335260116,
+      "grad_norm": 0.9030479192733765,
+      "learning_rate": 3.1515557329276654e-05,
+      "loss": 0.1571,
+      "step": 2888
+    },
+    {
+      "epoch": 1.7578338910860967,
+      "grad_norm": 0.8152849674224854,
+      "learning_rate": 3.150353789042919e-05,
+      "loss": 0.1772,
+      "step": 2889
+    },
+    {
+      "epoch": 1.7584423486461818,
+      "grad_norm": 0.8419698476791382,
+      "learning_rate": 3.149151683897541e-05,
+      "loss": 0.1603,
+      "step": 2890
+    },
+    {
+      "epoch": 1.759050806206267,
+      "grad_norm": 1.0735416412353516,
+      "learning_rate": 3.147949417789604e-05,
+      "loss": 0.1659,
+      "step": 2891
+    },
+    {
+      "epoch": 1.7596592637663524,
+      "grad_norm": 0.8551287651062012,
+      "learning_rate": 3.1467469910172184e-05,
+      "loss": 0.199,
+      "step": 2892
+    },
+    {
+      "epoch": 1.7602677213264375,
+      "grad_norm": 1.121510624885559,
+      "learning_rate": 3.145544403878538e-05,
+      "loss": 0.1591,
+      "step": 2893
+    },
+    {
+      "epoch": 1.7608761788865226,
+      "grad_norm": 0.9003729224205017,
+      "learning_rate": 3.144341656671751e-05,
+      "loss": 0.1713,
+      "step": 2894
+    },
+    {
+      "epoch": 1.7614846364466077,
+      "grad_norm": 0.8984287977218628,
+      "learning_rate": 3.143138749695091e-05,
+      "loss": 0.2174,
+      "step": 2895
+    },
+    {
+      "epoch": 1.762093094006693,
+      "grad_norm": 0.8700388669967651,
+      "learning_rate": 3.14193568324683e-05,
+      "loss": 0.2276,
+      "step": 2896
+    },
+    {
+      "epoch": 1.7627015515667783,
+      "grad_norm": 0.7582150101661682,
+      "learning_rate": 3.140732457625276e-05,
+      "loss": 0.1604,
+      "step": 2897
+    },
+    {
+      "epoch": 1.7633100091268634,
+      "grad_norm": 0.8540982007980347,
+      "learning_rate": 3.13952907312878e-05,
+      "loss": 0.1728,
+      "step": 2898
+    },
+    {
+      "epoch": 1.7639184666869485,
+      "grad_norm": 0.8038197159767151,
+      "learning_rate": 3.1383255300557293e-05,
+      "loss": 0.171,
+      "step": 2899
+    },
+    {
+      "epoch": 1.7645269242470336,
+      "grad_norm": 0.8544119000434875,
+      "learning_rate": 3.137121828704555e-05,
+      "loss": 0.1521,
+      "step": 2900
+    },
+    {
+      "epoch": 1.765135381807119,
+      "grad_norm": 0.8337200284004211,
+      "learning_rate": 3.135917969373724e-05,
+      "loss": 0.1895,
+      "step": 2901
+    },
+    {
+      "epoch": 1.7657438393672042,
+      "grad_norm": 0.8394232988357544,
+      "learning_rate": 3.134713952361742e-05,
+      "loss": 0.2031,
+      "step": 2902
+    },
+    {
+      "epoch": 1.7663522969272893,
+      "grad_norm": 0.7780349850654602,
+      "learning_rate": 3.1335097779671564e-05,
+      "loss": 0.1559,
+      "step": 2903
+    },
+    {
+      "epoch": 1.7669607544873744,
+      "grad_norm": 0.7052262425422668,
+      "learning_rate": 3.132305446488552e-05,
+      "loss": 0.1905,
+      "step": 2904
+    },
+    {
+      "epoch": 1.7675692120474595,
+      "grad_norm": 0.7838650941848755,
+      "learning_rate": 3.1311009582245525e-05,
+      "loss": 0.1404,
+      "step": 2905
+    },
+    {
+      "epoch": 1.7681776696075449,
+      "grad_norm": 0.9082717299461365,
+      "learning_rate": 3.1298963134738214e-05,
+      "loss": 0.2033,
+      "step": 2906
+    },
+    {
+      "epoch": 1.7687861271676302,
+      "grad_norm": 0.882872462272644,
+      "learning_rate": 3.128691512535059e-05,
+      "loss": 0.186,
+      "step": 2907
+    },
+    {
+      "epoch": 1.7693945847277153,
+      "grad_norm": 1.0027045011520386,
+      "learning_rate": 3.127486555707007e-05,
+      "loss": 0.1539,
+      "step": 2908
+    },
+    {
+      "epoch": 1.7700030422878004,
+      "grad_norm": 0.9138793349266052,
+      "learning_rate": 3.126281443288445e-05,
+      "loss": 0.1769,
+      "step": 2909
+    },
+    {
+      "epoch": 1.7706114998478855,
+      "grad_norm": 0.8712251782417297,
+      "learning_rate": 3.125076175578189e-05,
+      "loss": 0.1787,
+      "step": 2910
+    },
+    {
+      "epoch": 1.7712199574079708,
+      "grad_norm": 1.6176162958145142,
+      "learning_rate": 3.123870752875096e-05,
+      "loss": 0.174,
+      "step": 2911
+    },
+    {
+      "epoch": 1.771828414968056,
+      "grad_norm": 0.8051137924194336,
+      "learning_rate": 3.12266517547806e-05,
+      "loss": 0.177,
+      "step": 2912
+    },
+    {
+      "epoch": 1.7724368725281412,
+      "grad_norm": 0.8419429063796997,
+      "learning_rate": 3.121459443686015e-05,
+      "loss": 0.1795,
+      "step": 2913
+    },
+    {
+      "epoch": 1.7730453300882263,
+      "grad_norm": 0.8457537889480591,
+      "learning_rate": 3.120253557797932e-05,
+      "loss": 0.1616,
+      "step": 2914
+    },
+    {
+      "epoch": 1.7736537876483114,
+      "grad_norm": 0.7322502732276917,
+      "learning_rate": 3.1190475181128194e-05,
+      "loss": 0.1396,
+      "step": 2915
+    },
+    {
+      "epoch": 1.7742622452083967,
+      "grad_norm": 0.755520761013031,
+      "learning_rate": 3.1178413249297255e-05,
+      "loss": 0.1479,
+      "step": 2916
+    },
+    {
+      "epoch": 1.774870702768482,
+      "grad_norm": 0.8889976739883423,
+      "learning_rate": 3.116634978547737e-05,
+      "loss": 0.2115,
+      "step": 2917
+    },
+    {
+      "epoch": 1.7754791603285671,
+      "grad_norm": 0.8492110967636108,
+      "learning_rate": 3.115428479265975e-05,
+      "loss": 0.182,
+      "step": 2918
+    },
+    {
+      "epoch": 1.7760876178886522,
+      "grad_norm": 0.7823325395584106,
+      "learning_rate": 3.1142218273836025e-05,
+      "loss": 0.1563,
+      "step": 2919
+    },
+    {
+      "epoch": 1.7766960754487373,
+      "grad_norm": 0.774039626121521,
+      "learning_rate": 3.11301502319982e-05,
+      "loss": 0.1498,
+      "step": 2920
+    },
+    {
+      "epoch": 1.7773045330088226,
+      "grad_norm": 0.8219572901725769,
+      "learning_rate": 3.111808067013863e-05,
+      "loss": 0.1536,
+      "step": 2921
+    },
+    {
+      "epoch": 1.777912990568908,
+      "grad_norm": 0.7709328532218933,
+      "learning_rate": 3.1106009591250066e-05,
+      "loss": 0.135,
+      "step": 2922
+    },
+    {
+      "epoch": 1.778521448128993,
+      "grad_norm": 0.7555780410766602,
+      "learning_rate": 3.109393699832564e-05,
+      "loss": 0.1303,
+      "step": 2923
+    },
+    {
+      "epoch": 1.7791299056890781,
+      "grad_norm": 0.58537757396698,
+      "learning_rate": 3.108186289435884e-05,
+      "loss": 0.1188,
+      "step": 2924
+    },
+    {
+      "epoch": 1.7797383632491632,
+      "grad_norm": 0.8996820449829102,
+      "learning_rate": 3.106978728234354e-05,
+      "loss": 0.1952,
+      "step": 2925
+    },
+    {
+      "epoch": 1.7803468208092486,
+      "grad_norm": 1.680888056755066,
+      "learning_rate": 3.1057710165274004e-05,
+      "loss": 0.2346,
+      "step": 2926
+    },
+    {
+      "epoch": 1.7809552783693339,
+      "grad_norm": 0.6698862314224243,
+      "learning_rate": 3.1045631546144846e-05,
+      "loss": 0.1375,
+      "step": 2927
+    },
+    {
+      "epoch": 1.781563735929419,
+      "grad_norm": 0.7749894261360168,
+      "learning_rate": 3.1033551427951064e-05,
+      "loss": 0.1948,
+      "step": 2928
+    },
+    {
+      "epoch": 1.782172193489504,
+      "grad_norm": 0.8020423054695129,
+      "learning_rate": 3.102146981368801e-05,
+      "loss": 0.1814,
+      "step": 2929
+    },
+    {
+      "epoch": 1.7827806510495892,
+      "grad_norm": 0.8180800676345825,
+      "learning_rate": 3.100938670635143e-05,
+      "loss": 0.1584,
+      "step": 2930
+    },
+    {
+      "epoch": 1.7833891086096745,
+      "grad_norm": 1.1050705909729004,
+      "learning_rate": 3.099730210893743e-05,
+      "loss": 0.1605,
+      "step": 2931
+    },
+    {
+      "epoch": 1.7839975661697598,
+      "grad_norm": 0.8110726475715637,
+      "learning_rate": 3.0985216024442484e-05,
+      "loss": 0.1342,
+      "step": 2932
+    },
+    {
+      "epoch": 1.784606023729845,
+      "grad_norm": 0.7474027872085571,
+      "learning_rate": 3.097312845586345e-05,
+      "loss": 0.1579,
+      "step": 2933
+    },
+    {
+      "epoch": 1.78521448128993,
+      "grad_norm": 0.908333957195282,
+      "learning_rate": 3.096103940619752e-05,
+      "loss": 0.2047,
+      "step": 2934
+    },
+    {
+      "epoch": 1.785822938850015,
+      "grad_norm": 1.506729006767273,
+      "learning_rate": 3.0948948878442293e-05,
+      "loss": 0.189,
+      "step": 2935
+    },
+    {
+      "epoch": 1.7864313964101004,
+      "grad_norm": 0.7309430837631226,
+      "learning_rate": 3.093685687559571e-05,
+      "loss": 0.1383,
+      "step": 2936
+    },
+    {
+      "epoch": 1.7870398539701857,
+      "grad_norm": 0.866033136844635,
+      "learning_rate": 3.092476340065608e-05,
+      "loss": 0.1515,
+      "step": 2937
+    },
+    {
+      "epoch": 1.7876483115302708,
+      "grad_norm": 0.8974893689155579,
+      "learning_rate": 3.091266845662208e-05,
+      "loss": 0.1721,
+      "step": 2938
+    },
+    {
+      "epoch": 1.788256769090356,
+      "grad_norm": 0.8083369135856628,
+      "learning_rate": 3.090057204649276e-05,
+      "loss": 0.1714,
+      "step": 2939
+    },
+    {
+      "epoch": 1.788865226650441,
+      "grad_norm": 1.0619298219680786,
+      "learning_rate": 3.088847417326752e-05,
+      "loss": 0.172,
+      "step": 2940
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.9113779067993164,
+      "learning_rate": 3.087637483994612e-05,
+      "loss": 0.1679,
+      "step": 2941
+    },
+    {
+      "epoch": 1.7900821417706116,
+      "grad_norm": 0.8694355487823486,
+      "learning_rate": 3.086427404952871e-05,
+      "loss": 0.1498,
+      "step": 2942
+    },
+    {
+      "epoch": 1.7906905993306967,
+      "grad_norm": 0.789814829826355,
+      "learning_rate": 3.085217180501576e-05,
+      "loss": 0.1598,
+      "step": 2943
+    },
+    {
+      "epoch": 1.7912990568907818,
+      "grad_norm": 0.7697128653526306,
+      "learning_rate": 3.084006810940814e-05,
+      "loss": 0.1259,
+      "step": 2944
+    },
+    {
+      "epoch": 1.791907514450867,
+      "grad_norm": 0.8780056834220886,
+      "learning_rate": 3.082796296570706e-05,
+      "loss": 0.1966,
+      "step": 2945
+    },
+    {
+      "epoch": 1.7925159720109523,
+      "grad_norm": 0.790896475315094,
+      "learning_rate": 3.081585637691407e-05,
+      "loss": 0.1333,
+      "step": 2946
+    },
+    {
+      "epoch": 1.7931244295710376,
+      "grad_norm": 0.8985186815261841,
+      "learning_rate": 3.080374834603113e-05,
+      "loss": 0.127,
+      "step": 2947
+    },
+    {
+      "epoch": 1.7937328871311227,
+      "grad_norm": 0.9989727735519409,
+      "learning_rate": 3.079163887606051e-05,
+      "loss": 0.1706,
+      "step": 2948
+    },
+    {
+      "epoch": 1.7943413446912078,
+      "grad_norm": 0.7312899827957153,
+      "learning_rate": 3.077952797000485e-05,
+      "loss": 0.1619,
+      "step": 2949
+    },
+    {
+      "epoch": 1.7949498022512929,
+      "grad_norm": 0.9777790307998657,
+      "learning_rate": 3.0767415630867165e-05,
+      "loss": 0.178,
+      "step": 2950
+    },
+    {
+      "epoch": 1.7955582598113782,
+      "grad_norm": 1.0288338661193848,
+      "learning_rate": 3.0755301861650794e-05,
+      "loss": 0.1798,
+      "step": 2951
+    },
+    {
+      "epoch": 1.7961667173714635,
+      "grad_norm": 0.7798663973808289,
+      "learning_rate": 3.074318666535946e-05,
+      "loss": 0.1747,
+      "step": 2952
+    },
+    {
+      "epoch": 1.7967751749315486,
+      "grad_norm": 0.8773430585861206,
+      "learning_rate": 3.0731070044997215e-05,
+      "loss": 0.1644,
+      "step": 2953
+    },
+    {
+      "epoch": 1.7973836324916337,
+      "grad_norm": 0.828058123588562,
+      "learning_rate": 3.071895200356848e-05,
+      "loss": 0.163,
+      "step": 2954
+    },
+    {
+      "epoch": 1.7979920900517188,
+      "grad_norm": 0.8165954947471619,
+      "learning_rate": 3.070683254407803e-05,
+      "loss": 0.1583,
+      "step": 2955
+    },
+    {
+      "epoch": 1.798600547611804,
+      "grad_norm": 0.6299397349357605,
+      "learning_rate": 3.069471166953098e-05,
+      "loss": 0.1258,
+      "step": 2956
+    },
+    {
+      "epoch": 1.7992090051718894,
+      "grad_norm": 0.7792373895645142,
+      "learning_rate": 3.068258938293281e-05,
+      "loss": 0.1431,
+      "step": 2957
+    },
+    {
+      "epoch": 1.7998174627319745,
+      "grad_norm": 0.6819939613342285,
+      "learning_rate": 3.0670465687289325e-05,
+      "loss": 0.1071,
+      "step": 2958
+    },
+    {
+      "epoch": 1.8004259202920596,
+      "grad_norm": 1.1301064491271973,
+      "learning_rate": 3.065834058560671e-05,
+      "loss": 0.1438,
+      "step": 2959
+    },
+    {
+      "epoch": 1.8010343778521447,
+      "grad_norm": 1.0041139125823975,
+      "learning_rate": 3.064621408089148e-05,
+      "loss": 0.1515,
+      "step": 2960
+    },
+    {
+      "epoch": 1.80164283541223,
+      "grad_norm": 0.7390562295913696,
+      "learning_rate": 3.0634086176150504e-05,
+      "loss": 0.1325,
+      "step": 2961
+    },
+    {
+      "epoch": 1.8022512929723153,
+      "grad_norm": 0.8052049279212952,
+      "learning_rate": 3.0621956874391e-05,
+      "loss": 0.1349,
+      "step": 2962
+    },
+    {
+      "epoch": 1.8028597505324004,
+      "grad_norm": 0.7416874170303345,
+      "learning_rate": 3.060982617862053e-05,
+      "loss": 0.1505,
+      "step": 2963
+    },
+    {
+      "epoch": 1.8034682080924855,
+      "grad_norm": 1.035781979560852,
+      "learning_rate": 3.0597694091846985e-05,
+      "loss": 0.1632,
+      "step": 2964
+    },
+    {
+      "epoch": 1.8040766656525706,
+      "grad_norm": 1.187511682510376,
+      "learning_rate": 3.058556061707863e-05,
+      "loss": 0.26,
+      "step": 2965
+    },
+    {
+      "epoch": 1.804685123212656,
+      "grad_norm": 0.9462476372718811,
+      "learning_rate": 3.057342575732406e-05,
+      "loss": 0.1846,
+      "step": 2966
+    },
+    {
+      "epoch": 1.805293580772741,
+      "grad_norm": 0.9190717339515686,
+      "learning_rate": 3.0561289515592226e-05,
+      "loss": 0.1966,
+      "step": 2967
+    },
+    {
+      "epoch": 1.8059020383328264,
+      "grad_norm": 0.8364342451095581,
+      "learning_rate": 3.054915189489239e-05,
+      "loss": 0.1795,
+      "step": 2968
+    },
+    {
+      "epoch": 1.8065104958929115,
+      "grad_norm": 0.7401247620582581,
+      "learning_rate": 3.053701289823418e-05,
+      "loss": 0.1576,
+      "step": 2969
+    },
+    {
+      "epoch": 1.8071189534529966,
+      "grad_norm": 0.7658497095108032,
+      "learning_rate": 3.052487252862758e-05,
+      "loss": 0.1873,
+      "step": 2970
+    },
+    {
+      "epoch": 1.8077274110130819,
+      "grad_norm": 0.765739381313324,
+      "learning_rate": 3.0512730789082862e-05,
+      "loss": 0.1662,
+      "step": 2971
+    },
+    {
+      "epoch": 1.808335868573167,
+      "grad_norm": 0.7397434711456299,
+      "learning_rate": 3.0500587682610694e-05,
+      "loss": 0.154,
+      "step": 2972
+    },
+    {
+      "epoch": 1.8089443261332523,
+      "grad_norm": 0.8113225102424622,
+      "learning_rate": 3.0488443212222067e-05,
+      "loss": 0.1421,
+      "step": 2973
+    },
+    {
+      "epoch": 1.8095527836933374,
+      "grad_norm": 0.9399985074996948,
+      "learning_rate": 3.047629738092828e-05,
+      "loss": 0.1484,
+      "step": 2974
+    },
+    {
+      "epoch": 1.8101612412534225,
+      "grad_norm": 0.8929994106292725,
+      "learning_rate": 3.046415019174102e-05,
+      "loss": 0.1615,
+      "step": 2975
+    },
+    {
+      "epoch": 1.8107696988135078,
+      "grad_norm": 0.9099370837211609,
+      "learning_rate": 3.0452001647672256e-05,
+      "loss": 0.1902,
+      "step": 2976
+    },
+    {
+      "epoch": 1.811378156373593,
+      "grad_norm": 0.7755023837089539,
+      "learning_rate": 3.043985175173434e-05,
+      "loss": 0.1494,
+      "step": 2977
+    },
+    {
+      "epoch": 1.8119866139336782,
+      "grad_norm": 0.868893027305603,
+      "learning_rate": 3.042770050693994e-05,
+      "loss": 0.1345,
+      "step": 2978
+    },
+    {
+      "epoch": 1.8125950714937633,
+      "grad_norm": 0.9862670302391052,
+      "learning_rate": 3.0415547916302044e-05,
+      "loss": 0.1903,
+      "step": 2979
+    },
+    {
+      "epoch": 1.8132035290538484,
+      "grad_norm": 0.7265908718109131,
+      "learning_rate": 3.0403393982834e-05,
+      "loss": 0.1474,
+      "step": 2980
+    },
+    {
+      "epoch": 1.8138119866139337,
+      "grad_norm": 0.7525795698165894,
+      "learning_rate": 3.039123870954947e-05,
+      "loss": 0.1376,
+      "step": 2981
+    },
+    {
+      "epoch": 1.8144204441740188,
+      "grad_norm": 0.8074904680252075,
+      "learning_rate": 3.0379082099462454e-05,
+      "loss": 0.1699,
+      "step": 2982
+    },
+    {
+      "epoch": 1.8150289017341041,
+      "grad_norm": 0.7635090351104736,
+      "learning_rate": 3.0366924155587296e-05,
+      "loss": 0.1865,
+      "step": 2983
+    },
+    {
+      "epoch": 1.8156373592941892,
+      "grad_norm": 0.8797259330749512,
+      "learning_rate": 3.0354764880938647e-05,
+      "loss": 0.1832,
+      "step": 2984
+    },
+    {
+      "epoch": 1.8162458168542743,
+      "grad_norm": 0.7532181739807129,
+      "learning_rate": 3.0342604278531512e-05,
+      "loss": 0.1289,
+      "step": 2985
+    },
+    {
+      "epoch": 1.8168542744143596,
+      "grad_norm": 1.013737440109253,
+      "learning_rate": 3.0330442351381198e-05,
+      "loss": 0.1913,
+      "step": 2986
+    },
+    {
+      "epoch": 1.8174627319744447,
+      "grad_norm": 0.8055927753448486,
+      "learning_rate": 3.0318279102503367e-05,
+      "loss": 0.1773,
+      "step": 2987
+    },
+    {
+      "epoch": 1.81807118953453,
+      "grad_norm": 0.8605936765670776,
+      "learning_rate": 3.0306114534913998e-05,
+      "loss": 0.1499,
+      "step": 2988
+    },
+    {
+      "epoch": 1.8186796470946152,
+      "grad_norm": 0.8703776001930237,
+      "learning_rate": 3.0293948651629388e-05,
+      "loss": 0.2272,
+      "step": 2989
+    },
+    {
+      "epoch": 1.8192881046547003,
+      "grad_norm": 0.7393494844436646,
+      "learning_rate": 3.0281781455666182e-05,
+      "loss": 0.1363,
+      "step": 2990
+    },
+    {
+      "epoch": 1.8198965622147856,
+      "grad_norm": 0.9417717456817627,
+      "learning_rate": 3.026961295004133e-05,
+      "loss": 0.1626,
+      "step": 2991
+    },
+    {
+      "epoch": 1.8205050197748707,
+      "grad_norm": 0.72954922914505,
+      "learning_rate": 3.025744313777211e-05,
+      "loss": 0.1273,
+      "step": 2992
+    },
+    {
+      "epoch": 1.821113477334956,
+      "grad_norm": 0.8940945863723755,
+      "learning_rate": 3.0245272021876144e-05,
+      "loss": 0.1814,
+      "step": 2993
+    },
+    {
+      "epoch": 1.821721934895041,
+      "grad_norm": 0.8653423190116882,
+      "learning_rate": 3.023309960537134e-05,
+      "loss": 0.1743,
+      "step": 2994
+    },
+    {
+      "epoch": 1.8223303924551262,
+      "grad_norm": 0.9504120945930481,
+      "learning_rate": 3.0220925891275957e-05,
+      "loss": 0.18,
+      "step": 2995
+    },
+    {
+      "epoch": 1.8229388500152113,
+      "grad_norm": 0.9393471479415894,
+      "learning_rate": 3.0208750882608583e-05,
+      "loss": 0.1825,
+      "step": 2996
+    },
+    {
+      "epoch": 1.8235473075752966,
+      "grad_norm": 0.9305369257926941,
+      "learning_rate": 3.0196574582388095e-05,
+      "loss": 0.1618,
+      "step": 2997
+    },
+    {
+      "epoch": 1.824155765135382,
+      "grad_norm": 0.7652149796485901,
+      "learning_rate": 3.0184396993633718e-05,
+      "loss": 0.1558,
+      "step": 2998
+    },
+    {
+      "epoch": 1.824764222695467,
+      "grad_norm": 0.9691605567932129,
+      "learning_rate": 3.0172218119364975e-05,
+      "loss": 0.1435,
+      "step": 2999
+    },
+    {
+      "epoch": 1.825372680255552,
+      "grad_norm": 1.8901022672653198,
+      "learning_rate": 3.0160037962601727e-05,
+      "loss": 0.1259,
+      "step": 3000
+    },
+    {
+      "epoch": 1.8259811378156372,
+      "grad_norm": 0.7600100636482239,
+      "learning_rate": 3.014785652636416e-05,
+      "loss": 0.1442,
+      "step": 3001
+    },
+    {
+      "epoch": 1.8265895953757225,
+      "grad_norm": 0.904800295829773,
+      "learning_rate": 3.0135673813672734e-05,
+      "loss": 0.1831,
+      "step": 3002
+    },
+    {
+      "epoch": 1.8271980529358078,
+      "grad_norm": 0.9065991640090942,
+      "learning_rate": 3.0123489827548273e-05,
+      "loss": 0.1632,
+      "step": 3003
+    },
+    {
+      "epoch": 1.827806510495893,
+      "grad_norm": 0.7399230599403381,
+      "learning_rate": 3.0111304571011888e-05,
+      "loss": 0.1456,
+      "step": 3004
+    },
+    {
+      "epoch": 1.828414968055978,
+      "grad_norm": 0.6830241680145264,
+      "learning_rate": 3.0099118047085024e-05,
+      "loss": 0.1322,
+      "step": 3005
+    },
+    {
+      "epoch": 1.8290234256160631,
+      "grad_norm": 0.8357345461845398,
+      "learning_rate": 3.0086930258789426e-05,
+      "loss": 0.1839,
+      "step": 3006
+    },
+    {
+      "epoch": 1.8296318831761484,
+      "grad_norm": 0.8393163681030273,
+      "learning_rate": 3.0074741209147157e-05,
+      "loss": 0.1388,
+      "step": 3007
+    },
+    {
+      "epoch": 1.8302403407362338,
+      "grad_norm": 0.7641516923904419,
+      "learning_rate": 3.006255090118059e-05,
+      "loss": 0.134,
+      "step": 3008
+    },
+    {
+      "epoch": 1.8308487982963189,
+      "grad_norm": 0.8829279541969299,
+      "learning_rate": 3.005035933791242e-05,
+      "loss": 0.165,
+      "step": 3009
+    },
+    {
+      "epoch": 1.831457255856404,
+      "grad_norm": 0.7792665958404541,
+      "learning_rate": 3.0038166522365642e-05,
+      "loss": 0.1351,
+      "step": 3010
+    },
+    {
+      "epoch": 1.832065713416489,
+      "grad_norm": 0.7733790278434753,
+      "learning_rate": 3.0025972457563573e-05,
+      "loss": 0.1098,
+      "step": 3011
+    },
+    {
+      "epoch": 1.8326741709765744,
+      "grad_norm": 0.7731171250343323,
+      "learning_rate": 3.001377714652982e-05,
+      "loss": 0.1458,
+      "step": 3012
+    },
+    {
+      "epoch": 1.8332826285366597,
+      "grad_norm": 0.7515898942947388,
+      "learning_rate": 3.000158059228832e-05,
+      "loss": 0.1682,
+      "step": 3013
+    },
+    {
+      "epoch": 1.8338910860967448,
+      "grad_norm": 0.7409710884094238,
+      "learning_rate": 2.9989382797863313e-05,
+      "loss": 0.1413,
+      "step": 3014
+    },
+    {
+      "epoch": 1.8344995436568299,
+      "grad_norm": 0.8755896687507629,
+      "learning_rate": 2.997718376627934e-05,
+      "loss": 0.1678,
+      "step": 3015
+    },
+    {
+      "epoch": 1.835108001216915,
+      "grad_norm": 0.7442592978477478,
+      "learning_rate": 2.996498350056125e-05,
+      "loss": 0.1285,
+      "step": 3016
+    },
+    {
+      "epoch": 1.8357164587770003,
+      "grad_norm": 0.7244062423706055,
+      "learning_rate": 2.9952782003734202e-05,
+      "loss": 0.1296,
+      "step": 3017
+    },
+    {
+      "epoch": 1.8363249163370856,
+      "grad_norm": 0.7951245903968811,
+      "learning_rate": 2.994057927882366e-05,
+      "loss": 0.1638,
+      "step": 3018
+    },
+    {
+      "epoch": 1.8369333738971707,
+      "grad_norm": 0.778948962688446,
+      "learning_rate": 2.9928375328855396e-05,
+      "loss": 0.1594,
+      "step": 3019
+    },
+    {
+      "epoch": 1.8375418314572558,
+      "grad_norm": 1.0201570987701416,
+      "learning_rate": 2.9916170156855467e-05,
+      "loss": 0.1291,
+      "step": 3020
+    },
+    {
+      "epoch": 1.838150289017341,
+      "grad_norm": 0.8786885738372803,
+      "learning_rate": 2.9903963765850263e-05,
+      "loss": 0.1835,
+      "step": 3021
+    },
+    {
+      "epoch": 1.8387587465774262,
+      "grad_norm": 0.8748721480369568,
+      "learning_rate": 2.989175615886644e-05,
+      "loss": 0.1542,
+      "step": 3022
+    },
+    {
+      "epoch": 1.8393672041375115,
+      "grad_norm": 0.7832545042037964,
+      "learning_rate": 2.9879547338930997e-05,
+      "loss": 0.1622,
+      "step": 3023
+    },
+    {
+      "epoch": 1.8399756616975966,
+      "grad_norm": 0.902432382106781,
+      "learning_rate": 2.98673373090712e-05,
+      "loss": 0.1641,
+      "step": 3024
+    },
+    {
+      "epoch": 1.8405841192576817,
+      "grad_norm": 0.9937904477119446,
+      "learning_rate": 2.9855126072314638e-05,
+      "loss": 0.1468,
+      "step": 3025
+    },
+    {
+      "epoch": 1.8411925768177668,
+      "grad_norm": 0.8179795145988464,
+      "learning_rate": 2.984291363168918e-05,
+      "loss": 0.1675,
+      "step": 3026
+    },
+    {
+      "epoch": 1.8418010343778521,
+      "grad_norm": 0.7362605929374695,
+      "learning_rate": 2.9830699990222992e-05,
+      "loss": 0.1494,
+      "step": 3027
+    },
+    {
+      "epoch": 1.8424094919379375,
+      "grad_norm": 0.7124605178833008,
+      "learning_rate": 2.981848515094457e-05,
+      "loss": 0.117,
+      "step": 3028
+    },
+    {
+      "epoch": 1.8430179494980226,
+      "grad_norm": 0.9013100266456604,
+      "learning_rate": 2.9806269116882678e-05,
+      "loss": 0.1282,
+      "step": 3029
+    },
+    {
+      "epoch": 1.8436264070581077,
+      "grad_norm": 0.7831732630729675,
+      "learning_rate": 2.979405189106637e-05,
+      "loss": 0.1539,
+      "step": 3030
+    },
+    {
+      "epoch": 1.8442348646181927,
+      "grad_norm": 0.8050282597541809,
+      "learning_rate": 2.9781833476525022e-05,
+      "loss": 0.1714,
+      "step": 3031
+    },
+    {
+      "epoch": 1.844843322178278,
+      "grad_norm": 0.7580960988998413,
+      "learning_rate": 2.97696138762883e-05,
+      "loss": 0.133,
+      "step": 3032
+    },
+    {
+      "epoch": 1.8454517797383634,
+      "grad_norm": 0.5984554290771484,
+      "learning_rate": 2.9757393093386133e-05,
+      "loss": 0.1178,
+      "step": 3033
+    },
+    {
+      "epoch": 1.8460602372984485,
+      "grad_norm": 0.9737577438354492,
+      "learning_rate": 2.974517113084878e-05,
+      "loss": 0.2278,
+      "step": 3034
+    },
+    {
+      "epoch": 1.8466686948585336,
+      "grad_norm": 0.7336621284484863,
+      "learning_rate": 2.973294799170677e-05,
+      "loss": 0.1532,
+      "step": 3035
+    },
+    {
+      "epoch": 1.8472771524186187,
+      "grad_norm": 0.9183807373046875,
+      "learning_rate": 2.9720723678990943e-05,
+      "loss": 0.1445,
+      "step": 3036
+    },
+    {
+      "epoch": 1.847885609978704,
+      "grad_norm": 0.8194828629493713,
+      "learning_rate": 2.970849819573241e-05,
+      "loss": 0.1511,
+      "step": 3037
+    },
+    {
+      "epoch": 1.8484940675387893,
+      "grad_norm": 0.9400119185447693,
+      "learning_rate": 2.9696271544962583e-05,
+      "loss": 0.1763,
+      "step": 3038
+    },
+    {
+      "epoch": 1.8491025250988744,
+      "grad_norm": 0.9202710390090942,
+      "learning_rate": 2.968404372971316e-05,
+      "loss": 0.1323,
+      "step": 3039
+    },
+    {
+      "epoch": 1.8497109826589595,
+      "grad_norm": 1.0977171659469604,
+      "learning_rate": 2.9671814753016147e-05,
+      "loss": 0.1417,
+      "step": 3040
+    },
+    {
+      "epoch": 1.8503194402190446,
+      "grad_norm": 0.8560418486595154,
+      "learning_rate": 2.9659584617903795e-05,
+      "loss": 0.16,
+      "step": 3041
+    },
+    {
+      "epoch": 1.85092789777913,
+      "grad_norm": 0.9262466430664062,
+      "learning_rate": 2.9647353327408678e-05,
+      "loss": 0.1722,
+      "step": 3042
+    },
+    {
+      "epoch": 1.8515363553392152,
+      "grad_norm": 0.9393658638000488,
+      "learning_rate": 2.9635120884563654e-05,
+      "loss": 0.1889,
+      "step": 3043
+    },
+    {
+      "epoch": 1.8521448128993003,
+      "grad_norm": 0.6739933490753174,
+      "learning_rate": 2.9622887292401847e-05,
+      "loss": 0.1257,
+      "step": 3044
+    },
+    {
+      "epoch": 1.8527532704593854,
+      "grad_norm": 0.8402737975120544,
+      "learning_rate": 2.9610652553956688e-05,
+      "loss": 0.1763,
+      "step": 3045
+    },
+    {
+      "epoch": 1.8533617280194705,
+      "grad_norm": 0.7270654439926147,
+      "learning_rate": 2.959841667226187e-05,
+      "loss": 0.1321,
+      "step": 3046
+    },
+    {
+      "epoch": 1.8539701855795558,
+      "grad_norm": 0.8025201559066772,
+      "learning_rate": 2.9586179650351386e-05,
+      "loss": 0.1524,
+      "step": 3047
+    },
+    {
+      "epoch": 1.8545786431396412,
+      "grad_norm": 0.7017185091972351,
+      "learning_rate": 2.9573941491259523e-05,
+      "loss": 0.1274,
+      "step": 3048
+    },
+    {
+      "epoch": 1.8551871006997263,
+      "grad_norm": 0.7935231328010559,
+      "learning_rate": 2.9561702198020813e-05,
+      "loss": 0.1738,
+      "step": 3049
+    },
+    {
+      "epoch": 1.8557955582598114,
+      "grad_norm": 0.830694317817688,
+      "learning_rate": 2.9549461773670094e-05,
+      "loss": 0.1483,
+      "step": 3050
+    },
+    {
+      "epoch": 1.8564040158198964,
+      "grad_norm": 0.8252891302108765,
+      "learning_rate": 2.9537220221242496e-05,
+      "loss": 0.2033,
+      "step": 3051
+    },
+    {
+      "epoch": 1.8570124733799818,
+      "grad_norm": 1.0744190216064453,
+      "learning_rate": 2.9524977543773397e-05,
+      "loss": 0.1949,
+      "step": 3052
+    },
+    {
+      "epoch": 1.857620930940067,
+      "grad_norm": 0.7989039421081543,
+      "learning_rate": 2.9512733744298482e-05,
+      "loss": 0.1604,
+      "step": 3053
+    },
+    {
+      "epoch": 1.8582293885001522,
+      "grad_norm": 0.768959105014801,
+      "learning_rate": 2.9500488825853702e-05,
+      "loss": 0.1412,
+      "step": 3054
+    },
+    {
+      "epoch": 1.8588378460602373,
+      "grad_norm": 0.8371010422706604,
+      "learning_rate": 2.9488242791475272e-05,
+      "loss": 0.1422,
+      "step": 3055
+    },
+    {
+      "epoch": 1.8594463036203224,
+      "grad_norm": 0.8141250610351562,
+      "learning_rate": 2.947599564419971e-05,
+      "loss": 0.2141,
+      "step": 3056
+    },
+    {
+      "epoch": 1.8600547611804077,
+      "grad_norm": 0.7183147072792053,
+      "learning_rate": 2.9463747387063807e-05,
+      "loss": 0.1782,
+      "step": 3057
+    },
+    {
+      "epoch": 1.860663218740493,
+      "grad_norm": 0.7062293887138367,
+      "learning_rate": 2.94514980231046e-05,
+      "loss": 0.1377,
+      "step": 3058
+    },
+    {
+      "epoch": 1.861271676300578,
+      "grad_norm": 0.7800901532173157,
+      "learning_rate": 2.943924755535944e-05,
+      "loss": 0.16,
+      "step": 3059
+    },
+    {
+      "epoch": 1.8618801338606632,
+      "grad_norm": 0.8286892175674438,
+      "learning_rate": 2.9426995986865918e-05,
+      "loss": 0.1543,
+      "step": 3060
+    },
+    {
+      "epoch": 1.8624885914207483,
+      "grad_norm": 0.7443828582763672,
+      "learning_rate": 2.941474332066192e-05,
+      "loss": 0.1543,
+      "step": 3061
+    },
+    {
+      "epoch": 1.8630970489808336,
+      "grad_norm": 0.6537884473800659,
+      "learning_rate": 2.9402489559785594e-05,
+      "loss": 0.1595,
+      "step": 3062
+    },
+    {
+      "epoch": 1.863705506540919,
+      "grad_norm": 0.7740469574928284,
+      "learning_rate": 2.9390234707275355e-05,
+      "loss": 0.1529,
+      "step": 3063
+    },
+    {
+      "epoch": 1.864313964101004,
+      "grad_norm": 0.6976113319396973,
+      "learning_rate": 2.9377978766169912e-05,
+      "loss": 0.1339,
+      "step": 3064
+    },
+    {
+      "epoch": 1.8649224216610891,
+      "grad_norm": 0.7111392021179199,
+      "learning_rate": 2.9365721739508213e-05,
+      "loss": 0.142,
+      "step": 3065
+    },
+    {
+      "epoch": 1.8655308792211742,
+      "grad_norm": 0.671752393245697,
+      "learning_rate": 2.9353463630329493e-05,
+      "loss": 0.1487,
+      "step": 3066
+    },
+    {
+      "epoch": 1.8661393367812595,
+      "grad_norm": 0.8591635823249817,
+      "learning_rate": 2.9341204441673266e-05,
+      "loss": 0.1523,
+      "step": 3067
+    },
+    {
+      "epoch": 1.8667477943413449,
+      "grad_norm": 1.018031358718872,
+      "learning_rate": 2.932894417657927e-05,
+      "loss": 0.1422,
+      "step": 3068
+    },
+    {
+      "epoch": 1.86735625190143,
+      "grad_norm": 1.00792396068573,
+      "learning_rate": 2.9316682838087565e-05,
+      "loss": 0.1496,
+      "step": 3069
+    },
+    {
+      "epoch": 1.867964709461515,
+      "grad_norm": 0.8856237530708313,
+      "learning_rate": 2.930442042923845e-05,
+      "loss": 0.1572,
+      "step": 3070
+    },
+    {
+      "epoch": 1.8685731670216001,
+      "grad_norm": 0.9289006590843201,
+      "learning_rate": 2.929215695307248e-05,
+      "loss": 0.1744,
+      "step": 3071
+    },
+    {
+      "epoch": 1.8691816245816855,
+      "grad_norm": 0.7653868794441223,
+      "learning_rate": 2.9279892412630493e-05,
+      "loss": 0.17,
+      "step": 3072
+    },
+    {
+      "epoch": 1.8697900821417706,
+      "grad_norm": 0.700298011302948,
+      "learning_rate": 2.9267626810953584e-05,
+      "loss": 0.1628,
+      "step": 3073
+    },
+    {
+      "epoch": 1.8703985397018559,
+      "grad_norm": 0.7971607446670532,
+      "learning_rate": 2.9255360151083107e-05,
+      "loss": 0.1727,
+      "step": 3074
+    },
+    {
+      "epoch": 1.871006997261941,
+      "grad_norm": 0.8248562812805176,
+      "learning_rate": 2.924309243606069e-05,
+      "loss": 0.1263,
+      "step": 3075
+    },
+    {
+      "epoch": 1.871615454822026,
+      "grad_norm": 0.8164551854133606,
+      "learning_rate": 2.9230823668928198e-05,
+      "loss": 0.1476,
+      "step": 3076
+    },
+    {
+      "epoch": 1.8722239123821114,
+      "grad_norm": 0.7059839963912964,
+      "learning_rate": 2.9218553852727794e-05,
+      "loss": 0.1399,
+      "step": 3077
+    },
+    {
+      "epoch": 1.8728323699421965,
+      "grad_norm": 0.8123924732208252,
+      "learning_rate": 2.920628299050187e-05,
+      "loss": 0.1431,
+      "step": 3078
+    },
+    {
+      "epoch": 1.8734408275022818,
+      "grad_norm": 0.6520960927009583,
+      "learning_rate": 2.9194011085293093e-05,
+      "loss": 0.1423,
+      "step": 3079
+    },
+    {
+      "epoch": 1.874049285062367,
+      "grad_norm": 0.7921064496040344,
+      "learning_rate": 2.9181738140144382e-05,
+      "loss": 0.1541,
+      "step": 3080
+    },
+    {
+      "epoch": 1.874657742622452,
+      "grad_norm": 0.8729209303855896,
+      "learning_rate": 2.9169464158098914e-05,
+      "loss": 0.1544,
+      "step": 3081
+    },
+    {
+      "epoch": 1.8752662001825373,
+      "grad_norm": 0.745010256767273,
+      "learning_rate": 2.9157189142200124e-05,
+      "loss": 0.1548,
+      "step": 3082
+    },
+    {
+      "epoch": 1.8758746577426224,
+      "grad_norm": 0.7852745056152344,
+      "learning_rate": 2.914491309549171e-05,
+      "loss": 0.1595,
+      "step": 3083
+    },
+    {
+      "epoch": 1.8764831153027077,
+      "grad_norm": 0.7547739744186401,
+      "learning_rate": 2.9132636021017616e-05,
+      "loss": 0.14,
+      "step": 3084
+    },
+    {
+      "epoch": 1.8770915728627928,
+      "grad_norm": 0.9375334978103638,
+      "learning_rate": 2.912035792182205e-05,
+      "loss": 0.1149,
+      "step": 3085
+    },
+    {
+      "epoch": 1.877700030422878,
+      "grad_norm": 0.7835490703582764,
+      "learning_rate": 2.9108078800949456e-05,
+      "loss": 0.1577,
+      "step": 3086
+    },
+    {
+      "epoch": 1.8783084879829632,
+      "grad_norm": 0.7294179797172546,
+      "learning_rate": 2.9095798661444557e-05,
+      "loss": 0.1358,
+      "step": 3087
+    },
+    {
+      "epoch": 1.8789169455430483,
+      "grad_norm": 0.7559580206871033,
+      "learning_rate": 2.9083517506352315e-05,
+      "loss": 0.144,
+      "step": 3088
+    },
+    {
+      "epoch": 1.8795254031031337,
+      "grad_norm": 0.7099955677986145,
+      "learning_rate": 2.9071235338717935e-05,
+      "loss": 0.1271,
+      "step": 3089
+    },
+    {
+      "epoch": 1.8801338606632187,
+      "grad_norm": 0.8515664339065552,
+      "learning_rate": 2.9058952161586896e-05,
+      "loss": 0.1777,
+      "step": 3090
+    },
+    {
+      "epoch": 1.8807423182233038,
+      "grad_norm": 0.8556992411613464,
+      "learning_rate": 2.90466679780049e-05,
+      "loss": 0.1626,
+      "step": 3091
+    },
+    {
+      "epoch": 1.8813507757833892,
+      "grad_norm": 0.872840404510498,
+      "learning_rate": 2.9034382791017918e-05,
+      "loss": 0.1781,
+      "step": 3092
+    },
+    {
+      "epoch": 1.8819592333434743,
+      "grad_norm": 0.6905577182769775,
+      "learning_rate": 2.9022096603672166e-05,
+      "loss": 0.171,
+      "step": 3093
+    },
+    {
+      "epoch": 1.8825676909035596,
+      "grad_norm": 0.7203934788703918,
+      "learning_rate": 2.9009809419014107e-05,
+      "loss": 0.1366,
+      "step": 3094
+    },
+    {
+      "epoch": 1.8831761484636447,
+      "grad_norm": 0.8403582572937012,
+      "learning_rate": 2.8997521240090448e-05,
+      "loss": 0.1349,
+      "step": 3095
+    },
+    {
+      "epoch": 1.8837846060237298,
+      "grad_norm": 0.7579777836799622,
+      "learning_rate": 2.898523206994815e-05,
+      "loss": 0.1664,
+      "step": 3096
+    },
+    {
+      "epoch": 1.8843930635838149,
+      "grad_norm": 0.8160215616226196,
+      "learning_rate": 2.8972941911634406e-05,
+      "loss": 0.1475,
+      "step": 3097
+    },
+    {
+      "epoch": 1.8850015211439002,
+      "grad_norm": 0.7258248925209045,
+      "learning_rate": 2.8960650768196672e-05,
+      "loss": 0.1625,
+      "step": 3098
+    },
+    {
+      "epoch": 1.8856099787039855,
+      "grad_norm": 0.9643956422805786,
+      "learning_rate": 2.894835864268263e-05,
+      "loss": 0.1672,
+      "step": 3099
+    },
+    {
+      "epoch": 1.8862184362640706,
+      "grad_norm": 0.7425087690353394,
+      "learning_rate": 2.8936065538140228e-05,
+      "loss": 0.1373,
+      "step": 3100
+    },
+    {
+      "epoch": 1.8868268938241557,
+      "grad_norm": 0.8233270645141602,
+      "learning_rate": 2.8923771457617634e-05,
+      "loss": 0.2127,
+      "step": 3101
+    },
+    {
+      "epoch": 1.8874353513842408,
+      "grad_norm": 0.7719386219978333,
+      "learning_rate": 2.891147640416327e-05,
+      "loss": 0.1492,
+      "step": 3102
+    },
+    {
+      "epoch": 1.888043808944326,
+      "grad_norm": 0.8687146902084351,
+      "learning_rate": 2.889918038082579e-05,
+      "loss": 0.1577,
+      "step": 3103
+    },
+    {
+      "epoch": 1.8886522665044114,
+      "grad_norm": 0.8466896414756775,
+      "learning_rate": 2.8886883390654106e-05,
+      "loss": 0.1688,
+      "step": 3104
+    },
+    {
+      "epoch": 1.8892607240644965,
+      "grad_norm": 0.7307385206222534,
+      "learning_rate": 2.8874585436697355e-05,
+      "loss": 0.0943,
+      "step": 3105
+    },
+    {
+      "epoch": 1.8898691816245816,
+      "grad_norm": 0.9252493977546692,
+      "learning_rate": 2.8862286522004916e-05,
+      "loss": 0.1971,
+      "step": 3106
+    },
+    {
+      "epoch": 1.8904776391846667,
+      "grad_norm": 0.7721157073974609,
+      "learning_rate": 2.8849986649626405e-05,
+      "loss": 0.1684,
+      "step": 3107
+    },
+    {
+      "epoch": 1.891086096744752,
+      "grad_norm": 0.8666673302650452,
+      "learning_rate": 2.8837685822611682e-05,
+      "loss": 0.1636,
+      "step": 3108
+    },
+    {
+      "epoch": 1.8916945543048374,
+      "grad_norm": 0.8242167234420776,
+      "learning_rate": 2.882538404401084e-05,
+      "loss": 0.1706,
+      "step": 3109
+    },
+    {
+      "epoch": 1.8923030118649224,
+      "grad_norm": 0.8100599050521851,
+      "learning_rate": 2.8813081316874197e-05,
+      "loss": 0.1448,
+      "step": 3110
+    },
+    {
+      "epoch": 1.8929114694250075,
+      "grad_norm": 0.7981263399124146,
+      "learning_rate": 2.8800777644252336e-05,
+      "loss": 0.1324,
+      "step": 3111
+    },
+    {
+      "epoch": 1.8935199269850926,
+      "grad_norm": 0.7660496830940247,
+      "learning_rate": 2.8788473029196033e-05,
+      "loss": 0.1506,
+      "step": 3112
+    },
+    {
+      "epoch": 1.894128384545178,
+      "grad_norm": 0.8053763508796692,
+      "learning_rate": 2.877616747475634e-05,
+      "loss": 0.1377,
+      "step": 3113
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.7818877696990967,
+      "learning_rate": 2.8763860983984502e-05,
+      "loss": 0.1491,
+      "step": 3114
+    },
+    {
+      "epoch": 1.8953452996653484,
+      "grad_norm": 0.8091548085212708,
+      "learning_rate": 2.8751553559932033e-05,
+      "loss": 0.1398,
+      "step": 3115
+    },
+    {
+      "epoch": 1.8959537572254335,
+      "grad_norm": 0.8318586349487305,
+      "learning_rate": 2.873924520565065e-05,
+      "loss": 0.1602,
+      "step": 3116
+    },
+    {
+      "epoch": 1.8965622147855186,
+      "grad_norm": 0.8310206532478333,
+      "learning_rate": 2.872693592419231e-05,
+      "loss": 0.1797,
+      "step": 3117
+    },
+    {
+      "epoch": 1.8971706723456039,
+      "grad_norm": 0.802715003490448,
+      "learning_rate": 2.8714625718609213e-05,
+      "loss": 0.156,
+      "step": 3118
+    },
+    {
+      "epoch": 1.8977791299056892,
+      "grad_norm": 0.7569795846939087,
+      "learning_rate": 2.8702314591953776e-05,
+      "loss": 0.1636,
+      "step": 3119
+    },
+    {
+      "epoch": 1.8983875874657743,
+      "grad_norm": 0.6739256381988525,
+      "learning_rate": 2.8690002547278633e-05,
+      "loss": 0.1341,
+      "step": 3120
+    },
+    {
+      "epoch": 1.8989960450258594,
+      "grad_norm": 0.8993747234344482,
+      "learning_rate": 2.8677689587636668e-05,
+      "loss": 0.1541,
+      "step": 3121
+    },
+    {
+      "epoch": 1.8996045025859445,
+      "grad_norm": 0.7927727699279785,
+      "learning_rate": 2.866537571608098e-05,
+      "loss": 0.1603,
+      "step": 3122
+    },
+    {
+      "epoch": 1.9002129601460298,
+      "grad_norm": 0.7829379439353943,
+      "learning_rate": 2.8653060935664888e-05,
+      "loss": 0.1674,
+      "step": 3123
+    },
+    {
+      "epoch": 1.9008214177061151,
+      "grad_norm": 0.8112666606903076,
+      "learning_rate": 2.8640745249441958e-05,
+      "loss": 0.1731,
+      "step": 3124
+    },
+    {
+      "epoch": 1.9014298752662002,
+      "grad_norm": 0.8178973197937012,
+      "learning_rate": 2.8628428660465957e-05,
+      "loss": 0.1408,
+      "step": 3125
+    },
+    {
+      "epoch": 1.9020383328262853,
+      "grad_norm": 0.7954561114311218,
+      "learning_rate": 2.8616111171790894e-05,
+      "loss": 0.1557,
+      "step": 3126
+    },
+    {
+      "epoch": 1.9026467903863704,
+      "grad_norm": 0.8346177935600281,
+      "learning_rate": 2.860379278647098e-05,
+      "loss": 0.1281,
+      "step": 3127
+    },
+    {
+      "epoch": 1.9032552479464557,
+      "grad_norm": 0.7774602174758911,
+      "learning_rate": 2.8591473507560667e-05,
+      "loss": 0.1308,
+      "step": 3128
+    },
+    {
+      "epoch": 1.903863705506541,
+      "grad_norm": 0.8558622598648071,
+      "learning_rate": 2.8579153338114635e-05,
+      "loss": 0.1745,
+      "step": 3129
+    },
+    {
+      "epoch": 1.9044721630666261,
+      "grad_norm": 0.8205204606056213,
+      "learning_rate": 2.856683228118775e-05,
+      "loss": 0.1451,
+      "step": 3130
+    },
+    {
+      "epoch": 1.9050806206267112,
+      "grad_norm": 0.8976646661758423,
+      "learning_rate": 2.8554510339835144e-05,
+      "loss": 0.1594,
+      "step": 3131
+    },
+    {
+      "epoch": 1.9056890781867963,
+      "grad_norm": 0.6619974970817566,
+      "learning_rate": 2.8542187517112124e-05,
+      "loss": 0.1453,
+      "step": 3132
+    },
+    {
+      "epoch": 1.9062975357468817,
+      "grad_norm": 0.9032138586044312,
+      "learning_rate": 2.8529863816074244e-05,
+      "loss": 0.1222,
+      "step": 3133
+    },
+    {
+      "epoch": 1.906905993306967,
+      "grad_norm": 0.7461523413658142,
+      "learning_rate": 2.851753923977728e-05,
+      "loss": 0.129,
+      "step": 3134
+    },
+    {
+      "epoch": 1.907514450867052,
+      "grad_norm": 0.8470805883407593,
+      "learning_rate": 2.850521379127719e-05,
+      "loss": 0.1702,
+      "step": 3135
+    },
+    {
+      "epoch": 1.9081229084271372,
+      "grad_norm": 0.7833938002586365,
+      "learning_rate": 2.8492887473630193e-05,
+      "loss": 0.1323,
+      "step": 3136
+    },
+    {
+      "epoch": 1.9087313659872223,
+      "grad_norm": 0.8622560501098633,
+      "learning_rate": 2.848056028989269e-05,
+      "loss": 0.1663,
+      "step": 3137
+    },
+    {
+      "epoch": 1.9093398235473076,
+      "grad_norm": 0.7503676414489746,
+      "learning_rate": 2.8468232243121313e-05,
+      "loss": 0.1282,
+      "step": 3138
+    },
+    {
+      "epoch": 1.909948281107393,
+      "grad_norm": 0.7636743187904358,
+      "learning_rate": 2.8455903336372902e-05,
+      "loss": 0.1245,
+      "step": 3139
+    },
+    {
+      "epoch": 1.910556738667478,
+      "grad_norm": 0.8729671239852905,
+      "learning_rate": 2.844357357270451e-05,
+      "loss": 0.1747,
+      "step": 3140
+    },
+    {
+      "epoch": 1.911165196227563,
+      "grad_norm": 0.6713892817497253,
+      "learning_rate": 2.8431242955173408e-05,
+      "loss": 0.1449,
+      "step": 3141
+    },
+    {
+      "epoch": 1.9117736537876482,
+      "grad_norm": 0.7802202105522156,
+      "learning_rate": 2.841891148683708e-05,
+      "loss": 0.1552,
+      "step": 3142
+    },
+    {
+      "epoch": 1.9123821113477335,
+      "grad_norm": 0.8736104369163513,
+      "learning_rate": 2.8406579170753205e-05,
+      "loss": 0.1342,
+      "step": 3143
+    },
+    {
+      "epoch": 1.9129905689078188,
+      "grad_norm": 0.730114758014679,
+      "learning_rate": 2.8394246009979697e-05,
+      "loss": 0.1631,
+      "step": 3144
+    },
+    {
+      "epoch": 1.913599026467904,
+      "grad_norm": 0.7180743217468262,
+      "learning_rate": 2.8381912007574653e-05,
+      "loss": 0.124,
+      "step": 3145
+    },
+    {
+      "epoch": 1.914207484027989,
+      "grad_norm": 0.6744349002838135,
+      "learning_rate": 2.836957716659639e-05,
+      "loss": 0.1257,
+      "step": 3146
+    },
+    {
+      "epoch": 1.914815941588074,
+      "grad_norm": 0.8679302930831909,
+      "learning_rate": 2.8357241490103447e-05,
+      "loss": 0.1492,
+      "step": 3147
+    },
+    {
+      "epoch": 1.9154243991481594,
+      "grad_norm": 0.821484386920929,
+      "learning_rate": 2.8344904981154548e-05,
+      "loss": 0.1498,
+      "step": 3148
+    },
+    {
+      "epoch": 1.9160328567082447,
+      "grad_norm": 0.8600984811782837,
+      "learning_rate": 2.833256764280864e-05,
+      "loss": 0.1466,
+      "step": 3149
+    },
+    {
+      "epoch": 1.9166413142683298,
+      "grad_norm": 0.7836093902587891,
+      "learning_rate": 2.832022947812486e-05,
+      "loss": 0.1434,
+      "step": 3150
+    },
+    {
+      "epoch": 1.917249771828415,
+      "grad_norm": 0.8246343731880188,
+      "learning_rate": 2.8307890490162564e-05,
+      "loss": 0.1209,
+      "step": 3151
+    },
+    {
+      "epoch": 1.9178582293885,
+      "grad_norm": 0.8240684866905212,
+      "learning_rate": 2.8295550681981314e-05,
+      "loss": 0.1811,
+      "step": 3152
+    },
+    {
+      "epoch": 1.9184666869485854,
+      "grad_norm": 0.7745963931083679,
+      "learning_rate": 2.828321005664085e-05,
+      "loss": 0.1375,
+      "step": 3153
+    },
+    {
+      "epoch": 1.9190751445086707,
+      "grad_norm": 0.8456833362579346,
+      "learning_rate": 2.827086861720115e-05,
+      "loss": 0.1742,
+      "step": 3154
+    },
+    {
+      "epoch": 1.9196836020687558,
+      "grad_norm": 0.8081454634666443,
+      "learning_rate": 2.8258526366722364e-05,
+      "loss": 0.1807,
+      "step": 3155
+    },
+    {
+      "epoch": 1.9202920596288409,
+      "grad_norm": 0.7536942362785339,
+      "learning_rate": 2.8246183308264862e-05,
+      "loss": 0.144,
+      "step": 3156
+    },
+    {
+      "epoch": 1.920900517188926,
+      "grad_norm": 0.7224938869476318,
+      "learning_rate": 2.8233839444889216e-05,
+      "loss": 0.1139,
+      "step": 3157
+    },
+    {
+      "epoch": 1.9215089747490113,
+      "grad_norm": 0.6938396692276001,
+      "learning_rate": 2.822149477965617e-05,
+      "loss": 0.1324,
+      "step": 3158
+    },
+    {
+      "epoch": 1.9221174323090966,
+      "grad_norm": 0.8000092506408691,
+      "learning_rate": 2.8209149315626697e-05,
+      "loss": 0.1448,
+      "step": 3159
+    },
+    {
+      "epoch": 1.9227258898691817,
+      "grad_norm": 0.6927624344825745,
+      "learning_rate": 2.8196803055861964e-05,
+      "loss": 0.1666,
+      "step": 3160
+    },
+    {
+      "epoch": 1.9233343474292668,
+      "grad_norm": 0.6752164363861084,
+      "learning_rate": 2.818445600342332e-05,
+      "loss": 0.1415,
+      "step": 3161
+    },
+    {
+      "epoch": 1.9239428049893519,
+      "grad_norm": 0.8532926440238953,
+      "learning_rate": 2.817210816137232e-05,
+      "loss": 0.2002,
+      "step": 3162
+    },
+    {
+      "epoch": 1.9245512625494372,
+      "grad_norm": 0.6797624230384827,
+      "learning_rate": 2.815975953277072e-05,
+      "loss": 0.1286,
+      "step": 3163
+    },
+    {
+      "epoch": 1.9251597201095225,
+      "grad_norm": 0.7706435918807983,
+      "learning_rate": 2.8147410120680455e-05,
+      "loss": 0.1423,
+      "step": 3164
+    },
+    {
+      "epoch": 1.9257681776696076,
+      "grad_norm": 0.7351078391075134,
+      "learning_rate": 2.8135059928163683e-05,
+      "loss": 0.1081,
+      "step": 3165
+    },
+    {
+      "epoch": 1.9263766352296927,
+      "grad_norm": 0.7624043226242065,
+      "learning_rate": 2.812270895828271e-05,
+      "loss": 0.1184,
+      "step": 3166
+    },
+    {
+      "epoch": 1.9269850927897778,
+      "grad_norm": 1.0484083890914917,
+      "learning_rate": 2.811035721410008e-05,
+      "loss": 0.169,
+      "step": 3167
+    },
+    {
+      "epoch": 1.9275935503498631,
+      "grad_norm": 0.7948042750358582,
+      "learning_rate": 2.8098004698678517e-05,
+      "loss": 0.1207,
+      "step": 3168
+    },
+    {
+      "epoch": 1.9282020079099484,
+      "grad_norm": 0.7036427855491638,
+      "learning_rate": 2.8085651415080903e-05,
+      "loss": 0.1205,
+      "step": 3169
+    },
+    {
+      "epoch": 1.9288104654700335,
+      "grad_norm": 0.7674738168716431,
+      "learning_rate": 2.8073297366370365e-05,
+      "loss": 0.1141,
+      "step": 3170
+    },
+    {
+      "epoch": 1.9294189230301186,
+      "grad_norm": 0.932759702205658,
+      "learning_rate": 2.806094255561018e-05,
+      "loss": 0.1654,
+      "step": 3171
+    },
+    {
+      "epoch": 1.9300273805902037,
+      "grad_norm": 2.075066566467285,
+      "learning_rate": 2.804858698586383e-05,
+      "loss": 0.1699,
+      "step": 3172
+    },
+    {
+      "epoch": 1.930635838150289,
+      "grad_norm": 0.8003236055374146,
+      "learning_rate": 2.8036230660194972e-05,
+      "loss": 0.1529,
+      "step": 3173
+    },
+    {
+      "epoch": 1.9312442957103741,
+      "grad_norm": 0.8078186511993408,
+      "learning_rate": 2.8023873581667476e-05,
+      "loss": 0.1817,
+      "step": 3174
+    },
+    {
+      "epoch": 1.9318527532704595,
+      "grad_norm": 0.8531659841537476,
+      "learning_rate": 2.8011515753345363e-05,
+      "loss": 0.1293,
+      "step": 3175
+    },
+    {
+      "epoch": 1.9324612108305446,
+      "grad_norm": 0.642602264881134,
+      "learning_rate": 2.7999157178292873e-05,
+      "loss": 0.1312,
+      "step": 3176
+    },
+    {
+      "epoch": 1.9330696683906297,
+      "grad_norm": 0.780268132686615,
+      "learning_rate": 2.798679785957442e-05,
+      "loss": 0.1632,
+      "step": 3177
+    },
+    {
+      "epoch": 1.933678125950715,
+      "grad_norm": 0.8036219477653503,
+      "learning_rate": 2.797443780025459e-05,
+      "loss": 0.1647,
+      "step": 3178
+    },
+    {
+      "epoch": 1.9342865835108,
+      "grad_norm": 0.6498638391494751,
+      "learning_rate": 2.7962077003398162e-05,
+      "loss": 0.1176,
+      "step": 3179
+    },
+    {
+      "epoch": 1.9348950410708854,
+      "grad_norm": 0.6818913817405701,
+      "learning_rate": 2.794971547207011e-05,
+      "loss": 0.1363,
+      "step": 3180
+    },
+    {
+      "epoch": 1.9355034986309705,
+      "grad_norm": 0.8026784062385559,
+      "learning_rate": 2.793735320933556e-05,
+      "loss": 0.1645,
+      "step": 3181
+    },
+    {
+      "epoch": 1.9361119561910556,
+      "grad_norm": 0.7087329030036926,
+      "learning_rate": 2.7924990218259862e-05,
+      "loss": 0.1045,
+      "step": 3182
+    },
+    {
+      "epoch": 1.936720413751141,
+      "grad_norm": 0.7351903319358826,
+      "learning_rate": 2.79126265019085e-05,
+      "loss": 0.1643,
+      "step": 3183
+    },
+    {
+      "epoch": 1.937328871311226,
+      "grad_norm": 0.7057482004165649,
+      "learning_rate": 2.7900262063347172e-05,
+      "loss": 0.1522,
+      "step": 3184
+    },
+    {
+      "epoch": 1.9379373288713113,
+      "grad_norm": 0.8340151309967041,
+      "learning_rate": 2.7887896905641746e-05,
+      "loss": 0.152,
+      "step": 3185
+    },
+    {
+      "epoch": 1.9385457864313964,
+      "grad_norm": 0.8494538068771362,
+      "learning_rate": 2.7875531031858253e-05,
+      "loss": 0.1919,
+      "step": 3186
+    },
+    {
+      "epoch": 1.9391542439914815,
+      "grad_norm": 0.6910502314567566,
+      "learning_rate": 2.7863164445062928e-05,
+      "loss": 0.1377,
+      "step": 3187
+    },
+    {
+      "epoch": 1.9397627015515668,
+      "grad_norm": 0.7750650644302368,
+      "learning_rate": 2.785079714832216e-05,
+      "loss": 0.1633,
+      "step": 3188
+    },
+    {
+      "epoch": 1.940371159111652,
+      "grad_norm": 0.7941158413887024,
+      "learning_rate": 2.7838429144702528e-05,
+      "loss": 0.1317,
+      "step": 3189
+    },
+    {
+      "epoch": 1.9409796166717372,
+      "grad_norm": 0.7237272262573242,
+      "learning_rate": 2.7826060437270786e-05,
+      "loss": 0.1287,
+      "step": 3190
+    },
+    {
+      "epoch": 1.9415880742318223,
+      "grad_norm": 0.6614869236946106,
+      "learning_rate": 2.781369102909384e-05,
+      "loss": 0.0973,
+      "step": 3191
+    },
+    {
+      "epoch": 1.9421965317919074,
+      "grad_norm": 0.8167053461074829,
+      "learning_rate": 2.78013209232388e-05,
+      "loss": 0.1562,
+      "step": 3192
+    },
+    {
+      "epoch": 1.9428049893519928,
+      "grad_norm": 0.6749504804611206,
+      "learning_rate": 2.7788950122772944e-05,
+      "loss": 0.1205,
+      "step": 3193
+    },
+    {
+      "epoch": 1.9434134469120778,
+      "grad_norm": 0.729088544845581,
+      "learning_rate": 2.77765786307637e-05,
+      "loss": 0.1377,
+      "step": 3194
+    },
+    {
+      "epoch": 1.9440219044721632,
+      "grad_norm": 0.7123382687568665,
+      "learning_rate": 2.7764206450278697e-05,
+      "loss": 0.1255,
+      "step": 3195
+    },
+    {
+      "epoch": 1.9446303620322483,
+      "grad_norm": 0.7942476272583008,
+      "learning_rate": 2.7751833584385707e-05,
+      "loss": 0.1207,
+      "step": 3196
+    },
+    {
+      "epoch": 1.9452388195923334,
+      "grad_norm": 0.6616600751876831,
+      "learning_rate": 2.7739460036152686e-05,
+      "loss": 0.1114,
+      "step": 3197
+    },
+    {
+      "epoch": 1.9458472771524185,
+      "grad_norm": 0.7162274718284607,
+      "learning_rate": 2.772708580864777e-05,
+      "loss": 0.1137,
+      "step": 3198
+    },
+    {
+      "epoch": 1.9464557347125038,
+      "grad_norm": 0.7372970581054688,
+      "learning_rate": 2.7714710904939238e-05,
+      "loss": 0.1576,
+      "step": 3199
+    },
+    {
+      "epoch": 1.947064192272589,
+      "grad_norm": 0.8062869906425476,
+      "learning_rate": 2.7702335328095562e-05,
+      "loss": 0.1603,
+      "step": 3200
+    },
+    {
+      "epoch": 1.9476726498326742,
+      "grad_norm": 0.7464812397956848,
+      "learning_rate": 2.7689959081185355e-05,
+      "loss": 0.1139,
+      "step": 3201
+    },
+    {
+      "epoch": 1.9482811073927593,
+      "grad_norm": 0.7284194231033325,
+      "learning_rate": 2.7677582167277428e-05,
+      "loss": 0.1258,
+      "step": 3202
+    },
+    {
+      "epoch": 1.9488895649528444,
+      "grad_norm": 0.8625699877738953,
+      "learning_rate": 2.766520458944073e-05,
+      "loss": 0.164,
+      "step": 3203
+    },
+    {
+      "epoch": 1.9494980225129297,
+      "grad_norm": 0.8282449841499329,
+      "learning_rate": 2.7652826350744375e-05,
+      "loss": 0.136,
+      "step": 3204
+    },
+    {
+      "epoch": 1.950106480073015,
+      "grad_norm": 0.8043208122253418,
+      "learning_rate": 2.7640447454257667e-05,
+      "loss": 0.1447,
+      "step": 3205
+    },
+    {
+      "epoch": 1.9507149376331,
+      "grad_norm": 0.7022130489349365,
+      "learning_rate": 2.7628067903050052e-05,
+      "loss": 0.1197,
+      "step": 3206
+    },
+    {
+      "epoch": 1.9513233951931852,
+      "grad_norm": 0.7067160606384277,
+      "learning_rate": 2.761568770019114e-05,
+      "loss": 0.1394,
+      "step": 3207
+    },
+    {
+      "epoch": 1.9519318527532703,
+      "grad_norm": 0.779850423336029,
+      "learning_rate": 2.7603306848750703e-05,
+      "loss": 0.1669,
+      "step": 3208
+    },
+    {
+      "epoch": 1.9525403103133556,
+      "grad_norm": 0.765367865562439,
+      "learning_rate": 2.759092535179868e-05,
+      "loss": 0.1147,
+      "step": 3209
+    },
+    {
+      "epoch": 1.953148767873441,
+      "grad_norm": 0.6569228172302246,
+      "learning_rate": 2.757854321240516e-05,
+      "loss": 0.1039,
+      "step": 3210
+    },
+    {
+      "epoch": 1.953757225433526,
+      "grad_norm": 0.7526434063911438,
+      "learning_rate": 2.756616043364041e-05,
+      "loss": 0.1418,
+      "step": 3211
+    },
+    {
+      "epoch": 1.9543656829936111,
+      "grad_norm": 0.7364466786384583,
+      "learning_rate": 2.7553777018574834e-05,
+      "loss": 0.1265,
+      "step": 3212
+    },
+    {
+      "epoch": 1.9549741405536962,
+      "grad_norm": 0.7600389122962952,
+      "learning_rate": 2.7541392970279e-05,
+      "loss": 0.1372,
+      "step": 3213
+    },
+    {
+      "epoch": 1.9555825981137815,
+      "grad_norm": 0.7566266059875488,
+      "learning_rate": 2.7529008291823642e-05,
+      "loss": 0.1094,
+      "step": 3214
+    },
+    {
+      "epoch": 1.9561910556738669,
+      "grad_norm": 0.8381999135017395,
+      "learning_rate": 2.7516622986279638e-05,
+      "loss": 0.1618,
+      "step": 3215
+    },
+    {
+      "epoch": 1.956799513233952,
+      "grad_norm": 0.8576135039329529,
+      "learning_rate": 2.7504237056718042e-05,
+      "loss": 0.149,
+      "step": 3216
+    },
+    {
+      "epoch": 1.957407970794037,
+      "grad_norm": 0.7162174582481384,
+      "learning_rate": 2.7491850506210027e-05,
+      "loss": 0.129,
+      "step": 3217
+    },
+    {
+      "epoch": 1.9580164283541222,
+      "grad_norm": 0.6965352892875671,
+      "learning_rate": 2.747946333782696e-05,
+      "loss": 0.0866,
+      "step": 3218
+    },
+    {
+      "epoch": 1.9586248859142075,
+      "grad_norm": 0.7965853810310364,
+      "learning_rate": 2.7467075554640326e-05,
+      "loss": 0.1564,
+      "step": 3219
+    },
+    {
+      "epoch": 1.9592333434742928,
+      "grad_norm": 0.6897141933441162,
+      "learning_rate": 2.745468715972179e-05,
+      "loss": 0.1425,
+      "step": 3220
+    },
+    {
+      "epoch": 1.9598418010343779,
+      "grad_norm": 0.6696418523788452,
+      "learning_rate": 2.744229815614316e-05,
+      "loss": 0.1209,
+      "step": 3221
+    },
+    {
+      "epoch": 1.960450258594463,
+      "grad_norm": 0.7108597159385681,
+      "learning_rate": 2.742990854697638e-05,
+      "loss": 0.1068,
+      "step": 3222
+    },
+    {
+      "epoch": 1.961058716154548,
+      "grad_norm": 0.716172456741333,
+      "learning_rate": 2.741751833529357e-05,
+      "loss": 0.1491,
+      "step": 3223
+    },
+    {
+      "epoch": 1.9616671737146334,
+      "grad_norm": 0.8014208674430847,
+      "learning_rate": 2.7405127524166973e-05,
+      "loss": 0.1468,
+      "step": 3224
+    },
+    {
+      "epoch": 1.9622756312747187,
+      "grad_norm": 0.7219249606132507,
+      "learning_rate": 2.7392736116669005e-05,
+      "loss": 0.123,
+      "step": 3225
+    },
+    {
+      "epoch": 1.9628840888348038,
+      "grad_norm": 0.8846964240074158,
+      "learning_rate": 2.738034411587222e-05,
+      "loss": 0.1553,
+      "step": 3226
+    },
+    {
+      "epoch": 1.963492546394889,
+      "grad_norm": 0.7260108590126038,
+      "learning_rate": 2.73679515248493e-05,
+      "loss": 0.1358,
+      "step": 3227
+    },
+    {
+      "epoch": 1.964101003954974,
+      "grad_norm": 0.7410035133361816,
+      "learning_rate": 2.735555834667311e-05,
+      "loss": 0.1261,
+      "step": 3228
+    },
+    {
+      "epoch": 1.9647094615150593,
+      "grad_norm": 0.8541549444198608,
+      "learning_rate": 2.7343164584416637e-05,
+      "loss": 0.1394,
+      "step": 3229
+    },
+    {
+      "epoch": 1.9653179190751446,
+      "grad_norm": 0.7482866048812866,
+      "learning_rate": 2.7330770241153008e-05,
+      "loss": 0.1007,
+      "step": 3230
+    },
+    {
+      "epoch": 1.9659263766352297,
+      "grad_norm": 0.9279438853263855,
+      "learning_rate": 2.7318375319955512e-05,
+      "loss": 0.1857,
+      "step": 3231
+    },
+    {
+      "epoch": 1.9665348341953148,
+      "grad_norm": 0.815904438495636,
+      "learning_rate": 2.7305979823897576e-05,
+      "loss": 0.1255,
+      "step": 3232
+    },
+    {
+      "epoch": 1.9671432917554,
+      "grad_norm": 0.8731521368026733,
+      "learning_rate": 2.7293583756052755e-05,
+      "loss": 0.1447,
+      "step": 3233
+    },
+    {
+      "epoch": 1.9677517493154852,
+      "grad_norm": 0.8212949633598328,
+      "learning_rate": 2.7281187119494773e-05,
+      "loss": 0.1359,
+      "step": 3234
+    },
+    {
+      "epoch": 1.9683602068755706,
+      "grad_norm": 0.7952809929847717,
+      "learning_rate": 2.726878991729746e-05,
+      "loss": 0.1302,
+      "step": 3235
+    },
+    {
+      "epoch": 1.9689686644356557,
+      "grad_norm": 0.8877054452896118,
+      "learning_rate": 2.7256392152534825e-05,
+      "loss": 0.1527,
+      "step": 3236
+    },
+    {
+      "epoch": 1.9695771219957408,
+      "grad_norm": 0.844585120677948,
+      "learning_rate": 2.724399382828098e-05,
+      "loss": 0.1602,
+      "step": 3237
+    },
+    {
+      "epoch": 1.9701855795558258,
+      "grad_norm": 0.7585496306419373,
+      "learning_rate": 2.72315949476102e-05,
+      "loss": 0.148,
+      "step": 3238
+    },
+    {
+      "epoch": 1.9707940371159112,
+      "grad_norm": 0.7337778210639954,
+      "learning_rate": 2.7219195513596894e-05,
+      "loss": 0.1213,
+      "step": 3239
+    },
+    {
+      "epoch": 1.9714024946759965,
+      "grad_norm": 0.8415425419807434,
+      "learning_rate": 2.72067955293156e-05,
+      "loss": 0.1589,
+      "step": 3240
+    },
+    {
+      "epoch": 1.9720109522360816,
+      "grad_norm": 0.7917572855949402,
+      "learning_rate": 2.7194394997841e-05,
+      "loss": 0.1552,
+      "step": 3241
+    },
+    {
+      "epoch": 1.9726194097961667,
+      "grad_norm": 0.7449464201927185,
+      "learning_rate": 2.7181993922247907e-05,
+      "loss": 0.1232,
+      "step": 3242
+    },
+    {
+      "epoch": 1.9732278673562518,
+      "grad_norm": 0.956392228603363,
+      "learning_rate": 2.7169592305611262e-05,
+      "loss": 0.1336,
+      "step": 3243
+    },
+    {
+      "epoch": 1.973836324916337,
+      "grad_norm": 0.8234003782272339,
+      "learning_rate": 2.715719015100617e-05,
+      "loss": 0.1144,
+      "step": 3244
+    },
+    {
+      "epoch": 1.9744447824764224,
+      "grad_norm": 0.7959507703781128,
+      "learning_rate": 2.714478746150783e-05,
+      "loss": 0.1309,
+      "step": 3245
+    },
+    {
+      "epoch": 1.9750532400365075,
+      "grad_norm": 0.8472030162811279,
+      "learning_rate": 2.71323842401916e-05,
+      "loss": 0.1262,
+      "step": 3246
+    },
+    {
+      "epoch": 1.9756616975965926,
+      "grad_norm": 0.9578633904457092,
+      "learning_rate": 2.711998049013297e-05,
+      "loss": 0.1577,
+      "step": 3247
+    },
+    {
+      "epoch": 1.9762701551566777,
+      "grad_norm": 0.6737805604934692,
+      "learning_rate": 2.710757621440753e-05,
+      "loss": 0.1057,
+      "step": 3248
+    },
+    {
+      "epoch": 1.976878612716763,
+      "grad_norm": 0.8866187334060669,
+      "learning_rate": 2.709517141609105e-05,
+      "loss": 0.1305,
+      "step": 3249
+    },
+    {
+      "epoch": 1.9774870702768483,
+      "grad_norm": 0.7365625500679016,
+      "learning_rate": 2.7082766098259377e-05,
+      "loss": 0.1691,
+      "step": 3250
+    },
+    {
+      "epoch": 1.9780955278369334,
+      "grad_norm": 0.7187755703926086,
+      "learning_rate": 2.7070360263988526e-05,
+      "loss": 0.1036,
+      "step": 3251
+    },
+    {
+      "epoch": 1.9787039853970185,
+      "grad_norm": 0.7651616334915161,
+      "learning_rate": 2.7057953916354638e-05,
+      "loss": 0.1153,
+      "step": 3252
+    },
+    {
+      "epoch": 1.9793124429571036,
+      "grad_norm": 0.7826346158981323,
+      "learning_rate": 2.7045547058433953e-05,
+      "loss": 0.1409,
+      "step": 3253
+    },
+    {
+      "epoch": 1.979920900517189,
+      "grad_norm": 0.7577517032623291,
+      "learning_rate": 2.7033139693302854e-05,
+      "loss": 0.1137,
+      "step": 3254
+    },
+    {
+      "epoch": 1.9805293580772743,
+      "grad_norm": 0.7163174152374268,
+      "learning_rate": 2.7020731824037865e-05,
+      "loss": 0.1258,
+      "step": 3255
+    },
+    {
+      "epoch": 1.9811378156373594,
+      "grad_norm": 0.7855852246284485,
+      "learning_rate": 2.7008323453715607e-05,
+      "loss": 0.1282,
+      "step": 3256
+    },
+    {
+      "epoch": 1.9817462731974445,
+      "grad_norm": 0.8346313834190369,
+      "learning_rate": 2.6995914585412847e-05,
+      "loss": 0.126,
+      "step": 3257
+    },
+    {
+      "epoch": 1.9823547307575295,
+      "grad_norm": 0.7510027885437012,
+      "learning_rate": 2.6983505222206456e-05,
+      "loss": 0.1334,
+      "step": 3258
+    },
+    {
+      "epoch": 1.9829631883176149,
+      "grad_norm": 0.8980780839920044,
+      "learning_rate": 2.697109536717346e-05,
+      "loss": 0.1224,
+      "step": 3259
+    },
+    {
+      "epoch": 1.9835716458777002,
+      "grad_norm": 0.7538368105888367,
+      "learning_rate": 2.6958685023390963e-05,
+      "loss": 0.1239,
+      "step": 3260
+    },
+    {
+      "epoch": 1.9841801034377853,
+      "grad_norm": 0.9021039605140686,
+      "learning_rate": 2.6946274193936222e-05,
+      "loss": 0.1808,
+      "step": 3261
+    },
+    {
+      "epoch": 1.9847885609978704,
+      "grad_norm": 0.6933613419532776,
+      "learning_rate": 2.6933862881886607e-05,
+      "loss": 0.1236,
+      "step": 3262
+    },
+    {
+      "epoch": 1.9853970185579555,
+      "grad_norm": 0.8434465527534485,
+      "learning_rate": 2.6921451090319603e-05,
+      "loss": 0.1461,
+      "step": 3263
+    },
+    {
+      "epoch": 1.9860054761180408,
+      "grad_norm": 0.692387580871582,
+      "learning_rate": 2.6909038822312826e-05,
+      "loss": 0.1007,
+      "step": 3264
+    },
+    {
+      "epoch": 1.986613933678126,
+      "grad_norm": 0.7423060536384583,
+      "learning_rate": 2.6896626080943983e-05,
+      "loss": 0.1531,
+      "step": 3265
+    },
+    {
+      "epoch": 1.9872223912382112,
+      "grad_norm": 0.7204686403274536,
+      "learning_rate": 2.6884212869290932e-05,
+      "loss": 0.132,
+      "step": 3266
+    },
+    {
+      "epoch": 1.9878308487982963,
+      "grad_norm": 0.7953908443450928,
+      "learning_rate": 2.6871799190431627e-05,
+      "loss": 0.1327,
+      "step": 3267
+    },
+    {
+      "epoch": 1.9884393063583814,
+      "grad_norm": 0.7700750827789307,
+      "learning_rate": 2.6859385047444146e-05,
+      "loss": 0.1351,
+      "step": 3268
+    },
+    {
+      "epoch": 1.9890477639184667,
+      "grad_norm": 0.8604716658592224,
+      "learning_rate": 2.684697044340667e-05,
+      "loss": 0.1616,
+      "step": 3269
+    },
+    {
+      "epoch": 1.989656221478552,
+      "grad_norm": 0.7228001356124878,
+      "learning_rate": 2.683455538139752e-05,
+      "loss": 0.1162,
+      "step": 3270
+    },
+    {
+      "epoch": 1.9902646790386371,
+      "grad_norm": 0.7633928060531616,
+      "learning_rate": 2.6822139864495092e-05,
+      "loss": 0.1104,
+      "step": 3271
+    },
+    {
+      "epoch": 1.9908731365987222,
+      "grad_norm": 0.7890945076942444,
+      "learning_rate": 2.680972389577794e-05,
+      "loss": 0.1464,
+      "step": 3272
+    },
+    {
+      "epoch": 1.9914815941588073,
+      "grad_norm": 0.6925262808799744,
+      "learning_rate": 2.6797307478324683e-05,
+      "loss": 0.1416,
+      "step": 3273
+    },
+    {
+      "epoch": 1.9920900517188926,
+      "grad_norm": 0.7635548710823059,
+      "learning_rate": 2.678489061521409e-05,
+      "loss": 0.0874,
+      "step": 3274
+    },
+    {
+      "epoch": 1.9926985092789777,
+      "grad_norm": 0.7441225051879883,
+      "learning_rate": 2.6772473309525027e-05,
+      "loss": 0.1141,
+      "step": 3275
+    },
+    {
+      "epoch": 1.993306966839063,
+      "grad_norm": 0.7360615730285645,
+      "learning_rate": 2.6760055564336462e-05,
+      "loss": 0.1356,
+      "step": 3276
+    },
+    {
+      "epoch": 1.9939154243991482,
+      "grad_norm": 0.7741050720214844,
+      "learning_rate": 2.674763738272748e-05,
+      "loss": 0.1568,
+      "step": 3277
+    },
+    {
+      "epoch": 1.9945238819592332,
+      "grad_norm": 0.7115604281425476,
+      "learning_rate": 2.673521876777727e-05,
+      "loss": 0.1049,
+      "step": 3278
+    },
+    {
+      "epoch": 1.9951323395193186,
+      "grad_norm": 0.750180721282959,
+      "learning_rate": 2.6722799722565127e-05,
+      "loss": 0.1387,
+      "step": 3279
+    },
+    {
+      "epoch": 1.9957407970794037,
+      "grad_norm": 0.7199980020523071,
+      "learning_rate": 2.6710380250170476e-05,
+      "loss": 0.1238,
+      "step": 3280
+    },
+    {
+      "epoch": 1.996349254639489,
+      "grad_norm": 0.7894657850265503,
+      "learning_rate": 2.6697960353672808e-05,
+      "loss": 0.1399,
+      "step": 3281
+    },
+    {
+      "epoch": 1.996957712199574,
+      "grad_norm": 0.7097627520561218,
+      "learning_rate": 2.6685540036151752e-05,
+      "loss": 0.1499,
+      "step": 3282
+    },
+    {
+      "epoch": 1.9975661697596592,
+      "grad_norm": 0.7155794501304626,
+      "learning_rate": 2.6673119300687015e-05,
+      "loss": 0.1142,
+      "step": 3283
+    },
+    {
+      "epoch": 1.9981746273197445,
+      "grad_norm": 0.6458375453948975,
+      "learning_rate": 2.6660698150358433e-05,
+      "loss": 0.0996,
+      "step": 3284
+    },
+    {
+      "epoch": 1.9987830848798296,
+      "grad_norm": 0.6683632135391235,
+      "learning_rate": 2.6648276588245935e-05,
+      "loss": 0.1444,
+      "step": 3285
+    },
+    {
+      "epoch": 1.999391542439915,
+      "grad_norm": 0.6919201016426086,
+      "learning_rate": 2.663585461742954e-05,
+      "loss": 0.1567,
+      "step": 3286
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.7347707152366638,
+      "learning_rate": 2.662343224098939e-05,
+      "loss": 0.1315,
+      "step": 3287
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.1706266403198242,
+      "eval_runtime": 105.2304,
+      "eval_samples_per_second": 7.241,
+      "eval_steps_per_second": 0.456,
+      "step": 3287
+    },
+    {
+      "epoch": 2.000608457560085,
+      "grad_norm": 0.6429162621498108,
+      "learning_rate": 2.6611009462005716e-05,
+      "loss": 0.0639,
+      "step": 3288
+    },
+    {
+      "epoch": 2.00121691512017,
+      "grad_norm": 0.9474073052406311,
+      "learning_rate": 2.659858628355884e-05,
+      "loss": 0.0587,
+      "step": 3289
+    },
+    {
+      "epoch": 2.0018253726802557,
+      "grad_norm": 0.5293480753898621,
+      "learning_rate": 2.6586162708729197e-05,
+      "loss": 0.041,
+      "step": 3290
+    },
+    {
+      "epoch": 2.002433830240341,
+      "grad_norm": 0.6578330993652344,
+      "learning_rate": 2.657373874059732e-05,
+      "loss": 0.0757,
+      "step": 3291
+    },
+    {
+      "epoch": 2.003042287800426,
+      "grad_norm": 0.4521423876285553,
+      "learning_rate": 2.6561314382243835e-05,
+      "loss": 0.0431,
+      "step": 3292
+    },
+    {
+      "epoch": 2.003650745360511,
+      "grad_norm": 0.5849077701568604,
+      "learning_rate": 2.654888963674945e-05,
+      "loss": 0.0604,
+      "step": 3293
+    },
+    {
+      "epoch": 2.004259202920596,
+      "grad_norm": 0.6320520043373108,
+      "learning_rate": 2.6536464507195015e-05,
+      "loss": 0.0662,
+      "step": 3294
+    },
+    {
+      "epoch": 2.0048676604806817,
+      "grad_norm": 0.6279209852218628,
+      "learning_rate": 2.652403899666141e-05,
+      "loss": 0.0622,
+      "step": 3295
+    },
+    {
+      "epoch": 2.0054761180407668,
+      "grad_norm": 0.7935817837715149,
+      "learning_rate": 2.651161310822966e-05,
+      "loss": 0.0558,
+      "step": 3296
+    },
+    {
+      "epoch": 2.006084575600852,
+      "grad_norm": 0.5588900446891785,
+      "learning_rate": 2.649918684498087e-05,
+      "loss": 0.0638,
+      "step": 3297
+    },
+    {
+      "epoch": 2.006693033160937,
+      "grad_norm": 0.5976481437683105,
+      "learning_rate": 2.6486760209996238e-05,
+      "loss": 0.0582,
+      "step": 3298
+    },
+    {
+      "epoch": 2.007301490721022,
+      "grad_norm": 0.7166075110435486,
+      "learning_rate": 2.6474333206357038e-05,
+      "loss": 0.0676,
+      "step": 3299
+    },
+    {
+      "epoch": 2.0079099482811076,
+      "grad_norm": 0.71381014585495,
+      "learning_rate": 2.6461905837144656e-05,
+      "loss": 0.0694,
+      "step": 3300
+    },
+    {
+      "epoch": 2.0085184058411927,
+      "grad_norm": 0.626546323299408,
+      "learning_rate": 2.644947810544056e-05,
+      "loss": 0.0456,
+      "step": 3301
+    },
+    {
+      "epoch": 2.0091268634012778,
+      "grad_norm": 0.5405663251876831,
+      "learning_rate": 2.6437050014326313e-05,
+      "loss": 0.0539,
+      "step": 3302
+    },
+    {
+      "epoch": 2.009735320961363,
+      "grad_norm": 0.56814044713974,
+      "learning_rate": 2.642462156688356e-05,
+      "loss": 0.0652,
+      "step": 3303
+    },
+    {
+      "epoch": 2.010343778521448,
+      "grad_norm": 0.6406037211418152,
+      "learning_rate": 2.6412192766194043e-05,
+      "loss": 0.0693,
+      "step": 3304
+    },
+    {
+      "epoch": 2.0109522360815335,
+      "grad_norm": 0.5662345886230469,
+      "learning_rate": 2.6399763615339583e-05,
+      "loss": 0.0526,
+      "step": 3305
+    },
+    {
+      "epoch": 2.0115606936416186,
+      "grad_norm": 0.5635650157928467,
+      "learning_rate": 2.638733411740209e-05,
+      "loss": 0.0641,
+      "step": 3306
+    },
+    {
+      "epoch": 2.0121691512017037,
+      "grad_norm": 0.6141881346702576,
+      "learning_rate": 2.6374904275463563e-05,
+      "loss": 0.0482,
+      "step": 3307
+    },
+    {
+      "epoch": 2.012777608761789,
+      "grad_norm": 0.6459068655967712,
+      "learning_rate": 2.6362474092606088e-05,
+      "loss": 0.0558,
+      "step": 3308
+    },
+    {
+      "epoch": 2.013386066321874,
+      "grad_norm": 0.6503152251243591,
+      "learning_rate": 2.635004357191182e-05,
+      "loss": 0.0591,
+      "step": 3309
+    },
+    {
+      "epoch": 2.0139945238819594,
+      "grad_norm": 0.6401180624961853,
+      "learning_rate": 2.633761271646302e-05,
+      "loss": 0.052,
+      "step": 3310
+    },
+    {
+      "epoch": 2.0146029814420445,
+      "grad_norm": 0.6508215069770813,
+      "learning_rate": 2.632518152934203e-05,
+      "loss": 0.0513,
+      "step": 3311
+    },
+    {
+      "epoch": 2.0152114390021296,
+      "grad_norm": 0.5980373024940491,
+      "learning_rate": 2.631275001363125e-05,
+      "loss": 0.0738,
+      "step": 3312
+    },
+    {
+      "epoch": 2.0158198965622147,
+      "grad_norm": 0.6737512946128845,
+      "learning_rate": 2.630031817241319e-05,
+      "loss": 0.0581,
+      "step": 3313
+    },
+    {
+      "epoch": 2.0164283541223,
+      "grad_norm": 0.5935463905334473,
+      "learning_rate": 2.6287886008770418e-05,
+      "loss": 0.0489,
+      "step": 3314
+    },
+    {
+      "epoch": 2.0170368116823854,
+      "grad_norm": 0.565272331237793,
+      "learning_rate": 2.62754535257856e-05,
+      "loss": 0.0601,
+      "step": 3315
+    },
+    {
+      "epoch": 2.0176452692424705,
+      "grad_norm": 0.5857186317443848,
+      "learning_rate": 2.626302072654147e-05,
+      "loss": 0.0419,
+      "step": 3316
+    },
+    {
+      "epoch": 2.0182537268025555,
+      "grad_norm": 0.6303470134735107,
+      "learning_rate": 2.625058761412085e-05,
+      "loss": 0.061,
+      "step": 3317
+    },
+    {
+      "epoch": 2.0188621843626406,
+      "grad_norm": 0.48001688718795776,
+      "learning_rate": 2.6238154191606625e-05,
+      "loss": 0.0497,
+      "step": 3318
+    },
+    {
+      "epoch": 2.0194706419227257,
+      "grad_norm": 0.6741344332695007,
+      "learning_rate": 2.6225720462081765e-05,
+      "loss": 0.0624,
+      "step": 3319
+    },
+    {
+      "epoch": 2.0200790994828113,
+      "grad_norm": 0.5248076915740967,
+      "learning_rate": 2.6213286428629324e-05,
+      "loss": 0.0398,
+      "step": 3320
+    },
+    {
+      "epoch": 2.0206875570428964,
+      "grad_norm": 0.5384348630905151,
+      "learning_rate": 2.6200852094332423e-05,
+      "loss": 0.0336,
+      "step": 3321
+    },
+    {
+      "epoch": 2.0212960146029815,
+      "grad_norm": 0.527263343334198,
+      "learning_rate": 2.618841746227425e-05,
+      "loss": 0.0464,
+      "step": 3322
+    },
+    {
+      "epoch": 2.0219044721630666,
+      "grad_norm": 0.6149359941482544,
+      "learning_rate": 2.6175982535538098e-05,
+      "loss": 0.0427,
+      "step": 3323
+    },
+    {
+      "epoch": 2.0225129297231517,
+      "grad_norm": 0.691123366355896,
+      "learning_rate": 2.6163547317207276e-05,
+      "loss": 0.0704,
+      "step": 3324
+    },
+    {
+      "epoch": 2.023121387283237,
+      "grad_norm": 0.6679997444152832,
+      "learning_rate": 2.6151111810365224e-05,
+      "loss": 0.0534,
+      "step": 3325
+    },
+    {
+      "epoch": 2.0237298448433223,
+      "grad_norm": 0.5870081186294556,
+      "learning_rate": 2.613867601809543e-05,
+      "loss": 0.0486,
+      "step": 3326
+    },
+    {
+      "epoch": 2.0243383024034074,
+      "grad_norm": 0.5979164838790894,
+      "learning_rate": 2.612623994348144e-05,
+      "loss": 0.0516,
+      "step": 3327
+    },
+    {
+      "epoch": 2.0249467599634925,
+      "grad_norm": 0.6477155685424805,
+      "learning_rate": 2.6113803589606894e-05,
+      "loss": 0.0495,
+      "step": 3328
+    },
+    {
+      "epoch": 2.0255552175235776,
+      "grad_norm": 0.6475479006767273,
+      "learning_rate": 2.6101366959555478e-05,
+      "loss": 0.0615,
+      "step": 3329
+    },
+    {
+      "epoch": 2.026163675083663,
+      "grad_norm": 0.48027321696281433,
+      "learning_rate": 2.608893005641096e-05,
+      "loss": 0.0438,
+      "step": 3330
+    },
+    {
+      "epoch": 2.0267721326437482,
+      "grad_norm": 0.7494072914123535,
+      "learning_rate": 2.6076492883257187e-05,
+      "loss": 0.0782,
+      "step": 3331
+    },
+    {
+      "epoch": 2.0273805902038333,
+      "grad_norm": 0.49906063079833984,
+      "learning_rate": 2.6064055443178036e-05,
+      "loss": 0.0474,
+      "step": 3332
+    },
+    {
+      "epoch": 2.0279890477639184,
+      "grad_norm": 0.6378363370895386,
+      "learning_rate": 2.6051617739257494e-05,
+      "loss": 0.0663,
+      "step": 3333
+    },
+    {
+      "epoch": 2.0285975053240035,
+      "grad_norm": 0.6063147783279419,
+      "learning_rate": 2.603917977457959e-05,
+      "loss": 0.0589,
+      "step": 3334
+    },
+    {
+      "epoch": 2.029205962884089,
+      "grad_norm": 0.575063169002533,
+      "learning_rate": 2.6026741552228417e-05,
+      "loss": 0.054,
+      "step": 3335
+    },
+    {
+      "epoch": 2.029814420444174,
+      "grad_norm": 0.569412112236023,
+      "learning_rate": 2.601430307528813e-05,
+      "loss": 0.0419,
+      "step": 3336
+    },
+    {
+      "epoch": 2.0304228780042592,
+      "grad_norm": 0.5765204429626465,
+      "learning_rate": 2.6001864346842964e-05,
+      "loss": 0.0443,
+      "step": 3337
+    },
+    {
+      "epoch": 2.0310313355643443,
+      "grad_norm": 0.5352804064750671,
+      "learning_rate": 2.5989425369977195e-05,
+      "loss": 0.0451,
+      "step": 3338
+    },
+    {
+      "epoch": 2.0316397931244294,
+      "grad_norm": 0.6071703433990479,
+      "learning_rate": 2.5976986147775178e-05,
+      "loss": 0.0492,
+      "step": 3339
+    },
+    {
+      "epoch": 2.032248250684515,
+      "grad_norm": 1.0402305126190186,
+      "learning_rate": 2.5964546683321317e-05,
+      "loss": 0.0945,
+      "step": 3340
+    },
+    {
+      "epoch": 2.0328567082446,
+      "grad_norm": 0.7192779183387756,
+      "learning_rate": 2.595210697970009e-05,
+      "loss": 0.041,
+      "step": 3341
+    },
+    {
+      "epoch": 2.033465165804685,
+      "grad_norm": 0.623833417892456,
+      "learning_rate": 2.5939667039996006e-05,
+      "loss": 0.0514,
+      "step": 3342
+    },
+    {
+      "epoch": 2.0340736233647703,
+      "grad_norm": 0.7127055525779724,
+      "learning_rate": 2.592722686729367e-05,
+      "loss": 0.0609,
+      "step": 3343
+    },
+    {
+      "epoch": 2.0346820809248554,
+      "grad_norm": 0.7052009701728821,
+      "learning_rate": 2.5914786464677728e-05,
+      "loss": 0.047,
+      "step": 3344
+    },
+    {
+      "epoch": 2.0352905384849405,
+      "grad_norm": 0.5680457949638367,
+      "learning_rate": 2.590234583523286e-05,
+      "loss": 0.0566,
+      "step": 3345
+    },
+    {
+      "epoch": 2.035898996045026,
+      "grad_norm": 0.505352795124054,
+      "learning_rate": 2.5889904982043845e-05,
+      "loss": 0.0526,
+      "step": 3346
+    },
+    {
+      "epoch": 2.036507453605111,
+      "grad_norm": 0.4761469066143036,
+      "learning_rate": 2.5877463908195475e-05,
+      "loss": 0.0408,
+      "step": 3347
+    },
+    {
+      "epoch": 2.037115911165196,
+      "grad_norm": 0.5836338996887207,
+      "learning_rate": 2.586502261677264e-05,
+      "loss": 0.0519,
+      "step": 3348
+    },
+    {
+      "epoch": 2.0377243687252813,
+      "grad_norm": 0.7985496520996094,
+      "learning_rate": 2.5852581110860253e-05,
+      "loss": 0.0479,
+      "step": 3349
+    },
+    {
+      "epoch": 2.0383328262853664,
+      "grad_norm": 0.645341157913208,
+      "learning_rate": 2.5840139393543285e-05,
+      "loss": 0.0608,
+      "step": 3350
+    },
+    {
+      "epoch": 2.038941283845452,
+      "grad_norm": 0.6739196181297302,
+      "learning_rate": 2.582769746790677e-05,
+      "loss": 0.0543,
+      "step": 3351
+    },
+    {
+      "epoch": 2.039549741405537,
+      "grad_norm": 0.6631483435630798,
+      "learning_rate": 2.5815255337035775e-05,
+      "loss": 0.0529,
+      "step": 3352
+    },
+    {
+      "epoch": 2.040158198965622,
+      "grad_norm": 0.596749484539032,
+      "learning_rate": 2.5802813004015443e-05,
+      "loss": 0.0584,
+      "step": 3353
+    },
+    {
+      "epoch": 2.040766656525707,
+      "grad_norm": 0.5741117596626282,
+      "learning_rate": 2.5790370471930953e-05,
+      "loss": 0.0559,
+      "step": 3354
+    },
+    {
+      "epoch": 2.0413751140857923,
+      "grad_norm": 0.5789355635643005,
+      "learning_rate": 2.577792774386752e-05,
+      "loss": 0.0371,
+      "step": 3355
+    },
+    {
+      "epoch": 2.041983571645878,
+      "grad_norm": 0.5400775074958801,
+      "learning_rate": 2.5765484822910434e-05,
+      "loss": 0.0524,
+      "step": 3356
+    },
+    {
+      "epoch": 2.042592029205963,
+      "grad_norm": 0.5711213946342468,
+      "learning_rate": 2.5753041712145032e-05,
+      "loss": 0.0506,
+      "step": 3357
+    },
+    {
+      "epoch": 2.043200486766048,
+      "grad_norm": 0.6942930817604065,
+      "learning_rate": 2.5740598414656658e-05,
+      "loss": 0.0607,
+      "step": 3358
+    },
+    {
+      "epoch": 2.043808944326133,
+      "grad_norm": 0.6652560234069824,
+      "learning_rate": 2.5728154933530755e-05,
+      "loss": 0.0663,
+      "step": 3359
+    },
+    {
+      "epoch": 2.0444174018862182,
+      "grad_norm": 0.6147892475128174,
+      "learning_rate": 2.5715711271852777e-05,
+      "loss": 0.0445,
+      "step": 3360
+    },
+    {
+      "epoch": 2.0450258594463038,
+      "grad_norm": 0.546258270740509,
+      "learning_rate": 2.570326743270824e-05,
+      "loss": 0.0399,
+      "step": 3361
+    },
+    {
+      "epoch": 2.045634317006389,
+      "grad_norm": 0.6004377007484436,
+      "learning_rate": 2.569082341918269e-05,
+      "loss": 0.048,
+      "step": 3362
+    },
+    {
+      "epoch": 2.046242774566474,
+      "grad_norm": 0.6591867208480835,
+      "learning_rate": 2.5678379234361728e-05,
+      "loss": 0.0637,
+      "step": 3363
+    },
+    {
+      "epoch": 2.046851232126559,
+      "grad_norm": 0.7136169075965881,
+      "learning_rate": 2.566593488133099e-05,
+      "loss": 0.0634,
+      "step": 3364
+    },
+    {
+      "epoch": 2.047459689686644,
+      "grad_norm": 0.5792864561080933,
+      "learning_rate": 2.5653490363176157e-05,
+      "loss": 0.0515,
+      "step": 3365
+    },
+    {
+      "epoch": 2.0480681472467297,
+      "grad_norm": 0.6176591515541077,
+      "learning_rate": 2.5641045682982957e-05,
+      "loss": 0.0665,
+      "step": 3366
+    },
+    {
+      "epoch": 2.048676604806815,
+      "grad_norm": 0.5917629599571228,
+      "learning_rate": 2.5628600843837147e-05,
+      "loss": 0.0364,
+      "step": 3367
+    },
+    {
+      "epoch": 2.0492850623669,
+      "grad_norm": 0.6469512581825256,
+      "learning_rate": 2.561615584882453e-05,
+      "loss": 0.0625,
+      "step": 3368
+    },
+    {
+      "epoch": 2.049893519926985,
+      "grad_norm": 0.7131155133247375,
+      "learning_rate": 2.5603710701030946e-05,
+      "loss": 0.0567,
+      "step": 3369
+    },
+    {
+      "epoch": 2.05050197748707,
+      "grad_norm": 0.7157228589057922,
+      "learning_rate": 2.559126540354227e-05,
+      "loss": 0.0609,
+      "step": 3370
+    },
+    {
+      "epoch": 2.0511104350471556,
+      "grad_norm": 0.5784830451011658,
+      "learning_rate": 2.5578819959444413e-05,
+      "loss": 0.0495,
+      "step": 3371
+    },
+    {
+      "epoch": 2.0517188926072407,
+      "grad_norm": 0.6077147126197815,
+      "learning_rate": 2.5566374371823342e-05,
+      "loss": 0.0473,
+      "step": 3372
+    },
+    {
+      "epoch": 2.052327350167326,
+      "grad_norm": 0.47868582606315613,
+      "learning_rate": 2.555392864376503e-05,
+      "loss": 0.0446,
+      "step": 3373
+    },
+    {
+      "epoch": 2.052935807727411,
+      "grad_norm": 0.646273136138916,
+      "learning_rate": 2.554148277835551e-05,
+      "loss": 0.0495,
+      "step": 3374
+    },
+    {
+      "epoch": 2.053544265287496,
+      "grad_norm": 0.5764486193656921,
+      "learning_rate": 2.552903677868082e-05,
+      "loss": 0.0493,
+      "step": 3375
+    },
+    {
+      "epoch": 2.0541527228475815,
+      "grad_norm": 0.5889171957969666,
+      "learning_rate": 2.551659064782706e-05,
+      "loss": 0.0579,
+      "step": 3376
+    },
+    {
+      "epoch": 2.0547611804076666,
+      "grad_norm": 0.6388028860092163,
+      "learning_rate": 2.5504144388880364e-05,
+      "loss": 0.0435,
+      "step": 3377
+    },
+    {
+      "epoch": 2.0553696379677517,
+      "grad_norm": 0.6067814230918884,
+      "learning_rate": 2.5491698004926862e-05,
+      "loss": 0.0472,
+      "step": 3378
+    },
+    {
+      "epoch": 2.055978095527837,
+      "grad_norm": 0.652549684047699,
+      "learning_rate": 2.5479251499052752e-05,
+      "loss": 0.0672,
+      "step": 3379
+    },
+    {
+      "epoch": 2.056586553087922,
+      "grad_norm": 0.6091573238372803,
+      "learning_rate": 2.5466804874344253e-05,
+      "loss": 0.0586,
+      "step": 3380
+    },
+    {
+      "epoch": 2.0571950106480075,
+      "grad_norm": 0.6867235898971558,
+      "learning_rate": 2.5454358133887594e-05,
+      "loss": 0.0604,
+      "step": 3381
+    },
+    {
+      "epoch": 2.0578034682080926,
+      "grad_norm": 0.5185385942459106,
+      "learning_rate": 2.5441911280769065e-05,
+      "loss": 0.0516,
+      "step": 3382
+    },
+    {
+      "epoch": 2.0584119257681777,
+      "grad_norm": 0.5316767692565918,
+      "learning_rate": 2.5429464318074952e-05,
+      "loss": 0.0468,
+      "step": 3383
+    },
+    {
+      "epoch": 2.0590203833282628,
+      "grad_norm": 0.6571846008300781,
+      "learning_rate": 2.541701724889159e-05,
+      "loss": 0.0547,
+      "step": 3384
+    },
+    {
+      "epoch": 2.059628840888348,
+      "grad_norm": 0.6461378335952759,
+      "learning_rate": 2.5404570076305334e-05,
+      "loss": 0.0463,
+      "step": 3385
+    },
+    {
+      "epoch": 2.0602372984484334,
+      "grad_norm": 0.5801909565925598,
+      "learning_rate": 2.5392122803402562e-05,
+      "loss": 0.0558,
+      "step": 3386
+    },
+    {
+      "epoch": 2.0608457560085185,
+      "grad_norm": 0.5066788196563721,
+      "learning_rate": 2.5379675433269684e-05,
+      "loss": 0.0393,
+      "step": 3387
+    },
+    {
+      "epoch": 2.0614542135686036,
+      "grad_norm": 0.6354464292526245,
+      "learning_rate": 2.5367227968993112e-05,
+      "loss": 0.0723,
+      "step": 3388
+    },
+    {
+      "epoch": 2.0620626711286887,
+      "grad_norm": 0.6328490972518921,
+      "learning_rate": 2.535478041365932e-05,
+      "loss": 0.0533,
+      "step": 3389
+    },
+    {
+      "epoch": 2.062671128688774,
+      "grad_norm": 0.6055465936660767,
+      "learning_rate": 2.5342332770354772e-05,
+      "loss": 0.0624,
+      "step": 3390
+    },
+    {
+      "epoch": 2.0632795862488593,
+      "grad_norm": 0.6035982370376587,
+      "learning_rate": 2.5329885042165963e-05,
+      "loss": 0.0491,
+      "step": 3391
+    },
+    {
+      "epoch": 2.0638880438089444,
+      "grad_norm": 0.5577566027641296,
+      "learning_rate": 2.531743723217942e-05,
+      "loss": 0.0433,
+      "step": 3392
+    },
+    {
+      "epoch": 2.0644965013690295,
+      "grad_norm": 0.6334397196769714,
+      "learning_rate": 2.5304989343481668e-05,
+      "loss": 0.0557,
+      "step": 3393
+    },
+    {
+      "epoch": 2.0651049589291146,
+      "grad_norm": 0.6014379262924194,
+      "learning_rate": 2.5292541379159273e-05,
+      "loss": 0.0428,
+      "step": 3394
+    },
+    {
+      "epoch": 2.0657134164891997,
+      "grad_norm": 0.624415934085846,
+      "learning_rate": 2.5280093342298817e-05,
+      "loss": 0.0525,
+      "step": 3395
+    },
+    {
+      "epoch": 2.0663218740492852,
+      "grad_norm": 0.5978030562400818,
+      "learning_rate": 2.5267645235986874e-05,
+      "loss": 0.0581,
+      "step": 3396
+    },
+    {
+      "epoch": 2.0669303316093703,
+      "grad_norm": 0.5354481935501099,
+      "learning_rate": 2.5255197063310075e-05,
+      "loss": 0.0519,
+      "step": 3397
+    },
+    {
+      "epoch": 2.0675387891694554,
+      "grad_norm": 0.6657739877700806,
+      "learning_rate": 2.5242748827355046e-05,
+      "loss": 0.0474,
+      "step": 3398
+    },
+    {
+      "epoch": 2.0681472467295405,
+      "grad_norm": 0.5657456517219543,
+      "learning_rate": 2.5230300531208417e-05,
+      "loss": 0.0368,
+      "step": 3399
+    },
+    {
+      "epoch": 2.0687557042896256,
+      "grad_norm": 0.572913646697998,
+      "learning_rate": 2.5217852177956862e-05,
+      "loss": 0.0542,
+      "step": 3400
+    },
+    {
+      "epoch": 2.069364161849711,
+      "grad_norm": 0.45607224106788635,
+      "learning_rate": 2.5205403770687048e-05,
+      "loss": 0.0383,
+      "step": 3401
+    },
+    {
+      "epoch": 2.0699726194097963,
+      "grad_norm": 0.5699754953384399,
+      "learning_rate": 2.5192955312485655e-05,
+      "loss": 0.0553,
+      "step": 3402
+    },
+    {
+      "epoch": 2.0705810769698814,
+      "grad_norm": 0.6078694462776184,
+      "learning_rate": 2.51805068064394e-05,
+      "loss": 0.0538,
+      "step": 3403
+    },
+    {
+      "epoch": 2.0711895345299665,
+      "grad_norm": 0.6573747396469116,
+      "learning_rate": 2.5168058255634963e-05,
+      "loss": 0.0367,
+      "step": 3404
+    },
+    {
+      "epoch": 2.0717979920900516,
+      "grad_norm": 0.5173442959785461,
+      "learning_rate": 2.5155609663159098e-05,
+      "loss": 0.0324,
+      "step": 3405
+    },
+    {
+      "epoch": 2.072406449650137,
+      "grad_norm": 0.6545349359512329,
+      "learning_rate": 2.514316103209851e-05,
+      "loss": 0.0446,
+      "step": 3406
+    },
+    {
+      "epoch": 2.073014907210222,
+      "grad_norm": 0.6538964509963989,
+      "learning_rate": 2.513071236553996e-05,
+      "loss": 0.0494,
+      "step": 3407
+    },
+    {
+      "epoch": 2.0736233647703073,
+      "grad_norm": 0.636489987373352,
+      "learning_rate": 2.5118263666570198e-05,
+      "loss": 0.048,
+      "step": 3408
+    },
+    {
+      "epoch": 2.0742318223303924,
+      "grad_norm": 0.6040591597557068,
+      "learning_rate": 2.5105814938275968e-05,
+      "loss": 0.065,
+      "step": 3409
+    },
+    {
+      "epoch": 2.0748402798904775,
+      "grad_norm": 0.624994695186615,
+      "learning_rate": 2.5093366183744045e-05,
+      "loss": 0.0609,
+      "step": 3410
+    },
+    {
+      "epoch": 2.075448737450563,
+      "grad_norm": 0.7067580223083496,
+      "learning_rate": 2.50809174060612e-05,
+      "loss": 0.0649,
+      "step": 3411
+    },
+    {
+      "epoch": 2.076057195010648,
+      "grad_norm": 0.6198956370353699,
+      "learning_rate": 2.5068468608314212e-05,
+      "loss": 0.0583,
+      "step": 3412
+    },
+    {
+      "epoch": 2.076665652570733,
+      "grad_norm": 0.6904608607292175,
+      "learning_rate": 2.5056019793589858e-05,
+      "loss": 0.0611,
+      "step": 3413
+    },
+    {
+      "epoch": 2.0772741101308183,
+      "grad_norm": 0.6566850543022156,
+      "learning_rate": 2.504357096497494e-05,
+      "loss": 0.0545,
+      "step": 3414
+    },
+    {
+      "epoch": 2.0778825676909034,
+      "grad_norm": 0.544262707233429,
+      "learning_rate": 2.5031122125556234e-05,
+      "loss": 0.0518,
+      "step": 3415
+    },
+    {
+      "epoch": 2.078491025250989,
+      "grad_norm": 0.5134237408638,
+      "learning_rate": 2.5018673278420534e-05,
+      "loss": 0.0523,
+      "step": 3416
+    },
+    {
+      "epoch": 2.079099482811074,
+      "grad_norm": 0.5637457370758057,
+      "learning_rate": 2.500622442665465e-05,
+      "loss": 0.0466,
+      "step": 3417
+    },
+    {
+      "epoch": 2.079707940371159,
+      "grad_norm": 0.6894789338111877,
+      "learning_rate": 2.4993775573345356e-05,
+      "loss": 0.0601,
+      "step": 3418
+    },
+    {
+      "epoch": 2.0803163979312442,
+      "grad_norm": 0.5236656069755554,
+      "learning_rate": 2.4981326721579472e-05,
+      "loss": 0.0439,
+      "step": 3419
+    },
+    {
+      "epoch": 2.0809248554913293,
+      "grad_norm": 0.5601134300231934,
+      "learning_rate": 2.496887787444377e-05,
+      "loss": 0.0547,
+      "step": 3420
+    },
+    {
+      "epoch": 2.081533313051415,
+      "grad_norm": 0.6053742170333862,
+      "learning_rate": 2.4956429035025063e-05,
+      "loss": 0.064,
+      "step": 3421
+    },
+    {
+      "epoch": 2.0821417706115,
+      "grad_norm": 0.5905669331550598,
+      "learning_rate": 2.4943980206410144e-05,
+      "loss": 0.0522,
+      "step": 3422
+    },
+    {
+      "epoch": 2.082750228171585,
+      "grad_norm": 0.5676148533821106,
+      "learning_rate": 2.4931531391685794e-05,
+      "loss": 0.054,
+      "step": 3423
+    },
+    {
+      "epoch": 2.08335868573167,
+      "grad_norm": 0.6174594759941101,
+      "learning_rate": 2.4919082593938802e-05,
+      "loss": 0.0686,
+      "step": 3424
+    },
+    {
+      "epoch": 2.0839671432917553,
+      "grad_norm": 0.5777302980422974,
+      "learning_rate": 2.4906633816255964e-05,
+      "loss": 0.0531,
+      "step": 3425
+    },
+    {
+      "epoch": 2.084575600851841,
+      "grad_norm": 0.5795081257820129,
+      "learning_rate": 2.489418506172404e-05,
+      "loss": 0.0629,
+      "step": 3426
+    },
+    {
+      "epoch": 2.085184058411926,
+      "grad_norm": 0.6431803107261658,
+      "learning_rate": 2.4881736333429808e-05,
+      "loss": 0.0645,
+      "step": 3427
+    },
+    {
+      "epoch": 2.085792515972011,
+      "grad_norm": 0.5932779312133789,
+      "learning_rate": 2.4869287634460045e-05,
+      "loss": 0.0682,
+      "step": 3428
+    },
+    {
+      "epoch": 2.086400973532096,
+      "grad_norm": 0.4519423246383667,
+      "learning_rate": 2.4856838967901492e-05,
+      "loss": 0.0399,
+      "step": 3429
+    },
+    {
+      "epoch": 2.087009431092181,
+      "grad_norm": 0.5230922698974609,
+      "learning_rate": 2.4844390336840908e-05,
+      "loss": 0.0462,
+      "step": 3430
+    },
+    {
+      "epoch": 2.0876178886522667,
+      "grad_norm": 0.5311405658721924,
+      "learning_rate": 2.4831941744365043e-05,
+      "loss": 0.0402,
+      "step": 3431
+    },
+    {
+      "epoch": 2.088226346212352,
+      "grad_norm": 0.658665657043457,
+      "learning_rate": 2.4819493193560618e-05,
+      "loss": 0.0498,
+      "step": 3432
+    },
+    {
+      "epoch": 2.088834803772437,
+      "grad_norm": 0.5413145422935486,
+      "learning_rate": 2.4807044687514344e-05,
+      "loss": 0.0499,
+      "step": 3433
+    },
+    {
+      "epoch": 2.089443261332522,
+      "grad_norm": 0.6543242931365967,
+      "learning_rate": 2.4794596229312958e-05,
+      "loss": 0.0605,
+      "step": 3434
+    },
+    {
+      "epoch": 2.090051718892607,
+      "grad_norm": 0.5483290553092957,
+      "learning_rate": 2.4782147822043144e-05,
+      "loss": 0.0289,
+      "step": 3435
+    },
+    {
+      "epoch": 2.090660176452692,
+      "grad_norm": 0.5126484036445618,
+      "learning_rate": 2.476969946879159e-05,
+      "loss": 0.0396,
+      "step": 3436
+    },
+    {
+      "epoch": 2.0912686340127777,
+      "grad_norm": 0.5537660121917725,
+      "learning_rate": 2.475725117264496e-05,
+      "loss": 0.0515,
+      "step": 3437
+    },
+    {
+      "epoch": 2.091877091572863,
+      "grad_norm": 0.5816277861595154,
+      "learning_rate": 2.474480293668993e-05,
+      "loss": 0.041,
+      "step": 3438
+    },
+    {
+      "epoch": 2.092485549132948,
+      "grad_norm": 0.6799830794334412,
+      "learning_rate": 2.473235476401313e-05,
+      "loss": 0.0589,
+      "step": 3439
+    },
+    {
+      "epoch": 2.093094006693033,
+      "grad_norm": 0.6604799032211304,
+      "learning_rate": 2.4719906657701192e-05,
+      "loss": 0.0506,
+      "step": 3440
+    },
+    {
+      "epoch": 2.093702464253118,
+      "grad_norm": 0.7943026423454285,
+      "learning_rate": 2.4707458620840732e-05,
+      "loss": 0.0622,
+      "step": 3441
+    },
+    {
+      "epoch": 2.0943109218132037,
+      "grad_norm": 0.5456538200378418,
+      "learning_rate": 2.469501065651834e-05,
+      "loss": 0.0451,
+      "step": 3442
+    },
+    {
+      "epoch": 2.0949193793732888,
+      "grad_norm": 0.6690013408660889,
+      "learning_rate": 2.4682562767820587e-05,
+      "loss": 0.0578,
+      "step": 3443
+    },
+    {
+      "epoch": 2.095527836933374,
+      "grad_norm": 0.645366907119751,
+      "learning_rate": 2.4670114957834043e-05,
+      "loss": 0.062,
+      "step": 3444
+    },
+    {
+      "epoch": 2.096136294493459,
+      "grad_norm": 0.5978085398674011,
+      "learning_rate": 2.4657667229645237e-05,
+      "loss": 0.0486,
+      "step": 3445
+    },
+    {
+      "epoch": 2.096744752053544,
+      "grad_norm": 0.5921434760093689,
+      "learning_rate": 2.4645219586340683e-05,
+      "loss": 0.0548,
+      "step": 3446
+    },
+    {
+      "epoch": 2.0973532096136296,
+      "grad_norm": 0.4595530033111572,
+      "learning_rate": 2.463277203100689e-05,
+      "loss": 0.0425,
+      "step": 3447
+    },
+    {
+      "epoch": 2.0979616671737147,
+      "grad_norm": 0.5539052486419678,
+      "learning_rate": 2.462032456673033e-05,
+      "loss": 0.0537,
+      "step": 3448
+    },
+    {
+      "epoch": 2.0985701247338,
+      "grad_norm": 0.5757131576538086,
+      "learning_rate": 2.4607877196597437e-05,
+      "loss": 0.0438,
+      "step": 3449
+    },
+    {
+      "epoch": 2.099178582293885,
+      "grad_norm": 0.569847047328949,
+      "learning_rate": 2.4595429923694668e-05,
+      "loss": 0.0506,
+      "step": 3450
+    },
+    {
+      "epoch": 2.09978703985397,
+      "grad_norm": 0.5236031413078308,
+      "learning_rate": 2.4582982751108417e-05,
+      "loss": 0.0345,
+      "step": 3451
+    },
+    {
+      "epoch": 2.1003954974140555,
+      "grad_norm": 0.6032043695449829,
+      "learning_rate": 2.4570535681925047e-05,
+      "loss": 0.0528,
+      "step": 3452
+    },
+    {
+      "epoch": 2.1010039549741406,
+      "grad_norm": 0.5206643342971802,
+      "learning_rate": 2.455808871923094e-05,
+      "loss": 0.046,
+      "step": 3453
+    },
+    {
+      "epoch": 2.1016124125342257,
+      "grad_norm": 0.5682435035705566,
+      "learning_rate": 2.4545641866112408e-05,
+      "loss": 0.0408,
+      "step": 3454
+    },
+    {
+      "epoch": 2.102220870094311,
+      "grad_norm": 0.6149486899375916,
+      "learning_rate": 2.453319512565576e-05,
+      "loss": 0.0507,
+      "step": 3455
+    },
+    {
+      "epoch": 2.102829327654396,
+      "grad_norm": 0.5755969882011414,
+      "learning_rate": 2.4520748500947247e-05,
+      "loss": 0.0593,
+      "step": 3456
+    },
+    {
+      "epoch": 2.1034377852144814,
+      "grad_norm": 0.6103445291519165,
+      "learning_rate": 2.4508301995073144e-05,
+      "loss": 0.0553,
+      "step": 3457
+    },
+    {
+      "epoch": 2.1040462427745665,
+      "grad_norm": 0.5596705675125122,
+      "learning_rate": 2.449585561111965e-05,
+      "loss": 0.0489,
+      "step": 3458
+    },
+    {
+      "epoch": 2.1046547003346516,
+      "grad_norm": 0.535492479801178,
+      "learning_rate": 2.4483409352172936e-05,
+      "loss": 0.0503,
+      "step": 3459
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.6022449731826782,
+      "learning_rate": 2.4470963221319188e-05,
+      "loss": 0.0457,
+      "step": 3460
+    },
+    {
+      "epoch": 2.105871615454822,
+      "grad_norm": 0.680794894695282,
+      "learning_rate": 2.4458517221644507e-05,
+      "loss": 0.0521,
+      "step": 3461
+    },
+    {
+      "epoch": 2.1064800730149074,
+      "grad_norm": 0.6097959280014038,
+      "learning_rate": 2.444607135623497e-05,
+      "loss": 0.0467,
+      "step": 3462
+    },
+    {
+      "epoch": 2.1070885305749925,
+      "grad_norm": 0.4896848499774933,
+      "learning_rate": 2.4433625628176663e-05,
+      "loss": 0.0311,
+      "step": 3463
+    },
+    {
+      "epoch": 2.1076969881350776,
+      "grad_norm": 0.6469544172286987,
+      "learning_rate": 2.442118004055559e-05,
+      "loss": 0.0737,
+      "step": 3464
+    },
+    {
+      "epoch": 2.1083054456951626,
+      "grad_norm": 0.5251388549804688,
+      "learning_rate": 2.4408734596457744e-05,
+      "loss": 0.0456,
+      "step": 3465
+    },
+    {
+      "epoch": 2.1089139032552477,
+      "grad_norm": 0.5152940154075623,
+      "learning_rate": 2.439628929896906e-05,
+      "loss": 0.0327,
+      "step": 3466
+    },
+    {
+      "epoch": 2.1095223608153333,
+      "grad_norm": 0.552280068397522,
+      "learning_rate": 2.4383844151175478e-05,
+      "loss": 0.0592,
+      "step": 3467
+    },
+    {
+      "epoch": 2.1101308183754184,
+      "grad_norm": 0.6947097182273865,
+      "learning_rate": 2.4371399156162862e-05,
+      "loss": 0.0865,
+      "step": 3468
+    },
+    {
+      "epoch": 2.1107392759355035,
+      "grad_norm": 0.5415664315223694,
+      "learning_rate": 2.4358954317017045e-05,
+      "loss": 0.0453,
+      "step": 3469
+    },
+    {
+      "epoch": 2.1113477334955886,
+      "grad_norm": 0.5632659196853638,
+      "learning_rate": 2.434650963682385e-05,
+      "loss": 0.0497,
+      "step": 3470
+    },
+    {
+      "epoch": 2.1119561910556737,
+      "grad_norm": 0.5756339430809021,
+      "learning_rate": 2.433406511866902e-05,
+      "loss": 0.0493,
+      "step": 3471
+    },
+    {
+      "epoch": 2.112564648615759,
+      "grad_norm": 0.5048176646232605,
+      "learning_rate": 2.4321620765638274e-05,
+      "loss": 0.0509,
+      "step": 3472
+    },
+    {
+      "epoch": 2.1131731061758443,
+      "grad_norm": 0.6544382572174072,
+      "learning_rate": 2.4309176580817318e-05,
+      "loss": 0.0654,
+      "step": 3473
+    },
+    {
+      "epoch": 2.1137815637359294,
+      "grad_norm": 0.5566495656967163,
+      "learning_rate": 2.429673256729177e-05,
+      "loss": 0.0493,
+      "step": 3474
+    },
+    {
+      "epoch": 2.1143900212960145,
+      "grad_norm": 0.53816819190979,
+      "learning_rate": 2.428428872814722e-05,
+      "loss": 0.0438,
+      "step": 3475
+    },
+    {
+      "epoch": 2.1149984788560996,
+      "grad_norm": 0.5756782293319702,
+      "learning_rate": 2.4271845066469247e-05,
+      "loss": 0.0514,
+      "step": 3476
+    },
+    {
+      "epoch": 2.115606936416185,
+      "grad_norm": 0.6478017568588257,
+      "learning_rate": 2.4259401585343344e-05,
+      "loss": 0.0584,
+      "step": 3477
+    },
+    {
+      "epoch": 2.1162153939762702,
+      "grad_norm": 0.6319558024406433,
+      "learning_rate": 2.424695828785498e-05,
+      "loss": 0.0524,
+      "step": 3478
+    },
+    {
+      "epoch": 2.1168238515363553,
+      "grad_norm": 0.5530511736869812,
+      "learning_rate": 2.4234515177089562e-05,
+      "loss": 0.0523,
+      "step": 3479
+    },
+    {
+      "epoch": 2.1174323090964404,
+      "grad_norm": 0.5840123295783997,
+      "learning_rate": 2.4222072256132483e-05,
+      "loss": 0.0473,
+      "step": 3480
+    },
+    {
+      "epoch": 2.1180407666565255,
+      "grad_norm": 0.5127155184745789,
+      "learning_rate": 2.4209629528069063e-05,
+      "loss": 0.0456,
+      "step": 3481
+    },
+    {
+      "epoch": 2.118649224216611,
+      "grad_norm": 0.6958695650100708,
+      "learning_rate": 2.419718699598456e-05,
+      "loss": 0.0579,
+      "step": 3482
+    },
+    {
+      "epoch": 2.119257681776696,
+      "grad_norm": 0.6823195815086365,
+      "learning_rate": 2.418474466296423e-05,
+      "loss": 0.0523,
+      "step": 3483
+    },
+    {
+      "epoch": 2.1198661393367813,
+      "grad_norm": 0.5722386240959167,
+      "learning_rate": 2.417230253209324e-05,
+      "loss": 0.0497,
+      "step": 3484
+    },
+    {
+      "epoch": 2.1204745968968663,
+      "grad_norm": 1.0104351043701172,
+      "learning_rate": 2.4159860606456718e-05,
+      "loss": 0.0758,
+      "step": 3485
+    },
+    {
+      "epoch": 2.1210830544569514,
+      "grad_norm": 0.5905888080596924,
+      "learning_rate": 2.414741888913975e-05,
+      "loss": 0.0535,
+      "step": 3486
+    },
+    {
+      "epoch": 2.121691512017037,
+      "grad_norm": 0.732254683971405,
+      "learning_rate": 2.4134977383227364e-05,
+      "loss": 0.0567,
+      "step": 3487
+    },
+    {
+      "epoch": 2.122299969577122,
+      "grad_norm": 0.5152473449707031,
+      "learning_rate": 2.412253609180453e-05,
+      "loss": 0.0359,
+      "step": 3488
+    },
+    {
+      "epoch": 2.122908427137207,
+      "grad_norm": 0.6493118405342102,
+      "learning_rate": 2.4110095017956164e-05,
+      "loss": 0.0565,
+      "step": 3489
+    },
+    {
+      "epoch": 2.1235168846972923,
+      "grad_norm": 0.5857681632041931,
+      "learning_rate": 2.4097654164767148e-05,
+      "loss": 0.0332,
+      "step": 3490
+    },
+    {
+      "epoch": 2.1241253422573774,
+      "grad_norm": 0.533673107624054,
+      "learning_rate": 2.4085213535322288e-05,
+      "loss": 0.0364,
+      "step": 3491
+    },
+    {
+      "epoch": 2.124733799817463,
+      "grad_norm": 0.6203796863555908,
+      "learning_rate": 2.4072773132706326e-05,
+      "loss": 0.0544,
+      "step": 3492
+    },
+    {
+      "epoch": 2.125342257377548,
+      "grad_norm": 0.5900121927261353,
+      "learning_rate": 2.4060332960003996e-05,
+      "loss": 0.0369,
+      "step": 3493
+    },
+    {
+      "epoch": 2.125950714937633,
+      "grad_norm": 0.6247245669364929,
+      "learning_rate": 2.4047893020299922e-05,
+      "loss": 0.0499,
+      "step": 3494
+    },
+    {
+      "epoch": 2.126559172497718,
+      "grad_norm": 0.5890246033668518,
+      "learning_rate": 2.403545331667868e-05,
+      "loss": 0.0455,
+      "step": 3495
+    },
+    {
+      "epoch": 2.1271676300578033,
+      "grad_norm": 0.7032641768455505,
+      "learning_rate": 2.4023013852224828e-05,
+      "loss": 0.0676,
+      "step": 3496
+    },
+    {
+      "epoch": 2.127776087617889,
+      "grad_norm": 0.5493380427360535,
+      "learning_rate": 2.401057463002281e-05,
+      "loss": 0.0512,
+      "step": 3497
+    },
+    {
+      "epoch": 2.128384545177974,
+      "grad_norm": 0.590238094329834,
+      "learning_rate": 2.3998135653157035e-05,
+      "loss": 0.0602,
+      "step": 3498
+    },
+    {
+      "epoch": 2.128993002738059,
+      "grad_norm": 0.6478054523468018,
+      "learning_rate": 2.398569692471187e-05,
+      "loss": 0.0459,
+      "step": 3499
+    },
+    {
+      "epoch": 2.129601460298144,
+      "grad_norm": 0.6916689872741699,
+      "learning_rate": 2.397325844777159e-05,
+      "loss": 0.0472,
+      "step": 3500
+    },
+    {
+      "epoch": 2.130209917858229,
+      "grad_norm": 0.5881983637809753,
+      "learning_rate": 2.3960820225420418e-05,
+      "loss": 0.0581,
+      "step": 3501
+    },
+    {
+      "epoch": 2.1308183754183148,
+      "grad_norm": 0.5607914924621582,
+      "learning_rate": 2.39483822607425e-05,
+      "loss": 0.0485,
+      "step": 3502
+    },
+    {
+      "epoch": 2.1314268329784,
+      "grad_norm": 0.601393461227417,
+      "learning_rate": 2.3935944556821966e-05,
+      "loss": 0.0679,
+      "step": 3503
+    },
+    {
+      "epoch": 2.132035290538485,
+      "grad_norm": 0.5792237520217896,
+      "learning_rate": 2.3923507116742826e-05,
+      "loss": 0.0596,
+      "step": 3504
+    },
+    {
+      "epoch": 2.13264374809857,
+      "grad_norm": 0.5309643745422363,
+      "learning_rate": 2.391106994358904e-05,
+      "loss": 0.0422,
+      "step": 3505
+    },
+    {
+      "epoch": 2.133252205658655,
+      "grad_norm": 0.61043381690979,
+      "learning_rate": 2.3898633040444528e-05,
+      "loss": 0.0519,
+      "step": 3506
+    },
+    {
+      "epoch": 2.1338606632187407,
+      "grad_norm": 0.5031585693359375,
+      "learning_rate": 2.388619641039312e-05,
+      "loss": 0.0334,
+      "step": 3507
+    },
+    {
+      "epoch": 2.134469120778826,
+      "grad_norm": 0.5993291139602661,
+      "learning_rate": 2.387376005651856e-05,
+      "loss": 0.0476,
+      "step": 3508
+    },
+    {
+      "epoch": 2.135077578338911,
+      "grad_norm": 0.49983397126197815,
+      "learning_rate": 2.3861323981904575e-05,
+      "loss": 0.0476,
+      "step": 3509
+    },
+    {
+      "epoch": 2.135686035898996,
+      "grad_norm": 0.6096305847167969,
+      "learning_rate": 2.3848888189634778e-05,
+      "loss": 0.0449,
+      "step": 3510
+    },
+    {
+      "epoch": 2.136294493459081,
+      "grad_norm": 0.6761244535446167,
+      "learning_rate": 2.3836452682792734e-05,
+      "loss": 0.0474,
+      "step": 3511
+    },
+    {
+      "epoch": 2.1369029510191666,
+      "grad_norm": 0.6972514390945435,
+      "learning_rate": 2.382401746446191e-05,
+      "loss": 0.0652,
+      "step": 3512
+    },
+    {
+      "epoch": 2.1375114085792517,
+      "grad_norm": 0.6121935844421387,
+      "learning_rate": 2.3811582537725753e-05,
+      "loss": 0.0501,
+      "step": 3513
+    },
+    {
+      "epoch": 2.138119866139337,
+      "grad_norm": 0.6078847646713257,
+      "learning_rate": 2.379914790566759e-05,
+      "loss": 0.0568,
+      "step": 3514
+    },
+    {
+      "epoch": 2.138728323699422,
+      "grad_norm": 0.6310839653015137,
+      "learning_rate": 2.378671357137068e-05,
+      "loss": 0.0451,
+      "step": 3515
+    },
+    {
+      "epoch": 2.139336781259507,
+      "grad_norm": 0.48389577865600586,
+      "learning_rate": 2.377427953791824e-05,
+      "loss": 0.0449,
+      "step": 3516
+    },
+    {
+      "epoch": 2.1399452388195925,
+      "grad_norm": 0.6960240006446838,
+      "learning_rate": 2.3761845808393388e-05,
+      "loss": 0.0582,
+      "step": 3517
+    },
+    {
+      "epoch": 2.1405536963796776,
+      "grad_norm": 0.571576714515686,
+      "learning_rate": 2.3749412385879154e-05,
+      "loss": 0.0661,
+      "step": 3518
+    },
+    {
+      "epoch": 2.1411621539397627,
+      "grad_norm": 0.6685907244682312,
+      "learning_rate": 2.3736979273458535e-05,
+      "loss": 0.0555,
+      "step": 3519
+    },
+    {
+      "epoch": 2.141770611499848,
+      "grad_norm": 0.6551706194877625,
+      "learning_rate": 2.3724546474214406e-05,
+      "loss": 0.0392,
+      "step": 3520
+    },
+    {
+      "epoch": 2.142379069059933,
+      "grad_norm": 0.609714686870575,
+      "learning_rate": 2.371211399122958e-05,
+      "loss": 0.0674,
+      "step": 3521
+    },
+    {
+      "epoch": 2.1429875266200185,
+      "grad_norm": 0.5921576619148254,
+      "learning_rate": 2.3699681827586813e-05,
+      "loss": 0.0498,
+      "step": 3522
+    },
+    {
+      "epoch": 2.1435959841801036,
+      "grad_norm": 0.5523459315299988,
+      "learning_rate": 2.3687249986368752e-05,
+      "loss": 0.0502,
+      "step": 3523
+    },
+    {
+      "epoch": 2.1442044417401886,
+      "grad_norm": 0.6032785177230835,
+      "learning_rate": 2.3674818470657975e-05,
+      "loss": 0.0567,
+      "step": 3524
+    },
+    {
+      "epoch": 2.1448128993002737,
+      "grad_norm": 0.5920435786247253,
+      "learning_rate": 2.3662387283536978e-05,
+      "loss": 0.0389,
+      "step": 3525
+    },
+    {
+      "epoch": 2.145421356860359,
+      "grad_norm": 0.5291022062301636,
+      "learning_rate": 2.3649956428088187e-05,
+      "loss": 0.042,
+      "step": 3526
+    },
+    {
+      "epoch": 2.1460298144204444,
+      "grad_norm": 0.6911409497261047,
+      "learning_rate": 2.363752590739392e-05,
+      "loss": 0.0591,
+      "step": 3527
+    },
+    {
+      "epoch": 2.1466382719805295,
+      "grad_norm": 0.6609728336334229,
+      "learning_rate": 2.362509572453644e-05,
+      "loss": 0.0463,
+      "step": 3528
+    },
+    {
+      "epoch": 2.1472467295406146,
+      "grad_norm": 0.5278157591819763,
+      "learning_rate": 2.3612665882597915e-05,
+      "loss": 0.0378,
+      "step": 3529
+    },
+    {
+      "epoch": 2.1478551871006997,
+      "grad_norm": 0.7173718810081482,
+      "learning_rate": 2.360023638466042e-05,
+      "loss": 0.0511,
+      "step": 3530
+    },
+    {
+      "epoch": 2.1484636446607848,
+      "grad_norm": 0.5184733271598816,
+      "learning_rate": 2.3587807233805956e-05,
+      "loss": 0.0448,
+      "step": 3531
+    },
+    {
+      "epoch": 2.1490721022208703,
+      "grad_norm": 0.7443384528160095,
+      "learning_rate": 2.3575378433116445e-05,
+      "loss": 0.0522,
+      "step": 3532
+    },
+    {
+      "epoch": 2.1496805597809554,
+      "grad_norm": 0.5677869915962219,
+      "learning_rate": 2.356294998567369e-05,
+      "loss": 0.0391,
+      "step": 3533
+    },
+    {
+      "epoch": 2.1502890173410405,
+      "grad_norm": 0.7566191554069519,
+      "learning_rate": 2.3550521894559446e-05,
+      "loss": 0.0478,
+      "step": 3534
+    },
+    {
+      "epoch": 2.1508974749011256,
+      "grad_norm": 0.5744057297706604,
+      "learning_rate": 2.353809416285535e-05,
+      "loss": 0.0428,
+      "step": 3535
+    },
+    {
+      "epoch": 2.1515059324612107,
+      "grad_norm": 0.6628245115280151,
+      "learning_rate": 2.3525666793642968e-05,
+      "loss": 0.0547,
+      "step": 3536
+    },
+    {
+      "epoch": 2.1521143900212962,
+      "grad_norm": 0.5155208706855774,
+      "learning_rate": 2.351323979000377e-05,
+      "loss": 0.047,
+      "step": 3537
+    },
+    {
+      "epoch": 2.1527228475813813,
+      "grad_norm": 0.6521499752998352,
+      "learning_rate": 2.350081315501913e-05,
+      "loss": 0.0421,
+      "step": 3538
+    },
+    {
+      "epoch": 2.1533313051414664,
+      "grad_norm": 0.6431947946548462,
+      "learning_rate": 2.3488386891770346e-05,
+      "loss": 0.0441,
+      "step": 3539
+    },
+    {
+      "epoch": 2.1539397627015515,
+      "grad_norm": 0.6007315516471863,
+      "learning_rate": 2.3475961003338594e-05,
+      "loss": 0.052,
+      "step": 3540
+    },
+    {
+      "epoch": 2.1545482202616366,
+      "grad_norm": 0.5834375023841858,
+      "learning_rate": 2.3463535492804994e-05,
+      "loss": 0.056,
+      "step": 3541
+    },
+    {
+      "epoch": 2.155156677821722,
+      "grad_norm": 0.6027958989143372,
+      "learning_rate": 2.345111036325055e-05,
+      "loss": 0.0483,
+      "step": 3542
+    },
+    {
+      "epoch": 2.1557651353818073,
+      "grad_norm": 0.7618597745895386,
+      "learning_rate": 2.3438685617756174e-05,
+      "loss": 0.0605,
+      "step": 3543
+    },
+    {
+      "epoch": 2.1563735929418923,
+      "grad_norm": 0.6837668418884277,
+      "learning_rate": 2.342626125940268e-05,
+      "loss": 0.0551,
+      "step": 3544
+    },
+    {
+      "epoch": 2.1569820505019774,
+      "grad_norm": 0.5220887064933777,
+      "learning_rate": 2.341383729127081e-05,
+      "loss": 0.0433,
+      "step": 3545
+    },
+    {
+      "epoch": 2.1575905080620625,
+      "grad_norm": 0.7032208442687988,
+      "learning_rate": 2.3401413716441166e-05,
+      "loss": 0.066,
+      "step": 3546
+    },
+    {
+      "epoch": 2.158198965622148,
+      "grad_norm": 0.5578327178955078,
+      "learning_rate": 2.3388990537994296e-05,
+      "loss": 0.0519,
+      "step": 3547
+    },
+    {
+      "epoch": 2.158807423182233,
+      "grad_norm": 0.686423122882843,
+      "learning_rate": 2.3376567759010614e-05,
+      "loss": 0.0479,
+      "step": 3548
+    },
+    {
+      "epoch": 2.1594158807423183,
+      "grad_norm": 0.7522620558738708,
+      "learning_rate": 2.3364145382570462e-05,
+      "loss": 0.0713,
+      "step": 3549
+    },
+    {
+      "epoch": 2.1600243383024034,
+      "grad_norm": 0.5565488934516907,
+      "learning_rate": 2.3351723411754074e-05,
+      "loss": 0.0474,
+      "step": 3550
+    },
+    {
+      "epoch": 2.1606327958624885,
+      "grad_norm": 0.5260975956916809,
+      "learning_rate": 2.3339301849641573e-05,
+      "loss": 0.0464,
+      "step": 3551
+    },
+    {
+      "epoch": 2.161241253422574,
+      "grad_norm": 0.574340283870697,
+      "learning_rate": 2.332688069931299e-05,
+      "loss": 0.0425,
+      "step": 3552
+    },
+    {
+      "epoch": 2.161849710982659,
+      "grad_norm": 0.5723605751991272,
+      "learning_rate": 2.3314459963848264e-05,
+      "loss": 0.0519,
+      "step": 3553
+    },
+    {
+      "epoch": 2.162458168542744,
+      "grad_norm": 0.6394765377044678,
+      "learning_rate": 2.33020396463272e-05,
+      "loss": 0.0454,
+      "step": 3554
+    },
+    {
+      "epoch": 2.1630666261028293,
+      "grad_norm": 0.6276997923851013,
+      "learning_rate": 2.328961974982953e-05,
+      "loss": 0.0485,
+      "step": 3555
+    },
+    {
+      "epoch": 2.1636750836629144,
+      "grad_norm": 0.6199513077735901,
+      "learning_rate": 2.3277200277434876e-05,
+      "loss": 0.0526,
+      "step": 3556
+    },
+    {
+      "epoch": 2.164283541223,
+      "grad_norm": 0.5897301435470581,
+      "learning_rate": 2.3264781232222742e-05,
+      "loss": 0.0539,
+      "step": 3557
+    },
+    {
+      "epoch": 2.164891998783085,
+      "grad_norm": 0.5835275053977966,
+      "learning_rate": 2.3252362617272527e-05,
+      "loss": 0.0548,
+      "step": 3558
+    },
+    {
+      "epoch": 2.16550045634317,
+      "grad_norm": 0.5806435346603394,
+      "learning_rate": 2.3239944435663547e-05,
+      "loss": 0.0499,
+      "step": 3559
+    },
+    {
+      "epoch": 2.166108913903255,
+      "grad_norm": 0.6462282538414001,
+      "learning_rate": 2.3227526690474982e-05,
+      "loss": 0.0611,
+      "step": 3560
+    },
+    {
+      "epoch": 2.1667173714633403,
+      "grad_norm": 0.58769690990448,
+      "learning_rate": 2.321510938478591e-05,
+      "loss": 0.0462,
+      "step": 3561
+    },
+    {
+      "epoch": 2.1673258290234254,
+      "grad_norm": 0.5471524596214294,
+      "learning_rate": 2.3202692521675323e-05,
+      "loss": 0.0348,
+      "step": 3562
+    },
+    {
+      "epoch": 2.167934286583511,
+      "grad_norm": 0.5656521320343018,
+      "learning_rate": 2.3190276104222073e-05,
+      "loss": 0.055,
+      "step": 3563
+    },
+    {
+      "epoch": 2.168542744143596,
+      "grad_norm": 0.6103416085243225,
+      "learning_rate": 2.3177860135504907e-05,
+      "loss": 0.0562,
+      "step": 3564
+    },
+    {
+      "epoch": 2.169151201703681,
+      "grad_norm": 0.6096418499946594,
+      "learning_rate": 2.316544461860249e-05,
+      "loss": 0.0513,
+      "step": 3565
+    },
+    {
+      "epoch": 2.1697596592637662,
+      "grad_norm": 0.5852376818656921,
+      "learning_rate": 2.3153029556593335e-05,
+      "loss": 0.0505,
+      "step": 3566
+    },
+    {
+      "epoch": 2.1703681168238513,
+      "grad_norm": 0.6946520805358887,
+      "learning_rate": 2.3140614952555856e-05,
+      "loss": 0.0659,
+      "step": 3567
+    },
+    {
+      "epoch": 2.170976574383937,
+      "grad_norm": 0.546222448348999,
+      "learning_rate": 2.3128200809568375e-05,
+      "loss": 0.057,
+      "step": 3568
+    },
+    {
+      "epoch": 2.171585031944022,
+      "grad_norm": 0.5653393864631653,
+      "learning_rate": 2.3115787130709074e-05,
+      "loss": 0.0415,
+      "step": 3569
+    },
+    {
+      "epoch": 2.172193489504107,
+      "grad_norm": 0.5774003863334656,
+      "learning_rate": 2.3103373919056026e-05,
+      "loss": 0.0561,
+      "step": 3570
+    },
+    {
+      "epoch": 2.172801947064192,
+      "grad_norm": 0.5911126136779785,
+      "learning_rate": 2.309096117768718e-05,
+      "loss": 0.0538,
+      "step": 3571
+    },
+    {
+      "epoch": 2.1734104046242773,
+      "grad_norm": 0.549182653427124,
+      "learning_rate": 2.3078548909680403e-05,
+      "loss": 0.0474,
+      "step": 3572
+    },
+    {
+      "epoch": 2.174018862184363,
+      "grad_norm": 0.5475966334342957,
+      "learning_rate": 2.3066137118113405e-05,
+      "loss": 0.0486,
+      "step": 3573
+    },
+    {
+      "epoch": 2.174627319744448,
+      "grad_norm": 0.6825101375579834,
+      "learning_rate": 2.305372580606378e-05,
+      "loss": 0.0569,
+      "step": 3574
+    },
+    {
+      "epoch": 2.175235777304533,
+      "grad_norm": 0.5372380614280701,
+      "learning_rate": 2.3041314976609043e-05,
+      "loss": 0.0525,
+      "step": 3575
+    },
+    {
+      "epoch": 2.175844234864618,
+      "grad_norm": 0.5879507064819336,
+      "learning_rate": 2.3028904632826555e-05,
+      "loss": 0.0485,
+      "step": 3576
+    },
+    {
+      "epoch": 2.176452692424703,
+      "grad_norm": 0.6694579124450684,
+      "learning_rate": 2.301649477779354e-05,
+      "loss": 0.0516,
+      "step": 3577
+    },
+    {
+      "epoch": 2.1770611499847887,
+      "grad_norm": 0.48730021715164185,
+      "learning_rate": 2.300408541458716e-05,
+      "loss": 0.0444,
+      "step": 3578
+    },
+    {
+      "epoch": 2.177669607544874,
+      "grad_norm": 0.6008870005607605,
+      "learning_rate": 2.29916765462844e-05,
+      "loss": 0.0413,
+      "step": 3579
+    },
+    {
+      "epoch": 2.178278065104959,
+      "grad_norm": 0.557727038860321,
+      "learning_rate": 2.2979268175962134e-05,
+      "loss": 0.0375,
+      "step": 3580
+    },
+    {
+      "epoch": 2.178886522665044,
+      "grad_norm": 0.5483887195587158,
+      "learning_rate": 2.2966860306697148e-05,
+      "loss": 0.0426,
+      "step": 3581
+    },
+    {
+      "epoch": 2.179494980225129,
+      "grad_norm": 0.6324334144592285,
+      "learning_rate": 2.2954452941566057e-05,
+      "loss": 0.0419,
+      "step": 3582
+    },
+    {
+      "epoch": 2.1801034377852146,
+      "grad_norm": 0.643545925617218,
+      "learning_rate": 2.2942046083645375e-05,
+      "loss": 0.0425,
+      "step": 3583
+    },
+    {
+      "epoch": 2.1807118953452997,
+      "grad_norm": 0.5663473010063171,
+      "learning_rate": 2.292963973601147e-05,
+      "loss": 0.0356,
+      "step": 3584
+    },
+    {
+      "epoch": 2.181320352905385,
+      "grad_norm": 0.6918662786483765,
+      "learning_rate": 2.291723390174063e-05,
+      "loss": 0.0698,
+      "step": 3585
+    },
+    {
+      "epoch": 2.18192881046547,
+      "grad_norm": 0.8219648003578186,
+      "learning_rate": 2.2904828583908967e-05,
+      "loss": 0.0711,
+      "step": 3586
+    },
+    {
+      "epoch": 2.182537268025555,
+      "grad_norm": 0.5029035806655884,
+      "learning_rate": 2.289242378559247e-05,
+      "loss": 0.0496,
+      "step": 3587
+    },
+    {
+      "epoch": 2.1831457255856406,
+      "grad_norm": 0.5251539945602417,
+      "learning_rate": 2.288001950986704e-05,
+      "loss": 0.0411,
+      "step": 3588
+    },
+    {
+      "epoch": 2.1837541831457257,
+      "grad_norm": 0.7362745404243469,
+      "learning_rate": 2.2867615759808403e-05,
+      "loss": 0.0523,
+      "step": 3589
+    },
+    {
+      "epoch": 2.1843626407058108,
+      "grad_norm": 0.5987080335617065,
+      "learning_rate": 2.2855212538492167e-05,
+      "loss": 0.0515,
+      "step": 3590
+    },
+    {
+      "epoch": 2.184971098265896,
+      "grad_norm": 0.5399122834205627,
+      "learning_rate": 2.2842809848993834e-05,
+      "loss": 0.0555,
+      "step": 3591
+    },
+    {
+      "epoch": 2.185579555825981,
+      "grad_norm": 0.6066261529922485,
+      "learning_rate": 2.2830407694388743e-05,
+      "loss": 0.0404,
+      "step": 3592
+    },
+    {
+      "epoch": 2.1861880133860665,
+      "grad_norm": 0.5191651582717896,
+      "learning_rate": 2.281800607775211e-05,
+      "loss": 0.0519,
+      "step": 3593
+    },
+    {
+      "epoch": 2.1867964709461516,
+      "grad_norm": 0.6118345856666565,
+      "learning_rate": 2.2805605002159007e-05,
+      "loss": 0.05,
+      "step": 3594
+    },
+    {
+      "epoch": 2.1874049285062367,
+      "grad_norm": 0.6577736139297485,
+      "learning_rate": 2.2793204470684406e-05,
+      "loss": 0.0558,
+      "step": 3595
+    },
+    {
+      "epoch": 2.188013386066322,
+      "grad_norm": 0.5983335375785828,
+      "learning_rate": 2.2780804486403115e-05,
+      "loss": 0.064,
+      "step": 3596
+    },
+    {
+      "epoch": 2.188621843626407,
+      "grad_norm": 0.6087679266929626,
+      "learning_rate": 2.2768405052389802e-05,
+      "loss": 0.0561,
+      "step": 3597
+    },
+    {
+      "epoch": 2.1892303011864924,
+      "grad_norm": 0.5540795922279358,
+      "learning_rate": 2.2756006171719027e-05,
+      "loss": 0.0431,
+      "step": 3598
+    },
+    {
+      "epoch": 2.1898387587465775,
+      "grad_norm": 0.5119609236717224,
+      "learning_rate": 2.2743607847465188e-05,
+      "loss": 0.0466,
+      "step": 3599
+    },
+    {
+      "epoch": 2.1904472163066626,
+      "grad_norm": 0.5084932446479797,
+      "learning_rate": 2.273121008270254e-05,
+      "loss": 0.0535,
+      "step": 3600
+    },
+    {
+      "epoch": 2.1910556738667477,
+      "grad_norm": 0.5782137513160706,
+      "learning_rate": 2.2718812880505232e-05,
+      "loss": 0.0635,
+      "step": 3601
+    },
+    {
+      "epoch": 2.191664131426833,
+      "grad_norm": 0.6178430318832397,
+      "learning_rate": 2.2706416243947248e-05,
+      "loss": 0.0442,
+      "step": 3602
+    },
+    {
+      "epoch": 2.1922725889869183,
+      "grad_norm": 0.5488747954368591,
+      "learning_rate": 2.2694020176102427e-05,
+      "loss": 0.0394,
+      "step": 3603
+    },
+    {
+      "epoch": 2.1928810465470034,
+      "grad_norm": 0.6640757322311401,
+      "learning_rate": 2.268162468004449e-05,
+      "loss": 0.0524,
+      "step": 3604
+    },
+    {
+      "epoch": 2.1934895041070885,
+      "grad_norm": 0.5369768738746643,
+      "learning_rate": 2.2669229758846998e-05,
+      "loss": 0.044,
+      "step": 3605
+    },
+    {
+      "epoch": 2.1940979616671736,
+      "grad_norm": 0.554397702217102,
+      "learning_rate": 2.265683541558338e-05,
+      "loss": 0.0384,
+      "step": 3606
+    },
+    {
+      "epoch": 2.1947064192272587,
+      "grad_norm": 0.6047532558441162,
+      "learning_rate": 2.2644441653326892e-05,
+      "loss": 0.0555,
+      "step": 3607
+    },
+    {
+      "epoch": 2.1953148767873443,
+      "grad_norm": 0.5616478323936462,
+      "learning_rate": 2.2632048475150705e-05,
+      "loss": 0.0516,
+      "step": 3608
+    },
+    {
+      "epoch": 2.1959233343474294,
+      "grad_norm": 0.5725663900375366,
+      "learning_rate": 2.2619655884127793e-05,
+      "loss": 0.0551,
+      "step": 3609
+    },
+    {
+      "epoch": 2.1965317919075145,
+      "grad_norm": 0.6617404222488403,
+      "learning_rate": 2.2607263883330994e-05,
+      "loss": 0.0457,
+      "step": 3610
+    },
+    {
+      "epoch": 2.1971402494675996,
+      "grad_norm": 1.2666044235229492,
+      "learning_rate": 2.259487247583303e-05,
+      "loss": 0.078,
+      "step": 3611
+    },
+    {
+      "epoch": 2.1977487070276847,
+      "grad_norm": 0.7400673031806946,
+      "learning_rate": 2.258248166470644e-05,
+      "loss": 0.0534,
+      "step": 3612
+    },
+    {
+      "epoch": 2.19835716458777,
+      "grad_norm": 0.6314421892166138,
+      "learning_rate": 2.257009145302362e-05,
+      "loss": 0.039,
+      "step": 3613
+    },
+    {
+      "epoch": 2.1989656221478553,
+      "grad_norm": 0.5742847919464111,
+      "learning_rate": 2.2557701843856847e-05,
+      "loss": 0.0468,
+      "step": 3614
+    },
+    {
+      "epoch": 2.1995740797079404,
+      "grad_norm": 0.66370689868927,
+      "learning_rate": 2.2545312840278214e-05,
+      "loss": 0.0853,
+      "step": 3615
+    },
+    {
+      "epoch": 2.2001825372680255,
+      "grad_norm": 0.5838887095451355,
+      "learning_rate": 2.2532924445359686e-05,
+      "loss": 0.0373,
+      "step": 3616
+    },
+    {
+      "epoch": 2.2007909948281106,
+      "grad_norm": 0.6079072952270508,
+      "learning_rate": 2.252053666217305e-05,
+      "loss": 0.0372,
+      "step": 3617
+    },
+    {
+      "epoch": 2.201399452388196,
+      "grad_norm": 0.7246667742729187,
+      "learning_rate": 2.250814949378998e-05,
+      "loss": 0.057,
+      "step": 3618
+    },
+    {
+      "epoch": 2.202007909948281,
+      "grad_norm": 0.5479733943939209,
+      "learning_rate": 2.2495762943281974e-05,
+      "loss": 0.0404,
+      "step": 3619
+    },
+    {
+      "epoch": 2.2026163675083663,
+      "grad_norm": 0.5117501020431519,
+      "learning_rate": 2.248337701372036e-05,
+      "loss": 0.0419,
+      "step": 3620
+    },
+    {
+      "epoch": 2.2032248250684514,
+      "grad_norm": 0.735734224319458,
+      "learning_rate": 2.2470991708176364e-05,
+      "loss": 0.0618,
+      "step": 3621
+    },
+    {
+      "epoch": 2.2038332826285365,
+      "grad_norm": 0.4875301122665405,
+      "learning_rate": 2.245860702972101e-05,
+      "loss": 0.0497,
+      "step": 3622
+    },
+    {
+      "epoch": 2.204441740188622,
+      "grad_norm": 0.6231479644775391,
+      "learning_rate": 2.244622298142517e-05,
+      "loss": 0.0457,
+      "step": 3623
+    },
+    {
+      "epoch": 2.205050197748707,
+      "grad_norm": 0.5008909106254578,
+      "learning_rate": 2.2433839566359593e-05,
+      "loss": 0.0479,
+      "step": 3624
+    },
+    {
+      "epoch": 2.2056586553087922,
+      "grad_norm": 0.49350735545158386,
+      "learning_rate": 2.2421456787594845e-05,
+      "loss": 0.0543,
+      "step": 3625
+    },
+    {
+      "epoch": 2.2062671128688773,
+      "grad_norm": 0.6941964626312256,
+      "learning_rate": 2.240907464820132e-05,
+      "loss": 0.0747,
+      "step": 3626
+    },
+    {
+      "epoch": 2.2068755704289624,
+      "grad_norm": 0.5441345572471619,
+      "learning_rate": 2.2396693151249303e-05,
+      "loss": 0.047,
+      "step": 3627
+    },
+    {
+      "epoch": 2.207484027989048,
+      "grad_norm": 0.5408318042755127,
+      "learning_rate": 2.2384312299808868e-05,
+      "loss": 0.0556,
+      "step": 3628
+    },
+    {
+      "epoch": 2.208092485549133,
+      "grad_norm": 0.6332383155822754,
+      "learning_rate": 2.2371932096949957e-05,
+      "loss": 0.0488,
+      "step": 3629
+    },
+    {
+      "epoch": 2.208700943109218,
+      "grad_norm": 0.6663236021995544,
+      "learning_rate": 2.2359552545742332e-05,
+      "loss": 0.0381,
+      "step": 3630
+    },
+    {
+      "epoch": 2.2093094006693033,
+      "grad_norm": 0.5764215588569641,
+      "learning_rate": 2.2347173649255627e-05,
+      "loss": 0.0576,
+      "step": 3631
+    },
+    {
+      "epoch": 2.2099178582293884,
+      "grad_norm": 0.4953424632549286,
+      "learning_rate": 2.2334795410559283e-05,
+      "loss": 0.0395,
+      "step": 3632
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.5949841737747192,
+      "learning_rate": 2.2322417832722574e-05,
+      "loss": 0.0563,
+      "step": 3633
+    },
+    {
+      "epoch": 2.211134773349559,
+      "grad_norm": 0.5839807391166687,
+      "learning_rate": 2.2310040918814647e-05,
+      "loss": 0.0493,
+      "step": 3634
+    },
+    {
+      "epoch": 2.211743230909644,
+      "grad_norm": 0.648003101348877,
+      "learning_rate": 2.2297664671904447e-05,
+      "loss": 0.0691,
+      "step": 3635
+    },
+    {
+      "epoch": 2.212351688469729,
+      "grad_norm": 0.582639217376709,
+      "learning_rate": 2.2285289095060764e-05,
+      "loss": 0.0542,
+      "step": 3636
+    },
+    {
+      "epoch": 2.2129601460298143,
+      "grad_norm": 0.6050490140914917,
+      "learning_rate": 2.2272914191352234e-05,
+      "loss": 0.0421,
+      "step": 3637
+    },
+    {
+      "epoch": 2.2135686035898994,
+      "grad_norm": 0.8048490285873413,
+      "learning_rate": 2.2260539963847317e-05,
+      "loss": 0.0838,
+      "step": 3638
+    },
+    {
+      "epoch": 2.214177061149985,
+      "grad_norm": 0.6212171316146851,
+      "learning_rate": 2.2248166415614305e-05,
+      "loss": 0.0601,
+      "step": 3639
+    },
+    {
+      "epoch": 2.21478551871007,
+      "grad_norm": 0.5588234663009644,
+      "learning_rate": 2.223579354972131e-05,
+      "loss": 0.0384,
+      "step": 3640
+    },
+    {
+      "epoch": 2.215393976270155,
+      "grad_norm": 0.5870838761329651,
+      "learning_rate": 2.2223421369236304e-05,
+      "loss": 0.0476,
+      "step": 3641
+    },
+    {
+      "epoch": 2.21600243383024,
+      "grad_norm": 0.6339545845985413,
+      "learning_rate": 2.2211049877227065e-05,
+      "loss": 0.0791,
+      "step": 3642
+    },
+    {
+      "epoch": 2.2166108913903253,
+      "grad_norm": 0.6535237431526184,
+      "learning_rate": 2.2198679076761196e-05,
+      "loss": 0.0458,
+      "step": 3643
+    },
+    {
+      "epoch": 2.217219348950411,
+      "grad_norm": 0.4435776174068451,
+      "learning_rate": 2.2186308970906166e-05,
+      "loss": 0.0386,
+      "step": 3644
+    },
+    {
+      "epoch": 2.217827806510496,
+      "grad_norm": 0.6395978331565857,
+      "learning_rate": 2.217393956272923e-05,
+      "loss": 0.0606,
+      "step": 3645
+    },
+    {
+      "epoch": 2.218436264070581,
+      "grad_norm": 0.6212596297264099,
+      "learning_rate": 2.2161570855297474e-05,
+      "loss": 0.0597,
+      "step": 3646
+    },
+    {
+      "epoch": 2.219044721630666,
+      "grad_norm": 0.49921971559524536,
+      "learning_rate": 2.2149202851677842e-05,
+      "loss": 0.041,
+      "step": 3647
+    },
+    {
+      "epoch": 2.2196531791907512,
+      "grad_norm": 0.45042991638183594,
+      "learning_rate": 2.213683555493708e-05,
+      "loss": 0.0454,
+      "step": 3648
+    },
+    {
+      "epoch": 2.2202616367508368,
+      "grad_norm": 0.5090251564979553,
+      "learning_rate": 2.2124468968141746e-05,
+      "loss": 0.0528,
+      "step": 3649
+    },
+    {
+      "epoch": 2.220870094310922,
+      "grad_norm": 0.6496664881706238,
+      "learning_rate": 2.211210309435826e-05,
+      "loss": 0.0516,
+      "step": 3650
+    },
+    {
+      "epoch": 2.221478551871007,
+      "grad_norm": 0.5032176375389099,
+      "learning_rate": 2.2099737936652834e-05,
+      "loss": 0.0365,
+      "step": 3651
+    },
+    {
+      "epoch": 2.222087009431092,
+      "grad_norm": 0.5386142730712891,
+      "learning_rate": 2.2087373498091505e-05,
+      "loss": 0.0489,
+      "step": 3652
+    },
+    {
+      "epoch": 2.222695466991177,
+      "grad_norm": 0.5866867303848267,
+      "learning_rate": 2.2075009781740144e-05,
+      "loss": 0.0507,
+      "step": 3653
+    },
+    {
+      "epoch": 2.2233039245512627,
+      "grad_norm": 0.6343091726303101,
+      "learning_rate": 2.2062646790664443e-05,
+      "loss": 0.05,
+      "step": 3654
+    },
+    {
+      "epoch": 2.223912382111348,
+      "grad_norm": 0.49765875935554504,
+      "learning_rate": 2.2050284527929897e-05,
+      "loss": 0.0349,
+      "step": 3655
+    },
+    {
+      "epoch": 2.224520839671433,
+      "grad_norm": 0.6474698185920715,
+      "learning_rate": 2.203792299660184e-05,
+      "loss": 0.0544,
+      "step": 3656
+    },
+    {
+      "epoch": 2.225129297231518,
+      "grad_norm": 0.5999079346656799,
+      "learning_rate": 2.202556219974542e-05,
+      "loss": 0.0387,
+      "step": 3657
+    },
+    {
+      "epoch": 2.225737754791603,
+      "grad_norm": 0.5736773014068604,
+      "learning_rate": 2.2013202140425584e-05,
+      "loss": 0.0553,
+      "step": 3658
+    },
+    {
+      "epoch": 2.2263462123516886,
+      "grad_norm": 0.6365640163421631,
+      "learning_rate": 2.2000842821707122e-05,
+      "loss": 0.046,
+      "step": 3659
+    },
+    {
+      "epoch": 2.2269546699117737,
+      "grad_norm": 0.5798011422157288,
+      "learning_rate": 2.198848424665464e-05,
+      "loss": 0.0466,
+      "step": 3660
+    },
+    {
+      "epoch": 2.227563127471859,
+      "grad_norm": 0.599982738494873,
+      "learning_rate": 2.197612641833253e-05,
+      "loss": 0.0608,
+      "step": 3661
+    },
+    {
+      "epoch": 2.228171585031944,
+      "grad_norm": 0.5781305432319641,
+      "learning_rate": 2.196376933980503e-05,
+      "loss": 0.0446,
+      "step": 3662
+    },
+    {
+      "epoch": 2.228780042592029,
+      "grad_norm": 0.6489582061767578,
+      "learning_rate": 2.1951413014136177e-05,
+      "loss": 0.0562,
+      "step": 3663
+    },
+    {
+      "epoch": 2.2293885001521145,
+      "grad_norm": 0.5983100533485413,
+      "learning_rate": 2.1939057444389822e-05,
+      "loss": 0.046,
+      "step": 3664
+    },
+    {
+      "epoch": 2.2299969577121996,
+      "grad_norm": 0.5980080366134644,
+      "learning_rate": 2.192670263362964e-05,
+      "loss": 0.0658,
+      "step": 3665
+    },
+    {
+      "epoch": 2.2306054152722847,
+      "grad_norm": 0.5348019003868103,
+      "learning_rate": 2.19143485849191e-05,
+      "loss": 0.0498,
+      "step": 3666
+    },
+    {
+      "epoch": 2.23121387283237,
+      "grad_norm": 0.5412232875823975,
+      "learning_rate": 2.1901995301321493e-05,
+      "loss": 0.0397,
+      "step": 3667
+    },
+    {
+      "epoch": 2.231822330392455,
+      "grad_norm": 0.5473434925079346,
+      "learning_rate": 2.1889642785899926e-05,
+      "loss": 0.0388,
+      "step": 3668
+    },
+    {
+      "epoch": 2.2324307879525405,
+      "grad_norm": 0.8068600296974182,
+      "learning_rate": 2.1877291041717294e-05,
+      "loss": 0.0524,
+      "step": 3669
+    },
+    {
+      "epoch": 2.2330392455126256,
+      "grad_norm": 0.5853879451751709,
+      "learning_rate": 2.1864940071836326e-05,
+      "loss": 0.0582,
+      "step": 3670
+    },
+    {
+      "epoch": 2.2336477030727107,
+      "grad_norm": 0.5249461531639099,
+      "learning_rate": 2.1852589879319547e-05,
+      "loss": 0.048,
+      "step": 3671
+    },
+    {
+      "epoch": 2.2342561606327958,
+      "grad_norm": 0.5302576422691345,
+      "learning_rate": 2.1840240467229283e-05,
+      "loss": 0.0553,
+      "step": 3672
+    },
+    {
+      "epoch": 2.234864618192881,
+      "grad_norm": 0.5220465660095215,
+      "learning_rate": 2.1827891838627687e-05,
+      "loss": 0.0431,
+      "step": 3673
+    },
+    {
+      "epoch": 2.2354730757529664,
+      "grad_norm": 0.5852304697036743,
+      "learning_rate": 2.1815543996576688e-05,
+      "loss": 0.0496,
+      "step": 3674
+    },
+    {
+      "epoch": 2.2360815333130515,
+      "grad_norm": 0.6456661820411682,
+      "learning_rate": 2.1803196944138045e-05,
+      "loss": 0.0519,
+      "step": 3675
+    },
+    {
+      "epoch": 2.2366899908731366,
+      "grad_norm": 0.5966864824295044,
+      "learning_rate": 2.1790850684373305e-05,
+      "loss": 0.0358,
+      "step": 3676
+    },
+    {
+      "epoch": 2.2372984484332217,
+      "grad_norm": 0.6074363589286804,
+      "learning_rate": 2.1778505220343836e-05,
+      "loss": 0.0507,
+      "step": 3677
+    },
+    {
+      "epoch": 2.2379069059933068,
+      "grad_norm": 0.5367887616157532,
+      "learning_rate": 2.17661605551108e-05,
+      "loss": 0.0602,
+      "step": 3678
+    },
+    {
+      "epoch": 2.2385153635533923,
+      "grad_norm": 0.5965633392333984,
+      "learning_rate": 2.175381669173514e-05,
+      "loss": 0.043,
+      "step": 3679
+    },
+    {
+      "epoch": 2.2391238211134774,
+      "grad_norm": 0.5841217041015625,
+      "learning_rate": 2.174147363327764e-05,
+      "loss": 0.0536,
+      "step": 3680
+    },
+    {
+      "epoch": 2.2397322786735625,
+      "grad_norm": 0.5816332101821899,
+      "learning_rate": 2.1729131382798858e-05,
+      "loss": 0.0578,
+      "step": 3681
+    },
+    {
+      "epoch": 2.2403407362336476,
+      "grad_norm": 0.5089899301528931,
+      "learning_rate": 2.1716789943359155e-05,
+      "loss": 0.0424,
+      "step": 3682
+    },
+    {
+      "epoch": 2.2409491937937327,
+      "grad_norm": 1.4226927757263184,
+      "learning_rate": 2.1704449318018692e-05,
+      "loss": 0.0814,
+      "step": 3683
+    },
+    {
+      "epoch": 2.2415576513538182,
+      "grad_norm": 0.6067140102386475,
+      "learning_rate": 2.1692109509837442e-05,
+      "loss": 0.0553,
+      "step": 3684
+    },
+    {
+      "epoch": 2.2421661089139033,
+      "grad_norm": 0.504880964756012,
+      "learning_rate": 2.167977052187515e-05,
+      "loss": 0.0329,
+      "step": 3685
+    },
+    {
+      "epoch": 2.2427745664739884,
+      "grad_norm": 0.5495906472206116,
+      "learning_rate": 2.1667432357191364e-05,
+      "loss": 0.0526,
+      "step": 3686
+    },
+    {
+      "epoch": 2.2433830240340735,
+      "grad_norm": 0.6342142820358276,
+      "learning_rate": 2.1655095018845455e-05,
+      "loss": 0.0491,
+      "step": 3687
+    },
+    {
+      "epoch": 2.2439914815941586,
+      "grad_norm": 0.5818422436714172,
+      "learning_rate": 2.1642758509896562e-05,
+      "loss": 0.0383,
+      "step": 3688
+    },
+    {
+      "epoch": 2.244599939154244,
+      "grad_norm": 0.6000524163246155,
+      "learning_rate": 2.1630422833403613e-05,
+      "loss": 0.0517,
+      "step": 3689
+    },
+    {
+      "epoch": 2.2452083967143293,
+      "grad_norm": 0.5182064175605774,
+      "learning_rate": 2.1618087992425356e-05,
+      "loss": 0.0441,
+      "step": 3690
+    },
+    {
+      "epoch": 2.2458168542744144,
+      "grad_norm": 0.6751769781112671,
+      "learning_rate": 2.1605753990020315e-05,
+      "loss": 0.0608,
+      "step": 3691
+    },
+    {
+      "epoch": 2.2464253118344994,
+      "grad_norm": 0.6493861079216003,
+      "learning_rate": 2.1593420829246794e-05,
+      "loss": 0.0533,
+      "step": 3692
+    },
+    {
+      "epoch": 2.2470337693945845,
+      "grad_norm": 0.6242604851722717,
+      "learning_rate": 2.1581088513162923e-05,
+      "loss": 0.0522,
+      "step": 3693
+    },
+    {
+      "epoch": 2.24764222695467,
+      "grad_norm": 0.5824303030967712,
+      "learning_rate": 2.1568757044826595e-05,
+      "loss": 0.0507,
+      "step": 3694
+    },
+    {
+      "epoch": 2.248250684514755,
+      "grad_norm": 0.6880918741226196,
+      "learning_rate": 2.1556426427295488e-05,
+      "loss": 0.0582,
+      "step": 3695
+    },
+    {
+      "epoch": 2.2488591420748403,
+      "grad_norm": 0.6002342104911804,
+      "learning_rate": 2.1544096663627104e-05,
+      "loss": 0.053,
+      "step": 3696
+    },
+    {
+      "epoch": 2.2494675996349254,
+      "grad_norm": 0.5104690790176392,
+      "learning_rate": 2.1531767756878696e-05,
+      "loss": 0.0394,
+      "step": 3697
+    },
+    {
+      "epoch": 2.2500760571950105,
+      "grad_norm": 0.5821288228034973,
+      "learning_rate": 2.151943971010732e-05,
+      "loss": 0.0422,
+      "step": 3698
+    },
+    {
+      "epoch": 2.250684514755096,
+      "grad_norm": 0.6568170785903931,
+      "learning_rate": 2.150711252636981e-05,
+      "loss": 0.0554,
+      "step": 3699
+    },
+    {
+      "epoch": 2.251292972315181,
+      "grad_norm": 0.5166933536529541,
+      "learning_rate": 2.1494786208722814e-05,
+      "loss": 0.0454,
+      "step": 3700
+    },
+    {
+      "epoch": 2.251901429875266,
+      "grad_norm": 0.6102219223976135,
+      "learning_rate": 2.1482460760222733e-05,
+      "loss": 0.0525,
+      "step": 3701
+    },
+    {
+      "epoch": 2.2525098874353513,
+      "grad_norm": 0.5432948470115662,
+      "learning_rate": 2.1470136183925755e-05,
+      "loss": 0.0538,
+      "step": 3702
+    },
+    {
+      "epoch": 2.2531183449954364,
+      "grad_norm": 0.5394953489303589,
+      "learning_rate": 2.1457812482887882e-05,
+      "loss": 0.0412,
+      "step": 3703
+    },
+    {
+      "epoch": 2.253726802555522,
+      "grad_norm": 0.5707597136497498,
+      "learning_rate": 2.1445489660164868e-05,
+      "loss": 0.0498,
+      "step": 3704
+    },
+    {
+      "epoch": 2.254335260115607,
+      "grad_norm": 0.5889186263084412,
+      "learning_rate": 2.1433167718812247e-05,
+      "loss": 0.0554,
+      "step": 3705
+    },
+    {
+      "epoch": 2.254943717675692,
+      "grad_norm": 0.5888684391975403,
+      "learning_rate": 2.142084666188537e-05,
+      "loss": 0.0512,
+      "step": 3706
+    },
+    {
+      "epoch": 2.2555521752357772,
+      "grad_norm": 0.4928998053073883,
+      "learning_rate": 2.1408526492439336e-05,
+      "loss": 0.0578,
+      "step": 3707
+    },
+    {
+      "epoch": 2.2561606327958623,
+      "grad_norm": 0.5299960374832153,
+      "learning_rate": 2.139620721352903e-05,
+      "loss": 0.0405,
+      "step": 3708
+    },
+    {
+      "epoch": 2.256769090355948,
+      "grad_norm": 0.5291889905929565,
+      "learning_rate": 2.138388882820911e-05,
+      "loss": 0.0555,
+      "step": 3709
+    },
+    {
+      "epoch": 2.257377547916033,
+      "grad_norm": 0.6831831932067871,
+      "learning_rate": 2.1371571339534046e-05,
+      "loss": 0.0863,
+      "step": 3710
+    },
+    {
+      "epoch": 2.257986005476118,
+      "grad_norm": 0.5145639181137085,
+      "learning_rate": 2.135925475055805e-05,
+      "loss": 0.0474,
+      "step": 3711
+    },
+    {
+      "epoch": 2.258594463036203,
+      "grad_norm": 0.5370884537696838,
+      "learning_rate": 2.134693906433511e-05,
+      "loss": 0.0416,
+      "step": 3712
+    },
+    {
+      "epoch": 2.2592029205962882,
+      "grad_norm": 0.5465013384819031,
+      "learning_rate": 2.1334624283919026e-05,
+      "loss": 0.0537,
+      "step": 3713
+    },
+    {
+      "epoch": 2.259811378156374,
+      "grad_norm": 0.5709042549133301,
+      "learning_rate": 2.132231041236334e-05,
+      "loss": 0.0366,
+      "step": 3714
+    },
+    {
+      "epoch": 2.260419835716459,
+      "grad_norm": 0.6409594416618347,
+      "learning_rate": 2.1309997452721366e-05,
+      "loss": 0.0586,
+      "step": 3715
+    },
+    {
+      "epoch": 2.261028293276544,
+      "grad_norm": 0.634441077709198,
+      "learning_rate": 2.129768540804623e-05,
+      "loss": 0.0634,
+      "step": 3716
+    },
+    {
+      "epoch": 2.261636750836629,
+      "grad_norm": 0.5557528734207153,
+      "learning_rate": 2.128537428139079e-05,
+      "loss": 0.0429,
+      "step": 3717
+    },
+    {
+      "epoch": 2.262245208396714,
+      "grad_norm": 0.678367018699646,
+      "learning_rate": 2.1273064075807686e-05,
+      "loss": 0.0686,
+      "step": 3718
+    },
+    {
+      "epoch": 2.2628536659567997,
+      "grad_norm": 0.5972436666488647,
+      "learning_rate": 2.1260754794349354e-05,
+      "loss": 0.0461,
+      "step": 3719
+    },
+    {
+      "epoch": 2.263462123516885,
+      "grad_norm": 0.664558470249176,
+      "learning_rate": 2.1248446440067976e-05,
+      "loss": 0.0467,
+      "step": 3720
+    },
+    {
+      "epoch": 2.26407058107697,
+      "grad_norm": 0.6116575598716736,
+      "learning_rate": 2.1236139016015507e-05,
+      "loss": 0.0517,
+      "step": 3721
+    },
+    {
+      "epoch": 2.264679038637055,
+      "grad_norm": 0.5580343008041382,
+      "learning_rate": 2.1223832525243663e-05,
+      "loss": 0.0533,
+      "step": 3722
+    },
+    {
+      "epoch": 2.26528749619714,
+      "grad_norm": 0.5420104265213013,
+      "learning_rate": 2.1211526970803973e-05,
+      "loss": 0.039,
+      "step": 3723
+    },
+    {
+      "epoch": 2.2658959537572256,
+      "grad_norm": 1.199730634689331,
+      "learning_rate": 2.1199222355747674e-05,
+      "loss": 0.0618,
+      "step": 3724
+    },
+    {
+      "epoch": 2.2665044113173107,
+      "grad_norm": 0.6017125844955444,
+      "learning_rate": 2.1186918683125802e-05,
+      "loss": 0.0559,
+      "step": 3725
+    },
+    {
+      "epoch": 2.267112868877396,
+      "grad_norm": 0.5252535343170166,
+      "learning_rate": 2.117461595598917e-05,
+      "loss": 0.0335,
+      "step": 3726
+    },
+    {
+      "epoch": 2.267721326437481,
+      "grad_norm": 0.6079027652740479,
+      "learning_rate": 2.1162314177388327e-05,
+      "loss": 0.0508,
+      "step": 3727
+    },
+    {
+      "epoch": 2.268329783997566,
+      "grad_norm": 0.49775218963623047,
+      "learning_rate": 2.1150013350373594e-05,
+      "loss": 0.0402,
+      "step": 3728
+    },
+    {
+      "epoch": 2.2689382415576516,
+      "grad_norm": 0.5613334774971008,
+      "learning_rate": 2.113771347799509e-05,
+      "loss": 0.0483,
+      "step": 3729
+    },
+    {
+      "epoch": 2.2695466991177367,
+      "grad_norm": 0.5892758965492249,
+      "learning_rate": 2.1125414563302654e-05,
+      "loss": 0.0589,
+      "step": 3730
+    },
+    {
+      "epoch": 2.2701551566778218,
+      "grad_norm": 0.6577876806259155,
+      "learning_rate": 2.1113116609345893e-05,
+      "loss": 0.0417,
+      "step": 3731
+    },
+    {
+      "epoch": 2.270763614237907,
+      "grad_norm": 0.6426102519035339,
+      "learning_rate": 2.110081961917421e-05,
+      "loss": 0.0514,
+      "step": 3732
+    },
+    {
+      "epoch": 2.271372071797992,
+      "grad_norm": 0.5062007308006287,
+      "learning_rate": 2.108852359583674e-05,
+      "loss": 0.0332,
+      "step": 3733
+    },
+    {
+      "epoch": 2.2719805293580775,
+      "grad_norm": 0.5272140502929688,
+      "learning_rate": 2.1076228542382376e-05,
+      "loss": 0.0393,
+      "step": 3734
+    },
+    {
+      "epoch": 2.2725889869181626,
+      "grad_norm": 0.6248688697814941,
+      "learning_rate": 2.1063934461859774e-05,
+      "loss": 0.0518,
+      "step": 3735
+    },
+    {
+      "epoch": 2.2731974444782477,
+      "grad_norm": 0.6348571181297302,
+      "learning_rate": 2.105164135731737e-05,
+      "loss": 0.0552,
+      "step": 3736
+    },
+    {
+      "epoch": 2.2738059020383328,
+      "grad_norm": 0.5358805656433105,
+      "learning_rate": 2.1039349231803337e-05,
+      "loss": 0.0462,
+      "step": 3737
+    },
+    {
+      "epoch": 2.274414359598418,
+      "grad_norm": 0.5874552130699158,
+      "learning_rate": 2.1027058088365593e-05,
+      "loss": 0.0503,
+      "step": 3738
+    },
+    {
+      "epoch": 2.2750228171585034,
+      "grad_norm": 0.5456854104995728,
+      "learning_rate": 2.1014767930051856e-05,
+      "loss": 0.0462,
+      "step": 3739
+    },
+    {
+      "epoch": 2.2756312747185885,
+      "grad_norm": 0.5097061991691589,
+      "learning_rate": 2.1002478759909558e-05,
+      "loss": 0.0488,
+      "step": 3740
+    },
+    {
+      "epoch": 2.2762397322786736,
+      "grad_norm": 0.5960326790809631,
+      "learning_rate": 2.0990190580985892e-05,
+      "loss": 0.057,
+      "step": 3741
+    },
+    {
+      "epoch": 2.2768481898387587,
+      "grad_norm": 0.6707984805107117,
+      "learning_rate": 2.0977903396327837e-05,
+      "loss": 0.0909,
+      "step": 3742
+    },
+    {
+      "epoch": 2.277456647398844,
+      "grad_norm": 0.5370257496833801,
+      "learning_rate": 2.096561720898209e-05,
+      "loss": 0.0396,
+      "step": 3743
+    },
+    {
+      "epoch": 2.2780651049589293,
+      "grad_norm": 0.5569398403167725,
+      "learning_rate": 2.0953332021995114e-05,
+      "loss": 0.0492,
+      "step": 3744
+    },
+    {
+      "epoch": 2.2786735625190144,
+      "grad_norm": 0.5827397108078003,
+      "learning_rate": 2.094104783841311e-05,
+      "loss": 0.0591,
+      "step": 3745
+    },
+    {
+      "epoch": 2.2792820200790995,
+      "grad_norm": 0.5374563336372375,
+      "learning_rate": 2.0928764661282068e-05,
+      "loss": 0.0438,
+      "step": 3746
+    },
+    {
+      "epoch": 2.2798904776391846,
+      "grad_norm": 0.41728445887565613,
+      "learning_rate": 2.0916482493647694e-05,
+      "loss": 0.0278,
+      "step": 3747
+    },
+    {
+      "epoch": 2.2804989351992697,
+      "grad_norm": 0.5607496500015259,
+      "learning_rate": 2.0904201338555442e-05,
+      "loss": 0.0335,
+      "step": 3748
+    },
+    {
+      "epoch": 2.2811073927593553,
+      "grad_norm": 0.6181521415710449,
+      "learning_rate": 2.0891921199050547e-05,
+      "loss": 0.0428,
+      "step": 3749
+    },
+    {
+      "epoch": 2.2817158503194404,
+      "grad_norm": 0.6066822409629822,
+      "learning_rate": 2.087964207817796e-05,
+      "loss": 0.051,
+      "step": 3750
+    },
+    {
+      "epoch": 2.2823243078795254,
+      "grad_norm": 0.5879988670349121,
+      "learning_rate": 2.0867363978982383e-05,
+      "loss": 0.0372,
+      "step": 3751
+    },
+    {
+      "epoch": 2.2829327654396105,
+      "grad_norm": 0.6240618228912354,
+      "learning_rate": 2.085508690450829e-05,
+      "loss": 0.0432,
+      "step": 3752
+    },
+    {
+      "epoch": 2.2835412229996956,
+      "grad_norm": 0.5587534308433533,
+      "learning_rate": 2.084281085779988e-05,
+      "loss": 0.0415,
+      "step": 3753
+    },
+    {
+      "epoch": 2.284149680559781,
+      "grad_norm": 0.545318067073822,
+      "learning_rate": 2.083053584190109e-05,
+      "loss": 0.0417,
+      "step": 3754
+    },
+    {
+      "epoch": 2.2847581381198663,
+      "grad_norm": 0.5254773497581482,
+      "learning_rate": 2.0818261859855624e-05,
+      "loss": 0.0284,
+      "step": 3755
+    },
+    {
+      "epoch": 2.2853665956799514,
+      "grad_norm": 0.6514537930488586,
+      "learning_rate": 2.0805988914706913e-05,
+      "loss": 0.0714,
+      "step": 3756
+    },
+    {
+      "epoch": 2.2859750532400365,
+      "grad_norm": 0.5673707127571106,
+      "learning_rate": 2.079371700949814e-05,
+      "loss": 0.0426,
+      "step": 3757
+    },
+    {
+      "epoch": 2.2865835108001216,
+      "grad_norm": 0.8576644062995911,
+      "learning_rate": 2.078144614727221e-05,
+      "loss": 0.0707,
+      "step": 3758
+    },
+    {
+      "epoch": 2.287191968360207,
+      "grad_norm": 0.6915388107299805,
+      "learning_rate": 2.0769176331071805e-05,
+      "loss": 0.052,
+      "step": 3759
+    },
+    {
+      "epoch": 2.287800425920292,
+      "grad_norm": 0.536024808883667,
+      "learning_rate": 2.0756907563939327e-05,
+      "loss": 0.041,
+      "step": 3760
+    },
+    {
+      "epoch": 2.2884088834803773,
+      "grad_norm": 0.5904122591018677,
+      "learning_rate": 2.0744639848916895e-05,
+      "loss": 0.047,
+      "step": 3761
+    },
+    {
+      "epoch": 2.2890173410404624,
+      "grad_norm": 0.49669381976127625,
+      "learning_rate": 2.073237318904642e-05,
+      "loss": 0.0417,
+      "step": 3762
+    },
+    {
+      "epoch": 2.2896257986005475,
+      "grad_norm": 0.540749192237854,
+      "learning_rate": 2.0720107587369513e-05,
+      "loss": 0.0603,
+      "step": 3763
+    },
+    {
+      "epoch": 2.290234256160633,
+      "grad_norm": 0.5298412442207336,
+      "learning_rate": 2.070784304692752e-05,
+      "loss": 0.056,
+      "step": 3764
+    },
+    {
+      "epoch": 2.290842713720718,
+      "grad_norm": 0.623453676700592,
+      "learning_rate": 2.0695579570761552e-05,
+      "loss": 0.055,
+      "step": 3765
+    },
+    {
+      "epoch": 2.2914511712808032,
+      "grad_norm": 0.5402014255523682,
+      "learning_rate": 2.0683317161912437e-05,
+      "loss": 0.0448,
+      "step": 3766
+    },
+    {
+      "epoch": 2.2920596288408883,
+      "grad_norm": 0.5555523633956909,
+      "learning_rate": 2.067105582342073e-05,
+      "loss": 0.0432,
+      "step": 3767
+    },
+    {
+      "epoch": 2.2926680864009734,
+      "grad_norm": 0.4748152792453766,
+      "learning_rate": 2.0658795558326743e-05,
+      "loss": 0.0386,
+      "step": 3768
+    },
+    {
+      "epoch": 2.293276543961059,
+      "grad_norm": 0.5108730792999268,
+      "learning_rate": 2.064653636967051e-05,
+      "loss": 0.0442,
+      "step": 3769
+    },
+    {
+      "epoch": 2.293885001521144,
+      "grad_norm": 0.6018065810203552,
+      "learning_rate": 2.063427826049179e-05,
+      "loss": 0.0436,
+      "step": 3770
+    },
+    {
+      "epoch": 2.294493459081229,
+      "grad_norm": 0.5590872168540955,
+      "learning_rate": 2.062202123383009e-05,
+      "loss": 0.0568,
+      "step": 3771
+    },
+    {
+      "epoch": 2.2951019166413142,
+      "grad_norm": 0.6751280426979065,
+      "learning_rate": 2.0609765292724647e-05,
+      "loss": 0.0736,
+      "step": 3772
+    },
+    {
+      "epoch": 2.2957103742013993,
+      "grad_norm": 0.5929213762283325,
+      "learning_rate": 2.059751044021441e-05,
+      "loss": 0.0584,
+      "step": 3773
+    },
+    {
+      "epoch": 2.296318831761485,
+      "grad_norm": 1.4224424362182617,
+      "learning_rate": 2.0585256679338083e-05,
+      "loss": 0.0544,
+      "step": 3774
+    },
+    {
+      "epoch": 2.29692728932157,
+      "grad_norm": 0.5579968690872192,
+      "learning_rate": 2.0573004013134088e-05,
+      "loss": 0.0644,
+      "step": 3775
+    },
+    {
+      "epoch": 2.297535746881655,
+      "grad_norm": 0.6264181137084961,
+      "learning_rate": 2.0560752444640567e-05,
+      "loss": 0.0596,
+      "step": 3776
+    },
+    {
+      "epoch": 2.29814420444174,
+      "grad_norm": 0.6934359669685364,
+      "learning_rate": 2.0548501976895395e-05,
+      "loss": 0.0666,
+      "step": 3777
+    },
+    {
+      "epoch": 2.2987526620018253,
+      "grad_norm": 0.5914580821990967,
+      "learning_rate": 2.0536252612936196e-05,
+      "loss": 0.0455,
+      "step": 3778
+    },
+    {
+      "epoch": 2.299361119561911,
+      "grad_norm": 0.4668748080730438,
+      "learning_rate": 2.0524004355800292e-05,
+      "loss": 0.0424,
+      "step": 3779
+    },
+    {
+      "epoch": 2.299969577121996,
+      "grad_norm": 0.5458466410636902,
+      "learning_rate": 2.0511757208524734e-05,
+      "loss": 0.0447,
+      "step": 3780
+    },
+    {
+      "epoch": 2.300578034682081,
+      "grad_norm": 0.6712079644203186,
+      "learning_rate": 2.0499511174146307e-05,
+      "loss": 0.0512,
+      "step": 3781
+    },
+    {
+      "epoch": 2.301186492242166,
+      "grad_norm": 0.5523056387901306,
+      "learning_rate": 2.0487266255701527e-05,
+      "loss": 0.0471,
+      "step": 3782
+    },
+    {
+      "epoch": 2.301794949802251,
+      "grad_norm": 0.612053632736206,
+      "learning_rate": 2.0475022456226606e-05,
+      "loss": 0.0547,
+      "step": 3783
+    },
+    {
+      "epoch": 2.3024034073623363,
+      "grad_norm": 0.6411483883857727,
+      "learning_rate": 2.0462779778757507e-05,
+      "loss": 0.0698,
+      "step": 3784
+    },
+    {
+      "epoch": 2.303011864922422,
+      "grad_norm": 0.5178706645965576,
+      "learning_rate": 2.045053822632991e-05,
+      "loss": 0.0485,
+      "step": 3785
+    },
+    {
+      "epoch": 2.303620322482507,
+      "grad_norm": 0.5511070489883423,
+      "learning_rate": 2.0438297801979196e-05,
+      "loss": 0.0442,
+      "step": 3786
+    },
+    {
+      "epoch": 2.304228780042592,
+      "grad_norm": 0.5382661819458008,
+      "learning_rate": 2.0426058508740483e-05,
+      "loss": 0.0564,
+      "step": 3787
+    },
+    {
+      "epoch": 2.304837237602677,
+      "grad_norm": 0.5051350593566895,
+      "learning_rate": 2.041382034964862e-05,
+      "loss": 0.0333,
+      "step": 3788
+    },
+    {
+      "epoch": 2.305445695162762,
+      "grad_norm": 0.6360859870910645,
+      "learning_rate": 2.040158332773814e-05,
+      "loss": 0.0551,
+      "step": 3789
+    },
+    {
+      "epoch": 2.3060541527228477,
+      "grad_norm": 0.651378333568573,
+      "learning_rate": 2.0389347446043325e-05,
+      "loss": 0.0489,
+      "step": 3790
+    },
+    {
+      "epoch": 2.306662610282933,
+      "grad_norm": 0.4725477397441864,
+      "learning_rate": 2.0377112707598163e-05,
+      "loss": 0.0325,
+      "step": 3791
+    },
+    {
+      "epoch": 2.307271067843018,
+      "grad_norm": 0.6588869690895081,
+      "learning_rate": 2.0364879115436352e-05,
+      "loss": 0.0461,
+      "step": 3792
+    },
+    {
+      "epoch": 2.307879525403103,
+      "grad_norm": 0.5729846954345703,
+      "learning_rate": 2.0352646672591328e-05,
+      "loss": 0.0658,
+      "step": 3793
+    },
+    {
+      "epoch": 2.308487982963188,
+      "grad_norm": 0.6155021786689758,
+      "learning_rate": 2.034041538209621e-05,
+      "loss": 0.0583,
+      "step": 3794
+    },
+    {
+      "epoch": 2.3090964405232737,
+      "grad_norm": 0.5684482455253601,
+      "learning_rate": 2.0328185246983862e-05,
+      "loss": 0.0488,
+      "step": 3795
+    },
+    {
+      "epoch": 2.3097048980833588,
+      "grad_norm": 0.5797809362411499,
+      "learning_rate": 2.031595627028684e-05,
+      "loss": 0.0394,
+      "step": 3796
+    },
+    {
+      "epoch": 2.310313355643444,
+      "grad_norm": 0.7263544201850891,
+      "learning_rate": 2.0303728455037422e-05,
+      "loss": 0.0685,
+      "step": 3797
+    },
+    {
+      "epoch": 2.310921813203529,
+      "grad_norm": 0.5464826226234436,
+      "learning_rate": 2.0291501804267592e-05,
+      "loss": 0.0483,
+      "step": 3798
+    },
+    {
+      "epoch": 2.311530270763614,
+      "grad_norm": 0.5334215760231018,
+      "learning_rate": 2.0279276321009067e-05,
+      "loss": 0.0416,
+      "step": 3799
+    },
+    {
+      "epoch": 2.3121387283236996,
+      "grad_norm": 0.5960489511489868,
+      "learning_rate": 2.026705200829323e-05,
+      "loss": 0.0465,
+      "step": 3800
+    },
+    {
+      "epoch": 2.3127471858837847,
+      "grad_norm": 0.5816206932067871,
+      "learning_rate": 2.0254828869151225e-05,
+      "loss": 0.0482,
+      "step": 3801
+    },
+    {
+      "epoch": 2.31335564344387,
+      "grad_norm": 0.5829434990882874,
+      "learning_rate": 2.0242606906613876e-05,
+      "loss": 0.0693,
+      "step": 3802
+    },
+    {
+      "epoch": 2.313964101003955,
+      "grad_norm": 0.5510879158973694,
+      "learning_rate": 2.0230386123711714e-05,
+      "loss": 0.0428,
+      "step": 3803
+    },
+    {
+      "epoch": 2.31457255856404,
+      "grad_norm": 0.6710597276687622,
+      "learning_rate": 2.0218166523474973e-05,
+      "loss": 0.0483,
+      "step": 3804
+    },
+    {
+      "epoch": 2.3151810161241255,
+      "grad_norm": 0.5956816673278809,
+      "learning_rate": 2.020594810893363e-05,
+      "loss": 0.0566,
+      "step": 3805
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.6392737030982971,
+      "learning_rate": 2.0193730883117335e-05,
+      "loss": 0.0547,
+      "step": 3806
+    },
+    {
+      "epoch": 2.3163979312442957,
+      "grad_norm": 0.548627495765686,
+      "learning_rate": 2.018151484905543e-05,
+      "loss": 0.0348,
+      "step": 3807
+    },
+    {
+      "epoch": 2.317006388804381,
+      "grad_norm": 0.560031533241272,
+      "learning_rate": 2.016930000977701e-05,
+      "loss": 0.0522,
+      "step": 3808
+    },
+    {
+      "epoch": 2.317614846364466,
+      "grad_norm": 0.6135550141334534,
+      "learning_rate": 2.0157086368310836e-05,
+      "loss": 0.0511,
+      "step": 3809
+    },
+    {
+      "epoch": 2.3182233039245514,
+      "grad_norm": 0.6668503284454346,
+      "learning_rate": 2.014487392768537e-05,
+      "loss": 0.0485,
+      "step": 3810
+    },
+    {
+      "epoch": 2.3188317614846365,
+      "grad_norm": 0.6775378584861755,
+      "learning_rate": 2.01326626909288e-05,
+      "loss": 0.0683,
+      "step": 3811
+    },
+    {
+      "epoch": 2.3194402190447216,
+      "grad_norm": 0.48050111532211304,
+      "learning_rate": 2.012045266106901e-05,
+      "loss": 0.0377,
+      "step": 3812
+    },
+    {
+      "epoch": 2.3200486766048067,
+      "grad_norm": 0.9371777772903442,
+      "learning_rate": 2.010824384113357e-05,
+      "loss": 0.0584,
+      "step": 3813
+    },
+    {
+      "epoch": 2.320657134164892,
+      "grad_norm": 0.6224126815795898,
+      "learning_rate": 2.0096036234149746e-05,
+      "loss": 0.0479,
+      "step": 3814
+    },
+    {
+      "epoch": 2.3212655917249774,
+      "grad_norm": 0.6318977475166321,
+      "learning_rate": 2.0083829843144543e-05,
+      "loss": 0.0657,
+      "step": 3815
+    },
+    {
+      "epoch": 2.3218740492850625,
+      "grad_norm": 0.5769774913787842,
+      "learning_rate": 2.0071624671144616e-05,
+      "loss": 0.0409,
+      "step": 3816
+    },
+    {
+      "epoch": 2.3224825068451476,
+      "grad_norm": 0.5656179189682007,
+      "learning_rate": 2.005942072117634e-05,
+      "loss": 0.0515,
+      "step": 3817
+    },
+    {
+      "epoch": 2.3230909644052327,
+      "grad_norm": 0.5796024799346924,
+      "learning_rate": 2.00472179962658e-05,
+      "loss": 0.0501,
+      "step": 3818
+    },
+    {
+      "epoch": 2.3236994219653178,
+      "grad_norm": 0.5330328345298767,
+      "learning_rate": 2.0035016499438758e-05,
+      "loss": 0.0495,
+      "step": 3819
+    },
+    {
+      "epoch": 2.324307879525403,
+      "grad_norm": 0.6313185691833496,
+      "learning_rate": 2.002281623372066e-05,
+      "loss": 0.045,
+      "step": 3820
+    },
+    {
+      "epoch": 2.3249163370854884,
+      "grad_norm": 0.5151857137680054,
+      "learning_rate": 2.001061720213669e-05,
+      "loss": 0.0319,
+      "step": 3821
+    },
+    {
+      "epoch": 2.3255247946455735,
+      "grad_norm": 0.6799402236938477,
+      "learning_rate": 1.9998419407711686e-05,
+      "loss": 0.0514,
+      "step": 3822
+    },
+    {
+      "epoch": 2.3261332522056586,
+      "grad_norm": 0.499461829662323,
+      "learning_rate": 1.998622285347018e-05,
+      "loss": 0.0368,
+      "step": 3823
+    },
+    {
+      "epoch": 2.3267417097657437,
+      "grad_norm": 0.5110387206077576,
+      "learning_rate": 1.9974027542436433e-05,
+      "loss": 0.0464,
+      "step": 3824
+    },
+    {
+      "epoch": 2.327350167325829,
+      "grad_norm": 0.7454649806022644,
+      "learning_rate": 1.996183347763436e-05,
+      "loss": 0.076,
+      "step": 3825
+    },
+    {
+      "epoch": 2.3279586248859143,
+      "grad_norm": 0.5358094573020935,
+      "learning_rate": 1.994964066208759e-05,
+      "loss": 0.0465,
+      "step": 3826
+    },
+    {
+      "epoch": 2.3285670824459994,
+      "grad_norm": 0.6061232686042786,
+      "learning_rate": 1.993744909881941e-05,
+      "loss": 0.0434,
+      "step": 3827
+    },
+    {
+      "epoch": 2.3291755400060845,
+      "grad_norm": 0.6150352358818054,
+      "learning_rate": 1.9925258790852852e-05,
+      "loss": 0.0528,
+      "step": 3828
+    },
+    {
+      "epoch": 2.3297839975661696,
+      "grad_norm": 0.5760040283203125,
+      "learning_rate": 1.9913069741210583e-05,
+      "loss": 0.0573,
+      "step": 3829
+    },
+    {
+      "epoch": 2.3303924551262547,
+      "grad_norm": 0.6556773781776428,
+      "learning_rate": 1.9900881952914975e-05,
+      "loss": 0.0423,
+      "step": 3830
+    },
+    {
+      "epoch": 2.3310009126863402,
+      "grad_norm": 0.5123857259750366,
+      "learning_rate": 1.9888695428988115e-05,
+      "loss": 0.0432,
+      "step": 3831
+    },
+    {
+      "epoch": 2.3316093702464253,
+      "grad_norm": 0.651202917098999,
+      "learning_rate": 1.9876510172451733e-05,
+      "loss": 0.0423,
+      "step": 3832
+    },
+    {
+      "epoch": 2.3322178278065104,
+      "grad_norm": 0.4787401258945465,
+      "learning_rate": 1.9864326186327265e-05,
+      "loss": 0.0382,
+      "step": 3833
+    },
+    {
+      "epoch": 2.3328262853665955,
+      "grad_norm": 0.5490178465843201,
+      "learning_rate": 1.9852143473635847e-05,
+      "loss": 0.0627,
+      "step": 3834
+    },
+    {
+      "epoch": 2.3334347429266806,
+      "grad_norm": 0.47695285081863403,
+      "learning_rate": 1.9839962037398276e-05,
+      "loss": 0.0515,
+      "step": 3835
+    },
+    {
+      "epoch": 2.334043200486766,
+      "grad_norm": 0.5950469374656677,
+      "learning_rate": 1.9827781880635034e-05,
+      "loss": 0.0443,
+      "step": 3836
+    },
+    {
+      "epoch": 2.3346516580468513,
+      "grad_norm": 0.5753946304321289,
+      "learning_rate": 1.981560300636629e-05,
+      "loss": 0.0452,
+      "step": 3837
+    },
+    {
+      "epoch": 2.3352601156069364,
+      "grad_norm": 0.5735435485839844,
+      "learning_rate": 1.9803425417611915e-05,
+      "loss": 0.0579,
+      "step": 3838
+    },
+    {
+      "epoch": 2.3358685731670215,
+      "grad_norm": 0.5695839524269104,
+      "learning_rate": 1.979124911739143e-05,
+      "loss": 0.0466,
+      "step": 3839
+    },
+    {
+      "epoch": 2.3364770307271066,
+      "grad_norm": 0.4902591407299042,
+      "learning_rate": 1.9779074108724042e-05,
+      "loss": 0.0431,
+      "step": 3840
+    },
+    {
+      "epoch": 2.337085488287192,
+      "grad_norm": 0.6744665503501892,
+      "learning_rate": 1.9766900394628668e-05,
+      "loss": 0.0583,
+      "step": 3841
+    },
+    {
+      "epoch": 2.337693945847277,
+      "grad_norm": 0.5691081285476685,
+      "learning_rate": 1.975472797812387e-05,
+      "loss": 0.0315,
+      "step": 3842
+    },
+    {
+      "epoch": 2.3383024034073623,
+      "grad_norm": 0.5368900299072266,
+      "learning_rate": 1.9742556862227888e-05,
+      "loss": 0.0418,
+      "step": 3843
+    },
+    {
+      "epoch": 2.3389108609674474,
+      "grad_norm": 0.5534371137619019,
+      "learning_rate": 1.9730387049958672e-05,
+      "loss": 0.0415,
+      "step": 3844
+    },
+    {
+      "epoch": 2.3395193185275325,
+      "grad_norm": 0.5242774486541748,
+      "learning_rate": 1.971821854433382e-05,
+      "loss": 0.0447,
+      "step": 3845
+    },
+    {
+      "epoch": 2.340127776087618,
+      "grad_norm": 0.6565991044044495,
+      "learning_rate": 1.9706051348370608e-05,
+      "loss": 0.0448,
+      "step": 3846
+    },
+    {
+      "epoch": 2.340736233647703,
+      "grad_norm": 0.4980752766132355,
+      "learning_rate": 1.9693885465086008e-05,
+      "loss": 0.044,
+      "step": 3847
+    },
+    {
+      "epoch": 2.341344691207788,
+      "grad_norm": 0.5127741694450378,
+      "learning_rate": 1.968172089749664e-05,
+      "loss": 0.0374,
+      "step": 3848
+    },
+    {
+      "epoch": 2.3419531487678733,
+      "grad_norm": 0.5670256614685059,
+      "learning_rate": 1.9669557648618815e-05,
+      "loss": 0.0474,
+      "step": 3849
+    },
+    {
+      "epoch": 2.3425616063279584,
+      "grad_norm": 0.6457985043525696,
+      "learning_rate": 1.9657395721468494e-05,
+      "loss": 0.06,
+      "step": 3850
+    },
+    {
+      "epoch": 2.343170063888044,
+      "grad_norm": 0.6051198244094849,
+      "learning_rate": 1.964523511906136e-05,
+      "loss": 0.0545,
+      "step": 3851
+    },
+    {
+      "epoch": 2.343778521448129,
+      "grad_norm": 0.6616335511207581,
+      "learning_rate": 1.9633075844412714e-05,
+      "loss": 0.0371,
+      "step": 3852
+    },
+    {
+      "epoch": 2.344386979008214,
+      "grad_norm": 0.5447221398353577,
+      "learning_rate": 1.9620917900537546e-05,
+      "loss": 0.0435,
+      "step": 3853
+    },
+    {
+      "epoch": 2.3449954365682992,
+      "grad_norm": 0.518090546131134,
+      "learning_rate": 1.960876129045054e-05,
+      "loss": 0.0454,
+      "step": 3854
+    },
+    {
+      "epoch": 2.3456038941283843,
+      "grad_norm": 0.5191083550453186,
+      "learning_rate": 1.959660601716601e-05,
+      "loss": 0.0364,
+      "step": 3855
+    },
+    {
+      "epoch": 2.34621235168847,
+      "grad_norm": 0.6573032140731812,
+      "learning_rate": 1.958445208369796e-05,
+      "loss": 0.0589,
+      "step": 3856
+    },
+    {
+      "epoch": 2.346820809248555,
+      "grad_norm": 0.5200658440589905,
+      "learning_rate": 1.9572299493060067e-05,
+      "loss": 0.0445,
+      "step": 3857
+    },
+    {
+      "epoch": 2.34742926680864,
+      "grad_norm": 0.5530250072479248,
+      "learning_rate": 1.9560148248265662e-05,
+      "loss": 0.0455,
+      "step": 3858
+    },
+    {
+      "epoch": 2.348037724368725,
+      "grad_norm": 0.7511923313140869,
+      "learning_rate": 1.954799835232774e-05,
+      "loss": 0.0459,
+      "step": 3859
+    },
+    {
+      "epoch": 2.3486461819288103,
+      "grad_norm": 0.48297080397605896,
+      "learning_rate": 1.9535849808258985e-05,
+      "loss": 0.0386,
+      "step": 3860
+    },
+    {
+      "epoch": 2.349254639488896,
+      "grad_norm": 0.4536675214767456,
+      "learning_rate": 1.9523702619071722e-05,
+      "loss": 0.0357,
+      "step": 3861
+    },
+    {
+      "epoch": 2.349863097048981,
+      "grad_norm": 0.522323727607727,
+      "learning_rate": 1.9511556787777945e-05,
+      "loss": 0.0429,
+      "step": 3862
+    },
+    {
+      "epoch": 2.350471554609066,
+      "grad_norm": 0.5316919088363647,
+      "learning_rate": 1.9499412317389305e-05,
+      "loss": 0.0322,
+      "step": 3863
+    },
+    {
+      "epoch": 2.351080012169151,
+      "grad_norm": 0.5202551484107971,
+      "learning_rate": 1.9487269210917144e-05,
+      "loss": 0.0453,
+      "step": 3864
+    },
+    {
+      "epoch": 2.351688469729236,
+      "grad_norm": 0.6849666833877563,
+      "learning_rate": 1.9475127471372436e-05,
+      "loss": 0.0488,
+      "step": 3865
+    },
+    {
+      "epoch": 2.3522969272893217,
+      "grad_norm": 0.531730592250824,
+      "learning_rate": 1.946298710176582e-05,
+      "loss": 0.0435,
+      "step": 3866
+    },
+    {
+      "epoch": 2.352905384849407,
+      "grad_norm": 0.6236624121665955,
+      "learning_rate": 1.9450848105107615e-05,
+      "loss": 0.0723,
+      "step": 3867
+    },
+    {
+      "epoch": 2.353513842409492,
+      "grad_norm": 0.502086877822876,
+      "learning_rate": 1.9438710484407786e-05,
+      "loss": 0.0486,
+      "step": 3868
+    },
+    {
+      "epoch": 2.354122299969577,
+      "grad_norm": 0.4249696731567383,
+      "learning_rate": 1.9426574242675932e-05,
+      "loss": 0.0282,
+      "step": 3869
+    },
+    {
+      "epoch": 2.354730757529662,
+      "grad_norm": 0.6139441132545471,
+      "learning_rate": 1.941443938292137e-05,
+      "loss": 0.0492,
+      "step": 3870
+    },
+    {
+      "epoch": 2.3553392150897476,
+      "grad_norm": 0.4936158061027527,
+      "learning_rate": 1.9402305908153025e-05,
+      "loss": 0.0392,
+      "step": 3871
+    },
+    {
+      "epoch": 2.3559476726498327,
+      "grad_norm": 0.5353962182998657,
+      "learning_rate": 1.9390173821379486e-05,
+      "loss": 0.0347,
+      "step": 3872
+    },
+    {
+      "epoch": 2.356556130209918,
+      "grad_norm": 0.46766939759254456,
+      "learning_rate": 1.9378043125609003e-05,
+      "loss": 0.041,
+      "step": 3873
+    },
+    {
+      "epoch": 2.357164587770003,
+      "grad_norm": 0.5060580372810364,
+      "learning_rate": 1.93659138238495e-05,
+      "loss": 0.0414,
+      "step": 3874
+    },
+    {
+      "epoch": 2.357773045330088,
+      "grad_norm": 0.5533692240715027,
+      "learning_rate": 1.935378591910853e-05,
+      "loss": 0.0365,
+      "step": 3875
+    },
+    {
+      "epoch": 2.3583815028901736,
+      "grad_norm": 0.5114442110061646,
+      "learning_rate": 1.934165941439329e-05,
+      "loss": 0.0452,
+      "step": 3876
+    },
+    {
+      "epoch": 2.3589899604502587,
+      "grad_norm": 0.563607394695282,
+      "learning_rate": 1.932953431271068e-05,
+      "loss": 0.06,
+      "step": 3877
+    },
+    {
+      "epoch": 2.3595984180103438,
+      "grad_norm": 0.4950112998485565,
+      "learning_rate": 1.93174106170672e-05,
+      "loss": 0.047,
+      "step": 3878
+    },
+    {
+      "epoch": 2.360206875570429,
+      "grad_norm": 0.6233085989952087,
+      "learning_rate": 1.930528833046902e-05,
+      "loss": 0.0455,
+      "step": 3879
+    },
+    {
+      "epoch": 2.360815333130514,
+      "grad_norm": 0.5214982628822327,
+      "learning_rate": 1.9293167455921972e-05,
+      "loss": 0.0359,
+      "step": 3880
+    },
+    {
+      "epoch": 2.3614237906905995,
+      "grad_norm": 0.5291669368743896,
+      "learning_rate": 1.9281047996431527e-05,
+      "loss": 0.041,
+      "step": 3881
+    },
+    {
+      "epoch": 2.3620322482506846,
+      "grad_norm": 0.7979464530944824,
+      "learning_rate": 1.9268929955002787e-05,
+      "loss": 0.0629,
+      "step": 3882
+    },
+    {
+      "epoch": 2.3626407058107697,
+      "grad_norm": 0.5536260008811951,
+      "learning_rate": 1.9256813334640546e-05,
+      "loss": 0.05,
+      "step": 3883
+    },
+    {
+      "epoch": 2.3632491633708548,
+      "grad_norm": 0.5184295177459717,
+      "learning_rate": 1.9244698138349212e-05,
+      "loss": 0.0453,
+      "step": 3884
+    },
+    {
+      "epoch": 2.36385762093094,
+      "grad_norm": 0.632335901260376,
+      "learning_rate": 1.9232584369132848e-05,
+      "loss": 0.0579,
+      "step": 3885
+    },
+    {
+      "epoch": 2.3644660784910254,
+      "grad_norm": 0.611529529094696,
+      "learning_rate": 1.922047202999515e-05,
+      "loss": 0.0749,
+      "step": 3886
+    },
+    {
+      "epoch": 2.3650745360511105,
+      "grad_norm": 0.6848330497741699,
+      "learning_rate": 1.9208361123939498e-05,
+      "loss": 0.0487,
+      "step": 3887
+    },
+    {
+      "epoch": 2.3656829936111956,
+      "grad_norm": 0.5429683923721313,
+      "learning_rate": 1.9196251653968877e-05,
+      "loss": 0.0382,
+      "step": 3888
+    },
+    {
+      "epoch": 2.3662914511712807,
+      "grad_norm": 0.5638923645019531,
+      "learning_rate": 1.9184143623085926e-05,
+      "loss": 0.0403,
+      "step": 3889
+    },
+    {
+      "epoch": 2.366899908731366,
+      "grad_norm": 0.5319681167602539,
+      "learning_rate": 1.917203703429295e-05,
+      "loss": 0.0563,
+      "step": 3890
+    },
+    {
+      "epoch": 2.3675083662914513,
+      "grad_norm": 0.5609192848205566,
+      "learning_rate": 1.9159931890591865e-05,
+      "loss": 0.0425,
+      "step": 3891
+    },
+    {
+      "epoch": 2.3681168238515364,
+      "grad_norm": 0.5878105759620667,
+      "learning_rate": 1.9147828194984236e-05,
+      "loss": 0.0424,
+      "step": 3892
+    },
+    {
+      "epoch": 2.3687252814116215,
+      "grad_norm": 0.5947871208190918,
+      "learning_rate": 1.91357259504713e-05,
+      "loss": 0.0574,
+      "step": 3893
+    },
+    {
+      "epoch": 2.3693337389717066,
+      "grad_norm": 0.4823724031448364,
+      "learning_rate": 1.9123625160053885e-05,
+      "loss": 0.0452,
+      "step": 3894
+    },
+    {
+      "epoch": 2.3699421965317917,
+      "grad_norm": 0.5193595886230469,
+      "learning_rate": 1.911152582673249e-05,
+      "loss": 0.0459,
+      "step": 3895
+    },
+    {
+      "epoch": 2.3705506540918773,
+      "grad_norm": 0.6103157997131348,
+      "learning_rate": 1.9099427953507245e-05,
+      "loss": 0.0555,
+      "step": 3896
+    },
+    {
+      "epoch": 2.3711591116519624,
+      "grad_norm": 0.5717165470123291,
+      "learning_rate": 1.9087331543377925e-05,
+      "loss": 0.0421,
+      "step": 3897
+    },
+    {
+      "epoch": 2.3717675692120475,
+      "grad_norm": 0.5617209076881409,
+      "learning_rate": 1.9075236599343927e-05,
+      "loss": 0.0419,
+      "step": 3898
+    },
+    {
+      "epoch": 2.3723760267721326,
+      "grad_norm": 0.503222644329071,
+      "learning_rate": 1.9063143124404293e-05,
+      "loss": 0.054,
+      "step": 3899
+    },
+    {
+      "epoch": 2.3729844843322176,
+      "grad_norm": 0.6546376943588257,
+      "learning_rate": 1.9051051121557712e-05,
+      "loss": 0.0527,
+      "step": 3900
+    },
+    {
+      "epoch": 2.373592941892303,
+      "grad_norm": 0.5000297427177429,
+      "learning_rate": 1.903896059380248e-05,
+      "loss": 0.0388,
+      "step": 3901
+    },
+    {
+      "epoch": 2.3742013994523883,
+      "grad_norm": 0.6017688512802124,
+      "learning_rate": 1.9026871544136554e-05,
+      "loss": 0.0479,
+      "step": 3902
+    },
+    {
+      "epoch": 2.3748098570124734,
+      "grad_norm": 0.5484517216682434,
+      "learning_rate": 1.901478397555752e-05,
+      "loss": 0.0428,
+      "step": 3903
+    },
+    {
+      "epoch": 2.3754183145725585,
+      "grad_norm": 0.6508259177207947,
+      "learning_rate": 1.9002697891062572e-05,
+      "loss": 0.0521,
+      "step": 3904
+    },
+    {
+      "epoch": 2.3760267721326436,
+      "grad_norm": 0.6614736914634705,
+      "learning_rate": 1.8990613293648572e-05,
+      "loss": 0.0626,
+      "step": 3905
+    },
+    {
+      "epoch": 2.376635229692729,
+      "grad_norm": 0.5715472102165222,
+      "learning_rate": 1.8978530186312e-05,
+      "loss": 0.0365,
+      "step": 3906
+    },
+    {
+      "epoch": 2.377243687252814,
+      "grad_norm": 0.4740070104598999,
+      "learning_rate": 1.8966448572048946e-05,
+      "loss": 0.0401,
+      "step": 3907
+    },
+    {
+      "epoch": 2.3778521448128993,
+      "grad_norm": 0.4682742655277252,
+      "learning_rate": 1.895436845385516e-05,
+      "loss": 0.0391,
+      "step": 3908
+    },
+    {
+      "epoch": 2.3784606023729844,
+      "grad_norm": 0.6115542054176331,
+      "learning_rate": 1.8942289834725995e-05,
+      "loss": 0.0662,
+      "step": 3909
+    },
+    {
+      "epoch": 2.3790690599330695,
+      "grad_norm": 0.45415550470352173,
+      "learning_rate": 1.8930212717656464e-05,
+      "loss": 0.0377,
+      "step": 3910
+    },
+    {
+      "epoch": 2.379677517493155,
+      "grad_norm": 0.6085162162780762,
+      "learning_rate": 1.891813710564117e-05,
+      "loss": 0.0598,
+      "step": 3911
+    },
+    {
+      "epoch": 2.38028597505324,
+      "grad_norm": 0.5516512393951416,
+      "learning_rate": 1.8906063001674368e-05,
+      "loss": 0.0526,
+      "step": 3912
+    },
+    {
+      "epoch": 2.3808944326133252,
+      "grad_norm": 0.5332078337669373,
+      "learning_rate": 1.8893990408749943e-05,
+      "loss": 0.0507,
+      "step": 3913
+    },
+    {
+      "epoch": 2.3815028901734103,
+      "grad_norm": 0.5707116723060608,
+      "learning_rate": 1.8881919329861377e-05,
+      "loss": 0.0486,
+      "step": 3914
+    },
+    {
+      "epoch": 2.3821113477334954,
+      "grad_norm": 0.5801469683647156,
+      "learning_rate": 1.8869849768001803e-05,
+      "loss": 0.05,
+      "step": 3915
+    },
+    {
+      "epoch": 2.382719805293581,
+      "grad_norm": 0.4456758201122284,
+      "learning_rate": 1.8857781726163977e-05,
+      "loss": 0.0382,
+      "step": 3916
+    },
+    {
+      "epoch": 2.383328262853666,
+      "grad_norm": 0.5199500322341919,
+      "learning_rate": 1.8845715207340254e-05,
+      "loss": 0.0443,
+      "step": 3917
+    },
+    {
+      "epoch": 2.383936720413751,
+      "grad_norm": 0.6407058835029602,
+      "learning_rate": 1.8833650214522643e-05,
+      "loss": 0.065,
+      "step": 3918
+    },
+    {
+      "epoch": 2.3845451779738362,
+      "grad_norm": 0.5200806856155396,
+      "learning_rate": 1.882158675070275e-05,
+      "loss": 0.0285,
+      "step": 3919
+    },
+    {
+      "epoch": 2.3851536355339213,
+      "grad_norm": 0.5379014611244202,
+      "learning_rate": 1.880952481887181e-05,
+      "loss": 0.0615,
+      "step": 3920
+    },
+    {
+      "epoch": 2.385762093094007,
+      "grad_norm": 0.46084079146385193,
+      "learning_rate": 1.879746442202069e-05,
+      "loss": 0.0379,
+      "step": 3921
+    },
+    {
+      "epoch": 2.386370550654092,
+      "grad_norm": 0.5246143937110901,
+      "learning_rate": 1.878540556313986e-05,
+      "loss": 0.0558,
+      "step": 3922
+    },
+    {
+      "epoch": 2.386979008214177,
+      "grad_norm": 0.5521069169044495,
+      "learning_rate": 1.8773348245219403e-05,
+      "loss": 0.0533,
+      "step": 3923
+    },
+    {
+      "epoch": 2.387587465774262,
+      "grad_norm": 0.6149189472198486,
+      "learning_rate": 1.876129247124905e-05,
+      "loss": 0.0434,
+      "step": 3924
+    },
+    {
+      "epoch": 2.3881959233343473,
+      "grad_norm": 0.6903012990951538,
+      "learning_rate": 1.874923824421812e-05,
+      "loss": 0.052,
+      "step": 3925
+    },
+    {
+      "epoch": 2.388804380894433,
+      "grad_norm": 0.8194772005081177,
+      "learning_rate": 1.8737185567115555e-05,
+      "loss": 0.0505,
+      "step": 3926
+    },
+    {
+      "epoch": 2.389412838454518,
+      "grad_norm": 0.4386597275733948,
+      "learning_rate": 1.872513444292993e-05,
+      "loss": 0.028,
+      "step": 3927
+    },
+    {
+      "epoch": 2.390021296014603,
+      "grad_norm": 0.4583303928375244,
+      "learning_rate": 1.871308487464941e-05,
+      "loss": 0.0417,
+      "step": 3928
+    },
+    {
+      "epoch": 2.390629753574688,
+      "grad_norm": 0.5294291973114014,
+      "learning_rate": 1.8701036865261792e-05,
+      "loss": 0.0372,
+      "step": 3929
+    },
+    {
+      "epoch": 2.391238211134773,
+      "grad_norm": 0.5232396721839905,
+      "learning_rate": 1.868899041775448e-05,
+      "loss": 0.0386,
+      "step": 3930
+    },
+    {
+      "epoch": 2.3918466686948587,
+      "grad_norm": 0.5549443960189819,
+      "learning_rate": 1.867694553511449e-05,
+      "loss": 0.053,
+      "step": 3931
+    },
+    {
+      "epoch": 2.392455126254944,
+      "grad_norm": 0.6172237396240234,
+      "learning_rate": 1.8664902220328435e-05,
+      "loss": 0.0574,
+      "step": 3932
+    },
+    {
+      "epoch": 2.393063583815029,
+      "grad_norm": 0.7529854774475098,
+      "learning_rate": 1.8652860476382584e-05,
+      "loss": 0.0613,
+      "step": 3933
+    },
+    {
+      "epoch": 2.393672041375114,
+      "grad_norm": 0.5938291549682617,
+      "learning_rate": 1.8640820306262774e-05,
+      "loss": 0.0592,
+      "step": 3934
+    },
+    {
+      "epoch": 2.394280498935199,
+      "grad_norm": 0.6293975114822388,
+      "learning_rate": 1.862878171295445e-05,
+      "loss": 0.0491,
+      "step": 3935
+    },
+    {
+      "epoch": 2.3948889564952847,
+      "grad_norm": 0.42942139506340027,
+      "learning_rate": 1.8616744699442706e-05,
+      "loss": 0.0368,
+      "step": 3936
+    },
+    {
+      "epoch": 2.3954974140553698,
+      "grad_norm": 0.5470993518829346,
+      "learning_rate": 1.8604709268712213e-05,
+      "loss": 0.0506,
+      "step": 3937
+    },
+    {
+      "epoch": 2.396105871615455,
+      "grad_norm": 1.0791350603103638,
+      "learning_rate": 1.859267542374724e-05,
+      "loss": 0.0415,
+      "step": 3938
+    },
+    {
+      "epoch": 2.39671432917554,
+      "grad_norm": 0.6190977096557617,
+      "learning_rate": 1.8580643167531703e-05,
+      "loss": 0.0691,
+      "step": 3939
+    },
+    {
+      "epoch": 2.397322786735625,
+      "grad_norm": 0.5365692973136902,
+      "learning_rate": 1.8568612503049086e-05,
+      "loss": 0.0547,
+      "step": 3940
+    },
+    {
+      "epoch": 2.3979312442957106,
+      "grad_norm": 0.5777300596237183,
+      "learning_rate": 1.8556583433282497e-05,
+      "loss": 0.0521,
+      "step": 3941
+    },
+    {
+      "epoch": 2.3985397018557957,
+      "grad_norm": 0.612249493598938,
+      "learning_rate": 1.8544555961214634e-05,
+      "loss": 0.0507,
+      "step": 3942
+    },
+    {
+      "epoch": 2.3991481594158808,
+      "grad_norm": 0.5814237594604492,
+      "learning_rate": 1.853253008982782e-05,
+      "loss": 0.0476,
+      "step": 3943
+    },
+    {
+      "epoch": 2.399756616975966,
+      "grad_norm": 0.5139836072921753,
+      "learning_rate": 1.8520505822103972e-05,
+      "loss": 0.033,
+      "step": 3944
+    },
+    {
+      "epoch": 2.400365074536051,
+      "grad_norm": 0.5720403790473938,
+      "learning_rate": 1.8508483161024586e-05,
+      "loss": 0.0567,
+      "step": 3945
+    },
+    {
+      "epoch": 2.4009735320961365,
+      "grad_norm": 0.6230379343032837,
+      "learning_rate": 1.849646210957081e-05,
+      "loss": 0.0538,
+      "step": 3946
+    },
+    {
+      "epoch": 2.4015819896562216,
+      "grad_norm": 0.4990738034248352,
+      "learning_rate": 1.848444267072335e-05,
+      "loss": 0.0483,
+      "step": 3947
+    },
+    {
+      "epoch": 2.4021904472163067,
+      "grad_norm": 0.600247323513031,
+      "learning_rate": 1.8472424847462517e-05,
+      "loss": 0.0729,
+      "step": 3948
+    },
+    {
+      "epoch": 2.402798904776392,
+      "grad_norm": 0.535139799118042,
+      "learning_rate": 1.8460408642768244e-05,
+      "loss": 0.0502,
+      "step": 3949
+    },
+    {
+      "epoch": 2.403407362336477,
+      "grad_norm": 0.4909554421901703,
+      "learning_rate": 1.844839405962005e-05,
+      "loss": 0.0368,
+      "step": 3950
+    },
+    {
+      "epoch": 2.4040158198965624,
+      "grad_norm": 0.5839510560035706,
+      "learning_rate": 1.8436381100997034e-05,
+      "loss": 0.0523,
+      "step": 3951
+    },
+    {
+      "epoch": 2.4046242774566475,
+      "grad_norm": 0.7578384280204773,
+      "learning_rate": 1.8424369769877927e-05,
+      "loss": 0.0429,
+      "step": 3952
+    },
+    {
+      "epoch": 2.4052327350167326,
+      "grad_norm": 0.5987670421600342,
+      "learning_rate": 1.8412360069241034e-05,
+      "loss": 0.0529,
+      "step": 3953
+    },
+    {
+      "epoch": 2.4058411925768177,
+      "grad_norm": 0.5466412901878357,
+      "learning_rate": 1.8400352002064253e-05,
+      "loss": 0.0491,
+      "step": 3954
+    },
+    {
+      "epoch": 2.406449650136903,
+      "grad_norm": 1.5153589248657227,
+      "learning_rate": 1.8388345571325083e-05,
+      "loss": 0.0481,
+      "step": 3955
+    },
+    {
+      "epoch": 2.4070581076969884,
+      "grad_norm": 0.5942850708961487,
+      "learning_rate": 1.8376340780000635e-05,
+      "loss": 0.0641,
+      "step": 3956
+    },
+    {
+      "epoch": 2.4076665652570735,
+      "grad_norm": 0.5600593090057373,
+      "learning_rate": 1.836433763106758e-05,
+      "loss": 0.0534,
+      "step": 3957
+    },
+    {
+      "epoch": 2.4082750228171586,
+      "grad_norm": 0.5812585353851318,
+      "learning_rate": 1.8352336127502198e-05,
+      "loss": 0.0532,
+      "step": 3958
+    },
+    {
+      "epoch": 2.4088834803772436,
+      "grad_norm": 0.4886872172355652,
+      "learning_rate": 1.8340336272280377e-05,
+      "loss": 0.0524,
+      "step": 3959
+    },
+    {
+      "epoch": 2.4094919379373287,
+      "grad_norm": 0.6744169592857361,
+      "learning_rate": 1.8328338068377578e-05,
+      "loss": 0.0497,
+      "step": 3960
+    },
+    {
+      "epoch": 2.4101003954974143,
+      "grad_norm": 0.6337938904762268,
+      "learning_rate": 1.831634151876883e-05,
+      "loss": 0.0431,
+      "step": 3961
+    },
+    {
+      "epoch": 2.4107088530574994,
+      "grad_norm": 0.6581575870513916,
+      "learning_rate": 1.8304346626428813e-05,
+      "loss": 0.0637,
+      "step": 3962
+    },
+    {
+      "epoch": 2.4113173106175845,
+      "grad_norm": 1.0695427656173706,
+      "learning_rate": 1.829235339433174e-05,
+      "loss": 0.0482,
+      "step": 3963
+    },
+    {
+      "epoch": 2.4119257681776696,
+      "grad_norm": 0.6076899766921997,
+      "learning_rate": 1.828036182545144e-05,
+      "loss": 0.0375,
+      "step": 3964
+    },
+    {
+      "epoch": 2.4125342257377547,
+      "grad_norm": 0.6036027669906616,
+      "learning_rate": 1.8268371922761308e-05,
+      "loss": 0.0442,
+      "step": 3965
+    },
+    {
+      "epoch": 2.41314268329784,
+      "grad_norm": 0.5249053239822388,
+      "learning_rate": 1.8256383689234362e-05,
+      "loss": 0.0415,
+      "step": 3966
+    },
+    {
+      "epoch": 2.4137511408579253,
+      "grad_norm": 0.5730521082878113,
+      "learning_rate": 1.824439712784317e-05,
+      "loss": 0.0438,
+      "step": 3967
+    },
+    {
+      "epoch": 2.4143595984180104,
+      "grad_norm": 0.5469090342521667,
+      "learning_rate": 1.8232412241559896e-05,
+      "loss": 0.0265,
+      "step": 3968
+    },
+    {
+      "epoch": 2.4149680559780955,
+      "grad_norm": 0.5718511939048767,
+      "learning_rate": 1.8220429033356305e-05,
+      "loss": 0.0395,
+      "step": 3969
+    },
+    {
+      "epoch": 2.4155765135381806,
+      "grad_norm": 0.5090388655662537,
+      "learning_rate": 1.8208447506203734e-05,
+      "loss": 0.0359,
+      "step": 3970
+    },
+    {
+      "epoch": 2.416184971098266,
+      "grad_norm": 0.4648432731628418,
+      "learning_rate": 1.819646766307308e-05,
+      "loss": 0.0351,
+      "step": 3971
+    },
+    {
+      "epoch": 2.4167934286583512,
+      "grad_norm": 0.5751369595527649,
+      "learning_rate": 1.8184489506934875e-05,
+      "loss": 0.035,
+      "step": 3972
+    },
+    {
+      "epoch": 2.4174018862184363,
+      "grad_norm": 0.607879638671875,
+      "learning_rate": 1.8172513040759183e-05,
+      "loss": 0.0605,
+      "step": 3973
+    },
+    {
+      "epoch": 2.4180103437785214,
+      "grad_norm": 0.5648092031478882,
+      "learning_rate": 1.8160538267515665e-05,
+      "loss": 0.0417,
+      "step": 3974
+    },
+    {
+      "epoch": 2.4186188013386065,
+      "grad_norm": 0.6005630493164062,
+      "learning_rate": 1.8148565190173586e-05,
+      "loss": 0.0441,
+      "step": 3975
+    },
+    {
+      "epoch": 2.419227258898692,
+      "grad_norm": 0.524438202381134,
+      "learning_rate": 1.813659381170176e-05,
+      "loss": 0.0344,
+      "step": 3976
+    },
+    {
+      "epoch": 2.419835716458777,
+      "grad_norm": 0.4712424874305725,
+      "learning_rate": 1.812462413506858e-05,
+      "loss": 0.0264,
+      "step": 3977
+    },
+    {
+      "epoch": 2.4204441740188622,
+      "grad_norm": 0.561529278755188,
+      "learning_rate": 1.811265616324203e-05,
+      "loss": 0.0522,
+      "step": 3978
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.6183802485466003,
+      "learning_rate": 1.8100689899189683e-05,
+      "loss": 0.0509,
+      "step": 3979
+    },
+    {
+      "epoch": 2.4216610891390324,
+      "grad_norm": 0.5042179226875305,
+      "learning_rate": 1.808872534587866e-05,
+      "loss": 0.0501,
+      "step": 3980
+    },
+    {
+      "epoch": 2.422269546699118,
+      "grad_norm": 0.6235750317573547,
+      "learning_rate": 1.8076762506275667e-05,
+      "loss": 0.0617,
+      "step": 3981
+    },
+    {
+      "epoch": 2.422878004259203,
+      "grad_norm": 0.5784009099006653,
+      "learning_rate": 1.8064801383347006e-05,
+      "loss": 0.0542,
+      "step": 3982
+    },
+    {
+      "epoch": 2.423486461819288,
+      "grad_norm": 0.6061785221099854,
+      "learning_rate": 1.8052841980058533e-05,
+      "loss": 0.0477,
+      "step": 3983
+    },
+    {
+      "epoch": 2.4240949193793733,
+      "grad_norm": 0.5976025462150574,
+      "learning_rate": 1.8040884299375665e-05,
+      "loss": 0.0554,
+      "step": 3984
+    },
+    {
+      "epoch": 2.4247033769394584,
+      "grad_norm": 0.5100542306900024,
+      "learning_rate": 1.802892834426343e-05,
+      "loss": 0.0437,
+      "step": 3985
+    },
+    {
+      "epoch": 2.425311834499544,
+      "grad_norm": 0.6174180507659912,
+      "learning_rate": 1.8016974117686398e-05,
+      "loss": 0.0661,
+      "step": 3986
+    },
+    {
+      "epoch": 2.425920292059629,
+      "grad_norm": 0.6207003593444824,
+      "learning_rate": 1.8005021622608717e-05,
+      "loss": 0.0463,
+      "step": 3987
+    },
+    {
+      "epoch": 2.426528749619714,
+      "grad_norm": 0.5053902864456177,
+      "learning_rate": 1.79930708619941e-05,
+      "loss": 0.043,
+      "step": 3988
+    },
+    {
+      "epoch": 2.427137207179799,
+      "grad_norm": 0.5556389689445496,
+      "learning_rate": 1.7981121838805852e-05,
+      "loss": 0.0522,
+      "step": 3989
+    },
+    {
+      "epoch": 2.4277456647398843,
+      "grad_norm": 0.522250235080719,
+      "learning_rate": 1.7969174556006832e-05,
+      "loss": 0.0527,
+      "step": 3990
+    },
+    {
+      "epoch": 2.4283541222999694,
+      "grad_norm": 0.5426846742630005,
+      "learning_rate": 1.795722901655945e-05,
+      "loss": 0.0481,
+      "step": 3991
+    },
+    {
+      "epoch": 2.428962579860055,
+      "grad_norm": 0.5758845806121826,
+      "learning_rate": 1.7945285223425724e-05,
+      "loss": 0.0414,
+      "step": 3992
+    },
+    {
+      "epoch": 2.42957103742014,
+      "grad_norm": 0.556006669998169,
+      "learning_rate": 1.7933343179567208e-05,
+      "loss": 0.0647,
+      "step": 3993
+    },
+    {
+      "epoch": 2.430179494980225,
+      "grad_norm": 0.6288806796073914,
+      "learning_rate": 1.792140288794502e-05,
+      "loss": 0.0633,
+      "step": 3994
+    },
+    {
+      "epoch": 2.43078795254031,
+      "grad_norm": 0.5981294512748718,
+      "learning_rate": 1.790946435151988e-05,
+      "loss": 0.0627,
+      "step": 3995
+    },
+    {
+      "epoch": 2.4313964101003953,
+      "grad_norm": 0.5165243148803711,
+      "learning_rate": 1.789752757325203e-05,
+      "loss": 0.0411,
+      "step": 3996
+    },
+    {
+      "epoch": 2.432004867660481,
+      "grad_norm": 0.464417964220047,
+      "learning_rate": 1.7885592556101282e-05,
+      "loss": 0.0387,
+      "step": 3997
+    },
+    {
+      "epoch": 2.432613325220566,
+      "grad_norm": 0.4882068932056427,
+      "learning_rate": 1.7873659303027052e-05,
+      "loss": 0.0376,
+      "step": 3998
+    },
+    {
+      "epoch": 2.433221782780651,
+      "grad_norm": 0.5661106109619141,
+      "learning_rate": 1.7861727816988275e-05,
+      "loss": 0.0457,
+      "step": 3999
+    },
+    {
+      "epoch": 2.433830240340736,
+      "grad_norm": 0.5687078833580017,
+      "learning_rate": 1.784979810094346e-05,
+      "loss": 0.0454,
+      "step": 4000
+    },
+    {
+      "epoch": 2.4344386979008212,
+      "grad_norm": 0.6012535691261292,
+      "learning_rate": 1.783787015785067e-05,
+      "loss": 0.0612,
+      "step": 4001
+    },
+    {
+      "epoch": 2.4350471554609068,
+      "grad_norm": 0.5028715133666992,
+      "learning_rate": 1.782594399066756e-05,
+      "loss": 0.0442,
+      "step": 4002
+    },
+    {
+      "epoch": 2.435655613020992,
+      "grad_norm": 0.5349405407905579,
+      "learning_rate": 1.781401960235131e-05,
+      "loss": 0.06,
+      "step": 4003
+    },
+    {
+      "epoch": 2.436264070581077,
+      "grad_norm": 0.5278841257095337,
+      "learning_rate": 1.7802096995858658e-05,
+      "loss": 0.051,
+      "step": 4004
+    },
+    {
+      "epoch": 2.436872528141162,
+      "grad_norm": 0.6222814917564392,
+      "learning_rate": 1.7790176174145944e-05,
+      "loss": 0.0548,
+      "step": 4005
+    },
+    {
+      "epoch": 2.437480985701247,
+      "grad_norm": 0.47187790274620056,
+      "learning_rate": 1.7778257140169016e-05,
+      "loss": 0.0439,
+      "step": 4006
+    },
+    {
+      "epoch": 2.4380894432613327,
+      "grad_norm": 0.48313429951667786,
+      "learning_rate": 1.7766339896883292e-05,
+      "loss": 0.0319,
+      "step": 4007
+    },
+    {
+      "epoch": 2.438697900821418,
+      "grad_norm": 0.5558401346206665,
+      "learning_rate": 1.7754424447243767e-05,
+      "loss": 0.0484,
+      "step": 4008
+    },
+    {
+      "epoch": 2.439306358381503,
+      "grad_norm": 0.45081377029418945,
+      "learning_rate": 1.774251079420497e-05,
+      "loss": 0.0323,
+      "step": 4009
+    },
+    {
+      "epoch": 2.439914815941588,
+      "grad_norm": 0.4787575900554657,
+      "learning_rate": 1.773059894072098e-05,
+      "loss": 0.0441,
+      "step": 4010
+    },
+    {
+      "epoch": 2.440523273501673,
+      "grad_norm": 0.5224798321723938,
+      "learning_rate": 1.7718688889745455e-05,
+      "loss": 0.0532,
+      "step": 4011
+    },
+    {
+      "epoch": 2.4411317310617586,
+      "grad_norm": 0.40973082184791565,
+      "learning_rate": 1.7706780644231592e-05,
+      "loss": 0.037,
+      "step": 4012
+    },
+    {
+      "epoch": 2.4417401886218437,
+      "grad_norm": 0.4948725998401642,
+      "learning_rate": 1.7694874207132127e-05,
+      "loss": 0.0422,
+      "step": 4013
+    },
+    {
+      "epoch": 2.442348646181929,
+      "grad_norm": 0.5685492753982544,
+      "learning_rate": 1.7682969581399358e-05,
+      "loss": 0.0348,
+      "step": 4014
+    },
+    {
+      "epoch": 2.442957103742014,
+      "grad_norm": 0.4990760385990143,
+      "learning_rate": 1.7671066769985155e-05,
+      "loss": 0.0444,
+      "step": 4015
+    },
+    {
+      "epoch": 2.443565561302099,
+      "grad_norm": 0.5178315043449402,
+      "learning_rate": 1.765916577584091e-05,
+      "loss": 0.0404,
+      "step": 4016
+    },
+    {
+      "epoch": 2.4441740188621845,
+      "grad_norm": 0.7539467811584473,
+      "learning_rate": 1.7647266601917557e-05,
+      "loss": 0.0713,
+      "step": 4017
+    },
+    {
+      "epoch": 2.4447824764222696,
+      "grad_norm": 0.4690307676792145,
+      "learning_rate": 1.7635369251165617e-05,
+      "loss": 0.0527,
+      "step": 4018
+    },
+    {
+      "epoch": 2.4453909339823547,
+      "grad_norm": 0.37490659952163696,
+      "learning_rate": 1.7623473726535136e-05,
+      "loss": 0.0346,
+      "step": 4019
+    },
+    {
+      "epoch": 2.44599939154244,
+      "grad_norm": 0.5802814960479736,
+      "learning_rate": 1.7611580030975688e-05,
+      "loss": 0.0545,
+      "step": 4020
+    },
+    {
+      "epoch": 2.446607849102525,
+      "grad_norm": 0.5390932559967041,
+      "learning_rate": 1.7599688167436442e-05,
+      "loss": 0.0484,
+      "step": 4021
+    },
+    {
+      "epoch": 2.4472163066626105,
+      "grad_norm": 0.5329564809799194,
+      "learning_rate": 1.7587798138866067e-05,
+      "loss": 0.0373,
+      "step": 4022
+    },
+    {
+      "epoch": 2.4478247642226956,
+      "grad_norm": 0.4844886064529419,
+      "learning_rate": 1.7575909948212795e-05,
+      "loss": 0.0382,
+      "step": 4023
+    },
+    {
+      "epoch": 2.4484332217827807,
+      "grad_norm": 0.4697389602661133,
+      "learning_rate": 1.7564023598424403e-05,
+      "loss": 0.0416,
+      "step": 4024
+    },
+    {
+      "epoch": 2.4490416793428658,
+      "grad_norm": 0.46831750869750977,
+      "learning_rate": 1.7552139092448216e-05,
+      "loss": 0.0292,
+      "step": 4025
+    },
+    {
+      "epoch": 2.449650136902951,
+      "grad_norm": 0.4629685878753662,
+      "learning_rate": 1.754025643323109e-05,
+      "loss": 0.0332,
+      "step": 4026
+    },
+    {
+      "epoch": 2.450258594463036,
+      "grad_norm": 0.6923487782478333,
+      "learning_rate": 1.7528375623719427e-05,
+      "loss": 0.0602,
+      "step": 4027
+    },
+    {
+      "epoch": 2.4508670520231215,
+      "grad_norm": 0.5200254917144775,
+      "learning_rate": 1.751649666685919e-05,
+      "loss": 0.0419,
+      "step": 4028
+    },
+    {
+      "epoch": 2.4514755095832066,
+      "grad_norm": 0.635124683380127,
+      "learning_rate": 1.7504619565595836e-05,
+      "loss": 0.0439,
+      "step": 4029
+    },
+    {
+      "epoch": 2.4520839671432917,
+      "grad_norm": 0.47341227531433105,
+      "learning_rate": 1.7492744322874414e-05,
+      "loss": 0.0356,
+      "step": 4030
+    },
+    {
+      "epoch": 2.452692424703377,
+      "grad_norm": 0.5724462270736694,
+      "learning_rate": 1.7480870941639485e-05,
+      "loss": 0.0399,
+      "step": 4031
+    },
+    {
+      "epoch": 2.453300882263462,
+      "grad_norm": 0.5328457355499268,
+      "learning_rate": 1.7468999424835144e-05,
+      "loss": 0.0526,
+      "step": 4032
+    },
+    {
+      "epoch": 2.4539093398235474,
+      "grad_norm": 0.5413589477539062,
+      "learning_rate": 1.745712977540504e-05,
+      "loss": 0.0466,
+      "step": 4033
+    },
+    {
+      "epoch": 2.4545177973836325,
+      "grad_norm": 0.5428217053413391,
+      "learning_rate": 1.744526199629235e-05,
+      "loss": 0.0372,
+      "step": 4034
+    },
+    {
+      "epoch": 2.4551262549437176,
+      "grad_norm": 0.5191711187362671,
+      "learning_rate": 1.743339609043979e-05,
+      "loss": 0.0486,
+      "step": 4035
+    },
+    {
+      "epoch": 2.4557347125038027,
+      "grad_norm": 0.5395163297653198,
+      "learning_rate": 1.742153206078961e-05,
+      "loss": 0.0505,
+      "step": 4036
+    },
+    {
+      "epoch": 2.456343170063888,
+      "grad_norm": 0.5724096894264221,
+      "learning_rate": 1.7409669910283583e-05,
+      "loss": 0.0428,
+      "step": 4037
+    },
+    {
+      "epoch": 2.4569516276239733,
+      "grad_norm": 0.6358472108840942,
+      "learning_rate": 1.739780964186304e-05,
+      "loss": 0.0344,
+      "step": 4038
+    },
+    {
+      "epoch": 2.4575600851840584,
+      "grad_norm": 0.6139354109764099,
+      "learning_rate": 1.738595125846884e-05,
+      "loss": 0.082,
+      "step": 4039
+    },
+    {
+      "epoch": 2.4581685427441435,
+      "grad_norm": 0.5228933095932007,
+      "learning_rate": 1.7374094763041347e-05,
+      "loss": 0.0266,
+      "step": 4040
+    },
+    {
+      "epoch": 2.4587770003042286,
+      "grad_norm": 0.5466329455375671,
+      "learning_rate": 1.73622401585205e-05,
+      "loss": 0.0438,
+      "step": 4041
+    },
+    {
+      "epoch": 2.4593854578643137,
+      "grad_norm": 0.5522369742393494,
+      "learning_rate": 1.735038744784573e-05,
+      "loss": 0.0568,
+      "step": 4042
+    },
+    {
+      "epoch": 2.4599939154243993,
+      "grad_norm": 0.7028734087944031,
+      "learning_rate": 1.733853663395602e-05,
+      "loss": 0.0511,
+      "step": 4043
+    },
+    {
+      "epoch": 2.4606023729844844,
+      "grad_norm": 0.48779624700546265,
+      "learning_rate": 1.732668771978988e-05,
+      "loss": 0.0265,
+      "step": 4044
+    },
+    {
+      "epoch": 2.4612108305445695,
+      "grad_norm": 0.5411667823791504,
+      "learning_rate": 1.7314840708285354e-05,
+      "loss": 0.0484,
+      "step": 4045
+    },
+    {
+      "epoch": 2.4618192881046546,
+      "grad_norm": 0.5587241649627686,
+      "learning_rate": 1.7302995602379996e-05,
+      "loss": 0.0461,
+      "step": 4046
+    },
+    {
+      "epoch": 2.4624277456647397,
+      "grad_norm": 0.49550849199295044,
+      "learning_rate": 1.7291152405010898e-05,
+      "loss": 0.051,
+      "step": 4047
+    },
+    {
+      "epoch": 2.463036203224825,
+      "grad_norm": 0.5094623565673828,
+      "learning_rate": 1.727931111911468e-05,
+      "loss": 0.0483,
+      "step": 4048
+    },
+    {
+      "epoch": 2.4636446607849103,
+      "grad_norm": 0.5672603249549866,
+      "learning_rate": 1.72674717476275e-05,
+      "loss": 0.0535,
+      "step": 4049
+    },
+    {
+      "epoch": 2.4642531183449954,
+      "grad_norm": 0.4778454005718231,
+      "learning_rate": 1.725563429348501e-05,
+      "loss": 0.0443,
+      "step": 4050
+    },
+    {
+      "epoch": 2.4648615759050805,
+      "grad_norm": 0.5189485549926758,
+      "learning_rate": 1.7243798759622414e-05,
+      "loss": 0.0483,
+      "step": 4051
+    },
+    {
+      "epoch": 2.4654700334651656,
+      "grad_norm": 0.5488086938858032,
+      "learning_rate": 1.7231965148974438e-05,
+      "loss": 0.0477,
+      "step": 4052
+    },
+    {
+      "epoch": 2.466078491025251,
+      "grad_norm": 0.5553414225578308,
+      "learning_rate": 1.7220133464475312e-05,
+      "loss": 0.0434,
+      "step": 4053
+    },
+    {
+      "epoch": 2.466686948585336,
+      "grad_norm": 0.5483329892158508,
+      "learning_rate": 1.7208303709058806e-05,
+      "loss": 0.0504,
+      "step": 4054
+    },
+    {
+      "epoch": 2.4672954061454213,
+      "grad_norm": 0.4734504818916321,
+      "learning_rate": 1.7196475885658214e-05,
+      "loss": 0.0363,
+      "step": 4055
+    },
+    {
+      "epoch": 2.4679038637055064,
+      "grad_norm": 0.4984531104564667,
+      "learning_rate": 1.7184649997206324e-05,
+      "loss": 0.039,
+      "step": 4056
+    },
+    {
+      "epoch": 2.4685123212655915,
+      "grad_norm": 0.5191166996955872,
+      "learning_rate": 1.7172826046635483e-05,
+      "loss": 0.044,
+      "step": 4057
+    },
+    {
+      "epoch": 2.469120778825677,
+      "grad_norm": 0.5484380722045898,
+      "learning_rate": 1.7161004036877526e-05,
+      "loss": 0.0433,
+      "step": 4058
+    },
+    {
+      "epoch": 2.469729236385762,
+      "grad_norm": 0.570693850517273,
+      "learning_rate": 1.714918397086383e-05,
+      "loss": 0.0468,
+      "step": 4059
+    },
+    {
+      "epoch": 2.4703376939458472,
+      "grad_norm": 0.5350715517997742,
+      "learning_rate": 1.7137365851525255e-05,
+      "loss": 0.0406,
+      "step": 4060
+    },
+    {
+      "epoch": 2.4709461515059323,
+      "grad_norm": 0.5469467639923096,
+      "learning_rate": 1.7125549681792233e-05,
+      "loss": 0.0499,
+      "step": 4061
+    },
+    {
+      "epoch": 2.4715546090660174,
+      "grad_norm": 0.5633114576339722,
+      "learning_rate": 1.7113735464594665e-05,
+      "loss": 0.0371,
+      "step": 4062
+    },
+    {
+      "epoch": 2.472163066626103,
+      "grad_norm": 0.6275270581245422,
+      "learning_rate": 1.7101923202861974e-05,
+      "loss": 0.0482,
+      "step": 4063
+    },
+    {
+      "epoch": 2.472771524186188,
+      "grad_norm": 0.6021205186843872,
+      "learning_rate": 1.7090112899523132e-05,
+      "loss": 0.0524,
+      "step": 4064
+    },
+    {
+      "epoch": 2.473379981746273,
+      "grad_norm": 0.6164666414260864,
+      "learning_rate": 1.7078304557506593e-05,
+      "loss": 0.0459,
+      "step": 4065
+    },
+    {
+      "epoch": 2.4739884393063583,
+      "grad_norm": 0.5763289332389832,
+      "learning_rate": 1.7066498179740318e-05,
+      "loss": 0.046,
+      "step": 4066
+    },
+    {
+      "epoch": 2.4745968968664434,
+      "grad_norm": 0.6294301152229309,
+      "learning_rate": 1.705469376915182e-05,
+      "loss": 0.0465,
+      "step": 4067
+    },
+    {
+      "epoch": 2.475205354426529,
+      "grad_norm": 0.409706711769104,
+      "learning_rate": 1.7042891328668094e-05,
+      "loss": 0.0301,
+      "step": 4068
+    },
+    {
+      "epoch": 2.475813811986614,
+      "grad_norm": 0.42503032088279724,
+      "learning_rate": 1.7031090861215646e-05,
+      "loss": 0.0307,
+      "step": 4069
+    },
+    {
+      "epoch": 2.476422269546699,
+      "grad_norm": 0.5593361258506775,
+      "learning_rate": 1.7019292369720493e-05,
+      "loss": 0.0366,
+      "step": 4070
+    },
+    {
+      "epoch": 2.477030727106784,
+      "grad_norm": 0.5310861468315125,
+      "learning_rate": 1.700749585710819e-05,
+      "loss": 0.0367,
+      "step": 4071
+    },
+    {
+      "epoch": 2.4776391846668693,
+      "grad_norm": 0.5321909189224243,
+      "learning_rate": 1.6995701326303776e-05,
+      "loss": 0.031,
+      "step": 4072
+    },
+    {
+      "epoch": 2.478247642226955,
+      "grad_norm": 0.5682234764099121,
+      "learning_rate": 1.6983908780231782e-05,
+      "loss": 0.0455,
+      "step": 4073
+    },
+    {
+      "epoch": 2.47885609978704,
+      "grad_norm": 0.46487629413604736,
+      "learning_rate": 1.6972118221816298e-05,
+      "loss": 0.032,
+      "step": 4074
+    },
+    {
+      "epoch": 2.479464557347125,
+      "grad_norm": 0.616710364818573,
+      "learning_rate": 1.696032965398087e-05,
+      "loss": 0.0516,
+      "step": 4075
+    },
+    {
+      "epoch": 2.48007301490721,
+      "grad_norm": 0.5947301387786865,
+      "learning_rate": 1.6948543079648575e-05,
+      "loss": 0.0376,
+      "step": 4076
+    },
+    {
+      "epoch": 2.480681472467295,
+      "grad_norm": 0.5795893669128418,
+      "learning_rate": 1.6936758501742e-05,
+      "loss": 0.0487,
+      "step": 4077
+    },
+    {
+      "epoch": 2.4812899300273807,
+      "grad_norm": 0.5290836095809937,
+      "learning_rate": 1.6924975923183228e-05,
+      "loss": 0.0292,
+      "step": 4078
+    },
+    {
+      "epoch": 2.481898387587466,
+      "grad_norm": 0.4861297309398651,
+      "learning_rate": 1.6913195346893828e-05,
+      "loss": 0.0336,
+      "step": 4079
+    },
+    {
+      "epoch": 2.482506845147551,
+      "grad_norm": 0.5103831887245178,
+      "learning_rate": 1.690141677579492e-05,
+      "loss": 0.0422,
+      "step": 4080
+    },
+    {
+      "epoch": 2.483115302707636,
+      "grad_norm": 0.42127537727355957,
+      "learning_rate": 1.688964021280709e-05,
+      "loss": 0.0317,
+      "step": 4081
+    },
+    {
+      "epoch": 2.483723760267721,
+      "grad_norm": 0.7771899104118347,
+      "learning_rate": 1.6877865660850427e-05,
+      "loss": 0.0537,
+      "step": 4082
+    },
+    {
+      "epoch": 2.4843322178278067,
+      "grad_norm": 0.5473512411117554,
+      "learning_rate": 1.6866093122844523e-05,
+      "loss": 0.0322,
+      "step": 4083
+    },
+    {
+      "epoch": 2.4849406753878918,
+      "grad_norm": 0.5636205077171326,
+      "learning_rate": 1.6854322601708495e-05,
+      "loss": 0.0538,
+      "step": 4084
+    },
+    {
+      "epoch": 2.485549132947977,
+      "grad_norm": 0.5534929037094116,
+      "learning_rate": 1.6842554100360937e-05,
+      "loss": 0.0604,
+      "step": 4085
+    },
+    {
+      "epoch": 2.486157590508062,
+      "grad_norm": 0.5635712742805481,
+      "learning_rate": 1.683078762171993e-05,
+      "loss": 0.0418,
+      "step": 4086
+    },
+    {
+      "epoch": 2.486766048068147,
+      "grad_norm": 0.5449815392494202,
+      "learning_rate": 1.6819023168703094e-05,
+      "loss": 0.0422,
+      "step": 4087
+    },
+    {
+      "epoch": 2.4873745056282326,
+      "grad_norm": 0.5356243252754211,
+      "learning_rate": 1.6807260744227513e-05,
+      "loss": 0.052,
+      "step": 4088
+    },
+    {
+      "epoch": 2.4879829631883177,
+      "grad_norm": 0.525938093662262,
+      "learning_rate": 1.6795500351209766e-05,
+      "loss": 0.0457,
+      "step": 4089
+    },
+    {
+      "epoch": 2.488591420748403,
+      "grad_norm": 0.5305033922195435,
+      "learning_rate": 1.6783741992565963e-05,
+      "loss": 0.0469,
+      "step": 4090
+    },
+    {
+      "epoch": 2.489199878308488,
+      "grad_norm": 0.47671419382095337,
+      "learning_rate": 1.6771985671211673e-05,
+      "loss": 0.0426,
+      "step": 4091
+    },
+    {
+      "epoch": 2.489808335868573,
+      "grad_norm": 0.6762577891349792,
+      "learning_rate": 1.676023139006198e-05,
+      "loss": 0.0523,
+      "step": 4092
+    },
+    {
+      "epoch": 2.4904167934286585,
+      "grad_norm": 0.5468366146087646,
+      "learning_rate": 1.6748479152031436e-05,
+      "loss": 0.0502,
+      "step": 4093
+    },
+    {
+      "epoch": 2.4910252509887436,
+      "grad_norm": 0.5539106130599976,
+      "learning_rate": 1.6736728960034137e-05,
+      "loss": 0.0457,
+      "step": 4094
+    },
+    {
+      "epoch": 2.4916337085488287,
+      "grad_norm": 0.5105188488960266,
+      "learning_rate": 1.6724980816983625e-05,
+      "loss": 0.0392,
+      "step": 4095
+    },
+    {
+      "epoch": 2.492242166108914,
+      "grad_norm": 0.5564658045768738,
+      "learning_rate": 1.6713234725792938e-05,
+      "loss": 0.0382,
+      "step": 4096
+    },
+    {
+      "epoch": 2.492850623668999,
+      "grad_norm": 0.6725486516952515,
+      "learning_rate": 1.6701490689374642e-05,
+      "loss": 0.0393,
+      "step": 4097
+    },
+    {
+      "epoch": 2.4934590812290844,
+      "grad_norm": 0.46235400438308716,
+      "learning_rate": 1.6689748710640756e-05,
+      "loss": 0.0401,
+      "step": 4098
+    },
+    {
+      "epoch": 2.4940675387891695,
+      "grad_norm": 0.5110651254653931,
+      "learning_rate": 1.6678008792502793e-05,
+      "loss": 0.035,
+      "step": 4099
+    },
+    {
+      "epoch": 2.4946759963492546,
+      "grad_norm": 0.6853166818618774,
+      "learning_rate": 1.6666270937871774e-05,
+      "loss": 0.0396,
+      "step": 4100
+    },
+    {
+      "epoch": 2.4952844539093397,
+      "grad_norm": 0.4850734770298004,
+      "learning_rate": 1.6654535149658203e-05,
+      "loss": 0.0599,
+      "step": 4101
+    },
+    {
+      "epoch": 2.495892911469425,
+      "grad_norm": 0.5870274305343628,
+      "learning_rate": 1.6642801430772048e-05,
+      "loss": 0.0399,
+      "step": 4102
+    },
+    {
+      "epoch": 2.4965013690295104,
+      "grad_norm": 0.561387300491333,
+      "learning_rate": 1.6631069784122803e-05,
+      "loss": 0.033,
+      "step": 4103
+    },
+    {
+      "epoch": 2.4971098265895955,
+      "grad_norm": 0.5187516212463379,
+      "learning_rate": 1.6619340212619417e-05,
+      "loss": 0.0425,
+      "step": 4104
+    },
+    {
+      "epoch": 2.4977182841496806,
+      "grad_norm": 0.4592403173446655,
+      "learning_rate": 1.660761271917033e-05,
+      "loss": 0.0334,
+      "step": 4105
+    },
+    {
+      "epoch": 2.4983267417097657,
+      "grad_norm": 0.5504740476608276,
+      "learning_rate": 1.6595887306683474e-05,
+      "loss": 0.0474,
+      "step": 4106
+    },
+    {
+      "epoch": 2.4989351992698507,
+      "grad_norm": 0.5082219243049622,
+      "learning_rate": 1.658416397806627e-05,
+      "loss": 0.0448,
+      "step": 4107
+    },
+    {
+      "epoch": 2.4995436568299363,
+      "grad_norm": 0.6115595102310181,
+      "learning_rate": 1.6572442736225614e-05,
+      "loss": 0.0633,
+      "step": 4108
+    },
+    {
+      "epoch": 2.5001521143900214,
+      "grad_norm": 0.4949551224708557,
+      "learning_rate": 1.656072358406787e-05,
+      "loss": 0.0471,
+      "step": 4109
+    },
+    {
+      "epoch": 2.5007605719501065,
+      "grad_norm": 0.6002260446548462,
+      "learning_rate": 1.654900652449892e-05,
+      "loss": 0.0593,
+      "step": 4110
+    },
+    {
+      "epoch": 2.5013690295101916,
+      "grad_norm": 0.5908356308937073,
+      "learning_rate": 1.6537291560424097e-05,
+      "loss": 0.0536,
+      "step": 4111
+    },
+    {
+      "epoch": 2.5019774870702767,
+      "grad_norm": 0.6577576398849487,
+      "learning_rate": 1.6525578694748217e-05,
+      "loss": 0.0594,
+      "step": 4112
+    },
+    {
+      "epoch": 2.502585944630362,
+      "grad_norm": 0.6429257392883301,
+      "learning_rate": 1.6513867930375596e-05,
+      "loss": 0.0613,
+      "step": 4113
+    },
+    {
+      "epoch": 2.5031944021904473,
+      "grad_norm": 0.5084056854248047,
+      "learning_rate": 1.650215927021001e-05,
+      "loss": 0.0447,
+      "step": 4114
+    },
+    {
+      "epoch": 2.5038028597505324,
+      "grad_norm": 0.6193863749504089,
+      "learning_rate": 1.649045271715472e-05,
+      "loss": 0.0486,
+      "step": 4115
+    },
+    {
+      "epoch": 2.5044113173106175,
+      "grad_norm": 0.5566953420639038,
+      "learning_rate": 1.6478748274112445e-05,
+      "loss": 0.0441,
+      "step": 4116
+    },
+    {
+      "epoch": 2.5050197748707026,
+      "grad_norm": 0.5336138606071472,
+      "learning_rate": 1.646704594398543e-05,
+      "loss": 0.0477,
+      "step": 4117
+    },
+    {
+      "epoch": 2.505628232430788,
+      "grad_norm": 0.544262170791626,
+      "learning_rate": 1.6455345729675348e-05,
+      "loss": 0.0542,
+      "step": 4118
+    },
+    {
+      "epoch": 2.5062366899908732,
+      "grad_norm": 0.46165311336517334,
+      "learning_rate": 1.6443647634083353e-05,
+      "loss": 0.0326,
+      "step": 4119
+    },
+    {
+      "epoch": 2.5068451475509583,
+      "grad_norm": 0.4707409143447876,
+      "learning_rate": 1.6431951660110113e-05,
+      "loss": 0.0401,
+      "step": 4120
+    },
+    {
+      "epoch": 2.5074536051110434,
+      "grad_norm": 0.5299643874168396,
+      "learning_rate": 1.6420257810655727e-05,
+      "loss": 0.0488,
+      "step": 4121
+    },
+    {
+      "epoch": 2.5080620626711285,
+      "grad_norm": 0.9495867490768433,
+      "learning_rate": 1.640856608861977e-05,
+      "loss": 0.0375,
+      "step": 4122
+    },
+    {
+      "epoch": 2.508670520231214,
+      "grad_norm": 0.5048344135284424,
+      "learning_rate": 1.6396876496901324e-05,
+      "loss": 0.0285,
+      "step": 4123
+    },
+    {
+      "epoch": 2.509278977791299,
+      "grad_norm": 0.6059793829917908,
+      "learning_rate": 1.6385189038398914e-05,
+      "loss": 0.0442,
+      "step": 4124
+    },
+    {
+      "epoch": 2.5098874353513843,
+      "grad_norm": 0.5671916604042053,
+      "learning_rate": 1.637350371601053e-05,
+      "loss": 0.0442,
+      "step": 4125
+    },
+    {
+      "epoch": 2.5104958929114694,
+      "grad_norm": 0.49703407287597656,
+      "learning_rate": 1.6361820532633665e-05,
+      "loss": 0.0431,
+      "step": 4126
+    },
+    {
+      "epoch": 2.5111043504715544,
+      "grad_norm": 0.5402058959007263,
+      "learning_rate": 1.6350139491165246e-05,
+      "loss": 0.0349,
+      "step": 4127
+    },
+    {
+      "epoch": 2.51171280803164,
+      "grad_norm": 0.5467166900634766,
+      "learning_rate": 1.6338460594501692e-05,
+      "loss": 0.0577,
+      "step": 4128
+    },
+    {
+      "epoch": 2.512321265591725,
+      "grad_norm": 0.5586373805999756,
+      "learning_rate": 1.632678384553887e-05,
+      "loss": 0.0426,
+      "step": 4129
+    },
+    {
+      "epoch": 2.51292972315181,
+      "grad_norm": 0.613847017288208,
+      "learning_rate": 1.6315109247172146e-05,
+      "loss": 0.0535,
+      "step": 4130
+    },
+    {
+      "epoch": 2.5135381807118953,
+      "grad_norm": 0.47474873065948486,
+      "learning_rate": 1.6303436802296325e-05,
+      "loss": 0.0426,
+      "step": 4131
+    },
+    {
+      "epoch": 2.5141466382719804,
+      "grad_norm": 0.5153751373291016,
+      "learning_rate": 1.629176651380568e-05,
+      "loss": 0.0478,
+      "step": 4132
+    },
+    {
+      "epoch": 2.514755095832066,
+      "grad_norm": 0.7585486769676208,
+      "learning_rate": 1.6280098384593966e-05,
+      "loss": 0.0642,
+      "step": 4133
+    },
+    {
+      "epoch": 2.515363553392151,
+      "grad_norm": 0.5414904952049255,
+      "learning_rate": 1.626843241755439e-05,
+      "loss": 0.0553,
+      "step": 4134
+    },
+    {
+      "epoch": 2.515972010952236,
+      "grad_norm": 0.8704773783683777,
+      "learning_rate": 1.625676861557962e-05,
+      "loss": 0.0435,
+      "step": 4135
+    },
+    {
+      "epoch": 2.516580468512321,
+      "grad_norm": 0.551547110080719,
+      "learning_rate": 1.6245106981561804e-05,
+      "loss": 0.0475,
+      "step": 4136
+    },
+    {
+      "epoch": 2.5171889260724063,
+      "grad_norm": 0.614238977432251,
+      "learning_rate": 1.6233447518392537e-05,
+      "loss": 0.0593,
+      "step": 4137
+    },
+    {
+      "epoch": 2.517797383632492,
+      "grad_norm": 0.7122321724891663,
+      "learning_rate": 1.622179022896287e-05,
+      "loss": 0.0735,
+      "step": 4138
+    },
+    {
+      "epoch": 2.518405841192577,
+      "grad_norm": 0.5139821767807007,
+      "learning_rate": 1.6210135116163333e-05,
+      "loss": 0.0405,
+      "step": 4139
+    },
+    {
+      "epoch": 2.519014298752662,
+      "grad_norm": 0.6365357637405396,
+      "learning_rate": 1.6198482182883912e-05,
+      "loss": 0.0634,
+      "step": 4140
+    },
+    {
+      "epoch": 2.519622756312747,
+      "grad_norm": 0.5459882616996765,
+      "learning_rate": 1.618683143201404e-05,
+      "loss": 0.0341,
+      "step": 4141
+    },
+    {
+      "epoch": 2.520231213872832,
+      "grad_norm": 0.45921745896339417,
+      "learning_rate": 1.6175182866442624e-05,
+      "loss": 0.0335,
+      "step": 4142
+    },
+    {
+      "epoch": 2.5208396714329178,
+      "grad_norm": 0.524821400642395,
+      "learning_rate": 1.6163536489058023e-05,
+      "loss": 0.0544,
+      "step": 4143
+    },
+    {
+      "epoch": 2.521448128993003,
+      "grad_norm": 0.5286942720413208,
+      "learning_rate": 1.6151892302748046e-05,
+      "loss": 0.046,
+      "step": 4144
+    },
+    {
+      "epoch": 2.522056586553088,
+      "grad_norm": 0.5625891089439392,
+      "learning_rate": 1.614025031039997e-05,
+      "loss": 0.0543,
+      "step": 4145
+    },
+    {
+      "epoch": 2.522665044113173,
+      "grad_norm": 0.5208843350410461,
+      "learning_rate": 1.612861051490053e-05,
+      "loss": 0.0391,
+      "step": 4146
+    },
+    {
+      "epoch": 2.523273501673258,
+      "grad_norm": 0.6042157411575317,
+      "learning_rate": 1.6116972919135907e-05,
+      "loss": 0.0726,
+      "step": 4147
+    },
+    {
+      "epoch": 2.5238819592333437,
+      "grad_norm": 0.43940305709838867,
+      "learning_rate": 1.6105337525991723e-05,
+      "loss": 0.0372,
+      "step": 4148
+    },
+    {
+      "epoch": 2.524490416793429,
+      "grad_norm": 0.4963323771953583,
+      "learning_rate": 1.6093704338353097e-05,
+      "loss": 0.0383,
+      "step": 4149
+    },
+    {
+      "epoch": 2.525098874353514,
+      "grad_norm": 0.4309125244617462,
+      "learning_rate": 1.6082073359104564e-05,
+      "loss": 0.0448,
+      "step": 4150
+    },
+    {
+      "epoch": 2.525707331913599,
+      "grad_norm": 0.4299990236759186,
+      "learning_rate": 1.6070444591130113e-05,
+      "loss": 0.0361,
+      "step": 4151
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.5838400721549988,
+      "learning_rate": 1.60588180373132e-05,
+      "loss": 0.0494,
+      "step": 4152
+    },
+    {
+      "epoch": 2.5269242470337696,
+      "grad_norm": 0.4594270586967468,
+      "learning_rate": 1.6047193700536734e-05,
+      "loss": 0.0425,
+      "step": 4153
+    },
+    {
+      "epoch": 2.5275327045938547,
+      "grad_norm": 0.6414621472358704,
+      "learning_rate": 1.6035571583683055e-05,
+      "loss": 0.0606,
+      "step": 4154
+    },
+    {
+      "epoch": 2.52814116215394,
+      "grad_norm": 0.4599505364894867,
+      "learning_rate": 1.6023951689633965e-05,
+      "loss": 0.0304,
+      "step": 4155
+    },
+    {
+      "epoch": 2.528749619714025,
+      "grad_norm": 0.4809854030609131,
+      "learning_rate": 1.601233402127072e-05,
+      "loss": 0.0463,
+      "step": 4156
+    },
+    {
+      "epoch": 2.52935807727411,
+      "grad_norm": 0.5729118585586548,
+      "learning_rate": 1.600071858147401e-05,
+      "loss": 0.0412,
+      "step": 4157
+    },
+    {
+      "epoch": 2.5299665348341955,
+      "grad_norm": 0.5415656566619873,
+      "learning_rate": 1.5989105373123986e-05,
+      "loss": 0.0426,
+      "step": 4158
+    },
+    {
+      "epoch": 2.5305749923942806,
+      "grad_norm": 0.46047234535217285,
+      "learning_rate": 1.5977494399100242e-05,
+      "loss": 0.0414,
+      "step": 4159
+    },
+    {
+      "epoch": 2.5311834499543657,
+      "grad_norm": 0.5674901604652405,
+      "learning_rate": 1.59658856622818e-05,
+      "loss": 0.061,
+      "step": 4160
+    },
+    {
+      "epoch": 2.531791907514451,
+      "grad_norm": 0.539481520652771,
+      "learning_rate": 1.5954279165547158e-05,
+      "loss": 0.0422,
+      "step": 4161
+    },
+    {
+      "epoch": 2.532400365074536,
+      "grad_norm": 0.4912717342376709,
+      "learning_rate": 1.594267491177424e-05,
+      "loss": 0.0409,
+      "step": 4162
+    },
+    {
+      "epoch": 2.5330088226346215,
+      "grad_norm": 0.5007967948913574,
+      "learning_rate": 1.5931072903840415e-05,
+      "loss": 0.045,
+      "step": 4163
+    },
+    {
+      "epoch": 2.5336172801947066,
+      "grad_norm": 0.5245926976203918,
+      "learning_rate": 1.5919473144622502e-05,
+      "loss": 0.0438,
+      "step": 4164
+    },
+    {
+      "epoch": 2.5342257377547917,
+      "grad_norm": 0.5035905241966248,
+      "learning_rate": 1.5907875636996748e-05,
+      "loss": 0.0442,
+      "step": 4165
+    },
+    {
+      "epoch": 2.5348341953148767,
+      "grad_norm": 0.573156476020813,
+      "learning_rate": 1.5896280383838856e-05,
+      "loss": 0.0642,
+      "step": 4166
+    },
+    {
+      "epoch": 2.535442652874962,
+      "grad_norm": 0.5619865655899048,
+      "learning_rate": 1.5884687388023975e-05,
+      "loss": 0.0432,
+      "step": 4167
+    },
+    {
+      "epoch": 2.5360511104350474,
+      "grad_norm": 0.4712476432323456,
+      "learning_rate": 1.587309665242667e-05,
+      "loss": 0.0344,
+      "step": 4168
+    },
+    {
+      "epoch": 2.5366595679951325,
+      "grad_norm": 0.5480785965919495,
+      "learning_rate": 1.586150817992097e-05,
+      "loss": 0.0429,
+      "step": 4169
+    },
+    {
+      "epoch": 2.5372680255552176,
+      "grad_norm": 0.602826714515686,
+      "learning_rate": 1.5849921973380332e-05,
+      "loss": 0.047,
+      "step": 4170
+    },
+    {
+      "epoch": 2.5378764831153027,
+      "grad_norm": 0.5343915820121765,
+      "learning_rate": 1.5838338035677645e-05,
+      "loss": 0.0414,
+      "step": 4171
+    },
+    {
+      "epoch": 2.5384849406753878,
+      "grad_norm": 0.5871137380599976,
+      "learning_rate": 1.582675636968525e-05,
+      "loss": 0.0423,
+      "step": 4172
+    },
+    {
+      "epoch": 2.5390933982354733,
+      "grad_norm": 0.47411757707595825,
+      "learning_rate": 1.5815176978274924e-05,
+      "loss": 0.0368,
+      "step": 4173
+    },
+    {
+      "epoch": 2.5397018557955584,
+      "grad_norm": 0.5202276110649109,
+      "learning_rate": 1.580359986431786e-05,
+      "loss": 0.0423,
+      "step": 4174
+    },
+    {
+      "epoch": 2.5403103133556435,
+      "grad_norm": 0.5111712217330933,
+      "learning_rate": 1.57920250306847e-05,
+      "loss": 0.0434,
+      "step": 4175
+    },
+    {
+      "epoch": 2.5409187709157286,
+      "grad_norm": 0.5714934468269348,
+      "learning_rate": 1.578045248024553e-05,
+      "loss": 0.0529,
+      "step": 4176
+    },
+    {
+      "epoch": 2.5415272284758137,
+      "grad_norm": 0.5358994603157043,
+      "learning_rate": 1.5768882215869858e-05,
+      "loss": 0.043,
+      "step": 4177
+    },
+    {
+      "epoch": 2.5421356860358992,
+      "grad_norm": 0.5511855483055115,
+      "learning_rate": 1.5757314240426613e-05,
+      "loss": 0.0419,
+      "step": 4178
+    },
+    {
+      "epoch": 2.5427441435959843,
+      "grad_norm": 0.6131108999252319,
+      "learning_rate": 1.5745748556784194e-05,
+      "loss": 0.0716,
+      "step": 4179
+    },
+    {
+      "epoch": 2.5433526011560694,
+      "grad_norm": 0.39662912487983704,
+      "learning_rate": 1.5734185167810395e-05,
+      "loss": 0.0278,
+      "step": 4180
+    },
+    {
+      "epoch": 2.5439610587161545,
+      "grad_norm": 0.5008838772773743,
+      "learning_rate": 1.572262407637245e-05,
+      "loss": 0.0401,
+      "step": 4181
+    },
+    {
+      "epoch": 2.5445695162762396,
+      "grad_norm": 0.6066216230392456,
+      "learning_rate": 1.5711065285337035e-05,
+      "loss": 0.0686,
+      "step": 4182
+    },
+    {
+      "epoch": 2.545177973836325,
+      "grad_norm": 0.5468981862068176,
+      "learning_rate": 1.5699508797570255e-05,
+      "loss": 0.028,
+      "step": 4183
+    },
+    {
+      "epoch": 2.5457864313964103,
+      "grad_norm": 0.5890219807624817,
+      "learning_rate": 1.5687954615937623e-05,
+      "loss": 0.0497,
+      "step": 4184
+    },
+    {
+      "epoch": 2.5463948889564954,
+      "grad_norm": 0.489978164434433,
+      "learning_rate": 1.56764027433041e-05,
+      "loss": 0.0387,
+      "step": 4185
+    },
+    {
+      "epoch": 2.5470033465165804,
+      "grad_norm": 0.5345786809921265,
+      "learning_rate": 1.5664853182534077e-05,
+      "loss": 0.0441,
+      "step": 4186
+    },
+    {
+      "epoch": 2.5476118040766655,
+      "grad_norm": 0.5965332984924316,
+      "learning_rate": 1.565330593649135e-05,
+      "loss": 0.0456,
+      "step": 4187
+    },
+    {
+      "epoch": 2.548220261636751,
+      "grad_norm": 0.5223818421363831,
+      "learning_rate": 1.564176100803916e-05,
+      "loss": 0.0447,
+      "step": 4188
+    },
+    {
+      "epoch": 2.548828719196836,
+      "grad_norm": 0.6047852635383606,
+      "learning_rate": 1.5630218400040174e-05,
+      "loss": 0.049,
+      "step": 4189
+    },
+    {
+      "epoch": 2.5494371767569213,
+      "grad_norm": 0.6201773285865784,
+      "learning_rate": 1.561867811535648e-05,
+      "loss": 0.0598,
+      "step": 4190
+    },
+    {
+      "epoch": 2.5500456343170064,
+      "grad_norm": 0.4544050693511963,
+      "learning_rate": 1.5607140156849564e-05,
+      "loss": 0.0313,
+      "step": 4191
+    },
+    {
+      "epoch": 2.5506540918770915,
+      "grad_norm": 0.5224777460098267,
+      "learning_rate": 1.5595604527380387e-05,
+      "loss": 0.0538,
+      "step": 4192
+    },
+    {
+      "epoch": 2.551262549437177,
+      "grad_norm": 0.57136070728302,
+      "learning_rate": 1.5584071229809294e-05,
+      "loss": 0.0452,
+      "step": 4193
+    },
+    {
+      "epoch": 2.551871006997262,
+      "grad_norm": 0.45948007702827454,
+      "learning_rate": 1.5572540266996047e-05,
+      "loss": 0.0289,
+      "step": 4194
+    },
+    {
+      "epoch": 2.552479464557347,
+      "grad_norm": 0.5531154274940491,
+      "learning_rate": 1.5561011641799872e-05,
+      "loss": 0.0417,
+      "step": 4195
+    },
+    {
+      "epoch": 2.5530879221174323,
+      "grad_norm": 0.6279348134994507,
+      "learning_rate": 1.5549485357079373e-05,
+      "loss": 0.0619,
+      "step": 4196
+    },
+    {
+      "epoch": 2.5536963796775174,
+      "grad_norm": 0.5489982962608337,
+      "learning_rate": 1.5537961415692585e-05,
+      "loss": 0.0395,
+      "step": 4197
+    },
+    {
+      "epoch": 2.554304837237603,
+      "grad_norm": 0.4770684540271759,
+      "learning_rate": 1.5526439820496965e-05,
+      "loss": 0.0422,
+      "step": 4198
+    },
+    {
+      "epoch": 2.5549132947976876,
+      "grad_norm": 0.5624024868011475,
+      "learning_rate": 1.5514920574349397e-05,
+      "loss": 0.0552,
+      "step": 4199
+    },
+    {
+      "epoch": 2.555521752357773,
+      "grad_norm": 0.5154407620429993,
+      "learning_rate": 1.5503403680106168e-05,
+      "loss": 0.0474,
+      "step": 4200
+    },
+    {
+      "epoch": 2.556130209917858,
+      "grad_norm": 0.48511430621147156,
+      "learning_rate": 1.549188914062298e-05,
+      "loss": 0.0311,
+      "step": 4201
+    },
+    {
+      "epoch": 2.5567386674779433,
+      "grad_norm": 0.4834255874156952,
+      "learning_rate": 1.5480376958754976e-05,
+      "loss": 0.0426,
+      "step": 4202
+    },
+    {
+      "epoch": 2.557347125038029,
+      "grad_norm": 0.5208700299263,
+      "learning_rate": 1.5468867137356696e-05,
+      "loss": 0.0357,
+      "step": 4203
+    },
+    {
+      "epoch": 2.5579555825981135,
+      "grad_norm": 0.4725053906440735,
+      "learning_rate": 1.545735967928207e-05,
+      "loss": 0.0489,
+      "step": 4204
+    },
+    {
+      "epoch": 2.558564040158199,
+      "grad_norm": 0.4903971552848816,
+      "learning_rate": 1.54458545873845e-05,
+      "loss": 0.048,
+      "step": 4205
+    },
+    {
+      "epoch": 2.559172497718284,
+      "grad_norm": 0.5080806612968445,
+      "learning_rate": 1.543435186451676e-05,
+      "loss": 0.0525,
+      "step": 4206
+    },
+    {
+      "epoch": 2.5597809552783692,
+      "grad_norm": 0.608734130859375,
+      "learning_rate": 1.5422851513531028e-05,
+      "loss": 0.0495,
+      "step": 4207
+    },
+    {
+      "epoch": 2.560389412838455,
+      "grad_norm": 0.4380703270435333,
+      "learning_rate": 1.5411353537278935e-05,
+      "loss": 0.0282,
+      "step": 4208
+    },
+    {
+      "epoch": 2.5609978703985394,
+      "grad_norm": 0.4923262894153595,
+      "learning_rate": 1.539985793861149e-05,
+      "loss": 0.0473,
+      "step": 4209
+    },
+    {
+      "epoch": 2.561606327958625,
+      "grad_norm": 0.45295250415802,
+      "learning_rate": 1.5388364720379124e-05,
+      "loss": 0.041,
+      "step": 4210
+    },
+    {
+      "epoch": 2.56221478551871,
+      "grad_norm": 0.568891704082489,
+      "learning_rate": 1.5376873885431663e-05,
+      "loss": 0.0386,
+      "step": 4211
+    },
+    {
+      "epoch": 2.562823243078795,
+      "grad_norm": 0.4208332598209381,
+      "learning_rate": 1.536538543661838e-05,
+      "loss": 0.0389,
+      "step": 4212
+    },
+    {
+      "epoch": 2.5634317006388807,
+      "grad_norm": 0.4591726064682007,
+      "learning_rate": 1.535389937678792e-05,
+      "loss": 0.0361,
+      "step": 4213
+    },
+    {
+      "epoch": 2.5640401581989654,
+      "grad_norm": 0.48485690355300903,
+      "learning_rate": 1.5342415708788326e-05,
+      "loss": 0.0322,
+      "step": 4214
+    },
+    {
+      "epoch": 2.564648615759051,
+      "grad_norm": 0.546795129776001,
+      "learning_rate": 1.5330934435467104e-05,
+      "loss": 0.0465,
+      "step": 4215
+    },
+    {
+      "epoch": 2.565257073319136,
+      "grad_norm": 0.5020440816879272,
+      "learning_rate": 1.5319455559671116e-05,
+      "loss": 0.0341,
+      "step": 4216
+    },
+    {
+      "epoch": 2.565865530879221,
+      "grad_norm": 0.592922031879425,
+      "learning_rate": 1.530797908424663e-05,
+      "loss": 0.063,
+      "step": 4217
+    },
+    {
+      "epoch": 2.5664739884393066,
+      "grad_norm": 0.5598524808883667,
+      "learning_rate": 1.5296505012039362e-05,
+      "loss": 0.0321,
+      "step": 4218
+    },
+    {
+      "epoch": 2.5670824459993913,
+      "grad_norm": 0.5240581631660461,
+      "learning_rate": 1.5285033345894392e-05,
+      "loss": 0.0521,
+      "step": 4219
+    },
+    {
+      "epoch": 2.567690903559477,
+      "grad_norm": 0.4785360097885132,
+      "learning_rate": 1.5273564088656208e-05,
+      "loss": 0.0447,
+      "step": 4220
+    },
+    {
+      "epoch": 2.568299361119562,
+      "grad_norm": 0.502627432346344,
+      "learning_rate": 1.5262097243168705e-05,
+      "loss": 0.0438,
+      "step": 4221
+    },
+    {
+      "epoch": 2.568907818679647,
+      "grad_norm": 0.5592759251594543,
+      "learning_rate": 1.5250632812275194e-05,
+      "loss": 0.0488,
+      "step": 4222
+    },
+    {
+      "epoch": 2.5695162762397326,
+      "grad_norm": 0.528427004814148,
+      "learning_rate": 1.5239170798818381e-05,
+      "loss": 0.0451,
+      "step": 4223
+    },
+    {
+      "epoch": 2.570124733799817,
+      "grad_norm": 0.5380241870880127,
+      "learning_rate": 1.5227711205640341e-05,
+      "loss": 0.0442,
+      "step": 4224
+    },
+    {
+      "epoch": 2.5707331913599027,
+      "grad_norm": 0.4866541624069214,
+      "learning_rate": 1.5216254035582605e-05,
+      "loss": 0.0364,
+      "step": 4225
+    },
+    {
+      "epoch": 2.571341648919988,
+      "grad_norm": 0.6870383620262146,
+      "learning_rate": 1.5204799291486063e-05,
+      "loss": 0.0662,
+      "step": 4226
+    },
+    {
+      "epoch": 2.571950106480073,
+      "grad_norm": 0.5109632611274719,
+      "learning_rate": 1.5193346976191003e-05,
+      "loss": 0.0363,
+      "step": 4227
+    },
+    {
+      "epoch": 2.572558564040158,
+      "grad_norm": 0.4471718370914459,
+      "learning_rate": 1.518189709253714e-05,
+      "loss": 0.028,
+      "step": 4228
+    },
+    {
+      "epoch": 2.573167021600243,
+      "grad_norm": 0.5469757318496704,
+      "learning_rate": 1.517044964336356e-05,
+      "loss": 0.0474,
+      "step": 4229
+    },
+    {
+      "epoch": 2.5737754791603287,
+      "grad_norm": 0.5387968420982361,
+      "learning_rate": 1.5159004631508739e-05,
+      "loss": 0.0518,
+      "step": 4230
+    },
+    {
+      "epoch": 2.5743839367204138,
+      "grad_norm": 0.4985266327857971,
+      "learning_rate": 1.514756205981059e-05,
+      "loss": 0.0372,
+      "step": 4231
+    },
+    {
+      "epoch": 2.574992394280499,
+      "grad_norm": 0.8584986925125122,
+      "learning_rate": 1.5136121931106378e-05,
+      "loss": 0.0439,
+      "step": 4232
+    },
+    {
+      "epoch": 2.575600851840584,
+      "grad_norm": 0.4893166720867157,
+      "learning_rate": 1.5124684248232784e-05,
+      "loss": 0.0351,
+      "step": 4233
+    },
+    {
+      "epoch": 2.576209309400669,
+      "grad_norm": 0.5629805326461792,
+      "learning_rate": 1.511324901402586e-05,
+      "loss": 0.046,
+      "step": 4234
+    },
+    {
+      "epoch": 2.5768177669607546,
+      "grad_norm": 0.593343198299408,
+      "learning_rate": 1.5101816231321092e-05,
+      "loss": 0.0604,
+      "step": 4235
+    },
+    {
+      "epoch": 2.5774262245208397,
+      "grad_norm": 0.45182177424430847,
+      "learning_rate": 1.5090385902953325e-05,
+      "loss": 0.0377,
+      "step": 4236
+    },
+    {
+      "epoch": 2.578034682080925,
+      "grad_norm": 0.4370990991592407,
+      "learning_rate": 1.5078958031756784e-05,
+      "loss": 0.0393,
+      "step": 4237
+    },
+    {
+      "epoch": 2.57864313964101,
+      "grad_norm": 0.5261332392692566,
+      "learning_rate": 1.506753262056514e-05,
+      "loss": 0.0346,
+      "step": 4238
+    },
+    {
+      "epoch": 2.579251597201095,
+      "grad_norm": 0.5731834173202515,
+      "learning_rate": 1.5056109672211394e-05,
+      "loss": 0.0374,
+      "step": 4239
+    },
+    {
+      "epoch": 2.5798600547611805,
+      "grad_norm": 0.4817114770412445,
+      "learning_rate": 1.5044689189527957e-05,
+      "loss": 0.0364,
+      "step": 4240
+    },
+    {
+      "epoch": 2.5804685123212656,
+      "grad_norm": 0.6388393044471741,
+      "learning_rate": 1.5033271175346658e-05,
+      "loss": 0.0571,
+      "step": 4241
+    },
+    {
+      "epoch": 2.5810769698813507,
+      "grad_norm": 0.4850732088088989,
+      "learning_rate": 1.5021855632498671e-05,
+      "loss": 0.0417,
+      "step": 4242
+    },
+    {
+      "epoch": 2.581685427441436,
+      "grad_norm": 0.5559534430503845,
+      "learning_rate": 1.5010442563814575e-05,
+      "loss": 0.0567,
+      "step": 4243
+    },
+    {
+      "epoch": 2.582293885001521,
+      "grad_norm": 0.4856293201446533,
+      "learning_rate": 1.4999031972124333e-05,
+      "loss": 0.0494,
+      "step": 4244
+    },
+    {
+      "epoch": 2.5829023425616064,
+      "grad_norm": 0.8523834347724915,
+      "learning_rate": 1.498762386025731e-05,
+      "loss": 0.0328,
+      "step": 4245
+    },
+    {
+      "epoch": 2.5835108001216915,
+      "grad_norm": 0.6239566802978516,
+      "learning_rate": 1.4976218231042233e-05,
+      "loss": 0.0505,
+      "step": 4246
+    },
+    {
+      "epoch": 2.5841192576817766,
+      "grad_norm": 0.5066648125648499,
+      "learning_rate": 1.496481508730721e-05,
+      "loss": 0.0503,
+      "step": 4247
+    },
+    {
+      "epoch": 2.5847277152418617,
+      "grad_norm": 0.6433850526809692,
+      "learning_rate": 1.495341443187977e-05,
+      "loss": 0.0472,
+      "step": 4248
+    },
+    {
+      "epoch": 2.585336172801947,
+      "grad_norm": 0.5411747097969055,
+      "learning_rate": 1.4942016267586789e-05,
+      "loss": 0.0525,
+      "step": 4249
+    },
+    {
+      "epoch": 2.5859446303620324,
+      "grad_norm": 0.5121768116950989,
+      "learning_rate": 1.4930620597254524e-05,
+      "loss": 0.0443,
+      "step": 4250
+    },
+    {
+      "epoch": 2.5865530879221175,
+      "grad_norm": 0.40693843364715576,
+      "learning_rate": 1.4919227423708653e-05,
+      "loss": 0.0301,
+      "step": 4251
+    },
+    {
+      "epoch": 2.5871615454822026,
+      "grad_norm": 0.49290603399276733,
+      "learning_rate": 1.490783674977419e-05,
+      "loss": 0.0389,
+      "step": 4252
+    },
+    {
+      "epoch": 2.5877700030422877,
+      "grad_norm": 0.5338709354400635,
+      "learning_rate": 1.489644857827554e-05,
+      "loss": 0.0364,
+      "step": 4253
+    },
+    {
+      "epoch": 2.5883784606023728,
+      "grad_norm": 0.5304934978485107,
+      "learning_rate": 1.4885062912036517e-05,
+      "loss": 0.0427,
+      "step": 4254
+    },
+    {
+      "epoch": 2.5889869181624583,
+      "grad_norm": 0.47712579369544983,
+      "learning_rate": 1.4873679753880284e-05,
+      "loss": 0.0442,
+      "step": 4255
+    },
+    {
+      "epoch": 2.5895953757225434,
+      "grad_norm": 0.5361647605895996,
+      "learning_rate": 1.4862299106629385e-05,
+      "loss": 0.0473,
+      "step": 4256
+    },
+    {
+      "epoch": 2.5902038332826285,
+      "grad_norm": 0.49737563729286194,
+      "learning_rate": 1.4850920973105736e-05,
+      "loss": 0.0344,
+      "step": 4257
+    },
+    {
+      "epoch": 2.5908122908427136,
+      "grad_norm": 0.5141322612762451,
+      "learning_rate": 1.483954535613066e-05,
+      "loss": 0.0588,
+      "step": 4258
+    },
+    {
+      "epoch": 2.5914207484027987,
+      "grad_norm": 0.4289071559906006,
+      "learning_rate": 1.4828172258524828e-05,
+      "loss": 0.0318,
+      "step": 4259
+    },
+    {
+      "epoch": 2.592029205962884,
+      "grad_norm": 0.4381569027900696,
+      "learning_rate": 1.4816801683108283e-05,
+      "loss": 0.0371,
+      "step": 4260
+    },
+    {
+      "epoch": 2.5926376635229693,
+      "grad_norm": 0.6164323091506958,
+      "learning_rate": 1.4805433632700475e-05,
+      "loss": 0.0378,
+      "step": 4261
+    },
+    {
+      "epoch": 2.5932461210830544,
+      "grad_norm": 0.4577103555202484,
+      "learning_rate": 1.4794068110120196e-05,
+      "loss": 0.0393,
+      "step": 4262
+    },
+    {
+      "epoch": 2.5938545786431395,
+      "grad_norm": 0.42026767134666443,
+      "learning_rate": 1.4782705118185608e-05,
+      "loss": 0.0397,
+      "step": 4263
+    },
+    {
+      "epoch": 2.5944630362032246,
+      "grad_norm": 0.4118916690349579,
+      "learning_rate": 1.4771344659714287e-05,
+      "loss": 0.0307,
+      "step": 4264
+    },
+    {
+      "epoch": 2.59507149376331,
+      "grad_norm": 0.4349444508552551,
+      "learning_rate": 1.4759986737523135e-05,
+      "loss": 0.0329,
+      "step": 4265
+    },
+    {
+      "epoch": 2.5956799513233952,
+      "grad_norm": 0.6133854389190674,
+      "learning_rate": 1.4748631354428444e-05,
+      "loss": 0.0446,
+      "step": 4266
+    },
+    {
+      "epoch": 2.5962884088834803,
+      "grad_norm": 0.5277345776557922,
+      "learning_rate": 1.473727851324588e-05,
+      "loss": 0.0432,
+      "step": 4267
+    },
+    {
+      "epoch": 2.5968968664435654,
+      "grad_norm": 0.6331156492233276,
+      "learning_rate": 1.472592821679048e-05,
+      "loss": 0.0598,
+      "step": 4268
+    },
+    {
+      "epoch": 2.5975053240036505,
+      "grad_norm": 0.5062826871871948,
+      "learning_rate": 1.4714580467876633e-05,
+      "loss": 0.0414,
+      "step": 4269
+    },
+    {
+      "epoch": 2.598113781563736,
+      "grad_norm": 0.43930584192276,
+      "learning_rate": 1.4703235269318107e-05,
+      "loss": 0.0425,
+      "step": 4270
+    },
+    {
+      "epoch": 2.598722239123821,
+      "grad_norm": 0.36331331729888916,
+      "learning_rate": 1.4691892623928052e-05,
+      "loss": 0.0169,
+      "step": 4271
+    },
+    {
+      "epoch": 2.5993306966839063,
+      "grad_norm": 0.49878576397895813,
+      "learning_rate": 1.468055253451896e-05,
+      "loss": 0.0363,
+      "step": 4272
+    },
+    {
+      "epoch": 2.5999391542439914,
+      "grad_norm": 0.5917057991027832,
+      "learning_rate": 1.4669215003902704e-05,
+      "loss": 0.063,
+      "step": 4273
+    },
+    {
+      "epoch": 2.6005476118040765,
+      "grad_norm": 0.6224356293678284,
+      "learning_rate": 1.465788003489052e-05,
+      "loss": 0.045,
+      "step": 4274
+    },
+    {
+      "epoch": 2.601156069364162,
+      "grad_norm": 0.6835285425186157,
+      "learning_rate": 1.4646547630293003e-05,
+      "loss": 0.0451,
+      "step": 4275
+    },
+    {
+      "epoch": 2.601764526924247,
+      "grad_norm": 0.5661000609397888,
+      "learning_rate": 1.463521779292012e-05,
+      "loss": 0.0451,
+      "step": 4276
+    },
+    {
+      "epoch": 2.602372984484332,
+      "grad_norm": 0.4806481897830963,
+      "learning_rate": 1.4623890525581204e-05,
+      "loss": 0.0418,
+      "step": 4277
+    },
+    {
+      "epoch": 2.6029814420444173,
+      "grad_norm": 0.5451545119285583,
+      "learning_rate": 1.4612565831084937e-05,
+      "loss": 0.0359,
+      "step": 4278
+    },
+    {
+      "epoch": 2.6035898996045024,
+      "grad_norm": 0.6044538021087646,
+      "learning_rate": 1.4601243712239376e-05,
+      "loss": 0.0408,
+      "step": 4279
+    },
+    {
+      "epoch": 2.604198357164588,
+      "grad_norm": 0.5169386863708496,
+      "learning_rate": 1.4589924171851926e-05,
+      "loss": 0.0313,
+      "step": 4280
+    },
+    {
+      "epoch": 2.604806814724673,
+      "grad_norm": 0.5409014821052551,
+      "learning_rate": 1.4578607212729384e-05,
+      "loss": 0.0353,
+      "step": 4281
+    },
+    {
+      "epoch": 2.605415272284758,
+      "grad_norm": 1.0071747303009033,
+      "learning_rate": 1.4567292837677855e-05,
+      "loss": 0.1323,
+      "step": 4282
+    },
+    {
+      "epoch": 2.606023729844843,
+      "grad_norm": 0.5074636936187744,
+      "learning_rate": 1.4555981049502849e-05,
+      "loss": 0.0353,
+      "step": 4283
+    },
+    {
+      "epoch": 2.6066321874049283,
+      "grad_norm": 0.4935401976108551,
+      "learning_rate": 1.4544671851009212e-05,
+      "loss": 0.0389,
+      "step": 4284
+    },
+    {
+      "epoch": 2.607240644965014,
+      "grad_norm": 0.46502718329429626,
+      "learning_rate": 1.4533365245001168e-05,
+      "loss": 0.0413,
+      "step": 4285
+    },
+    {
+      "epoch": 2.607849102525099,
+      "grad_norm": 0.49090376496315,
+      "learning_rate": 1.452206123428225e-05,
+      "loss": 0.0323,
+      "step": 4286
+    },
+    {
+      "epoch": 2.608457560085184,
+      "grad_norm": 0.3921135365962982,
+      "learning_rate": 1.4510759821655423e-05,
+      "loss": 0.0423,
+      "step": 4287
+    },
+    {
+      "epoch": 2.609066017645269,
+      "grad_norm": 0.6183372139930725,
+      "learning_rate": 1.449946100992294e-05,
+      "loss": 0.0389,
+      "step": 4288
+    },
+    {
+      "epoch": 2.6096744752053542,
+      "grad_norm": 0.4686357080936432,
+      "learning_rate": 1.4488164801886453e-05,
+      "loss": 0.0434,
+      "step": 4289
+    },
+    {
+      "epoch": 2.6102829327654398,
+      "grad_norm": 0.49797457456588745,
+      "learning_rate": 1.447687120034692e-05,
+      "loss": 0.0394,
+      "step": 4290
+    },
+    {
+      "epoch": 2.610891390325525,
+      "grad_norm": 0.5921012163162231,
+      "learning_rate": 1.4465580208104722e-05,
+      "loss": 0.0566,
+      "step": 4291
+    },
+    {
+      "epoch": 2.61149984788561,
+      "grad_norm": 0.5064262747764587,
+      "learning_rate": 1.4454291827959526e-05,
+      "loss": 0.0424,
+      "step": 4292
+    },
+    {
+      "epoch": 2.612108305445695,
+      "grad_norm": 0.45030635595321655,
+      "learning_rate": 1.4443006062710391e-05,
+      "loss": 0.0361,
+      "step": 4293
+    },
+    {
+      "epoch": 2.61271676300578,
+      "grad_norm": 0.8079550266265869,
+      "learning_rate": 1.4431722915155716e-05,
+      "loss": 0.045,
+      "step": 4294
+    },
+    {
+      "epoch": 2.6133252205658657,
+      "grad_norm": 0.5597800612449646,
+      "learning_rate": 1.4420442388093258e-05,
+      "loss": 0.0415,
+      "step": 4295
+    },
+    {
+      "epoch": 2.613933678125951,
+      "grad_norm": 0.5742923617362976,
+      "learning_rate": 1.4409164484320092e-05,
+      "loss": 0.059,
+      "step": 4296
+    },
+    {
+      "epoch": 2.614542135686036,
+      "grad_norm": 0.4569978713989258,
+      "learning_rate": 1.4397889206632706e-05,
+      "loss": 0.0445,
+      "step": 4297
+    },
+    {
+      "epoch": 2.615150593246121,
+      "grad_norm": 0.4496994614601135,
+      "learning_rate": 1.4386616557826869e-05,
+      "loss": 0.0386,
+      "step": 4298
+    },
+    {
+      "epoch": 2.615759050806206,
+      "grad_norm": 0.531541645526886,
+      "learning_rate": 1.4375346540697738e-05,
+      "loss": 0.0294,
+      "step": 4299
+    },
+    {
+      "epoch": 2.6163675083662916,
+      "grad_norm": 0.4035661816596985,
+      "learning_rate": 1.4364079158039807e-05,
+      "loss": 0.0337,
+      "step": 4300
+    },
+    {
+      "epoch": 2.6169759659263767,
+      "grad_norm": 0.42909929156303406,
+      "learning_rate": 1.4352814412646926e-05,
+      "loss": 0.0295,
+      "step": 4301
+    },
+    {
+      "epoch": 2.617584423486462,
+      "grad_norm": 0.6985564231872559,
+      "learning_rate": 1.4341552307312266e-05,
+      "loss": 0.0477,
+      "step": 4302
+    },
+    {
+      "epoch": 2.618192881046547,
+      "grad_norm": 0.5794458985328674,
+      "learning_rate": 1.4330292844828368e-05,
+      "loss": 0.0554,
+      "step": 4303
+    },
+    {
+      "epoch": 2.618801338606632,
+      "grad_norm": 0.6460904479026794,
+      "learning_rate": 1.431903602798711e-05,
+      "loss": 0.0393,
+      "step": 4304
+    },
+    {
+      "epoch": 2.6194097961667175,
+      "grad_norm": 0.5019935965538025,
+      "learning_rate": 1.4307781859579721e-05,
+      "loss": 0.0309,
+      "step": 4305
+    },
+    {
+      "epoch": 2.6200182537268026,
+      "grad_norm": 0.5265465974807739,
+      "learning_rate": 1.4296530342396741e-05,
+      "loss": 0.0407,
+      "step": 4306
+    },
+    {
+      "epoch": 2.6206267112868877,
+      "grad_norm": 0.6341969966888428,
+      "learning_rate": 1.4285281479228107e-05,
+      "loss": 0.0558,
+      "step": 4307
+    },
+    {
+      "epoch": 2.621235168846973,
+      "grad_norm": 0.5350863337516785,
+      "learning_rate": 1.4274035272863051e-05,
+      "loss": 0.0376,
+      "step": 4308
+    },
+    {
+      "epoch": 2.621843626407058,
+      "grad_norm": 0.5693925023078918,
+      "learning_rate": 1.4262791726090163e-05,
+      "loss": 0.0418,
+      "step": 4309
+    },
+    {
+      "epoch": 2.6224520839671435,
+      "grad_norm": 0.4471037983894348,
+      "learning_rate": 1.425155084169738e-05,
+      "loss": 0.0379,
+      "step": 4310
+    },
+    {
+      "epoch": 2.6230605415272286,
+      "grad_norm": 0.428018718957901,
+      "learning_rate": 1.424031262247198e-05,
+      "loss": 0.0334,
+      "step": 4311
+    },
+    {
+      "epoch": 2.6236689990873137,
+      "grad_norm": 0.5358444452285767,
+      "learning_rate": 1.4229077071200542e-05,
+      "loss": 0.0471,
+      "step": 4312
+    },
+    {
+      "epoch": 2.6242774566473988,
+      "grad_norm": 0.515167772769928,
+      "learning_rate": 1.4217844190669058e-05,
+      "loss": 0.0456,
+      "step": 4313
+    },
+    {
+      "epoch": 2.624885914207484,
+      "grad_norm": 0.5426524877548218,
+      "learning_rate": 1.4206613983662781e-05,
+      "loss": 0.0412,
+      "step": 4314
+    },
+    {
+      "epoch": 2.6254943717675694,
+      "grad_norm": 0.6369853019714355,
+      "learning_rate": 1.4195386452966359e-05,
+      "loss": 0.0569,
+      "step": 4315
+    },
+    {
+      "epoch": 2.6261028293276545,
+      "grad_norm": 0.5484540462493896,
+      "learning_rate": 1.4184161601363716e-05,
+      "loss": 0.0377,
+      "step": 4316
+    },
+    {
+      "epoch": 2.6267112868877396,
+      "grad_norm": 0.4859999418258667,
+      "learning_rate": 1.4172939431638188e-05,
+      "loss": 0.0353,
+      "step": 4317
+    },
+    {
+      "epoch": 2.6273197444478247,
+      "grad_norm": 0.5194180011749268,
+      "learning_rate": 1.4161719946572377e-05,
+      "loss": 0.0401,
+      "step": 4318
+    },
+    {
+      "epoch": 2.6279282020079098,
+      "grad_norm": 0.6942944526672363,
+      "learning_rate": 1.415050314894826e-05,
+      "loss": 0.0614,
+      "step": 4319
+    },
+    {
+      "epoch": 2.6285366595679953,
+      "grad_norm": 0.4476464092731476,
+      "learning_rate": 1.4139289041547132e-05,
+      "loss": 0.0307,
+      "step": 4320
+    },
+    {
+      "epoch": 2.6291451171280804,
+      "grad_norm": 0.5009614825248718,
+      "learning_rate": 1.4128077627149633e-05,
+      "loss": 0.0404,
+      "step": 4321
+    },
+    {
+      "epoch": 2.6297535746881655,
+      "grad_norm": 0.47980043292045593,
+      "learning_rate": 1.4116868908535702e-05,
+      "loss": 0.0377,
+      "step": 4322
+    },
+    {
+      "epoch": 2.6303620322482506,
+      "grad_norm": 0.5018118619918823,
+      "learning_rate": 1.4105662888484667e-05,
+      "loss": 0.0424,
+      "step": 4323
+    },
+    {
+      "epoch": 2.6309704898083357,
+      "grad_norm": 0.602103590965271,
+      "learning_rate": 1.4094459569775128e-05,
+      "loss": 0.0571,
+      "step": 4324
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.5462225675582886,
+      "learning_rate": 1.4083258955185053e-05,
+      "loss": 0.059,
+      "step": 4325
+    },
+    {
+      "epoch": 2.6321874049285063,
+      "grad_norm": 0.4753000736236572,
+      "learning_rate": 1.4072061047491721e-05,
+      "loss": 0.0439,
+      "step": 4326
+    },
+    {
+      "epoch": 2.6327958624885914,
+      "grad_norm": 0.45958274602890015,
+      "learning_rate": 1.4060865849471764e-05,
+      "loss": 0.0353,
+      "step": 4327
+    },
+    {
+      "epoch": 2.6334043200486765,
+      "grad_norm": 0.5752948522567749,
+      "learning_rate": 1.4049673363901097e-05,
+      "loss": 0.0462,
+      "step": 4328
+    },
+    {
+      "epoch": 2.6340127776087616,
+      "grad_norm": 0.4886283874511719,
+      "learning_rate": 1.4038483593555007e-05,
+      "loss": 0.0483,
+      "step": 4329
+    },
+    {
+      "epoch": 2.634621235168847,
+      "grad_norm": 0.4635407030582428,
+      "learning_rate": 1.4027296541208084e-05,
+      "loss": 0.038,
+      "step": 4330
+    },
+    {
+      "epoch": 2.6352296927289323,
+      "grad_norm": 0.4888119399547577,
+      "learning_rate": 1.4016112209634258e-05,
+      "loss": 0.0394,
+      "step": 4331
+    },
+    {
+      "epoch": 2.6358381502890174,
+      "grad_norm": 0.52603679895401,
+      "learning_rate": 1.400493060160677e-05,
+      "loss": 0.0452,
+      "step": 4332
+    },
+    {
+      "epoch": 2.6364466078491025,
+      "grad_norm": 0.5692835450172424,
+      "learning_rate": 1.3993751719898207e-05,
+      "loss": 0.0364,
+      "step": 4333
+    },
+    {
+      "epoch": 2.6370550654091875,
+      "grad_norm": 0.5845534205436707,
+      "learning_rate": 1.3982575567280442e-05,
+      "loss": 0.0503,
+      "step": 4334
+    },
+    {
+      "epoch": 2.637663522969273,
+      "grad_norm": 0.4657052755355835,
+      "learning_rate": 1.397140214652471e-05,
+      "loss": 0.0328,
+      "step": 4335
+    },
+    {
+      "epoch": 2.638271980529358,
+      "grad_norm": 0.4424096941947937,
+      "learning_rate": 1.3960231460401552e-05,
+      "loss": 0.0287,
+      "step": 4336
+    },
+    {
+      "epoch": 2.6388804380894433,
+      "grad_norm": 0.4316795766353607,
+      "learning_rate": 1.3949063511680837e-05,
+      "loss": 0.0282,
+      "step": 4337
+    },
+    {
+      "epoch": 2.6394888956495284,
+      "grad_norm": 0.5137184262275696,
+      "learning_rate": 1.3937898303131742e-05,
+      "loss": 0.0395,
+      "step": 4338
+    },
+    {
+      "epoch": 2.6400973532096135,
+      "grad_norm": 0.5804075002670288,
+      "learning_rate": 1.392673583752277e-05,
+      "loss": 0.0472,
+      "step": 4339
+    },
+    {
+      "epoch": 2.640705810769699,
+      "grad_norm": 0.43918368220329285,
+      "learning_rate": 1.3915576117621758e-05,
+      "loss": 0.0335,
+      "step": 4340
+    },
+    {
+      "epoch": 2.641314268329784,
+      "grad_norm": 0.4672797918319702,
+      "learning_rate": 1.3904419146195846e-05,
+      "loss": 0.0429,
+      "step": 4341
+    },
+    {
+      "epoch": 2.641922725889869,
+      "grad_norm": 0.515037477016449,
+      "learning_rate": 1.3893264926011502e-05,
+      "loss": 0.0467,
+      "step": 4342
+    },
+    {
+      "epoch": 2.6425311834499543,
+      "grad_norm": 0.5160783529281616,
+      "learning_rate": 1.3882113459834512e-05,
+      "loss": 0.0379,
+      "step": 4343
+    },
+    {
+      "epoch": 2.6431396410100394,
+      "grad_norm": 0.5049687623977661,
+      "learning_rate": 1.3870964750429954e-05,
+      "loss": 0.0434,
+      "step": 4344
+    },
+    {
+      "epoch": 2.643748098570125,
+      "grad_norm": 0.5025326013565063,
+      "learning_rate": 1.3859818800562263e-05,
+      "loss": 0.051,
+      "step": 4345
+    },
+    {
+      "epoch": 2.64435655613021,
+      "grad_norm": 0.5900479555130005,
+      "learning_rate": 1.384867561299516e-05,
+      "loss": 0.0468,
+      "step": 4346
+    },
+    {
+      "epoch": 2.644965013690295,
+      "grad_norm": 0.48302316665649414,
+      "learning_rate": 1.3837535190491696e-05,
+      "loss": 0.041,
+      "step": 4347
+    },
+    {
+      "epoch": 2.6455734712503802,
+      "grad_norm": 0.5139932632446289,
+      "learning_rate": 1.3826397535814242e-05,
+      "loss": 0.0532,
+      "step": 4348
+    },
+    {
+      "epoch": 2.6461819288104653,
+      "grad_norm": 0.40311169624328613,
+      "learning_rate": 1.3815262651724448e-05,
+      "loss": 0.0332,
+      "step": 4349
+    },
+    {
+      "epoch": 2.646790386370551,
+      "grad_norm": 0.512266218662262,
+      "learning_rate": 1.3804130540983317e-05,
+      "loss": 0.0447,
+      "step": 4350
+    },
+    {
+      "epoch": 2.647398843930636,
+      "grad_norm": 0.48747456073760986,
+      "learning_rate": 1.3793001206351142e-05,
+      "loss": 0.0312,
+      "step": 4351
+    },
+    {
+      "epoch": 2.648007301490721,
+      "grad_norm": 0.3904288709163666,
+      "learning_rate": 1.3781874650587536e-05,
+      "loss": 0.0294,
+      "step": 4352
+    },
+    {
+      "epoch": 2.648615759050806,
+      "grad_norm": 0.6491031050682068,
+      "learning_rate": 1.3770750876451427e-05,
+      "loss": 0.046,
+      "step": 4353
+    },
+    {
+      "epoch": 2.6492242166108912,
+      "grad_norm": 0.4241449534893036,
+      "learning_rate": 1.3759629886701047e-05,
+      "loss": 0.0422,
+      "step": 4354
+    },
+    {
+      "epoch": 2.649832674170977,
+      "grad_norm": 0.48629888892173767,
+      "learning_rate": 1.3748511684093926e-05,
+      "loss": 0.0308,
+      "step": 4355
+    },
+    {
+      "epoch": 2.650441131731062,
+      "grad_norm": 0.5005238056182861,
+      "learning_rate": 1.3737396271386921e-05,
+      "loss": 0.0566,
+      "step": 4356
+    },
+    {
+      "epoch": 2.651049589291147,
+      "grad_norm": 0.6130560040473938,
+      "learning_rate": 1.3726283651336194e-05,
+      "loss": 0.064,
+      "step": 4357
+    },
+    {
+      "epoch": 2.651658046851232,
+      "grad_norm": 0.578168511390686,
+      "learning_rate": 1.3715173826697209e-05,
+      "loss": 0.0479,
+      "step": 4358
+    },
+    {
+      "epoch": 2.652266504411317,
+      "grad_norm": 0.5515162944793701,
+      "learning_rate": 1.3704066800224741e-05,
+      "loss": 0.0452,
+      "step": 4359
+    },
+    {
+      "epoch": 2.6528749619714027,
+      "grad_norm": 0.5804712772369385,
+      "learning_rate": 1.369296257467288e-05,
+      "loss": 0.0467,
+      "step": 4360
+    },
+    {
+      "epoch": 2.653483419531488,
+      "grad_norm": 0.45583152770996094,
+      "learning_rate": 1.3681861152794984e-05,
+      "loss": 0.0318,
+      "step": 4361
+    },
+    {
+      "epoch": 2.654091877091573,
+      "grad_norm": 0.8064979910850525,
+      "learning_rate": 1.3670762537343765e-05,
+      "loss": 0.0576,
+      "step": 4362
+    },
+    {
+      "epoch": 2.654700334651658,
+      "grad_norm": 0.48464739322662354,
+      "learning_rate": 1.3659666731071207e-05,
+      "loss": 0.0436,
+      "step": 4363
+    },
+    {
+      "epoch": 2.655308792211743,
+      "grad_norm": 0.48261159658432007,
+      "learning_rate": 1.3648573736728627e-05,
+      "loss": 0.03,
+      "step": 4364
+    },
+    {
+      "epoch": 2.6559172497718286,
+      "grad_norm": 0.5928403735160828,
+      "learning_rate": 1.3637483557066583e-05,
+      "loss": 0.0423,
+      "step": 4365
+    },
+    {
+      "epoch": 2.6565257073319137,
+      "grad_norm": 0.521212637424469,
+      "learning_rate": 1.3626396194835026e-05,
+      "loss": 0.0518,
+      "step": 4366
+    },
+    {
+      "epoch": 2.657134164891999,
+      "grad_norm": 0.47702714800834656,
+      "learning_rate": 1.3615311652783127e-05,
+      "loss": 0.0327,
+      "step": 4367
+    },
+    {
+      "epoch": 2.657742622452084,
+      "grad_norm": 0.5482526421546936,
+      "learning_rate": 1.3604229933659402e-05,
+      "loss": 0.0494,
+      "step": 4368
+    },
+    {
+      "epoch": 2.658351080012169,
+      "grad_norm": 0.45253828167915344,
+      "learning_rate": 1.3593151040211654e-05,
+      "loss": 0.0375,
+      "step": 4369
+    },
+    {
+      "epoch": 2.6589595375722546,
+      "grad_norm": 0.5163220167160034,
+      "learning_rate": 1.3582074975186998e-05,
+      "loss": 0.034,
+      "step": 4370
+    },
+    {
+      "epoch": 2.6595679951323397,
+      "grad_norm": 0.4799509346485138,
+      "learning_rate": 1.3571001741331815e-05,
+      "loss": 0.035,
+      "step": 4371
+    },
+    {
+      "epoch": 2.6601764526924248,
+      "grad_norm": 3.0906639099121094,
+      "learning_rate": 1.3559931341391815e-05,
+      "loss": 0.0468,
+      "step": 4372
+    },
+    {
+      "epoch": 2.66078491025251,
+      "grad_norm": 0.5727887153625488,
+      "learning_rate": 1.3548863778111998e-05,
+      "loss": 0.0482,
+      "step": 4373
+    },
+    {
+      "epoch": 2.661393367812595,
+      "grad_norm": 0.5182603597640991,
+      "learning_rate": 1.3537799054236666e-05,
+      "loss": 0.0403,
+      "step": 4374
+    },
+    {
+      "epoch": 2.6620018253726805,
+      "grad_norm": 0.4546807110309601,
+      "learning_rate": 1.3526737172509383e-05,
+      "loss": 0.0396,
+      "step": 4375
+    },
+    {
+      "epoch": 2.6626102829327656,
+      "grad_norm": 0.5297713279724121,
+      "learning_rate": 1.3515678135673072e-05,
+      "loss": 0.0366,
+      "step": 4376
+    },
+    {
+      "epoch": 2.6632187404928507,
+      "grad_norm": 0.4167693853378296,
+      "learning_rate": 1.350462194646988e-05,
+      "loss": 0.0333,
+      "step": 4377
+    },
+    {
+      "epoch": 2.6638271980529358,
+      "grad_norm": 0.4427233040332794,
+      "learning_rate": 1.3493568607641294e-05,
+      "loss": 0.0403,
+      "step": 4378
+    },
+    {
+      "epoch": 2.664435655613021,
+      "grad_norm": 0.4933980405330658,
+      "learning_rate": 1.3482518121928083e-05,
+      "loss": 0.0355,
+      "step": 4379
+    },
+    {
+      "epoch": 2.6650441131731064,
+      "grad_norm": 0.6081718802452087,
+      "learning_rate": 1.3471470492070315e-05,
+      "loss": 0.0516,
+      "step": 4380
+    },
+    {
+      "epoch": 2.6656525707331915,
+      "grad_norm": 0.5365004539489746,
+      "learning_rate": 1.3460425720807316e-05,
+      "loss": 0.0451,
+      "step": 4381
+    },
+    {
+      "epoch": 2.6662610282932766,
+      "grad_norm": 0.4933543801307678,
+      "learning_rate": 1.3449383810877761e-05,
+      "loss": 0.0492,
+      "step": 4382
+    },
+    {
+      "epoch": 2.6668694858533617,
+      "grad_norm": 0.48126405477523804,
+      "learning_rate": 1.3438344765019558e-05,
+      "loss": 0.0424,
+      "step": 4383
+    },
+    {
+      "epoch": 2.667477943413447,
+      "grad_norm": 0.5277286171913147,
+      "learning_rate": 1.342730858596995e-05,
+      "loss": 0.0428,
+      "step": 4384
+    },
+    {
+      "epoch": 2.6680864009735323,
+      "grad_norm": 0.45332902669906616,
+      "learning_rate": 1.341627527646542e-05,
+      "loss": 0.037,
+      "step": 4385
+    },
+    {
+      "epoch": 2.6686948585336174,
+      "grad_norm": 0.4423314034938812,
+      "learning_rate": 1.340524483924181e-05,
+      "loss": 0.0283,
+      "step": 4386
+    },
+    {
+      "epoch": 2.6693033160937025,
+      "grad_norm": 0.6126822829246521,
+      "learning_rate": 1.3394217277034171e-05,
+      "loss": 0.0286,
+      "step": 4387
+    },
+    {
+      "epoch": 2.6699117736537876,
+      "grad_norm": 0.5217598080635071,
+      "learning_rate": 1.3383192592576898e-05,
+      "loss": 0.0424,
+      "step": 4388
+    },
+    {
+      "epoch": 2.6705202312138727,
+      "grad_norm": 0.5640128254890442,
+      "learning_rate": 1.3372170788603649e-05,
+      "loss": 0.0493,
+      "step": 4389
+    },
+    {
+      "epoch": 2.6711286887739583,
+      "grad_norm": 0.504953145980835,
+      "learning_rate": 1.3361151867847382e-05,
+      "loss": 0.0329,
+      "step": 4390
+    },
+    {
+      "epoch": 2.6717371463340434,
+      "grad_norm": 0.4337262511253357,
+      "learning_rate": 1.3350135833040305e-05,
+      "loss": 0.0376,
+      "step": 4391
+    },
+    {
+      "epoch": 2.6723456038941285,
+      "grad_norm": 0.5187790989875793,
+      "learning_rate": 1.3339122686913968e-05,
+      "loss": 0.0401,
+      "step": 4392
+    },
+    {
+      "epoch": 2.6729540614542135,
+      "grad_norm": 0.48012837767601013,
+      "learning_rate": 1.3328112432199144e-05,
+      "loss": 0.039,
+      "step": 4393
+    },
+    {
+      "epoch": 2.6735625190142986,
+      "grad_norm": 0.5445698499679565,
+      "learning_rate": 1.331710507162594e-05,
+      "loss": 0.062,
+      "step": 4394
+    },
+    {
+      "epoch": 2.674170976574384,
+      "grad_norm": 0.45752736926078796,
+      "learning_rate": 1.3306100607923687e-05,
+      "loss": 0.04,
+      "step": 4395
+    },
+    {
+      "epoch": 2.6747794341344693,
+      "grad_norm": 0.5074790120124817,
+      "learning_rate": 1.3295099043821085e-05,
+      "loss": 0.0531,
+      "step": 4396
+    },
+    {
+      "epoch": 2.6753878916945544,
+      "grad_norm": 0.46912896633148193,
+      "learning_rate": 1.3284100382046022e-05,
+      "loss": 0.0366,
+      "step": 4397
+    },
+    {
+      "epoch": 2.6759963492546395,
+      "grad_norm": 0.4760879576206207,
+      "learning_rate": 1.3273104625325722e-05,
+      "loss": 0.0374,
+      "step": 4398
+    },
+    {
+      "epoch": 2.6766048068147246,
+      "grad_norm": 0.6422132253646851,
+      "learning_rate": 1.326211177638667e-05,
+      "loss": 0.0607,
+      "step": 4399
+    },
+    {
+      "epoch": 2.67721326437481,
+      "grad_norm": 0.47899577021598816,
+      "learning_rate": 1.3251121837954655e-05,
+      "loss": 0.0411,
+      "step": 4400
+    },
+    {
+      "epoch": 2.677821721934895,
+      "grad_norm": 0.6358346939086914,
+      "learning_rate": 1.3240134812754681e-05,
+      "loss": 0.0489,
+      "step": 4401
+    },
+    {
+      "epoch": 2.6784301794949803,
+      "grad_norm": 0.5330213308334351,
+      "learning_rate": 1.3229150703511122e-05,
+      "loss": 0.0398,
+      "step": 4402
+    },
+    {
+      "epoch": 2.6790386370550654,
+      "grad_norm": 0.4606907367706299,
+      "learning_rate": 1.3218169512947542e-05,
+      "loss": 0.0269,
+      "step": 4403
+    },
+    {
+      "epoch": 2.6796470946151505,
+      "grad_norm": 0.5010808110237122,
+      "learning_rate": 1.3207191243786834e-05,
+      "loss": 0.0487,
+      "step": 4404
+    },
+    {
+      "epoch": 2.680255552175236,
+      "grad_norm": 0.4977697432041168,
+      "learning_rate": 1.319621589875115e-05,
+      "loss": 0.0378,
+      "step": 4405
+    },
+    {
+      "epoch": 2.6808640097353207,
+      "grad_norm": 0.41439104080200195,
+      "learning_rate": 1.3185243480561926e-05,
+      "loss": 0.0436,
+      "step": 4406
+    },
+    {
+      "epoch": 2.6814724672954062,
+      "grad_norm": 0.35358157753944397,
+      "learning_rate": 1.3174273991939845e-05,
+      "loss": 0.0269,
+      "step": 4407
+    },
+    {
+      "epoch": 2.6820809248554913,
+      "grad_norm": 0.4562048017978668,
+      "learning_rate": 1.3163307435604893e-05,
+      "loss": 0.0383,
+      "step": 4408
+    },
+    {
+      "epoch": 2.6826893824155764,
+      "grad_norm": 0.5168052911758423,
+      "learning_rate": 1.3152343814276318e-05,
+      "loss": 0.046,
+      "step": 4409
+    },
+    {
+      "epoch": 2.683297839975662,
+      "grad_norm": 0.5310634970664978,
+      "learning_rate": 1.3141383130672658e-05,
+      "loss": 0.0405,
+      "step": 4410
+    },
+    {
+      "epoch": 2.6839062975357466,
+      "grad_norm": 0.5165497660636902,
+      "learning_rate": 1.3130425387511667e-05,
+      "loss": 0.0516,
+      "step": 4411
+    },
+    {
+      "epoch": 2.684514755095832,
+      "grad_norm": 0.4916205406188965,
+      "learning_rate": 1.3119470587510451e-05,
+      "loss": 0.0372,
+      "step": 4412
+    },
+    {
+      "epoch": 2.6851232126559172,
+      "grad_norm": 0.40354472398757935,
+      "learning_rate": 1.3108518733385314e-05,
+      "loss": 0.0213,
+      "step": 4413
+    },
+    {
+      "epoch": 2.6857316702160023,
+      "grad_norm": 0.4055176079273224,
+      "learning_rate": 1.309756982785187e-05,
+      "loss": 0.0283,
+      "step": 4414
+    },
+    {
+      "epoch": 2.686340127776088,
+      "grad_norm": 0.5336512923240662,
+      "learning_rate": 1.3086623873624992e-05,
+      "loss": 0.041,
+      "step": 4415
+    },
+    {
+      "epoch": 2.6869485853361725,
+      "grad_norm": 0.4223940670490265,
+      "learning_rate": 1.3075680873418828e-05,
+      "loss": 0.0254,
+      "step": 4416
+    },
+    {
+      "epoch": 2.687557042896258,
+      "grad_norm": 0.6376833319664001,
+      "learning_rate": 1.306474082994677e-05,
+      "loss": 0.0409,
+      "step": 4417
+    },
+    {
+      "epoch": 2.688165500456343,
+      "grad_norm": 0.4131200611591339,
+      "learning_rate": 1.3053803745921498e-05,
+      "loss": 0.0309,
+      "step": 4418
+    },
+    {
+      "epoch": 2.6887739580164283,
+      "grad_norm": 0.5285510420799255,
+      "learning_rate": 1.3042869624054955e-05,
+      "loss": 0.0427,
+      "step": 4419
+    },
+    {
+      "epoch": 2.689382415576514,
+      "grad_norm": 0.4723327159881592,
+      "learning_rate": 1.3031938467058358e-05,
+      "loss": 0.0292,
+      "step": 4420
+    },
+    {
+      "epoch": 2.6899908731365985,
+      "grad_norm": 0.5857917070388794,
+      "learning_rate": 1.3021010277642145e-05,
+      "loss": 0.0493,
+      "step": 4421
+    },
+    {
+      "epoch": 2.690599330696684,
+      "grad_norm": 0.5390817523002625,
+      "learning_rate": 1.3010085058516097e-05,
+      "loss": 0.043,
+      "step": 4422
+    },
+    {
+      "epoch": 2.691207788256769,
+      "grad_norm": 0.45317864418029785,
+      "learning_rate": 1.2999162812389181e-05,
+      "loss": 0.029,
+      "step": 4423
+    },
+    {
+      "epoch": 2.691816245816854,
+      "grad_norm": 0.4660782217979431,
+      "learning_rate": 1.2988243541969667e-05,
+      "loss": 0.0258,
+      "step": 4424
+    },
+    {
+      "epoch": 2.6924247033769397,
+      "grad_norm": 0.5454357266426086,
+      "learning_rate": 1.297732724996508e-05,
+      "loss": 0.0432,
+      "step": 4425
+    },
+    {
+      "epoch": 2.6930331609370244,
+      "grad_norm": 0.4717608094215393,
+      "learning_rate": 1.2966413939082214e-05,
+      "loss": 0.0356,
+      "step": 4426
+    },
+    {
+      "epoch": 2.69364161849711,
+      "grad_norm": 0.43902650475502014,
+      "learning_rate": 1.2955503612027086e-05,
+      "loss": 0.0321,
+      "step": 4427
+    },
+    {
+      "epoch": 2.694250076057195,
+      "grad_norm": 0.5439276695251465,
+      "learning_rate": 1.2944596271505046e-05,
+      "loss": 0.0477,
+      "step": 4428
+    },
+    {
+      "epoch": 2.69485853361728,
+      "grad_norm": 0.5469129085540771,
+      "learning_rate": 1.2933691920220626e-05,
+      "loss": 0.0429,
+      "step": 4429
+    },
+    {
+      "epoch": 2.695466991177365,
+      "grad_norm": 0.6971737742424011,
+      "learning_rate": 1.2922790560877667e-05,
+      "loss": 0.067,
+      "step": 4430
+    },
+    {
+      "epoch": 2.6960754487374503,
+      "grad_norm": 0.5088085532188416,
+      "learning_rate": 1.2911892196179231e-05,
+      "loss": 0.0499,
+      "step": 4431
+    },
+    {
+      "epoch": 2.696683906297536,
+      "grad_norm": 0.502051591873169,
+      "learning_rate": 1.2900996828827693e-05,
+      "loss": 0.0361,
+      "step": 4432
+    },
+    {
+      "epoch": 2.697292363857621,
+      "grad_norm": 0.4716339707374573,
+      "learning_rate": 1.2890104461524619e-05,
+      "loss": 0.032,
+      "step": 4433
+    },
+    {
+      "epoch": 2.697900821417706,
+      "grad_norm": 0.4593643546104431,
+      "learning_rate": 1.2879215096970871e-05,
+      "loss": 0.0321,
+      "step": 4434
+    },
+    {
+      "epoch": 2.698509278977791,
+      "grad_norm": 0.42600390315055847,
+      "learning_rate": 1.2868328737866563e-05,
+      "loss": 0.049,
+      "step": 4435
+    },
+    {
+      "epoch": 2.6991177365378762,
+      "grad_norm": 0.5230850577354431,
+      "learning_rate": 1.2857445386911064e-05,
+      "loss": 0.0362,
+      "step": 4436
+    },
+    {
+      "epoch": 2.6997261940979618,
+      "grad_norm": 0.5436407923698425,
+      "learning_rate": 1.284656504680296e-05,
+      "loss": 0.0493,
+      "step": 4437
+    },
+    {
+      "epoch": 2.700334651658047,
+      "grad_norm": 0.5182242393493652,
+      "learning_rate": 1.283568772024017e-05,
+      "loss": 0.0354,
+      "step": 4438
+    },
+    {
+      "epoch": 2.700943109218132,
+      "grad_norm": 0.4881172776222229,
+      "learning_rate": 1.2824813409919777e-05,
+      "loss": 0.0473,
+      "step": 4439
+    },
+    {
+      "epoch": 2.701551566778217,
+      "grad_norm": 0.5372771620750427,
+      "learning_rate": 1.2813942118538181e-05,
+      "loss": 0.0472,
+      "step": 4440
+    },
+    {
+      "epoch": 2.702160024338302,
+      "grad_norm": 0.31403908133506775,
+      "learning_rate": 1.2803073848790983e-05,
+      "loss": 0.0213,
+      "step": 4441
+    },
+    {
+      "epoch": 2.7027684818983877,
+      "grad_norm": 0.5783568024635315,
+      "learning_rate": 1.2792208603373096e-05,
+      "loss": 0.046,
+      "step": 4442
+    },
+    {
+      "epoch": 2.703376939458473,
+      "grad_norm": 0.48002150654792786,
+      "learning_rate": 1.278134638497862e-05,
+      "loss": 0.0315,
+      "step": 4443
+    },
+    {
+      "epoch": 2.703985397018558,
+      "grad_norm": 0.4475983679294586,
+      "learning_rate": 1.277048719630094e-05,
+      "loss": 0.0338,
+      "step": 4444
+    },
+    {
+      "epoch": 2.704593854578643,
+      "grad_norm": 0.4856289029121399,
+      "learning_rate": 1.2759631040032688e-05,
+      "loss": 0.0384,
+      "step": 4445
+    },
+    {
+      "epoch": 2.705202312138728,
+      "grad_norm": 0.34708070755004883,
+      "learning_rate": 1.274877791886574e-05,
+      "loss": 0.0199,
+      "step": 4446
+    },
+    {
+      "epoch": 2.7058107696988136,
+      "grad_norm": 0.4791733920574188,
+      "learning_rate": 1.2737927835491196e-05,
+      "loss": 0.0384,
+      "step": 4447
+    },
+    {
+      "epoch": 2.7064192272588987,
+      "grad_norm": 0.4502567648887634,
+      "learning_rate": 1.2727080792599455e-05,
+      "loss": 0.0389,
+      "step": 4448
+    },
+    {
+      "epoch": 2.707027684818984,
+      "grad_norm": 0.583976149559021,
+      "learning_rate": 1.2716236792880112e-05,
+      "loss": 0.0401,
+      "step": 4449
+    },
+    {
+      "epoch": 2.707636142379069,
+      "grad_norm": 0.5464617013931274,
+      "learning_rate": 1.270539583902203e-05,
+      "loss": 0.0389,
+      "step": 4450
+    },
+    {
+      "epoch": 2.708244599939154,
+      "grad_norm": 0.4903790354728699,
+      "learning_rate": 1.2694557933713316e-05,
+      "loss": 0.0391,
+      "step": 4451
+    },
+    {
+      "epoch": 2.7088530574992395,
+      "grad_norm": 0.42102527618408203,
+      "learning_rate": 1.2683723079641329e-05,
+      "loss": 0.0348,
+      "step": 4452
+    },
+    {
+      "epoch": 2.7094615150593246,
+      "grad_norm": 0.44486379623413086,
+      "learning_rate": 1.2672891279492638e-05,
+      "loss": 0.0285,
+      "step": 4453
+    },
+    {
+      "epoch": 2.7100699726194097,
+      "grad_norm": 0.5711926221847534,
+      "learning_rate": 1.2662062535953095e-05,
+      "loss": 0.0396,
+      "step": 4454
+    },
+    {
+      "epoch": 2.710678430179495,
+      "grad_norm": 0.5700972676277161,
+      "learning_rate": 1.2651236851707768e-05,
+      "loss": 0.0475,
+      "step": 4455
+    },
+    {
+      "epoch": 2.71128688773958,
+      "grad_norm": 0.4372463822364807,
+      "learning_rate": 1.2640414229440983e-05,
+      "loss": 0.0364,
+      "step": 4456
+    },
+    {
+      "epoch": 2.7118953452996655,
+      "grad_norm": 0.5272347927093506,
+      "learning_rate": 1.2629594671836292e-05,
+      "loss": 0.0545,
+      "step": 4457
+    },
+    {
+      "epoch": 2.7125038028597506,
+      "grad_norm": 0.49635735154151917,
+      "learning_rate": 1.2618778181576513e-05,
+      "loss": 0.0401,
+      "step": 4458
+    },
+    {
+      "epoch": 2.7131122604198357,
+      "grad_norm": 0.430007666349411,
+      "learning_rate": 1.260796476134366e-05,
+      "loss": 0.0336,
+      "step": 4459
+    },
+    {
+      "epoch": 2.7137207179799208,
+      "grad_norm": 0.6025989055633545,
+      "learning_rate": 1.2597154413819018e-05,
+      "loss": 0.0432,
+      "step": 4460
+    },
+    {
+      "epoch": 2.714329175540006,
+      "grad_norm": 0.5454067587852478,
+      "learning_rate": 1.2586347141683108e-05,
+      "loss": 0.0383,
+      "step": 4461
+    },
+    {
+      "epoch": 2.7149376331000914,
+      "grad_norm": 0.5201361775398254,
+      "learning_rate": 1.2575542947615675e-05,
+      "loss": 0.0388,
+      "step": 4462
+    },
+    {
+      "epoch": 2.7155460906601765,
+      "grad_norm": 0.5323150753974915,
+      "learning_rate": 1.256474183429572e-05,
+      "loss": 0.0497,
+      "step": 4463
+    },
+    {
+      "epoch": 2.7161545482202616,
+      "grad_norm": 0.4000251889228821,
+      "learning_rate": 1.2553943804401472e-05,
+      "loss": 0.0306,
+      "step": 4464
+    },
+    {
+      "epoch": 2.7167630057803467,
+      "grad_norm": 0.47391611337661743,
+      "learning_rate": 1.254314886061037e-05,
+      "loss": 0.0445,
+      "step": 4465
+    },
+    {
+      "epoch": 2.717371463340432,
+      "grad_norm": 0.3792179524898529,
+      "learning_rate": 1.2532357005599126e-05,
+      "loss": 0.0361,
+      "step": 4466
+    },
+    {
+      "epoch": 2.7179799209005173,
+      "grad_norm": 0.4722985029220581,
+      "learning_rate": 1.2521568242043669e-05,
+      "loss": 0.0487,
+      "step": 4467
+    },
+    {
+      "epoch": 2.7185883784606024,
+      "grad_norm": 0.4448596239089966,
+      "learning_rate": 1.251078257261916e-05,
+      "loss": 0.0373,
+      "step": 4468
+    },
+    {
+      "epoch": 2.7191968360206875,
+      "grad_norm": 0.482346773147583,
+      "learning_rate": 1.2500000000000006e-05,
+      "loss": 0.0407,
+      "step": 4469
+    },
+    {
+      "epoch": 2.7198052935807726,
+      "grad_norm": 0.5089819431304932,
+      "learning_rate": 1.2489220526859816e-05,
+      "loss": 0.0476,
+      "step": 4470
+    },
+    {
+      "epoch": 2.7204137511408577,
+      "grad_norm": 0.3887785077095032,
+      "learning_rate": 1.247844415587146e-05,
+      "loss": 0.0293,
+      "step": 4471
+    },
+    {
+      "epoch": 2.7210222087009432,
+      "grad_norm": 0.5263569951057434,
+      "learning_rate": 1.2467670889707032e-05,
+      "loss": 0.0463,
+      "step": 4472
+    },
+    {
+      "epoch": 2.7216306662610283,
+      "grad_norm": 0.7234339714050293,
+      "learning_rate": 1.2456900731037849e-05,
+      "loss": 0.0493,
+      "step": 4473
+    },
+    {
+      "epoch": 2.7222391238211134,
+      "grad_norm": 0.565403938293457,
+      "learning_rate": 1.2446133682534473e-05,
+      "loss": 0.0406,
+      "step": 4474
+    },
+    {
+      "epoch": 2.7228475813811985,
+      "grad_norm": 0.476296067237854,
+      "learning_rate": 1.243536974686666e-05,
+      "loss": 0.0374,
+      "step": 4475
+    },
+    {
+      "epoch": 2.7234560389412836,
+      "grad_norm": 0.49702188372612,
+      "learning_rate": 1.2424608926703433e-05,
+      "loss": 0.0248,
+      "step": 4476
+    },
+    {
+      "epoch": 2.724064496501369,
+      "grad_norm": 0.4639529585838318,
+      "learning_rate": 1.2413851224713022e-05,
+      "loss": 0.0491,
+      "step": 4477
+    },
+    {
+      "epoch": 2.7246729540614543,
+      "grad_norm": 0.5371832251548767,
+      "learning_rate": 1.2403096643562891e-05,
+      "loss": 0.0396,
+      "step": 4478
+    },
+    {
+      "epoch": 2.7252814116215394,
+      "grad_norm": 0.7577241659164429,
+      "learning_rate": 1.2392345185919737e-05,
+      "loss": 0.0627,
+      "step": 4479
+    },
+    {
+      "epoch": 2.7258898691816245,
+      "grad_norm": 0.46722671389579773,
+      "learning_rate": 1.2381596854449457e-05,
+      "loss": 0.0446,
+      "step": 4480
+    },
+    {
+      "epoch": 2.7264983267417096,
+      "grad_norm": 0.39545291662216187,
+      "learning_rate": 1.2370851651817194e-05,
+      "loss": 0.0299,
+      "step": 4481
+    },
+    {
+      "epoch": 2.727106784301795,
+      "grad_norm": 0.4299526810646057,
+      "learning_rate": 1.2360109580687313e-05,
+      "loss": 0.0343,
+      "step": 4482
+    },
+    {
+      "epoch": 2.72771524186188,
+      "grad_norm": 0.3463142514228821,
+      "learning_rate": 1.2349370643723399e-05,
+      "loss": 0.0296,
+      "step": 4483
+    },
+    {
+      "epoch": 2.7283236994219653,
+      "grad_norm": 0.42831888794898987,
+      "learning_rate": 1.2338634843588263e-05,
+      "loss": 0.0388,
+      "step": 4484
+    },
+    {
+      "epoch": 2.7289321569820504,
+      "grad_norm": 0.5531302690505981,
+      "learning_rate": 1.232790218294394e-05,
+      "loss": 0.0563,
+      "step": 4485
+    },
+    {
+      "epoch": 2.7295406145421355,
+      "grad_norm": 0.537441611289978,
+      "learning_rate": 1.231717266445167e-05,
+      "loss": 0.0507,
+      "step": 4486
+    },
+    {
+      "epoch": 2.730149072102221,
+      "grad_norm": 0.5240452885627747,
+      "learning_rate": 1.2306446290771934e-05,
+      "loss": 0.0478,
+      "step": 4487
+    },
+    {
+      "epoch": 2.730757529662306,
+      "grad_norm": 0.5667134523391724,
+      "learning_rate": 1.2295723064564422e-05,
+      "loss": 0.0346,
+      "step": 4488
+    },
+    {
+      "epoch": 2.731365987222391,
+      "grad_norm": 0.6184573769569397,
+      "learning_rate": 1.228500298848806e-05,
+      "loss": 0.0501,
+      "step": 4489
+    },
+    {
+      "epoch": 2.7319744447824763,
+      "grad_norm": 0.47468021512031555,
+      "learning_rate": 1.227428606520095e-05,
+      "loss": 0.0299,
+      "step": 4490
+    },
+    {
+      "epoch": 2.7325829023425614,
+      "grad_norm": 0.534231424331665,
+      "learning_rate": 1.2263572297360478e-05,
+      "loss": 0.0534,
+      "step": 4491
+    },
+    {
+      "epoch": 2.733191359902647,
+      "grad_norm": 0.3944413959980011,
+      "learning_rate": 1.225286168762319e-05,
+      "loss": 0.0295,
+      "step": 4492
+    },
+    {
+      "epoch": 2.733799817462732,
+      "grad_norm": 0.45191413164138794,
+      "learning_rate": 1.2242154238644879e-05,
+      "loss": 0.0385,
+      "step": 4493
+    },
+    {
+      "epoch": 2.734408275022817,
+      "grad_norm": 0.5744283199310303,
+      "learning_rate": 1.223144995308054e-05,
+      "loss": 0.0453,
+      "step": 4494
+    },
+    {
+      "epoch": 2.7350167325829022,
+      "grad_norm": 0.542323887348175,
+      "learning_rate": 1.2220748833584403e-05,
+      "loss": 0.0428,
+      "step": 4495
+    },
+    {
+      "epoch": 2.7356251901429873,
+      "grad_norm": 0.4557442367076874,
+      "learning_rate": 1.221005088280987e-05,
+      "loss": 0.0253,
+      "step": 4496
+    },
+    {
+      "epoch": 2.736233647703073,
+      "grad_norm": 0.5451676845550537,
+      "learning_rate": 1.219935610340963e-05,
+      "loss": 0.0469,
+      "step": 4497
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.568136990070343,
+      "learning_rate": 1.2188664498035504e-05,
+      "loss": 0.0496,
+      "step": 4498
+    },
+    {
+      "epoch": 2.737450562823243,
+      "grad_norm": 0.451236754655838,
+      "learning_rate": 1.2177976069338592e-05,
+      "loss": 0.0302,
+      "step": 4499
+    },
+    {
+      "epoch": 2.738059020383328,
+      "grad_norm": 0.5220772624015808,
+      "learning_rate": 1.2167290819969149e-05,
+      "loss": 0.0521,
+      "step": 4500
+    },
+    {
+      "epoch": 2.7386674779434133,
+      "grad_norm": 0.508506715297699,
+      "learning_rate": 1.2156608752576707e-05,
+      "loss": 0.0526,
+      "step": 4501
+    },
+    {
+      "epoch": 2.739275935503499,
+      "grad_norm": 0.49221986532211304,
+      "learning_rate": 1.2145929869809944e-05,
+      "loss": 0.0402,
+      "step": 4502
+    },
+    {
+      "epoch": 2.739884393063584,
+      "grad_norm": 0.5821833610534668,
+      "learning_rate": 1.213525417431679e-05,
+      "loss": 0.0393,
+      "step": 4503
+    },
+    {
+      "epoch": 2.740492850623669,
+      "grad_norm": 0.4816054701805115,
+      "learning_rate": 1.2124581668744372e-05,
+      "loss": 0.0305,
+      "step": 4504
+    },
+    {
+      "epoch": 2.741101308183754,
+      "grad_norm": 0.5656616687774658,
+      "learning_rate": 1.2113912355739037e-05,
+      "loss": 0.0549,
+      "step": 4505
+    },
+    {
+      "epoch": 2.741709765743839,
+      "grad_norm": 0.5027511715888977,
+      "learning_rate": 1.21032462379463e-05,
+      "loss": 0.042,
+      "step": 4506
+    },
+    {
+      "epoch": 2.7423182233039247,
+      "grad_norm": 1.0891318321228027,
+      "learning_rate": 1.209258331801095e-05,
+      "loss": 0.1,
+      "step": 4507
+    },
+    {
+      "epoch": 2.74292668086401,
+      "grad_norm": 0.5207118988037109,
+      "learning_rate": 1.2081923598576921e-05,
+      "loss": 0.0473,
+      "step": 4508
+    },
+    {
+      "epoch": 2.743535138424095,
+      "grad_norm": 0.5301030278205872,
+      "learning_rate": 1.2071267082287388e-05,
+      "loss": 0.0409,
+      "step": 4509
+    },
+    {
+      "epoch": 2.74414359598418,
+      "grad_norm": 0.577142596244812,
+      "learning_rate": 1.2060613771784724e-05,
+      "loss": 0.0391,
+      "step": 4510
+    },
+    {
+      "epoch": 2.744752053544265,
+      "grad_norm": 0.4636692702770233,
+      "learning_rate": 1.204996366971051e-05,
+      "loss": 0.0381,
+      "step": 4511
+    },
+    {
+      "epoch": 2.7453605111043506,
+      "grad_norm": 0.43843039870262146,
+      "learning_rate": 1.2039316778705514e-05,
+      "loss": 0.0349,
+      "step": 4512
+    },
+    {
+      "epoch": 2.7459689686644357,
+      "grad_norm": 0.5346546769142151,
+      "learning_rate": 1.2028673101409729e-05,
+      "loss": 0.0466,
+      "step": 4513
+    },
+    {
+      "epoch": 2.746577426224521,
+      "grad_norm": 0.5091422200202942,
+      "learning_rate": 1.2018032640462345e-05,
+      "loss": 0.0382,
+      "step": 4514
+    },
+    {
+      "epoch": 2.747185883784606,
+      "grad_norm": 0.4113479256629944,
+      "learning_rate": 1.2007395398501759e-05,
+      "loss": 0.0386,
+      "step": 4515
+    },
+    {
+      "epoch": 2.747794341344691,
+      "grad_norm": 0.44784247875213623,
+      "learning_rate": 1.1996761378165535e-05,
+      "loss": 0.0397,
+      "step": 4516
+    },
+    {
+      "epoch": 2.7484027989047766,
+      "grad_norm": 0.44350409507751465,
+      "learning_rate": 1.198613058209051e-05,
+      "loss": 0.0327,
+      "step": 4517
+    },
+    {
+      "epoch": 2.7490112564648617,
+      "grad_norm": 0.4770455062389374,
+      "learning_rate": 1.1975503012912645e-05,
+      "loss": 0.0396,
+      "step": 4518
+    },
+    {
+      "epoch": 2.7496197140249468,
+      "grad_norm": 0.4313164949417114,
+      "learning_rate": 1.1964878673267143e-05,
+      "loss": 0.0279,
+      "step": 4519
+    },
+    {
+      "epoch": 2.750228171585032,
+      "grad_norm": 0.5598992705345154,
+      "learning_rate": 1.1954257565788402e-05,
+      "loss": 0.0391,
+      "step": 4520
+    },
+    {
+      "epoch": 2.750836629145117,
+      "grad_norm": 0.44698452949523926,
+      "learning_rate": 1.194363969311002e-05,
+      "loss": 0.0492,
+      "step": 4521
+    },
+    {
+      "epoch": 2.7514450867052025,
+      "grad_norm": 0.4371298551559448,
+      "learning_rate": 1.1933025057864769e-05,
+      "loss": 0.0342,
+      "step": 4522
+    },
+    {
+      "epoch": 2.7520535442652876,
+      "grad_norm": 0.4278777539730072,
+      "learning_rate": 1.1922413662684644e-05,
+      "loss": 0.0251,
+      "step": 4523
+    },
+    {
+      "epoch": 2.7526620018253727,
+      "grad_norm": 0.6282891631126404,
+      "learning_rate": 1.1911805510200833e-05,
+      "loss": 0.0435,
+      "step": 4524
+    },
+    {
+      "epoch": 2.753270459385458,
+      "grad_norm": 0.6077399253845215,
+      "learning_rate": 1.1901200603043718e-05,
+      "loss": 0.0501,
+      "step": 4525
+    },
+    {
+      "epoch": 2.753878916945543,
+      "grad_norm": 0.4395311772823334,
+      "learning_rate": 1.1890598943842854e-05,
+      "loss": 0.0423,
+      "step": 4526
+    },
+    {
+      "epoch": 2.7544873745056284,
+      "grad_norm": 0.5030261874198914,
+      "learning_rate": 1.188000053522704e-05,
+      "loss": 0.0456,
+      "step": 4527
+    },
+    {
+      "epoch": 2.7550958320657135,
+      "grad_norm": 0.49082061648368835,
+      "learning_rate": 1.186940537982422e-05,
+      "loss": 0.0404,
+      "step": 4528
+    },
+    {
+      "epoch": 2.7557042896257986,
+      "grad_norm": 0.476769357919693,
+      "learning_rate": 1.1858813480261552e-05,
+      "loss": 0.0286,
+      "step": 4529
+    },
+    {
+      "epoch": 2.7563127471858837,
+      "grad_norm": 0.4839482307434082,
+      "learning_rate": 1.1848224839165389e-05,
+      "loss": 0.0287,
+      "step": 4530
+    },
+    {
+      "epoch": 2.756921204745969,
+      "grad_norm": 0.5627169609069824,
+      "learning_rate": 1.1837639459161284e-05,
+      "loss": 0.0645,
+      "step": 4531
+    },
+    {
+      "epoch": 2.7575296623060543,
+      "grad_norm": 0.5222127437591553,
+      "learning_rate": 1.182705734287394e-05,
+      "loss": 0.0388,
+      "step": 4532
+    },
+    {
+      "epoch": 2.7581381198661394,
+      "grad_norm": 0.4021141827106476,
+      "learning_rate": 1.1816478492927316e-05,
+      "loss": 0.0307,
+      "step": 4533
+    },
+    {
+      "epoch": 2.7587465774262245,
+      "grad_norm": 0.41538554430007935,
+      "learning_rate": 1.1805902911944503e-05,
+      "loss": 0.032,
+      "step": 4534
+    },
+    {
+      "epoch": 2.7593550349863096,
+      "grad_norm": 0.517659604549408,
+      "learning_rate": 1.179533060254782e-05,
+      "loss": 0.0451,
+      "step": 4535
+    },
+    {
+      "epoch": 2.7599634925463947,
+      "grad_norm": 0.5641751885414124,
+      "learning_rate": 1.1784761567358729e-05,
+      "loss": 0.0513,
+      "step": 4536
+    },
+    {
+      "epoch": 2.7605719501064803,
+      "grad_norm": 0.5050565600395203,
+      "learning_rate": 1.177419580899795e-05,
+      "loss": 0.0318,
+      "step": 4537
+    },
+    {
+      "epoch": 2.7611804076665654,
+      "grad_norm": 0.5235201716423035,
+      "learning_rate": 1.1763633330085325e-05,
+      "loss": 0.0241,
+      "step": 4538
+    },
+    {
+      "epoch": 2.7617888652266505,
+      "grad_norm": 0.6112051010131836,
+      "learning_rate": 1.1753074133239914e-05,
+      "loss": 0.0513,
+      "step": 4539
+    },
+    {
+      "epoch": 2.7623973227867356,
+      "grad_norm": 0.3812270164489746,
+      "learning_rate": 1.1742518221079957e-05,
+      "loss": 0.0267,
+      "step": 4540
+    },
+    {
+      "epoch": 2.7630057803468207,
+      "grad_norm": 0.4638807773590088,
+      "learning_rate": 1.1731965596222893e-05,
+      "loss": 0.0362,
+      "step": 4541
+    },
+    {
+      "epoch": 2.763614237906906,
+      "grad_norm": 0.47099676728248596,
+      "learning_rate": 1.1721416261285303e-05,
+      "loss": 0.0516,
+      "step": 4542
+    },
+    {
+      "epoch": 2.7642226954669913,
+      "grad_norm": 0.6157107353210449,
+      "learning_rate": 1.1710870218883022e-05,
+      "loss": 0.0583,
+      "step": 4543
+    },
+    {
+      "epoch": 2.7648311530270764,
+      "grad_norm": 0.4588291049003601,
+      "learning_rate": 1.1700327471630995e-05,
+      "loss": 0.03,
+      "step": 4544
+    },
+    {
+      "epoch": 2.7654396105871615,
+      "grad_norm": 0.54620760679245,
+      "learning_rate": 1.1689788022143411e-05,
+      "loss": 0.0413,
+      "step": 4545
+    },
+    {
+      "epoch": 2.7660480681472466,
+      "grad_norm": 0.5352728366851807,
+      "learning_rate": 1.167925187303358e-05,
+      "loss": 0.0469,
+      "step": 4546
+    },
+    {
+      "epoch": 2.766656525707332,
+      "grad_norm": 0.560219407081604,
+      "learning_rate": 1.1668719026914068e-05,
+      "loss": 0.032,
+      "step": 4547
+    },
+    {
+      "epoch": 2.767264983267417,
+      "grad_norm": 0.5410435199737549,
+      "learning_rate": 1.165818948639655e-05,
+      "loss": 0.0457,
+      "step": 4548
+    },
+    {
+      "epoch": 2.7678734408275023,
+      "grad_norm": 0.46575361490249634,
+      "learning_rate": 1.1647663254091928e-05,
+      "loss": 0.0321,
+      "step": 4549
+    },
+    {
+      "epoch": 2.7684818983875874,
+      "grad_norm": 0.571999728679657,
+      "learning_rate": 1.1637140332610267e-05,
+      "loss": 0.05,
+      "step": 4550
+    },
+    {
+      "epoch": 2.7690903559476725,
+      "grad_norm": 0.4646073281764984,
+      "learning_rate": 1.1626620724560819e-05,
+      "loss": 0.04,
+      "step": 4551
+    },
+    {
+      "epoch": 2.769698813507758,
+      "grad_norm": 0.44628095626831055,
+      "learning_rate": 1.1616104432551982e-05,
+      "loss": 0.0328,
+      "step": 4552
+    },
+    {
+      "epoch": 2.770307271067843,
+      "grad_norm": 0.49966830015182495,
+      "learning_rate": 1.16055914591914e-05,
+      "loss": 0.0392,
+      "step": 4553
+    },
+    {
+      "epoch": 2.7709157286279282,
+      "grad_norm": 0.4285971522331238,
+      "learning_rate": 1.1595081807085816e-05,
+      "loss": 0.0387,
+      "step": 4554
+    },
+    {
+      "epoch": 2.7715241861880133,
+      "grad_norm": 0.46280503273010254,
+      "learning_rate": 1.1584575478841204e-05,
+      "loss": 0.0361,
+      "step": 4555
+    },
+    {
+      "epoch": 2.7721326437480984,
+      "grad_norm": 0.48611634969711304,
+      "learning_rate": 1.1574072477062686e-05,
+      "loss": 0.0432,
+      "step": 4556
+    },
+    {
+      "epoch": 2.772741101308184,
+      "grad_norm": 0.4622981548309326,
+      "learning_rate": 1.1563572804354586e-05,
+      "loss": 0.0429,
+      "step": 4557
+    },
+    {
+      "epoch": 2.773349558868269,
+      "grad_norm": 0.500038206577301,
+      "learning_rate": 1.1553076463320363e-05,
+      "loss": 0.0648,
+      "step": 4558
+    },
+    {
+      "epoch": 2.773958016428354,
+      "grad_norm": 0.5276514291763306,
+      "learning_rate": 1.154258345656268e-05,
+      "loss": 0.0452,
+      "step": 4559
+    },
+    {
+      "epoch": 2.7745664739884393,
+      "grad_norm": 0.6001972556114197,
+      "learning_rate": 1.1532093786683368e-05,
+      "loss": 0.0578,
+      "step": 4560
+    },
+    {
+      "epoch": 2.7751749315485243,
+      "grad_norm": 0.5367801785469055,
+      "learning_rate": 1.1521607456283437e-05,
+      "loss": 0.0388,
+      "step": 4561
+    },
+    {
+      "epoch": 2.77578338910861,
+      "grad_norm": 0.45866864919662476,
+      "learning_rate": 1.151112446796303e-05,
+      "loss": 0.0373,
+      "step": 4562
+    },
+    {
+      "epoch": 2.776391846668695,
+      "grad_norm": 0.549251139163971,
+      "learning_rate": 1.1500644824321529e-05,
+      "loss": 0.05,
+      "step": 4563
+    },
+    {
+      "epoch": 2.77700030422878,
+      "grad_norm": 0.6965852379798889,
+      "learning_rate": 1.1490168527957423e-05,
+      "loss": 0.0401,
+      "step": 4564
+    },
+    {
+      "epoch": 2.777608761788865,
+      "grad_norm": 0.4115385413169861,
+      "learning_rate": 1.1479695581468405e-05,
+      "loss": 0.0365,
+      "step": 4565
+    },
+    {
+      "epoch": 2.7782172193489503,
+      "grad_norm": 0.4784400761127472,
+      "learning_rate": 1.1469225987451327e-05,
+      "loss": 0.0356,
+      "step": 4566
+    },
+    {
+      "epoch": 2.778825676909036,
+      "grad_norm": 0.49439680576324463,
+      "learning_rate": 1.1458759748502223e-05,
+      "loss": 0.0499,
+      "step": 4567
+    },
+    {
+      "epoch": 2.779434134469121,
+      "grad_norm": 0.39347532391548157,
+      "learning_rate": 1.1448296867216268e-05,
+      "loss": 0.0287,
+      "step": 4568
+    },
+    {
+      "epoch": 2.780042592029206,
+      "grad_norm": 0.5476490259170532,
+      "learning_rate": 1.1437837346187824e-05,
+      "loss": 0.0491,
+      "step": 4569
+    },
+    {
+      "epoch": 2.780651049589291,
+      "grad_norm": 0.5295094847679138,
+      "learning_rate": 1.142738118801042e-05,
+      "loss": 0.0493,
+      "step": 4570
+    },
+    {
+      "epoch": 2.781259507149376,
+      "grad_norm": 0.5610222816467285,
+      "learning_rate": 1.1416928395276757e-05,
+      "loss": 0.0442,
+      "step": 4571
+    },
+    {
+      "epoch": 2.7818679647094617,
+      "grad_norm": 0.443450003862381,
+      "learning_rate": 1.1406478970578662e-05,
+      "loss": 0.0362,
+      "step": 4572
+    },
+    {
+      "epoch": 2.782476422269547,
+      "grad_norm": 0.5407084226608276,
+      "learning_rate": 1.1396032916507196e-05,
+      "loss": 0.046,
+      "step": 4573
+    },
+    {
+      "epoch": 2.783084879829632,
+      "grad_norm": 0.5253666043281555,
+      "learning_rate": 1.1385590235652515e-05,
+      "loss": 0.0456,
+      "step": 4574
+    },
+    {
+      "epoch": 2.783693337389717,
+      "grad_norm": 0.5159477591514587,
+      "learning_rate": 1.1375150930603982e-05,
+      "loss": 0.0368,
+      "step": 4575
+    },
+    {
+      "epoch": 2.784301794949802,
+      "grad_norm": 0.49028316140174866,
+      "learning_rate": 1.1364715003950102e-05,
+      "loss": 0.0431,
+      "step": 4576
+    },
+    {
+      "epoch": 2.7849102525098877,
+      "grad_norm": 0.4681493937969208,
+      "learning_rate": 1.1354282458278567e-05,
+      "loss": 0.0318,
+      "step": 4577
+    },
+    {
+      "epoch": 2.7855187100699728,
+      "grad_norm": 0.44730204343795776,
+      "learning_rate": 1.134385329617618e-05,
+      "loss": 0.0361,
+      "step": 4578
+    },
+    {
+      "epoch": 2.786127167630058,
+      "grad_norm": 0.47637778520584106,
+      "learning_rate": 1.1333427520228979e-05,
+      "loss": 0.037,
+      "step": 4579
+    },
+    {
+      "epoch": 2.786735625190143,
+      "grad_norm": 0.42056208848953247,
+      "learning_rate": 1.1323005133022095e-05,
+      "loss": 0.0391,
+      "step": 4580
+    },
+    {
+      "epoch": 2.787344082750228,
+      "grad_norm": 0.4972124695777893,
+      "learning_rate": 1.131258613713985e-05,
+      "loss": 0.0455,
+      "step": 4581
+    },
+    {
+      "epoch": 2.7879525403103136,
+      "grad_norm": 0.5506184697151184,
+      "learning_rate": 1.130217053516572e-05,
+      "loss": 0.0345,
+      "step": 4582
+    },
+    {
+      "epoch": 2.7885609978703987,
+      "grad_norm": 0.5120692253112793,
+      "learning_rate": 1.1291758329682358e-05,
+      "loss": 0.0371,
+      "step": 4583
+    },
+    {
+      "epoch": 2.789169455430484,
+      "grad_norm": 0.4274442791938782,
+      "learning_rate": 1.1281349523271534e-05,
+      "loss": 0.0344,
+      "step": 4584
+    },
+    {
+      "epoch": 2.789777912990569,
+      "grad_norm": 0.5016206502914429,
+      "learning_rate": 1.1270944118514203e-05,
+      "loss": 0.0366,
+      "step": 4585
+    },
+    {
+      "epoch": 2.790386370550654,
+      "grad_norm": 0.5108194351196289,
+      "learning_rate": 1.1260542117990478e-05,
+      "loss": 0.0465,
+      "step": 4586
+    },
+    {
+      "epoch": 2.7909948281107395,
+      "grad_norm": 0.5079597234725952,
+      "learning_rate": 1.1250143524279618e-05,
+      "loss": 0.0422,
+      "step": 4587
+    },
+    {
+      "epoch": 2.7916032856708246,
+      "grad_norm": 0.42185690999031067,
+      "learning_rate": 1.1239748339960043e-05,
+      "loss": 0.0339,
+      "step": 4588
+    },
+    {
+      "epoch": 2.7922117432309097,
+      "grad_norm": 0.5816838145256042,
+      "learning_rate": 1.122935656760933e-05,
+      "loss": 0.065,
+      "step": 4589
+    },
+    {
+      "epoch": 2.792820200790995,
+      "grad_norm": 0.5395743250846863,
+      "learning_rate": 1.1218968209804192e-05,
+      "loss": 0.0482,
+      "step": 4590
+    },
+    {
+      "epoch": 2.79342865835108,
+      "grad_norm": 0.5038939118385315,
+      "learning_rate": 1.1208583269120517e-05,
+      "loss": 0.0371,
+      "step": 4591
+    },
+    {
+      "epoch": 2.7940371159111654,
+      "grad_norm": 0.43629151582717896,
+      "learning_rate": 1.1198201748133338e-05,
+      "loss": 0.0306,
+      "step": 4592
+    },
+    {
+      "epoch": 2.7946455734712505,
+      "grad_norm": 0.3945944607257843,
+      "learning_rate": 1.1187823649416836e-05,
+      "loss": 0.0291,
+      "step": 4593
+    },
+    {
+      "epoch": 2.7952540310313356,
+      "grad_norm": 0.3942541778087616,
+      "learning_rate": 1.1177448975544361e-05,
+      "loss": 0.0294,
+      "step": 4594
+    },
+    {
+      "epoch": 2.7958624885914207,
+      "grad_norm": 0.4381770193576813,
+      "learning_rate": 1.1167077729088374e-05,
+      "loss": 0.0294,
+      "step": 4595
+    },
+    {
+      "epoch": 2.796470946151506,
+      "grad_norm": 0.7765640616416931,
+      "learning_rate": 1.1156709912620529e-05,
+      "loss": 0.0846,
+      "step": 4596
+    },
+    {
+      "epoch": 2.7970794037115914,
+      "grad_norm": 0.38119301199913025,
+      "learning_rate": 1.1146345528711608e-05,
+      "loss": 0.0215,
+      "step": 4597
+    },
+    {
+      "epoch": 2.7976878612716765,
+      "grad_norm": 0.4524082541465759,
+      "learning_rate": 1.1135984579931547e-05,
+      "loss": 0.0351,
+      "step": 4598
+    },
+    {
+      "epoch": 2.7982963188317616,
+      "grad_norm": 0.4416946768760681,
+      "learning_rate": 1.1125627068849428e-05,
+      "loss": 0.037,
+      "step": 4599
+    },
+    {
+      "epoch": 2.7989047763918466,
+      "grad_norm": 0.5243279933929443,
+      "learning_rate": 1.1115272998033496e-05,
+      "loss": 0.037,
+      "step": 4600
+    },
+    {
+      "epoch": 2.7995132339519317,
+      "grad_norm": 0.5025190711021423,
+      "learning_rate": 1.1104922370051105e-05,
+      "loss": 0.0333,
+      "step": 4601
+    },
+    {
+      "epoch": 2.8001216915120173,
+      "grad_norm": 0.5427147746086121,
+      "learning_rate": 1.1094575187468792e-05,
+      "loss": 0.0279,
+      "step": 4602
+    },
+    {
+      "epoch": 2.8007301490721024,
+      "grad_norm": 0.369625985622406,
+      "learning_rate": 1.1084231452852226e-05,
+      "loss": 0.029,
+      "step": 4603
+    },
+    {
+      "epoch": 2.8013386066321875,
+      "grad_norm": 0.39575597643852234,
+      "learning_rate": 1.1073891168766229e-05,
+      "loss": 0.0311,
+      "step": 4604
+    },
+    {
+      "epoch": 2.8019470641922726,
+      "grad_norm": 0.6142522692680359,
+      "learning_rate": 1.1063554337774745e-05,
+      "loss": 0.0341,
+      "step": 4605
+    },
+    {
+      "epoch": 2.8025555217523577,
+      "grad_norm": 0.4445866048336029,
+      "learning_rate": 1.1053220962440889e-05,
+      "loss": 0.0396,
+      "step": 4606
+    },
+    {
+      "epoch": 2.803163979312443,
+      "grad_norm": 0.5825973749160767,
+      "learning_rate": 1.10428910453269e-05,
+      "loss": 0.033,
+      "step": 4607
+    },
+    {
+      "epoch": 2.8037724368725283,
+      "grad_norm": 0.4755668342113495,
+      "learning_rate": 1.103256458899417e-05,
+      "loss": 0.0303,
+      "step": 4608
+    },
+    {
+      "epoch": 2.8043808944326134,
+      "grad_norm": 0.5290636420249939,
+      "learning_rate": 1.1022241596003232e-05,
+      "loss": 0.045,
+      "step": 4609
+    },
+    {
+      "epoch": 2.8049893519926985,
+      "grad_norm": 0.44851988554000854,
+      "learning_rate": 1.101192206891377e-05,
+      "loss": 0.0318,
+      "step": 4610
+    },
+    {
+      "epoch": 2.8055978095527836,
+      "grad_norm": 0.5043997764587402,
+      "learning_rate": 1.1001606010284569e-05,
+      "loss": 0.0462,
+      "step": 4611
+    },
+    {
+      "epoch": 2.806206267112869,
+      "grad_norm": 0.41543683409690857,
+      "learning_rate": 1.0991293422673596e-05,
+      "loss": 0.0359,
+      "step": 4612
+    },
+    {
+      "epoch": 2.806814724672954,
+      "grad_norm": 0.5744189620018005,
+      "learning_rate": 1.0980984308637944e-05,
+      "loss": 0.0454,
+      "step": 4613
+    },
+    {
+      "epoch": 2.8074231822330393,
+      "grad_norm": 0.4493009150028229,
+      "learning_rate": 1.0970678670733839e-05,
+      "loss": 0.031,
+      "step": 4614
+    },
+    {
+      "epoch": 2.8080316397931244,
+      "grad_norm": 0.466262549161911,
+      "learning_rate": 1.0960376511516655e-05,
+      "loss": 0.0362,
+      "step": 4615
+    },
+    {
+      "epoch": 2.8086400973532095,
+      "grad_norm": 0.5800133943557739,
+      "learning_rate": 1.0950077833540906e-05,
+      "loss": 0.0544,
+      "step": 4616
+    },
+    {
+      "epoch": 2.809248554913295,
+      "grad_norm": 0.32791540026664734,
+      "learning_rate": 1.0939782639360214e-05,
+      "loss": 0.0188,
+      "step": 4617
+    },
+    {
+      "epoch": 2.8098570124733797,
+      "grad_norm": 0.45963072776794434,
+      "learning_rate": 1.0929490931527369e-05,
+      "loss": 0.0387,
+      "step": 4618
+    },
+    {
+      "epoch": 2.8104654700334653,
+      "grad_norm": 0.46646299958229065,
+      "learning_rate": 1.0919202712594284e-05,
+      "loss": 0.0344,
+      "step": 4619
+    },
+    {
+      "epoch": 2.8110739275935503,
+      "grad_norm": 0.39890867471694946,
+      "learning_rate": 1.0908917985112021e-05,
+      "loss": 0.0238,
+      "step": 4620
+    },
+    {
+      "epoch": 2.8116823851536354,
+      "grad_norm": 0.439327210187912,
+      "learning_rate": 1.0898636751630733e-05,
+      "loss": 0.0415,
+      "step": 4621
+    },
+    {
+      "epoch": 2.812290842713721,
+      "grad_norm": 0.3781556785106659,
+      "learning_rate": 1.0888359014699776e-05,
+      "loss": 0.0332,
+      "step": 4622
+    },
+    {
+      "epoch": 2.8128993002738056,
+      "grad_norm": 0.42909082770347595,
+      "learning_rate": 1.087808477686757e-05,
+      "loss": 0.0419,
+      "step": 4623
+    },
+    {
+      "epoch": 2.813507757833891,
+      "grad_norm": 0.4752819538116455,
+      "learning_rate": 1.0867814040681711e-05,
+      "loss": 0.0308,
+      "step": 4624
+    },
+    {
+      "epoch": 2.8141162153939763,
+      "grad_norm": 0.5056931376457214,
+      "learning_rate": 1.0857546808688912e-05,
+      "loss": 0.0384,
+      "step": 4625
+    },
+    {
+      "epoch": 2.8147246729540614,
+      "grad_norm": 0.9717746376991272,
+      "learning_rate": 1.0847283083435026e-05,
+      "loss": 0.0731,
+      "step": 4626
+    },
+    {
+      "epoch": 2.815333130514147,
+      "grad_norm": 0.4587456285953522,
+      "learning_rate": 1.083702286746501e-05,
+      "loss": 0.0435,
+      "step": 4627
+    },
+    {
+      "epoch": 2.8159415880742316,
+      "grad_norm": 0.5184073448181152,
+      "learning_rate": 1.0826766163322982e-05,
+      "loss": 0.0337,
+      "step": 4628
+    },
+    {
+      "epoch": 2.816550045634317,
+      "grad_norm": 0.5532636046409607,
+      "learning_rate": 1.0816512973552178e-05,
+      "loss": 0.0415,
+      "step": 4629
+    },
+    {
+      "epoch": 2.817158503194402,
+      "grad_norm": 0.5456773638725281,
+      "learning_rate": 1.0806263300694966e-05,
+      "loss": 0.0372,
+      "step": 4630
+    },
+    {
+      "epoch": 2.8177669607544873,
+      "grad_norm": 0.5305153727531433,
+      "learning_rate": 1.0796017147292817e-05,
+      "loss": 0.0458,
+      "step": 4631
+    },
+    {
+      "epoch": 2.818375418314573,
+      "grad_norm": 0.4857366681098938,
+      "learning_rate": 1.0785774515886379e-05,
+      "loss": 0.0372,
+      "step": 4632
+    },
+    {
+      "epoch": 2.8189838758746575,
+      "grad_norm": 0.5140689611434937,
+      "learning_rate": 1.0775535409015374e-05,
+      "loss": 0.0521,
+      "step": 4633
+    },
+    {
+      "epoch": 2.819592333434743,
+      "grad_norm": 0.5720868110656738,
+      "learning_rate": 1.0765299829218683e-05,
+      "loss": 0.0484,
+      "step": 4634
+    },
+    {
+      "epoch": 2.820200790994828,
+      "grad_norm": 0.4893862009048462,
+      "learning_rate": 1.0755067779034302e-05,
+      "loss": 0.0446,
+      "step": 4635
+    },
+    {
+      "epoch": 2.820809248554913,
+      "grad_norm": 0.4779762923717499,
+      "learning_rate": 1.0744839260999368e-05,
+      "loss": 0.0327,
+      "step": 4636
+    },
+    {
+      "epoch": 2.8214177061149983,
+      "grad_norm": 0.4866357743740082,
+      "learning_rate": 1.073461427765009e-05,
+      "loss": 0.0258,
+      "step": 4637
+    },
+    {
+      "epoch": 2.8220261636750834,
+      "grad_norm": 0.42479321360588074,
+      "learning_rate": 1.0724392831521878e-05,
+      "loss": 0.0353,
+      "step": 4638
+    },
+    {
+      "epoch": 2.822634621235169,
+      "grad_norm": 0.4593687951564789,
+      "learning_rate": 1.0714174925149201e-05,
+      "loss": 0.0423,
+      "step": 4639
+    },
+    {
+      "epoch": 2.823243078795254,
+      "grad_norm": 0.5180348753929138,
+      "learning_rate": 1.0703960561065688e-05,
+      "loss": 0.0448,
+      "step": 4640
+    },
+    {
+      "epoch": 2.823851536355339,
+      "grad_norm": 0.5316763520240784,
+      "learning_rate": 1.0693749741804048e-05,
+      "loss": 0.038,
+      "step": 4641
+    },
+    {
+      "epoch": 2.8244599939154242,
+      "grad_norm": 0.28423795104026794,
+      "learning_rate": 1.068354246989618e-05,
+      "loss": 0.0152,
+      "step": 4642
+    },
+    {
+      "epoch": 2.8250684514755093,
+      "grad_norm": 0.3734150230884552,
+      "learning_rate": 1.0673338747873027e-05,
+      "loss": 0.0258,
+      "step": 4643
+    },
+    {
+      "epoch": 2.825676909035595,
+      "grad_norm": 0.4261896312236786,
+      "learning_rate": 1.06631385782647e-05,
+      "loss": 0.0279,
+      "step": 4644
+    },
+    {
+      "epoch": 2.82628536659568,
+      "grad_norm": 0.35567089915275574,
+      "learning_rate": 1.0652941963600418e-05,
+      "loss": 0.027,
+      "step": 4645
+    },
+    {
+      "epoch": 2.826893824155765,
+      "grad_norm": 0.39518773555755615,
+      "learning_rate": 1.0642748906408522e-05,
+      "loss": 0.0355,
+      "step": 4646
+    },
+    {
+      "epoch": 2.82750228171585,
+      "grad_norm": 0.5049468874931335,
+      "learning_rate": 1.0632559409216442e-05,
+      "loss": 0.0393,
+      "step": 4647
+    },
+    {
+      "epoch": 2.8281107392759353,
+      "grad_norm": 0.4476812481880188,
+      "learning_rate": 1.062237347455078e-05,
+      "loss": 0.0258,
+      "step": 4648
+    },
+    {
+      "epoch": 2.828719196836021,
+      "grad_norm": 0.41204380989074707,
+      "learning_rate": 1.0612191104937198e-05,
+      "loss": 0.0345,
+      "step": 4649
+    },
+    {
+      "epoch": 2.829327654396106,
+      "grad_norm": 0.5308571457862854,
+      "learning_rate": 1.0602012302900516e-05,
+      "loss": 0.0422,
+      "step": 4650
+    },
+    {
+      "epoch": 2.829936111956191,
+      "grad_norm": 0.43992185592651367,
+      "learning_rate": 1.059183707096463e-05,
+      "loss": 0.0342,
+      "step": 4651
+    },
+    {
+      "epoch": 2.830544569516276,
+      "grad_norm": 0.482450395822525,
+      "learning_rate": 1.0581665411652605e-05,
+      "loss": 0.0282,
+      "step": 4652
+    },
+    {
+      "epoch": 2.831153027076361,
+      "grad_norm": 0.41237378120422363,
+      "learning_rate": 1.0571497327486563e-05,
+      "loss": 0.033,
+      "step": 4653
+    },
+    {
+      "epoch": 2.8317614846364467,
+      "grad_norm": 0.5276917219161987,
+      "learning_rate": 1.0561332820987774e-05,
+      "loss": 0.0362,
+      "step": 4654
+    },
+    {
+      "epoch": 2.832369942196532,
+      "grad_norm": 0.4543791115283966,
+      "learning_rate": 1.0551171894676615e-05,
+      "loss": 0.0247,
+      "step": 4655
+    },
+    {
+      "epoch": 2.832978399756617,
+      "grad_norm": 0.4912313222885132,
+      "learning_rate": 1.0541014551072576e-05,
+      "loss": 0.0258,
+      "step": 4656
+    },
+    {
+      "epoch": 2.833586857316702,
+      "grad_norm": 0.46352067589759827,
+      "learning_rate": 1.053086079269423e-05,
+      "loss": 0.0371,
+      "step": 4657
+    },
+    {
+      "epoch": 2.834195314876787,
+      "grad_norm": 0.5429676175117493,
+      "learning_rate": 1.052071062205932e-05,
+      "loss": 0.0499,
+      "step": 4658
+    },
+    {
+      "epoch": 2.8348037724368726,
+      "grad_norm": 0.4813297688961029,
+      "learning_rate": 1.0510564041684645e-05,
+      "loss": 0.0556,
+      "step": 4659
+    },
+    {
+      "epoch": 2.8354122299969577,
+      "grad_norm": 0.5711151957511902,
+      "learning_rate": 1.0500421054086135e-05,
+      "loss": 0.0396,
+      "step": 4660
+    },
+    {
+      "epoch": 2.836020687557043,
+      "grad_norm": 0.48414355516433716,
+      "learning_rate": 1.049028166177883e-05,
+      "loss": 0.0347,
+      "step": 4661
+    },
+    {
+      "epoch": 2.836629145117128,
+      "grad_norm": 0.5011805295944214,
+      "learning_rate": 1.0480145867276892e-05,
+      "loss": 0.0439,
+      "step": 4662
+    },
+    {
+      "epoch": 2.837237602677213,
+      "grad_norm": 0.49640247225761414,
+      "learning_rate": 1.0470013673093548e-05,
+      "loss": 0.0395,
+      "step": 4663
+    },
+    {
+      "epoch": 2.8378460602372986,
+      "grad_norm": 0.48717397451400757,
+      "learning_rate": 1.0459885081741175e-05,
+      "loss": 0.0382,
+      "step": 4664
+    },
+    {
+      "epoch": 2.8384545177973837,
+      "grad_norm": 0.4565880596637726,
+      "learning_rate": 1.044976009573124e-05,
+      "loss": 0.0269,
+      "step": 4665
+    },
+    {
+      "epoch": 2.8390629753574688,
+      "grad_norm": 0.6080870032310486,
+      "learning_rate": 1.0439638717574327e-05,
+      "loss": 0.0385,
+      "step": 4666
+    },
+    {
+      "epoch": 2.839671432917554,
+      "grad_norm": 0.6103970408439636,
+      "learning_rate": 1.0429520949780085e-05,
+      "loss": 0.0523,
+      "step": 4667
+    },
+    {
+      "epoch": 2.840279890477639,
+      "grad_norm": 0.45440056920051575,
+      "learning_rate": 1.0419406794857343e-05,
+      "loss": 0.0385,
+      "step": 4668
+    },
+    {
+      "epoch": 2.8408883480377245,
+      "grad_norm": 0.3543758690357208,
+      "learning_rate": 1.0409296255313955e-05,
+      "loss": 0.0286,
+      "step": 4669
+    },
+    {
+      "epoch": 2.8414968055978096,
+      "grad_norm": 0.5107575058937073,
+      "learning_rate": 1.0399189333656925e-05,
+      "loss": 0.0439,
+      "step": 4670
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.45144620537757874,
+      "learning_rate": 1.0389086032392348e-05,
+      "loss": 0.0318,
+      "step": 4671
+    },
+    {
+      "epoch": 2.84271372071798,
+      "grad_norm": 0.41293710470199585,
+      "learning_rate": 1.0378986354025433e-05,
+      "loss": 0.0224,
+      "step": 4672
+    },
+    {
+      "epoch": 2.843322178278065,
+      "grad_norm": 0.4346982538700104,
+      "learning_rate": 1.0368890301060457e-05,
+      "loss": 0.0418,
+      "step": 4673
+    },
+    {
+      "epoch": 2.8439306358381504,
+      "grad_norm": 0.47238776087760925,
+      "learning_rate": 1.0358797876000834e-05,
+      "loss": 0.0314,
+      "step": 4674
+    },
+    {
+      "epoch": 2.8445390933982355,
+      "grad_norm": 0.4769528806209564,
+      "learning_rate": 1.0348709081349062e-05,
+      "loss": 0.0416,
+      "step": 4675
+    },
+    {
+      "epoch": 2.8451475509583206,
+      "grad_norm": 0.4458237886428833,
+      "learning_rate": 1.033862391960675e-05,
+      "loss": 0.0395,
+      "step": 4676
+    },
+    {
+      "epoch": 2.8457560085184057,
+      "grad_norm": 0.3304067850112915,
+      "learning_rate": 1.0328542393274571e-05,
+      "loss": 0.0189,
+      "step": 4677
+    },
+    {
+      "epoch": 2.846364466078491,
+      "grad_norm": 0.3914128243923187,
+      "learning_rate": 1.031846450485236e-05,
+      "loss": 0.0229,
+      "step": 4678
+    },
+    {
+      "epoch": 2.8469729236385763,
+      "grad_norm": 0.4684884250164032,
+      "learning_rate": 1.0308390256838987e-05,
+      "loss": 0.0423,
+      "step": 4679
+    },
+    {
+      "epoch": 2.8475813811986614,
+      "grad_norm": 0.5104598999023438,
+      "learning_rate": 1.0298319651732455e-05,
+      "loss": 0.0385,
+      "step": 4680
+    },
+    {
+      "epoch": 2.8481898387587465,
+      "grad_norm": 0.5490165948867798,
+      "learning_rate": 1.0288252692029851e-05,
+      "loss": 0.0506,
+      "step": 4681
+    },
+    {
+      "epoch": 2.8487982963188316,
+      "grad_norm": 0.3741062581539154,
+      "learning_rate": 1.0278189380227373e-05,
+      "loss": 0.0251,
+      "step": 4682
+    },
+    {
+      "epoch": 2.8494067538789167,
+      "grad_norm": 0.5128927230834961,
+      "learning_rate": 1.0268129718820277e-05,
+      "loss": 0.0526,
+      "step": 4683
+    },
+    {
+      "epoch": 2.8500152114390023,
+      "grad_norm": 0.43258655071258545,
+      "learning_rate": 1.0258073710302973e-05,
+      "loss": 0.0312,
+      "step": 4684
+    },
+    {
+      "epoch": 2.8506236689990874,
+      "grad_norm": 0.5292765498161316,
+      "learning_rate": 1.0248021357168908e-05,
+      "loss": 0.0539,
+      "step": 4685
+    },
+    {
+      "epoch": 2.8512321265591725,
+      "grad_norm": 0.5508760809898376,
+      "learning_rate": 1.023797266191066e-05,
+      "loss": 0.0453,
+      "step": 4686
+    },
+    {
+      "epoch": 2.8518405841192576,
+      "grad_norm": 0.45886969566345215,
+      "learning_rate": 1.0227927627019862e-05,
+      "loss": 0.0382,
+      "step": 4687
+    },
+    {
+      "epoch": 2.8524490416793427,
+      "grad_norm": 0.587009847164154,
+      "learning_rate": 1.0217886254987303e-05,
+      "loss": 0.0589,
+      "step": 4688
+    },
+    {
+      "epoch": 2.853057499239428,
+      "grad_norm": 0.5119481682777405,
+      "learning_rate": 1.0207848548302793e-05,
+      "loss": 0.0305,
+      "step": 4689
+    },
+    {
+      "epoch": 2.8536659567995133,
+      "grad_norm": 0.4871687889099121,
+      "learning_rate": 1.0197814509455275e-05,
+      "loss": 0.0551,
+      "step": 4690
+    },
+    {
+      "epoch": 2.8542744143595984,
+      "grad_norm": 0.5183733701705933,
+      "learning_rate": 1.0187784140932774e-05,
+      "loss": 0.0435,
+      "step": 4691
+    },
+    {
+      "epoch": 2.8548828719196835,
+      "grad_norm": 0.46334466338157654,
+      "learning_rate": 1.0177757445222411e-05,
+      "loss": 0.0406,
+      "step": 4692
+    },
+    {
+      "epoch": 2.8554913294797686,
+      "grad_norm": 0.42779427766799927,
+      "learning_rate": 1.0167734424810363e-05,
+      "loss": 0.024,
+      "step": 4693
+    },
+    {
+      "epoch": 2.856099787039854,
+      "grad_norm": 0.5750960111618042,
+      "learning_rate": 1.0157715082181956e-05,
+      "loss": 0.0467,
+      "step": 4694
+    },
+    {
+      "epoch": 2.856708244599939,
+      "grad_norm": 0.6426092982292175,
+      "learning_rate": 1.0147699419821543e-05,
+      "loss": 0.0576,
+      "step": 4695
+    },
+    {
+      "epoch": 2.8573167021600243,
+      "grad_norm": 0.39237314462661743,
+      "learning_rate": 1.0137687440212598e-05,
+      "loss": 0.0315,
+      "step": 4696
+    },
+    {
+      "epoch": 2.8579251597201094,
+      "grad_norm": 0.4717332124710083,
+      "learning_rate": 1.012767914583768e-05,
+      "loss": 0.029,
+      "step": 4697
+    },
+    {
+      "epoch": 2.8585336172801945,
+      "grad_norm": 0.4430489242076874,
+      "learning_rate": 1.0117674539178428e-05,
+      "loss": 0.0343,
+      "step": 4698
+    },
+    {
+      "epoch": 2.85914207484028,
+      "grad_norm": 0.42511722445487976,
+      "learning_rate": 1.010767362271556e-05,
+      "loss": 0.0306,
+      "step": 4699
+    },
+    {
+      "epoch": 2.859750532400365,
+      "grad_norm": 0.5086604952812195,
+      "learning_rate": 1.009767639892889e-05,
+      "loss": 0.0396,
+      "step": 4700
+    },
+    {
+      "epoch": 2.8603589899604502,
+      "grad_norm": 0.4445130527019501,
+      "learning_rate": 1.0087682870297313e-05,
+      "loss": 0.0335,
+      "step": 4701
+    },
+    {
+      "epoch": 2.8609674475205353,
+      "grad_norm": 0.4963395595550537,
+      "learning_rate": 1.0077693039298808e-05,
+      "loss": 0.0442,
+      "step": 4702
+    },
+    {
+      "epoch": 2.8615759050806204,
+      "grad_norm": 0.5074029564857483,
+      "learning_rate": 1.0067706908410437e-05,
+      "loss": 0.0469,
+      "step": 4703
+    },
+    {
+      "epoch": 2.862184362640706,
+      "grad_norm": 0.4995264708995819,
+      "learning_rate": 1.0057724480108352e-05,
+      "loss": 0.0369,
+      "step": 4704
+    },
+    {
+      "epoch": 2.862792820200791,
+      "grad_norm": 0.41249075531959534,
+      "learning_rate": 1.0047745756867762e-05,
+      "loss": 0.0245,
+      "step": 4705
+    },
+    {
+      "epoch": 2.863401277760876,
+      "grad_norm": 0.4585972726345062,
+      "learning_rate": 1.0037770741162986e-05,
+      "loss": 0.0362,
+      "step": 4706
+    },
+    {
+      "epoch": 2.8640097353209613,
+      "grad_norm": 0.4985556900501251,
+      "learning_rate": 1.0027799435467409e-05,
+      "loss": 0.0433,
+      "step": 4707
+    },
+    {
+      "epoch": 2.8646181928810464,
+      "grad_norm": 0.46208369731903076,
+      "learning_rate": 1.0017831842253509e-05,
+      "loss": 0.0252,
+      "step": 4708
+    },
+    {
+      "epoch": 2.865226650441132,
+      "grad_norm": 0.5662164688110352,
+      "learning_rate": 1.0007867963992814e-05,
+      "loss": 0.0602,
+      "step": 4709
+    },
+    {
+      "epoch": 2.865835108001217,
+      "grad_norm": 0.5789265036582947,
+      "learning_rate": 9.997907803155962e-06,
+      "loss": 0.051,
+      "step": 4710
+    },
+    {
+      "epoch": 2.866443565561302,
+      "grad_norm": 0.4753771126270294,
+      "learning_rate": 9.987951362212658e-06,
+      "loss": 0.0505,
+      "step": 4711
+    },
+    {
+      "epoch": 2.867052023121387,
+      "grad_norm": 0.5585101842880249,
+      "learning_rate": 9.977998643631684e-06,
+      "loss": 0.0621,
+      "step": 4712
+    },
+    {
+      "epoch": 2.8676604806814723,
+      "grad_norm": 0.396291583776474,
+      "learning_rate": 9.968049649880895e-06,
+      "loss": 0.031,
+      "step": 4713
+    },
+    {
+      "epoch": 2.868268938241558,
+      "grad_norm": 0.5255879163742065,
+      "learning_rate": 9.95810438342724e-06,
+      "loss": 0.0365,
+      "step": 4714
+    },
+    {
+      "epoch": 2.868877395801643,
+      "grad_norm": 0.5379154086112976,
+      "learning_rate": 9.948162846736715e-06,
+      "loss": 0.0466,
+      "step": 4715
+    },
+    {
+      "epoch": 2.869485853361728,
+      "grad_norm": 0.5413723587989807,
+      "learning_rate": 9.93822504227441e-06,
+      "loss": 0.0422,
+      "step": 4716
+    },
+    {
+      "epoch": 2.870094310921813,
+      "grad_norm": 0.6132776141166687,
+      "learning_rate": 9.928290972504489e-06,
+      "loss": 0.044,
+      "step": 4717
+    },
+    {
+      "epoch": 2.870702768481898,
+      "grad_norm": 0.5400335192680359,
+      "learning_rate": 9.918360639890187e-06,
+      "loss": 0.0462,
+      "step": 4718
+    },
+    {
+      "epoch": 2.8713112260419837,
+      "grad_norm": 0.37978994846343994,
+      "learning_rate": 9.908434046893822e-06,
+      "loss": 0.0303,
+      "step": 4719
+    },
+    {
+      "epoch": 2.871919683602069,
+      "grad_norm": 0.49378716945648193,
+      "learning_rate": 9.898511195976756e-06,
+      "loss": 0.0331,
+      "step": 4720
+    },
+    {
+      "epoch": 2.872528141162154,
+      "grad_norm": 0.44092321395874023,
+      "learning_rate": 9.888592089599452e-06,
+      "loss": 0.0358,
+      "step": 4721
+    },
+    {
+      "epoch": 2.873136598722239,
+      "grad_norm": 0.4932500123977661,
+      "learning_rate": 9.878676730221437e-06,
+      "loss": 0.0459,
+      "step": 4722
+    },
+    {
+      "epoch": 2.873745056282324,
+      "grad_norm": 0.36311423778533936,
+      "learning_rate": 9.868765120301305e-06,
+      "loss": 0.0274,
+      "step": 4723
+    },
+    {
+      "epoch": 2.8743535138424097,
+      "grad_norm": 0.4552887976169586,
+      "learning_rate": 9.858857262296719e-06,
+      "loss": 0.0377,
+      "step": 4724
+    },
+    {
+      "epoch": 2.8749619714024948,
+      "grad_norm": 0.4563083052635193,
+      "learning_rate": 9.848953158664431e-06,
+      "loss": 0.0407,
+      "step": 4725
+    },
+    {
+      "epoch": 2.87557042896258,
+      "grad_norm": 0.5411676168441772,
+      "learning_rate": 9.839052811860222e-06,
+      "loss": 0.04,
+      "step": 4726
+    },
+    {
+      "epoch": 2.876178886522665,
+      "grad_norm": 0.34208500385284424,
+      "learning_rate": 9.829156224338975e-06,
+      "loss": 0.0262,
+      "step": 4727
+    },
+    {
+      "epoch": 2.87678734408275,
+      "grad_norm": 0.4771437644958496,
+      "learning_rate": 9.819263398554634e-06,
+      "loss": 0.0411,
+      "step": 4728
+    },
+    {
+      "epoch": 2.8773958016428356,
+      "grad_norm": 0.43631598353385925,
+      "learning_rate": 9.809374336960208e-06,
+      "loss": 0.0348,
+      "step": 4729
+    },
+    {
+      "epoch": 2.8780042592029207,
+      "grad_norm": 0.457810640335083,
+      "learning_rate": 9.799489042007767e-06,
+      "loss": 0.0462,
+      "step": 4730
+    },
+    {
+      "epoch": 2.878612716763006,
+      "grad_norm": 0.5080879926681519,
+      "learning_rate": 9.789607516148467e-06,
+      "loss": 0.0379,
+      "step": 4731
+    },
+    {
+      "epoch": 2.879221174323091,
+      "grad_norm": 0.5162711143493652,
+      "learning_rate": 9.77972976183249e-06,
+      "loss": 0.0444,
+      "step": 4732
+    },
+    {
+      "epoch": 2.879829631883176,
+      "grad_norm": 0.4640320837497711,
+      "learning_rate": 9.769855781509121e-06,
+      "loss": 0.0301,
+      "step": 4733
+    },
+    {
+      "epoch": 2.8804380894432615,
+      "grad_norm": 0.457856148481369,
+      "learning_rate": 9.759985577626695e-06,
+      "loss": 0.0409,
+      "step": 4734
+    },
+    {
+      "epoch": 2.8810465470033466,
+      "grad_norm": 0.43068334460258484,
+      "learning_rate": 9.75011915263262e-06,
+      "loss": 0.0367,
+      "step": 4735
+    },
+    {
+      "epoch": 2.8816550045634317,
+      "grad_norm": 0.5189123153686523,
+      "learning_rate": 9.74025650897333e-06,
+      "loss": 0.0427,
+      "step": 4736
+    },
+    {
+      "epoch": 2.882263462123517,
+      "grad_norm": 0.4327934682369232,
+      "learning_rate": 9.730397649094386e-06,
+      "loss": 0.0368,
+      "step": 4737
+    },
+    {
+      "epoch": 2.882871919683602,
+      "grad_norm": 0.3771323263645172,
+      "learning_rate": 9.72054257544035e-06,
+      "loss": 0.0311,
+      "step": 4738
+    },
+    {
+      "epoch": 2.8834803772436874,
+      "grad_norm": 0.5253776907920837,
+      "learning_rate": 9.710691290454874e-06,
+      "loss": 0.0475,
+      "step": 4739
+    },
+    {
+      "epoch": 2.8840888348037725,
+      "grad_norm": 0.47808972001075745,
+      "learning_rate": 9.70084379658067e-06,
+      "loss": 0.0314,
+      "step": 4740
+    },
+    {
+      "epoch": 2.8846972923638576,
+      "grad_norm": 0.4568268656730652,
+      "learning_rate": 9.691000096259512e-06,
+      "loss": 0.0294,
+      "step": 4741
+    },
+    {
+      "epoch": 2.8853057499239427,
+      "grad_norm": 0.417096346616745,
+      "learning_rate": 9.68116019193221e-06,
+      "loss": 0.0287,
+      "step": 4742
+    },
+    {
+      "epoch": 2.885914207484028,
+      "grad_norm": 0.4551639258861542,
+      "learning_rate": 9.671324086038658e-06,
+      "loss": 0.0252,
+      "step": 4743
+    },
+    {
+      "epoch": 2.8865226650441134,
+      "grad_norm": 0.4030291736125946,
+      "learning_rate": 9.661491781017806e-06,
+      "loss": 0.039,
+      "step": 4744
+    },
+    {
+      "epoch": 2.8871311226041985,
+      "grad_norm": 0.49422821402549744,
+      "learning_rate": 9.651663279307657e-06,
+      "loss": 0.0379,
+      "step": 4745
+    },
+    {
+      "epoch": 2.8877395801642836,
+      "grad_norm": 0.4428713321685791,
+      "learning_rate": 9.64183858334525e-06,
+      "loss": 0.0335,
+      "step": 4746
+    },
+    {
+      "epoch": 2.8883480377243687,
+      "grad_norm": 0.4136061370372772,
+      "learning_rate": 9.632017695566731e-06,
+      "loss": 0.0354,
+      "step": 4747
+    },
+    {
+      "epoch": 2.8889564952844538,
+      "grad_norm": 0.44465282559394836,
+      "learning_rate": 9.622200618407246e-06,
+      "loss": 0.0353,
+      "step": 4748
+    },
+    {
+      "epoch": 2.8895649528445393,
+      "grad_norm": 0.42466622591018677,
+      "learning_rate": 9.612387354301028e-06,
+      "loss": 0.0322,
+      "step": 4749
+    },
+    {
+      "epoch": 2.8901734104046244,
+      "grad_norm": 0.5076179504394531,
+      "learning_rate": 9.602577905681358e-06,
+      "loss": 0.0531,
+      "step": 4750
+    },
+    {
+      "epoch": 2.8907818679647095,
+      "grad_norm": 0.42435580492019653,
+      "learning_rate": 9.592772274980582e-06,
+      "loss": 0.0338,
+      "step": 4751
+    },
+    {
+      "epoch": 2.8913903255247946,
+      "grad_norm": 0.43116191029548645,
+      "learning_rate": 9.58297046463006e-06,
+      "loss": 0.0269,
+      "step": 4752
+    },
+    {
+      "epoch": 2.8919987830848797,
+      "grad_norm": 0.4200409948825836,
+      "learning_rate": 9.573172477060266e-06,
+      "loss": 0.0255,
+      "step": 4753
+    },
+    {
+      "epoch": 2.892607240644965,
+      "grad_norm": 0.44302016496658325,
+      "learning_rate": 9.563378314700665e-06,
+      "loss": 0.036,
+      "step": 4754
+    },
+    {
+      "epoch": 2.8932156982050503,
+      "grad_norm": 0.4883388876914978,
+      "learning_rate": 9.553587979979826e-06,
+      "loss": 0.0334,
+      "step": 4755
+    },
+    {
+      "epoch": 2.8938241557651354,
+      "grad_norm": 0.5191654562950134,
+      "learning_rate": 9.543801475325306e-06,
+      "loss": 0.0383,
+      "step": 4756
+    },
+    {
+      "epoch": 2.8944326133252205,
+      "grad_norm": 0.47728797793388367,
+      "learning_rate": 9.534018803163794e-06,
+      "loss": 0.031,
+      "step": 4757
+    },
+    {
+      "epoch": 2.8950410708853056,
+      "grad_norm": 0.42437100410461426,
+      "learning_rate": 9.524239965920958e-06,
+      "loss": 0.0288,
+      "step": 4758
+    },
+    {
+      "epoch": 2.895649528445391,
+      "grad_norm": 0.46275144815444946,
+      "learning_rate": 9.514464966021547e-06,
+      "loss": 0.0391,
+      "step": 4759
+    },
+    {
+      "epoch": 2.8962579860054762,
+      "grad_norm": 0.5176787376403809,
+      "learning_rate": 9.504693805889355e-06,
+      "loss": 0.0275,
+      "step": 4760
+    },
+    {
+      "epoch": 2.8968664435655613,
+      "grad_norm": 0.39485782384872437,
+      "learning_rate": 9.494926487947233e-06,
+      "loss": 0.0326,
+      "step": 4761
+    },
+    {
+      "epoch": 2.8974749011256464,
+      "grad_norm": 1.0066802501678467,
+      "learning_rate": 9.485163014617043e-06,
+      "loss": 0.0387,
+      "step": 4762
+    },
+    {
+      "epoch": 2.8980833586857315,
+      "grad_norm": 0.38768336176872253,
+      "learning_rate": 9.475403388319752e-06,
+      "loss": 0.029,
+      "step": 4763
+    },
+    {
+      "epoch": 2.898691816245817,
+      "grad_norm": 0.3965464234352112,
+      "learning_rate": 9.465647611475312e-06,
+      "loss": 0.0288,
+      "step": 4764
+    },
+    {
+      "epoch": 2.899300273805902,
+      "grad_norm": 0.5666676163673401,
+      "learning_rate": 9.455895686502762e-06,
+      "loss": 0.0437,
+      "step": 4765
+    },
+    {
+      "epoch": 2.8999087313659873,
+      "grad_norm": 0.49036771059036255,
+      "learning_rate": 9.446147615820169e-06,
+      "loss": 0.0332,
+      "step": 4766
+    },
+    {
+      "epoch": 2.9005171889260724,
+      "grad_norm": 0.49707627296447754,
+      "learning_rate": 9.436403401844663e-06,
+      "loss": 0.0317,
+      "step": 4767
+    },
+    {
+      "epoch": 2.9011256464861574,
+      "grad_norm": 0.5176650881767273,
+      "learning_rate": 9.426663046992381e-06,
+      "loss": 0.0347,
+      "step": 4768
+    },
+    {
+      "epoch": 2.901734104046243,
+      "grad_norm": 0.536285936832428,
+      "learning_rate": 9.416926553678535e-06,
+      "loss": 0.0404,
+      "step": 4769
+    },
+    {
+      "epoch": 2.902342561606328,
+      "grad_norm": 0.5301802158355713,
+      "learning_rate": 9.40719392431737e-06,
+      "loss": 0.0418,
+      "step": 4770
+    },
+    {
+      "epoch": 2.902951019166413,
+      "grad_norm": 0.5603205561637878,
+      "learning_rate": 9.397465161322177e-06,
+      "loss": 0.0421,
+      "step": 4771
+    },
+    {
+      "epoch": 2.9035594767264983,
+      "grad_norm": 0.4085746109485626,
+      "learning_rate": 9.387740267105263e-06,
+      "loss": 0.0362,
+      "step": 4772
+    },
+    {
+      "epoch": 2.9041679342865834,
+      "grad_norm": 0.4638700485229492,
+      "learning_rate": 9.378019244078028e-06,
+      "loss": 0.0435,
+      "step": 4773
+    },
+    {
+      "epoch": 2.904776391846669,
+      "grad_norm": 0.5153172612190247,
+      "learning_rate": 9.368302094650858e-06,
+      "loss": 0.0505,
+      "step": 4774
+    },
+    {
+      "epoch": 2.905384849406754,
+      "grad_norm": 0.3367726802825928,
+      "learning_rate": 9.358588821233207e-06,
+      "loss": 0.0292,
+      "step": 4775
+    },
+    {
+      "epoch": 2.905993306966839,
+      "grad_norm": 0.4512867033481598,
+      "learning_rate": 9.348879426233561e-06,
+      "loss": 0.0333,
+      "step": 4776
+    },
+    {
+      "epoch": 2.906601764526924,
+      "grad_norm": 0.4421890676021576,
+      "learning_rate": 9.339173912059455e-06,
+      "loss": 0.0347,
+      "step": 4777
+    },
+    {
+      "epoch": 2.9072102220870093,
+      "grad_norm": 0.379062294960022,
+      "learning_rate": 9.329472281117434e-06,
+      "loss": 0.0256,
+      "step": 4778
+    },
+    {
+      "epoch": 2.907818679647095,
+      "grad_norm": 0.4036319851875305,
+      "learning_rate": 9.319774535813109e-06,
+      "loss": 0.0392,
+      "step": 4779
+    },
+    {
+      "epoch": 2.90842713720718,
+      "grad_norm": 0.47172126173973083,
+      "learning_rate": 9.310080678551114e-06,
+      "loss": 0.0408,
+      "step": 4780
+    },
+    {
+      "epoch": 2.909035594767265,
+      "grad_norm": 0.494945228099823,
+      "learning_rate": 9.300390711735132e-06,
+      "loss": 0.0257,
+      "step": 4781
+    },
+    {
+      "epoch": 2.90964405232735,
+      "grad_norm": 0.4051055610179901,
+      "learning_rate": 9.290704637767843e-06,
+      "loss": 0.0298,
+      "step": 4782
+    },
+    {
+      "epoch": 2.9102525098874352,
+      "grad_norm": 0.5722257494926453,
+      "learning_rate": 9.281022459051029e-06,
+      "loss": 0.0371,
+      "step": 4783
+    },
+    {
+      "epoch": 2.9108609674475208,
+      "grad_norm": 0.4758877754211426,
+      "learning_rate": 9.271344177985433e-06,
+      "loss": 0.045,
+      "step": 4784
+    },
+    {
+      "epoch": 2.911469425007606,
+      "grad_norm": 0.5374777913093567,
+      "learning_rate": 9.261669796970881e-06,
+      "loss": 0.0507,
+      "step": 4785
+    },
+    {
+      "epoch": 2.912077882567691,
+      "grad_norm": 0.35278376936912537,
+      "learning_rate": 9.251999318406213e-06,
+      "loss": 0.0239,
+      "step": 4786
+    },
+    {
+      "epoch": 2.912686340127776,
+      "grad_norm": 0.40819981694221497,
+      "learning_rate": 9.242332744689314e-06,
+      "loss": 0.026,
+      "step": 4787
+    },
+    {
+      "epoch": 2.913294797687861,
+      "grad_norm": 0.5325673222541809,
+      "learning_rate": 9.232670078217062e-06,
+      "loss": 0.0409,
+      "step": 4788
+    },
+    {
+      "epoch": 2.9139032552479467,
+      "grad_norm": 0.44101467728614807,
+      "learning_rate": 9.223011321385435e-06,
+      "loss": 0.0378,
+      "step": 4789
+    },
+    {
+      "epoch": 2.914511712808032,
+      "grad_norm": 0.47359776496887207,
+      "learning_rate": 9.213356476589375e-06,
+      "loss": 0.0324,
+      "step": 4790
+    },
+    {
+      "epoch": 2.915120170368117,
+      "grad_norm": 0.3451775908470154,
+      "learning_rate": 9.2037055462229e-06,
+      "loss": 0.0211,
+      "step": 4791
+    },
+    {
+      "epoch": 2.915728627928202,
+      "grad_norm": 0.3723897635936737,
+      "learning_rate": 9.194058532679006e-06,
+      "loss": 0.0358,
+      "step": 4792
+    },
+    {
+      "epoch": 2.916337085488287,
+      "grad_norm": 0.5562166571617126,
+      "learning_rate": 9.184415438349788e-06,
+      "loss": 0.0396,
+      "step": 4793
+    },
+    {
+      "epoch": 2.9169455430483726,
+      "grad_norm": 0.41815292835235596,
+      "learning_rate": 9.174776265626312e-06,
+      "loss": 0.0194,
+      "step": 4794
+    },
+    {
+      "epoch": 2.9175540006084577,
+      "grad_norm": 0.5843579173088074,
+      "learning_rate": 9.16514101689869e-06,
+      "loss": 0.0494,
+      "step": 4795
+    },
+    {
+      "epoch": 2.918162458168543,
+      "grad_norm": 0.42992332577705383,
+      "learning_rate": 9.15550969455607e-06,
+      "loss": 0.033,
+      "step": 4796
+    },
+    {
+      "epoch": 2.918770915728628,
+      "grad_norm": 0.4272845387458801,
+      "learning_rate": 9.14588230098662e-06,
+      "loss": 0.0286,
+      "step": 4797
+    },
+    {
+      "epoch": 2.919379373288713,
+      "grad_norm": 0.4098367691040039,
+      "learning_rate": 9.136258838577511e-06,
+      "loss": 0.0226,
+      "step": 4798
+    },
+    {
+      "epoch": 2.9199878308487985,
+      "grad_norm": 0.47262877225875854,
+      "learning_rate": 9.126639309714997e-06,
+      "loss": 0.0416,
+      "step": 4799
+    },
+    {
+      "epoch": 2.9205962884088836,
+      "grad_norm": 0.5105498433113098,
+      "learning_rate": 9.117023716784287e-06,
+      "loss": 0.043,
+      "step": 4800
+    },
+    {
+      "epoch": 2.9212047459689687,
+      "grad_norm": 0.4782066345214844,
+      "learning_rate": 9.10741206216967e-06,
+      "loss": 0.0292,
+      "step": 4801
+    },
+    {
+      "epoch": 2.921813203529054,
+      "grad_norm": 0.5072314143180847,
+      "learning_rate": 9.09780434825441e-06,
+      "loss": 0.0364,
+      "step": 4802
+    },
+    {
+      "epoch": 2.922421661089139,
+      "grad_norm": 0.46522530913352966,
+      "learning_rate": 9.08820057742085e-06,
+      "loss": 0.0315,
+      "step": 4803
+    },
+    {
+      "epoch": 2.9230301186492245,
+      "grad_norm": 0.5069283246994019,
+      "learning_rate": 9.078600752050304e-06,
+      "loss": 0.0451,
+      "step": 4804
+    },
+    {
+      "epoch": 2.9236385762093096,
+      "grad_norm": 0.3636358678340912,
+      "learning_rate": 9.069004874523136e-06,
+      "loss": 0.0215,
+      "step": 4805
+    },
+    {
+      "epoch": 2.9242470337693947,
+      "grad_norm": 0.43121346831321716,
+      "learning_rate": 9.059412947218718e-06,
+      "loss": 0.0287,
+      "step": 4806
+    },
+    {
+      "epoch": 2.9248554913294798,
+      "grad_norm": 0.35692986845970154,
+      "learning_rate": 9.049824972515464e-06,
+      "loss": 0.0224,
+      "step": 4807
+    },
+    {
+      "epoch": 2.925463948889565,
+      "grad_norm": 0.4458438456058502,
+      "learning_rate": 9.040240952790765e-06,
+      "loss": 0.0279,
+      "step": 4808
+    },
+    {
+      "epoch": 2.9260724064496504,
+      "grad_norm": 0.4637523591518402,
+      "learning_rate": 9.030660890421089e-06,
+      "loss": 0.0411,
+      "step": 4809
+    },
+    {
+      "epoch": 2.9266808640097355,
+      "grad_norm": 0.4921475648880005,
+      "learning_rate": 9.02108478778187e-06,
+      "loss": 0.0346,
+      "step": 4810
+    },
+    {
+      "epoch": 2.9272893215698206,
+      "grad_norm": 0.4898071587085724,
+      "learning_rate": 9.011512647247588e-06,
+      "loss": 0.0455,
+      "step": 4811
+    },
+    {
+      "epoch": 2.9278977791299057,
+      "grad_norm": 0.3920941650867462,
+      "learning_rate": 9.00194447119174e-06,
+      "loss": 0.0236,
+      "step": 4812
+    },
+    {
+      "epoch": 2.9285062366899908,
+      "grad_norm": 0.4192269444465637,
+      "learning_rate": 8.992380261986837e-06,
+      "loss": 0.0303,
+      "step": 4813
+    },
+    {
+      "epoch": 2.9291146942500763,
+      "grad_norm": 0.44664058089256287,
+      "learning_rate": 8.98282002200439e-06,
+      "loss": 0.0363,
+      "step": 4814
+    },
+    {
+      "epoch": 2.929723151810161,
+      "grad_norm": 0.5392205715179443,
+      "learning_rate": 8.973263753614949e-06,
+      "loss": 0.0398,
+      "step": 4815
+    },
+    {
+      "epoch": 2.9303316093702465,
+      "grad_norm": 0.43230360746383667,
+      "learning_rate": 8.96371145918807e-06,
+      "loss": 0.0439,
+      "step": 4816
+    },
+    {
+      "epoch": 2.9309400669303316,
+      "grad_norm": 0.3646244704723358,
+      "learning_rate": 8.954163141092333e-06,
+      "loss": 0.0249,
+      "step": 4817
+    },
+    {
+      "epoch": 2.9315485244904167,
+      "grad_norm": 0.5149335861206055,
+      "learning_rate": 8.944618801695295e-06,
+      "loss": 0.0292,
+      "step": 4818
+    },
+    {
+      "epoch": 2.9321569820505022,
+      "grad_norm": 0.3785509765148163,
+      "learning_rate": 8.935078443363592e-06,
+      "loss": 0.0376,
+      "step": 4819
+    },
+    {
+      "epoch": 2.932765439610587,
+      "grad_norm": 0.5124600529670715,
+      "learning_rate": 8.925542068462806e-06,
+      "loss": 0.0407,
+      "step": 4820
+    },
+    {
+      "epoch": 2.9333738971706724,
+      "grad_norm": 0.4565979838371277,
+      "learning_rate": 8.91600967935757e-06,
+      "loss": 0.0334,
+      "step": 4821
+    },
+    {
+      "epoch": 2.9339823547307575,
+      "grad_norm": 0.5674571394920349,
+      "learning_rate": 8.906481278411522e-06,
+      "loss": 0.0486,
+      "step": 4822
+    },
+    {
+      "epoch": 2.9345908122908426,
+      "grad_norm": 0.5254626870155334,
+      "learning_rate": 8.896956867987313e-06,
+      "loss": 0.0281,
+      "step": 4823
+    },
+    {
+      "epoch": 2.935199269850928,
+      "grad_norm": 0.380880743265152,
+      "learning_rate": 8.887436450446584e-06,
+      "loss": 0.0277,
+      "step": 4824
+    },
+    {
+      "epoch": 2.935807727411013,
+      "grad_norm": 0.41296547651290894,
+      "learning_rate": 8.877920028150013e-06,
+      "loss": 0.028,
+      "step": 4825
+    },
+    {
+      "epoch": 2.9364161849710984,
+      "grad_norm": 0.5368117094039917,
+      "learning_rate": 8.868407603457272e-06,
+      "loss": 0.0386,
+      "step": 4826
+    },
+    {
+      "epoch": 2.9370246425311834,
+      "grad_norm": 0.5776300430297852,
+      "learning_rate": 8.858899178727045e-06,
+      "loss": 0.0416,
+      "step": 4827
+    },
+    {
+      "epoch": 2.9376331000912685,
+      "grad_norm": 0.38237324357032776,
+      "learning_rate": 8.84939475631703e-06,
+      "loss": 0.0324,
+      "step": 4828
+    },
+    {
+      "epoch": 2.938241557651354,
+      "grad_norm": 0.42977848649024963,
+      "learning_rate": 8.839894338583935e-06,
+      "loss": 0.0297,
+      "step": 4829
+    },
+    {
+      "epoch": 2.9388500152114387,
+      "grad_norm": 0.5375818610191345,
+      "learning_rate": 8.830397927883446e-06,
+      "loss": 0.0374,
+      "step": 4830
+    },
+    {
+      "epoch": 2.9394584727715243,
+      "grad_norm": 0.6935356259346008,
+      "learning_rate": 8.820905526570289e-06,
+      "loss": 0.0601,
+      "step": 4831
+    },
+    {
+      "epoch": 2.9400669303316094,
+      "grad_norm": 0.5644581317901611,
+      "learning_rate": 8.811417136998184e-06,
+      "loss": 0.0488,
+      "step": 4832
+    },
+    {
+      "epoch": 2.9406753878916945,
+      "grad_norm": 0.4846365451812744,
+      "learning_rate": 8.801932761519857e-06,
+      "loss": 0.0472,
+      "step": 4833
+    },
+    {
+      "epoch": 2.94128384545178,
+      "grad_norm": 0.4495841860771179,
+      "learning_rate": 8.792452402487037e-06,
+      "loss": 0.033,
+      "step": 4834
+    },
+    {
+      "epoch": 2.9418923030118647,
+      "grad_norm": 0.42976799607276917,
+      "learning_rate": 8.782976062250464e-06,
+      "loss": 0.0334,
+      "step": 4835
+    },
+    {
+      "epoch": 2.94250076057195,
+      "grad_norm": 0.5387213230133057,
+      "learning_rate": 8.77350374315986e-06,
+      "loss": 0.0398,
+      "step": 4836
+    },
+    {
+      "epoch": 2.9431092181320353,
+      "grad_norm": 0.46354934573173523,
+      "learning_rate": 8.764035447563976e-06,
+      "loss": 0.0363,
+      "step": 4837
+    },
+    {
+      "epoch": 2.9437176756921204,
+      "grad_norm": 0.471663236618042,
+      "learning_rate": 8.75457117781055e-06,
+      "loss": 0.0491,
+      "step": 4838
+    },
+    {
+      "epoch": 2.944326133252206,
+      "grad_norm": 0.584274172782898,
+      "learning_rate": 8.745110936246331e-06,
+      "loss": 0.0437,
+      "step": 4839
+    },
+    {
+      "epoch": 2.9449345908122906,
+      "grad_norm": 0.5374387502670288,
+      "learning_rate": 8.73565472521707e-06,
+      "loss": 0.0475,
+      "step": 4840
+    },
+    {
+      "epoch": 2.945543048372376,
+      "grad_norm": 0.46591848134994507,
+      "learning_rate": 8.726202547067496e-06,
+      "loss": 0.0427,
+      "step": 4841
+    },
+    {
+      "epoch": 2.9461515059324612,
+      "grad_norm": 0.3446124792098999,
+      "learning_rate": 8.716754404141368e-06,
+      "loss": 0.0233,
+      "step": 4842
+    },
+    {
+      "epoch": 2.9467599634925463,
+      "grad_norm": 0.43566611409187317,
+      "learning_rate": 8.707310298781427e-06,
+      "loss": 0.0277,
+      "step": 4843
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.38527342677116394,
+      "learning_rate": 8.69787023332942e-06,
+      "loss": 0.0224,
+      "step": 4844
+    },
+    {
+      "epoch": 2.9479768786127165,
+      "grad_norm": 0.36706897616386414,
+      "learning_rate": 8.688434210126099e-06,
+      "loss": 0.0239,
+      "step": 4845
+    },
+    {
+      "epoch": 2.948585336172802,
+      "grad_norm": 0.43865934014320374,
+      "learning_rate": 8.679002231511182e-06,
+      "loss": 0.0322,
+      "step": 4846
+    },
+    {
+      "epoch": 2.949193793732887,
+      "grad_norm": 0.5478273034095764,
+      "learning_rate": 8.669574299823422e-06,
+      "loss": 0.0374,
+      "step": 4847
+    },
+    {
+      "epoch": 2.9498022512929722,
+      "grad_norm": 0.4504847824573517,
+      "learning_rate": 8.660150417400547e-06,
+      "loss": 0.0363,
+      "step": 4848
+    },
+    {
+      "epoch": 2.9504107088530573,
+      "grad_norm": 0.4088916778564453,
+      "learning_rate": 8.650730586579292e-06,
+      "loss": 0.036,
+      "step": 4849
+    },
+    {
+      "epoch": 2.9510191664131424,
+      "grad_norm": 0.4998932480812073,
+      "learning_rate": 8.641314809695389e-06,
+      "loss": 0.0482,
+      "step": 4850
+    },
+    {
+      "epoch": 2.951627623973228,
+      "grad_norm": 0.506369411945343,
+      "learning_rate": 8.63190308908354e-06,
+      "loss": 0.0313,
+      "step": 4851
+    },
+    {
+      "epoch": 2.952236081533313,
+      "grad_norm": 0.4469977915287018,
+      "learning_rate": 8.622495427077468e-06,
+      "loss": 0.0239,
+      "step": 4852
+    },
+    {
+      "epoch": 2.952844539093398,
+      "grad_norm": 0.4590071439743042,
+      "learning_rate": 8.613091826009884e-06,
+      "loss": 0.0423,
+      "step": 4853
+    },
+    {
+      "epoch": 2.9534529966534833,
+      "grad_norm": 0.614018440246582,
+      "learning_rate": 8.603692288212486e-06,
+      "loss": 0.0576,
+      "step": 4854
+    },
+    {
+      "epoch": 2.9540614542135684,
+      "grad_norm": 0.45783132314682007,
+      "learning_rate": 8.59429681601597e-06,
+      "loss": 0.0325,
+      "step": 4855
+    },
+    {
+      "epoch": 2.954669911773654,
+      "grad_norm": 0.4062897562980652,
+      "learning_rate": 8.58490541175003e-06,
+      "loss": 0.0286,
+      "step": 4856
+    },
+    {
+      "epoch": 2.955278369333739,
+      "grad_norm": 0.561907947063446,
+      "learning_rate": 8.575518077743322e-06,
+      "loss": 0.0356,
+      "step": 4857
+    },
+    {
+      "epoch": 2.955886826893824,
+      "grad_norm": 0.42090898752212524,
+      "learning_rate": 8.566134816323526e-06,
+      "loss": 0.0326,
+      "step": 4858
+    },
+    {
+      "epoch": 2.956495284453909,
+      "grad_norm": 0.45067688822746277,
+      "learning_rate": 8.556755629817295e-06,
+      "loss": 0.0378,
+      "step": 4859
+    },
+    {
+      "epoch": 2.9571037420139943,
+      "grad_norm": 0.4070775806903839,
+      "learning_rate": 8.54738052055029e-06,
+      "loss": 0.0318,
+      "step": 4860
+    },
+    {
+      "epoch": 2.95771219957408,
+      "grad_norm": 0.40776872634887695,
+      "learning_rate": 8.538009490847124e-06,
+      "loss": 0.0308,
+      "step": 4861
+    },
+    {
+      "epoch": 2.958320657134165,
+      "grad_norm": 0.4585787355899811,
+      "learning_rate": 8.528642543031448e-06,
+      "loss": 0.0288,
+      "step": 4862
+    },
+    {
+      "epoch": 2.95892911469425,
+      "grad_norm": 0.4308152496814728,
+      "learning_rate": 8.519279679425851e-06,
+      "loss": 0.0528,
+      "step": 4863
+    },
+    {
+      "epoch": 2.959537572254335,
+      "grad_norm": 0.3569679856300354,
+      "learning_rate": 8.509920902351948e-06,
+      "loss": 0.0242,
+      "step": 4864
+    },
+    {
+      "epoch": 2.96014602981442,
+      "grad_norm": 0.38262954354286194,
+      "learning_rate": 8.500566214130317e-06,
+      "loss": 0.0322,
+      "step": 4865
+    },
+    {
+      "epoch": 2.9607544873745058,
+      "grad_norm": 0.4387529492378235,
+      "learning_rate": 8.491215617080545e-06,
+      "loss": 0.0275,
+      "step": 4866
+    },
+    {
+      "epoch": 2.961362944934591,
+      "grad_norm": 0.41847920417785645,
+      "learning_rate": 8.481869113521163e-06,
+      "loss": 0.0294,
+      "step": 4867
+    },
+    {
+      "epoch": 2.961971402494676,
+      "grad_norm": 0.41633474826812744,
+      "learning_rate": 8.472526705769746e-06,
+      "loss": 0.0315,
+      "step": 4868
+    },
+    {
+      "epoch": 2.962579860054761,
+      "grad_norm": 0.5311760902404785,
+      "learning_rate": 8.463188396142799e-06,
+      "loss": 0.0422,
+      "step": 4869
+    },
+    {
+      "epoch": 2.963188317614846,
+      "grad_norm": 0.4289592504501343,
+      "learning_rate": 8.453854186955851e-06,
+      "loss": 0.0289,
+      "step": 4870
+    },
+    {
+      "epoch": 2.9637967751749317,
+      "grad_norm": 0.6336865425109863,
+      "learning_rate": 8.444524080523373e-06,
+      "loss": 0.0491,
+      "step": 4871
+    },
+    {
+      "epoch": 2.9644052327350168,
+      "grad_norm": 0.3597281873226166,
+      "learning_rate": 8.435198079158867e-06,
+      "loss": 0.0251,
+      "step": 4872
+    },
+    {
+      "epoch": 2.965013690295102,
+      "grad_norm": 0.5553586483001709,
+      "learning_rate": 8.42587618517478e-06,
+      "loss": 0.0345,
+      "step": 4873
+    },
+    {
+      "epoch": 2.965622147855187,
+      "grad_norm": 0.5425621271133423,
+      "learning_rate": 8.416558400882552e-06,
+      "loss": 0.0432,
+      "step": 4874
+    },
+    {
+      "epoch": 2.966230605415272,
+      "grad_norm": 0.3684383034706116,
+      "learning_rate": 8.407244728592612e-06,
+      "loss": 0.0214,
+      "step": 4875
+    },
+    {
+      "epoch": 2.9668390629753576,
+      "grad_norm": 0.42099565267562866,
+      "learning_rate": 8.397935170614366e-06,
+      "loss": 0.0292,
+      "step": 4876
+    },
+    {
+      "epoch": 2.9674475205354427,
+      "grad_norm": 0.48442554473876953,
+      "learning_rate": 8.38862972925617e-06,
+      "loss": 0.0449,
+      "step": 4877
+    },
+    {
+      "epoch": 2.968055978095528,
+      "grad_norm": 0.491372287273407,
+      "learning_rate": 8.379328406825426e-06,
+      "loss": 0.0444,
+      "step": 4878
+    },
+    {
+      "epoch": 2.968664435655613,
+      "grad_norm": 0.4108073115348816,
+      "learning_rate": 8.370031205628442e-06,
+      "loss": 0.0385,
+      "step": 4879
+    },
+    {
+      "epoch": 2.969272893215698,
+      "grad_norm": 0.4929574429988861,
+      "learning_rate": 8.36073812797055e-06,
+      "loss": 0.0421,
+      "step": 4880
+    },
+    {
+      "epoch": 2.9698813507757835,
+      "grad_norm": 0.4480864405632019,
+      "learning_rate": 8.351449176156043e-06,
+      "loss": 0.0321,
+      "step": 4881
+    },
+    {
+      "epoch": 2.9704898083358686,
+      "grad_norm": 0.39888888597488403,
+      "learning_rate": 8.342164352488202e-06,
+      "loss": 0.0195,
+      "step": 4882
+    },
+    {
+      "epoch": 2.9710982658959537,
+      "grad_norm": 0.475404292345047,
+      "learning_rate": 8.332883659269261e-06,
+      "loss": 0.031,
+      "step": 4883
+    },
+    {
+      "epoch": 2.971706723456039,
+      "grad_norm": 0.509152352809906,
+      "learning_rate": 8.323607098800454e-06,
+      "loss": 0.0327,
+      "step": 4884
+    },
+    {
+      "epoch": 2.972315181016124,
+      "grad_norm": 0.4782547652721405,
+      "learning_rate": 8.314334673381976e-06,
+      "loss": 0.0375,
+      "step": 4885
+    },
+    {
+      "epoch": 2.9729236385762094,
+      "grad_norm": 0.3744230568408966,
+      "learning_rate": 8.305066385313021e-06,
+      "loss": 0.0262,
+      "step": 4886
+    },
+    {
+      "epoch": 2.9735320961362945,
+      "grad_norm": 0.5393217206001282,
+      "learning_rate": 8.2958022368917e-06,
+      "loss": 0.0503,
+      "step": 4887
+    },
+    {
+      "epoch": 2.9741405536963796,
+      "grad_norm": 0.44529324769973755,
+      "learning_rate": 8.286542230415182e-06,
+      "loss": 0.0344,
+      "step": 4888
+    },
+    {
+      "epoch": 2.9747490112564647,
+      "grad_norm": 0.46004465222358704,
+      "learning_rate": 8.277286368179526e-06,
+      "loss": 0.0374,
+      "step": 4889
+    },
+    {
+      "epoch": 2.97535746881655,
+      "grad_norm": 0.46514490246772766,
+      "learning_rate": 8.268034652479817e-06,
+      "loss": 0.0374,
+      "step": 4890
+    },
+    {
+      "epoch": 2.9759659263766354,
+      "grad_norm": 0.4334418475627899,
+      "learning_rate": 8.25878708561009e-06,
+      "loss": 0.031,
+      "step": 4891
+    },
+    {
+      "epoch": 2.9765743839367205,
+      "grad_norm": 0.3778439164161682,
+      "learning_rate": 8.249543669863366e-06,
+      "loss": 0.0354,
+      "step": 4892
+    },
+    {
+      "epoch": 2.9771828414968056,
+      "grad_norm": 0.4430623948574066,
+      "learning_rate": 8.240304407531602e-06,
+      "loss": 0.0393,
+      "step": 4893
+    },
+    {
+      "epoch": 2.9777912990568907,
+      "grad_norm": 0.3321574628353119,
+      "learning_rate": 8.231069300905783e-06,
+      "loss": 0.0243,
+      "step": 4894
+    },
+    {
+      "epoch": 2.9783997566169758,
+      "grad_norm": 0.6121447086334229,
+      "learning_rate": 8.221838352275807e-06,
+      "loss": 0.0376,
+      "step": 4895
+    },
+    {
+      "epoch": 2.9790082141770613,
+      "grad_norm": 0.48445889353752136,
+      "learning_rate": 8.212611563930577e-06,
+      "loss": 0.0425,
+      "step": 4896
+    },
+    {
+      "epoch": 2.9796166717371464,
+      "grad_norm": 0.43182381987571716,
+      "learning_rate": 8.20338893815793e-06,
+      "loss": 0.0318,
+      "step": 4897
+    },
+    {
+      "epoch": 2.9802251292972315,
+      "grad_norm": 0.4070221185684204,
+      "learning_rate": 8.194170477244729e-06,
+      "loss": 0.0291,
+      "step": 4898
+    },
+    {
+      "epoch": 2.9808335868573166,
+      "grad_norm": 0.4428864121437073,
+      "learning_rate": 8.184956183476735e-06,
+      "loss": 0.0336,
+      "step": 4899
+    },
+    {
+      "epoch": 2.9814420444174017,
+      "grad_norm": 0.4261319637298584,
+      "learning_rate": 8.175746059138725e-06,
+      "loss": 0.03,
+      "step": 4900
+    },
+    {
+      "epoch": 2.9820505019774872,
+      "grad_norm": 0.45288318395614624,
+      "learning_rate": 8.166540106514422e-06,
+      "loss": 0.0374,
+      "step": 4901
+    },
+    {
+      "epoch": 2.9826589595375723,
+      "grad_norm": 0.4366372227668762,
+      "learning_rate": 8.15733832788653e-06,
+      "loss": 0.037,
+      "step": 4902
+    },
+    {
+      "epoch": 2.9832674170976574,
+      "grad_norm": 0.5225448608398438,
+      "learning_rate": 8.148140725536676e-06,
+      "loss": 0.0371,
+      "step": 4903
+    },
+    {
+      "epoch": 2.9838758746577425,
+      "grad_norm": 0.47045180201530457,
+      "learning_rate": 8.138947301745519e-06,
+      "loss": 0.0357,
+      "step": 4904
+    },
+    {
+      "epoch": 2.9844843322178276,
+      "grad_norm": 0.5777865648269653,
+      "learning_rate": 8.129758058792623e-06,
+      "loss": 0.0439,
+      "step": 4905
+    },
+    {
+      "epoch": 2.985092789777913,
+      "grad_norm": 0.5593705773353577,
+      "learning_rate": 8.120572998956546e-06,
+      "loss": 0.0355,
+      "step": 4906
+    },
+    {
+      "epoch": 2.9857012473379982,
+      "grad_norm": 0.41948768496513367,
+      "learning_rate": 8.111392124514783e-06,
+      "loss": 0.0446,
+      "step": 4907
+    },
+    {
+      "epoch": 2.9863097048980833,
+      "grad_norm": 0.44301876425743103,
+      "learning_rate": 8.102215437743835e-06,
+      "loss": 0.0342,
+      "step": 4908
+    },
+    {
+      "epoch": 2.9869181624581684,
+      "grad_norm": 0.4251943826675415,
+      "learning_rate": 8.093042940919118e-06,
+      "loss": 0.0345,
+      "step": 4909
+    },
+    {
+      "epoch": 2.9875266200182535,
+      "grad_norm": 0.28501009941101074,
+      "learning_rate": 8.083874636315034e-06,
+      "loss": 0.0155,
+      "step": 4910
+    },
+    {
+      "epoch": 2.988135077578339,
+      "grad_norm": 0.4958946108818054,
+      "learning_rate": 8.07471052620494e-06,
+      "loss": 0.0373,
+      "step": 4911
+    },
+    {
+      "epoch": 2.988743535138424,
+      "grad_norm": 0.48140478134155273,
+      "learning_rate": 8.065550612861164e-06,
+      "loss": 0.0397,
+      "step": 4912
+    },
+    {
+      "epoch": 2.9893519926985093,
+      "grad_norm": 0.556651771068573,
+      "learning_rate": 8.05639489855496e-06,
+      "loss": 0.0472,
+      "step": 4913
+    },
+    {
+      "epoch": 2.9899604502585944,
+      "grad_norm": 0.4993399381637573,
+      "learning_rate": 8.047243385556588e-06,
+      "loss": 0.0437,
+      "step": 4914
+    },
+    {
+      "epoch": 2.9905689078186795,
+      "grad_norm": 0.33873289823532104,
+      "learning_rate": 8.038096076135227e-06,
+      "loss": 0.0172,
+      "step": 4915
+    },
+    {
+      "epoch": 2.991177365378765,
+      "grad_norm": 0.624120831489563,
+      "learning_rate": 8.028952972559028e-06,
+      "loss": 0.0467,
+      "step": 4916
+    },
+    {
+      "epoch": 2.99178582293885,
+      "grad_norm": 0.4880430996417999,
+      "learning_rate": 8.019814077095107e-06,
+      "loss": 0.0364,
+      "step": 4917
+    },
+    {
+      "epoch": 2.992394280498935,
+      "grad_norm": 0.40606868267059326,
+      "learning_rate": 8.010679392009532e-06,
+      "loss": 0.0334,
+      "step": 4918
+    },
+    {
+      "epoch": 2.9930027380590203,
+      "grad_norm": 0.6444923281669617,
+      "learning_rate": 8.001548919567312e-06,
+      "loss": 0.0422,
+      "step": 4919
+    },
+    {
+      "epoch": 2.9936111956191054,
+      "grad_norm": 0.45829200744628906,
+      "learning_rate": 7.992422662032429e-06,
+      "loss": 0.0284,
+      "step": 4920
+    },
+    {
+      "epoch": 2.994219653179191,
+      "grad_norm": 0.38635167479515076,
+      "learning_rate": 7.983300621667814e-06,
+      "loss": 0.0256,
+      "step": 4921
+    },
+    {
+      "epoch": 2.994828110739276,
+      "grad_norm": 0.47939708828926086,
+      "learning_rate": 7.974182800735361e-06,
+      "loss": 0.041,
+      "step": 4922
+    },
+    {
+      "epoch": 2.995436568299361,
+      "grad_norm": 0.343191534280777,
+      "learning_rate": 7.965069201495887e-06,
+      "loss": 0.0212,
+      "step": 4923
+    },
+    {
+      "epoch": 2.996045025859446,
+      "grad_norm": 0.40407371520996094,
+      "learning_rate": 7.955959826209217e-06,
+      "loss": 0.0233,
+      "step": 4924
+    },
+    {
+      "epoch": 2.9966534834195313,
+      "grad_norm": 0.5077594518661499,
+      "learning_rate": 7.946854677134072e-06,
+      "loss": 0.0448,
+      "step": 4925
+    },
+    {
+      "epoch": 2.997261940979617,
+      "grad_norm": 0.41432538628578186,
+      "learning_rate": 7.937753756528155e-06,
+      "loss": 0.0365,
+      "step": 4926
+    },
+    {
+      "epoch": 2.997870398539702,
+      "grad_norm": 0.3649181127548218,
+      "learning_rate": 7.928657066648118e-06,
+      "loss": 0.0347,
+      "step": 4927
+    },
+    {
+      "epoch": 2.998478856099787,
+      "grad_norm": 0.4499366283416748,
+      "learning_rate": 7.919564609749568e-06,
+      "loss": 0.0343,
+      "step": 4928
+    },
+    {
+      "epoch": 2.999087313659872,
+      "grad_norm": 0.4373056888580322,
+      "learning_rate": 7.910476388087038e-06,
+      "loss": 0.035,
+      "step": 4929
+    },
+    {
+      "epoch": 2.9996957712199572,
+      "grad_norm": 0.3516384959220886,
+      "learning_rate": 7.90139240391404e-06,
+      "loss": 0.0229,
+      "step": 4930
+    },
+    {
+      "epoch": 2.9996957712199572,
+      "eval_loss": 1.429209589958191,
+      "eval_runtime": 105.3655,
+      "eval_samples_per_second": 7.232,
+      "eval_steps_per_second": 0.456,
+      "step": 4930
+    },
+    {
+      "epoch": 3.0003042287800428,
+      "grad_norm": 0.4027365744113922,
+      "learning_rate": 7.892312659483017e-06,
+      "loss": 0.0212,
+      "step": 4931
+    },
+    {
+      "epoch": 3.000912686340128,
+      "grad_norm": 0.18064260482788086,
+      "learning_rate": 7.88323715704538e-06,
+      "loss": 0.0077,
+      "step": 4932
+    },
+    {
+      "epoch": 3.001521143900213,
+      "grad_norm": 0.20326943695545197,
+      "learning_rate": 7.874165898851454e-06,
+      "loss": 0.0088,
+      "step": 4933
+    },
+    {
+      "epoch": 3.002129601460298,
+      "grad_norm": 0.21207283437252045,
+      "learning_rate": 7.865098887150557e-06,
+      "loss": 0.0098,
+      "step": 4934
+    },
+    {
+      "epoch": 3.002738059020383,
+      "grad_norm": 0.25013598799705505,
+      "learning_rate": 7.856036124190913e-06,
+      "loss": 0.0099,
+      "step": 4935
+    },
+    {
+      "epoch": 3.0033465165804687,
+      "grad_norm": 0.16713546216487885,
+      "learning_rate": 7.84697761221971e-06,
+      "loss": 0.0071,
+      "step": 4936
+    },
+    {
+      "epoch": 3.003954974140554,
+      "grad_norm": 0.20086368918418884,
+      "learning_rate": 7.83792335348309e-06,
+      "loss": 0.0092,
+      "step": 4937
+    },
+    {
+      "epoch": 3.004563431700639,
+      "grad_norm": 0.19693416357040405,
+      "learning_rate": 7.828873350226135e-06,
+      "loss": 0.0084,
+      "step": 4938
+    },
+    {
+      "epoch": 3.005171889260724,
+      "grad_norm": 0.22012776136398315,
+      "learning_rate": 7.819827604692845e-06,
+      "loss": 0.0081,
+      "step": 4939
+    },
+    {
+      "epoch": 3.005780346820809,
+      "grad_norm": 0.15234141051769257,
+      "learning_rate": 7.81078611912622e-06,
+      "loss": 0.0062,
+      "step": 4940
+    },
+    {
+      "epoch": 3.0063888043808946,
+      "grad_norm": 0.23004913330078125,
+      "learning_rate": 7.801748895768148e-06,
+      "loss": 0.0081,
+      "step": 4941
+    },
+    {
+      "epoch": 3.0069972619409797,
+      "grad_norm": 0.32940471172332764,
+      "learning_rate": 7.79271593685949e-06,
+      "loss": 0.0068,
+      "step": 4942
+    },
+    {
+      "epoch": 3.007605719501065,
+      "grad_norm": 0.19095928966999054,
+      "learning_rate": 7.783687244640048e-06,
+      "loss": 0.0077,
+      "step": 4943
+    },
+    {
+      "epoch": 3.00821417706115,
+      "grad_norm": 0.25689762830734253,
+      "learning_rate": 7.774662821348563e-06,
+      "loss": 0.0107,
+      "step": 4944
+    },
+    {
+      "epoch": 3.008822634621235,
+      "grad_norm": 0.22656317055225372,
+      "learning_rate": 7.765642669222706e-06,
+      "loss": 0.0056,
+      "step": 4945
+    },
+    {
+      "epoch": 3.0094310921813205,
+      "grad_norm": 0.20489272475242615,
+      "learning_rate": 7.7566267904991e-06,
+      "loss": 0.0086,
+      "step": 4946
+    },
+    {
+      "epoch": 3.0100395497414056,
+      "grad_norm": 0.24489440023899078,
+      "learning_rate": 7.747615187413312e-06,
+      "loss": 0.009,
+      "step": 4947
+    },
+    {
+      "epoch": 3.0106480073014907,
+      "grad_norm": 0.32784104347229004,
+      "learning_rate": 7.738607862199851e-06,
+      "loss": 0.0107,
+      "step": 4948
+    },
+    {
+      "epoch": 3.011256464861576,
+      "grad_norm": 0.25375160574913025,
+      "learning_rate": 7.729604817092134e-06,
+      "loss": 0.0066,
+      "step": 4949
+    },
+    {
+      "epoch": 3.011864922421661,
+      "grad_norm": 0.33918920159339905,
+      "learning_rate": 7.720606054322574e-06,
+      "loss": 0.0123,
+      "step": 4950
+    },
+    {
+      "epoch": 3.0124733799817465,
+      "grad_norm": 0.28510206937789917,
+      "learning_rate": 7.71161157612246e-06,
+      "loss": 0.0111,
+      "step": 4951
+    },
+    {
+      "epoch": 3.0130818375418316,
+      "grad_norm": 0.18262362480163574,
+      "learning_rate": 7.702621384722064e-06,
+      "loss": 0.0048,
+      "step": 4952
+    },
+    {
+      "epoch": 3.0136902951019167,
+      "grad_norm": 0.3342544138431549,
+      "learning_rate": 7.693635482350572e-06,
+      "loss": 0.013,
+      "step": 4953
+    },
+    {
+      "epoch": 3.0142987526620018,
+      "grad_norm": 0.2429402768611908,
+      "learning_rate": 7.684653871236125e-06,
+      "loss": 0.0059,
+      "step": 4954
+    },
+    {
+      "epoch": 3.014907210222087,
+      "grad_norm": 0.30693817138671875,
+      "learning_rate": 7.675676553605768e-06,
+      "loss": 0.007,
+      "step": 4955
+    },
+    {
+      "epoch": 3.0155156677821724,
+      "grad_norm": 0.27962803840637207,
+      "learning_rate": 7.666703531685516e-06,
+      "loss": 0.0086,
+      "step": 4956
+    },
+    {
+      "epoch": 3.0161241253422575,
+      "grad_norm": 0.5548810958862305,
+      "learning_rate": 7.657734807700297e-06,
+      "loss": 0.0059,
+      "step": 4957
+    },
+    {
+      "epoch": 3.0167325829023426,
+      "grad_norm": 0.20625406503677368,
+      "learning_rate": 7.648770383873988e-06,
+      "loss": 0.0054,
+      "step": 4958
+    },
+    {
+      "epoch": 3.0173410404624277,
+      "grad_norm": 0.19295968115329742,
+      "learning_rate": 7.639810262429386e-06,
+      "loss": 0.0042,
+      "step": 4959
+    },
+    {
+      "epoch": 3.0179494980225128,
+      "grad_norm": 0.231733500957489,
+      "learning_rate": 7.630854445588239e-06,
+      "loss": 0.0065,
+      "step": 4960
+    },
+    {
+      "epoch": 3.0185579555825983,
+      "grad_norm": 0.42650073766708374,
+      "learning_rate": 7.621902935571201e-06,
+      "loss": 0.0063,
+      "step": 4961
+    },
+    {
+      "epoch": 3.0191664131426834,
+      "grad_norm": 0.21923278272151947,
+      "learning_rate": 7.612955734597879e-06,
+      "loss": 0.0045,
+      "step": 4962
+    },
+    {
+      "epoch": 3.0197748707027685,
+      "grad_norm": 0.3130083978176117,
+      "learning_rate": 7.6040128448868096e-06,
+      "loss": 0.0065,
+      "step": 4963
+    },
+    {
+      "epoch": 3.0203833282628536,
+      "grad_norm": 0.2901938557624817,
+      "learning_rate": 7.5950742686554525e-06,
+      "loss": 0.0072,
+      "step": 4964
+    },
+    {
+      "epoch": 3.0209917858229387,
+      "grad_norm": 0.2919492721557617,
+      "learning_rate": 7.58614000812021e-06,
+      "loss": 0.0072,
+      "step": 4965
+    },
+    {
+      "epoch": 3.0216002433830242,
+      "grad_norm": 0.2188805788755417,
+      "learning_rate": 7.577210065496396e-06,
+      "loss": 0.0032,
+      "step": 4966
+    },
+    {
+      "epoch": 3.0222087009431093,
+      "grad_norm": 0.24292194843292236,
+      "learning_rate": 7.568284442998263e-06,
+      "loss": 0.0065,
+      "step": 4967
+    },
+    {
+      "epoch": 3.0228171585031944,
+      "grad_norm": 0.26626601815223694,
+      "learning_rate": 7.559363142839002e-06,
+      "loss": 0.011,
+      "step": 4968
+    },
+    {
+      "epoch": 3.0234256160632795,
+      "grad_norm": 0.2926938235759735,
+      "learning_rate": 7.550446167230718e-06,
+      "loss": 0.0076,
+      "step": 4969
+    },
+    {
+      "epoch": 3.0240340736233646,
+      "grad_norm": 0.26158931851387024,
+      "learning_rate": 7.5415335183844525e-06,
+      "loss": 0.0093,
+      "step": 4970
+    },
+    {
+      "epoch": 3.02464253118345,
+      "grad_norm": 0.2527156472206116,
+      "learning_rate": 7.53262519851018e-06,
+      "loss": 0.0066,
+      "step": 4971
+    },
+    {
+      "epoch": 3.0252509887435353,
+      "grad_norm": 0.21547862887382507,
+      "learning_rate": 7.52372120981677e-06,
+      "loss": 0.0039,
+      "step": 4972
+    },
+    {
+      "epoch": 3.0258594463036204,
+      "grad_norm": 0.1877271980047226,
+      "learning_rate": 7.514821554512056e-06,
+      "loss": 0.0046,
+      "step": 4973
+    },
+    {
+      "epoch": 3.0264679038637055,
+      "grad_norm": 0.31582483649253845,
+      "learning_rate": 7.505926234802777e-06,
+      "loss": 0.0115,
+      "step": 4974
+    },
+    {
+      "epoch": 3.0270763614237906,
+      "grad_norm": 0.2469736635684967,
+      "learning_rate": 7.49703525289461e-06,
+      "loss": 0.0073,
+      "step": 4975
+    },
+    {
+      "epoch": 3.027684818983876,
+      "grad_norm": 0.2651013731956482,
+      "learning_rate": 7.488148610992127e-06,
+      "loss": 0.0059,
+      "step": 4976
+    },
+    {
+      "epoch": 3.028293276543961,
+      "grad_norm": 0.3526837229728699,
+      "learning_rate": 7.479266311298871e-06,
+      "loss": 0.0103,
+      "step": 4977
+    },
+    {
+      "epoch": 3.0289017341040463,
+      "grad_norm": 0.4271092712879181,
+      "learning_rate": 7.470388356017266e-06,
+      "loss": 0.0092,
+      "step": 4978
+    },
+    {
+      "epoch": 3.0295101916641314,
+      "grad_norm": 0.1828882098197937,
+      "learning_rate": 7.461514747348674e-06,
+      "loss": 0.0029,
+      "step": 4979
+    },
+    {
+      "epoch": 3.0301186492242165,
+      "grad_norm": 0.2836543917655945,
+      "learning_rate": 7.452645487493387e-06,
+      "loss": 0.008,
+      "step": 4980
+    },
+    {
+      "epoch": 3.030727106784302,
+      "grad_norm": 0.5397449731826782,
+      "learning_rate": 7.4437805786506164e-06,
+      "loss": 0.0153,
+      "step": 4981
+    },
+    {
+      "epoch": 3.031335564344387,
+      "grad_norm": 0.24427630007266998,
+      "learning_rate": 7.434920023018476e-06,
+      "loss": 0.004,
+      "step": 4982
+    },
+    {
+      "epoch": 3.031944021904472,
+      "grad_norm": 0.320932000875473,
+      "learning_rate": 7.426063822794022e-06,
+      "loss": 0.0092,
+      "step": 4983
+    },
+    {
+      "epoch": 3.0325524794645573,
+      "grad_norm": 0.280251145362854,
+      "learning_rate": 7.417211980173222e-06,
+      "loss": 0.0093,
+      "step": 4984
+    },
+    {
+      "epoch": 3.0331609370246424,
+      "grad_norm": 0.19520291686058044,
+      "learning_rate": 7.408364497350964e-06,
+      "loss": 0.0055,
+      "step": 4985
+    },
+    {
+      "epoch": 3.0337693945847275,
+      "grad_norm": 0.28560855984687805,
+      "learning_rate": 7.3995213765210545e-06,
+      "loss": 0.0072,
+      "step": 4986
+    },
+    {
+      "epoch": 3.034377852144813,
+      "grad_norm": 0.226197749376297,
+      "learning_rate": 7.390682619876227e-06,
+      "loss": 0.0038,
+      "step": 4987
+    },
+    {
+      "epoch": 3.034986309704898,
+      "grad_norm": 0.3306165039539337,
+      "learning_rate": 7.38184822960811e-06,
+      "loss": 0.0087,
+      "step": 4988
+    },
+    {
+      "epoch": 3.0355947672649832,
+      "grad_norm": 0.20668867230415344,
+      "learning_rate": 7.37301820790727e-06,
+      "loss": 0.0043,
+      "step": 4989
+    },
+    {
+      "epoch": 3.0362032248250683,
+      "grad_norm": 0.28162097930908203,
+      "learning_rate": 7.364192556963187e-06,
+      "loss": 0.0085,
+      "step": 4990
+    },
+    {
+      "epoch": 3.0368116823851534,
+      "grad_norm": 0.2704124450683594,
+      "learning_rate": 7.35537127896426e-06,
+      "loss": 0.0047,
+      "step": 4991
+    },
+    {
+      "epoch": 3.037420139945239,
+      "grad_norm": 0.18822601437568665,
+      "learning_rate": 7.346554376097778e-06,
+      "loss": 0.0044,
+      "step": 4992
+    },
+    {
+      "epoch": 3.038028597505324,
+      "grad_norm": 0.28500285744667053,
+      "learning_rate": 7.337741850549992e-06,
+      "loss": 0.0073,
+      "step": 4993
+    },
+    {
+      "epoch": 3.038637055065409,
+      "grad_norm": 0.394612193107605,
+      "learning_rate": 7.328933704506022e-06,
+      "loss": 0.0124,
+      "step": 4994
+    },
+    {
+      "epoch": 3.0392455126254942,
+      "grad_norm": 0.2029077261686325,
+      "learning_rate": 7.320129940149925e-06,
+      "loss": 0.0046,
+      "step": 4995
+    },
+    {
+      "epoch": 3.0398539701855793,
+      "grad_norm": 0.3149600625038147,
+      "learning_rate": 7.311330559664673e-06,
+      "loss": 0.0061,
+      "step": 4996
+    },
+    {
+      "epoch": 3.040462427745665,
+      "grad_norm": 0.18582724034786224,
+      "learning_rate": 7.3025355652321464e-06,
+      "loss": 0.0046,
+      "step": 4997
+    },
+    {
+      "epoch": 3.04107088530575,
+      "grad_norm": 0.2304268330335617,
+      "learning_rate": 7.293744959033124e-06,
+      "loss": 0.0065,
+      "step": 4998
+    },
+    {
+      "epoch": 3.041679342865835,
+      "grad_norm": 0.3009764850139618,
+      "learning_rate": 7.284958743247322e-06,
+      "loss": 0.0064,
+      "step": 4999
+    },
+    {
+      "epoch": 3.04228780042592,
+      "grad_norm": 0.2738197147846222,
+      "learning_rate": 7.276176920053351e-06,
+      "loss": 0.0107,
+      "step": 5000
+    },
+    {
+      "epoch": 3.0428962579860053,
+      "grad_norm": 0.28810760378837585,
+      "learning_rate": 7.267399491628748e-06,
+      "loss": 0.0055,
+      "step": 5001
+    },
+    {
+      "epoch": 3.043504715546091,
+      "grad_norm": 0.3494376838207245,
+      "learning_rate": 7.258626460149922e-06,
+      "loss": 0.0112,
+      "step": 5002
+    },
+    {
+      "epoch": 3.044113173106176,
+      "grad_norm": 0.20463795959949493,
+      "learning_rate": 7.249857827792253e-06,
+      "loss": 0.0053,
+      "step": 5003
+    },
+    {
+      "epoch": 3.044721630666261,
+      "grad_norm": 0.22482581436634064,
+      "learning_rate": 7.241093596729976e-06,
+      "loss": 0.0088,
+      "step": 5004
+    },
+    {
+      "epoch": 3.045330088226346,
+      "grad_norm": 0.29432442784309387,
+      "learning_rate": 7.232333769136254e-06,
+      "loss": 0.0081,
+      "step": 5005
+    },
+    {
+      "epoch": 3.045938545786431,
+      "grad_norm": 0.273941308259964,
+      "learning_rate": 7.223578347183166e-06,
+      "loss": 0.0092,
+      "step": 5006
+    },
+    {
+      "epoch": 3.0465470033465167,
+      "grad_norm": 0.3165578842163086,
+      "learning_rate": 7.2148273330416985e-06,
+      "loss": 0.0134,
+      "step": 5007
+    },
+    {
+      "epoch": 3.047155460906602,
+      "grad_norm": 0.293646901845932,
+      "learning_rate": 7.206080728881715e-06,
+      "loss": 0.009,
+      "step": 5008
+    },
+    {
+      "epoch": 3.047763918466687,
+      "grad_norm": 0.24793358147144318,
+      "learning_rate": 7.197338536872039e-06,
+      "loss": 0.008,
+      "step": 5009
+    },
+    {
+      "epoch": 3.048372376026772,
+      "grad_norm": 0.23065130412578583,
+      "learning_rate": 7.188600759180347e-06,
+      "loss": 0.0052,
+      "step": 5010
+    },
+    {
+      "epoch": 3.048980833586857,
+      "grad_norm": 0.32581770420074463,
+      "learning_rate": 7.1798673979732585e-06,
+      "loss": 0.0089,
+      "step": 5011
+    },
+    {
+      "epoch": 3.0495892911469427,
+      "grad_norm": 0.2617538571357727,
+      "learning_rate": 7.17113845541626e-06,
+      "loss": 0.0074,
+      "step": 5012
+    },
+    {
+      "epoch": 3.0501977487070278,
+      "grad_norm": 0.2555639445781708,
+      "learning_rate": 7.162413933673795e-06,
+      "loss": 0.008,
+      "step": 5013
+    },
+    {
+      "epoch": 3.050806206267113,
+      "grad_norm": 0.1875854879617691,
+      "learning_rate": 7.153693834909161e-06,
+      "loss": 0.005,
+      "step": 5014
+    },
+    {
+      "epoch": 3.051414663827198,
+      "grad_norm": 0.19386368989944458,
+      "learning_rate": 7.144978161284585e-06,
+      "loss": 0.0039,
+      "step": 5015
+    },
+    {
+      "epoch": 3.052023121387283,
+      "grad_norm": 0.23817914724349976,
+      "learning_rate": 7.13626691496119e-06,
+      "loss": 0.0068,
+      "step": 5016
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.215720534324646,
+      "learning_rate": 7.127560098099012e-06,
+      "loss": 0.006,
+      "step": 5017
+    },
+    {
+      "epoch": 3.0532400365074537,
+      "grad_norm": 0.35492417216300964,
+      "learning_rate": 7.118857712856952e-06,
+      "loss": 0.0065,
+      "step": 5018
+    },
+    {
+      "epoch": 3.0538484940675388,
+      "grad_norm": 0.30515995621681213,
+      "learning_rate": 7.110159761392876e-06,
+      "loss": 0.0106,
+      "step": 5019
+    },
+    {
+      "epoch": 3.054456951627624,
+      "grad_norm": 0.1642477661371231,
+      "learning_rate": 7.101466245863483e-06,
+      "loss": 0.0038,
+      "step": 5020
+    },
+    {
+      "epoch": 3.055065409187709,
+      "grad_norm": 0.3018817901611328,
+      "learning_rate": 7.092777168424422e-06,
+      "loss": 0.0085,
+      "step": 5021
+    },
+    {
+      "epoch": 3.0556738667477945,
+      "grad_norm": 0.4562574625015259,
+      "learning_rate": 7.084092531230196e-06,
+      "loss": 0.0108,
+      "step": 5022
+    },
+    {
+      "epoch": 3.0562823243078796,
+      "grad_norm": 0.3128783702850342,
+      "learning_rate": 7.07541233643427e-06,
+      "loss": 0.007,
+      "step": 5023
+    },
+    {
+      "epoch": 3.0568907818679647,
+      "grad_norm": 0.2551473379135132,
+      "learning_rate": 7.066736586188941e-06,
+      "loss": 0.0075,
+      "step": 5024
+    },
+    {
+      "epoch": 3.05749923942805,
+      "grad_norm": 0.20833851397037506,
+      "learning_rate": 7.058065282645443e-06,
+      "loss": 0.0061,
+      "step": 5025
+    },
+    {
+      "epoch": 3.058107696988135,
+      "grad_norm": 0.2601064443588257,
+      "learning_rate": 7.049398427953899e-06,
+      "loss": 0.005,
+      "step": 5026
+    },
+    {
+      "epoch": 3.0587161545482204,
+      "grad_norm": 0.2717060148715973,
+      "learning_rate": 7.040736024263334e-06,
+      "loss": 0.007,
+      "step": 5027
+    },
+    {
+      "epoch": 3.0593246121083055,
+      "grad_norm": 0.1747554987668991,
+      "learning_rate": 7.03207807372164e-06,
+      "loss": 0.0047,
+      "step": 5028
+    },
+    {
+      "epoch": 3.0599330696683906,
+      "grad_norm": 0.3020252585411072,
+      "learning_rate": 7.023424578475659e-06,
+      "loss": 0.008,
+      "step": 5029
+    },
+    {
+      "epoch": 3.0605415272284757,
+      "grad_norm": 0.20203709602355957,
+      "learning_rate": 7.014775540671076e-06,
+      "loss": 0.0062,
+      "step": 5030
+    },
+    {
+      "epoch": 3.061149984788561,
+      "grad_norm": 0.30242085456848145,
+      "learning_rate": 7.006130962452498e-06,
+      "loss": 0.0091,
+      "step": 5031
+    },
+    {
+      "epoch": 3.0617584423486464,
+      "grad_norm": 0.25199273228645325,
+      "learning_rate": 6.997490845963417e-06,
+      "loss": 0.0062,
+      "step": 5032
+    },
+    {
+      "epoch": 3.0623668999087315,
+      "grad_norm": 0.2626577615737915,
+      "learning_rate": 6.988855193346236e-06,
+      "loss": 0.0067,
+      "step": 5033
+    },
+    {
+      "epoch": 3.0629753574688166,
+      "grad_norm": 0.21209199726581573,
+      "learning_rate": 6.980224006742214e-06,
+      "loss": 0.0036,
+      "step": 5034
+    },
+    {
+      "epoch": 3.0635838150289016,
+      "grad_norm": 0.14695894718170166,
+      "learning_rate": 6.97159728829154e-06,
+      "loss": 0.0028,
+      "step": 5035
+    },
+    {
+      "epoch": 3.0641922725889867,
+      "grad_norm": 0.28200700879096985,
+      "learning_rate": 6.9629750401332766e-06,
+      "loss": 0.0076,
+      "step": 5036
+    },
+    {
+      "epoch": 3.0648007301490723,
+      "grad_norm": 0.26668545603752136,
+      "learning_rate": 6.954357264405392e-06,
+      "loss": 0.0058,
+      "step": 5037
+    },
+    {
+      "epoch": 3.0654091877091574,
+      "grad_norm": 0.2839964032173157,
+      "learning_rate": 6.945743963244711e-06,
+      "loss": 0.0102,
+      "step": 5038
+    },
+    {
+      "epoch": 3.0660176452692425,
+      "grad_norm": 0.29744213819503784,
+      "learning_rate": 6.937135138787001e-06,
+      "loss": 0.0084,
+      "step": 5039
+    },
+    {
+      "epoch": 3.0666261028293276,
+      "grad_norm": 0.3763432800769806,
+      "learning_rate": 6.928530793166874e-06,
+      "loss": 0.0099,
+      "step": 5040
+    },
+    {
+      "epoch": 3.0672345603894127,
+      "grad_norm": 0.23103450238704681,
+      "learning_rate": 6.919930928517854e-06,
+      "loss": 0.0071,
+      "step": 5041
+    },
+    {
+      "epoch": 3.067843017949498,
+      "grad_norm": 0.3754606246948242,
+      "learning_rate": 6.91133554697235e-06,
+      "loss": 0.0078,
+      "step": 5042
+    },
+    {
+      "epoch": 3.0684514755095833,
+      "grad_norm": 0.31299889087677,
+      "learning_rate": 6.902744650661663e-06,
+      "loss": 0.0066,
+      "step": 5043
+    },
+    {
+      "epoch": 3.0690599330696684,
+      "grad_norm": 0.26163333654403687,
+      "learning_rate": 6.894158241715959e-06,
+      "loss": 0.0059,
+      "step": 5044
+    },
+    {
+      "epoch": 3.0696683906297535,
+      "grad_norm": 0.25679513812065125,
+      "learning_rate": 6.885576322264336e-06,
+      "loss": 0.0043,
+      "step": 5045
+    },
+    {
+      "epoch": 3.0702768481898386,
+      "grad_norm": 0.38195693492889404,
+      "learning_rate": 6.87699889443473e-06,
+      "loss": 0.0088,
+      "step": 5046
+    },
+    {
+      "epoch": 3.070885305749924,
+      "grad_norm": 0.3322720229625702,
+      "learning_rate": 6.868425960354005e-06,
+      "loss": 0.0103,
+      "step": 5047
+    },
+    {
+      "epoch": 3.0714937633100092,
+      "grad_norm": 0.3915663957595825,
+      "learning_rate": 6.859857522147864e-06,
+      "loss": 0.0069,
+      "step": 5048
+    },
+    {
+      "epoch": 3.0721022208700943,
+      "grad_norm": 0.26933425664901733,
+      "learning_rate": 6.851293581940954e-06,
+      "loss": 0.0077,
+      "step": 5049
+    },
+    {
+      "epoch": 3.0727106784301794,
+      "grad_norm": 0.24823108315467834,
+      "learning_rate": 6.842734141856755e-06,
+      "loss": 0.0065,
+      "step": 5050
+    },
+    {
+      "epoch": 3.0733191359902645,
+      "grad_norm": 0.31463244557380676,
+      "learning_rate": 6.834179204017655e-06,
+      "loss": 0.0049,
+      "step": 5051
+    },
+    {
+      "epoch": 3.07392759355035,
+      "grad_norm": 0.23442064225673676,
+      "learning_rate": 6.82562877054492e-06,
+      "loss": 0.0064,
+      "step": 5052
+    },
+    {
+      "epoch": 3.074536051110435,
+      "grad_norm": 0.36819660663604736,
+      "learning_rate": 6.817082843558717e-06,
+      "loss": 0.0045,
+      "step": 5053
+    },
+    {
+      "epoch": 3.0751445086705202,
+      "grad_norm": 0.24585528671741486,
+      "learning_rate": 6.80854142517805e-06,
+      "loss": 0.0048,
+      "step": 5054
+    },
+    {
+      "epoch": 3.0757529662306053,
+      "grad_norm": 0.18307526409626007,
+      "learning_rate": 6.80000451752087e-06,
+      "loss": 0.0048,
+      "step": 5055
+    },
+    {
+      "epoch": 3.0763614237906904,
+      "grad_norm": 0.20536406338214874,
+      "learning_rate": 6.791472122703946e-06,
+      "loss": 0.0045,
+      "step": 5056
+    },
+    {
+      "epoch": 3.076969881350776,
+      "grad_norm": 0.3933330774307251,
+      "learning_rate": 6.782944242842976e-06,
+      "loss": 0.0126,
+      "step": 5057
+    },
+    {
+      "epoch": 3.077578338910861,
+      "grad_norm": 0.2311299741268158,
+      "learning_rate": 6.774420880052496e-06,
+      "loss": 0.0061,
+      "step": 5058
+    },
+    {
+      "epoch": 3.078186796470946,
+      "grad_norm": 0.2359190732240677,
+      "learning_rate": 6.7659020364459705e-06,
+      "loss": 0.0067,
+      "step": 5059
+    },
+    {
+      "epoch": 3.0787952540310313,
+      "grad_norm": 0.2867051959037781,
+      "learning_rate": 6.757387714135696e-06,
+      "loss": 0.0055,
+      "step": 5060
+    },
+    {
+      "epoch": 3.0794037115911164,
+      "grad_norm": 0.2318498194217682,
+      "learning_rate": 6.748877915232882e-06,
+      "loss": 0.0062,
+      "step": 5061
+    },
+    {
+      "epoch": 3.080012169151202,
+      "grad_norm": 0.15246573090553284,
+      "learning_rate": 6.7403726418476005e-06,
+      "loss": 0.003,
+      "step": 5062
+    },
+    {
+      "epoch": 3.080620626711287,
+      "grad_norm": 0.30910444259643555,
+      "learning_rate": 6.731871896088812e-06,
+      "loss": 0.009,
+      "step": 5063
+    },
+    {
+      "epoch": 3.081229084271372,
+      "grad_norm": 0.32887697219848633,
+      "learning_rate": 6.723375680064325e-06,
+      "loss": 0.007,
+      "step": 5064
+    },
+    {
+      "epoch": 3.081837541831457,
+      "grad_norm": 0.38914939761161804,
+      "learning_rate": 6.714883995880877e-06,
+      "loss": 0.0124,
+      "step": 5065
+    },
+    {
+      "epoch": 3.0824459993915423,
+      "grad_norm": 0.3048250079154968,
+      "learning_rate": 6.706396845644031e-06,
+      "loss": 0.0073,
+      "step": 5066
+    },
+    {
+      "epoch": 3.083054456951628,
+      "grad_norm": 0.28059664368629456,
+      "learning_rate": 6.697914231458249e-06,
+      "loss": 0.0043,
+      "step": 5067
+    },
+    {
+      "epoch": 3.083662914511713,
+      "grad_norm": 0.23873847723007202,
+      "learning_rate": 6.689436155426873e-06,
+      "loss": 0.0053,
+      "step": 5068
+    },
+    {
+      "epoch": 3.084271372071798,
+      "grad_norm": 0.4285385012626648,
+      "learning_rate": 6.680962619652115e-06,
+      "loss": 0.0055,
+      "step": 5069
+    },
+    {
+      "epoch": 3.084879829631883,
+      "grad_norm": 0.2138291299343109,
+      "learning_rate": 6.672493626235044e-06,
+      "loss": 0.0059,
+      "step": 5070
+    },
+    {
+      "epoch": 3.085488287191968,
+      "grad_norm": 0.591597318649292,
+      "learning_rate": 6.664029177275624e-06,
+      "loss": 0.0073,
+      "step": 5071
+    },
+    {
+      "epoch": 3.0860967447520538,
+      "grad_norm": 0.301170289516449,
+      "learning_rate": 6.655569274872689e-06,
+      "loss": 0.0088,
+      "step": 5072
+    },
+    {
+      "epoch": 3.086705202312139,
+      "grad_norm": 0.309881329536438,
+      "learning_rate": 6.647113921123941e-06,
+      "loss": 0.0096,
+      "step": 5073
+    },
+    {
+      "epoch": 3.087313659872224,
+      "grad_norm": 0.17016667127609253,
+      "learning_rate": 6.638663118125951e-06,
+      "loss": 0.0056,
+      "step": 5074
+    },
+    {
+      "epoch": 3.087922117432309,
+      "grad_norm": 0.18052910268306732,
+      "learning_rate": 6.6302168679741785e-06,
+      "loss": 0.0032,
+      "step": 5075
+    },
+    {
+      "epoch": 3.088530574992394,
+      "grad_norm": 0.22513878345489502,
+      "learning_rate": 6.6217751727629285e-06,
+      "loss": 0.0041,
+      "step": 5076
+    },
+    {
+      "epoch": 3.0891390325524792,
+      "grad_norm": 0.2399200052022934,
+      "learning_rate": 6.61333803458539e-06,
+      "loss": 0.0082,
+      "step": 5077
+    },
+    {
+      "epoch": 3.0897474901125648,
+      "grad_norm": 0.20230963826179504,
+      "learning_rate": 6.604905455533625e-06,
+      "loss": 0.0073,
+      "step": 5078
+    },
+    {
+      "epoch": 3.09035594767265,
+      "grad_norm": 0.24099519848823547,
+      "learning_rate": 6.596477437698565e-06,
+      "loss": 0.0054,
+      "step": 5079
+    },
+    {
+      "epoch": 3.090964405232735,
+      "grad_norm": 0.16781140863895416,
+      "learning_rate": 6.588053983170006e-06,
+      "loss": 0.0036,
+      "step": 5080
+    },
+    {
+      "epoch": 3.09157286279282,
+      "grad_norm": 0.24010391533374786,
+      "learning_rate": 6.579635094036607e-06,
+      "loss": 0.0067,
+      "step": 5081
+    },
+    {
+      "epoch": 3.092181320352905,
+      "grad_norm": 0.3611644208431244,
+      "learning_rate": 6.571220772385905e-06,
+      "loss": 0.009,
+      "step": 5082
+    },
+    {
+      "epoch": 3.0927897779129907,
+      "grad_norm": 0.2840390205383301,
+      "learning_rate": 6.562811020304305e-06,
+      "loss": 0.008,
+      "step": 5083
+    },
+    {
+      "epoch": 3.093398235473076,
+      "grad_norm": 0.29138410091400146,
+      "learning_rate": 6.55440583987707e-06,
+      "loss": 0.0054,
+      "step": 5084
+    },
+    {
+      "epoch": 3.094006693033161,
+      "grad_norm": 0.23507152497768402,
+      "learning_rate": 6.546005233188343e-06,
+      "loss": 0.0051,
+      "step": 5085
+    },
+    {
+      "epoch": 3.094615150593246,
+      "grad_norm": 0.2997680902481079,
+      "learning_rate": 6.537609202321113e-06,
+      "loss": 0.007,
+      "step": 5086
+    },
+    {
+      "epoch": 3.095223608153331,
+      "grad_norm": 0.2853991389274597,
+      "learning_rate": 6.529217749357247e-06,
+      "loss": 0.0089,
+      "step": 5087
+    },
+    {
+      "epoch": 3.0958320657134166,
+      "grad_norm": 0.18508751690387726,
+      "learning_rate": 6.520830876377482e-06,
+      "loss": 0.0029,
+      "step": 5088
+    },
+    {
+      "epoch": 3.0964405232735017,
+      "grad_norm": 0.1906099170446396,
+      "learning_rate": 6.5124485854614086e-06,
+      "loss": 0.0041,
+      "step": 5089
+    },
+    {
+      "epoch": 3.097048980833587,
+      "grad_norm": 0.3323763906955719,
+      "learning_rate": 6.504070878687485e-06,
+      "loss": 0.0109,
+      "step": 5090
+    },
+    {
+      "epoch": 3.097657438393672,
+      "grad_norm": 0.2999829947948456,
+      "learning_rate": 6.495697758133046e-06,
+      "loss": 0.0077,
+      "step": 5091
+    },
+    {
+      "epoch": 3.098265895953757,
+      "grad_norm": 0.3580150604248047,
+      "learning_rate": 6.487329225874256e-06,
+      "loss": 0.0135,
+      "step": 5092
+    },
+    {
+      "epoch": 3.0988743535138426,
+      "grad_norm": 0.17444658279418945,
+      "learning_rate": 6.478965283986174e-06,
+      "loss": 0.0035,
+      "step": 5093
+    },
+    {
+      "epoch": 3.0994828110739276,
+      "grad_norm": 0.20913676917552948,
+      "learning_rate": 6.470605934542703e-06,
+      "loss": 0.0045,
+      "step": 5094
+    },
+    {
+      "epoch": 3.1000912686340127,
+      "grad_norm": 0.2408442199230194,
+      "learning_rate": 6.462251179616621e-06,
+      "loss": 0.0065,
+      "step": 5095
+    },
+    {
+      "epoch": 3.100699726194098,
+      "grad_norm": 0.18831227719783783,
+      "learning_rate": 6.453901021279559e-06,
+      "loss": 0.0035,
+      "step": 5096
+    },
+    {
+      "epoch": 3.101308183754183,
+      "grad_norm": 0.3297005295753479,
+      "learning_rate": 6.445555461602001e-06,
+      "loss": 0.0078,
+      "step": 5097
+    },
+    {
+      "epoch": 3.1019166413142685,
+      "grad_norm": 0.21692611277103424,
+      "learning_rate": 6.437214502653299e-06,
+      "loss": 0.0051,
+      "step": 5098
+    },
+    {
+      "epoch": 3.1025250988743536,
+      "grad_norm": 0.24288775026798248,
+      "learning_rate": 6.428878146501666e-06,
+      "loss": 0.0051,
+      "step": 5099
+    },
+    {
+      "epoch": 3.1031335564344387,
+      "grad_norm": 0.17630305886268616,
+      "learning_rate": 6.420546395214167e-06,
+      "loss": 0.0034,
+      "step": 5100
+    },
+    {
+      "epoch": 3.1037420139945238,
+      "grad_norm": 0.2515610158443451,
+      "learning_rate": 6.412219250856735e-06,
+      "loss": 0.006,
+      "step": 5101
+    },
+    {
+      "epoch": 3.104350471554609,
+      "grad_norm": 0.3058002293109894,
+      "learning_rate": 6.403896715494159e-06,
+      "loss": 0.01,
+      "step": 5102
+    },
+    {
+      "epoch": 3.1049589291146944,
+      "grad_norm": 0.29932624101638794,
+      "learning_rate": 6.395578791190066e-06,
+      "loss": 0.0046,
+      "step": 5103
+    },
+    {
+      "epoch": 3.1055673866747795,
+      "grad_norm": 0.2611561119556427,
+      "learning_rate": 6.3872654800069626e-06,
+      "loss": 0.0097,
+      "step": 5104
+    },
+    {
+      "epoch": 3.1061758442348646,
+      "grad_norm": 0.2513970136642456,
+      "learning_rate": 6.378956784006204e-06,
+      "loss": 0.006,
+      "step": 5105
+    },
+    {
+      "epoch": 3.1067843017949497,
+      "grad_norm": 0.2295328974723816,
+      "learning_rate": 6.370652705248007e-06,
+      "loss": 0.0055,
+      "step": 5106
+    },
+    {
+      "epoch": 3.107392759355035,
+      "grad_norm": 0.3369624614715576,
+      "learning_rate": 6.362353245791411e-06,
+      "loss": 0.013,
+      "step": 5107
+    },
+    {
+      "epoch": 3.1080012169151203,
+      "grad_norm": 0.21619094908237457,
+      "learning_rate": 6.354058407694374e-06,
+      "loss": 0.0059,
+      "step": 5108
+    },
+    {
+      "epoch": 3.1086096744752054,
+      "grad_norm": 0.2147800624370575,
+      "learning_rate": 6.34576819301364e-06,
+      "loss": 0.0071,
+      "step": 5109
+    },
+    {
+      "epoch": 3.1092181320352905,
+      "grad_norm": 0.18836335837841034,
+      "learning_rate": 6.337482603804851e-06,
+      "loss": 0.0051,
+      "step": 5110
+    },
+    {
+      "epoch": 3.1098265895953756,
+      "grad_norm": 0.15471284091472626,
+      "learning_rate": 6.329201642122481e-06,
+      "loss": 0.0046,
+      "step": 5111
+    },
+    {
+      "epoch": 3.1104350471554607,
+      "grad_norm": 0.2526698708534241,
+      "learning_rate": 6.320925310019876e-06,
+      "loss": 0.0071,
+      "step": 5112
+    },
+    {
+      "epoch": 3.1110435047155462,
+      "grad_norm": 0.2389310896396637,
+      "learning_rate": 6.312653609549196e-06,
+      "loss": 0.0065,
+      "step": 5113
+    },
+    {
+      "epoch": 3.1116519622756313,
+      "grad_norm": 0.22375918924808502,
+      "learning_rate": 6.304386542761509e-06,
+      "loss": 0.0057,
+      "step": 5114
+    },
+    {
+      "epoch": 3.1122604198357164,
+      "grad_norm": 0.29425686597824097,
+      "learning_rate": 6.29612411170668e-06,
+      "loss": 0.005,
+      "step": 5115
+    },
+    {
+      "epoch": 3.1128688773958015,
+      "grad_norm": 0.37902581691741943,
+      "learning_rate": 6.287866318433464e-06,
+      "loss": 0.0065,
+      "step": 5116
+    },
+    {
+      "epoch": 3.1134773349558866,
+      "grad_norm": 0.3976466655731201,
+      "learning_rate": 6.279613164989426e-06,
+      "loss": 0.0142,
+      "step": 5117
+    },
+    {
+      "epoch": 3.114085792515972,
+      "grad_norm": 0.2854953706264496,
+      "learning_rate": 6.27136465342103e-06,
+      "loss": 0.0062,
+      "step": 5118
+    },
+    {
+      "epoch": 3.1146942500760573,
+      "grad_norm": 0.6086744070053101,
+      "learning_rate": 6.263120785773549e-06,
+      "loss": 0.0085,
+      "step": 5119
+    },
+    {
+      "epoch": 3.1153027076361424,
+      "grad_norm": 0.2866113483905792,
+      "learning_rate": 6.254881564091119e-06,
+      "loss": 0.0042,
+      "step": 5120
+    },
+    {
+      "epoch": 3.1159111651962275,
+      "grad_norm": 0.20252281427383423,
+      "learning_rate": 6.246646990416727e-06,
+      "loss": 0.0056,
+      "step": 5121
+    },
+    {
+      "epoch": 3.1165196227563126,
+      "grad_norm": 0.21693319082260132,
+      "learning_rate": 6.238417066792212e-06,
+      "loss": 0.0052,
+      "step": 5122
+    },
+    {
+      "epoch": 3.117128080316398,
+      "grad_norm": 0.2629375159740448,
+      "learning_rate": 6.230191795258228e-06,
+      "loss": 0.0105,
+      "step": 5123
+    },
+    {
+      "epoch": 3.117736537876483,
+      "grad_norm": 0.18073906004428864,
+      "learning_rate": 6.2219711778543274e-06,
+      "loss": 0.0055,
+      "step": 5124
+    },
+    {
+      "epoch": 3.1183449954365683,
+      "grad_norm": 0.3107531666755676,
+      "learning_rate": 6.213755216618861e-06,
+      "loss": 0.0077,
+      "step": 5125
+    },
+    {
+      "epoch": 3.1189534529966534,
+      "grad_norm": 0.3121819496154785,
+      "learning_rate": 6.205543913589059e-06,
+      "loss": 0.0042,
+      "step": 5126
+    },
+    {
+      "epoch": 3.1195619105567385,
+      "grad_norm": 0.3623504638671875,
+      "learning_rate": 6.19733727080096e-06,
+      "loss": 0.0076,
+      "step": 5127
+    },
+    {
+      "epoch": 3.120170368116824,
+      "grad_norm": 0.3374268412590027,
+      "learning_rate": 6.189135290289499e-06,
+      "loss": 0.0076,
+      "step": 5128
+    },
+    {
+      "epoch": 3.120778825676909,
+      "grad_norm": 0.2626723051071167,
+      "learning_rate": 6.180937974088405e-06,
+      "loss": 0.0045,
+      "step": 5129
+    },
+    {
+      "epoch": 3.121387283236994,
+      "grad_norm": 0.39012977480888367,
+      "learning_rate": 6.172745324230275e-06,
+      "loss": 0.0123,
+      "step": 5130
+    },
+    {
+      "epoch": 3.1219957407970793,
+      "grad_norm": 0.1949702948331833,
+      "learning_rate": 6.164557342746547e-06,
+      "loss": 0.0064,
+      "step": 5131
+    },
+    {
+      "epoch": 3.1226041983571644,
+      "grad_norm": 0.3511587679386139,
+      "learning_rate": 6.156374031667503e-06,
+      "loss": 0.0094,
+      "step": 5132
+    },
+    {
+      "epoch": 3.12321265591725,
+      "grad_norm": 0.2506071627140045,
+      "learning_rate": 6.1481953930222435e-06,
+      "loss": 0.0068,
+      "step": 5133
+    },
+    {
+      "epoch": 3.123821113477335,
+      "grad_norm": 0.2543926537036896,
+      "learning_rate": 6.140021428838761e-06,
+      "loss": 0.0035,
+      "step": 5134
+    },
+    {
+      "epoch": 3.12442957103742,
+      "grad_norm": 0.2524166703224182,
+      "learning_rate": 6.131852141143834e-06,
+      "loss": 0.0059,
+      "step": 5135
+    },
+    {
+      "epoch": 3.1250380285975052,
+      "grad_norm": 0.3138072192668915,
+      "learning_rate": 6.123687531963113e-06,
+      "loss": 0.0061,
+      "step": 5136
+    },
+    {
+      "epoch": 3.1256464861575903,
+      "grad_norm": 0.32244062423706055,
+      "learning_rate": 6.115527603321081e-06,
+      "loss": 0.0081,
+      "step": 5137
+    },
+    {
+      "epoch": 3.126254943717676,
+      "grad_norm": 0.21048331260681152,
+      "learning_rate": 6.1073723572410645e-06,
+      "loss": 0.0037,
+      "step": 5138
+    },
+    {
+      "epoch": 3.126863401277761,
+      "grad_norm": 0.255831778049469,
+      "learning_rate": 6.099221795745213e-06,
+      "loss": 0.0074,
+      "step": 5139
+    },
+    {
+      "epoch": 3.127471858837846,
+      "grad_norm": 0.3076886236667633,
+      "learning_rate": 6.091075920854536e-06,
+      "loss": 0.0058,
+      "step": 5140
+    },
+    {
+      "epoch": 3.128080316397931,
+      "grad_norm": 0.21628208458423615,
+      "learning_rate": 6.082934734588866e-06,
+      "loss": 0.0051,
+      "step": 5141
+    },
+    {
+      "epoch": 3.1286887739580163,
+      "grad_norm": 0.2794947326183319,
+      "learning_rate": 6.074798238966886e-06,
+      "loss": 0.0078,
+      "step": 5142
+    },
+    {
+      "epoch": 3.129297231518102,
+      "grad_norm": 0.38834357261657715,
+      "learning_rate": 6.066666436006088e-06,
+      "loss": 0.0055,
+      "step": 5143
+    },
+    {
+      "epoch": 3.129905689078187,
+      "grad_norm": 0.28329214453697205,
+      "learning_rate": 6.058539327722848e-06,
+      "loss": 0.0069,
+      "step": 5144
+    },
+    {
+      "epoch": 3.130514146638272,
+      "grad_norm": 0.2482929527759552,
+      "learning_rate": 6.050416916132329e-06,
+      "loss": 0.0069,
+      "step": 5145
+    },
+    {
+      "epoch": 3.131122604198357,
+      "grad_norm": 0.22554363310337067,
+      "learning_rate": 6.042299203248555e-06,
+      "loss": 0.0045,
+      "step": 5146
+    },
+    {
+      "epoch": 3.131731061758442,
+      "grad_norm": 0.3571288287639618,
+      "learning_rate": 6.034186191084384e-06,
+      "loss": 0.0082,
+      "step": 5147
+    },
+    {
+      "epoch": 3.1323395193185277,
+      "grad_norm": 0.24375028908252716,
+      "learning_rate": 6.026077881651513e-06,
+      "loss": 0.0037,
+      "step": 5148
+    },
+    {
+      "epoch": 3.132947976878613,
+      "grad_norm": 0.33513811230659485,
+      "learning_rate": 6.017974276960445e-06,
+      "loss": 0.0047,
+      "step": 5149
+    },
+    {
+      "epoch": 3.133556434438698,
+      "grad_norm": 0.32763129472732544,
+      "learning_rate": 6.0098753790205494e-06,
+      "loss": 0.0091,
+      "step": 5150
+    },
+    {
+      "epoch": 3.134164891998783,
+      "grad_norm": 0.18572641909122467,
+      "learning_rate": 6.001781189840011e-06,
+      "loss": 0.0065,
+      "step": 5151
+    },
+    {
+      "epoch": 3.134773349558868,
+      "grad_norm": 0.24770891666412354,
+      "learning_rate": 5.9936917114258585e-06,
+      "loss": 0.007,
+      "step": 5152
+    },
+    {
+      "epoch": 3.1353818071189536,
+      "grad_norm": 0.38206419348716736,
+      "learning_rate": 5.985606945783926e-06,
+      "loss": 0.005,
+      "step": 5153
+    },
+    {
+      "epoch": 3.1359902646790387,
+      "grad_norm": 0.24334634840488434,
+      "learning_rate": 5.977526894918928e-06,
+      "loss": 0.0062,
+      "step": 5154
+    },
+    {
+      "epoch": 3.136598722239124,
+      "grad_norm": 0.23436123132705688,
+      "learning_rate": 5.969451560834355e-06,
+      "loss": 0.0062,
+      "step": 5155
+    },
+    {
+      "epoch": 3.137207179799209,
+      "grad_norm": 0.20244354009628296,
+      "learning_rate": 5.9613809455325635e-06,
+      "loss": 0.0046,
+      "step": 5156
+    },
+    {
+      "epoch": 3.137815637359294,
+      "grad_norm": 0.20892736315727234,
+      "learning_rate": 5.95331505101473e-06,
+      "loss": 0.0037,
+      "step": 5157
+    },
+    {
+      "epoch": 3.1384240949193796,
+      "grad_norm": 0.32044070959091187,
+      "learning_rate": 5.945253879280862e-06,
+      "loss": 0.0097,
+      "step": 5158
+    },
+    {
+      "epoch": 3.1390325524794647,
+      "grad_norm": 0.27256110310554504,
+      "learning_rate": 5.93719743232978e-06,
+      "loss": 0.0057,
+      "step": 5159
+    },
+    {
+      "epoch": 3.1396410100395498,
+      "grad_norm": 0.22417287528514862,
+      "learning_rate": 5.929145712159173e-06,
+      "loss": 0.0039,
+      "step": 5160
+    },
+    {
+      "epoch": 3.140249467599635,
+      "grad_norm": 0.28702032566070557,
+      "learning_rate": 5.921098720765508e-06,
+      "loss": 0.0038,
+      "step": 5161
+    },
+    {
+      "epoch": 3.14085792515972,
+      "grad_norm": 0.3003734350204468,
+      "learning_rate": 5.91305646014412e-06,
+      "loss": 0.0079,
+      "step": 5162
+    },
+    {
+      "epoch": 3.1414663827198055,
+      "grad_norm": 0.3371279835700989,
+      "learning_rate": 5.905018932289133e-06,
+      "loss": 0.0085,
+      "step": 5163
+    },
+    {
+      "epoch": 3.1420748402798906,
+      "grad_norm": 0.24667906761169434,
+      "learning_rate": 5.896986139193547e-06,
+      "loss": 0.0089,
+      "step": 5164
+    },
+    {
+      "epoch": 3.1426832978399757,
+      "grad_norm": 0.20292513072490692,
+      "learning_rate": 5.888958082849139e-06,
+      "loss": 0.0047,
+      "step": 5165
+    },
+    {
+      "epoch": 3.143291755400061,
+      "grad_norm": 0.1873752474784851,
+      "learning_rate": 5.880934765246537e-06,
+      "loss": 0.0056,
+      "step": 5166
+    },
+    {
+      "epoch": 3.143900212960146,
+      "grad_norm": 0.2573238015174866,
+      "learning_rate": 5.872916188375194e-06,
+      "loss": 0.0035,
+      "step": 5167
+    },
+    {
+      "epoch": 3.1445086705202314,
+      "grad_norm": 0.22613008320331573,
+      "learning_rate": 5.864902354223384e-06,
+      "loss": 0.0063,
+      "step": 5168
+    },
+    {
+      "epoch": 3.1451171280803165,
+      "grad_norm": 0.2717379927635193,
+      "learning_rate": 5.856893264778188e-06,
+      "loss": 0.0072,
+      "step": 5169
+    },
+    {
+      "epoch": 3.1457255856404016,
+      "grad_norm": 0.18429331481456757,
+      "learning_rate": 5.848888922025553e-06,
+      "loss": 0.005,
+      "step": 5170
+    },
+    {
+      "epoch": 3.1463340432004867,
+      "grad_norm": 0.2671307921409607,
+      "learning_rate": 5.8408893279502e-06,
+      "loss": 0.0073,
+      "step": 5171
+    },
+    {
+      "epoch": 3.146942500760572,
+      "grad_norm": 0.263640820980072,
+      "learning_rate": 5.832894484535709e-06,
+      "loss": 0.0094,
+      "step": 5172
+    },
+    {
+      "epoch": 3.1475509583206573,
+      "grad_norm": 0.17915846407413483,
+      "learning_rate": 5.8249043937644465e-06,
+      "loss": 0.0039,
+      "step": 5173
+    },
+    {
+      "epoch": 3.1481594158807424,
+      "grad_norm": 0.30745455622673035,
+      "learning_rate": 5.816919057617653e-06,
+      "loss": 0.0067,
+      "step": 5174
+    },
+    {
+      "epoch": 3.1487678734408275,
+      "grad_norm": 0.2966357171535492,
+      "learning_rate": 5.808938478075335e-06,
+      "loss": 0.0063,
+      "step": 5175
+    },
+    {
+      "epoch": 3.1493763310009126,
+      "grad_norm": 0.3174609839916229,
+      "learning_rate": 5.800962657116351e-06,
+      "loss": 0.0088,
+      "step": 5176
+    },
+    {
+      "epoch": 3.1499847885609977,
+      "grad_norm": 0.25037750601768494,
+      "learning_rate": 5.792991596718375e-06,
+      "loss": 0.0063,
+      "step": 5177
+    },
+    {
+      "epoch": 3.1505932461210833,
+      "grad_norm": 0.30678263306617737,
+      "learning_rate": 5.785025298857902e-06,
+      "loss": 0.0043,
+      "step": 5178
+    },
+    {
+      "epoch": 3.1512017036811684,
+      "grad_norm": 0.25345203280448914,
+      "learning_rate": 5.777063765510219e-06,
+      "loss": 0.0061,
+      "step": 5179
+    },
+    {
+      "epoch": 3.1518101612412535,
+      "grad_norm": 0.5578262209892273,
+      "learning_rate": 5.769106998649488e-06,
+      "loss": 0.0109,
+      "step": 5180
+    },
+    {
+      "epoch": 3.1524186188013386,
+      "grad_norm": 0.2420710027217865,
+      "learning_rate": 5.761155000248627e-06,
+      "loss": 0.007,
+      "step": 5181
+    },
+    {
+      "epoch": 3.1530270763614237,
+      "grad_norm": 0.21901202201843262,
+      "learning_rate": 5.753207772279415e-06,
+      "loss": 0.0049,
+      "step": 5182
+    },
+    {
+      "epoch": 3.153635533921509,
+      "grad_norm": 0.26763221621513367,
+      "learning_rate": 5.745265316712428e-06,
+      "loss": 0.0065,
+      "step": 5183
+    },
+    {
+      "epoch": 3.1542439914815943,
+      "grad_norm": 0.23685556650161743,
+      "learning_rate": 5.737327635517073e-06,
+      "loss": 0.007,
+      "step": 5184
+    },
+    {
+      "epoch": 3.1548524490416794,
+      "grad_norm": 0.318765789270401,
+      "learning_rate": 5.729394730661547e-06,
+      "loss": 0.0065,
+      "step": 5185
+    },
+    {
+      "epoch": 3.1554609066017645,
+      "grad_norm": 0.289823055267334,
+      "learning_rate": 5.721466604112893e-06,
+      "loss": 0.0079,
+      "step": 5186
+    },
+    {
+      "epoch": 3.1560693641618496,
+      "grad_norm": 0.23716124892234802,
+      "learning_rate": 5.713543257836951e-06,
+      "loss": 0.0055,
+      "step": 5187
+    },
+    {
+      "epoch": 3.156677821721935,
+      "grad_norm": 0.25619250535964966,
+      "learning_rate": 5.705624693798389e-06,
+      "loss": 0.0063,
+      "step": 5188
+    },
+    {
+      "epoch": 3.15728627928202,
+      "grad_norm": 0.21789373457431793,
+      "learning_rate": 5.6977109139606605e-06,
+      "loss": 0.0036,
+      "step": 5189
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.2993996739387512,
+      "learning_rate": 5.689801920286078e-06,
+      "loss": 0.0066,
+      "step": 5190
+    },
+    {
+      "epoch": 3.1585031944021904,
+      "grad_norm": 0.3557785749435425,
+      "learning_rate": 5.681897714735726e-06,
+      "loss": 0.0102,
+      "step": 5191
+    },
+    {
+      "epoch": 3.1591116519622755,
+      "grad_norm": 0.21176862716674805,
+      "learning_rate": 5.673998299269523e-06,
+      "loss": 0.0035,
+      "step": 5192
+    },
+    {
+      "epoch": 3.159720109522361,
+      "grad_norm": 0.23229624330997467,
+      "learning_rate": 5.666103675846191e-06,
+      "loss": 0.0055,
+      "step": 5193
+    },
+    {
+      "epoch": 3.160328567082446,
+      "grad_norm": 0.1171480268239975,
+      "learning_rate": 5.65821384642328e-06,
+      "loss": 0.0017,
+      "step": 5194
+    },
+    {
+      "epoch": 3.1609370246425312,
+      "grad_norm": 0.20375606417655945,
+      "learning_rate": 5.650328812957115e-06,
+      "loss": 0.0049,
+      "step": 5195
+    },
+    {
+      "epoch": 3.1615454822026163,
+      "grad_norm": 0.3008398711681366,
+      "learning_rate": 5.642448577402887e-06,
+      "loss": 0.0067,
+      "step": 5196
+    },
+    {
+      "epoch": 3.1621539397627014,
+      "grad_norm": 0.3283615708351135,
+      "learning_rate": 5.6345731417145385e-06,
+      "loss": 0.005,
+      "step": 5197
+    },
+    {
+      "epoch": 3.162762397322787,
+      "grad_norm": 0.2205916792154312,
+      "learning_rate": 5.6267025078448634e-06,
+      "loss": 0.0053,
+      "step": 5198
+    },
+    {
+      "epoch": 3.163370854882872,
+      "grad_norm": 0.23091734945774078,
+      "learning_rate": 5.618836677745445e-06,
+      "loss": 0.0067,
+      "step": 5199
+    },
+    {
+      "epoch": 3.163979312442957,
+      "grad_norm": 0.3182157278060913,
+      "learning_rate": 5.610975653366693e-06,
+      "loss": 0.0087,
+      "step": 5200
+    },
+    {
+      "epoch": 3.1645877700030423,
+      "grad_norm": 0.3846273422241211,
+      "learning_rate": 5.603119436657794e-06,
+      "loss": 0.0171,
+      "step": 5201
+    },
+    {
+      "epoch": 3.1651962275631274,
+      "grad_norm": 0.26661422848701477,
+      "learning_rate": 5.595268029566777e-06,
+      "loss": 0.0062,
+      "step": 5202
+    },
+    {
+      "epoch": 3.165804685123213,
+      "grad_norm": 0.4423938989639282,
+      "learning_rate": 5.587421434040457e-06,
+      "loss": 0.0079,
+      "step": 5203
+    },
+    {
+      "epoch": 3.166413142683298,
+      "grad_norm": 0.3375749886035919,
+      "learning_rate": 5.579579652024467e-06,
+      "loss": 0.0085,
+      "step": 5204
+    },
+    {
+      "epoch": 3.167021600243383,
+      "grad_norm": 0.2887289822101593,
+      "learning_rate": 5.571742685463238e-06,
+      "loss": 0.0104,
+      "step": 5205
+    },
+    {
+      "epoch": 3.167630057803468,
+      "grad_norm": 0.23182101547718048,
+      "learning_rate": 5.563910536300021e-06,
+      "loss": 0.005,
+      "step": 5206
+    },
+    {
+      "epoch": 3.1682385153635533,
+      "grad_norm": 0.252333402633667,
+      "learning_rate": 5.55608320647685e-06,
+      "loss": 0.0059,
+      "step": 5207
+    },
+    {
+      "epoch": 3.1688469729236384,
+      "grad_norm": 0.22592368721961975,
+      "learning_rate": 5.548260697934579e-06,
+      "loss": 0.0045,
+      "step": 5208
+    },
+    {
+      "epoch": 3.169455430483724,
+      "grad_norm": 0.2711806893348694,
+      "learning_rate": 5.540443012612867e-06,
+      "loss": 0.0061,
+      "step": 5209
+    },
+    {
+      "epoch": 3.170063888043809,
+      "grad_norm": 0.23822571337223053,
+      "learning_rate": 5.532630152450175e-06,
+      "loss": 0.0053,
+      "step": 5210
+    },
+    {
+      "epoch": 3.170672345603894,
+      "grad_norm": 0.3533017039299011,
+      "learning_rate": 5.5248221193837715e-06,
+      "loss": 0.0102,
+      "step": 5211
+    },
+    {
+      "epoch": 3.171280803163979,
+      "grad_norm": 0.2815355360507965,
+      "learning_rate": 5.5170189153497085e-06,
+      "loss": 0.0073,
+      "step": 5212
+    },
+    {
+      "epoch": 3.1718892607240643,
+      "grad_norm": 0.3320218622684479,
+      "learning_rate": 5.509220542282864e-06,
+      "loss": 0.0076,
+      "step": 5213
+    },
+    {
+      "epoch": 3.17249771828415,
+      "grad_norm": 0.2814382314682007,
+      "learning_rate": 5.501427002116913e-06,
+      "loss": 0.0074,
+      "step": 5214
+    },
+    {
+      "epoch": 3.173106175844235,
+      "grad_norm": 0.29092541337013245,
+      "learning_rate": 5.4936382967843206e-06,
+      "loss": 0.0057,
+      "step": 5215
+    },
+    {
+      "epoch": 3.17371463340432,
+      "grad_norm": 0.1667691469192505,
+      "learning_rate": 5.485854428216375e-06,
+      "loss": 0.0039,
+      "step": 5216
+    },
+    {
+      "epoch": 3.174323090964405,
+      "grad_norm": 0.2939755618572235,
+      "learning_rate": 5.478075398343133e-06,
+      "loss": 0.0082,
+      "step": 5217
+    },
+    {
+      "epoch": 3.17493154852449,
+      "grad_norm": 0.21566243469715118,
+      "learning_rate": 5.470301209093478e-06,
+      "loss": 0.0053,
+      "step": 5218
+    },
+    {
+      "epoch": 3.1755400060845758,
+      "grad_norm": 0.26064324378967285,
+      "learning_rate": 5.462531862395087e-06,
+      "loss": 0.0046,
+      "step": 5219
+    },
+    {
+      "epoch": 3.176148463644661,
+      "grad_norm": 0.22674903273582458,
+      "learning_rate": 5.454767360174431e-06,
+      "loss": 0.0043,
+      "step": 5220
+    },
+    {
+      "epoch": 3.176756921204746,
+      "grad_norm": 0.2666117548942566,
+      "learning_rate": 5.447007704356791e-06,
+      "loss": 0.0051,
+      "step": 5221
+    },
+    {
+      "epoch": 3.177365378764831,
+      "grad_norm": 0.27889886498451233,
+      "learning_rate": 5.439252896866226e-06,
+      "loss": 0.0053,
+      "step": 5222
+    },
+    {
+      "epoch": 3.177973836324916,
+      "grad_norm": 0.273966908454895,
+      "learning_rate": 5.431502939625608e-06,
+      "loss": 0.0053,
+      "step": 5223
+    },
+    {
+      "epoch": 3.1785822938850017,
+      "grad_norm": 0.29631203413009644,
+      "learning_rate": 5.423757834556606e-06,
+      "loss": 0.0079,
+      "step": 5224
+    },
+    {
+      "epoch": 3.179190751445087,
+      "grad_norm": 0.3811946213245392,
+      "learning_rate": 5.416017583579686e-06,
+      "loss": 0.0081,
+      "step": 5225
+    },
+    {
+      "epoch": 3.179799209005172,
+      "grad_norm": 0.285285621881485,
+      "learning_rate": 5.408282188614103e-06,
+      "loss": 0.0083,
+      "step": 5226
+    },
+    {
+      "epoch": 3.180407666565257,
+      "grad_norm": 0.2812410593032837,
+      "learning_rate": 5.40055165157792e-06,
+      "loss": 0.0051,
+      "step": 5227
+    },
+    {
+      "epoch": 3.181016124125342,
+      "grad_norm": 0.3686572015285492,
+      "learning_rate": 5.392825974387977e-06,
+      "loss": 0.0078,
+      "step": 5228
+    },
+    {
+      "epoch": 3.1816245816854276,
+      "grad_norm": 0.34885480999946594,
+      "learning_rate": 5.385105158959924e-06,
+      "loss": 0.0081,
+      "step": 5229
+    },
+    {
+      "epoch": 3.1822330392455127,
+      "grad_norm": 0.2072272151708603,
+      "learning_rate": 5.377389207208203e-06,
+      "loss": 0.0027,
+      "step": 5230
+    },
+    {
+      "epoch": 3.182841496805598,
+      "grad_norm": 0.22844795882701874,
+      "learning_rate": 5.369678121046054e-06,
+      "loss": 0.0068,
+      "step": 5231
+    },
+    {
+      "epoch": 3.183449954365683,
+      "grad_norm": 0.3148641884326935,
+      "learning_rate": 5.361971902385485e-06,
+      "loss": 0.0111,
+      "step": 5232
+    },
+    {
+      "epoch": 3.184058411925768,
+      "grad_norm": 0.2549585700035095,
+      "learning_rate": 5.354270553137347e-06,
+      "loss": 0.008,
+      "step": 5233
+    },
+    {
+      "epoch": 3.1846668694858535,
+      "grad_norm": 0.26395851373672485,
+      "learning_rate": 5.346574075211228e-06,
+      "loss": 0.0077,
+      "step": 5234
+    },
+    {
+      "epoch": 3.1852753270459386,
+      "grad_norm": 0.24896180629730225,
+      "learning_rate": 5.338882470515544e-06,
+      "loss": 0.0076,
+      "step": 5235
+    },
+    {
+      "epoch": 3.1858837846060237,
+      "grad_norm": 0.240103617310524,
+      "learning_rate": 5.331195740957493e-06,
+      "loss": 0.0065,
+      "step": 5236
+    },
+    {
+      "epoch": 3.186492242166109,
+      "grad_norm": 0.1803087294101715,
+      "learning_rate": 5.3235138884430655e-06,
+      "loss": 0.0052,
+      "step": 5237
+    },
+    {
+      "epoch": 3.187100699726194,
+      "grad_norm": 0.25235310196876526,
+      "learning_rate": 5.315836914877026e-06,
+      "loss": 0.0046,
+      "step": 5238
+    },
+    {
+      "epoch": 3.1877091572862795,
+      "grad_norm": 0.2217434048652649,
+      "learning_rate": 5.30816482216297e-06,
+      "loss": 0.0073,
+      "step": 5239
+    },
+    {
+      "epoch": 3.1883176148463646,
+      "grad_norm": 0.1752319633960724,
+      "learning_rate": 5.300497612203231e-06,
+      "loss": 0.0029,
+      "step": 5240
+    },
+    {
+      "epoch": 3.1889260724064497,
+      "grad_norm": 0.2559622824192047,
+      "learning_rate": 5.292835286898973e-06,
+      "loss": 0.006,
+      "step": 5241
+    },
+    {
+      "epoch": 3.1895345299665347,
+      "grad_norm": 0.30886584520339966,
+      "learning_rate": 5.285177848150127e-06,
+      "loss": 0.0065,
+      "step": 5242
+    },
+    {
+      "epoch": 3.19014298752662,
+      "grad_norm": 0.32938501238822937,
+      "learning_rate": 5.277525297855427e-06,
+      "loss": 0.0087,
+      "step": 5243
+    },
+    {
+      "epoch": 3.1907514450867054,
+      "grad_norm": 0.30601125955581665,
+      "learning_rate": 5.269877637912374e-06,
+      "loss": 0.0065,
+      "step": 5244
+    },
+    {
+      "epoch": 3.1913599026467905,
+      "grad_norm": 0.1772543340921402,
+      "learning_rate": 5.262234870217272e-06,
+      "loss": 0.0038,
+      "step": 5245
+    },
+    {
+      "epoch": 3.1919683602068756,
+      "grad_norm": 0.31185612082481384,
+      "learning_rate": 5.254596996665215e-06,
+      "loss": 0.0116,
+      "step": 5246
+    },
+    {
+      "epoch": 3.1925768177669607,
+      "grad_norm": 0.2804357409477234,
+      "learning_rate": 5.246964019150077e-06,
+      "loss": 0.0092,
+      "step": 5247
+    },
+    {
+      "epoch": 3.1931852753270458,
+      "grad_norm": 0.31903064250946045,
+      "learning_rate": 5.239335939564505e-06,
+      "loss": 0.0092,
+      "step": 5248
+    },
+    {
+      "epoch": 3.1937937328871313,
+      "grad_norm": 0.27632802724838257,
+      "learning_rate": 5.231712759799967e-06,
+      "loss": 0.0074,
+      "step": 5249
+    },
+    {
+      "epoch": 3.1944021904472164,
+      "grad_norm": 0.15155087411403656,
+      "learning_rate": 5.224094481746675e-06,
+      "loss": 0.0031,
+      "step": 5250
+    },
+    {
+      "epoch": 3.1950106480073015,
+      "grad_norm": 0.322252094745636,
+      "learning_rate": 5.216481107293653e-06,
+      "loss": 0.0045,
+      "step": 5251
+    },
+    {
+      "epoch": 3.1956191055673866,
+      "grad_norm": 0.12668286263942719,
+      "learning_rate": 5.2088726383286965e-06,
+      "loss": 0.0028,
+      "step": 5252
+    },
+    {
+      "epoch": 3.1962275631274717,
+      "grad_norm": 0.3014906048774719,
+      "learning_rate": 5.2012690767383964e-06,
+      "loss": 0.008,
+      "step": 5253
+    },
+    {
+      "epoch": 3.1968360206875572,
+      "grad_norm": 0.21465936303138733,
+      "learning_rate": 5.193670424408109e-06,
+      "loss": 0.0053,
+      "step": 5254
+    },
+    {
+      "epoch": 3.1974444782476423,
+      "grad_norm": 0.24435143172740936,
+      "learning_rate": 5.186076683221988e-06,
+      "loss": 0.0062,
+      "step": 5255
+    },
+    {
+      "epoch": 3.1980529358077274,
+      "grad_norm": 0.28407108783721924,
+      "learning_rate": 5.1784878550629635e-06,
+      "loss": 0.0073,
+      "step": 5256
+    },
+    {
+      "epoch": 3.1986613933678125,
+      "grad_norm": 0.3360306918621063,
+      "learning_rate": 5.1709039418127575e-06,
+      "loss": 0.0102,
+      "step": 5257
+    },
+    {
+      "epoch": 3.1992698509278976,
+      "grad_norm": 0.15760943293571472,
+      "learning_rate": 5.163324945351841e-06,
+      "loss": 0.0029,
+      "step": 5258
+    },
+    {
+      "epoch": 3.199878308487983,
+      "grad_norm": 0.31353893876075745,
+      "learning_rate": 5.15575086755952e-06,
+      "loss": 0.0061,
+      "step": 5259
+    },
+    {
+      "epoch": 3.2004867660480683,
+      "grad_norm": 0.2500174641609192,
+      "learning_rate": 5.148181710313827e-06,
+      "loss": 0.0055,
+      "step": 5260
+    },
+    {
+      "epoch": 3.2010952236081534,
+      "grad_norm": 0.1809082180261612,
+      "learning_rate": 5.140617475491605e-06,
+      "loss": 0.0049,
+      "step": 5261
+    },
+    {
+      "epoch": 3.2017036811682384,
+      "grad_norm": 0.1834014654159546,
+      "learning_rate": 5.1330581649684715e-06,
+      "loss": 0.0044,
+      "step": 5262
+    },
+    {
+      "epoch": 3.2023121387283235,
+      "grad_norm": 0.3861825466156006,
+      "learning_rate": 5.125503780618824e-06,
+      "loss": 0.01,
+      "step": 5263
+    },
+    {
+      "epoch": 3.202920596288409,
+      "grad_norm": 0.33081695437431335,
+      "learning_rate": 5.117954324315813e-06,
+      "loss": 0.0065,
+      "step": 5264
+    },
+    {
+      "epoch": 3.203529053848494,
+      "grad_norm": 0.1698094755411148,
+      "learning_rate": 5.11040979793142e-06,
+      "loss": 0.0031,
+      "step": 5265
+    },
+    {
+      "epoch": 3.2041375114085793,
+      "grad_norm": 0.2472512573003769,
+      "learning_rate": 5.102870203336352e-06,
+      "loss": 0.0062,
+      "step": 5266
+    },
+    {
+      "epoch": 3.2047459689686644,
+      "grad_norm": 0.19104048609733582,
+      "learning_rate": 5.095335542400129e-06,
+      "loss": 0.0045,
+      "step": 5267
+    },
+    {
+      "epoch": 3.2053544265287495,
+      "grad_norm": 0.23127706348896027,
+      "learning_rate": 5.087805816991006e-06,
+      "loss": 0.0056,
+      "step": 5268
+    },
+    {
+      "epoch": 3.205962884088835,
+      "grad_norm": 0.28482139110565186,
+      "learning_rate": 5.080281028976078e-06,
+      "loss": 0.0062,
+      "step": 5269
+    },
+    {
+      "epoch": 3.20657134164892,
+      "grad_norm": 0.3393114507198334,
+      "learning_rate": 5.07276118022115e-06,
+      "loss": 0.0099,
+      "step": 5270
+    },
+    {
+      "epoch": 3.207179799209005,
+      "grad_norm": 0.25069156289100647,
+      "learning_rate": 5.06524627259084e-06,
+      "loss": 0.0043,
+      "step": 5271
+    },
+    {
+      "epoch": 3.2077882567690903,
+      "grad_norm": 0.28203946352005005,
+      "learning_rate": 5.057736307948535e-06,
+      "loss": 0.0068,
+      "step": 5272
+    },
+    {
+      "epoch": 3.2083967143291754,
+      "grad_norm": 0.2758210301399231,
+      "learning_rate": 5.050231288156398e-06,
+      "loss": 0.0066,
+      "step": 5273
+    },
+    {
+      "epoch": 3.2090051718892605,
+      "grad_norm": 0.2686399519443512,
+      "learning_rate": 5.042731215075341e-06,
+      "loss": 0.0049,
+      "step": 5274
+    },
+    {
+      "epoch": 3.209613629449346,
+      "grad_norm": 0.25653237104415894,
+      "learning_rate": 5.035236090565093e-06,
+      "loss": 0.0062,
+      "step": 5275
+    },
+    {
+      "epoch": 3.210222087009431,
+      "grad_norm": 0.21267692744731903,
+      "learning_rate": 5.027745916484119e-06,
+      "loss": 0.0056,
+      "step": 5276
+    },
+    {
+      "epoch": 3.210830544569516,
+      "grad_norm": 0.3413097858428955,
+      "learning_rate": 5.02026069468968e-06,
+      "loss": 0.0055,
+      "step": 5277
+    },
+    {
+      "epoch": 3.2114390021296013,
+      "grad_norm": 0.4562237560749054,
+      "learning_rate": 5.012780427037775e-06,
+      "loss": 0.0205,
+      "step": 5278
+    },
+    {
+      "epoch": 3.2120474596896864,
+      "grad_norm": 0.28480085730552673,
+      "learning_rate": 5.005305115383233e-06,
+      "loss": 0.0078,
+      "step": 5279
+    },
+    {
+      "epoch": 3.212655917249772,
+      "grad_norm": 0.24477575719356537,
+      "learning_rate": 4.997834761579595e-06,
+      "loss": 0.0043,
+      "step": 5280
+    },
+    {
+      "epoch": 3.213264374809857,
+      "grad_norm": 0.27294448018074036,
+      "learning_rate": 4.990369367479203e-06,
+      "loss": 0.0061,
+      "step": 5281
+    },
+    {
+      "epoch": 3.213872832369942,
+      "grad_norm": 0.21413196623325348,
+      "learning_rate": 4.982908934933167e-06,
+      "loss": 0.0047,
+      "step": 5282
+    },
+    {
+      "epoch": 3.2144812899300272,
+      "grad_norm": 0.31827402114868164,
+      "learning_rate": 4.975453465791366e-06,
+      "loss": 0.0074,
+      "step": 5283
+    },
+    {
+      "epoch": 3.2150897474901123,
+      "grad_norm": 0.27179375290870667,
+      "learning_rate": 4.9680029619024295e-06,
+      "loss": 0.0051,
+      "step": 5284
+    },
+    {
+      "epoch": 3.215698205050198,
+      "grad_norm": 0.16175347566604614,
+      "learning_rate": 4.9605574251137985e-06,
+      "loss": 0.0041,
+      "step": 5285
+    },
+    {
+      "epoch": 3.216306662610283,
+      "grad_norm": 0.3044746220111847,
+      "learning_rate": 4.953116857271634e-06,
+      "loss": 0.0106,
+      "step": 5286
+    },
+    {
+      "epoch": 3.216915120170368,
+      "grad_norm": 0.263638436794281,
+      "learning_rate": 4.945681260220891e-06,
+      "loss": 0.0108,
+      "step": 5287
+    },
+    {
+      "epoch": 3.217523577730453,
+      "grad_norm": 0.26071223616600037,
+      "learning_rate": 4.9382506358052916e-06,
+      "loss": 0.0084,
+      "step": 5288
+    },
+    {
+      "epoch": 3.2181320352905383,
+      "grad_norm": 0.20165440440177917,
+      "learning_rate": 4.930824985867328e-06,
+      "loss": 0.0038,
+      "step": 5289
+    },
+    {
+      "epoch": 3.218740492850624,
+      "grad_norm": 0.29097187519073486,
+      "learning_rate": 4.923404312248234e-06,
+      "loss": 0.0078,
+      "step": 5290
+    },
+    {
+      "epoch": 3.219348950410709,
+      "grad_norm": 0.21615852415561676,
+      "learning_rate": 4.915988616788039e-06,
+      "loss": 0.0036,
+      "step": 5291
+    },
+    {
+      "epoch": 3.219957407970794,
+      "grad_norm": 0.2644842565059662,
+      "learning_rate": 4.9085779013255225e-06,
+      "loss": 0.0066,
+      "step": 5292
+    },
+    {
+      "epoch": 3.220565865530879,
+      "grad_norm": 0.330758273601532,
+      "learning_rate": 4.901172167698242e-06,
+      "loss": 0.0079,
+      "step": 5293
+    },
+    {
+      "epoch": 3.221174323090964,
+      "grad_norm": 0.2616957128047943,
+      "learning_rate": 4.89377141774249e-06,
+      "loss": 0.0041,
+      "step": 5294
+    },
+    {
+      "epoch": 3.2217827806510497,
+      "grad_norm": 0.38702672719955444,
+      "learning_rate": 4.886375653293371e-06,
+      "loss": 0.0084,
+      "step": 5295
+    },
+    {
+      "epoch": 3.222391238211135,
+      "grad_norm": 0.3087538778781891,
+      "learning_rate": 4.878984876184706e-06,
+      "loss": 0.0064,
+      "step": 5296
+    },
+    {
+      "epoch": 3.22299969577122,
+      "grad_norm": 0.2493748515844345,
+      "learning_rate": 4.871599088249107e-06,
+      "loss": 0.0052,
+      "step": 5297
+    },
+    {
+      "epoch": 3.223608153331305,
+      "grad_norm": 0.3203872740268707,
+      "learning_rate": 4.86421829131794e-06,
+      "loss": 0.0077,
+      "step": 5298
+    },
+    {
+      "epoch": 3.22421661089139,
+      "grad_norm": 0.21354947984218597,
+      "learning_rate": 4.856842487221344e-06,
+      "loss": 0.0052,
+      "step": 5299
+    },
+    {
+      "epoch": 3.2248250684514757,
+      "grad_norm": 0.41666483879089355,
+      "learning_rate": 4.849471677788195e-06,
+      "loss": 0.0138,
+      "step": 5300
+    },
+    {
+      "epoch": 3.2254335260115607,
+      "grad_norm": 0.1607670783996582,
+      "learning_rate": 4.842105864846155e-06,
+      "loss": 0.0021,
+      "step": 5301
+    },
+    {
+      "epoch": 3.226041983571646,
+      "grad_norm": 0.25365200638771057,
+      "learning_rate": 4.83474505022164e-06,
+      "loss": 0.008,
+      "step": 5302
+    },
+    {
+      "epoch": 3.226650441131731,
+      "grad_norm": 0.23420965671539307,
+      "learning_rate": 4.82738923573983e-06,
+      "loss": 0.0039,
+      "step": 5303
+    },
+    {
+      "epoch": 3.227258898691816,
+      "grad_norm": 0.24002118408679962,
+      "learning_rate": 4.820038423224638e-06,
+      "loss": 0.0039,
+      "step": 5304
+    },
+    {
+      "epoch": 3.2278673562519016,
+      "grad_norm": 0.17119701206684113,
+      "learning_rate": 4.812692614498787e-06,
+      "loss": 0.0022,
+      "step": 5305
+    },
+    {
+      "epoch": 3.2284758138119867,
+      "grad_norm": 0.16908611357212067,
+      "learning_rate": 4.805351811383716e-06,
+      "loss": 0.0036,
+      "step": 5306
+    },
+    {
+      "epoch": 3.2290842713720718,
+      "grad_norm": 0.24062961339950562,
+      "learning_rate": 4.798016015699638e-06,
+      "loss": 0.0045,
+      "step": 5307
+    },
+    {
+      "epoch": 3.229692728932157,
+      "grad_norm": 0.26717016100883484,
+      "learning_rate": 4.790685229265529e-06,
+      "loss": 0.0064,
+      "step": 5308
+    },
+    {
+      "epoch": 3.230301186492242,
+      "grad_norm": 0.30753692984580994,
+      "learning_rate": 4.783359453899125e-06,
+      "loss": 0.0083,
+      "step": 5309
+    },
+    {
+      "epoch": 3.2309096440523275,
+      "grad_norm": 0.2237798273563385,
+      "learning_rate": 4.776038691416892e-06,
+      "loss": 0.0078,
+      "step": 5310
+    },
+    {
+      "epoch": 3.2315181016124126,
+      "grad_norm": 0.24201908707618713,
+      "learning_rate": 4.768722943634099e-06,
+      "loss": 0.0034,
+      "step": 5311
+    },
+    {
+      "epoch": 3.2321265591724977,
+      "grad_norm": 0.14591459929943085,
+      "learning_rate": 4.7614122123647295e-06,
+      "loss": 0.002,
+      "step": 5312
+    },
+    {
+      "epoch": 3.232735016732583,
+      "grad_norm": 0.1915680468082428,
+      "learning_rate": 4.754106499421545e-06,
+      "loss": 0.0033,
+      "step": 5313
+    },
+    {
+      "epoch": 3.233343474292668,
+      "grad_norm": 0.24156653881072998,
+      "learning_rate": 4.746805806616059e-06,
+      "loss": 0.0066,
+      "step": 5314
+    },
+    {
+      "epoch": 3.2339519318527534,
+      "grad_norm": 0.2683356702327728,
+      "learning_rate": 4.739510135758546e-06,
+      "loss": 0.0054,
+      "step": 5315
+    },
+    {
+      "epoch": 3.2345603894128385,
+      "grad_norm": 0.29896220564842224,
+      "learning_rate": 4.7322194886580105e-06,
+      "loss": 0.0082,
+      "step": 5316
+    },
+    {
+      "epoch": 3.2351688469729236,
+      "grad_norm": 0.2597740590572357,
+      "learning_rate": 4.724933867122242e-06,
+      "loss": 0.0076,
+      "step": 5317
+    },
+    {
+      "epoch": 3.2357773045330087,
+      "grad_norm": 0.17904557287693024,
+      "learning_rate": 4.717653272957767e-06,
+      "loss": 0.0035,
+      "step": 5318
+    },
+    {
+      "epoch": 3.236385762093094,
+      "grad_norm": 0.26095980405807495,
+      "learning_rate": 4.710377707969876e-06,
+      "loss": 0.0066,
+      "step": 5319
+    },
+    {
+      "epoch": 3.2369942196531793,
+      "grad_norm": 0.14280088245868683,
+      "learning_rate": 4.703107173962587e-06,
+      "loss": 0.0027,
+      "step": 5320
+    },
+    {
+      "epoch": 3.2376026772132644,
+      "grad_norm": 0.2902020812034607,
+      "learning_rate": 4.695841672738718e-06,
+      "loss": 0.0059,
+      "step": 5321
+    },
+    {
+      "epoch": 3.2382111347733495,
+      "grad_norm": 0.34895089268684387,
+      "learning_rate": 4.688581206099787e-06,
+      "loss": 0.009,
+      "step": 5322
+    },
+    {
+      "epoch": 3.2388195923334346,
+      "grad_norm": 0.21605385839939117,
+      "learning_rate": 4.681325775846096e-06,
+      "loss": 0.0052,
+      "step": 5323
+    },
+    {
+      "epoch": 3.2394280498935197,
+      "grad_norm": 0.21634212136268616,
+      "learning_rate": 4.674075383776689e-06,
+      "loss": 0.0058,
+      "step": 5324
+    },
+    {
+      "epoch": 3.2400365074536053,
+      "grad_norm": 0.25203219056129456,
+      "learning_rate": 4.666830031689365e-06,
+      "loss": 0.0048,
+      "step": 5325
+    },
+    {
+      "epoch": 3.2406449650136904,
+      "grad_norm": 0.26320701837539673,
+      "learning_rate": 4.659589721380661e-06,
+      "loss": 0.0049,
+      "step": 5326
+    },
+    {
+      "epoch": 3.2412534225737755,
+      "grad_norm": 0.268195241689682,
+      "learning_rate": 4.652354454645874e-06,
+      "loss": 0.0058,
+      "step": 5327
+    },
+    {
+      "epoch": 3.2418618801338606,
+      "grad_norm": 0.2631436288356781,
+      "learning_rate": 4.64512423327905e-06,
+      "loss": 0.0049,
+      "step": 5328
+    },
+    {
+      "epoch": 3.2424703376939457,
+      "grad_norm": 0.3505512773990631,
+      "learning_rate": 4.637899059072984e-06,
+      "loss": 0.0039,
+      "step": 5329
+    },
+    {
+      "epoch": 3.243078795254031,
+      "grad_norm": 0.2612294554710388,
+      "learning_rate": 4.630678933819218e-06,
+      "loss": 0.0065,
+      "step": 5330
+    },
+    {
+      "epoch": 3.2436872528141163,
+      "grad_norm": 0.2038453072309494,
+      "learning_rate": 4.623463859308047e-06,
+      "loss": 0.0045,
+      "step": 5331
+    },
+    {
+      "epoch": 3.2442957103742014,
+      "grad_norm": 0.26115256547927856,
+      "learning_rate": 4.616253837328497e-06,
+      "loss": 0.0074,
+      "step": 5332
+    },
+    {
+      "epoch": 3.2449041679342865,
+      "grad_norm": 0.31222108006477356,
+      "learning_rate": 4.6090488696683585e-06,
+      "loss": 0.008,
+      "step": 5333
+    },
+    {
+      "epoch": 3.2455126254943716,
+      "grad_norm": 0.2225833237171173,
+      "learning_rate": 4.601848958114164e-06,
+      "loss": 0.0064,
+      "step": 5334
+    },
+    {
+      "epoch": 3.246121083054457,
+      "grad_norm": 0.2462644726037979,
+      "learning_rate": 4.5946541044511905e-06,
+      "loss": 0.0049,
+      "step": 5335
+    },
+    {
+      "epoch": 3.246729540614542,
+      "grad_norm": 0.27759093046188354,
+      "learning_rate": 4.5874643104634685e-06,
+      "loss": 0.0099,
+      "step": 5336
+    },
+    {
+      "epoch": 3.2473379981746273,
+      "grad_norm": 0.24177397787570953,
+      "learning_rate": 4.580279577933755e-06,
+      "loss": 0.0059,
+      "step": 5337
+    },
+    {
+      "epoch": 3.2479464557347124,
+      "grad_norm": 0.2759784162044525,
+      "learning_rate": 4.573099908643572e-06,
+      "loss": 0.0062,
+      "step": 5338
+    },
+    {
+      "epoch": 3.2485549132947975,
+      "grad_norm": 0.400174617767334,
+      "learning_rate": 4.565925304373176e-06,
+      "loss": 0.0064,
+      "step": 5339
+    },
+    {
+      "epoch": 3.249163370854883,
+      "grad_norm": 0.3525528609752655,
+      "learning_rate": 4.558755766901568e-06,
+      "loss": 0.0097,
+      "step": 5340
+    },
+    {
+      "epoch": 3.249771828414968,
+      "grad_norm": 0.41769731044769287,
+      "learning_rate": 4.551591298006497e-06,
+      "loss": 0.0096,
+      "step": 5341
+    },
+    {
+      "epoch": 3.2503802859750532,
+      "grad_norm": 0.2166384756565094,
+      "learning_rate": 4.54443189946446e-06,
+      "loss": 0.0031,
+      "step": 5342
+    },
+    {
+      "epoch": 3.2509887435351383,
+      "grad_norm": 0.2720005214214325,
+      "learning_rate": 4.537277573050674e-06,
+      "loss": 0.0067,
+      "step": 5343
+    },
+    {
+      "epoch": 3.2515972010952234,
+      "grad_norm": 0.2838521897792816,
+      "learning_rate": 4.530128320539126e-06,
+      "loss": 0.01,
+      "step": 5344
+    },
+    {
+      "epoch": 3.252205658655309,
+      "grad_norm": 0.22926031053066254,
+      "learning_rate": 4.522984143702524e-06,
+      "loss": 0.0049,
+      "step": 5345
+    },
+    {
+      "epoch": 3.252814116215394,
+      "grad_norm": 0.19992421567440033,
+      "learning_rate": 4.515845044312331e-06,
+      "loss": 0.0041,
+      "step": 5346
+    },
+    {
+      "epoch": 3.253422573775479,
+      "grad_norm": 0.28706929087638855,
+      "learning_rate": 4.508711024138746e-06,
+      "loss": 0.0085,
+      "step": 5347
+    },
+    {
+      "epoch": 3.2540310313355643,
+      "grad_norm": 0.1090521439909935,
+      "learning_rate": 4.501582084950715e-06,
+      "loss": 0.0021,
+      "step": 5348
+    },
+    {
+      "epoch": 3.2546394888956494,
+      "grad_norm": 0.18653200566768646,
+      "learning_rate": 4.494458228515902e-06,
+      "loss": 0.0045,
+      "step": 5349
+    },
+    {
+      "epoch": 3.255247946455735,
+      "grad_norm": 0.2562299370765686,
+      "learning_rate": 4.487339456600736e-06,
+      "loss": 0.0074,
+      "step": 5350
+    },
+    {
+      "epoch": 3.25585640401582,
+      "grad_norm": 0.21422119438648224,
+      "learning_rate": 4.480225770970378e-06,
+      "loss": 0.0036,
+      "step": 5351
+    },
+    {
+      "epoch": 3.256464861575905,
+      "grad_norm": 0.16707159578800201,
+      "learning_rate": 4.4731171733887245e-06,
+      "loss": 0.0038,
+      "step": 5352
+    },
+    {
+      "epoch": 3.25707331913599,
+      "grad_norm": 0.27651339769363403,
+      "learning_rate": 4.466013665618407e-06,
+      "loss": 0.0049,
+      "step": 5353
+    },
+    {
+      "epoch": 3.2576817766960753,
+      "grad_norm": 0.18352994322776794,
+      "learning_rate": 4.458915249420798e-06,
+      "loss": 0.0034,
+      "step": 5354
+    },
+    {
+      "epoch": 3.258290234256161,
+      "grad_norm": 0.24244019389152527,
+      "learning_rate": 4.451821926556016e-06,
+      "loss": 0.006,
+      "step": 5355
+    },
+    {
+      "epoch": 3.258898691816246,
+      "grad_norm": 0.2059207558631897,
+      "learning_rate": 4.444733698782902e-06,
+      "loss": 0.0051,
+      "step": 5356
+    },
+    {
+      "epoch": 3.259507149376331,
+      "grad_norm": 0.32662034034729004,
+      "learning_rate": 4.437650567859047e-06,
+      "loss": 0.0062,
+      "step": 5357
+    },
+    {
+      "epoch": 3.260115606936416,
+      "grad_norm": 0.2572704553604126,
+      "learning_rate": 4.430572535540778e-06,
+      "loss": 0.0055,
+      "step": 5358
+    },
+    {
+      "epoch": 3.260724064496501,
+      "grad_norm": 0.15180638432502747,
+      "learning_rate": 4.423499603583137e-06,
+      "loss": 0.0031,
+      "step": 5359
+    },
+    {
+      "epoch": 3.2613325220565867,
+      "grad_norm": 0.20388486981391907,
+      "learning_rate": 4.416431773739924e-06,
+      "loss": 0.004,
+      "step": 5360
+    },
+    {
+      "epoch": 3.261940979616672,
+      "grad_norm": 0.25693532824516296,
+      "learning_rate": 4.409369047763664e-06,
+      "loss": 0.0053,
+      "step": 5361
+    },
+    {
+      "epoch": 3.262549437176757,
+      "grad_norm": 0.311781644821167,
+      "learning_rate": 4.402311427405628e-06,
+      "loss": 0.0063,
+      "step": 5362
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.23537062108516693,
+      "learning_rate": 4.39525891441579e-06,
+      "loss": 0.0053,
+      "step": 5363
+    },
+    {
+      "epoch": 3.263766352296927,
+      "grad_norm": 0.3553119897842407,
+      "learning_rate": 4.388211510542906e-06,
+      "loss": 0.0094,
+      "step": 5364
+    },
+    {
+      "epoch": 3.2643748098570127,
+      "grad_norm": 0.20130811631679535,
+      "learning_rate": 4.38116921753442e-06,
+      "loss": 0.0049,
+      "step": 5365
+    },
+    {
+      "epoch": 3.2649832674170978,
+      "grad_norm": 0.24931548535823822,
+      "learning_rate": 4.374132037136533e-06,
+      "loss": 0.004,
+      "step": 5366
+    },
+    {
+      "epoch": 3.265591724977183,
+      "grad_norm": 0.29390352964401245,
+      "learning_rate": 4.367099971094174e-06,
+      "loss": 0.008,
+      "step": 5367
+    },
+    {
+      "epoch": 3.266200182537268,
+      "grad_norm": 0.3790845274925232,
+      "learning_rate": 4.360073021151004e-06,
+      "loss": 0.0104,
+      "step": 5368
+    },
+    {
+      "epoch": 3.266808640097353,
+      "grad_norm": 0.2828552722930908,
+      "learning_rate": 4.353051189049398e-06,
+      "loss": 0.0077,
+      "step": 5369
+    },
+    {
+      "epoch": 3.2674170976574386,
+      "grad_norm": 0.13647259771823883,
+      "learning_rate": 4.346034476530503e-06,
+      "loss": 0.0035,
+      "step": 5370
+    },
+    {
+      "epoch": 3.2680255552175237,
+      "grad_norm": 0.28553807735443115,
+      "learning_rate": 4.339022885334154e-06,
+      "loss": 0.0087,
+      "step": 5371
+    },
+    {
+      "epoch": 3.268634012777609,
+      "grad_norm": 0.37478142976760864,
+      "learning_rate": 4.332016417198942e-06,
+      "loss": 0.0084,
+      "step": 5372
+    },
+    {
+      "epoch": 3.269242470337694,
+      "grad_norm": 0.2366034984588623,
+      "learning_rate": 4.32501507386216e-06,
+      "loss": 0.0061,
+      "step": 5373
+    },
+    {
+      "epoch": 3.269850927897779,
+      "grad_norm": 0.25971001386642456,
+      "learning_rate": 4.318018857059878e-06,
+      "loss": 0.0082,
+      "step": 5374
+    },
+    {
+      "epoch": 3.2704593854578645,
+      "grad_norm": 0.24216106534004211,
+      "learning_rate": 4.311027768526846e-06,
+      "loss": 0.0056,
+      "step": 5375
+    },
+    {
+      "epoch": 3.2710678430179496,
+      "grad_norm": 0.32058781385421753,
+      "learning_rate": 4.3040418099965674e-06,
+      "loss": 0.0085,
+      "step": 5376
+    },
+    {
+      "epoch": 3.2716763005780347,
+      "grad_norm": 0.25957608222961426,
+      "learning_rate": 4.297060983201273e-06,
+      "loss": 0.005,
+      "step": 5377
+    },
+    {
+      "epoch": 3.27228475813812,
+      "grad_norm": 0.2595154047012329,
+      "learning_rate": 4.290085289871917e-06,
+      "loss": 0.005,
+      "step": 5378
+    },
+    {
+      "epoch": 3.272893215698205,
+      "grad_norm": 0.25397470593452454,
+      "learning_rate": 4.283114731738166e-06,
+      "loss": 0.0034,
+      "step": 5379
+    },
+    {
+      "epoch": 3.2735016732582904,
+      "grad_norm": 0.2799084782600403,
+      "learning_rate": 4.276149310528452e-06,
+      "loss": 0.0068,
+      "step": 5380
+    },
+    {
+      "epoch": 3.2741101308183755,
+      "grad_norm": 0.27302873134613037,
+      "learning_rate": 4.269189027969889e-06,
+      "loss": 0.0046,
+      "step": 5381
+    },
+    {
+      "epoch": 3.2747185883784606,
+      "grad_norm": 0.27299267053604126,
+      "learning_rate": 4.26223388578835e-06,
+      "loss": 0.0083,
+      "step": 5382
+    },
+    {
+      "epoch": 3.2753270459385457,
+      "grad_norm": 0.3280734717845917,
+      "learning_rate": 4.255283885708403e-06,
+      "loss": 0.0097,
+      "step": 5383
+    },
+    {
+      "epoch": 3.275935503498631,
+      "grad_norm": 0.2109125405550003,
+      "learning_rate": 4.2483390294533825e-06,
+      "loss": 0.005,
+      "step": 5384
+    },
+    {
+      "epoch": 3.2765439610587164,
+      "grad_norm": 0.3015410304069519,
+      "learning_rate": 4.2413993187453035e-06,
+      "loss": 0.005,
+      "step": 5385
+    },
+    {
+      "epoch": 3.2771524186188015,
+      "grad_norm": 0.239542618393898,
+      "learning_rate": 4.234464755304934e-06,
+      "loss": 0.0064,
+      "step": 5386
+    },
+    {
+      "epoch": 3.2777608761788866,
+      "grad_norm": 0.3143784999847412,
+      "learning_rate": 4.227535340851754e-06,
+      "loss": 0.0056,
+      "step": 5387
+    },
+    {
+      "epoch": 3.2783693337389717,
+      "grad_norm": 0.3585420250892639,
+      "learning_rate": 4.220611077103978e-06,
+      "loss": 0.0104,
+      "step": 5388
+    },
+    {
+      "epoch": 3.2789777912990568,
+      "grad_norm": 0.3050104081630707,
+      "learning_rate": 4.213691965778516e-06,
+      "loss": 0.0094,
+      "step": 5389
+    },
+    {
+      "epoch": 3.2795862488591423,
+      "grad_norm": 0.22966350615024567,
+      "learning_rate": 4.206778008591039e-06,
+      "loss": 0.0047,
+      "step": 5390
+    },
+    {
+      "epoch": 3.2801947064192274,
+      "grad_norm": 0.2953983247280121,
+      "learning_rate": 4.199869207255907e-06,
+      "loss": 0.0051,
+      "step": 5391
+    },
+    {
+      "epoch": 3.2808031639793125,
+      "grad_norm": 0.22054730355739594,
+      "learning_rate": 4.192965563486221e-06,
+      "loss": 0.0047,
+      "step": 5392
+    },
+    {
+      "epoch": 3.2814116215393976,
+      "grad_norm": 0.21199162304401398,
+      "learning_rate": 4.186067078993794e-06,
+      "loss": 0.0051,
+      "step": 5393
+    },
+    {
+      "epoch": 3.2820200790994827,
+      "grad_norm": 0.184243842959404,
+      "learning_rate": 4.179173755489171e-06,
+      "loss": 0.0028,
+      "step": 5394
+    },
+    {
+      "epoch": 3.282628536659568,
+      "grad_norm": 0.15772341191768646,
+      "learning_rate": 4.172285594681594e-06,
+      "loss": 0.0034,
+      "step": 5395
+    },
+    {
+      "epoch": 3.2832369942196533,
+      "grad_norm": 0.2756187617778778,
+      "learning_rate": 4.165402598279048e-06,
+      "loss": 0.005,
+      "step": 5396
+    },
+    {
+      "epoch": 3.2838454517797384,
+      "grad_norm": 0.2757441997528076,
+      "learning_rate": 4.1585247679882315e-06,
+      "loss": 0.0049,
+      "step": 5397
+    },
+    {
+      "epoch": 3.2844539093398235,
+      "grad_norm": 0.2898223400115967,
+      "learning_rate": 4.151652105514559e-06,
+      "loss": 0.0065,
+      "step": 5398
+    },
+    {
+      "epoch": 3.2850623668999086,
+      "grad_norm": 1.003989815711975,
+      "learning_rate": 4.144784612562152e-06,
+      "loss": 0.0069,
+      "step": 5399
+    },
+    {
+      "epoch": 3.285670824459994,
+      "grad_norm": 0.230075865983963,
+      "learning_rate": 4.1379222908338844e-06,
+      "loss": 0.0046,
+      "step": 5400
+    },
+    {
+      "epoch": 3.2862792820200792,
+      "grad_norm": 0.24226723611354828,
+      "learning_rate": 4.13106514203131e-06,
+      "loss": 0.0061,
+      "step": 5401
+    },
+    {
+      "epoch": 3.2868877395801643,
+      "grad_norm": 0.2802073657512665,
+      "learning_rate": 4.124213167854721e-06,
+      "loss": 0.0073,
+      "step": 5402
+    },
+    {
+      "epoch": 3.2874961971402494,
+      "grad_norm": 0.2725411355495453,
+      "learning_rate": 4.1173663700031174e-06,
+      "loss": 0.0034,
+      "step": 5403
+    },
+    {
+      "epoch": 3.2881046547003345,
+      "grad_norm": 0.20064030587673187,
+      "learning_rate": 4.11052475017423e-06,
+      "loss": 0.0034,
+      "step": 5404
+    },
+    {
+      "epoch": 3.28871311226042,
+      "grad_norm": 0.2537406086921692,
+      "learning_rate": 4.103688310064485e-06,
+      "loss": 0.0069,
+      "step": 5405
+    },
+    {
+      "epoch": 3.289321569820505,
+      "grad_norm": 0.22151683270931244,
+      "learning_rate": 4.096857051369035e-06,
+      "loss": 0.006,
+      "step": 5406
+    },
+    {
+      "epoch": 3.2899300273805903,
+      "grad_norm": 0.14316684007644653,
+      "learning_rate": 4.090030975781753e-06,
+      "loss": 0.003,
+      "step": 5407
+    },
+    {
+      "epoch": 3.2905384849406754,
+      "grad_norm": 0.3060914874076843,
+      "learning_rate": 4.08321008499522e-06,
+      "loss": 0.0093,
+      "step": 5408
+    },
+    {
+      "epoch": 3.2911469425007605,
+      "grad_norm": 0.23609232902526855,
+      "learning_rate": 4.076394380700724e-06,
+      "loss": 0.0054,
+      "step": 5409
+    },
+    {
+      "epoch": 3.291755400060846,
+      "grad_norm": 0.34055647253990173,
+      "learning_rate": 4.069583864588291e-06,
+      "loss": 0.007,
+      "step": 5410
+    },
+    {
+      "epoch": 3.292363857620931,
+      "grad_norm": 0.20988178253173828,
+      "learning_rate": 4.062778538346634e-06,
+      "loss": 0.0051,
+      "step": 5411
+    },
+    {
+      "epoch": 3.292972315181016,
+      "grad_norm": 0.23799017071723938,
+      "learning_rate": 4.055978403663191e-06,
+      "loss": 0.0048,
+      "step": 5412
+    },
+    {
+      "epoch": 3.2935807727411013,
+      "grad_norm": 0.28818562626838684,
+      "learning_rate": 4.049183462224115e-06,
+      "loss": 0.0076,
+      "step": 5413
+    },
+    {
+      "epoch": 3.2941892303011864,
+      "grad_norm": 0.2099044919013977,
+      "learning_rate": 4.042393715714274e-06,
+      "loss": 0.0033,
+      "step": 5414
+    },
+    {
+      "epoch": 3.294797687861272,
+      "grad_norm": 0.27117833495140076,
+      "learning_rate": 4.0356091658172225e-06,
+      "loss": 0.0066,
+      "step": 5415
+    },
+    {
+      "epoch": 3.295406145421357,
+      "grad_norm": 0.3128974437713623,
+      "learning_rate": 4.028829814215271e-06,
+      "loss": 0.0082,
+      "step": 5416
+    },
+    {
+      "epoch": 3.296014602981442,
+      "grad_norm": 0.20908896625041962,
+      "learning_rate": 4.022055662589397e-06,
+      "loss": 0.0045,
+      "step": 5417
+    },
+    {
+      "epoch": 3.296623060541527,
+      "grad_norm": 0.2693098783493042,
+      "learning_rate": 4.01528671261932e-06,
+      "loss": 0.0045,
+      "step": 5418
+    },
+    {
+      "epoch": 3.2972315181016123,
+      "grad_norm": 0.25489383935928345,
+      "learning_rate": 4.008522965983444e-06,
+      "loss": 0.0049,
+      "step": 5419
+    },
+    {
+      "epoch": 3.297839975661698,
+      "grad_norm": 0.27955323457717896,
+      "learning_rate": 4.001764424358914e-06,
+      "loss": 0.0091,
+      "step": 5420
+    },
+    {
+      "epoch": 3.298448433221783,
+      "grad_norm": 0.1617559790611267,
+      "learning_rate": 3.995011089421552e-06,
+      "loss": 0.0024,
+      "step": 5421
+    },
+    {
+      "epoch": 3.299056890781868,
+      "grad_norm": 0.1988738626241684,
+      "learning_rate": 3.9882629628459095e-06,
+      "loss": 0.0048,
+      "step": 5422
+    },
+    {
+      "epoch": 3.299665348341953,
+      "grad_norm": 0.32020819187164307,
+      "learning_rate": 3.98152004630524e-06,
+      "loss": 0.0103,
+      "step": 5423
+    },
+    {
+      "epoch": 3.3002738059020382,
+      "grad_norm": 0.2752763628959656,
+      "learning_rate": 3.974782341471508e-06,
+      "loss": 0.0072,
+      "step": 5424
+    },
+    {
+      "epoch": 3.3008822634621233,
+      "grad_norm": 0.177403524518013,
+      "learning_rate": 3.968049850015374e-06,
+      "loss": 0.0036,
+      "step": 5425
+    },
+    {
+      "epoch": 3.301490721022209,
+      "grad_norm": 0.2594372630119324,
+      "learning_rate": 3.96132257360623e-06,
+      "loss": 0.0054,
+      "step": 5426
+    },
+    {
+      "epoch": 3.302099178582294,
+      "grad_norm": 0.24807630479335785,
+      "learning_rate": 3.954600513912149e-06,
+      "loss": 0.0048,
+      "step": 5427
+    },
+    {
+      "epoch": 3.302707636142379,
+      "grad_norm": 0.24896809458732605,
+      "learning_rate": 3.9478836725999306e-06,
+      "loss": 0.0062,
+      "step": 5428
+    },
+    {
+      "epoch": 3.303316093702464,
+      "grad_norm": 0.28816282749176025,
+      "learning_rate": 3.941172051335054e-06,
+      "loss": 0.0076,
+      "step": 5429
+    },
+    {
+      "epoch": 3.3039245512625492,
+      "grad_norm": 0.27258211374282837,
+      "learning_rate": 3.934465651781746e-06,
+      "loss": 0.0078,
+      "step": 5430
+    },
+    {
+      "epoch": 3.304533008822635,
+      "grad_norm": 0.2567967176437378,
+      "learning_rate": 3.927764475602893e-06,
+      "loss": 0.0085,
+      "step": 5431
+    },
+    {
+      "epoch": 3.30514146638272,
+      "grad_norm": 0.2778639793395996,
+      "learning_rate": 3.9210685244601195e-06,
+      "loss": 0.0084,
+      "step": 5432
+    },
+    {
+      "epoch": 3.305749923942805,
+      "grad_norm": 0.2634139060974121,
+      "learning_rate": 3.914377800013738e-06,
+      "loss": 0.0034,
+      "step": 5433
+    },
+    {
+      "epoch": 3.30635838150289,
+      "grad_norm": 0.3199003338813782,
+      "learning_rate": 3.907692303922775e-06,
+      "loss": 0.0084,
+      "step": 5434
+    },
+    {
+      "epoch": 3.306966839062975,
+      "grad_norm": 0.22238844633102417,
+      "learning_rate": 3.9010120378449365e-06,
+      "loss": 0.004,
+      "step": 5435
+    },
+    {
+      "epoch": 3.3075752966230607,
+      "grad_norm": 0.23715293407440186,
+      "learning_rate": 3.894337003436679e-06,
+      "loss": 0.0041,
+      "step": 5436
+    },
+    {
+      "epoch": 3.308183754183146,
+      "grad_norm": 0.2940605580806732,
+      "learning_rate": 3.88766720235311e-06,
+      "loss": 0.007,
+      "step": 5437
+    },
+    {
+      "epoch": 3.308792211743231,
+      "grad_norm": 0.3147020637989044,
+      "learning_rate": 3.8810026362480684e-06,
+      "loss": 0.0059,
+      "step": 5438
+    },
+    {
+      "epoch": 3.309400669303316,
+      "grad_norm": 0.19217897951602936,
+      "learning_rate": 3.8743433067740895e-06,
+      "loss": 0.0067,
+      "step": 5439
+    },
+    {
+      "epoch": 3.310009126863401,
+      "grad_norm": 0.36218491196632385,
+      "learning_rate": 3.867689215582415e-06,
+      "loss": 0.0147,
+      "step": 5440
+    },
+    {
+      "epoch": 3.3106175844234866,
+      "grad_norm": 0.23334696888923645,
+      "learning_rate": 3.861040364322974e-06,
+      "loss": 0.0067,
+      "step": 5441
+    },
+    {
+      "epoch": 3.3112260419835717,
+      "grad_norm": 0.3586207926273346,
+      "learning_rate": 3.854396754644405e-06,
+      "loss": 0.0062,
+      "step": 5442
+    },
+    {
+      "epoch": 3.311834499543657,
+      "grad_norm": 0.3108196556568146,
+      "learning_rate": 3.84775838819405e-06,
+      "loss": 0.0105,
+      "step": 5443
+    },
+    {
+      "epoch": 3.312442957103742,
+      "grad_norm": 0.21571455895900726,
+      "learning_rate": 3.841125266617946e-06,
+      "loss": 0.0057,
+      "step": 5444
+    },
+    {
+      "epoch": 3.313051414663827,
+      "grad_norm": 0.2929924726486206,
+      "learning_rate": 3.834497391560829e-06,
+      "loss": 0.0084,
+      "step": 5445
+    },
+    {
+      "epoch": 3.3136598722239126,
+      "grad_norm": 0.1458842009305954,
+      "learning_rate": 3.827874764666145e-06,
+      "loss": 0.0021,
+      "step": 5446
+    },
+    {
+      "epoch": 3.3142683297839977,
+      "grad_norm": 0.25384876132011414,
+      "learning_rate": 3.821257387576014e-06,
+      "loss": 0.0074,
+      "step": 5447
+    },
+    {
+      "epoch": 3.3148767873440828,
+      "grad_norm": 0.29817652702331543,
+      "learning_rate": 3.8146452619312768e-06,
+      "loss": 0.0078,
+      "step": 5448
+    },
+    {
+      "epoch": 3.315485244904168,
+      "grad_norm": 0.2839571535587311,
+      "learning_rate": 3.8080383893714684e-06,
+      "loss": 0.006,
+      "step": 5449
+    },
+    {
+      "epoch": 3.316093702464253,
+      "grad_norm": 0.3559333384037018,
+      "learning_rate": 3.801436771534814e-06,
+      "loss": 0.0114,
+      "step": 5450
+    },
+    {
+      "epoch": 3.3167021600243385,
+      "grad_norm": 0.46538999676704407,
+      "learning_rate": 3.7948404100582453e-06,
+      "loss": 0.0078,
+      "step": 5451
+    },
+    {
+      "epoch": 3.3173106175844236,
+      "grad_norm": 0.41122740507125854,
+      "learning_rate": 3.788249306577374e-06,
+      "loss": 0.0127,
+      "step": 5452
+    },
+    {
+      "epoch": 3.3179190751445087,
+      "grad_norm": 0.22749930620193481,
+      "learning_rate": 3.7816634627265285e-06,
+      "loss": 0.0075,
+      "step": 5453
+    },
+    {
+      "epoch": 3.3185275327045938,
+      "grad_norm": 0.257730633020401,
+      "learning_rate": 3.7750828801387167e-06,
+      "loss": 0.0062,
+      "step": 5454
+    },
+    {
+      "epoch": 3.319135990264679,
+      "grad_norm": 0.27139922976493835,
+      "learning_rate": 3.7685075604456553e-06,
+      "loss": 0.0042,
+      "step": 5455
+    },
+    {
+      "epoch": 3.3197444478247644,
+      "grad_norm": 0.23728352785110474,
+      "learning_rate": 3.7619375052777518e-06,
+      "loss": 0.0036,
+      "step": 5456
+    },
+    {
+      "epoch": 3.3203529053848495,
+      "grad_norm": 0.35973915457725525,
+      "learning_rate": 3.755372716264094e-06,
+      "loss": 0.0105,
+      "step": 5457
+    },
+    {
+      "epoch": 3.3209613629449346,
+      "grad_norm": 0.15002618730068207,
+      "learning_rate": 3.748813195032483e-06,
+      "loss": 0.0033,
+      "step": 5458
+    },
+    {
+      "epoch": 3.3215698205050197,
+      "grad_norm": 0.14451484382152557,
+      "learning_rate": 3.7422589432094064e-06,
+      "loss": 0.0016,
+      "step": 5459
+    },
+    {
+      "epoch": 3.322178278065105,
+      "grad_norm": 0.19004569947719574,
+      "learning_rate": 3.7357099624200475e-06,
+      "loss": 0.0041,
+      "step": 5460
+    },
+    {
+      "epoch": 3.3227867356251903,
+      "grad_norm": 0.27472901344299316,
+      "learning_rate": 3.729166254288277e-06,
+      "loss": 0.006,
+      "step": 5461
+    },
+    {
+      "epoch": 3.3233951931852754,
+      "grad_norm": 0.19888129830360413,
+      "learning_rate": 3.7226278204366695e-06,
+      "loss": 0.0047,
+      "step": 5462
+    },
+    {
+      "epoch": 3.3240036507453605,
+      "grad_norm": 0.267270565032959,
+      "learning_rate": 3.7160946624864716e-06,
+      "loss": 0.0077,
+      "step": 5463
+    },
+    {
+      "epoch": 3.3246121083054456,
+      "grad_norm": 0.24350838363170624,
+      "learning_rate": 3.7095667820576415e-06,
+      "loss": 0.0058,
+      "step": 5464
+    },
+    {
+      "epoch": 3.3252205658655307,
+      "grad_norm": 0.28811752796173096,
+      "learning_rate": 3.7030441807688193e-06,
+      "loss": 0.007,
+      "step": 5465
+    },
+    {
+      "epoch": 3.325829023425616,
+      "grad_norm": 0.1940048485994339,
+      "learning_rate": 3.69652686023734e-06,
+      "loss": 0.0038,
+      "step": 5466
+    },
+    {
+      "epoch": 3.3264374809857014,
+      "grad_norm": 0.2174157351255417,
+      "learning_rate": 3.6900148220792323e-06,
+      "loss": 0.0034,
+      "step": 5467
+    },
+    {
+      "epoch": 3.3270459385457865,
+      "grad_norm": 0.18492846190929413,
+      "learning_rate": 3.6835080679091998e-06,
+      "loss": 0.0027,
+      "step": 5468
+    },
+    {
+      "epoch": 3.3276543961058715,
+      "grad_norm": 0.30942097306251526,
+      "learning_rate": 3.6770065993406517e-06,
+      "loss": 0.0077,
+      "step": 5469
+    },
+    {
+      "epoch": 3.3282628536659566,
+      "grad_norm": 0.3265366852283478,
+      "learning_rate": 3.6705104179856785e-06,
+      "loss": 0.0071,
+      "step": 5470
+    },
+    {
+      "epoch": 3.3288713112260417,
+      "grad_norm": 0.2290954440832138,
+      "learning_rate": 3.6640195254550676e-06,
+      "loss": 0.0061,
+      "step": 5471
+    },
+    {
+      "epoch": 3.3294797687861273,
+      "grad_norm": 0.21861323714256287,
+      "learning_rate": 3.657533923358286e-06,
+      "loss": 0.0048,
+      "step": 5472
+    },
+    {
+      "epoch": 3.3300882263462124,
+      "grad_norm": 0.25013089179992676,
+      "learning_rate": 3.6510536133034985e-06,
+      "loss": 0.0059,
+      "step": 5473
+    },
+    {
+      "epoch": 3.3306966839062975,
+      "grad_norm": 0.26969480514526367,
+      "learning_rate": 3.644578596897541e-06,
+      "loss": 0.0078,
+      "step": 5474
+    },
+    {
+      "epoch": 3.3313051414663826,
+      "grad_norm": 0.23772546648979187,
+      "learning_rate": 3.638108875745955e-06,
+      "loss": 0.0055,
+      "step": 5475
+    },
+    {
+      "epoch": 3.3319135990264677,
+      "grad_norm": 0.245829239487648,
+      "learning_rate": 3.631644451452959e-06,
+      "loss": 0.007,
+      "step": 5476
+    },
+    {
+      "epoch": 3.332522056586553,
+      "grad_norm": 0.23014789819717407,
+      "learning_rate": 3.625185325621469e-06,
+      "loss": 0.0052,
+      "step": 5477
+    },
+    {
+      "epoch": 3.3331305141466383,
+      "grad_norm": 0.2805229425430298,
+      "learning_rate": 3.618731499853059e-06,
+      "loss": 0.0076,
+      "step": 5478
+    },
+    {
+      "epoch": 3.3337389717067234,
+      "grad_norm": 0.23805426061153412,
+      "learning_rate": 3.6122829757480354e-06,
+      "loss": 0.0048,
+      "step": 5479
+    },
+    {
+      "epoch": 3.3343474292668085,
+      "grad_norm": 0.3036494553089142,
+      "learning_rate": 3.605839754905341e-06,
+      "loss": 0.0084,
+      "step": 5480
+    },
+    {
+      "epoch": 3.3349558868268936,
+      "grad_norm": 0.2554338872432709,
+      "learning_rate": 3.5994018389226365e-06,
+      "loss": 0.0045,
+      "step": 5481
+    },
+    {
+      "epoch": 3.335564344386979,
+      "grad_norm": 0.26599669456481934,
+      "learning_rate": 3.5929692293962562e-06,
+      "loss": 0.0069,
+      "step": 5482
+    },
+    {
+      "epoch": 3.3361728019470642,
+      "grad_norm": 0.1806708723306656,
+      "learning_rate": 3.586541927921222e-06,
+      "loss": 0.003,
+      "step": 5483
+    },
+    {
+      "epoch": 3.3367812595071493,
+      "grad_norm": 0.21414224803447723,
+      "learning_rate": 3.5801199360912226e-06,
+      "loss": 0.0046,
+      "step": 5484
+    },
+    {
+      "epoch": 3.3373897170672344,
+      "grad_norm": 0.23333413898944855,
+      "learning_rate": 3.573703255498667e-06,
+      "loss": 0.0036,
+      "step": 5485
+    },
+    {
+      "epoch": 3.3379981746273195,
+      "grad_norm": 0.19560351967811584,
+      "learning_rate": 3.567291887734603e-06,
+      "loss": 0.0052,
+      "step": 5486
+    },
+    {
+      "epoch": 3.338606632187405,
+      "grad_norm": 0.48690447211265564,
+      "learning_rate": 3.5608858343887997e-06,
+      "loss": 0.0215,
+      "step": 5487
+    },
+    {
+      "epoch": 3.33921508974749,
+      "grad_norm": 0.27765342593193054,
+      "learning_rate": 3.55448509704967e-06,
+      "loss": 0.0068,
+      "step": 5488
+    },
+    {
+      "epoch": 3.3398235473075752,
+      "grad_norm": 0.2087118774652481,
+      "learning_rate": 3.548089677304356e-06,
+      "loss": 0.0043,
+      "step": 5489
+    },
+    {
+      "epoch": 3.3404320048676603,
+      "grad_norm": 0.3580057919025421,
+      "learning_rate": 3.5416995767386357e-06,
+      "loss": 0.0066,
+      "step": 5490
+    },
+    {
+      "epoch": 3.3410404624277454,
+      "grad_norm": 0.2551713287830353,
+      "learning_rate": 3.5353147969369948e-06,
+      "loss": 0.0061,
+      "step": 5491
+    },
+    {
+      "epoch": 3.341648919987831,
+      "grad_norm": 0.26948314905166626,
+      "learning_rate": 3.5289353394825947e-06,
+      "loss": 0.0082,
+      "step": 5492
+    },
+    {
+      "epoch": 3.342257377547916,
+      "grad_norm": 0.438873827457428,
+      "learning_rate": 3.5225612059572727e-06,
+      "loss": 0.0112,
+      "step": 5493
+    },
+    {
+      "epoch": 3.342865835108001,
+      "grad_norm": 0.18292073905467987,
+      "learning_rate": 3.5161923979415395e-06,
+      "loss": 0.0039,
+      "step": 5494
+    },
+    {
+      "epoch": 3.3434742926680863,
+      "grad_norm": 0.2990265488624573,
+      "learning_rate": 3.509828917014615e-06,
+      "loss": 0.0097,
+      "step": 5495
+    },
+    {
+      "epoch": 3.3440827502281714,
+      "grad_norm": 0.23416587710380554,
+      "learning_rate": 3.5034707647543576e-06,
+      "loss": 0.0061,
+      "step": 5496
+    },
+    {
+      "epoch": 3.344691207788257,
+      "grad_norm": 0.28342893719673157,
+      "learning_rate": 3.4971179427373295e-06,
+      "loss": 0.0072,
+      "step": 5497
+    },
+    {
+      "epoch": 3.345299665348342,
+      "grad_norm": 0.22702716290950775,
+      "learning_rate": 3.490770452538769e-06,
+      "loss": 0.0053,
+      "step": 5498
+    },
+    {
+      "epoch": 3.345908122908427,
+      "grad_norm": 0.24197997152805328,
+      "learning_rate": 3.484428295732592e-06,
+      "loss": 0.0071,
+      "step": 5499
+    },
+    {
+      "epoch": 3.346516580468512,
+      "grad_norm": 0.16110782325267792,
+      "learning_rate": 3.4780914738913816e-06,
+      "loss": 0.0035,
+      "step": 5500
+    },
+    {
+      "epoch": 3.3471250380285973,
+      "grad_norm": 0.3154269754886627,
+      "learning_rate": 3.471759988586404e-06,
+      "loss": 0.0057,
+      "step": 5501
+    },
+    {
+      "epoch": 3.347733495588683,
+      "grad_norm": 0.24822941422462463,
+      "learning_rate": 3.465433841387611e-06,
+      "loss": 0.0036,
+      "step": 5502
+    },
+    {
+      "epoch": 3.348341953148768,
+      "grad_norm": 0.26176613569259644,
+      "learning_rate": 3.4591130338636257e-06,
+      "loss": 0.0049,
+      "step": 5503
+    },
+    {
+      "epoch": 3.348950410708853,
+      "grad_norm": 0.31120842695236206,
+      "learning_rate": 3.4527975675817282e-06,
+      "loss": 0.0095,
+      "step": 5504
+    },
+    {
+      "epoch": 3.349558868268938,
+      "grad_norm": 0.20220082998275757,
+      "learning_rate": 3.4464874441079126e-06,
+      "loss": 0.005,
+      "step": 5505
+    },
+    {
+      "epoch": 3.350167325829023,
+      "grad_norm": 0.29147812724113464,
+      "learning_rate": 3.440182665006811e-06,
+      "loss": 0.0047,
+      "step": 5506
+    },
+    {
+      "epoch": 3.3507757833891088,
+      "grad_norm": 0.3253108561038971,
+      "learning_rate": 3.43388323184175e-06,
+      "loss": 0.01,
+      "step": 5507
+    },
+    {
+      "epoch": 3.351384240949194,
+      "grad_norm": 0.3161683976650238,
+      "learning_rate": 3.4275891461747283e-06,
+      "loss": 0.007,
+      "step": 5508
+    },
+    {
+      "epoch": 3.351992698509279,
+      "grad_norm": 0.3239922821521759,
+      "learning_rate": 3.421300409566422e-06,
+      "loss": 0.0054,
+      "step": 5509
+    },
+    {
+      "epoch": 3.352601156069364,
+      "grad_norm": 0.30600473284721375,
+      "learning_rate": 3.415017023576164e-06,
+      "loss": 0.0044,
+      "step": 5510
+    },
+    {
+      "epoch": 3.353209613629449,
+      "grad_norm": 0.2611445486545563,
+      "learning_rate": 3.408738989761978e-06,
+      "loss": 0.0062,
+      "step": 5511
+    },
+    {
+      "epoch": 3.3538180711895347,
+      "grad_norm": 0.1755179911851883,
+      "learning_rate": 3.4024663096805547e-06,
+      "loss": 0.0036,
+      "step": 5512
+    },
+    {
+      "epoch": 3.3544265287496198,
+      "grad_norm": 0.2571423649787903,
+      "learning_rate": 3.396198984887261e-06,
+      "loss": 0.0089,
+      "step": 5513
+    },
+    {
+      "epoch": 3.355034986309705,
+      "grad_norm": 0.23700016736984253,
+      "learning_rate": 3.389937016936118e-06,
+      "loss": 0.0037,
+      "step": 5514
+    },
+    {
+      "epoch": 3.35564344386979,
+      "grad_norm": 0.31217527389526367,
+      "learning_rate": 3.383680407379855e-06,
+      "loss": 0.0048,
+      "step": 5515
+    },
+    {
+      "epoch": 3.356251901429875,
+      "grad_norm": 0.2647400200366974,
+      "learning_rate": 3.377429157769832e-06,
+      "loss": 0.0099,
+      "step": 5516
+    },
+    {
+      "epoch": 3.3568603589899606,
+      "grad_norm": 0.199229896068573,
+      "learning_rate": 3.3711832696561056e-06,
+      "loss": 0.0034,
+      "step": 5517
+    },
+    {
+      "epoch": 3.3574688165500457,
+      "grad_norm": 0.24624145030975342,
+      "learning_rate": 3.3649427445873934e-06,
+      "loss": 0.0058,
+      "step": 5518
+    },
+    {
+      "epoch": 3.358077274110131,
+      "grad_norm": 0.21135321259498596,
+      "learning_rate": 3.3587075841110927e-06,
+      "loss": 0.0047,
+      "step": 5519
+    },
+    {
+      "epoch": 3.358685731670216,
+      "grad_norm": 0.31276053190231323,
+      "learning_rate": 3.3524777897732452e-06,
+      "loss": 0.0068,
+      "step": 5520
+    },
+    {
+      "epoch": 3.359294189230301,
+      "grad_norm": 0.20705646276474,
+      "learning_rate": 3.3462533631186066e-06,
+      "loss": 0.0041,
+      "step": 5521
+    },
+    {
+      "epoch": 3.3599026467903865,
+      "grad_norm": 0.373995840549469,
+      "learning_rate": 3.340034305690554e-06,
+      "loss": 0.0137,
+      "step": 5522
+    },
+    {
+      "epoch": 3.3605111043504716,
+      "grad_norm": 0.1533001810312271,
+      "learning_rate": 3.3338206190311667e-06,
+      "loss": 0.0022,
+      "step": 5523
+    },
+    {
+      "epoch": 3.3611195619105567,
+      "grad_norm": 0.3650040328502655,
+      "learning_rate": 3.327612304681166e-06,
+      "loss": 0.0084,
+      "step": 5524
+    },
+    {
+      "epoch": 3.361728019470642,
+      "grad_norm": 0.2500038146972656,
+      "learning_rate": 3.321409364179975e-06,
+      "loss": 0.0052,
+      "step": 5525
+    },
+    {
+      "epoch": 3.362336477030727,
+      "grad_norm": 0.23246996104717255,
+      "learning_rate": 3.3152117990656456e-06,
+      "loss": 0.0054,
+      "step": 5526
+    },
+    {
+      "epoch": 3.3629449345908125,
+      "grad_norm": 0.21729208528995514,
+      "learning_rate": 3.309019610874925e-06,
+      "loss": 0.0031,
+      "step": 5527
+    },
+    {
+      "epoch": 3.3635533921508975,
+      "grad_norm": 0.29215723276138306,
+      "learning_rate": 3.3028328011432157e-06,
+      "loss": 0.0071,
+      "step": 5528
+    },
+    {
+      "epoch": 3.3641618497109826,
+      "grad_norm": 0.21962468326091766,
+      "learning_rate": 3.2966513714045967e-06,
+      "loss": 0.0037,
+      "step": 5529
+    },
+    {
+      "epoch": 3.3647703072710677,
+      "grad_norm": 0.24007774889469147,
+      "learning_rate": 3.2904753231917857e-06,
+      "loss": 0.0048,
+      "step": 5530
+    },
+    {
+      "epoch": 3.365378764831153,
+      "grad_norm": 0.2580210864543915,
+      "learning_rate": 3.284304658036208e-06,
+      "loss": 0.0049,
+      "step": 5531
+    },
+    {
+      "epoch": 3.3659872223912384,
+      "grad_norm": 0.1776011735200882,
+      "learning_rate": 3.278139377467912e-06,
+      "loss": 0.0027,
+      "step": 5532
+    },
+    {
+      "epoch": 3.3665956799513235,
+      "grad_norm": 0.22608719766139984,
+      "learning_rate": 3.271979483015647e-06,
+      "loss": 0.0057,
+      "step": 5533
+    },
+    {
+      "epoch": 3.3672041375114086,
+      "grad_norm": 0.29510682821273804,
+      "learning_rate": 3.265824976206791e-06,
+      "loss": 0.0056,
+      "step": 5534
+    },
+    {
+      "epoch": 3.3678125950714937,
+      "grad_norm": 0.2816297113895416,
+      "learning_rate": 3.2596758585674238e-06,
+      "loss": 0.0058,
+      "step": 5535
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.21519798040390015,
+      "learning_rate": 3.253532131622261e-06,
+      "loss": 0.0077,
+      "step": 5536
+    },
+    {
+      "epoch": 3.3690295101916643,
+      "grad_norm": 0.1678488701581955,
+      "learning_rate": 3.247393796894688e-06,
+      "loss": 0.003,
+      "step": 5537
+    },
+    {
+      "epoch": 3.3696379677517494,
+      "grad_norm": 0.26176977157592773,
+      "learning_rate": 3.2412608559067638e-06,
+      "loss": 0.0042,
+      "step": 5538
+    },
+    {
+      "epoch": 3.3702464253118345,
+      "grad_norm": 0.2218218296766281,
+      "learning_rate": 3.2351333101792005e-06,
+      "loss": 0.0036,
+      "step": 5539
+    },
+    {
+      "epoch": 3.3708548828719196,
+      "grad_norm": 0.2444075644016266,
+      "learning_rate": 3.229011161231363e-06,
+      "loss": 0.0051,
+      "step": 5540
+    },
+    {
+      "epoch": 3.3714633404320047,
+      "grad_norm": 0.18922793865203857,
+      "learning_rate": 3.2228944105813093e-06,
+      "loss": 0.0036,
+      "step": 5541
+    },
+    {
+      "epoch": 3.3720717979920902,
+      "grad_norm": 0.24314537644386292,
+      "learning_rate": 3.2167830597457205e-06,
+      "loss": 0.0046,
+      "step": 5542
+    },
+    {
+      "epoch": 3.3726802555521753,
+      "grad_norm": 0.12830045819282532,
+      "learning_rate": 3.2106771102399625e-06,
+      "loss": 0.0019,
+      "step": 5543
+    },
+    {
+      "epoch": 3.3732887131122604,
+      "grad_norm": 0.26987224817276,
+      "learning_rate": 3.20457656357806e-06,
+      "loss": 0.0062,
+      "step": 5544
+    },
+    {
+      "epoch": 3.3738971706723455,
+      "grad_norm": 0.21220239996910095,
+      "learning_rate": 3.198481421272698e-06,
+      "loss": 0.0038,
+      "step": 5545
+    },
+    {
+      "epoch": 3.3745056282324306,
+      "grad_norm": 0.2044731080532074,
+      "learning_rate": 3.1923916848352025e-06,
+      "loss": 0.0044,
+      "step": 5546
+    },
+    {
+      "epoch": 3.375114085792516,
+      "grad_norm": 0.3579948842525482,
+      "learning_rate": 3.186307355775586e-06,
+      "loss": 0.0061,
+      "step": 5547
+    },
+    {
+      "epoch": 3.3757225433526012,
+      "grad_norm": 0.6677601337432861,
+      "learning_rate": 3.180228435602503e-06,
+      "loss": 0.0061,
+      "step": 5548
+    },
+    {
+      "epoch": 3.3763310009126863,
+      "grad_norm": 0.21300187706947327,
+      "learning_rate": 3.1741549258232833e-06,
+      "loss": 0.005,
+      "step": 5549
+    },
+    {
+      "epoch": 3.3769394584727714,
+      "grad_norm": 0.24835187196731567,
+      "learning_rate": 3.1680868279438835e-06,
+      "loss": 0.0057,
+      "step": 5550
+    },
+    {
+      "epoch": 3.3775479160328565,
+      "grad_norm": 0.4124807119369507,
+      "learning_rate": 3.162024143468964e-06,
+      "loss": 0.0088,
+      "step": 5551
+    },
+    {
+      "epoch": 3.378156373592942,
+      "grad_norm": 0.2930738925933838,
+      "learning_rate": 3.1559668739017974e-06,
+      "loss": 0.0062,
+      "step": 5552
+    },
+    {
+      "epoch": 3.378764831153027,
+      "grad_norm": 0.18829555809497833,
+      "learning_rate": 3.149915020744343e-06,
+      "loss": 0.0042,
+      "step": 5553
+    },
+    {
+      "epoch": 3.3793732887131123,
+      "grad_norm": 0.24336494505405426,
+      "learning_rate": 3.143868585497206e-06,
+      "loss": 0.0039,
+      "step": 5554
+    },
+    {
+      "epoch": 3.3799817462731974,
+      "grad_norm": 0.2603733241558075,
+      "learning_rate": 3.1378275696596597e-06,
+      "loss": 0.0061,
+      "step": 5555
+    },
+    {
+      "epoch": 3.3805902038332825,
+      "grad_norm": 0.17050331830978394,
+      "learning_rate": 3.131791974729609e-06,
+      "loss": 0.0035,
+      "step": 5556
+    },
+    {
+      "epoch": 3.381198661393368,
+      "grad_norm": 0.263378381729126,
+      "learning_rate": 3.125761802203636e-06,
+      "loss": 0.0063,
+      "step": 5557
+    },
+    {
+      "epoch": 3.381807118953453,
+      "grad_norm": 0.3855268359184265,
+      "learning_rate": 3.119737053576971e-06,
+      "loss": 0.0063,
+      "step": 5558
+    },
+    {
+      "epoch": 3.382415576513538,
+      "grad_norm": 0.17356686294078827,
+      "learning_rate": 3.11371773034351e-06,
+      "loss": 0.0025,
+      "step": 5559
+    },
+    {
+      "epoch": 3.3830240340736233,
+      "grad_norm": 0.2114478200674057,
+      "learning_rate": 3.107703833995776e-06,
+      "loss": 0.0054,
+      "step": 5560
+    },
+    {
+      "epoch": 3.3836324916337084,
+      "grad_norm": 0.31965020298957825,
+      "learning_rate": 3.101695366024987e-06,
+      "loss": 0.0067,
+      "step": 5561
+    },
+    {
+      "epoch": 3.384240949193794,
+      "grad_norm": 0.2662268579006195,
+      "learning_rate": 3.095692327920974e-06,
+      "loss": 0.0069,
+      "step": 5562
+    },
+    {
+      "epoch": 3.384849406753879,
+      "grad_norm": 0.204771026968956,
+      "learning_rate": 3.0896947211722484e-06,
+      "loss": 0.004,
+      "step": 5563
+    },
+    {
+      "epoch": 3.385457864313964,
+      "grad_norm": 0.22940324246883392,
+      "learning_rate": 3.083702547265965e-06,
+      "loss": 0.0058,
+      "step": 5564
+    },
+    {
+      "epoch": 3.386066321874049,
+      "grad_norm": 0.2345285266637802,
+      "learning_rate": 3.077715807687939e-06,
+      "loss": 0.0132,
+      "step": 5565
+    },
+    {
+      "epoch": 3.3866747794341343,
+      "grad_norm": 1.2732864618301392,
+      "learning_rate": 3.071734503922616e-06,
+      "loss": 0.0262,
+      "step": 5566
+    },
+    {
+      "epoch": 3.38728323699422,
+      "grad_norm": 0.24260850250720978,
+      "learning_rate": 3.065758637453131e-06,
+      "loss": 0.0033,
+      "step": 5567
+    },
+    {
+      "epoch": 3.387891694554305,
+      "grad_norm": 0.28996798396110535,
+      "learning_rate": 3.059788209761233e-06,
+      "loss": 0.0037,
+      "step": 5568
+    },
+    {
+      "epoch": 3.38850015211439,
+      "grad_norm": 0.2517380714416504,
+      "learning_rate": 3.0538232223273482e-06,
+      "loss": 0.0056,
+      "step": 5569
+    },
+    {
+      "epoch": 3.389108609674475,
+      "grad_norm": 0.2352709025144577,
+      "learning_rate": 3.047863676630541e-06,
+      "loss": 0.0056,
+      "step": 5570
+    },
+    {
+      "epoch": 3.3897170672345602,
+      "grad_norm": 0.24700261652469635,
+      "learning_rate": 3.041909574148538e-06,
+      "loss": 0.0038,
+      "step": 5571
+    },
+    {
+      "epoch": 3.3903255247946458,
+      "grad_norm": 0.20761002600193024,
+      "learning_rate": 3.035960916357697e-06,
+      "loss": 0.0046,
+      "step": 5572
+    },
+    {
+      "epoch": 3.390933982354731,
+      "grad_norm": 0.24791277945041656,
+      "learning_rate": 3.030017704733043e-06,
+      "loss": 0.0048,
+      "step": 5573
+    },
+    {
+      "epoch": 3.391542439914816,
+      "grad_norm": 0.20745399594306946,
+      "learning_rate": 3.0240799407482452e-06,
+      "loss": 0.0044,
+      "step": 5574
+    },
+    {
+      "epoch": 3.392150897474901,
+      "grad_norm": 0.14504899084568024,
+      "learning_rate": 3.018147625875617e-06,
+      "loss": 0.002,
+      "step": 5575
+    },
+    {
+      "epoch": 3.392759355034986,
+      "grad_norm": 0.3044413626194,
+      "learning_rate": 3.012220761586132e-06,
+      "loss": 0.0064,
+      "step": 5576
+    },
+    {
+      "epoch": 3.3933678125950717,
+      "grad_norm": 0.2477070391178131,
+      "learning_rate": 3.006299349349406e-06,
+      "loss": 0.0065,
+      "step": 5577
+    },
+    {
+      "epoch": 3.393976270155157,
+      "grad_norm": 0.28918352723121643,
+      "learning_rate": 3.000383390633696e-06,
+      "loss": 0.0062,
+      "step": 5578
+    },
+    {
+      "epoch": 3.394584727715242,
+      "grad_norm": 0.2834755480289459,
+      "learning_rate": 2.9944728869059136e-06,
+      "loss": 0.0063,
+      "step": 5579
+    },
+    {
+      "epoch": 3.395193185275327,
+      "grad_norm": 0.2524814009666443,
+      "learning_rate": 2.98856783963162e-06,
+      "loss": 0.0038,
+      "step": 5580
+    },
+    {
+      "epoch": 3.395801642835412,
+      "grad_norm": 0.2619141638278961,
+      "learning_rate": 2.98266825027502e-06,
+      "loss": 0.0057,
+      "step": 5581
+    },
+    {
+      "epoch": 3.3964101003954976,
+      "grad_norm": 0.2823928892612457,
+      "learning_rate": 2.9767741202989723e-06,
+      "loss": 0.0061,
+      "step": 5582
+    },
+    {
+      "epoch": 3.3970185579555827,
+      "grad_norm": 0.20528404414653778,
+      "learning_rate": 2.970885451164965e-06,
+      "loss": 0.0046,
+      "step": 5583
+    },
+    {
+      "epoch": 3.397627015515668,
+      "grad_norm": 0.2534896433353424,
+      "learning_rate": 2.9650022443331453e-06,
+      "loss": 0.0045,
+      "step": 5584
+    },
+    {
+      "epoch": 3.398235473075753,
+      "grad_norm": 0.26259493827819824,
+      "learning_rate": 2.9591245012623058e-06,
+      "loss": 0.0071,
+      "step": 5585
+    },
+    {
+      "epoch": 3.398843930635838,
+      "grad_norm": 0.3380330204963684,
+      "learning_rate": 2.9532522234098803e-06,
+      "loss": 0.0089,
+      "step": 5586
+    },
+    {
+      "epoch": 3.3994523881959235,
+      "grad_norm": 0.18706291913986206,
+      "learning_rate": 2.947385412231951e-06,
+      "loss": 0.0031,
+      "step": 5587
+    },
+    {
+      "epoch": 3.4000608457560086,
+      "grad_norm": 0.2913658916950226,
+      "learning_rate": 2.9415240691832463e-06,
+      "loss": 0.0085,
+      "step": 5588
+    },
+    {
+      "epoch": 3.4006693033160937,
+      "grad_norm": 0.3028022050857544,
+      "learning_rate": 2.9356681957171227e-06,
+      "loss": 0.0066,
+      "step": 5589
+    },
+    {
+      "epoch": 3.401277760876179,
+      "grad_norm": 0.2161863148212433,
+      "learning_rate": 2.9298177932856025e-06,
+      "loss": 0.0045,
+      "step": 5590
+    },
+    {
+      "epoch": 3.401886218436264,
+      "grad_norm": 0.25855207443237305,
+      "learning_rate": 2.923972863339336e-06,
+      "loss": 0.007,
+      "step": 5591
+    },
+    {
+      "epoch": 3.4024946759963495,
+      "grad_norm": 0.2582375109195709,
+      "learning_rate": 2.9181334073276334e-06,
+      "loss": 0.0055,
+      "step": 5592
+    },
+    {
+      "epoch": 3.4031031335564346,
+      "grad_norm": 0.24739745259284973,
+      "learning_rate": 2.9122994266984226e-06,
+      "loss": 0.0048,
+      "step": 5593
+    },
+    {
+      "epoch": 3.4037115911165197,
+      "grad_norm": 0.2663595676422119,
+      "learning_rate": 2.906470922898291e-06,
+      "loss": 0.0044,
+      "step": 5594
+    },
+    {
+      "epoch": 3.4043200486766048,
+      "grad_norm": 0.21414734423160553,
+      "learning_rate": 2.900647897372469e-06,
+      "loss": 0.0041,
+      "step": 5595
+    },
+    {
+      "epoch": 3.40492850623669,
+      "grad_norm": 0.33169928193092346,
+      "learning_rate": 2.894830351564823e-06,
+      "loss": 0.0088,
+      "step": 5596
+    },
+    {
+      "epoch": 3.4055369637967754,
+      "grad_norm": 0.3833865523338318,
+      "learning_rate": 2.889018286917858e-06,
+      "loss": 0.0072,
+      "step": 5597
+    },
+    {
+      "epoch": 3.4061454213568605,
+      "grad_norm": 0.30657726526260376,
+      "learning_rate": 2.8832117048727348e-06,
+      "loss": 0.009,
+      "step": 5598
+    },
+    {
+      "epoch": 3.4067538789169456,
+      "grad_norm": 0.1277196705341339,
+      "learning_rate": 2.877410606869227e-06,
+      "loss": 0.0017,
+      "step": 5599
+    },
+    {
+      "epoch": 3.4073623364770307,
+      "grad_norm": 0.29806557297706604,
+      "learning_rate": 2.8716149943457755e-06,
+      "loss": 0.0072,
+      "step": 5600
+    },
+    {
+      "epoch": 3.407970794037116,
+      "grad_norm": 0.24409450590610504,
+      "learning_rate": 2.86582486873945e-06,
+      "loss": 0.0052,
+      "step": 5601
+    },
+    {
+      "epoch": 3.4085792515972013,
+      "grad_norm": 0.23363740742206573,
+      "learning_rate": 2.8600402314859636e-06,
+      "loss": 0.0072,
+      "step": 5602
+    },
+    {
+      "epoch": 3.4091877091572864,
+      "grad_norm": 0.23178331553936005,
+      "learning_rate": 2.8542610840196493e-06,
+      "loss": 0.0055,
+      "step": 5603
+    },
+    {
+      "epoch": 3.4097961667173715,
+      "grad_norm": 0.2069847732782364,
+      "learning_rate": 2.848487427773519e-06,
+      "loss": 0.0039,
+      "step": 5604
+    },
+    {
+      "epoch": 3.4104046242774566,
+      "grad_norm": 0.17575199902057648,
+      "learning_rate": 2.842719264179178e-06,
+      "loss": 0.0028,
+      "step": 5605
+    },
+    {
+      "epoch": 3.4110130818375417,
+      "grad_norm": 0.1965407133102417,
+      "learning_rate": 2.836956594666901e-06,
+      "loss": 0.0036,
+      "step": 5606
+    },
+    {
+      "epoch": 3.4116215393976272,
+      "grad_norm": 0.25646457076072693,
+      "learning_rate": 2.8311994206655867e-06,
+      "loss": 0.0067,
+      "step": 5607
+    },
+    {
+      "epoch": 3.4122299969577123,
+      "grad_norm": 0.2868242859840393,
+      "learning_rate": 2.8254477436027797e-06,
+      "loss": 0.0063,
+      "step": 5608
+    },
+    {
+      "epoch": 3.4128384545177974,
+      "grad_norm": 0.277547150850296,
+      "learning_rate": 2.8197015649046393e-06,
+      "loss": 0.0067,
+      "step": 5609
+    },
+    {
+      "epoch": 3.4134469120778825,
+      "grad_norm": 0.20469379425048828,
+      "learning_rate": 2.813960885996003e-06,
+      "loss": 0.0029,
+      "step": 5610
+    },
+    {
+      "epoch": 3.4140553696379676,
+      "grad_norm": 0.39242902398109436,
+      "learning_rate": 2.808225708300302e-06,
+      "loss": 0.004,
+      "step": 5611
+    },
+    {
+      "epoch": 3.414663827198053,
+      "grad_norm": 0.2715418338775635,
+      "learning_rate": 2.8024960332396266e-06,
+      "loss": 0.0084,
+      "step": 5612
+    },
+    {
+      "epoch": 3.4152722847581383,
+      "grad_norm": 0.1898062378168106,
+      "learning_rate": 2.7967718622346957e-06,
+      "loss": 0.0042,
+      "step": 5613
+    },
+    {
+      "epoch": 3.4158807423182234,
+      "grad_norm": 0.2265051305294037,
+      "learning_rate": 2.7910531967048736e-06,
+      "loss": 0.0029,
+      "step": 5614
+    },
+    {
+      "epoch": 3.4164891998783085,
+      "grad_norm": 0.3068977892398834,
+      "learning_rate": 2.7853400380681404e-06,
+      "loss": 0.0095,
+      "step": 5615
+    },
+    {
+      "epoch": 3.4170976574383936,
+      "grad_norm": 0.22495239973068237,
+      "learning_rate": 2.779632387741121e-06,
+      "loss": 0.0059,
+      "step": 5616
+    },
+    {
+      "epoch": 3.417706114998479,
+      "grad_norm": 0.2805343568325043,
+      "learning_rate": 2.7739302471390836e-06,
+      "loss": 0.0073,
+      "step": 5617
+    },
+    {
+      "epoch": 3.418314572558564,
+      "grad_norm": 0.2405698299407959,
+      "learning_rate": 2.7682336176759195e-06,
+      "loss": 0.0081,
+      "step": 5618
+    },
+    {
+      "epoch": 3.4189230301186493,
+      "grad_norm": 0.3347167372703552,
+      "learning_rate": 2.7625425007641425e-06,
+      "loss": 0.0061,
+      "step": 5619
+    },
+    {
+      "epoch": 3.4195314876787344,
+      "grad_norm": 0.25292569398880005,
+      "learning_rate": 2.756856897814933e-06,
+      "loss": 0.0041,
+      "step": 5620
+    },
+    {
+      "epoch": 3.4201399452388195,
+      "grad_norm": 0.1868637651205063,
+      "learning_rate": 2.7511768102380654e-06,
+      "loss": 0.0028,
+      "step": 5621
+    },
+    {
+      "epoch": 3.420748402798905,
+      "grad_norm": 0.175969198346138,
+      "learning_rate": 2.7455022394419746e-06,
+      "loss": 0.0043,
+      "step": 5622
+    },
+    {
+      "epoch": 3.42135686035899,
+      "grad_norm": 0.31663379073143005,
+      "learning_rate": 2.739833186833715e-06,
+      "loss": 0.0037,
+      "step": 5623
+    },
+    {
+      "epoch": 3.421965317919075,
+      "grad_norm": 0.3171955347061157,
+      "learning_rate": 2.73416965381898e-06,
+      "loss": 0.0114,
+      "step": 5624
+    },
+    {
+      "epoch": 3.4225737754791603,
+      "grad_norm": 0.22550025582313538,
+      "learning_rate": 2.728511641802076e-06,
+      "loss": 0.0033,
+      "step": 5625
+    },
+    {
+      "epoch": 3.4231822330392454,
+      "grad_norm": 0.16939371824264526,
+      "learning_rate": 2.722859152185972e-06,
+      "loss": 0.0042,
+      "step": 5626
+    },
+    {
+      "epoch": 3.423790690599331,
+      "grad_norm": 0.3074353337287903,
+      "learning_rate": 2.7172121863722366e-06,
+      "loss": 0.0072,
+      "step": 5627
+    },
+    {
+      "epoch": 3.424399148159416,
+      "grad_norm": 0.3095559775829315,
+      "learning_rate": 2.711570745761094e-06,
+      "loss": 0.0081,
+      "step": 5628
+    },
+    {
+      "epoch": 3.425007605719501,
+      "grad_norm": 0.24687674641609192,
+      "learning_rate": 2.7059348317513665e-06,
+      "loss": 0.0073,
+      "step": 5629
+    },
+    {
+      "epoch": 3.4256160632795862,
+      "grad_norm": 0.28120920062065125,
+      "learning_rate": 2.700304445740551e-06,
+      "loss": 0.0062,
+      "step": 5630
+    },
+    {
+      "epoch": 3.4262245208396713,
+      "grad_norm": 0.18883898854255676,
+      "learning_rate": 2.6946795891247266e-06,
+      "loss": 0.0037,
+      "step": 5631
+    },
+    {
+      "epoch": 3.4268329783997564,
+      "grad_norm": 0.21660275757312775,
+      "learning_rate": 2.6890602632986373e-06,
+      "loss": 0.0056,
+      "step": 5632
+    },
+    {
+      "epoch": 3.427441435959842,
+      "grad_norm": 0.2389414757490158,
+      "learning_rate": 2.6834464696556343e-06,
+      "loss": 0.0043,
+      "step": 5633
+    },
+    {
+      "epoch": 3.428049893519927,
+      "grad_norm": 0.23681369423866272,
+      "learning_rate": 2.6778382095877104e-06,
+      "loss": 0.0051,
+      "step": 5634
+    },
+    {
+      "epoch": 3.428658351080012,
+      "grad_norm": 0.251058965921402,
+      "learning_rate": 2.6722354844854693e-06,
+      "loss": 0.0031,
+      "step": 5635
+    },
+    {
+      "epoch": 3.4292668086400973,
+      "grad_norm": 0.27185845375061035,
+      "learning_rate": 2.666638295738169e-06,
+      "loss": 0.0058,
+      "step": 5636
+    },
+    {
+      "epoch": 3.4298752662001823,
+      "grad_norm": 0.2628857493400574,
+      "learning_rate": 2.6610466447336697e-06,
+      "loss": 0.0045,
+      "step": 5637
+    },
+    {
+      "epoch": 3.430483723760268,
+      "grad_norm": 0.17281438410282135,
+      "learning_rate": 2.65546053285847e-06,
+      "loss": 0.0036,
+      "step": 5638
+    },
+    {
+      "epoch": 3.431092181320353,
+      "grad_norm": 0.16873584687709808,
+      "learning_rate": 2.649879961497684e-06,
+      "loss": 0.0028,
+      "step": 5639
+    },
+    {
+      "epoch": 3.431700638880438,
+      "grad_norm": 0.23468784987926483,
+      "learning_rate": 2.64430493203508e-06,
+      "loss": 0.0062,
+      "step": 5640
+    },
+    {
+      "epoch": 3.432309096440523,
+      "grad_norm": 0.20407533645629883,
+      "learning_rate": 2.6387354458530134e-06,
+      "loss": 0.0034,
+      "step": 5641
+    },
+    {
+      "epoch": 3.4329175540006083,
+      "grad_norm": 0.17050844430923462,
+      "learning_rate": 2.6331715043324935e-06,
+      "loss": 0.0037,
+      "step": 5642
+    },
+    {
+      "epoch": 3.433526011560694,
+      "grad_norm": 0.26065853238105774,
+      "learning_rate": 2.627613108853147e-06,
+      "loss": 0.004,
+      "step": 5643
+    },
+    {
+      "epoch": 3.434134469120779,
+      "grad_norm": 0.31893685460090637,
+      "learning_rate": 2.622060260793227e-06,
+      "loss": 0.0089,
+      "step": 5644
+    },
+    {
+      "epoch": 3.434742926680864,
+      "grad_norm": 0.16491414606571198,
+      "learning_rate": 2.616512961529591e-06,
+      "loss": 0.0041,
+      "step": 5645
+    },
+    {
+      "epoch": 3.435351384240949,
+      "grad_norm": 0.30626043677330017,
+      "learning_rate": 2.610971212437763e-06,
+      "loss": 0.0042,
+      "step": 5646
+    },
+    {
+      "epoch": 3.435959841801034,
+      "grad_norm": 0.3586437702178955,
+      "learning_rate": 2.6054350148918493e-06,
+      "loss": 0.0068,
+      "step": 5647
+    },
+    {
+      "epoch": 3.4365682993611197,
+      "grad_norm": 0.2317257821559906,
+      "learning_rate": 2.5999043702646e-06,
+      "loss": 0.0054,
+      "step": 5648
+    },
+    {
+      "epoch": 3.437176756921205,
+      "grad_norm": 0.2033744603395462,
+      "learning_rate": 2.5943792799273838e-06,
+      "loss": 0.0042,
+      "step": 5649
+    },
+    {
+      "epoch": 3.43778521448129,
+      "grad_norm": 0.18324805796146393,
+      "learning_rate": 2.5888597452501994e-06,
+      "loss": 0.0049,
+      "step": 5650
+    },
+    {
+      "epoch": 3.438393672041375,
+      "grad_norm": 0.2906474769115448,
+      "learning_rate": 2.5833457676016526e-06,
+      "loss": 0.0073,
+      "step": 5651
+    },
+    {
+      "epoch": 3.43900212960146,
+      "grad_norm": 0.16437779366970062,
+      "learning_rate": 2.5778373483489827e-06,
+      "loss": 0.0042,
+      "step": 5652
+    },
+    {
+      "epoch": 3.4396105871615457,
+      "grad_norm": 0.2783378064632416,
+      "learning_rate": 2.572334488858047e-06,
+      "loss": 0.009,
+      "step": 5653
+    },
+    {
+      "epoch": 3.4402190447216308,
+      "grad_norm": 0.21060001850128174,
+      "learning_rate": 2.566837190493332e-06,
+      "loss": 0.0041,
+      "step": 5654
+    },
+    {
+      "epoch": 3.440827502281716,
+      "grad_norm": 0.2840571403503418,
+      "learning_rate": 2.561345454617925e-06,
+      "loss": 0.0058,
+      "step": 5655
+    },
+    {
+      "epoch": 3.441435959841801,
+      "grad_norm": 0.20874862372875214,
+      "learning_rate": 2.5558592825935645e-06,
+      "loss": 0.0034,
+      "step": 5656
+    },
+    {
+      "epoch": 3.442044417401886,
+      "grad_norm": 0.237037792801857,
+      "learning_rate": 2.5503786757805794e-06,
+      "loss": 0.0045,
+      "step": 5657
+    },
+    {
+      "epoch": 3.4426528749619716,
+      "grad_norm": 0.25760436058044434,
+      "learning_rate": 2.5449036355379347e-06,
+      "loss": 0.0053,
+      "step": 5658
+    },
+    {
+      "epoch": 3.4432613325220567,
+      "grad_norm": 0.2317677140235901,
+      "learning_rate": 2.539434163223217e-06,
+      "loss": 0.0052,
+      "step": 5659
+    },
+    {
+      "epoch": 3.443869790082142,
+      "grad_norm": 0.3090670704841614,
+      "learning_rate": 2.533970260192628e-06,
+      "loss": 0.0103,
+      "step": 5660
+    },
+    {
+      "epoch": 3.444478247642227,
+      "grad_norm": 0.23244629800319672,
+      "learning_rate": 2.528511927800978e-06,
+      "loss": 0.0052,
+      "step": 5661
+    },
+    {
+      "epoch": 3.445086705202312,
+      "grad_norm": 0.20796987414360046,
+      "learning_rate": 2.5230591674017145e-06,
+      "loss": 0.0032,
+      "step": 5662
+    },
+    {
+      "epoch": 3.4456951627623975,
+      "grad_norm": 0.2811738848686218,
+      "learning_rate": 2.51761198034689e-06,
+      "loss": 0.007,
+      "step": 5663
+    },
+    {
+      "epoch": 3.4463036203224826,
+      "grad_norm": 0.2699826955795288,
+      "learning_rate": 2.5121703679871907e-06,
+      "loss": 0.0047,
+      "step": 5664
+    },
+    {
+      "epoch": 3.4469120778825677,
+      "grad_norm": 0.21873120963573456,
+      "learning_rate": 2.5067343316718866e-06,
+      "loss": 0.0041,
+      "step": 5665
+    },
+    {
+      "epoch": 3.447520535442653,
+      "grad_norm": 0.21390405297279358,
+      "learning_rate": 2.501303872748917e-06,
+      "loss": 0.0045,
+      "step": 5666
+    },
+    {
+      "epoch": 3.448128993002738,
+      "grad_norm": 0.18419447541236877,
+      "learning_rate": 2.4958789925647873e-06,
+      "loss": 0.0044,
+      "step": 5667
+    },
+    {
+      "epoch": 3.4487374505628234,
+      "grad_norm": 0.19194208085536957,
+      "learning_rate": 2.49045969246465e-06,
+      "loss": 0.0049,
+      "step": 5668
+    },
+    {
+      "epoch": 3.4493459081229085,
+      "grad_norm": 0.18368539214134216,
+      "learning_rate": 2.485045973792266e-06,
+      "loss": 0.0046,
+      "step": 5669
+    },
+    {
+      "epoch": 3.4499543656829936,
+      "grad_norm": 0.33034926652908325,
+      "learning_rate": 2.4796378378900142e-06,
+      "loss": 0.0104,
+      "step": 5670
+    },
+    {
+      "epoch": 3.4505628232430787,
+      "grad_norm": 0.23672623932361603,
+      "learning_rate": 2.474235286098878e-06,
+      "loss": 0.0038,
+      "step": 5671
+    },
+    {
+      "epoch": 3.451171280803164,
+      "grad_norm": 0.23982059955596924,
+      "learning_rate": 2.4688383197584762e-06,
+      "loss": 0.0046,
+      "step": 5672
+    },
+    {
+      "epoch": 3.451779738363249,
+      "grad_norm": 0.3084208369255066,
+      "learning_rate": 2.4634469402070233e-06,
+      "loss": 0.0091,
+      "step": 5673
+    },
+    {
+      "epoch": 3.4523881959233345,
+      "grad_norm": 0.14161497354507446,
+      "learning_rate": 2.458061148781363e-06,
+      "loss": 0.0026,
+      "step": 5674
+    },
+    {
+      "epoch": 3.4529966534834196,
+      "grad_norm": 0.3343401253223419,
+      "learning_rate": 2.452680946816932e-06,
+      "loss": 0.0078,
+      "step": 5675
+    },
+    {
+      "epoch": 3.4536051110435046,
+      "grad_norm": 0.23120488226413727,
+      "learning_rate": 2.4473063356478198e-06,
+      "loss": 0.0054,
+      "step": 5676
+    },
+    {
+      "epoch": 3.4542135686035897,
+      "grad_norm": 0.2343619167804718,
+      "learning_rate": 2.4419373166066904e-06,
+      "loss": 0.0056,
+      "step": 5677
+    },
+    {
+      "epoch": 3.454822026163675,
+      "grad_norm": 0.37773486971855164,
+      "learning_rate": 2.4365738910248375e-06,
+      "loss": 0.0102,
+      "step": 5678
+    },
+    {
+      "epoch": 3.4554304837237604,
+      "grad_norm": 0.19855354726314545,
+      "learning_rate": 2.431216060232169e-06,
+      "loss": 0.0056,
+      "step": 5679
+    },
+    {
+      "epoch": 3.4560389412838455,
+      "grad_norm": 0.2364191859960556,
+      "learning_rate": 2.425863825557212e-06,
+      "loss": 0.0064,
+      "step": 5680
+    },
+    {
+      "epoch": 3.4566473988439306,
+      "grad_norm": 0.4233652353286743,
+      "learning_rate": 2.420517188327079e-06,
+      "loss": 0.0106,
+      "step": 5681
+    },
+    {
+      "epoch": 3.4572558564040157,
+      "grad_norm": 0.2906593084335327,
+      "learning_rate": 2.4151761498675345e-06,
+      "loss": 0.0053,
+      "step": 5682
+    },
+    {
+      "epoch": 3.4578643139641008,
+      "grad_norm": 0.2784557044506073,
+      "learning_rate": 2.409840711502917e-06,
+      "loss": 0.0079,
+      "step": 5683
+    },
+    {
+      "epoch": 3.4584727715241863,
+      "grad_norm": 0.2974449396133423,
+      "learning_rate": 2.4045108745561985e-06,
+      "loss": 0.0071,
+      "step": 5684
+    },
+    {
+      "epoch": 3.4590812290842714,
+      "grad_norm": 0.3297038674354553,
+      "learning_rate": 2.3991866403489577e-06,
+      "loss": 0.0073,
+      "step": 5685
+    },
+    {
+      "epoch": 3.4596896866443565,
+      "grad_norm": 0.2977392077445984,
+      "learning_rate": 2.393868010201386e-06,
+      "loss": 0.0072,
+      "step": 5686
+    },
+    {
+      "epoch": 3.4602981442044416,
+      "grad_norm": 0.3237118721008301,
+      "learning_rate": 2.388554985432273e-06,
+      "loss": 0.005,
+      "step": 5687
+    },
+    {
+      "epoch": 3.4609066017645267,
+      "grad_norm": 0.23049971461296082,
+      "learning_rate": 2.3832475673590316e-06,
+      "loss": 0.0063,
+      "step": 5688
+    },
+    {
+      "epoch": 3.4615150593246122,
+      "grad_norm": 0.2660450339317322,
+      "learning_rate": 2.377945757297681e-06,
+      "loss": 0.0061,
+      "step": 5689
+    },
+    {
+      "epoch": 3.4621235168846973,
+      "grad_norm": 0.2178175449371338,
+      "learning_rate": 2.3726495565628506e-06,
+      "loss": 0.0058,
+      "step": 5690
+    },
+    {
+      "epoch": 3.4627319744447824,
+      "grad_norm": 0.47303879261016846,
+      "learning_rate": 2.3673589664677727e-06,
+      "loss": 0.0134,
+      "step": 5691
+    },
+    {
+      "epoch": 3.4633404320048675,
+      "grad_norm": 0.2600957453250885,
+      "learning_rate": 2.362073988324304e-06,
+      "loss": 0.0049,
+      "step": 5692
+    },
+    {
+      "epoch": 3.4639488895649526,
+      "grad_norm": 0.2990921139717102,
+      "learning_rate": 2.3567946234428844e-06,
+      "loss": 0.0068,
+      "step": 5693
+    },
+    {
+      "epoch": 3.464557347125038,
+      "grad_norm": 0.19195446372032166,
+      "learning_rate": 2.3515208731325815e-06,
+      "loss": 0.004,
+      "step": 5694
+    },
+    {
+      "epoch": 3.4651658046851233,
+      "grad_norm": 0.3154228627681732,
+      "learning_rate": 2.346252738701071e-06,
+      "loss": 0.005,
+      "step": 5695
+    },
+    {
+      "epoch": 3.4657742622452083,
+      "grad_norm": 0.19232860207557678,
+      "learning_rate": 2.340990221454628e-06,
+      "loss": 0.0035,
+      "step": 5696
+    },
+    {
+      "epoch": 3.4663827198052934,
+      "grad_norm": 0.28526800870895386,
+      "learning_rate": 2.3357333226981333e-06,
+      "loss": 0.0065,
+      "step": 5697
+    },
+    {
+      "epoch": 3.4669911773653785,
+      "grad_norm": 0.20558756589889526,
+      "learning_rate": 2.33048204373508e-06,
+      "loss": 0.0026,
+      "step": 5698
+    },
+    {
+      "epoch": 3.467599634925464,
+      "grad_norm": 0.21182219684123993,
+      "learning_rate": 2.3252363858675684e-06,
+      "loss": 0.0044,
+      "step": 5699
+    },
+    {
+      "epoch": 3.468208092485549,
+      "grad_norm": 0.22109881043434143,
+      "learning_rate": 2.3199963503963e-06,
+      "loss": 0.0043,
+      "step": 5700
+    },
+    {
+      "epoch": 3.4688165500456343,
+      "grad_norm": 0.26969975233078003,
+      "learning_rate": 2.314761938620591e-06,
+      "loss": 0.0073,
+      "step": 5701
+    },
+    {
+      "epoch": 3.4694250076057194,
+      "grad_norm": 0.20506982505321503,
+      "learning_rate": 2.3095331518383582e-06,
+      "loss": 0.0061,
+      "step": 5702
+    },
+    {
+      "epoch": 3.4700334651658045,
+      "grad_norm": 0.20657236874103546,
+      "learning_rate": 2.3043099913461125e-06,
+      "loss": 0.0033,
+      "step": 5703
+    },
+    {
+      "epoch": 3.47064192272589,
+      "grad_norm": 0.22096757590770721,
+      "learning_rate": 2.2990924584389868e-06,
+      "loss": 0.0039,
+      "step": 5704
+    },
+    {
+      "epoch": 3.471250380285975,
+      "grad_norm": 0.221493199467659,
+      "learning_rate": 2.29388055441071e-06,
+      "loss": 0.0034,
+      "step": 5705
+    },
+    {
+      "epoch": 3.47185883784606,
+      "grad_norm": 0.2772558331489563,
+      "learning_rate": 2.2886742805536183e-06,
+      "loss": 0.0058,
+      "step": 5706
+    },
+    {
+      "epoch": 3.4724672954061453,
+      "grad_norm": 0.26506444811820984,
+      "learning_rate": 2.283473638158656e-06,
+      "loss": 0.0064,
+      "step": 5707
+    },
+    {
+      "epoch": 3.4730757529662304,
+      "grad_norm": 0.18250389397144318,
+      "learning_rate": 2.278278628515354e-06,
+      "loss": 0.0037,
+      "step": 5708
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.21926113963127136,
+      "learning_rate": 2.2730892529118643e-06,
+      "loss": 0.0046,
+      "step": 5709
+    },
+    {
+      "epoch": 3.474292668086401,
+      "grad_norm": 0.2898429334163666,
+      "learning_rate": 2.267905512634935e-06,
+      "loss": 0.0055,
+      "step": 5710
+    },
+    {
+      "epoch": 3.474901125646486,
+      "grad_norm": 0.21994440257549286,
+      "learning_rate": 2.2627274089699195e-06,
+      "loss": 0.005,
+      "step": 5711
+    },
+    {
+      "epoch": 3.475509583206571,
+      "grad_norm": 0.3623565137386322,
+      "learning_rate": 2.257554943200771e-06,
+      "loss": 0.0111,
+      "step": 5712
+    },
+    {
+      "epoch": 3.4761180407666563,
+      "grad_norm": 0.2396956980228424,
+      "learning_rate": 2.252388116610046e-06,
+      "loss": 0.0043,
+      "step": 5713
+    },
+    {
+      "epoch": 3.476726498326742,
+      "grad_norm": 0.25912052392959595,
+      "learning_rate": 2.247226930478899e-06,
+      "loss": 0.0062,
+      "step": 5714
+    },
+    {
+      "epoch": 3.477334955886827,
+      "grad_norm": 0.22874920070171356,
+      "learning_rate": 2.2420713860870914e-06,
+      "loss": 0.0055,
+      "step": 5715
+    },
+    {
+      "epoch": 3.477943413446912,
+      "grad_norm": 0.3113996386528015,
+      "learning_rate": 2.2369214847129812e-06,
+      "loss": 0.0056,
+      "step": 5716
+    },
+    {
+      "epoch": 3.478551871006997,
+      "grad_norm": 0.21554897725582123,
+      "learning_rate": 2.231777227633533e-06,
+      "loss": 0.0053,
+      "step": 5717
+    },
+    {
+      "epoch": 3.4791603285670822,
+      "grad_norm": 0.29633817076683044,
+      "learning_rate": 2.2266386161243045e-06,
+      "loss": 0.0057,
+      "step": 5718
+    },
+    {
+      "epoch": 3.479768786127168,
+      "grad_norm": 0.14219973981380463,
+      "learning_rate": 2.221505651459463e-06,
+      "loss": 0.0035,
+      "step": 5719
+    },
+    {
+      "epoch": 3.480377243687253,
+      "grad_norm": 0.2941674590110779,
+      "learning_rate": 2.2163783349117616e-06,
+      "loss": 0.0053,
+      "step": 5720
+    },
+    {
+      "epoch": 3.480985701247338,
+      "grad_norm": 0.30135074257850647,
+      "learning_rate": 2.211256667752565e-06,
+      "loss": 0.0099,
+      "step": 5721
+    },
+    {
+      "epoch": 3.481594158807423,
+      "grad_norm": 0.3033702075481415,
+      "learning_rate": 2.206140651251834e-06,
+      "loss": 0.0064,
+      "step": 5722
+    },
+    {
+      "epoch": 3.482202616367508,
+      "grad_norm": 0.26715901494026184,
+      "learning_rate": 2.2010302866781296e-06,
+      "loss": 0.0059,
+      "step": 5723
+    },
+    {
+      "epoch": 3.4828110739275937,
+      "grad_norm": 0.3949749767780304,
+      "learning_rate": 2.1959255752986017e-06,
+      "loss": 0.0104,
+      "step": 5724
+    },
+    {
+      "epoch": 3.483419531487679,
+      "grad_norm": 0.2797585129737854,
+      "learning_rate": 2.1908265183790105e-06,
+      "loss": 0.0073,
+      "step": 5725
+    },
+    {
+      "epoch": 3.484027989047764,
+      "grad_norm": 0.23626194894313812,
+      "learning_rate": 2.1857331171837107e-06,
+      "loss": 0.009,
+      "step": 5726
+    },
+    {
+      "epoch": 3.484636446607849,
+      "grad_norm": 0.23918910324573517,
+      "learning_rate": 2.180645372975651e-06,
+      "loss": 0.0059,
+      "step": 5727
+    },
+    {
+      "epoch": 3.485244904167934,
+      "grad_norm": 0.16909535229206085,
+      "learning_rate": 2.1755632870163828e-06,
+      "loss": 0.0032,
+      "step": 5728
+    },
+    {
+      "epoch": 3.4858533617280196,
+      "grad_norm": 0.2599836587905884,
+      "learning_rate": 2.170486860566054e-06,
+      "loss": 0.008,
+      "step": 5729
+    },
+    {
+      "epoch": 3.4864618192881047,
+      "grad_norm": 0.3136194944381714,
+      "learning_rate": 2.1654160948833963e-06,
+      "loss": 0.0058,
+      "step": 5730
+    },
+    {
+      "epoch": 3.48707027684819,
+      "grad_norm": 0.2240346372127533,
+      "learning_rate": 2.160350991225757e-06,
+      "loss": 0.0049,
+      "step": 5731
+    },
+    {
+      "epoch": 3.487678734408275,
+      "grad_norm": 0.34254634380340576,
+      "learning_rate": 2.1552915508490675e-06,
+      "loss": 0.0119,
+      "step": 5732
+    },
+    {
+      "epoch": 3.48828719196836,
+      "grad_norm": 0.22509586811065674,
+      "learning_rate": 2.1502377750078627e-06,
+      "loss": 0.0043,
+      "step": 5733
+    },
+    {
+      "epoch": 3.4888956495284456,
+      "grad_norm": 0.19744735956192017,
+      "learning_rate": 2.145189664955258e-06,
+      "loss": 0.0055,
+      "step": 5734
+    },
+    {
+      "epoch": 3.4895041070885306,
+      "grad_norm": 0.24436576664447784,
+      "learning_rate": 2.1401472219429867e-06,
+      "loss": 0.0053,
+      "step": 5735
+    },
+    {
+      "epoch": 3.4901125646486157,
+      "grad_norm": 0.29822880029678345,
+      "learning_rate": 2.1351104472213585e-06,
+      "loss": 0.0054,
+      "step": 5736
+    },
+    {
+      "epoch": 3.490721022208701,
+      "grad_norm": 0.33873575925827026,
+      "learning_rate": 2.1300793420392815e-06,
+      "loss": 0.0053,
+      "step": 5737
+    },
+    {
+      "epoch": 3.491329479768786,
+      "grad_norm": 0.3289222717285156,
+      "learning_rate": 2.1250539076442617e-06,
+      "loss": 0.0073,
+      "step": 5738
+    },
+    {
+      "epoch": 3.4919379373288715,
+      "grad_norm": 0.23619408905506134,
+      "learning_rate": 2.120034145282404e-06,
+      "loss": 0.0044,
+      "step": 5739
+    },
+    {
+      "epoch": 3.4925463948889566,
+      "grad_norm": 0.2359532117843628,
+      "learning_rate": 2.115020056198383e-06,
+      "loss": 0.0052,
+      "step": 5740
+    },
+    {
+      "epoch": 3.4931548524490417,
+      "grad_norm": 0.20343980193138123,
+      "learning_rate": 2.1100116416355063e-06,
+      "loss": 0.0051,
+      "step": 5741
+    },
+    {
+      "epoch": 3.4937633100091268,
+      "grad_norm": 0.2578510642051697,
+      "learning_rate": 2.1050089028356366e-06,
+      "loss": 0.0055,
+      "step": 5742
+    },
+    {
+      "epoch": 3.494371767569212,
+      "grad_norm": 0.2600227892398834,
+      "learning_rate": 2.100011841039251e-06,
+      "loss": 0.0072,
+      "step": 5743
+    },
+    {
+      "epoch": 3.4949802251292974,
+      "grad_norm": 0.23431195318698883,
+      "learning_rate": 2.0950204574854027e-06,
+      "loss": 0.0045,
+      "step": 5744
+    },
+    {
+      "epoch": 3.4955886826893825,
+      "grad_norm": 0.265465646982193,
+      "learning_rate": 2.0900347534117627e-06,
+      "loss": 0.007,
+      "step": 5745
+    },
+    {
+      "epoch": 3.4961971402494676,
+      "grad_norm": 0.22313670814037323,
+      "learning_rate": 2.0850547300545668e-06,
+      "loss": 0.0041,
+      "step": 5746
+    },
+    {
+      "epoch": 3.4968055978095527,
+      "grad_norm": 0.3893153965473175,
+      "learning_rate": 2.080080388648653e-06,
+      "loss": 0.0111,
+      "step": 5747
+    },
+    {
+      "epoch": 3.497414055369638,
+      "grad_norm": 0.31730809807777405,
+      "learning_rate": 2.0751117304274563e-06,
+      "loss": 0.0099,
+      "step": 5748
+    },
+    {
+      "epoch": 3.4980225129297233,
+      "grad_norm": 0.30250468850135803,
+      "learning_rate": 2.070148756622997e-06,
+      "loss": 0.0057,
+      "step": 5749
+    },
+    {
+      "epoch": 3.4986309704898084,
+      "grad_norm": 0.30746808648109436,
+      "learning_rate": 2.065191468465874e-06,
+      "loss": 0.0084,
+      "step": 5750
+    },
+    {
+      "epoch": 3.4992394280498935,
+      "grad_norm": 0.1614643782377243,
+      "learning_rate": 2.060239867185304e-06,
+      "loss": 0.0021,
+      "step": 5751
+    },
+    {
+      "epoch": 3.4998478856099786,
+      "grad_norm": 0.2963274121284485,
+      "learning_rate": 2.0552939540090687e-06,
+      "loss": 0.006,
+      "step": 5752
+    },
+    {
+      "epoch": 3.5004563431700637,
+      "grad_norm": 0.20009563863277435,
+      "learning_rate": 2.050353730163554e-06,
+      "loss": 0.0052,
+      "step": 5753
+    },
+    {
+      "epoch": 3.5010648007301493,
+      "grad_norm": 0.20086035132408142,
+      "learning_rate": 2.045419196873716e-06,
+      "loss": 0.0028,
+      "step": 5754
+    },
+    {
+      "epoch": 3.5016732582902343,
+      "grad_norm": 0.22441764175891876,
+      "learning_rate": 2.0404903553631337e-06,
+      "loss": 0.0052,
+      "step": 5755
+    },
+    {
+      "epoch": 3.5022817158503194,
+      "grad_norm": 0.2553858160972595,
+      "learning_rate": 2.0355672068539387e-06,
+      "loss": 0.0045,
+      "step": 5756
+    },
+    {
+      "epoch": 3.5028901734104045,
+      "grad_norm": 0.32011979818344116,
+      "learning_rate": 2.030649752566871e-06,
+      "loss": 0.0076,
+      "step": 5757
+    },
+    {
+      "epoch": 3.5034986309704896,
+      "grad_norm": 0.21500691771507263,
+      "learning_rate": 2.025737993721255e-06,
+      "loss": 0.0053,
+      "step": 5758
+    },
+    {
+      "epoch": 3.504107088530575,
+      "grad_norm": 0.32461220026016235,
+      "learning_rate": 2.020831931535008e-06,
+      "loss": 0.0096,
+      "step": 5759
+    },
+    {
+      "epoch": 3.5047155460906603,
+      "grad_norm": 0.28721633553504944,
+      "learning_rate": 2.015931567224613e-06,
+      "loss": 0.0088,
+      "step": 5760
+    },
+    {
+      "epoch": 3.5053240036507454,
+      "grad_norm": 0.17419946193695068,
+      "learning_rate": 2.0110369020051755e-06,
+      "loss": 0.0027,
+      "step": 5761
+    },
+    {
+      "epoch": 3.5059324612108305,
+      "grad_norm": 0.33820995688438416,
+      "learning_rate": 2.006147937090355e-06,
+      "loss": 0.0107,
+      "step": 5762
+    },
+    {
+      "epoch": 3.5065409187709156,
+      "grad_norm": 0.3115573227405548,
+      "learning_rate": 2.001264673692413e-06,
+      "loss": 0.0061,
+      "step": 5763
+    },
+    {
+      "epoch": 3.507149376331001,
+      "grad_norm": 0.2463875263929367,
+      "learning_rate": 1.9963871130221997e-06,
+      "loss": 0.0061,
+      "step": 5764
+    },
+    {
+      "epoch": 3.507757833891086,
+      "grad_norm": 0.19313430786132812,
+      "learning_rate": 1.9915152562891476e-06,
+      "loss": 0.0046,
+      "step": 5765
+    },
+    {
+      "epoch": 3.5083662914511713,
+      "grad_norm": 0.27370426058769226,
+      "learning_rate": 1.9866491047012687e-06,
+      "loss": 0.0052,
+      "step": 5766
+    },
+    {
+      "epoch": 3.5089747490112564,
+      "grad_norm": 0.22294797003269196,
+      "learning_rate": 1.981788659465164e-06,
+      "loss": 0.0045,
+      "step": 5767
+    },
+    {
+      "epoch": 3.5095832065713415,
+      "grad_norm": 0.24368692934513092,
+      "learning_rate": 1.976933921786028e-06,
+      "loss": 0.0047,
+      "step": 5768
+    },
+    {
+      "epoch": 3.510191664131427,
+      "grad_norm": 0.29862114787101746,
+      "learning_rate": 1.972084892867637e-06,
+      "loss": 0.0081,
+      "step": 5769
+    },
+    {
+      "epoch": 3.510800121691512,
+      "grad_norm": 0.16069160401821136,
+      "learning_rate": 1.967241573912329e-06,
+      "loss": 0.0023,
+      "step": 5770
+    },
+    {
+      "epoch": 3.511408579251597,
+      "grad_norm": 0.20247617363929749,
+      "learning_rate": 1.962403966121071e-06,
+      "loss": 0.0043,
+      "step": 5771
+    },
+    {
+      "epoch": 3.5120170368116823,
+      "grad_norm": 0.2329401671886444,
+      "learning_rate": 1.957572070693367e-06,
+      "loss": 0.0039,
+      "step": 5772
+    },
+    {
+      "epoch": 3.5126254943717674,
+      "grad_norm": 0.2965979278087616,
+      "learning_rate": 1.952745888827337e-06,
+      "loss": 0.0107,
+      "step": 5773
+    },
+    {
+      "epoch": 3.513233951931853,
+      "grad_norm": 0.10749952495098114,
+      "learning_rate": 1.947925421719668e-06,
+      "loss": 0.0021,
+      "step": 5774
+    },
+    {
+      "epoch": 3.513842409491938,
+      "grad_norm": 0.2879178822040558,
+      "learning_rate": 1.9431106705656397e-06,
+      "loss": 0.0084,
+      "step": 5775
+    },
+    {
+      "epoch": 3.514450867052023,
+      "grad_norm": 0.34235483407974243,
+      "learning_rate": 1.938301636559098e-06,
+      "loss": 0.0071,
+      "step": 5776
+    },
+    {
+      "epoch": 3.5150593246121082,
+      "grad_norm": 0.21877804398536682,
+      "learning_rate": 1.9334983208925017e-06,
+      "loss": 0.007,
+      "step": 5777
+    },
+    {
+      "epoch": 3.5156677821721933,
+      "grad_norm": 0.36894896626472473,
+      "learning_rate": 1.9287007247568573e-06,
+      "loss": 0.007,
+      "step": 5778
+    },
+    {
+      "epoch": 3.516276239732279,
+      "grad_norm": 0.3058571517467499,
+      "learning_rate": 1.9239088493417796e-06,
+      "loss": 0.0046,
+      "step": 5779
+    },
+    {
+      "epoch": 3.516884697292364,
+      "grad_norm": 0.24490618705749512,
+      "learning_rate": 1.9191226958354403e-06,
+      "loss": 0.0057,
+      "step": 5780
+    },
+    {
+      "epoch": 3.517493154852449,
+      "grad_norm": 0.304394006729126,
+      "learning_rate": 1.9143422654246205e-06,
+      "loss": 0.0098,
+      "step": 5781
+    },
+    {
+      "epoch": 3.518101612412534,
+      "grad_norm": 0.28492802381515503,
+      "learning_rate": 1.9095675592946587e-06,
+      "loss": 0.0061,
+      "step": 5782
+    },
+    {
+      "epoch": 3.5187100699726193,
+      "grad_norm": 0.2946596145629883,
+      "learning_rate": 1.9047985786294853e-06,
+      "loss": 0.0061,
+      "step": 5783
+    },
+    {
+      "epoch": 3.519318527532705,
+      "grad_norm": 0.18398457765579224,
+      "learning_rate": 1.900035324611607e-06,
+      "loss": 0.0037,
+      "step": 5784
+    },
+    {
+      "epoch": 3.51992698509279,
+      "grad_norm": 0.23536744713783264,
+      "learning_rate": 1.895277798422121e-06,
+      "loss": 0.004,
+      "step": 5785
+    },
+    {
+      "epoch": 3.520535442652875,
+      "grad_norm": 0.24130214750766754,
+      "learning_rate": 1.8905260012406778e-06,
+      "loss": 0.0049,
+      "step": 5786
+    },
+    {
+      "epoch": 3.52114390021296,
+      "grad_norm": 0.2898615300655365,
+      "learning_rate": 1.8857799342455462e-06,
+      "loss": 0.0057,
+      "step": 5787
+    },
+    {
+      "epoch": 3.521752357773045,
+      "grad_norm": 0.25997620820999146,
+      "learning_rate": 1.8810395986135377e-06,
+      "loss": 0.0057,
+      "step": 5788
+    },
+    {
+      "epoch": 3.5223608153331307,
+      "grad_norm": 0.2553774416446686,
+      "learning_rate": 1.8763049955200674e-06,
+      "loss": 0.0057,
+      "step": 5789
+    },
+    {
+      "epoch": 3.522969272893216,
+      "grad_norm": 0.26607316732406616,
+      "learning_rate": 1.8715761261391074e-06,
+      "loss": 0.0045,
+      "step": 5790
+    },
+    {
+      "epoch": 3.523577730453301,
+      "grad_norm": 0.24370121955871582,
+      "learning_rate": 1.8668529916432365e-06,
+      "loss": 0.0047,
+      "step": 5791
+    },
+    {
+      "epoch": 3.524186188013386,
+      "grad_norm": 0.2677232325077057,
+      "learning_rate": 1.8621355932035788e-06,
+      "loss": 0.0061,
+      "step": 5792
+    },
+    {
+      "epoch": 3.524794645573471,
+      "grad_norm": 0.29431116580963135,
+      "learning_rate": 1.8574239319898657e-06,
+      "loss": 0.0068,
+      "step": 5793
+    },
+    {
+      "epoch": 3.5254031031335566,
+      "grad_norm": 0.2646505832672119,
+      "learning_rate": 1.8527180091703843e-06,
+      "loss": 0.005,
+      "step": 5794
+    },
+    {
+      "epoch": 3.5260115606936417,
+      "grad_norm": 0.2067060023546219,
+      "learning_rate": 1.848017825912013e-06,
+      "loss": 0.0051,
+      "step": 5795
+    },
+    {
+      "epoch": 3.526620018253727,
+      "grad_norm": 0.19585838913917542,
+      "learning_rate": 1.843323383380194e-06,
+      "loss": 0.0028,
+      "step": 5796
+    },
+    {
+      "epoch": 3.527228475813812,
+      "grad_norm": 0.27853307127952576,
+      "learning_rate": 1.8386346827389629e-06,
+      "loss": 0.006,
+      "step": 5797
+    },
+    {
+      "epoch": 3.527836933373897,
+      "grad_norm": 0.260459840297699,
+      "learning_rate": 1.8339517251509146e-06,
+      "loss": 0.0058,
+      "step": 5798
+    },
+    {
+      "epoch": 3.5284453909339826,
+      "grad_norm": 0.21536771953105927,
+      "learning_rate": 1.829274511777232e-06,
+      "loss": 0.006,
+      "step": 5799
+    },
+    {
+      "epoch": 3.5290538484940677,
+      "grad_norm": 0.3227105736732483,
+      "learning_rate": 1.8246030437776645e-06,
+      "loss": 0.0062,
+      "step": 5800
+    },
+    {
+      "epoch": 3.5296623060541528,
+      "grad_norm": 0.21509003639221191,
+      "learning_rate": 1.8199373223105498e-06,
+      "loss": 0.0041,
+      "step": 5801
+    },
+    {
+      "epoch": 3.530270763614238,
+      "grad_norm": 0.3301815092563629,
+      "learning_rate": 1.8152773485327818e-06,
+      "loss": 0.0088,
+      "step": 5802
+    },
+    {
+      "epoch": 3.530879221174323,
+      "grad_norm": 0.13643626868724823,
+      "learning_rate": 1.8106231235998444e-06,
+      "loss": 0.0032,
+      "step": 5803
+    },
+    {
+      "epoch": 3.5314876787344085,
+      "grad_norm": 0.3046422600746155,
+      "learning_rate": 1.8059746486657896e-06,
+      "loss": 0.0091,
+      "step": 5804
+    },
+    {
+      "epoch": 3.5320961362944936,
+      "grad_norm": 0.2640552222728729,
+      "learning_rate": 1.8013319248832538e-06,
+      "loss": 0.0067,
+      "step": 5805
+    },
+    {
+      "epoch": 3.5327045938545787,
+      "grad_norm": 0.32940080761909485,
+      "learning_rate": 1.7966949534034243e-06,
+      "loss": 0.0086,
+      "step": 5806
+    },
+    {
+      "epoch": 3.533313051414664,
+      "grad_norm": 0.2538890242576599,
+      "learning_rate": 1.7920637353760928e-06,
+      "loss": 0.0057,
+      "step": 5807
+    },
+    {
+      "epoch": 3.533921508974749,
+      "grad_norm": 0.6640288233757019,
+      "learning_rate": 1.7874382719495958e-06,
+      "loss": 0.0576,
+      "step": 5808
+    },
+    {
+      "epoch": 3.5345299665348344,
+      "grad_norm": 0.3605937063694,
+      "learning_rate": 1.7828185642708605e-06,
+      "loss": 0.0078,
+      "step": 5809
+    },
+    {
+      "epoch": 3.5351384240949195,
+      "grad_norm": 0.2806709408760071,
+      "learning_rate": 1.7782046134853792e-06,
+      "loss": 0.0088,
+      "step": 5810
+    },
+    {
+      "epoch": 3.5357468816550046,
+      "grad_norm": 0.17481489479541779,
+      "learning_rate": 1.773596420737228e-06,
+      "loss": 0.0039,
+      "step": 5811
+    },
+    {
+      "epoch": 3.5363553392150897,
+      "grad_norm": 0.15608619153499603,
+      "learning_rate": 1.7689939871690375e-06,
+      "loss": 0.0043,
+      "step": 5812
+    },
+    {
+      "epoch": 3.536963796775175,
+      "grad_norm": 0.2716093361377716,
+      "learning_rate": 1.7643973139220198e-06,
+      "loss": 0.0101,
+      "step": 5813
+    },
+    {
+      "epoch": 3.5375722543352603,
+      "grad_norm": 0.13356530666351318,
+      "learning_rate": 1.7598064021359607e-06,
+      "loss": 0.0015,
+      "step": 5814
+    },
+    {
+      "epoch": 3.5381807118953454,
+      "grad_norm": 0.17770147323608398,
+      "learning_rate": 1.7552212529492158e-06,
+      "loss": 0.0029,
+      "step": 5815
+    },
+    {
+      "epoch": 3.5387891694554305,
+      "grad_norm": 0.1851194053888321,
+      "learning_rate": 1.750641867498709e-06,
+      "loss": 0.0034,
+      "step": 5816
+    },
+    {
+      "epoch": 3.5393976270155156,
+      "grad_norm": 0.3669975697994232,
+      "learning_rate": 1.7460682469199435e-06,
+      "loss": 0.0077,
+      "step": 5817
+    },
+    {
+      "epoch": 3.5400060845756007,
+      "grad_norm": 0.24136383831501007,
+      "learning_rate": 1.7415003923469787e-06,
+      "loss": 0.0053,
+      "step": 5818
+    },
+    {
+      "epoch": 3.5406145421356863,
+      "grad_norm": 0.2976721227169037,
+      "learning_rate": 1.7369383049124498e-06,
+      "loss": 0.0061,
+      "step": 5819
+    },
+    {
+      "epoch": 3.5412229996957714,
+      "grad_norm": 0.20041820406913757,
+      "learning_rate": 1.7323819857475721e-06,
+      "loss": 0.0051,
+      "step": 5820
+    },
+    {
+      "epoch": 3.5418314572558565,
+      "grad_norm": 0.2120663970708847,
+      "learning_rate": 1.727831435982119e-06,
+      "loss": 0.0047,
+      "step": 5821
+    },
+    {
+      "epoch": 3.5424399148159416,
+      "grad_norm": 0.18316033482551575,
+      "learning_rate": 1.7232866567444384e-06,
+      "loss": 0.0035,
+      "step": 5822
+    },
+    {
+      "epoch": 3.5430483723760267,
+      "grad_norm": 0.22955255210399628,
+      "learning_rate": 1.7187476491614507e-06,
+      "loss": 0.0058,
+      "step": 5823
+    },
+    {
+      "epoch": 3.543656829936112,
+      "grad_norm": 0.22870229184627533,
+      "learning_rate": 1.7142144143586308e-06,
+      "loss": 0.0033,
+      "step": 5824
+    },
+    {
+      "epoch": 3.5442652874961973,
+      "grad_norm": 0.24868161976337433,
+      "learning_rate": 1.7096869534600352e-06,
+      "loss": 0.0076,
+      "step": 5825
+    },
+    {
+      "epoch": 3.5448737450562824,
+      "grad_norm": 0.32703396677970886,
+      "learning_rate": 1.7051652675882878e-06,
+      "loss": 0.0066,
+      "step": 5826
+    },
+    {
+      "epoch": 3.5454822026163675,
+      "grad_norm": 0.28158140182495117,
+      "learning_rate": 1.7006493578645838e-06,
+      "loss": 0.0064,
+      "step": 5827
+    },
+    {
+      "epoch": 3.5460906601764526,
+      "grad_norm": 0.2724444270133972,
+      "learning_rate": 1.6961392254086689e-06,
+      "loss": 0.0065,
+      "step": 5828
+    },
+    {
+      "epoch": 3.546699117736538,
+      "grad_norm": 0.312369704246521,
+      "learning_rate": 1.6916348713388708e-06,
+      "loss": 0.0063,
+      "step": 5829
+    },
+    {
+      "epoch": 3.547307575296623,
+      "grad_norm": 0.2701030373573303,
+      "learning_rate": 1.6871362967720878e-06,
+      "loss": 0.006,
+      "step": 5830
+    },
+    {
+      "epoch": 3.5479160328567083,
+      "grad_norm": 0.16680403053760529,
+      "learning_rate": 1.682643502823772e-06,
+      "loss": 0.0028,
+      "step": 5831
+    },
+    {
+      "epoch": 3.5485244904167934,
+      "grad_norm": 0.25249308347702026,
+      "learning_rate": 1.6781564906079545e-06,
+      "loss": 0.0045,
+      "step": 5832
+    },
+    {
+      "epoch": 3.5491329479768785,
+      "grad_norm": 0.1920885592699051,
+      "learning_rate": 1.6736752612372286e-06,
+      "loss": 0.0041,
+      "step": 5833
+    },
+    {
+      "epoch": 3.549741405536964,
+      "grad_norm": 0.28763115406036377,
+      "learning_rate": 1.6691998158227446e-06,
+      "loss": 0.007,
+      "step": 5834
+    },
+    {
+      "epoch": 3.550349863097049,
+      "grad_norm": 0.479855477809906,
+      "learning_rate": 1.6647301554742312e-06,
+      "loss": 0.0151,
+      "step": 5835
+    },
+    {
+      "epoch": 3.5509583206571342,
+      "grad_norm": 0.4631776213645935,
+      "learning_rate": 1.6602662812999742e-06,
+      "loss": 0.0091,
+      "step": 5836
+    },
+    {
+      "epoch": 3.5515667782172193,
+      "grad_norm": 0.3424180746078491,
+      "learning_rate": 1.6558081944068354e-06,
+      "loss": 0.0108,
+      "step": 5837
+    },
+    {
+      "epoch": 3.5521752357773044,
+      "grad_norm": 0.24957185983657837,
+      "learning_rate": 1.6513558959002334e-06,
+      "loss": 0.0073,
+      "step": 5838
+    },
+    {
+      "epoch": 3.55278369333739,
+      "grad_norm": 0.23904968798160553,
+      "learning_rate": 1.6469093868841434e-06,
+      "loss": 0.0071,
+      "step": 5839
+    },
+    {
+      "epoch": 3.553392150897475,
+      "grad_norm": 0.1692766696214676,
+      "learning_rate": 1.6424686684611224e-06,
+      "loss": 0.0027,
+      "step": 5840
+    },
+    {
+      "epoch": 3.55400060845756,
+      "grad_norm": 0.28650224208831787,
+      "learning_rate": 1.638033741732281e-06,
+      "loss": 0.0097,
+      "step": 5841
+    },
+    {
+      "epoch": 3.5546090660176453,
+      "grad_norm": 0.14089111983776093,
+      "learning_rate": 1.6336046077972983e-06,
+      "loss": 0.0019,
+      "step": 5842
+    },
+    {
+      "epoch": 3.5552175235777304,
+      "grad_norm": 0.148437961935997,
+      "learning_rate": 1.6291812677544121e-06,
+      "loss": 0.0024,
+      "step": 5843
+    },
+    {
+      "epoch": 3.555825981137816,
+      "grad_norm": 0.26019591093063354,
+      "learning_rate": 1.6247637227004342e-06,
+      "loss": 0.007,
+      "step": 5844
+    },
+    {
+      "epoch": 3.5564344386979005,
+      "grad_norm": 0.2771880626678467,
+      "learning_rate": 1.6203519737307187e-06,
+      "loss": 0.0042,
+      "step": 5845
+    },
+    {
+      "epoch": 3.557042896257986,
+      "grad_norm": 0.27632075548171997,
+      "learning_rate": 1.615946021939202e-06,
+      "loss": 0.0059,
+      "step": 5846
+    },
+    {
+      "epoch": 3.557651353818071,
+      "grad_norm": 0.2640374004840851,
+      "learning_rate": 1.6115458684183793e-06,
+      "loss": 0.0031,
+      "step": 5847
+    },
+    {
+      "epoch": 3.5582598113781563,
+      "grad_norm": 0.23111113905906677,
+      "learning_rate": 1.607151514259303e-06,
+      "loss": 0.0044,
+      "step": 5848
+    },
+    {
+      "epoch": 3.558868268938242,
+      "grad_norm": 0.40425482392311096,
+      "learning_rate": 1.602762960551582e-06,
+      "loss": 0.0084,
+      "step": 5849
+    },
+    {
+      "epoch": 3.5594767264983265,
+      "grad_norm": 0.17525078356266022,
+      "learning_rate": 1.5983802083834126e-06,
+      "loss": 0.0032,
+      "step": 5850
+    },
+    {
+      "epoch": 3.560085184058412,
+      "grad_norm": 0.2451610118150711,
+      "learning_rate": 1.5940032588415171e-06,
+      "loss": 0.0077,
+      "step": 5851
+    },
+    {
+      "epoch": 3.560693641618497,
+      "grad_norm": 0.2035626322031021,
+      "learning_rate": 1.5896321130112023e-06,
+      "loss": 0.0035,
+      "step": 5852
+    },
+    {
+      "epoch": 3.561302099178582,
+      "grad_norm": 0.16917142271995544,
+      "learning_rate": 1.5852667719763348e-06,
+      "loss": 0.0026,
+      "step": 5853
+    },
+    {
+      "epoch": 3.5619105567386677,
+      "grad_norm": 0.40496107935905457,
+      "learning_rate": 1.5809072368193345e-06,
+      "loss": 0.027,
+      "step": 5854
+    },
+    {
+      "epoch": 3.5625190142987524,
+      "grad_norm": 0.2576707899570465,
+      "learning_rate": 1.5765535086211786e-06,
+      "loss": 0.0059,
+      "step": 5855
+    },
+    {
+      "epoch": 3.563127471858838,
+      "grad_norm": 0.27489739656448364,
+      "learning_rate": 1.57220558846142e-06,
+      "loss": 0.0037,
+      "step": 5856
+    },
+    {
+      "epoch": 3.563735929418923,
+      "grad_norm": 0.16968370974063873,
+      "learning_rate": 1.567863477418155e-06,
+      "loss": 0.0036,
+      "step": 5857
+    },
+    {
+      "epoch": 3.564344386979008,
+      "grad_norm": 0.271075040102005,
+      "learning_rate": 1.5635271765680525e-06,
+      "loss": 0.0051,
+      "step": 5858
+    },
+    {
+      "epoch": 3.5649528445390937,
+      "grad_norm": 0.27260100841522217,
+      "learning_rate": 1.5591966869863196e-06,
+      "loss": 0.0062,
+      "step": 5859
+    },
+    {
+      "epoch": 3.5655613020991783,
+      "grad_norm": 0.15648119151592255,
+      "learning_rate": 1.554872009746758e-06,
+      "loss": 0.0032,
+      "step": 5860
+    },
+    {
+      "epoch": 3.566169759659264,
+      "grad_norm": 0.23940731585025787,
+      "learning_rate": 1.5505531459216904e-06,
+      "loss": 0.004,
+      "step": 5861
+    },
+    {
+      "epoch": 3.566778217219349,
+      "grad_norm": 0.3243446350097656,
+      "learning_rate": 1.5462400965820218e-06,
+      "loss": 0.0109,
+      "step": 5862
+    },
+    {
+      "epoch": 3.567386674779434,
+      "grad_norm": 0.14726978540420532,
+      "learning_rate": 1.5419328627972103e-06,
+      "loss": 0.0021,
+      "step": 5863
+    },
+    {
+      "epoch": 3.5679951323395196,
+      "grad_norm": 0.2737961411476135,
+      "learning_rate": 1.5376314456352708e-06,
+      "loss": 0.0065,
+      "step": 5864
+    },
+    {
+      "epoch": 3.5686035898996042,
+      "grad_norm": 0.2394101470708847,
+      "learning_rate": 1.5333358461627673e-06,
+      "loss": 0.0036,
+      "step": 5865
+    },
+    {
+      "epoch": 3.56921204745969,
+      "grad_norm": 0.20798607170581818,
+      "learning_rate": 1.5290460654448418e-06,
+      "loss": 0.0051,
+      "step": 5866
+    },
+    {
+      "epoch": 3.569820505019775,
+      "grad_norm": 0.21632260084152222,
+      "learning_rate": 1.5247621045451688e-06,
+      "loss": 0.0044,
+      "step": 5867
+    },
+    {
+      "epoch": 3.57042896257986,
+      "grad_norm": 0.28413519263267517,
+      "learning_rate": 1.5204839645259983e-06,
+      "loss": 0.005,
+      "step": 5868
+    },
+    {
+      "epoch": 3.571037420139945,
+      "grad_norm": 0.28667712211608887,
+      "learning_rate": 1.5162116464481318e-06,
+      "loss": 0.0077,
+      "step": 5869
+    },
+    {
+      "epoch": 3.57164587770003,
+      "grad_norm": 0.25631603598594666,
+      "learning_rate": 1.5119451513709277e-06,
+      "loss": 0.0039,
+      "step": 5870
+    },
+    {
+      "epoch": 3.5722543352601157,
+      "grad_norm": 0.19717690348625183,
+      "learning_rate": 1.5076844803522922e-06,
+      "loss": 0.0024,
+      "step": 5871
+    },
+    {
+      "epoch": 3.572862792820201,
+      "grad_norm": 0.28270480036735535,
+      "learning_rate": 1.503429634448697e-06,
+      "loss": 0.0071,
+      "step": 5872
+    },
+    {
+      "epoch": 3.573471250380286,
+      "grad_norm": 0.23655082285404205,
+      "learning_rate": 1.4991806147151677e-06,
+      "loss": 0.0069,
+      "step": 5873
+    },
+    {
+      "epoch": 3.574079707940371,
+      "grad_norm": 0.22452490031719208,
+      "learning_rate": 1.4949374222052864e-06,
+      "loss": 0.0042,
+      "step": 5874
+    },
+    {
+      "epoch": 3.574688165500456,
+      "grad_norm": 0.21213477849960327,
+      "learning_rate": 1.4907000579711782e-06,
+      "loss": 0.0043,
+      "step": 5875
+    },
+    {
+      "epoch": 3.5752966230605416,
+      "grad_norm": 0.27884024381637573,
+      "learning_rate": 1.4864685230635473e-06,
+      "loss": 0.0057,
+      "step": 5876
+    },
+    {
+      "epoch": 3.5759050806206267,
+      "grad_norm": 0.23877516388893127,
+      "learning_rate": 1.4822428185316261e-06,
+      "loss": 0.0067,
+      "step": 5877
+    },
+    {
+      "epoch": 3.576513538180712,
+      "grad_norm": 0.20616334676742554,
+      "learning_rate": 1.4780229454232158e-06,
+      "loss": 0.0053,
+      "step": 5878
+    },
+    {
+      "epoch": 3.577121995740797,
+      "grad_norm": 0.18893709778785706,
+      "learning_rate": 1.4738089047846736e-06,
+      "loss": 0.0038,
+      "step": 5879
+    },
+    {
+      "epoch": 3.577730453300882,
+      "grad_norm": 0.2203339785337448,
+      "learning_rate": 1.4696006976609057e-06,
+      "loss": 0.0054,
+      "step": 5880
+    },
+    {
+      "epoch": 3.5783389108609676,
+      "grad_norm": 0.15522873401641846,
+      "learning_rate": 1.4653983250953657e-06,
+      "loss": 0.0029,
+      "step": 5881
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.14982964098453522,
+      "learning_rate": 1.4612017881300704e-06,
+      "loss": 0.0032,
+      "step": 5882
+    },
+    {
+      "epoch": 3.5795558259811378,
+      "grad_norm": 0.26596924662590027,
+      "learning_rate": 1.4570110878055877e-06,
+      "loss": 0.0044,
+      "step": 5883
+    },
+    {
+      "epoch": 3.580164283541223,
+      "grad_norm": 0.2501273453235626,
+      "learning_rate": 1.4528262251610358e-06,
+      "loss": 0.007,
+      "step": 5884
+    },
+    {
+      "epoch": 3.580772741101308,
+      "grad_norm": 0.19661864638328552,
+      "learning_rate": 1.4486472012340824e-06,
+      "loss": 0.0042,
+      "step": 5885
+    },
+    {
+      "epoch": 3.5813811986613935,
+      "grad_norm": 0.22375038266181946,
+      "learning_rate": 1.4444740170609567e-06,
+      "loss": 0.0061,
+      "step": 5886
+    },
+    {
+      "epoch": 3.5819896562214786,
+      "grad_norm": 0.28383755683898926,
+      "learning_rate": 1.440306673676431e-06,
+      "loss": 0.0056,
+      "step": 5887
+    },
+    {
+      "epoch": 3.5825981137815637,
+      "grad_norm": 0.17903096973896027,
+      "learning_rate": 1.436145172113834e-06,
+      "loss": 0.0037,
+      "step": 5888
+    },
+    {
+      "epoch": 3.5832065713416488,
+      "grad_norm": 0.24099615216255188,
+      "learning_rate": 1.4319895134050437e-06,
+      "loss": 0.0047,
+      "step": 5889
+    },
+    {
+      "epoch": 3.583815028901734,
+      "grad_norm": 0.20468921959400177,
+      "learning_rate": 1.4278396985804966e-06,
+      "loss": 0.005,
+      "step": 5890
+    },
+    {
+      "epoch": 3.5844234864618194,
+      "grad_norm": 0.26146969199180603,
+      "learning_rate": 1.4236957286691581e-06,
+      "loss": 0.0064,
+      "step": 5891
+    },
+    {
+      "epoch": 3.5850319440219045,
+      "grad_norm": 0.328022837638855,
+      "learning_rate": 1.4195576046985793e-06,
+      "loss": 0.0064,
+      "step": 5892
+    },
+    {
+      "epoch": 3.5856404015819896,
+      "grad_norm": 0.26720380783081055,
+      "learning_rate": 1.4154253276948276e-06,
+      "loss": 0.0044,
+      "step": 5893
+    },
+    {
+      "epoch": 3.5862488591420747,
+      "grad_norm": 0.20053601264953613,
+      "learning_rate": 1.4112988986825476e-06,
+      "loss": 0.0033,
+      "step": 5894
+    },
+    {
+      "epoch": 3.58685731670216,
+      "grad_norm": 0.251700758934021,
+      "learning_rate": 1.407178318684907e-06,
+      "loss": 0.0063,
+      "step": 5895
+    },
+    {
+      "epoch": 3.5874657742622453,
+      "grad_norm": 0.4364267587661743,
+      "learning_rate": 1.403063588723652e-06,
+      "loss": 0.0114,
+      "step": 5896
+    },
+    {
+      "epoch": 3.5880742318223304,
+      "grad_norm": 0.25533613562583923,
+      "learning_rate": 1.3989547098190559e-06,
+      "loss": 0.0049,
+      "step": 5897
+    },
+    {
+      "epoch": 3.5886826893824155,
+      "grad_norm": 0.22121703624725342,
+      "learning_rate": 1.3948516829899505e-06,
+      "loss": 0.0056,
+      "step": 5898
+    },
+    {
+      "epoch": 3.5892911469425006,
+      "grad_norm": 0.16689860820770264,
+      "learning_rate": 1.3907545092537166e-06,
+      "loss": 0.0029,
+      "step": 5899
+    },
+    {
+      "epoch": 3.5898996045025857,
+      "grad_norm": 0.22792737185955048,
+      "learning_rate": 1.386663189626286e-06,
+      "loss": 0.0051,
+      "step": 5900
+    },
+    {
+      "epoch": 3.5905080620626713,
+      "grad_norm": 0.275244802236557,
+      "learning_rate": 1.3825777251221278e-06,
+      "loss": 0.0072,
+      "step": 5901
+    },
+    {
+      "epoch": 3.5911165196227564,
+      "grad_norm": 0.34632372856140137,
+      "learning_rate": 1.378498116754276e-06,
+      "loss": 0.0078,
+      "step": 5902
+    },
+    {
+      "epoch": 3.5917249771828414,
+      "grad_norm": 0.20081204175949097,
+      "learning_rate": 1.3744243655342937e-06,
+      "loss": 0.006,
+      "step": 5903
+    },
+    {
+      "epoch": 3.5923334347429265,
+      "grad_norm": 0.11412859708070755,
+      "learning_rate": 1.3703564724723116e-06,
+      "loss": 0.0016,
+      "step": 5904
+    },
+    {
+      "epoch": 3.5929418923030116,
+      "grad_norm": 0.22430390119552612,
+      "learning_rate": 1.3662944385769843e-06,
+      "loss": 0.004,
+      "step": 5905
+    },
+    {
+      "epoch": 3.593550349863097,
+      "grad_norm": 0.25405508279800415,
+      "learning_rate": 1.362238264855542e-06,
+      "loss": 0.007,
+      "step": 5906
+    },
+    {
+      "epoch": 3.5941588074231823,
+      "grad_norm": 0.447488397359848,
+      "learning_rate": 1.3581879523137386e-06,
+      "loss": 0.0064,
+      "step": 5907
+    },
+    {
+      "epoch": 3.5947672649832674,
+      "grad_norm": 0.2718401253223419,
+      "learning_rate": 1.354143501955879e-06,
+      "loss": 0.0067,
+      "step": 5908
+    },
+    {
+      "epoch": 3.5953757225433525,
+      "grad_norm": 0.18165600299835205,
+      "learning_rate": 1.3501049147848277e-06,
+      "loss": 0.0047,
+      "step": 5909
+    },
+    {
+      "epoch": 3.5959841801034376,
+      "grad_norm": 0.2794153392314911,
+      "learning_rate": 1.346072191801981e-06,
+      "loss": 0.0047,
+      "step": 5910
+    },
+    {
+      "epoch": 3.596592637663523,
+      "grad_norm": 0.21536363661289215,
+      "learning_rate": 1.3420453340072832e-06,
+      "loss": 0.0067,
+      "step": 5911
+    },
+    {
+      "epoch": 3.597201095223608,
+      "grad_norm": 0.2198522537946701,
+      "learning_rate": 1.3380243423992328e-06,
+      "loss": 0.0053,
+      "step": 5912
+    },
+    {
+      "epoch": 3.5978095527836933,
+      "grad_norm": 0.17715135216712952,
+      "learning_rate": 1.3340092179748658e-06,
+      "loss": 0.004,
+      "step": 5913
+    },
+    {
+      "epoch": 3.5984180103437784,
+      "grad_norm": 0.2602989375591278,
+      "learning_rate": 1.3299999617297637e-06,
+      "loss": 0.0056,
+      "step": 5914
+    },
+    {
+      "epoch": 3.5990264679038635,
+      "grad_norm": 0.20000648498535156,
+      "learning_rate": 1.3259965746580588e-06,
+      "loss": 0.0018,
+      "step": 5915
+    },
+    {
+      "epoch": 3.599634925463949,
+      "grad_norm": 0.25586432218551636,
+      "learning_rate": 1.3219990577524239e-06,
+      "loss": 0.008,
+      "step": 5916
+    },
+    {
+      "epoch": 3.600243383024034,
+      "grad_norm": 0.2492225617170334,
+      "learning_rate": 1.3180074120040741e-06,
+      "loss": 0.0054,
+      "step": 5917
+    },
+    {
+      "epoch": 3.6008518405841192,
+      "grad_norm": 0.21238954365253448,
+      "learning_rate": 1.3140216384027682e-06,
+      "loss": 0.0044,
+      "step": 5918
+    },
+    {
+      "epoch": 3.6014602981442043,
+      "grad_norm": 0.13737817108631134,
+      "learning_rate": 1.3100417379368179e-06,
+      "loss": 0.0033,
+      "step": 5919
+    },
+    {
+      "epoch": 3.6020687557042894,
+      "grad_norm": 0.34668809175491333,
+      "learning_rate": 1.306067711593076e-06,
+      "loss": 0.0071,
+      "step": 5920
+    },
+    {
+      "epoch": 3.602677213264375,
+      "grad_norm": 0.2114226520061493,
+      "learning_rate": 1.3020995603569203e-06,
+      "loss": 0.0044,
+      "step": 5921
+    },
+    {
+      "epoch": 3.60328567082446,
+      "grad_norm": 0.3277231454849243,
+      "learning_rate": 1.298137285212303e-06,
+      "loss": 0.008,
+      "step": 5922
+    },
+    {
+      "epoch": 3.603894128384545,
+      "grad_norm": 0.2284201830625534,
+      "learning_rate": 1.2941808871416938e-06,
+      "loss": 0.0035,
+      "step": 5923
+    },
+    {
+      "epoch": 3.6045025859446302,
+      "grad_norm": 0.2673366069793701,
+      "learning_rate": 1.290230367126119e-06,
+      "loss": 0.005,
+      "step": 5924
+    },
+    {
+      "epoch": 3.6051110435047153,
+      "grad_norm": 0.21040508151054382,
+      "learning_rate": 1.2862857261451395e-06,
+      "loss": 0.0037,
+      "step": 5925
+    },
+    {
+      "epoch": 3.605719501064801,
+      "grad_norm": 0.2350999116897583,
+      "learning_rate": 1.282346965176867e-06,
+      "loss": 0.0021,
+      "step": 5926
+    },
+    {
+      "epoch": 3.606327958624886,
+      "grad_norm": 0.33141762018203735,
+      "learning_rate": 1.2784140851979404e-06,
+      "loss": 0.0068,
+      "step": 5927
+    },
+    {
+      "epoch": 3.606936416184971,
+      "grad_norm": 0.2573280334472656,
+      "learning_rate": 1.2744870871835623e-06,
+      "loss": 0.0072,
+      "step": 5928
+    },
+    {
+      "epoch": 3.607544873745056,
+      "grad_norm": 0.21989892423152924,
+      "learning_rate": 1.270565972107457e-06,
+      "loss": 0.0088,
+      "step": 5929
+    },
+    {
+      "epoch": 3.6081533313051413,
+      "grad_norm": 0.2301366925239563,
+      "learning_rate": 1.2666507409418993e-06,
+      "loss": 0.0065,
+      "step": 5930
+    },
+    {
+      "epoch": 3.608761788865227,
+      "grad_norm": 0.18148985505104065,
+      "learning_rate": 1.2627413946576989e-06,
+      "loss": 0.0043,
+      "step": 5931
+    },
+    {
+      "epoch": 3.609370246425312,
+      "grad_norm": 0.24122391641139984,
+      "learning_rate": 1.2588379342242218e-06,
+      "loss": 0.0051,
+      "step": 5932
+    },
+    {
+      "epoch": 3.609978703985397,
+      "grad_norm": 0.4258638322353363,
+      "learning_rate": 1.2549403606093525e-06,
+      "loss": 0.0056,
+      "step": 5933
+    },
+    {
+      "epoch": 3.610587161545482,
+      "grad_norm": 0.2916209399700165,
+      "learning_rate": 1.2510486747795286e-06,
+      "loss": 0.006,
+      "step": 5934
+    },
+    {
+      "epoch": 3.611195619105567,
+      "grad_norm": 0.2829422056674957,
+      "learning_rate": 1.2471628776997312e-06,
+      "loss": 0.0039,
+      "step": 5935
+    },
+    {
+      "epoch": 3.6118040766656527,
+      "grad_norm": 0.2982374429702759,
+      "learning_rate": 1.2432829703334759e-06,
+      "loss": 0.006,
+      "step": 5936
+    },
+    {
+      "epoch": 3.612412534225738,
+      "grad_norm": 0.19474712014198303,
+      "learning_rate": 1.2394089536428067e-06,
+      "loss": 0.006,
+      "step": 5937
+    },
+    {
+      "epoch": 3.613020991785823,
+      "grad_norm": 0.2293567806482315,
+      "learning_rate": 1.2355408285883357e-06,
+      "loss": 0.0076,
+      "step": 5938
+    },
+    {
+      "epoch": 3.613629449345908,
+      "grad_norm": 0.26767081022262573,
+      "learning_rate": 1.2316785961291849e-06,
+      "loss": 0.0035,
+      "step": 5939
+    },
+    {
+      "epoch": 3.614237906905993,
+      "grad_norm": 0.2060486376285553,
+      "learning_rate": 1.2278222572230268e-06,
+      "loss": 0.0043,
+      "step": 5940
+    },
+    {
+      "epoch": 3.6148463644660787,
+      "grad_norm": 0.21085241436958313,
+      "learning_rate": 1.2239718128260774e-06,
+      "loss": 0.0062,
+      "step": 5941
+    },
+    {
+      "epoch": 3.6154548220261638,
+      "grad_norm": 0.24178573489189148,
+      "learning_rate": 1.2201272638930894e-06,
+      "loss": 0.0058,
+      "step": 5942
+    },
+    {
+      "epoch": 3.616063279586249,
+      "grad_norm": 0.20811717212200165,
+      "learning_rate": 1.216288611377342e-06,
+      "loss": 0.0049,
+      "step": 5943
+    },
+    {
+      "epoch": 3.616671737146334,
+      "grad_norm": 0.3802126348018646,
+      "learning_rate": 1.2124558562306625e-06,
+      "loss": 0.0064,
+      "step": 5944
+    },
+    {
+      "epoch": 3.617280194706419,
+      "grad_norm": 0.2511650323867798,
+      "learning_rate": 1.2086289994034217e-06,
+      "loss": 0.0051,
+      "step": 5945
+    },
+    {
+      "epoch": 3.6178886522665046,
+      "grad_norm": 0.27000531554222107,
+      "learning_rate": 1.204808041844513e-06,
+      "loss": 0.0072,
+      "step": 5946
+    },
+    {
+      "epoch": 3.6184971098265897,
+      "grad_norm": 0.26983216404914856,
+      "learning_rate": 1.2009929845013757e-06,
+      "loss": 0.0072,
+      "step": 5947
+    },
+    {
+      "epoch": 3.6191055673866748,
+      "grad_norm": 0.19793999195098877,
+      "learning_rate": 1.197183828319992e-06,
+      "loss": 0.0045,
+      "step": 5948
+    },
+    {
+      "epoch": 3.61971402494676,
+      "grad_norm": 0.238411545753479,
+      "learning_rate": 1.193380574244865e-06,
+      "loss": 0.0068,
+      "step": 5949
+    },
+    {
+      "epoch": 3.620322482506845,
+      "grad_norm": 0.1789894700050354,
+      "learning_rate": 1.1895832232190485e-06,
+      "loss": 0.0026,
+      "step": 5950
+    },
+    {
+      "epoch": 3.6209309400669305,
+      "grad_norm": 0.27357521653175354,
+      "learning_rate": 1.1857917761841225e-06,
+      "loss": 0.0085,
+      "step": 5951
+    },
+    {
+      "epoch": 3.6215393976270156,
+      "grad_norm": 0.31104499101638794,
+      "learning_rate": 1.182006234080213e-06,
+      "loss": 0.0063,
+      "step": 5952
+    },
+    {
+      "epoch": 3.6221478551871007,
+      "grad_norm": 0.23851986229419708,
+      "learning_rate": 1.1782265978459771e-06,
+      "loss": 0.0035,
+      "step": 5953
+    },
+    {
+      "epoch": 3.622756312747186,
+      "grad_norm": 0.48092883825302124,
+      "learning_rate": 1.1744528684186018e-06,
+      "loss": 0.0056,
+      "step": 5954
+    },
+    {
+      "epoch": 3.623364770307271,
+      "grad_norm": 0.19673019647598267,
+      "learning_rate": 1.170685046733816e-06,
+      "loss": 0.0052,
+      "step": 5955
+    },
+    {
+      "epoch": 3.6239732278673564,
+      "grad_norm": 0.24021019041538239,
+      "learning_rate": 1.1669231337258862e-06,
+      "loss": 0.0064,
+      "step": 5956
+    },
+    {
+      "epoch": 3.6245816854274415,
+      "grad_norm": 0.2394445836544037,
+      "learning_rate": 1.1631671303276054e-06,
+      "loss": 0.0043,
+      "step": 5957
+    },
+    {
+      "epoch": 3.6251901429875266,
+      "grad_norm": 0.24804948270320892,
+      "learning_rate": 1.1594170374703088e-06,
+      "loss": 0.0048,
+      "step": 5958
+    },
+    {
+      "epoch": 3.6257986005476117,
+      "grad_norm": 0.3146321475505829,
+      "learning_rate": 1.155672856083867e-06,
+      "loss": 0.0046,
+      "step": 5959
+    },
+    {
+      "epoch": 3.626407058107697,
+      "grad_norm": 0.2864423990249634,
+      "learning_rate": 1.1519345870966703e-06,
+      "loss": 0.0097,
+      "step": 5960
+    },
+    {
+      "epoch": 3.6270155156677824,
+      "grad_norm": 0.26554256677627563,
+      "learning_rate": 1.1482022314356606e-06,
+      "loss": 0.007,
+      "step": 5961
+    },
+    {
+      "epoch": 3.6276239732278674,
+      "grad_norm": 0.2531816363334656,
+      "learning_rate": 1.144475790026306e-06,
+      "loss": 0.0054,
+      "step": 5962
+    },
+    {
+      "epoch": 3.6282324307879525,
+      "grad_norm": 0.2108616828918457,
+      "learning_rate": 1.1407552637926117e-06,
+      "loss": 0.0055,
+      "step": 5963
+    },
+    {
+      "epoch": 3.6288408883480376,
+      "grad_norm": 0.19454440474510193,
+      "learning_rate": 1.1370406536571066e-06,
+      "loss": 0.0038,
+      "step": 5964
+    },
+    {
+      "epoch": 3.6294493459081227,
+      "grad_norm": 0.2602161169052124,
+      "learning_rate": 1.1333319605408622e-06,
+      "loss": 0.0079,
+      "step": 5965
+    },
+    {
+      "epoch": 3.6300578034682083,
+      "grad_norm": 0.22308175265789032,
+      "learning_rate": 1.1296291853634816e-06,
+      "loss": 0.0068,
+      "step": 5966
+    },
+    {
+      "epoch": 3.6306662610282934,
+      "grad_norm": 0.19663815200328827,
+      "learning_rate": 1.1259323290430944e-06,
+      "loss": 0.0054,
+      "step": 5967
+    },
+    {
+      "epoch": 3.6312747185883785,
+      "grad_norm": 0.23746468126773834,
+      "learning_rate": 1.1222413924963705e-06,
+      "loss": 0.0043,
+      "step": 5968
+    },
+    {
+      "epoch": 3.6318831761484636,
+      "grad_norm": 0.19152231514453888,
+      "learning_rate": 1.1185563766385077e-06,
+      "loss": 0.0048,
+      "step": 5969
+    },
+    {
+      "epoch": 3.6324916337085487,
+      "grad_norm": 0.21245577931404114,
+      "learning_rate": 1.1148772823832365e-06,
+      "loss": 0.0034,
+      "step": 5970
+    },
+    {
+      "epoch": 3.633100091268634,
+      "grad_norm": 0.29071682691574097,
+      "learning_rate": 1.1112041106428162e-06,
+      "loss": 0.008,
+      "step": 5971
+    },
+    {
+      "epoch": 3.6337085488287193,
+      "grad_norm": 0.23585431277751923,
+      "learning_rate": 1.107536862328043e-06,
+      "loss": 0.0071,
+      "step": 5972
+    },
+    {
+      "epoch": 3.6343170063888044,
+      "grad_norm": 0.29736191034317017,
+      "learning_rate": 1.1038755383482397e-06,
+      "loss": 0.0092,
+      "step": 5973
+    },
+    {
+      "epoch": 3.6349254639488895,
+      "grad_norm": 0.24729853868484497,
+      "learning_rate": 1.100220139611266e-06,
+      "loss": 0.0049,
+      "step": 5974
+    },
+    {
+      "epoch": 3.6355339215089746,
+      "grad_norm": 0.15655024349689484,
+      "learning_rate": 1.0965706670235081e-06,
+      "loss": 0.0027,
+      "step": 5975
+    },
+    {
+      "epoch": 3.63614237906906,
+      "grad_norm": 0.15523763000965118,
+      "learning_rate": 1.0929271214898755e-06,
+      "loss": 0.0044,
+      "step": 5976
+    },
+    {
+      "epoch": 3.6367508366291452,
+      "grad_norm": 0.33551907539367676,
+      "learning_rate": 1.0892895039138234e-06,
+      "loss": 0.0069,
+      "step": 5977
+    },
+    {
+      "epoch": 3.6373592941892303,
+      "grad_norm": 0.2048521190881729,
+      "learning_rate": 1.0856578151973246e-06,
+      "loss": 0.0038,
+      "step": 5978
+    },
+    {
+      "epoch": 3.6379677517493154,
+      "grad_norm": 0.4035876393318176,
+      "learning_rate": 1.0820320562408947e-06,
+      "loss": 0.013,
+      "step": 5979
+    },
+    {
+      "epoch": 3.6385762093094005,
+      "grad_norm": 0.3847244083881378,
+      "learning_rate": 1.0784122279435565e-06,
+      "loss": 0.0125,
+      "step": 5980
+    },
+    {
+      "epoch": 3.639184666869486,
+      "grad_norm": 0.2535593807697296,
+      "learning_rate": 1.074798331202892e-06,
+      "loss": 0.0073,
+      "step": 5981
+    },
+    {
+      "epoch": 3.639793124429571,
+      "grad_norm": 0.3089277148246765,
+      "learning_rate": 1.0711903669149843e-06,
+      "loss": 0.0079,
+      "step": 5982
+    },
+    {
+      "epoch": 3.6404015819896562,
+      "grad_norm": 0.3164043724536896,
+      "learning_rate": 1.067588335974465e-06,
+      "loss": 0.0064,
+      "step": 5983
+    },
+    {
+      "epoch": 3.6410100395497413,
+      "grad_norm": 0.30803948640823364,
+      "learning_rate": 1.0639922392744889e-06,
+      "loss": 0.0103,
+      "step": 5984
+    },
+    {
+      "epoch": 3.6416184971098264,
+      "grad_norm": 0.32882362604141235,
+      "learning_rate": 1.0604020777067347e-06,
+      "loss": 0.0082,
+      "step": 5985
+    },
+    {
+      "epoch": 3.642226954669912,
+      "grad_norm": 0.9143747091293335,
+      "learning_rate": 1.0568178521614125e-06,
+      "loss": 0.0232,
+      "step": 5986
+    },
+    {
+      "epoch": 3.642835412229997,
+      "grad_norm": 0.43392911553382874,
+      "learning_rate": 1.0532395635272613e-06,
+      "loss": 0.0092,
+      "step": 5987
+    },
+    {
+      "epoch": 3.643443869790082,
+      "grad_norm": 0.18668395280838013,
+      "learning_rate": 1.0496672126915492e-06,
+      "loss": 0.0029,
+      "step": 5988
+    },
+    {
+      "epoch": 3.6440523273501673,
+      "grad_norm": 0.2678108513355255,
+      "learning_rate": 1.046100800540073e-06,
+      "loss": 0.0064,
+      "step": 5989
+    },
+    {
+      "epoch": 3.6446607849102524,
+      "grad_norm": 0.30396801233291626,
+      "learning_rate": 1.0425403279571422e-06,
+      "loss": 0.0081,
+      "step": 5990
+    },
+    {
+      "epoch": 3.645269242470338,
+      "grad_norm": 0.17665399610996246,
+      "learning_rate": 1.0389857958256227e-06,
+      "loss": 0.0039,
+      "step": 5991
+    },
+    {
+      "epoch": 3.645877700030423,
+      "grad_norm": 0.24349285662174225,
+      "learning_rate": 1.0354372050268762e-06,
+      "loss": 0.003,
+      "step": 5992
+    },
+    {
+      "epoch": 3.646486157590508,
+      "grad_norm": 0.22671709954738617,
+      "learning_rate": 1.03189455644081e-06,
+      "loss": 0.0066,
+      "step": 5993
+    },
+    {
+      "epoch": 3.647094615150593,
+      "grad_norm": 0.2378721684217453,
+      "learning_rate": 1.0283578509458548e-06,
+      "loss": 0.0059,
+      "step": 5994
+    },
+    {
+      "epoch": 3.6477030727106783,
+      "grad_norm": 0.3115555942058563,
+      "learning_rate": 1.0248270894189698e-06,
+      "loss": 0.0051,
+      "step": 5995
+    },
+    {
+      "epoch": 3.648311530270764,
+      "grad_norm": 0.3361762464046478,
+      "learning_rate": 1.0213022727356247e-06,
+      "loss": 0.0077,
+      "step": 5996
+    },
+    {
+      "epoch": 3.648919987830849,
+      "grad_norm": 0.27221810817718506,
+      "learning_rate": 1.0177834017698423e-06,
+      "loss": 0.0068,
+      "step": 5997
+    },
+    {
+      "epoch": 3.649528445390934,
+      "grad_norm": 0.2501876652240753,
+      "learning_rate": 1.0142704773941443e-06,
+      "loss": 0.0063,
+      "step": 5998
+    },
+    {
+      "epoch": 3.650136902951019,
+      "grad_norm": 0.23100760579109192,
+      "learning_rate": 1.0107635004795946e-06,
+      "loss": 0.0052,
+      "step": 5999
+    },
+    {
+      "epoch": 3.650745360511104,
+      "grad_norm": 0.18547479808330536,
+      "learning_rate": 1.007262471895773e-06,
+      "loss": 0.004,
+      "step": 6000
+    },
+    {
+      "epoch": 3.6513538180711897,
+      "grad_norm": 0.22204366326332092,
+      "learning_rate": 1.003767392510796e-06,
+      "loss": 0.0041,
+      "step": 6001
+    },
+    {
+      "epoch": 3.651962275631275,
+      "grad_norm": 0.1914484053850174,
+      "learning_rate": 1.00027826319129e-06,
+      "loss": 0.0057,
+      "step": 6002
+    },
+    {
+      "epoch": 3.65257073319136,
+      "grad_norm": 0.18928459286689758,
+      "learning_rate": 9.967950848024183e-07,
+      "loss": 0.0041,
+      "step": 6003
+    },
+    {
+      "epoch": 3.653179190751445,
+      "grad_norm": 0.2991024851799011,
+      "learning_rate": 9.933178582078624e-07,
+      "loss": 0.005,
+      "step": 6004
+    },
+    {
+      "epoch": 3.65378764831153,
+      "grad_norm": 0.12466610223054886,
+      "learning_rate": 9.898465842698323e-07,
+      "loss": 0.0017,
+      "step": 6005
+    },
+    {
+      "epoch": 3.6543961058716157,
+      "grad_norm": 0.5989731550216675,
+      "learning_rate": 9.863812638490511e-07,
+      "loss": 0.016,
+      "step": 6006
+    },
+    {
+      "epoch": 3.6550045634317008,
+      "grad_norm": 0.22727476060390472,
+      "learning_rate": 9.829218978047839e-07,
+      "loss": 0.0059,
+      "step": 6007
+    },
+    {
+      "epoch": 3.655613020991786,
+      "grad_norm": 0.26156729459762573,
+      "learning_rate": 9.794684869948056e-07,
+      "loss": 0.0066,
+      "step": 6008
+    },
+    {
+      "epoch": 3.656221478551871,
+      "grad_norm": 0.24984483420848846,
+      "learning_rate": 9.760210322754175e-07,
+      "loss": 0.0052,
+      "step": 6009
+    },
+    {
+      "epoch": 3.656829936111956,
+      "grad_norm": 0.1444278508424759,
+      "learning_rate": 9.725795345014387e-07,
+      "loss": 0.0034,
+      "step": 6010
+    },
+    {
+      "epoch": 3.6574383936720416,
+      "grad_norm": 0.2517237663269043,
+      "learning_rate": 9.69143994526231e-07,
+      "loss": 0.0065,
+      "step": 6011
+    },
+    {
+      "epoch": 3.6580468512321267,
+      "grad_norm": 0.2072015106678009,
+      "learning_rate": 9.657144132016517e-07,
+      "loss": 0.0028,
+      "step": 6012
+    },
+    {
+      "epoch": 3.658655308792212,
+      "grad_norm": 0.21760626137256622,
+      "learning_rate": 9.62290791378101e-07,
+      "loss": 0.0055,
+      "step": 6013
+    },
+    {
+      "epoch": 3.659263766352297,
+      "grad_norm": 0.22900134325027466,
+      "learning_rate": 9.588731299044945e-07,
+      "loss": 0.0089,
+      "step": 6014
+    },
+    {
+      "epoch": 3.659872223912382,
+      "grad_norm": 0.23601651191711426,
+      "learning_rate": 9.554614296282682e-07,
+      "loss": 0.008,
+      "step": 6015
+    },
+    {
+      "epoch": 3.6604806814724675,
+      "grad_norm": 0.20876143872737885,
+      "learning_rate": 9.52055691395376e-07,
+      "loss": 0.0056,
+      "step": 6016
+    },
+    {
+      "epoch": 3.6610891390325526,
+      "grad_norm": 0.20409031212329865,
+      "learning_rate": 9.486559160503117e-07,
+      "loss": 0.0051,
+      "step": 6017
+    },
+    {
+      "epoch": 3.6616975965926377,
+      "grad_norm": 0.17437663674354553,
+      "learning_rate": 9.452621044360676e-07,
+      "loss": 0.0045,
+      "step": 6018
+    },
+    {
+      "epoch": 3.662306054152723,
+      "grad_norm": 0.7772207260131836,
+      "learning_rate": 9.418742573941707e-07,
+      "loss": 0.042,
+      "step": 6019
+    },
+    {
+      "epoch": 3.662914511712808,
+      "grad_norm": 0.21668609976768494,
+      "learning_rate": 9.384923757646657e-07,
+      "loss": 0.0037,
+      "step": 6020
+    },
+    {
+      "epoch": 3.6635229692728934,
+      "grad_norm": 0.22653992474079132,
+      "learning_rate": 9.351164603861234e-07,
+      "loss": 0.0067,
+      "step": 6021
+    },
+    {
+      "epoch": 3.6641314268329785,
+      "grad_norm": 0.19564014673233032,
+      "learning_rate": 9.317465120956215e-07,
+      "loss": 0.0054,
+      "step": 6022
+    },
+    {
+      "epoch": 3.6647398843930636,
+      "grad_norm": 0.17543290555477142,
+      "learning_rate": 9.283825317287692e-07,
+      "loss": 0.0033,
+      "step": 6023
+    },
+    {
+      "epoch": 3.6653483419531487,
+      "grad_norm": 0.16604088246822357,
+      "learning_rate": 9.250245201196938e-07,
+      "loss": 0.0054,
+      "step": 6024
+    },
+    {
+      "epoch": 3.665956799513234,
+      "grad_norm": 0.25378942489624023,
+      "learning_rate": 9.216724781010461e-07,
+      "loss": 0.005,
+      "step": 6025
+    },
+    {
+      "epoch": 3.6665652570733194,
+      "grad_norm": 0.2088557928800583,
+      "learning_rate": 9.183264065039859e-07,
+      "loss": 0.0033,
+      "step": 6026
+    },
+    {
+      "epoch": 3.6671737146334045,
+      "grad_norm": 0.22516481578350067,
+      "learning_rate": 9.149863061582053e-07,
+      "loss": 0.0092,
+      "step": 6027
+    },
+    {
+      "epoch": 3.6677821721934896,
+      "grad_norm": 0.31587862968444824,
+      "learning_rate": 9.116521778919085e-07,
+      "loss": 0.0089,
+      "step": 6028
+    },
+    {
+      "epoch": 3.6683906297535747,
+      "grad_norm": 0.34137919545173645,
+      "learning_rate": 9.083240225318202e-07,
+      "loss": 0.0135,
+      "step": 6029
+    },
+    {
+      "epoch": 3.6689990873136598,
+      "grad_norm": 0.21958407759666443,
+      "learning_rate": 9.050018409031801e-07,
+      "loss": 0.0044,
+      "step": 6030
+    },
+    {
+      "epoch": 3.6696075448737453,
+      "grad_norm": 0.32528388500213623,
+      "learning_rate": 9.016856338297602e-07,
+      "loss": 0.0076,
+      "step": 6031
+    },
+    {
+      "epoch": 3.6702160024338304,
+      "grad_norm": 0.21904724836349487,
+      "learning_rate": 8.983754021338331e-07,
+      "loss": 0.0048,
+      "step": 6032
+    },
+    {
+      "epoch": 3.6708244599939155,
+      "grad_norm": 0.3020154535770416,
+      "learning_rate": 8.950711466362005e-07,
+      "loss": 0.006,
+      "step": 6033
+    },
+    {
+      "epoch": 3.6714329175540006,
+      "grad_norm": 0.2894566059112549,
+      "learning_rate": 8.917728681561793e-07,
+      "loss": 0.0057,
+      "step": 6034
+    },
+    {
+      "epoch": 3.6720413751140857,
+      "grad_norm": 0.2588036060333252,
+      "learning_rate": 8.884805675116098e-07,
+      "loss": 0.0056,
+      "step": 6035
+    },
+    {
+      "epoch": 3.672649832674171,
+      "grad_norm": 0.2685299813747406,
+      "learning_rate": 8.851942455188362e-07,
+      "loss": 0.007,
+      "step": 6036
+    },
+    {
+      "epoch": 3.6732582902342563,
+      "grad_norm": 0.19193175435066223,
+      "learning_rate": 8.819139029927425e-07,
+      "loss": 0.0032,
+      "step": 6037
+    },
+    {
+      "epoch": 3.6738667477943414,
+      "grad_norm": 0.20610147714614868,
+      "learning_rate": 8.786395407467062e-07,
+      "loss": 0.0036,
+      "step": 6038
+    },
+    {
+      "epoch": 3.6744752053544265,
+      "grad_norm": 0.17886589467525482,
+      "learning_rate": 8.753711595926334e-07,
+      "loss": 0.0046,
+      "step": 6039
+    },
+    {
+      "epoch": 3.6750836629145116,
+      "grad_norm": 0.2853025794029236,
+      "learning_rate": 8.721087603409506e-07,
+      "loss": 0.0057,
+      "step": 6040
+    },
+    {
+      "epoch": 3.675692120474597,
+      "grad_norm": 0.22553856670856476,
+      "learning_rate": 8.688523438005996e-07,
+      "loss": 0.0043,
+      "step": 6041
+    },
+    {
+      "epoch": 3.6763005780346822,
+      "grad_norm": 0.20292532444000244,
+      "learning_rate": 8.656019107790237e-07,
+      "loss": 0.0061,
+      "step": 6042
+    },
+    {
+      "epoch": 3.6769090355947673,
+      "grad_norm": 0.24954308569431305,
+      "learning_rate": 8.623574620822083e-07,
+      "loss": 0.0065,
+      "step": 6043
+    },
+    {
+      "epoch": 3.6775174931548524,
+      "grad_norm": 0.36442530155181885,
+      "learning_rate": 8.591189985146352e-07,
+      "loss": 0.0074,
+      "step": 6044
+    },
+    {
+      "epoch": 3.6781259507149375,
+      "grad_norm": 0.26411545276641846,
+      "learning_rate": 8.558865208793093e-07,
+      "loss": 0.0049,
+      "step": 6045
+    },
+    {
+      "epoch": 3.678734408275023,
+      "grad_norm": 0.2690979838371277,
+      "learning_rate": 8.526600299777448e-07,
+      "loss": 0.0055,
+      "step": 6046
+    },
+    {
+      "epoch": 3.679342865835108,
+      "grad_norm": 0.14458657801151276,
+      "learning_rate": 8.49439526609988e-07,
+      "loss": 0.0029,
+      "step": 6047
+    },
+    {
+      "epoch": 3.6799513233951933,
+      "grad_norm": 0.2396114468574524,
+      "learning_rate": 8.462250115745807e-07,
+      "loss": 0.0076,
+      "step": 6048
+    },
+    {
+      "epoch": 3.6805597809552784,
+      "grad_norm": 0.22277654707431793,
+      "learning_rate": 8.430164856685935e-07,
+      "loss": 0.0046,
+      "step": 6049
+    },
+    {
+      "epoch": 3.6811682385153635,
+      "grad_norm": 0.22810159623622894,
+      "learning_rate": 8.398139496876012e-07,
+      "loss": 0.0072,
+      "step": 6050
+    },
+    {
+      "epoch": 3.681776696075449,
+      "grad_norm": 0.21732278168201447,
+      "learning_rate": 8.366174044257103e-07,
+      "loss": 0.0048,
+      "step": 6051
+    },
+    {
+      "epoch": 3.6823851536355336,
+      "grad_norm": 0.22662782669067383,
+      "learning_rate": 8.334268506755144e-07,
+      "loss": 0.0038,
+      "step": 6052
+    },
+    {
+      "epoch": 3.682993611195619,
+      "grad_norm": 0.26345643401145935,
+      "learning_rate": 8.302422892281558e-07,
+      "loss": 0.0067,
+      "step": 6053
+    },
+    {
+      "epoch": 3.6836020687557043,
+      "grad_norm": 0.24986715614795685,
+      "learning_rate": 8.270637208732585e-07,
+      "loss": 0.0056,
+      "step": 6054
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.31368815898895264,
+      "learning_rate": 8.238911463989835e-07,
+      "loss": 0.0067,
+      "step": 6055
+    },
+    {
+      "epoch": 3.684818983875875,
+      "grad_norm": 0.6407575607299805,
+      "learning_rate": 8.207245665919932e-07,
+      "loss": 0.0278,
+      "step": 6056
+    },
+    {
+      "epoch": 3.6854274414359596,
+      "grad_norm": 0.259106308221817,
+      "learning_rate": 8.175639822374709e-07,
+      "loss": 0.0082,
+      "step": 6057
+    },
+    {
+      "epoch": 3.686035898996045,
+      "grad_norm": 0.1873266100883484,
+      "learning_rate": 8.144093941191061e-07,
+      "loss": 0.0042,
+      "step": 6058
+    },
+    {
+      "epoch": 3.68664435655613,
+      "grad_norm": 0.3827762007713318,
+      "learning_rate": 8.112608030191037e-07,
+      "loss": 0.0117,
+      "step": 6059
+    },
+    {
+      "epoch": 3.6872528141162153,
+      "grad_norm": 0.3754812479019165,
+      "learning_rate": 8.081182097181894e-07,
+      "loss": 0.0145,
+      "step": 6060
+    },
+    {
+      "epoch": 3.687861271676301,
+      "grad_norm": 0.26298126578330994,
+      "learning_rate": 8.049816149955896e-07,
+      "loss": 0.0056,
+      "step": 6061
+    },
+    {
+      "epoch": 3.6884697292363855,
+      "grad_norm": 0.1487468034029007,
+      "learning_rate": 8.018510196290519e-07,
+      "loss": 0.0029,
+      "step": 6062
+    },
+    {
+      "epoch": 3.689078186796471,
+      "grad_norm": 0.30363762378692627,
+      "learning_rate": 7.987264243948356e-07,
+      "loss": 0.0086,
+      "step": 6063
+    },
+    {
+      "epoch": 3.689686644356556,
+      "grad_norm": 0.1642022281885147,
+      "learning_rate": 7.956078300677045e-07,
+      "loss": 0.0029,
+      "step": 6064
+    },
+    {
+      "epoch": 3.6902951019166412,
+      "grad_norm": 0.2749357521533966,
+      "learning_rate": 7.924952374209399e-07,
+      "loss": 0.008,
+      "step": 6065
+    },
+    {
+      "epoch": 3.6909035594767268,
+      "grad_norm": 0.14298266172409058,
+      "learning_rate": 7.893886472263412e-07,
+      "loss": 0.0026,
+      "step": 6066
+    },
+    {
+      "epoch": 3.6915120170368114,
+      "grad_norm": 0.2198246717453003,
+      "learning_rate": 7.862880602542116e-07,
+      "loss": 0.005,
+      "step": 6067
+    },
+    {
+      "epoch": 3.692120474596897,
+      "grad_norm": 0.21324285864830017,
+      "learning_rate": 7.83193477273364e-07,
+      "loss": 0.0043,
+      "step": 6068
+    },
+    {
+      "epoch": 3.692728932156982,
+      "grad_norm": 0.3092917203903198,
+      "learning_rate": 7.801048990511262e-07,
+      "loss": 0.0179,
+      "step": 6069
+    },
+    {
+      "epoch": 3.693337389717067,
+      "grad_norm": 0.24335362017154694,
+      "learning_rate": 7.770223263533411e-07,
+      "loss": 0.0066,
+      "step": 6070
+    },
+    {
+      "epoch": 3.6939458472771527,
+      "grad_norm": 0.27647829055786133,
+      "learning_rate": 7.739457599443528e-07,
+      "loss": 0.0069,
+      "step": 6071
+    },
+    {
+      "epoch": 3.6945543048372373,
+      "grad_norm": 0.27333924174308777,
+      "learning_rate": 7.708752005870263e-07,
+      "loss": 0.0076,
+      "step": 6072
+    },
+    {
+      "epoch": 3.695162762397323,
+      "grad_norm": 0.21789668500423431,
+      "learning_rate": 7.6781064904273e-07,
+      "loss": 0.0057,
+      "step": 6073
+    },
+    {
+      "epoch": 3.695771219957408,
+      "grad_norm": 0.25622355937957764,
+      "learning_rate": 7.647521060713452e-07,
+      "loss": 0.0054,
+      "step": 6074
+    },
+    {
+      "epoch": 3.696379677517493,
+      "grad_norm": 0.28123655915260315,
+      "learning_rate": 7.616995724312626e-07,
+      "loss": 0.0085,
+      "step": 6075
+    },
+    {
+      "epoch": 3.696988135077578,
+      "grad_norm": 0.2564314305782318,
+      "learning_rate": 7.586530488793847e-07,
+      "loss": 0.0069,
+      "step": 6076
+    },
+    {
+      "epoch": 3.6975965926376633,
+      "grad_norm": 0.30161136388778687,
+      "learning_rate": 7.556125361711214e-07,
+      "loss": 0.0059,
+      "step": 6077
+    },
+    {
+      "epoch": 3.698205050197749,
+      "grad_norm": 0.17149339616298676,
+      "learning_rate": 7.525780350603917e-07,
+      "loss": 0.0035,
+      "step": 6078
+    },
+    {
+      "epoch": 3.698813507757834,
+      "grad_norm": 0.2607922852039337,
+      "learning_rate": 7.495495462996327e-07,
+      "loss": 0.0156,
+      "step": 6079
+    },
+    {
+      "epoch": 3.699421965317919,
+      "grad_norm": 0.30373069643974304,
+      "learning_rate": 7.465270706397714e-07,
+      "loss": 0.0071,
+      "step": 6080
+    },
+    {
+      "epoch": 3.700030422878004,
+      "grad_norm": 0.17537471652030945,
+      "learning_rate": 7.43510608830264e-07,
+      "loss": 0.0047,
+      "step": 6081
+    },
+    {
+      "epoch": 3.700638880438089,
+      "grad_norm": 0.2823267877101898,
+      "learning_rate": 7.405001616190649e-07,
+      "loss": 0.0044,
+      "step": 6082
+    },
+    {
+      "epoch": 3.7012473379981747,
+      "grad_norm": 0.3912118375301361,
+      "learning_rate": 7.374957297526408e-07,
+      "loss": 0.006,
+      "step": 6083
+    },
+    {
+      "epoch": 3.70185579555826,
+      "grad_norm": 0.2731781303882599,
+      "learning_rate": 7.344973139759654e-07,
+      "loss": 0.0076,
+      "step": 6084
+    },
+    {
+      "epoch": 3.702464253118345,
+      "grad_norm": 0.23557397723197937,
+      "learning_rate": 7.315049150325187e-07,
+      "loss": 0.0049,
+      "step": 6085
+    },
+    {
+      "epoch": 3.70307271067843,
+      "grad_norm": 0.19650277495384216,
+      "learning_rate": 7.285185336642908e-07,
+      "loss": 0.0042,
+      "step": 6086
+    },
+    {
+      "epoch": 3.703681168238515,
+      "grad_norm": 0.18600913882255554,
+      "learning_rate": 7.255381706117837e-07,
+      "loss": 0.0031,
+      "step": 6087
+    },
+    {
+      "epoch": 3.7042896257986007,
+      "grad_norm": 0.21114131808280945,
+      "learning_rate": 7.225638266140006e-07,
+      "loss": 0.0051,
+      "step": 6088
+    },
+    {
+      "epoch": 3.7048980833586858,
+      "grad_norm": 0.2112278938293457,
+      "learning_rate": 7.195955024084544e-07,
+      "loss": 0.0045,
+      "step": 6089
+    },
+    {
+      "epoch": 3.705506540918771,
+      "grad_norm": 0.16948889195919037,
+      "learning_rate": 7.166331987311675e-07,
+      "loss": 0.0044,
+      "step": 6090
+    },
+    {
+      "epoch": 3.706114998478856,
+      "grad_norm": 0.19342589378356934,
+      "learning_rate": 7.136769163166662e-07,
+      "loss": 0.0037,
+      "step": 6091
+    },
+    {
+      "epoch": 3.706723456038941,
+      "grad_norm": 0.2661781311035156,
+      "learning_rate": 7.107266558979864e-07,
+      "loss": 0.0043,
+      "step": 6092
+    },
+    {
+      "epoch": 3.7073319135990266,
+      "grad_norm": 0.23141947388648987,
+      "learning_rate": 7.077824182066678e-07,
+      "loss": 0.0044,
+      "step": 6093
+    },
+    {
+      "epoch": 3.7079403711591117,
+      "grad_norm": 0.25309503078460693,
+      "learning_rate": 7.048442039727627e-07,
+      "loss": 0.004,
+      "step": 6094
+    },
+    {
+      "epoch": 3.7085488287191968,
+      "grad_norm": 0.2142305076122284,
+      "learning_rate": 7.019120139248187e-07,
+      "loss": 0.005,
+      "step": 6095
+    },
+    {
+      "epoch": 3.709157286279282,
+      "grad_norm": 0.2936217188835144,
+      "learning_rate": 6.989858487899043e-07,
+      "loss": 0.0058,
+      "step": 6096
+    },
+    {
+      "epoch": 3.709765743839367,
+      "grad_norm": 0.22409573197364807,
+      "learning_rate": 6.960657092935807e-07,
+      "loss": 0.0069,
+      "step": 6097
+    },
+    {
+      "epoch": 3.7103742013994525,
+      "grad_norm": 0.21282601356506348,
+      "learning_rate": 6.931515961599244e-07,
+      "loss": 0.0028,
+      "step": 6098
+    },
+    {
+      "epoch": 3.7109826589595376,
+      "grad_norm": 0.22673866152763367,
+      "learning_rate": 6.902435101115129e-07,
+      "loss": 0.0061,
+      "step": 6099
+    },
+    {
+      "epoch": 3.7115911165196227,
+      "grad_norm": 0.181802436709404,
+      "learning_rate": 6.873414518694332e-07,
+      "loss": 0.0035,
+      "step": 6100
+    },
+    {
+      "epoch": 3.712199574079708,
+      "grad_norm": 0.24081139266490936,
+      "learning_rate": 6.844454221532682e-07,
+      "loss": 0.0051,
+      "step": 6101
+    },
+    {
+      "epoch": 3.712808031639793,
+      "grad_norm": 0.19030770659446716,
+      "learning_rate": 6.815554216811182e-07,
+      "loss": 0.0031,
+      "step": 6102
+    },
+    {
+      "epoch": 3.7134164891998784,
+      "grad_norm": 0.2784750163555145,
+      "learning_rate": 6.786714511695796e-07,
+      "loss": 0.0089,
+      "step": 6103
+    },
+    {
+      "epoch": 3.7140249467599635,
+      "grad_norm": 0.26827266812324524,
+      "learning_rate": 6.75793511333761e-07,
+      "loss": 0.0047,
+      "step": 6104
+    },
+    {
+      "epoch": 3.7146334043200486,
+      "grad_norm": 0.2427385449409485,
+      "learning_rate": 6.729216028872637e-07,
+      "loss": 0.0047,
+      "step": 6105
+    },
+    {
+      "epoch": 3.7152418618801337,
+      "grad_norm": 0.19240151345729828,
+      "learning_rate": 6.700557265422097e-07,
+      "loss": 0.0039,
+      "step": 6106
+    },
+    {
+      "epoch": 3.715850319440219,
+      "grad_norm": 0.27845215797424316,
+      "learning_rate": 6.67195883009214e-07,
+      "loss": 0.0066,
+      "step": 6107
+    },
+    {
+      "epoch": 3.7164587770003044,
+      "grad_norm": 0.23421838879585266,
+      "learning_rate": 6.643420729973954e-07,
+      "loss": 0.0073,
+      "step": 6108
+    },
+    {
+      "epoch": 3.7170672345603895,
+      "grad_norm": 0.23832927644252777,
+      "learning_rate": 6.614942972143822e-07,
+      "loss": 0.0045,
+      "step": 6109
+    },
+    {
+      "epoch": 3.7176756921204746,
+      "grad_norm": 0.47682875394821167,
+      "learning_rate": 6.586525563663099e-07,
+      "loss": 0.0115,
+      "step": 6110
+    },
+    {
+      "epoch": 3.7182841496805596,
+      "grad_norm": 0.2529081702232361,
+      "learning_rate": 6.558168511577978e-07,
+      "loss": 0.0061,
+      "step": 6111
+    },
+    {
+      "epoch": 3.7188926072406447,
+      "grad_norm": 0.3189617693424225,
+      "learning_rate": 6.529871822919975e-07,
+      "loss": 0.0086,
+      "step": 6112
+    },
+    {
+      "epoch": 3.7195010648007303,
+      "grad_norm": 0.21395058929920197,
+      "learning_rate": 6.501635504705422e-07,
+      "loss": 0.0042,
+      "step": 6113
+    },
+    {
+      "epoch": 3.7201095223608154,
+      "grad_norm": 0.1643410176038742,
+      "learning_rate": 6.473459563935747e-07,
+      "loss": 0.0024,
+      "step": 6114
+    },
+    {
+      "epoch": 3.7207179799209005,
+      "grad_norm": 0.15923701226711273,
+      "learning_rate": 6.445344007597387e-07,
+      "loss": 0.0026,
+      "step": 6115
+    },
+    {
+      "epoch": 3.7213264374809856,
+      "grad_norm": 0.2772526741027832,
+      "learning_rate": 6.417288842661878e-07,
+      "loss": 0.0062,
+      "step": 6116
+    },
+    {
+      "epoch": 3.7219348950410707,
+      "grad_norm": 0.21529936790466309,
+      "learning_rate": 6.389294076085684e-07,
+      "loss": 0.0061,
+      "step": 6117
+    },
+    {
+      "epoch": 3.722543352601156,
+      "grad_norm": 0.23362010717391968,
+      "learning_rate": 6.361359714810389e-07,
+      "loss": 0.0059,
+      "step": 6118
+    },
+    {
+      "epoch": 3.7231518101612413,
+      "grad_norm": 0.23200243711471558,
+      "learning_rate": 6.333485765762509e-07,
+      "loss": 0.0065,
+      "step": 6119
+    },
+    {
+      "epoch": 3.7237602677213264,
+      "grad_norm": 0.22773078083992004,
+      "learning_rate": 6.305672235853682e-07,
+      "loss": 0.0042,
+      "step": 6120
+    },
+    {
+      "epoch": 3.7243687252814115,
+      "grad_norm": 0.29405245184898376,
+      "learning_rate": 6.277919131980392e-07,
+      "loss": 0.0076,
+      "step": 6121
+    },
+    {
+      "epoch": 3.7249771828414966,
+      "grad_norm": 0.16285215318202972,
+      "learning_rate": 6.250226461024383e-07,
+      "loss": 0.0032,
+      "step": 6122
+    },
+    {
+      "epoch": 3.725585640401582,
+      "grad_norm": 0.2497759461402893,
+      "learning_rate": 6.222594229852163e-07,
+      "loss": 0.0056,
+      "step": 6123
+    },
+    {
+      "epoch": 3.7261940979616672,
+      "grad_norm": 0.18337193131446838,
+      "learning_rate": 6.195022445315474e-07,
+      "loss": 0.0032,
+      "step": 6124
+    },
+    {
+      "epoch": 3.7268025555217523,
+      "grad_norm": 0.2235439568758011,
+      "learning_rate": 6.167511114250901e-07,
+      "loss": 0.0057,
+      "step": 6125
+    },
+    {
+      "epoch": 3.7274110130818374,
+      "grad_norm": 0.26169970631599426,
+      "learning_rate": 6.140060243480156e-07,
+      "loss": 0.0074,
+      "step": 6126
+    },
+    {
+      "epoch": 3.7280194706419225,
+      "grad_norm": 0.41065341234207153,
+      "learning_rate": 6.112669839809876e-07,
+      "loss": 0.0077,
+      "step": 6127
+    },
+    {
+      "epoch": 3.728627928202008,
+      "grad_norm": 0.22760555148124695,
+      "learning_rate": 6.08533991003174e-07,
+      "loss": 0.0051,
+      "step": 6128
+    },
+    {
+      "epoch": 3.729236385762093,
+      "grad_norm": 0.3023330867290497,
+      "learning_rate": 6.058070460922466e-07,
+      "loss": 0.0043,
+      "step": 6129
+    },
+    {
+      "epoch": 3.7298448433221782,
+      "grad_norm": 0.17906133830547333,
+      "learning_rate": 6.030861499243701e-07,
+      "loss": 0.0027,
+      "step": 6130
+    },
+    {
+      "epoch": 3.7304533008822633,
+      "grad_norm": 0.18436305224895477,
+      "learning_rate": 6.003713031742131e-07,
+      "loss": 0.0043,
+      "step": 6131
+    },
+    {
+      "epoch": 3.7310617584423484,
+      "grad_norm": 0.16411884129047394,
+      "learning_rate": 5.97662506514951e-07,
+      "loss": 0.003,
+      "step": 6132
+    },
+    {
+      "epoch": 3.731670216002434,
+      "grad_norm": 0.28377214074134827,
+      "learning_rate": 5.949597606182439e-07,
+      "loss": 0.0052,
+      "step": 6133
+    },
+    {
+      "epoch": 3.732278673562519,
+      "grad_norm": 0.24661727249622345,
+      "learning_rate": 5.922630661542639e-07,
+      "loss": 0.0072,
+      "step": 6134
+    },
+    {
+      "epoch": 3.732887131122604,
+      "grad_norm": 0.23685060441493988,
+      "learning_rate": 5.895724237916816e-07,
+      "loss": 0.0057,
+      "step": 6135
+    },
+    {
+      "epoch": 3.7334955886826893,
+      "grad_norm": 0.24538807570934296,
+      "learning_rate": 5.868878341976608e-07,
+      "loss": 0.006,
+      "step": 6136
+    },
+    {
+      "epoch": 3.7341040462427744,
+      "grad_norm": 0.23321305215358734,
+      "learning_rate": 5.842092980378688e-07,
+      "loss": 0.0036,
+      "step": 6137
+    },
+    {
+      "epoch": 3.73471250380286,
+      "grad_norm": 0.2515392005443573,
+      "learning_rate": 5.815368159764689e-07,
+      "loss": 0.0044,
+      "step": 6138
+    },
+    {
+      "epoch": 3.735320961362945,
+      "grad_norm": 0.3265810012817383,
+      "learning_rate": 5.788703886761254e-07,
+      "loss": 0.0064,
+      "step": 6139
+    },
+    {
+      "epoch": 3.73592941892303,
+      "grad_norm": 0.2389630675315857,
+      "learning_rate": 5.762100167980067e-07,
+      "loss": 0.0038,
+      "step": 6140
+    },
+    {
+      "epoch": 3.736537876483115,
+      "grad_norm": 0.26586848497390747,
+      "learning_rate": 5.735557010017656e-07,
+      "loss": 0.0071,
+      "step": 6141
+    },
+    {
+      "epoch": 3.7371463340432003,
+      "grad_norm": 0.1601967066526413,
+      "learning_rate": 5.709074419455701e-07,
+      "loss": 0.0019,
+      "step": 6142
+    },
+    {
+      "epoch": 3.737754791603286,
+      "grad_norm": 0.29823827743530273,
+      "learning_rate": 5.682652402860727e-07,
+      "loss": 0.0091,
+      "step": 6143
+    },
+    {
+      "epoch": 3.738363249163371,
+      "grad_norm": 0.24698449671268463,
+      "learning_rate": 5.6562909667843e-07,
+      "loss": 0.0061,
+      "step": 6144
+    },
+    {
+      "epoch": 3.738971706723456,
+      "grad_norm": 0.23663544654846191,
+      "learning_rate": 5.629990117762968e-07,
+      "loss": 0.0065,
+      "step": 6145
+    },
+    {
+      "epoch": 3.739580164283541,
+      "grad_norm": 0.300060898065567,
+      "learning_rate": 5.603749862318292e-07,
+      "loss": 0.0047,
+      "step": 6146
+    },
+    {
+      "epoch": 3.740188621843626,
+      "grad_norm": 0.13921256363391876,
+      "learning_rate": 5.577570206956623e-07,
+      "loss": 0.0032,
+      "step": 6147
+    },
+    {
+      "epoch": 3.7407970794037118,
+      "grad_norm": 0.15815310180187225,
+      "learning_rate": 5.551451158169602e-07,
+      "loss": 0.0035,
+      "step": 6148
+    },
+    {
+      "epoch": 3.741405536963797,
+      "grad_norm": 0.21306031942367554,
+      "learning_rate": 5.52539272243352e-07,
+      "loss": 0.0043,
+      "step": 6149
+    },
+    {
+      "epoch": 3.742013994523882,
+      "grad_norm": 0.2478276789188385,
+      "learning_rate": 5.499394906209876e-07,
+      "loss": 0.0076,
+      "step": 6150
+    },
+    {
+      "epoch": 3.742622452083967,
+      "grad_norm": 0.18793390691280365,
+      "learning_rate": 5.473457715944957e-07,
+      "loss": 0.0035,
+      "step": 6151
+    },
+    {
+      "epoch": 3.743230909644052,
+      "grad_norm": 0.2907355725765228,
+      "learning_rate": 5.447581158070203e-07,
+      "loss": 0.0041,
+      "step": 6152
+    },
+    {
+      "epoch": 3.7438393672041377,
+      "grad_norm": 0.31171515583992004,
+      "learning_rate": 5.42176523900187e-07,
+      "loss": 0.0065,
+      "step": 6153
+    },
+    {
+      "epoch": 3.7444478247642228,
+      "grad_norm": 0.27889809012413025,
+      "learning_rate": 5.396009965141197e-07,
+      "loss": 0.0062,
+      "step": 6154
+    },
+    {
+      "epoch": 3.745056282324308,
+      "grad_norm": 0.2419702708721161,
+      "learning_rate": 5.370315342874494e-07,
+      "loss": 0.0047,
+      "step": 6155
+    },
+    {
+      "epoch": 3.745664739884393,
+      "grad_norm": 0.22070972621440887,
+      "learning_rate": 5.344681378572913e-07,
+      "loss": 0.0048,
+      "step": 6156
+    },
+    {
+      "epoch": 3.746273197444478,
+      "grad_norm": 0.2477099597454071,
+      "learning_rate": 5.319108078592567e-07,
+      "loss": 0.0039,
+      "step": 6157
+    },
+    {
+      "epoch": 3.7468816550045636,
+      "grad_norm": 0.1542275846004486,
+      "learning_rate": 5.293595449274685e-07,
+      "loss": 0.0033,
+      "step": 6158
+    },
+    {
+      "epoch": 3.7474901125646487,
+      "grad_norm": 0.17303107678890228,
+      "learning_rate": 5.268143496945239e-07,
+      "loss": 0.0031,
+      "step": 6159
+    },
+    {
+      "epoch": 3.748098570124734,
+      "grad_norm": 0.26074641942977905,
+      "learning_rate": 5.242752227915287e-07,
+      "loss": 0.0071,
+      "step": 6160
+    },
+    {
+      "epoch": 3.748707027684819,
+      "grad_norm": 0.20203956961631775,
+      "learning_rate": 5.217421648480769e-07,
+      "loss": 0.0051,
+      "step": 6161
+    },
+    {
+      "epoch": 3.749315485244904,
+      "grad_norm": 0.18085290491580963,
+      "learning_rate": 5.192151764922659e-07,
+      "loss": 0.0027,
+      "step": 6162
+    },
+    {
+      "epoch": 3.7499239428049895,
+      "grad_norm": 0.31731289625167847,
+      "learning_rate": 5.166942583506806e-07,
+      "loss": 0.0055,
+      "step": 6163
+    },
+    {
+      "epoch": 3.7505324003650746,
+      "grad_norm": 0.1756841391324997,
+      "learning_rate": 5.141794110484071e-07,
+      "loss": 0.0023,
+      "step": 6164
+    },
+    {
+      "epoch": 3.7511408579251597,
+      "grad_norm": 0.17156068980693817,
+      "learning_rate": 5.116706352090189e-07,
+      "loss": 0.003,
+      "step": 6165
+    },
+    {
+      "epoch": 3.751749315485245,
+      "grad_norm": 0.2008720338344574,
+      "learning_rate": 5.091679314545905e-07,
+      "loss": 0.0039,
+      "step": 6166
+    },
+    {
+      "epoch": 3.75235777304533,
+      "grad_norm": 0.1687059849500656,
+      "learning_rate": 5.066713004056839e-07,
+      "loss": 0.0052,
+      "step": 6167
+    },
+    {
+      "epoch": 3.7529662306054155,
+      "grad_norm": 0.2529749572277069,
+      "learning_rate": 5.041807426813649e-07,
+      "loss": 0.0085,
+      "step": 6168
+    },
+    {
+      "epoch": 3.7535746881655006,
+      "grad_norm": 0.2726878225803375,
+      "learning_rate": 5.01696258899187e-07,
+      "loss": 0.0089,
+      "step": 6169
+    },
+    {
+      "epoch": 3.7541831457255856,
+      "grad_norm": 0.14040397107601166,
+      "learning_rate": 4.992178496751931e-07,
+      "loss": 0.0023,
+      "step": 6170
+    },
+    {
+      "epoch": 3.7547916032856707,
+      "grad_norm": 0.2372034788131714,
+      "learning_rate": 4.967455156239337e-07,
+      "loss": 0.0057,
+      "step": 6171
+    },
+    {
+      "epoch": 3.755400060845756,
+      "grad_norm": 0.22318537533283234,
+      "learning_rate": 4.942792573584404e-07,
+      "loss": 0.0033,
+      "step": 6172
+    },
+    {
+      "epoch": 3.7560085184058414,
+      "grad_norm": 0.22246554493904114,
+      "learning_rate": 4.918190754902408e-07,
+      "loss": 0.0038,
+      "step": 6173
+    },
+    {
+      "epoch": 3.7566169759659265,
+      "grad_norm": 0.18072238564491272,
+      "learning_rate": 4.89364970629358e-07,
+      "loss": 0.0038,
+      "step": 6174
+    },
+    {
+      "epoch": 3.7572254335260116,
+      "grad_norm": 0.2487388253211975,
+      "learning_rate": 4.86916943384308e-07,
+      "loss": 0.0067,
+      "step": 6175
+    },
+    {
+      "epoch": 3.7578338910860967,
+      "grad_norm": 0.1718432605266571,
+      "learning_rate": 4.844749943621052e-07,
+      "loss": 0.0029,
+      "step": 6176
+    },
+    {
+      "epoch": 3.7584423486461818,
+      "grad_norm": 0.20204098522663116,
+      "learning_rate": 4.820391241682404e-07,
+      "loss": 0.0034,
+      "step": 6177
+    },
+    {
+      "epoch": 3.7590508062062673,
+      "grad_norm": 0.20857053995132446,
+      "learning_rate": 4.796093334067192e-07,
+      "loss": 0.0042,
+      "step": 6178
+    },
+    {
+      "epoch": 3.7596592637663524,
+      "grad_norm": 0.2744029760360718,
+      "learning_rate": 4.771856226800209e-07,
+      "loss": 0.0074,
+      "step": 6179
+    },
+    {
+      "epoch": 3.7602677213264375,
+      "grad_norm": 0.33221253752708435,
+      "learning_rate": 4.7476799258912574e-07,
+      "loss": 0.0068,
+      "step": 6180
+    },
+    {
+      "epoch": 3.7608761788865226,
+      "grad_norm": 0.3299937844276428,
+      "learning_rate": 4.7235644373350707e-07,
+      "loss": 0.0058,
+      "step": 6181
+    },
+    {
+      "epoch": 3.7614846364466077,
+      "grad_norm": 0.33729061484336853,
+      "learning_rate": 4.69950976711131e-07,
+      "loss": 0.0079,
+      "step": 6182
+    },
+    {
+      "epoch": 3.7620930940066932,
+      "grad_norm": 0.26489993929862976,
+      "learning_rate": 4.675515921184481e-07,
+      "loss": 0.0064,
+      "step": 6183
+    },
+    {
+      "epoch": 3.7627015515667783,
+      "grad_norm": 0.14743492007255554,
+      "learning_rate": 4.651582905504048e-07,
+      "loss": 0.0022,
+      "step": 6184
+    },
+    {
+      "epoch": 3.7633100091268634,
+      "grad_norm": 0.21888674795627594,
+      "learning_rate": 4.627710726004458e-07,
+      "loss": 0.0074,
+      "step": 6185
+    },
+    {
+      "epoch": 3.7639184666869485,
+      "grad_norm": 0.21569287776947021,
+      "learning_rate": 4.6038993886049484e-07,
+      "loss": 0.005,
+      "step": 6186
+    },
+    {
+      "epoch": 3.7645269242470336,
+      "grad_norm": 0.22505217790603638,
+      "learning_rate": 4.5801488992098243e-07,
+      "loss": 0.0038,
+      "step": 6187
+    },
+    {
+      "epoch": 3.765135381807119,
+      "grad_norm": 0.196213498711586,
+      "learning_rate": 4.5564592637081517e-07,
+      "loss": 0.0047,
+      "step": 6188
+    },
+    {
+      "epoch": 3.7657438393672042,
+      "grad_norm": 0.26138564944267273,
+      "learning_rate": 4.5328304879739823e-07,
+      "loss": 0.0047,
+      "step": 6189
+    },
+    {
+      "epoch": 3.7663522969272893,
+      "grad_norm": 0.30381667613983154,
+      "learning_rate": 4.509262577866269e-07,
+      "loss": 0.0102,
+      "step": 6190
+    },
+    {
+      "epoch": 3.7669607544873744,
+      "grad_norm": 0.17448806762695312,
+      "learning_rate": 4.4857555392288917e-07,
+      "loss": 0.0023,
+      "step": 6191
+    },
+    {
+      "epoch": 3.7675692120474595,
+      "grad_norm": 0.2946946620941162,
+      "learning_rate": 4.4623093778906053e-07,
+      "loss": 0.0049,
+      "step": 6192
+    },
+    {
+      "epoch": 3.768177669607545,
+      "grad_norm": 0.23686204850673676,
+      "learning_rate": 4.438924099665065e-07,
+      "loss": 0.0048,
+      "step": 6193
+    },
+    {
+      "epoch": 3.76878612716763,
+      "grad_norm": 0.2854866087436676,
+      "learning_rate": 4.4155997103508817e-07,
+      "loss": 0.0049,
+      "step": 6194
+    },
+    {
+      "epoch": 3.7693945847277153,
+      "grad_norm": 0.22638092935085297,
+      "learning_rate": 4.392336215731513e-07,
+      "loss": 0.0056,
+      "step": 6195
+    },
+    {
+      "epoch": 3.7700030422878004,
+      "grad_norm": 0.1855926811695099,
+      "learning_rate": 4.369133621575289e-07,
+      "loss": 0.0034,
+      "step": 6196
+    },
+    {
+      "epoch": 3.7706114998478855,
+      "grad_norm": 0.21199901401996613,
+      "learning_rate": 4.3459919336355514e-07,
+      "loss": 0.0035,
+      "step": 6197
+    },
+    {
+      "epoch": 3.771219957407971,
+      "grad_norm": 0.1068754643201828,
+      "learning_rate": 4.322911157650433e-07,
+      "loss": 0.0015,
+      "step": 6198
+    },
+    {
+      "epoch": 3.771828414968056,
+      "grad_norm": 0.24557361006736755,
+      "learning_rate": 4.2998912993430785e-07,
+      "loss": 0.0054,
+      "step": 6199
+    },
+    {
+      "epoch": 3.772436872528141,
+      "grad_norm": 0.3791668117046356,
+      "learning_rate": 4.2769323644213375e-07,
+      "loss": 0.0127,
+      "step": 6200
+    },
+    {
+      "epoch": 3.7730453300882263,
+      "grad_norm": 0.19211210310459137,
+      "learning_rate": 4.2540343585781573e-07,
+      "loss": 0.0034,
+      "step": 6201
+    },
+    {
+      "epoch": 3.7736537876483114,
+      "grad_norm": 0.23425185680389404,
+      "learning_rate": 4.2311972874912453e-07,
+      "loss": 0.0062,
+      "step": 6202
+    },
+    {
+      "epoch": 3.774262245208397,
+      "grad_norm": 0.2918088138103485,
+      "learning_rate": 4.208421156823239e-07,
+      "loss": 0.005,
+      "step": 6203
+    },
+    {
+      "epoch": 3.774870702768482,
+      "grad_norm": 0.17967435717582703,
+      "learning_rate": 4.1857059722217316e-07,
+      "loss": 0.0033,
+      "step": 6204
+    },
+    {
+      "epoch": 3.775479160328567,
+      "grad_norm": 0.2613130211830139,
+      "learning_rate": 4.1630517393190794e-07,
+      "loss": 0.0055,
+      "step": 6205
+    },
+    {
+      "epoch": 3.776087617888652,
+      "grad_norm": 0.3404596149921417,
+      "learning_rate": 4.1404584637325936e-07,
+      "loss": 0.0088,
+      "step": 6206
+    },
+    {
+      "epoch": 3.7766960754487373,
+      "grad_norm": 0.286954402923584,
+      "learning_rate": 4.117926151064488e-07,
+      "loss": 0.0082,
+      "step": 6207
+    },
+    {
+      "epoch": 3.777304533008823,
+      "grad_norm": 0.19479084014892578,
+      "learning_rate": 4.0954548069018217e-07,
+      "loss": 0.0041,
+      "step": 6208
+    },
+    {
+      "epoch": 3.777912990568908,
+      "grad_norm": 0.20791840553283691,
+      "learning_rate": 4.073044436816581e-07,
+      "loss": 0.0048,
+      "step": 6209
+    },
+    {
+      "epoch": 3.778521448128993,
+      "grad_norm": 0.2315748929977417,
+      "learning_rate": 4.0506950463655713e-07,
+      "loss": 0.0048,
+      "step": 6210
+    },
+    {
+      "epoch": 3.779129905689078,
+      "grad_norm": 0.2659534811973572,
+      "learning_rate": 4.028406641090499e-07,
+      "loss": 0.004,
+      "step": 6211
+    },
+    {
+      "epoch": 3.7797383632491632,
+      "grad_norm": 0.2190115749835968,
+      "learning_rate": 4.0061792265179696e-07,
+      "loss": 0.0029,
+      "step": 6212
+    },
+    {
+      "epoch": 3.7803468208092488,
+      "grad_norm": 0.20221631228923798,
+      "learning_rate": 3.984012808159493e-07,
+      "loss": 0.0021,
+      "step": 6213
+    },
+    {
+      "epoch": 3.780955278369334,
+      "grad_norm": 0.2616676390171051,
+      "learning_rate": 3.961907391511366e-07,
+      "loss": 0.0075,
+      "step": 6214
+    },
+    {
+      "epoch": 3.781563735929419,
+      "grad_norm": 0.27306365966796875,
+      "learning_rate": 3.9398629820548703e-07,
+      "loss": 0.0063,
+      "step": 6215
+    },
+    {
+      "epoch": 3.782172193489504,
+      "grad_norm": 0.18841589987277985,
+      "learning_rate": 3.9178795852560236e-07,
+      "loss": 0.0023,
+      "step": 6216
+    },
+    {
+      "epoch": 3.782780651049589,
+      "grad_norm": 0.27202504873275757,
+      "learning_rate": 3.8959572065658535e-07,
+      "loss": 0.0057,
+      "step": 6217
+    },
+    {
+      "epoch": 3.7833891086096747,
+      "grad_norm": 0.24065257608890533,
+      "learning_rate": 3.874095851420151e-07,
+      "loss": 0.0048,
+      "step": 6218
+    },
+    {
+      "epoch": 3.78399756616976,
+      "grad_norm": 0.31770059466362,
+      "learning_rate": 3.852295525239663e-07,
+      "loss": 0.0073,
+      "step": 6219
+    },
+    {
+      "epoch": 3.784606023729845,
+      "grad_norm": 0.20511983335018158,
+      "learning_rate": 3.830556233429927e-07,
+      "loss": 0.0039,
+      "step": 6220
+    },
+    {
+      "epoch": 3.78521448128993,
+      "grad_norm": 0.33700552582740784,
+      "learning_rate": 3.808877981381437e-07,
+      "loss": 0.0049,
+      "step": 6221
+    },
+    {
+      "epoch": 3.785822938850015,
+      "grad_norm": 0.2216031849384308,
+      "learning_rate": 3.78726077446942e-07,
+      "loss": 0.0034,
+      "step": 6222
+    },
+    {
+      "epoch": 3.7864313964101006,
+      "grad_norm": 0.1690044105052948,
+      "learning_rate": 3.7657046180540887e-07,
+      "loss": 0.0042,
+      "step": 6223
+    },
+    {
+      "epoch": 3.7870398539701857,
+      "grad_norm": 0.3780827224254608,
+      "learning_rate": 3.744209517480446e-07,
+      "loss": 0.0068,
+      "step": 6224
+    },
+    {
+      "epoch": 3.787648311530271,
+      "grad_norm": 0.2709617018699646,
+      "learning_rate": 3.722775478078422e-07,
+      "loss": 0.0064,
+      "step": 6225
+    },
+    {
+      "epoch": 3.788256769090356,
+      "grad_norm": 0.24110420048236847,
+      "learning_rate": 3.70140250516271e-07,
+      "loss": 0.0058,
+      "step": 6226
+    },
+    {
+      "epoch": 3.788865226650441,
+      "grad_norm": 0.31499508023262024,
+      "learning_rate": 3.6800906040329595e-07,
+      "loss": 0.0075,
+      "step": 6227
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.2634466588497162,
+      "learning_rate": 3.658839779973611e-07,
+      "loss": 0.0054,
+      "step": 6228
+    },
+    {
+      "epoch": 3.7900821417706116,
+      "grad_norm": 0.20416542887687683,
+      "learning_rate": 3.637650038254004e-07,
+      "loss": 0.0045,
+      "step": 6229
+    },
+    {
+      "epoch": 3.7906905993306967,
+      "grad_norm": 0.3133479356765747,
+      "learning_rate": 3.6165213841282966e-07,
+      "loss": 0.0248,
+      "step": 6230
+    },
+    {
+      "epoch": 3.791299056890782,
+      "grad_norm": 0.16795231401920319,
+      "learning_rate": 3.5954538228355205e-07,
+      "loss": 0.0035,
+      "step": 6231
+    },
+    {
+      "epoch": 3.791907514450867,
+      "grad_norm": 0.26573193073272705,
+      "learning_rate": 3.5744473595995533e-07,
+      "loss": 0.0077,
+      "step": 6232
+    },
+    {
+      "epoch": 3.7925159720109525,
+      "grad_norm": 0.3238775432109833,
+      "learning_rate": 3.5535019996290885e-07,
+      "loss": 0.0108,
+      "step": 6233
+    },
+    {
+      "epoch": 3.7931244295710376,
+      "grad_norm": 0.1755254566669464,
+      "learning_rate": 3.5326177481177505e-07,
+      "loss": 0.0038,
+      "step": 6234
+    },
+    {
+      "epoch": 3.7937328871311227,
+      "grad_norm": 0.2531315088272095,
+      "learning_rate": 3.5117946102439513e-07,
+      "loss": 0.0048,
+      "step": 6235
+    },
+    {
+      "epoch": 3.7943413446912078,
+      "grad_norm": 0.28363898396492004,
+      "learning_rate": 3.491032591170951e-07,
+      "loss": 0.0049,
+      "step": 6236
+    },
+    {
+      "epoch": 3.794949802251293,
+      "grad_norm": 0.24325834214687347,
+      "learning_rate": 3.4703316960468524e-07,
+      "loss": 0.0048,
+      "step": 6237
+    },
+    {
+      "epoch": 3.7955582598113784,
+      "grad_norm": 0.22880126535892487,
+      "learning_rate": 3.4496919300046617e-07,
+      "loss": 0.0046,
+      "step": 6238
+    },
+    {
+      "epoch": 3.7961667173714635,
+      "grad_norm": 0.30837124586105347,
+      "learning_rate": 3.4291132981621174e-07,
+      "loss": 0.002,
+      "step": 6239
+    },
+    {
+      "epoch": 3.7967751749315486,
+      "grad_norm": 0.16037946939468384,
+      "learning_rate": 3.408595805621889e-07,
+      "loss": 0.0026,
+      "step": 6240
+    },
+    {
+      "epoch": 3.7973836324916337,
+      "grad_norm": 0.1504233032464981,
+      "learning_rate": 3.3881394574715174e-07,
+      "loss": 0.0035,
+      "step": 6241
+    },
+    {
+      "epoch": 3.797992090051719,
+      "grad_norm": 0.25224924087524414,
+      "learning_rate": 3.367744258783223e-07,
+      "loss": 0.006,
+      "step": 6242
+    },
+    {
+      "epoch": 3.7986005476118043,
+      "grad_norm": 0.25352349877357483,
+      "learning_rate": 3.347410214614211e-07,
+      "loss": 0.0061,
+      "step": 6243
+    },
+    {
+      "epoch": 3.7992090051718894,
+      "grad_norm": 0.27149248123168945,
+      "learning_rate": 3.327137330006502e-07,
+      "loss": 0.0158,
+      "step": 6244
+    },
+    {
+      "epoch": 3.7998174627319745,
+      "grad_norm": 0.23779939115047455,
+      "learning_rate": 3.3069256099869105e-07,
+      "loss": 0.0074,
+      "step": 6245
+    },
+    {
+      "epoch": 3.8004259202920596,
+      "grad_norm": 0.16755706071853638,
+      "learning_rate": 3.2867750595670657e-07,
+      "loss": 0.0032,
+      "step": 6246
+    },
+    {
+      "epoch": 3.8010343778521447,
+      "grad_norm": 0.19651807844638824,
+      "learning_rate": 3.266685683743498e-07,
+      "loss": 0.0047,
+      "step": 6247
+    },
+    {
+      "epoch": 3.8016428354122302,
+      "grad_norm": 0.277131587266922,
+      "learning_rate": 3.2466574874975565e-07,
+      "loss": 0.0041,
+      "step": 6248
+    },
+    {
+      "epoch": 3.8022512929723153,
+      "grad_norm": 0.27917802333831787,
+      "learning_rate": 3.226690475795324e-07,
+      "loss": 0.0071,
+      "step": 6249
+    },
+    {
+      "epoch": 3.8028597505324004,
+      "grad_norm": 0.2759116291999817,
+      "learning_rate": 3.206784653587869e-07,
+      "loss": 0.0063,
+      "step": 6250
+    },
+    {
+      "epoch": 3.8034682080924855,
+      "grad_norm": 0.37202462553977966,
+      "learning_rate": 3.186940025810992e-07,
+      "loss": 0.0067,
+      "step": 6251
+    },
+    {
+      "epoch": 3.8040766656525706,
+      "grad_norm": 0.1423821896314621,
+      "learning_rate": 3.1671565973852567e-07,
+      "loss": 0.0034,
+      "step": 6252
+    },
+    {
+      "epoch": 3.804685123212656,
+      "grad_norm": 0.6763217449188232,
+      "learning_rate": 3.14743437321624e-07,
+      "loss": 0.0252,
+      "step": 6253
+    },
+    {
+      "epoch": 3.805293580772741,
+      "grad_norm": 0.34884288907051086,
+      "learning_rate": 3.1277733581941416e-07,
+      "loss": 0.0078,
+      "step": 6254
+    },
+    {
+      "epoch": 3.8059020383328264,
+      "grad_norm": 0.24357537925243378,
+      "learning_rate": 3.1081735571941437e-07,
+      "loss": 0.0061,
+      "step": 6255
+    },
+    {
+      "epoch": 3.8065104958929115,
+      "grad_norm": 0.27209481596946716,
+      "learning_rate": 3.088634975076082e-07,
+      "loss": 0.0054,
+      "step": 6256
+    },
+    {
+      "epoch": 3.8071189534529966,
+      "grad_norm": 0.24121397733688354,
+      "learning_rate": 3.0691576166848314e-07,
+      "loss": 0.0054,
+      "step": 6257
+    },
+    {
+      "epoch": 3.807727411013082,
+      "grad_norm": 0.24442483484745026,
+      "learning_rate": 3.0497414868498884e-07,
+      "loss": 0.0058,
+      "step": 6258
+    },
+    {
+      "epoch": 3.8083358685731667,
+      "grad_norm": 0.24496260285377502,
+      "learning_rate": 3.030386590385653e-07,
+      "loss": 0.0083,
+      "step": 6259
+    },
+    {
+      "epoch": 3.8089443261332523,
+      "grad_norm": 0.14519323408603668,
+      "learning_rate": 3.0110929320913694e-07,
+      "loss": 0.0029,
+      "step": 6260
+    },
+    {
+      "epoch": 3.8095527836933374,
+      "grad_norm": 0.32007092237472534,
+      "learning_rate": 2.991860516751016e-07,
+      "loss": 0.0097,
+      "step": 6261
+    },
+    {
+      "epoch": 3.8101612412534225,
+      "grad_norm": 0.3110872209072113,
+      "learning_rate": 2.972689349133417e-07,
+      "loss": 0.0085,
+      "step": 6262
+    },
+    {
+      "epoch": 3.810769698813508,
+      "grad_norm": 0.27134284377098083,
+      "learning_rate": 2.9535794339922984e-07,
+      "loss": 0.0026,
+      "step": 6263
+    },
+    {
+      "epoch": 3.8113781563735927,
+      "grad_norm": 0.2500753700733185,
+      "learning_rate": 2.9345307760660637e-07,
+      "loss": 0.0059,
+      "step": 6264
+    },
+    {
+      "epoch": 3.811986613933678,
+      "grad_norm": 0.3057078421115875,
+      "learning_rate": 2.9155433800780176e-07,
+      "loss": 0.0049,
+      "step": 6265
+    },
+    {
+      "epoch": 3.8125950714937633,
+      "grad_norm": 0.2372872531414032,
+      "learning_rate": 2.8966172507362e-07,
+      "loss": 0.0049,
+      "step": 6266
+    },
+    {
+      "epoch": 3.8132035290538484,
+      "grad_norm": 0.2749864161014557,
+      "learning_rate": 2.8777523927335515e-07,
+      "loss": 0.0056,
+      "step": 6267
+    },
+    {
+      "epoch": 3.813811986613934,
+      "grad_norm": 0.26591286063194275,
+      "learning_rate": 2.8589488107477194e-07,
+      "loss": 0.0046,
+      "step": 6268
+    },
+    {
+      "epoch": 3.8144204441740186,
+      "grad_norm": 0.21569810807704926,
+      "learning_rate": 2.840206509441251e-07,
+      "loss": 0.0049,
+      "step": 6269
+    },
+    {
+      "epoch": 3.815028901734104,
+      "grad_norm": 0.19832342863082886,
+      "learning_rate": 2.82152549346143e-07,
+      "loss": 0.0056,
+      "step": 6270
+    },
+    {
+      "epoch": 3.8156373592941892,
+      "grad_norm": 0.3168319761753082,
+      "learning_rate": 2.8029057674404115e-07,
+      "loss": 0.0104,
+      "step": 6271
+    },
+    {
+      "epoch": 3.8162458168542743,
+      "grad_norm": 0.1959853619337082,
+      "learning_rate": 2.7843473359950303e-07,
+      "loss": 0.0045,
+      "step": 6272
+    },
+    {
+      "epoch": 3.81685427441436,
+      "grad_norm": 0.3243808150291443,
+      "learning_rate": 2.7658502037270774e-07,
+      "loss": 0.0067,
+      "step": 6273
+    },
+    {
+      "epoch": 3.8174627319744445,
+      "grad_norm": 0.2744379937648773,
+      "learning_rate": 2.747414375223051e-07,
+      "loss": 0.0057,
+      "step": 6274
+    },
+    {
+      "epoch": 3.81807118953453,
+      "grad_norm": 0.26081326603889465,
+      "learning_rate": 2.7290398550542664e-07,
+      "loss": 0.0047,
+      "step": 6275
+    },
+    {
+      "epoch": 3.818679647094615,
+      "grad_norm": 0.24908572435379028,
+      "learning_rate": 2.710726647776829e-07,
+      "loss": 0.0062,
+      "step": 6276
+    },
+    {
+      "epoch": 3.8192881046547003,
+      "grad_norm": 0.23935918509960175,
+      "learning_rate": 2.692474757931662e-07,
+      "loss": 0.0075,
+      "step": 6277
+    },
+    {
+      "epoch": 3.819896562214786,
+      "grad_norm": 0.1932537704706192,
+      "learning_rate": 2.674284190044479e-07,
+      "loss": 0.0034,
+      "step": 6278
+    },
+    {
+      "epoch": 3.8205050197748704,
+      "grad_norm": 0.3207075893878937,
+      "learning_rate": 2.6561549486257556e-07,
+      "loss": 0.0078,
+      "step": 6279
+    },
+    {
+      "epoch": 3.821113477334956,
+      "grad_norm": 0.2624523639678955,
+      "learning_rate": 2.638087038170811e-07,
+      "loss": 0.0082,
+      "step": 6280
+    },
+    {
+      "epoch": 3.821721934895041,
+      "grad_norm": 0.16647395491600037,
+      "learning_rate": 2.62008046315973e-07,
+      "loss": 0.003,
+      "step": 6281
+    },
+    {
+      "epoch": 3.822330392455126,
+      "grad_norm": 0.2986607253551483,
+      "learning_rate": 2.602135228057384e-07,
+      "loss": 0.0101,
+      "step": 6282
+    },
+    {
+      "epoch": 3.8229388500152113,
+      "grad_norm": 0.35330304503440857,
+      "learning_rate": 2.5842513373134645e-07,
+      "loss": 0.0103,
+      "step": 6283
+    },
+    {
+      "epoch": 3.8235473075752964,
+      "grad_norm": 0.2447318136692047,
+      "learning_rate": 2.566428795362397e-07,
+      "loss": 0.0059,
+      "step": 6284
+    },
+    {
+      "epoch": 3.824155765135382,
+      "grad_norm": 0.16872382164001465,
+      "learning_rate": 2.5486676066234504e-07,
+      "loss": 0.0039,
+      "step": 6285
+    },
+    {
+      "epoch": 3.824764222695467,
+      "grad_norm": 0.2803313434123993,
+      "learning_rate": 2.5309677755006867e-07,
+      "loss": 0.0081,
+      "step": 6286
+    },
+    {
+      "epoch": 3.825372680255552,
+      "grad_norm": 0.36526891589164734,
+      "learning_rate": 2.5133293063828724e-07,
+      "loss": 0.0137,
+      "step": 6287
+    },
+    {
+      "epoch": 3.825981137815637,
+      "grad_norm": 0.2731226086616516,
+      "learning_rate": 2.4957522036436474e-07,
+      "loss": 0.0092,
+      "step": 6288
+    },
+    {
+      "epoch": 3.8265895953757223,
+      "grad_norm": 0.19423212110996246,
+      "learning_rate": 2.4782364716413873e-07,
+      "loss": 0.0049,
+      "step": 6289
+    },
+    {
+      "epoch": 3.827198052935808,
+      "grad_norm": 0.16490328311920166,
+      "learning_rate": 2.460782114719257e-07,
+      "loss": 0.0032,
+      "step": 6290
+    },
+    {
+      "epoch": 3.827806510495893,
+      "grad_norm": 0.30284324288368225,
+      "learning_rate": 2.4433891372052376e-07,
+      "loss": 0.0055,
+      "step": 6291
+    },
+    {
+      "epoch": 3.828414968055978,
+      "grad_norm": 0.18749664723873138,
+      "learning_rate": 2.426057543412019e-07,
+      "loss": 0.0038,
+      "step": 6292
+    },
+    {
+      "epoch": 3.829023425616063,
+      "grad_norm": 0.2562442719936371,
+      "learning_rate": 2.408787337637164e-07,
+      "loss": 0.0059,
+      "step": 6293
+    },
+    {
+      "epoch": 3.829631883176148,
+      "grad_norm": 0.1709728091955185,
+      "learning_rate": 2.3915785241629406e-07,
+      "loss": 0.0035,
+      "step": 6294
+    },
+    {
+      "epoch": 3.8302403407362338,
+      "grad_norm": 0.3805026113986969,
+      "learning_rate": 2.3744311072563808e-07,
+      "loss": 0.0078,
+      "step": 6295
+    },
+    {
+      "epoch": 3.830848798296319,
+      "grad_norm": 0.20816834270954132,
+      "learning_rate": 2.3573450911693883e-07,
+      "loss": 0.0035,
+      "step": 6296
+    },
+    {
+      "epoch": 3.831457255856404,
+      "grad_norm": 0.16338828206062317,
+      "learning_rate": 2.3403204801385746e-07,
+      "loss": 0.0023,
+      "step": 6297
+    },
+    {
+      "epoch": 3.832065713416489,
+      "grad_norm": 0.21396136283874512,
+      "learning_rate": 2.3233572783852854e-07,
+      "loss": 0.0054,
+      "step": 6298
+    },
+    {
+      "epoch": 3.832674170976574,
+      "grad_norm": 0.20282062888145447,
+      "learning_rate": 2.3064554901157388e-07,
+      "loss": 0.0051,
+      "step": 6299
+    },
+    {
+      "epoch": 3.8332826285366597,
+      "grad_norm": 0.2224971354007721,
+      "learning_rate": 2.2896151195208603e-07,
+      "loss": 0.004,
+      "step": 6300
+    },
+    {
+      "epoch": 3.833891086096745,
+      "grad_norm": 0.3189062476158142,
+      "learning_rate": 2.272836170776338e-07,
+      "loss": 0.0039,
+      "step": 6301
+    },
+    {
+      "epoch": 3.83449954365683,
+      "grad_norm": 0.14660929143428802,
+      "learning_rate": 2.2561186480426766e-07,
+      "loss": 0.0035,
+      "step": 6302
+    },
+    {
+      "epoch": 3.835108001216915,
+      "grad_norm": 0.30386772751808167,
+      "learning_rate": 2.239462555465116e-07,
+      "loss": 0.0065,
+      "step": 6303
+    },
+    {
+      "epoch": 3.835716458777,
+      "grad_norm": 0.18635998666286469,
+      "learning_rate": 2.222867897173686e-07,
+      "loss": 0.0032,
+      "step": 6304
+    },
+    {
+      "epoch": 3.8363249163370856,
+      "grad_norm": 0.12038879841566086,
+      "learning_rate": 2.206334677283123e-07,
+      "loss": 0.0022,
+      "step": 6305
+    },
+    {
+      "epoch": 3.8369333738971707,
+      "grad_norm": 0.14553184807300568,
+      "learning_rate": 2.189862899893036e-07,
+      "loss": 0.0035,
+      "step": 6306
+    },
+    {
+      "epoch": 3.837541831457256,
+      "grad_norm": 0.21486513316631317,
+      "learning_rate": 2.173452569087714e-07,
+      "loss": 0.0051,
+      "step": 6307
+    },
+    {
+      "epoch": 3.838150289017341,
+      "grad_norm": 0.25689902901649475,
+      "learning_rate": 2.157103688936235e-07,
+      "loss": 0.0056,
+      "step": 6308
+    },
+    {
+      "epoch": 3.838758746577426,
+      "grad_norm": 0.22431597113609314,
+      "learning_rate": 2.14081626349244e-07,
+      "loss": 0.0025,
+      "step": 6309
+    },
+    {
+      "epoch": 3.8393672041375115,
+      "grad_norm": 0.23348523676395416,
+      "learning_rate": 2.1245902967949315e-07,
+      "loss": 0.0043,
+      "step": 6310
+    },
+    {
+      "epoch": 3.8399756616975966,
+      "grad_norm": 0.21763131022453308,
+      "learning_rate": 2.1084257928670748e-07,
+      "loss": 0.0034,
+      "step": 6311
+    },
+    {
+      "epoch": 3.8405841192576817,
+      "grad_norm": 0.2803143560886383,
+      "learning_rate": 2.092322755716969e-07,
+      "loss": 0.015,
+      "step": 6312
+    },
+    {
+      "epoch": 3.841192576817767,
+      "grad_norm": 0.1908082216978073,
+      "learning_rate": 2.0762811893375588e-07,
+      "loss": 0.0037,
+      "step": 6313
+    },
+    {
+      "epoch": 3.841801034377852,
+      "grad_norm": 0.24294763803482056,
+      "learning_rate": 2.060301097706413e-07,
+      "loss": 0.0044,
+      "step": 6314
+    },
+    {
+      "epoch": 3.8424094919379375,
+      "grad_norm": 0.2627767026424408,
+      "learning_rate": 2.0443824847859727e-07,
+      "loss": 0.0064,
+      "step": 6315
+    },
+    {
+      "epoch": 3.8430179494980226,
+      "grad_norm": 0.24739083647727966,
+      "learning_rate": 2.0285253545233585e-07,
+      "loss": 0.0071,
+      "step": 6316
+    },
+    {
+      "epoch": 3.8436264070581077,
+      "grad_norm": 0.31518417596817017,
+      "learning_rate": 2.0127297108505082e-07,
+      "loss": 0.0102,
+      "step": 6317
+    },
+    {
+      "epoch": 3.8442348646181927,
+      "grad_norm": 0.40057793259620667,
+      "learning_rate": 1.996995557684067e-07,
+      "loss": 0.006,
+      "step": 6318
+    },
+    {
+      "epoch": 3.844843322178278,
+      "grad_norm": 0.1834147572517395,
+      "learning_rate": 1.9813228989254695e-07,
+      "loss": 0.0039,
+      "step": 6319
+    },
+    {
+      "epoch": 3.8454517797383634,
+      "grad_norm": 0.2752745449542999,
+      "learning_rate": 1.9657117384608569e-07,
+      "loss": 0.0065,
+      "step": 6320
+    },
+    {
+      "epoch": 3.8460602372984485,
+      "grad_norm": 0.3888533413410187,
+      "learning_rate": 1.950162080161161e-07,
+      "loss": 0.0083,
+      "step": 6321
+    },
+    {
+      "epoch": 3.8466686948585336,
+      "grad_norm": 0.3024943172931671,
+      "learning_rate": 1.9346739278820192e-07,
+      "loss": 0.0117,
+      "step": 6322
+    },
+    {
+      "epoch": 3.8472771524186187,
+      "grad_norm": 0.2485586553812027,
+      "learning_rate": 1.9192472854638875e-07,
+      "loss": 0.0073,
+      "step": 6323
+    },
+    {
+      "epoch": 3.8478856099787038,
+      "grad_norm": 0.19068674743175507,
+      "learning_rate": 1.9038821567319286e-07,
+      "loss": 0.0028,
+      "step": 6324
+    },
+    {
+      "epoch": 3.8484940675387893,
+      "grad_norm": 0.18191272020339966,
+      "learning_rate": 1.8885785454960115e-07,
+      "loss": 0.0039,
+      "step": 6325
+    },
+    {
+      "epoch": 3.8491025250988744,
+      "grad_norm": 0.28522956371307373,
+      "learning_rate": 1.8733364555508225e-07,
+      "loss": 0.0054,
+      "step": 6326
+    },
+    {
+      "epoch": 3.8497109826589595,
+      "grad_norm": 0.24018339812755585,
+      "learning_rate": 1.8581558906757557e-07,
+      "loss": 0.0064,
+      "step": 6327
+    },
+    {
+      "epoch": 3.8503194402190446,
+      "grad_norm": 0.3272731304168701,
+      "learning_rate": 1.8430368546349942e-07,
+      "loss": 0.0079,
+      "step": 6328
+    },
+    {
+      "epoch": 3.8509278977791297,
+      "grad_norm": 0.30089452862739563,
+      "learning_rate": 1.8279793511774e-07,
+      "loss": 0.0038,
+      "step": 6329
+    },
+    {
+      "epoch": 3.8515363553392152,
+      "grad_norm": 0.2161916345357895,
+      "learning_rate": 1.8129833840365985e-07,
+      "loss": 0.0054,
+      "step": 6330
+    },
+    {
+      "epoch": 3.8521448128993003,
+      "grad_norm": 0.2529414892196655,
+      "learning_rate": 1.7980489569309755e-07,
+      "loss": 0.005,
+      "step": 6331
+    },
+    {
+      "epoch": 3.8527532704593854,
+      "grad_norm": 0.3203285336494446,
+      "learning_rate": 1.7831760735636248e-07,
+      "loss": 0.0057,
+      "step": 6332
+    },
+    {
+      "epoch": 3.8533617280194705,
+      "grad_norm": 0.20392923057079315,
+      "learning_rate": 1.76836473762243e-07,
+      "loss": 0.004,
+      "step": 6333
+    },
+    {
+      "epoch": 3.8539701855795556,
+      "grad_norm": 0.3112758994102478,
+      "learning_rate": 1.7536149527800082e-07,
+      "loss": 0.0069,
+      "step": 6334
+    },
+    {
+      "epoch": 3.854578643139641,
+      "grad_norm": 0.2692670226097107,
+      "learning_rate": 1.738926722693629e-07,
+      "loss": 0.0065,
+      "step": 6335
+    },
+    {
+      "epoch": 3.8551871006997263,
+      "grad_norm": 0.1982191503047943,
+      "learning_rate": 1.7243000510053787e-07,
+      "loss": 0.0045,
+      "step": 6336
+    },
+    {
+      "epoch": 3.8557955582598114,
+      "grad_norm": 0.21474218368530273,
+      "learning_rate": 1.7097349413420781e-07,
+      "loss": 0.0056,
+      "step": 6337
+    },
+    {
+      "epoch": 3.8564040158198964,
+      "grad_norm": 0.16877877712249756,
+      "learning_rate": 1.6952313973152834e-07,
+      "loss": 0.0047,
+      "step": 6338
+    },
+    {
+      "epoch": 3.8570124733799815,
+      "grad_norm": 0.35990428924560547,
+      "learning_rate": 1.6807894225212283e-07,
+      "loss": 0.0087,
+      "step": 6339
+    },
+    {
+      "epoch": 3.857620930940067,
+      "grad_norm": 0.3100360333919525,
+      "learning_rate": 1.6664090205409656e-07,
+      "loss": 0.0051,
+      "step": 6340
+    },
+    {
+      "epoch": 3.858229388500152,
+      "grad_norm": 0.1680872142314911,
+      "learning_rate": 1.652090194940198e-07,
+      "loss": 0.0032,
+      "step": 6341
+    },
+    {
+      "epoch": 3.8588378460602373,
+      "grad_norm": 0.22448216378688812,
+      "learning_rate": 1.637832949269419e-07,
+      "loss": 0.0045,
+      "step": 6342
+    },
+    {
+      "epoch": 3.8594463036203224,
+      "grad_norm": 0.20747099816799164,
+      "learning_rate": 1.6236372870638284e-07,
+      "loss": 0.0045,
+      "step": 6343
+    },
+    {
+      "epoch": 3.8600547611804075,
+      "grad_norm": 0.23858143389225006,
+      "learning_rate": 1.609503211843333e-07,
+      "loss": 0.0066,
+      "step": 6344
+    },
+    {
+      "epoch": 3.860663218740493,
+      "grad_norm": 0.2025892287492752,
+      "learning_rate": 1.595430727112629e-07,
+      "loss": 0.0042,
+      "step": 6345
+    },
+    {
+      "epoch": 3.861271676300578,
+      "grad_norm": 0.3395010530948639,
+      "learning_rate": 1.581419836361092e-07,
+      "loss": 0.0097,
+      "step": 6346
+    },
+    {
+      "epoch": 3.861880133860663,
+      "grad_norm": 0.23754116892814636,
+      "learning_rate": 1.5674705430628323e-07,
+      "loss": 0.0032,
+      "step": 6347
+    },
+    {
+      "epoch": 3.8624885914207483,
+      "grad_norm": 0.21957284212112427,
+      "learning_rate": 1.5535828506766936e-07,
+      "loss": 0.0063,
+      "step": 6348
+    },
+    {
+      "epoch": 3.8630970489808334,
+      "grad_norm": 0.18481026589870453,
+      "learning_rate": 1.539756762646255e-07,
+      "loss": 0.0039,
+      "step": 6349
+    },
+    {
+      "epoch": 3.863705506540919,
+      "grad_norm": 0.3242833912372589,
+      "learning_rate": 1.5259922823998297e-07,
+      "loss": 0.0067,
+      "step": 6350
+    },
+    {
+      "epoch": 3.864313964101004,
+      "grad_norm": 0.24175108969211578,
+      "learning_rate": 1.5122894133503817e-07,
+      "loss": 0.0048,
+      "step": 6351
+    },
+    {
+      "epoch": 3.864922421661089,
+      "grad_norm": 0.27556535601615906,
+      "learning_rate": 1.4986481588956934e-07,
+      "loss": 0.0043,
+      "step": 6352
+    },
+    {
+      "epoch": 3.865530879221174,
+      "grad_norm": 0.27273163199424744,
+      "learning_rate": 1.485068522418226e-07,
+      "loss": 0.0042,
+      "step": 6353
+    },
+    {
+      "epoch": 3.8661393367812593,
+      "grad_norm": 0.277964323759079,
+      "learning_rate": 1.4715505072851188e-07,
+      "loss": 0.015,
+      "step": 6354
+    },
+    {
+      "epoch": 3.866747794341345,
+      "grad_norm": 0.3170606791973114,
+      "learning_rate": 1.4580941168483298e-07,
+      "loss": 0.0103,
+      "step": 6355
+    },
+    {
+      "epoch": 3.86735625190143,
+      "grad_norm": 0.23001722991466522,
+      "learning_rate": 1.4446993544444954e-07,
+      "loss": 0.0063,
+      "step": 6356
+    },
+    {
+      "epoch": 3.867964709461515,
+      "grad_norm": 0.240625262260437,
+      "learning_rate": 1.4313662233948755e-07,
+      "loss": 0.0072,
+      "step": 6357
+    },
+    {
+      "epoch": 3.8685731670216,
+      "grad_norm": 0.22342710196971893,
+      "learning_rate": 1.4180947270056032e-07,
+      "loss": 0.0043,
+      "step": 6358
+    },
+    {
+      "epoch": 3.8691816245816852,
+      "grad_norm": 0.2697957754135132,
+      "learning_rate": 1.4048848685674354e-07,
+      "loss": 0.0061,
+      "step": 6359
+    },
+    {
+      "epoch": 3.869790082141771,
+      "grad_norm": 0.21805325150489807,
+      "learning_rate": 1.3917366513558629e-07,
+      "loss": 0.004,
+      "step": 6360
+    },
+    {
+      "epoch": 3.870398539701856,
+      "grad_norm": 0.23698726296424866,
+      "learning_rate": 1.378650078631083e-07,
+      "loss": 0.0069,
+      "step": 6361
+    },
+    {
+      "epoch": 3.871006997261941,
+      "grad_norm": 0.2618488669395447,
+      "learning_rate": 1.365625153638056e-07,
+      "loss": 0.0074,
+      "step": 6362
+    },
+    {
+      "epoch": 3.871615454822026,
+      "grad_norm": 0.2968907654285431,
+      "learning_rate": 1.352661879606393e-07,
+      "loss": 0.0097,
+      "step": 6363
+    },
+    {
+      "epoch": 3.872223912382111,
+      "grad_norm": 0.31761443614959717,
+      "learning_rate": 1.339760259750439e-07,
+      "loss": 0.0071,
+      "step": 6364
+    },
+    {
+      "epoch": 3.8728323699421967,
+      "grad_norm": 0.18307559192180634,
+      "learning_rate": 1.3269202972692741e-07,
+      "loss": 0.0043,
+      "step": 6365
+    },
+    {
+      "epoch": 3.873440827502282,
+      "grad_norm": 0.2085932344198227,
+      "learning_rate": 1.314141995346685e-07,
+      "loss": 0.0047,
+      "step": 6366
+    },
+    {
+      "epoch": 3.874049285062367,
+      "grad_norm": 0.21864819526672363,
+      "learning_rate": 1.301425357151137e-07,
+      "loss": 0.0072,
+      "step": 6367
+    },
+    {
+      "epoch": 3.874657742622452,
+      "grad_norm": 0.15939606726169586,
+      "learning_rate": 1.2887703858358302e-07,
+      "loss": 0.0028,
+      "step": 6368
+    },
+    {
+      "epoch": 3.875266200182537,
+      "grad_norm": 0.21072454750537872,
+      "learning_rate": 1.2761770845386712e-07,
+      "loss": 0.0034,
+      "step": 6369
+    },
+    {
+      "epoch": 3.8758746577426226,
+      "grad_norm": 0.3073256313800812,
+      "learning_rate": 1.2636454563823009e-07,
+      "loss": 0.0075,
+      "step": 6370
+    },
+    {
+      "epoch": 3.8764831153027077,
+      "grad_norm": 0.2551986873149872,
+      "learning_rate": 1.2511755044739836e-07,
+      "loss": 0.0072,
+      "step": 6371
+    },
+    {
+      "epoch": 3.877091572862793,
+      "grad_norm": 0.2689383327960968,
+      "learning_rate": 1.238767231905774e-07,
+      "loss": 0.0041,
+      "step": 6372
+    },
+    {
+      "epoch": 3.877700030422878,
+      "grad_norm": 0.2862478792667389,
+      "learning_rate": 1.2264206417544333e-07,
+      "loss": 0.0051,
+      "step": 6373
+    },
+    {
+      "epoch": 3.878308487982963,
+      "grad_norm": 0.34870097041130066,
+      "learning_rate": 1.2141357370813732e-07,
+      "loss": 0.0063,
+      "step": 6374
+    },
+    {
+      "epoch": 3.8789169455430486,
+      "grad_norm": 0.27742713689804077,
+      "learning_rate": 1.2019125209327409e-07,
+      "loss": 0.0047,
+      "step": 6375
+    },
+    {
+      "epoch": 3.8795254031031337,
+      "grad_norm": 0.19228945672512054,
+      "learning_rate": 1.1897509963394171e-07,
+      "loss": 0.0043,
+      "step": 6376
+    },
+    {
+      "epoch": 3.8801338606632187,
+      "grad_norm": 0.2762080430984497,
+      "learning_rate": 1.1776511663168788e-07,
+      "loss": 0.0081,
+      "step": 6377
+    },
+    {
+      "epoch": 3.880742318223304,
+      "grad_norm": 0.3993409276008606,
+      "learning_rate": 1.165613033865448e-07,
+      "loss": 0.0087,
+      "step": 6378
+    },
+    {
+      "epoch": 3.881350775783389,
+      "grad_norm": 0.26734060049057007,
+      "learning_rate": 1.1536366019700428e-07,
+      "loss": 0.0054,
+      "step": 6379
+    },
+    {
+      "epoch": 3.8819592333434745,
+      "grad_norm": 0.2520960867404938,
+      "learning_rate": 1.1417218736003432e-07,
+      "loss": 0.0048,
+      "step": 6380
+    },
+    {
+      "epoch": 3.8825676909035596,
+      "grad_norm": 0.24928855895996094,
+      "learning_rate": 1.1298688517107081e-07,
+      "loss": 0.0046,
+      "step": 6381
+    },
+    {
+      "epoch": 3.8831761484636447,
+      "grad_norm": 0.20136448740959167,
+      "learning_rate": 1.1180775392401754e-07,
+      "loss": 0.0037,
+      "step": 6382
+    },
+    {
+      "epoch": 3.8837846060237298,
+      "grad_norm": 0.2751845121383667,
+      "learning_rate": 1.1063479391124898e-07,
+      "loss": 0.0036,
+      "step": 6383
+    },
+    {
+      "epoch": 3.884393063583815,
+      "grad_norm": 0.25994813442230225,
+      "learning_rate": 1.0946800542361025e-07,
+      "loss": 0.0059,
+      "step": 6384
+    },
+    {
+      "epoch": 3.8850015211439004,
+      "grad_norm": 0.2973096966743469,
+      "learning_rate": 1.083073887504199e-07,
+      "loss": 0.0103,
+      "step": 6385
+    },
+    {
+      "epoch": 3.8856099787039855,
+      "grad_norm": 0.1909116506576538,
+      "learning_rate": 1.0715294417946164e-07,
+      "loss": 0.0021,
+      "step": 6386
+    },
+    {
+      "epoch": 3.8862184362640706,
+      "grad_norm": 0.17645008862018585,
+      "learning_rate": 1.0600467199698427e-07,
+      "loss": 0.0044,
+      "step": 6387
+    },
+    {
+      "epoch": 3.8868268938241557,
+      "grad_norm": 0.24159935116767883,
+      "learning_rate": 1.0486257248771835e-07,
+      "loss": 0.0088,
+      "step": 6388
+    },
+    {
+      "epoch": 3.887435351384241,
+      "grad_norm": 0.30770307779312134,
+      "learning_rate": 1.0372664593485403e-07,
+      "loss": 0.0055,
+      "step": 6389
+    },
+    {
+      "epoch": 3.8880438089443263,
+      "grad_norm": 0.7140611410140991,
+      "learning_rate": 1.0259689262005212e-07,
+      "loss": 0.0058,
+      "step": 6390
+    },
+    {
+      "epoch": 3.8886522665044114,
+      "grad_norm": 0.249837726354599,
+      "learning_rate": 1.0147331282344686e-07,
+      "loss": 0.0047,
+      "step": 6391
+    },
+    {
+      "epoch": 3.8892607240644965,
+      "grad_norm": 0.1296958178281784,
+      "learning_rate": 1.003559068236376e-07,
+      "loss": 0.0019,
+      "step": 6392
+    },
+    {
+      "epoch": 3.8898691816245816,
+      "grad_norm": 0.31125199794769287,
+      "learning_rate": 9.924467489769717e-08,
+      "loss": 0.0084,
+      "step": 6393
+    },
+    {
+      "epoch": 3.8904776391846667,
+      "grad_norm": 0.25288596749305725,
+      "learning_rate": 9.813961732116073e-08,
+      "loss": 0.0079,
+      "step": 6394
+    },
+    {
+      "epoch": 3.8910860967447523,
+      "grad_norm": 0.21936556696891785,
+      "learning_rate": 9.704073436803685e-08,
+      "loss": 0.0032,
+      "step": 6395
+    },
+    {
+      "epoch": 3.8916945543048374,
+      "grad_norm": 0.3556055724620819,
+      "learning_rate": 9.594802631080756e-08,
+      "loss": 0.0119,
+      "step": 6396
+    },
+    {
+      "epoch": 3.8923030118649224,
+      "grad_norm": 0.2905397117137909,
+      "learning_rate": 9.48614934204145e-08,
+      "loss": 0.007,
+      "step": 6397
+    },
+    {
+      "epoch": 3.8929114694250075,
+      "grad_norm": 0.27061349153518677,
+      "learning_rate": 9.378113596627546e-08,
+      "loss": 0.0089,
+      "step": 6398
+    },
+    {
+      "epoch": 3.8935199269850926,
+      "grad_norm": 0.182336688041687,
+      "learning_rate": 9.270695421626784e-08,
+      "loss": 0.0036,
+      "step": 6399
+    },
+    {
+      "epoch": 3.894128384545178,
+      "grad_norm": 0.07943737506866455,
+      "learning_rate": 9.163894843675357e-08,
+      "loss": 0.0007,
+      "step": 6400
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.24037736654281616,
+      "learning_rate": 9.057711889254584e-08,
+      "loss": 0.0051,
+      "step": 6401
+    },
+    {
+      "epoch": 3.8953452996653484,
+      "grad_norm": 0.28533557057380676,
+      "learning_rate": 8.952146584693677e-08,
+      "loss": 0.0073,
+      "step": 6402
+    },
+    {
+      "epoch": 3.8959537572254335,
+      "grad_norm": 0.1930028647184372,
+      "learning_rate": 8.847198956168368e-08,
+      "loss": 0.0045,
+      "step": 6403
+    },
+    {
+      "epoch": 3.8965622147855186,
+      "grad_norm": 0.24128195643424988,
+      "learning_rate": 8.742869029701451e-08,
+      "loss": 0.0045,
+      "step": 6404
+    },
+    {
+      "epoch": 3.897170672345604,
+      "grad_norm": 0.2623744308948517,
+      "learning_rate": 8.639156831162231e-08,
+      "loss": 0.0079,
+      "step": 6405
+    },
+    {
+      "epoch": 3.897779129905689,
+      "grad_norm": 0.31534165143966675,
+      "learning_rate": 8.536062386267362e-08,
+      "loss": 0.0061,
+      "step": 6406
+    },
+    {
+      "epoch": 3.8983875874657743,
+      "grad_norm": 0.22992461919784546,
+      "learning_rate": 8.433585720579173e-08,
+      "loss": 0.0066,
+      "step": 6407
+    },
+    {
+      "epoch": 3.8989960450258594,
+      "grad_norm": 0.17727993428707123,
+      "learning_rate": 8.331726859508726e-08,
+      "loss": 0.0031,
+      "step": 6408
+    },
+    {
+      "epoch": 3.8996045025859445,
+      "grad_norm": 0.2012277990579605,
+      "learning_rate": 8.230485828311651e-08,
+      "loss": 0.0044,
+      "step": 6409
+    },
+    {
+      "epoch": 3.90021296014603,
+      "grad_norm": 0.253017783164978,
+      "learning_rate": 8.129862652092313e-08,
+      "loss": 0.0045,
+      "step": 6410
+    },
+    {
+      "epoch": 3.900821417706115,
+      "grad_norm": 0.19332891702651978,
+      "learning_rate": 8.029857355800475e-08,
+      "loss": 0.0044,
+      "step": 6411
+    },
+    {
+      "epoch": 3.9014298752662,
+      "grad_norm": 0.27358150482177734,
+      "learning_rate": 7.930469964234078e-08,
+      "loss": 0.0062,
+      "step": 6412
+    },
+    {
+      "epoch": 3.9020383328262853,
+      "grad_norm": 0.23901762068271637,
+      "learning_rate": 7.83170050203591e-08,
+      "loss": 0.0046,
+      "step": 6413
+    },
+    {
+      "epoch": 3.9026467903863704,
+      "grad_norm": 0.23956961929798126,
+      "learning_rate": 7.73354899369777e-08,
+      "loss": 0.0065,
+      "step": 6414
+    },
+    {
+      "epoch": 3.903255247946456,
+      "grad_norm": 0.25745102763175964,
+      "learning_rate": 7.636015463556578e-08,
+      "loss": 0.0074,
+      "step": 6415
+    },
+    {
+      "epoch": 3.903863705506541,
+      "grad_norm": 0.23999325931072235,
+      "learning_rate": 7.539099935796879e-08,
+      "loss": 0.0059,
+      "step": 6416
+    },
+    {
+      "epoch": 3.904472163066626,
+      "grad_norm": 0.18391796946525574,
+      "learning_rate": 7.442802434449169e-08,
+      "loss": 0.0043,
+      "step": 6417
+    },
+    {
+      "epoch": 3.9050806206267112,
+      "grad_norm": 0.2279115468263626,
+      "learning_rate": 7.347122983391851e-08,
+      "loss": 0.0061,
+      "step": 6418
+    },
+    {
+      "epoch": 3.9056890781867963,
+      "grad_norm": 0.3626938760280609,
+      "learning_rate": 7.252061606349003e-08,
+      "loss": 0.0105,
+      "step": 6419
+    },
+    {
+      "epoch": 3.906297535746882,
+      "grad_norm": 0.26662662625312805,
+      "learning_rate": 7.157618326892046e-08,
+      "loss": 0.0039,
+      "step": 6420
+    },
+    {
+      "epoch": 3.906905993306967,
+      "grad_norm": 0.23431222140789032,
+      "learning_rate": 7.063793168438915e-08,
+      "loss": 0.0097,
+      "step": 6421
+    },
+    {
+      "epoch": 3.907514450867052,
+      "grad_norm": 0.23029586672782898,
+      "learning_rate": 6.970586154254333e-08,
+      "loss": 0.0044,
+      "step": 6422
+    },
+    {
+      "epoch": 3.908122908427137,
+      "grad_norm": 0.30431026220321655,
+      "learning_rate": 6.877997307449813e-08,
+      "loss": 0.003,
+      "step": 6423
+    },
+    {
+      "epoch": 3.9087313659872223,
+      "grad_norm": 0.31623929738998413,
+      "learning_rate": 6.786026650983657e-08,
+      "loss": 0.003,
+      "step": 6424
+    },
+    {
+      "epoch": 3.909339823547308,
+      "grad_norm": 0.2245815098285675,
+      "learning_rate": 6.6946742076604e-08,
+      "loss": 0.0021,
+      "step": 6425
+    },
+    {
+      "epoch": 3.909948281107393,
+      "grad_norm": 0.20678408443927765,
+      "learning_rate": 6.603940000132203e-08,
+      "loss": 0.0039,
+      "step": 6426
+    },
+    {
+      "epoch": 3.910556738667478,
+      "grad_norm": 0.32013562321662903,
+      "learning_rate": 6.513824050896622e-08,
+      "loss": 0.0073,
+      "step": 6427
+    },
+    {
+      "epoch": 3.911165196227563,
+      "grad_norm": 0.26314496994018555,
+      "learning_rate": 6.424326382299394e-08,
+      "loss": 0.0059,
+      "step": 6428
+    },
+    {
+      "epoch": 3.911773653787648,
+      "grad_norm": 0.2609120011329651,
+      "learning_rate": 6.33544701653166e-08,
+      "loss": 0.0074,
+      "step": 6429
+    },
+    {
+      "epoch": 3.9123821113477337,
+      "grad_norm": 0.22066128253936768,
+      "learning_rate": 6.247185975631897e-08,
+      "loss": 0.0059,
+      "step": 6430
+    },
+    {
+      "epoch": 3.912990568907819,
+      "grad_norm": 0.24796698987483978,
+      "learning_rate": 6.15954328148538e-08,
+      "loss": 0.0056,
+      "step": 6431
+    },
+    {
+      "epoch": 3.913599026467904,
+      "grad_norm": 0.32350313663482666,
+      "learning_rate": 6.072518955823891e-08,
+      "loss": 0.0097,
+      "step": 6432
+    },
+    {
+      "epoch": 3.914207484027989,
+      "grad_norm": 0.24313661456108093,
+      "learning_rate": 5.986113020225448e-08,
+      "loss": 0.0052,
+      "step": 6433
+    },
+    {
+      "epoch": 3.914815941588074,
+      "grad_norm": 0.15304477512836456,
+      "learning_rate": 5.900325496115411e-08,
+      "loss": 0.0037,
+      "step": 6434
+    },
+    {
+      "epoch": 3.9154243991481597,
+      "grad_norm": 0.189406618475914,
+      "learning_rate": 5.815156404765654e-08,
+      "loss": 0.003,
+      "step": 6435
+    },
+    {
+      "epoch": 3.9160328567082447,
+      "grad_norm": 0.23600012063980103,
+      "learning_rate": 5.7306057672942833e-08,
+      "loss": 0.0029,
+      "step": 6436
+    },
+    {
+      "epoch": 3.91664131426833,
+      "grad_norm": 0.20945847034454346,
+      "learning_rate": 5.6466736046661974e-08,
+      "loss": 0.0038,
+      "step": 6437
+    },
+    {
+      "epoch": 3.917249771828415,
+      "grad_norm": 0.2792292535305023,
+      "learning_rate": 5.5633599376936353e-08,
+      "loss": 0.0071,
+      "step": 6438
+    },
+    {
+      "epoch": 3.9178582293885,
+      "grad_norm": 0.1970946043729782,
+      "learning_rate": 5.48066478703424e-08,
+      "loss": 0.0035,
+      "step": 6439
+    },
+    {
+      "epoch": 3.9184666869485856,
+      "grad_norm": 0.23822718858718872,
+      "learning_rate": 5.398588173193275e-08,
+      "loss": 0.0058,
+      "step": 6440
+    },
+    {
+      "epoch": 3.9190751445086707,
+      "grad_norm": 0.30124393105506897,
+      "learning_rate": 5.317130116522517e-08,
+      "loss": 0.0082,
+      "step": 6441
+    },
+    {
+      "epoch": 3.9196836020687558,
+      "grad_norm": 0.22608311474323273,
+      "learning_rate": 5.2362906372199764e-08,
+      "loss": 0.0037,
+      "step": 6442
+    },
+    {
+      "epoch": 3.920292059628841,
+      "grad_norm": 0.23782384395599365,
+      "learning_rate": 5.156069755330451e-08,
+      "loss": 0.0057,
+      "step": 6443
+    },
+    {
+      "epoch": 3.920900517188926,
+      "grad_norm": 0.20344187319278717,
+      "learning_rate": 5.0764674907452516e-08,
+      "loss": 0.0058,
+      "step": 6444
+    },
+    {
+      "epoch": 3.9215089747490115,
+      "grad_norm": 0.2951229214668274,
+      "learning_rate": 4.997483863202757e-08,
+      "loss": 0.0071,
+      "step": 6445
+    },
+    {
+      "epoch": 3.9221174323090966,
+      "grad_norm": 0.18063601851463318,
+      "learning_rate": 4.9191188922875775e-08,
+      "loss": 0.0046,
+      "step": 6446
+    },
+    {
+      "epoch": 3.9227258898691817,
+      "grad_norm": 0.2033570259809494,
+      "learning_rate": 4.8413725974305604e-08,
+      "loss": 0.0046,
+      "step": 6447
+    },
+    {
+      "epoch": 3.923334347429267,
+      "grad_norm": 0.3676776587963104,
+      "learning_rate": 4.764244997909895e-08,
+      "loss": 0.0053,
+      "step": 6448
+    },
+    {
+      "epoch": 3.923942804989352,
+      "grad_norm": 0.1831228882074356,
+      "learning_rate": 4.6877361128497276e-08,
+      "loss": 0.0039,
+      "step": 6449
+    },
+    {
+      "epoch": 3.9245512625494374,
+      "grad_norm": 0.17482344806194305,
+      "learning_rate": 4.61184596122155e-08,
+      "loss": 0.0048,
+      "step": 6450
+    },
+    {
+      "epoch": 3.9251597201095225,
+      "grad_norm": 0.25266796350479126,
+      "learning_rate": 4.5365745618425325e-08,
+      "loss": 0.007,
+      "step": 6451
+    },
+    {
+      "epoch": 3.9257681776696076,
+      "grad_norm": 0.25007006525993347,
+      "learning_rate": 4.4619219333769115e-08,
+      "loss": 0.0068,
+      "step": 6452
+    },
+    {
+      "epoch": 3.9263766352296927,
+      "grad_norm": 0.2621901035308838,
+      "learning_rate": 4.3878880943357124e-08,
+      "loss": 0.0078,
+      "step": 6453
+    },
+    {
+      "epoch": 3.926985092789778,
+      "grad_norm": 0.31776395440101624,
+      "learning_rate": 4.314473063075919e-08,
+      "loss": 0.0099,
+      "step": 6454
+    },
+    {
+      "epoch": 3.9275935503498633,
+      "grad_norm": 0.19145816564559937,
+      "learning_rate": 4.24167685780158e-08,
+      "loss": 0.0034,
+      "step": 6455
+    },
+    {
+      "epoch": 3.9282020079099484,
+      "grad_norm": 0.24003374576568604,
+      "learning_rate": 4.169499496562701e-08,
+      "loss": 0.004,
+      "step": 6456
+    },
+    {
+      "epoch": 3.9288104654700335,
+      "grad_norm": 0.19230146706104279,
+      "learning_rate": 4.097940997256911e-08,
+      "loss": 0.005,
+      "step": 6457
+    },
+    {
+      "epoch": 3.9294189230301186,
+      "grad_norm": 0.22203390300273895,
+      "learning_rate": 4.0270013776275153e-08,
+      "loss": 0.005,
+      "step": 6458
+    },
+    {
+      "epoch": 3.9300273805902037,
+      "grad_norm": 0.28375643491744995,
+      "learning_rate": 3.9566806552643335e-08,
+      "loss": 0.0063,
+      "step": 6459
+    },
+    {
+      "epoch": 3.9306358381502893,
+      "grad_norm": 0.20378822088241577,
+      "learning_rate": 3.8869788476039725e-08,
+      "loss": 0.0035,
+      "step": 6460
+    },
+    {
+      "epoch": 3.931244295710374,
+      "grad_norm": 0.611138641834259,
+      "learning_rate": 3.817895971930108e-08,
+      "loss": 0.0211,
+      "step": 6461
+    },
+    {
+      "epoch": 3.9318527532704595,
+      "grad_norm": 0.20204846560955048,
+      "learning_rate": 3.749432045371815e-08,
+      "loss": 0.005,
+      "step": 6462
+    },
+    {
+      "epoch": 3.9324612108305446,
+      "grad_norm": 0.2756820619106293,
+      "learning_rate": 3.6815870849055134e-08,
+      "loss": 0.0047,
+      "step": 6463
+    },
+    {
+      "epoch": 3.9330696683906297,
+      "grad_norm": 0.26881980895996094,
+      "learning_rate": 3.614361107354136e-08,
+      "loss": 0.004,
+      "step": 6464
+    },
+    {
+      "epoch": 3.933678125950715,
+      "grad_norm": 0.30591699481010437,
+      "learning_rate": 3.547754129386571e-08,
+      "loss": 0.0054,
+      "step": 6465
+    },
+    {
+      "epoch": 3.9342865835108,
+      "grad_norm": 0.3070509433746338,
+      "learning_rate": 3.481766167518774e-08,
+      "loss": 0.0074,
+      "step": 6466
+    },
+    {
+      "epoch": 3.9348950410708854,
+      "grad_norm": 0.21430547535419464,
+      "learning_rate": 3.416397238112934e-08,
+      "loss": 0.0054,
+      "step": 6467
+    },
+    {
+      "epoch": 3.9355034986309705,
+      "grad_norm": 0.15337048470973969,
+      "learning_rate": 3.351647357377752e-08,
+      "loss": 0.0031,
+      "step": 6468
+    },
+    {
+      "epoch": 3.9361119561910556,
+      "grad_norm": 0.17072342336177826,
+      "learning_rate": 3.2875165413687184e-08,
+      "loss": 0.0035,
+      "step": 6469
+    },
+    {
+      "epoch": 3.936720413751141,
+      "grad_norm": 0.22262074053287506,
+      "learning_rate": 3.2240048059872793e-08,
+      "loss": 0.0051,
+      "step": 6470
+    },
+    {
+      "epoch": 3.9373288713112258,
+      "grad_norm": 0.22603394091129303,
+      "learning_rate": 3.161112166982227e-08,
+      "loss": 0.0052,
+      "step": 6471
+    },
+    {
+      "epoch": 3.9379373288713113,
+      "grad_norm": 0.20640049874782562,
+      "learning_rate": 3.098838639947754e-08,
+      "loss": 0.0047,
+      "step": 6472
+    },
+    {
+      "epoch": 3.9385457864313964,
+      "grad_norm": 0.17780143022537231,
+      "learning_rate": 3.037184240325397e-08,
+      "loss": 0.0027,
+      "step": 6473
+    },
+    {
+      "epoch": 3.9391542439914815,
+      "grad_norm": 0.1945071816444397,
+      "learning_rate": 2.976148983402649e-08,
+      "loss": 0.0056,
+      "step": 6474
+    },
+    {
+      "epoch": 3.939762701551567,
+      "grad_norm": 0.2443944811820984,
+      "learning_rate": 2.9157328843140708e-08,
+      "loss": 0.005,
+      "step": 6475
+    },
+    {
+      "epoch": 3.9403711591116517,
+      "grad_norm": 0.1970926970243454,
+      "learning_rate": 2.855935958040179e-08,
+      "loss": 0.0036,
+      "step": 6476
+    },
+    {
+      "epoch": 3.9409796166717372,
+      "grad_norm": 0.2412329912185669,
+      "learning_rate": 2.7967582194080022e-08,
+      "loss": 0.0042,
+      "step": 6477
+    },
+    {
+      "epoch": 3.9415880742318223,
+      "grad_norm": 0.15232165157794952,
+      "learning_rate": 2.7381996830910805e-08,
+      "loss": 0.0041,
+      "step": 6478
+    },
+    {
+      "epoch": 3.9421965317919074,
+      "grad_norm": 0.2518268823623657,
+      "learning_rate": 2.6802603636097435e-08,
+      "loss": 0.0074,
+      "step": 6479
+    },
+    {
+      "epoch": 3.942804989351993,
+      "grad_norm": 0.25513625144958496,
+      "learning_rate": 2.6229402753305544e-08,
+      "loss": 0.0088,
+      "step": 6480
+    },
+    {
+      "epoch": 3.9434134469120776,
+      "grad_norm": 0.2241678684949875,
+      "learning_rate": 2.5662394324663108e-08,
+      "loss": 0.0053,
+      "step": 6481
+    },
+    {
+      "epoch": 3.944021904472163,
+      "grad_norm": 0.1274229884147644,
+      "learning_rate": 2.5101578490763223e-08,
+      "loss": 0.0015,
+      "step": 6482
+    },
+    {
+      "epoch": 3.9446303620322483,
+      "grad_norm": 0.20924502611160278,
+      "learning_rate": 2.4546955390669645e-08,
+      "loss": 0.003,
+      "step": 6483
+    },
+    {
+      "epoch": 3.9452388195923334,
+      "grad_norm": 0.23410069942474365,
+      "learning_rate": 2.3998525161900153e-08,
+      "loss": 0.0072,
+      "step": 6484
+    },
+    {
+      "epoch": 3.9458472771524185,
+      "grad_norm": 0.18146799504756927,
+      "learning_rate": 2.345628794044874e-08,
+      "loss": 0.0023,
+      "step": 6485
+    },
+    {
+      "epoch": 3.9464557347125035,
+      "grad_norm": 0.3447299003601074,
+      "learning_rate": 2.292024386076064e-08,
+      "loss": 0.0028,
+      "step": 6486
+    },
+    {
+      "epoch": 3.947064192272589,
+      "grad_norm": 0.2687872052192688,
+      "learning_rate": 2.2390393055757297e-08,
+      "loss": 0.0052,
+      "step": 6487
+    },
+    {
+      "epoch": 3.947672649832674,
+      "grad_norm": 0.20712141692638397,
+      "learning_rate": 2.1866735656819736e-08,
+      "loss": 0.0062,
+      "step": 6488
+    },
+    {
+      "epoch": 3.9482811073927593,
+      "grad_norm": 0.31326937675476074,
+      "learning_rate": 2.1349271793791313e-08,
+      "loss": 0.0079,
+      "step": 6489
+    },
+    {
+      "epoch": 3.9488895649528444,
+      "grad_norm": 0.3461199104785919,
+      "learning_rate": 2.0838001594980504e-08,
+      "loss": 0.009,
+      "step": 6490
+    },
+    {
+      "epoch": 3.9494980225129295,
+      "grad_norm": 0.18858326971530914,
+      "learning_rate": 2.0332925187163676e-08,
+      "loss": 0.0052,
+      "step": 6491
+    },
+    {
+      "epoch": 3.950106480073015,
+      "grad_norm": 0.26478853821754456,
+      "learning_rate": 1.983404269557676e-08,
+      "loss": 0.0061,
+      "step": 6492
+    },
+    {
+      "epoch": 3.9507149376331,
+      "grad_norm": 0.2934277057647705,
+      "learning_rate": 1.9341354243923583e-08,
+      "loss": 0.0034,
+      "step": 6493
+    },
+    {
+      "epoch": 3.951323395193185,
+      "grad_norm": 0.18972255289554596,
+      "learning_rate": 1.8854859954370306e-08,
+      "loss": 0.0042,
+      "step": 6494
+    },
+    {
+      "epoch": 3.9519318527532703,
+      "grad_norm": 0.3544655740261078,
+      "learning_rate": 1.8374559947545446e-08,
+      "loss": 0.0206,
+      "step": 6495
+    },
+    {
+      "epoch": 3.9525403103133554,
+      "grad_norm": 0.1487325131893158,
+      "learning_rate": 1.7900454342542616e-08,
+      "loss": 0.0032,
+      "step": 6496
+    },
+    {
+      "epoch": 3.953148767873441,
+      "grad_norm": 0.3513227701187134,
+      "learning_rate": 1.743254325692334e-08,
+      "loss": 0.0078,
+      "step": 6497
+    },
+    {
+      "epoch": 3.953757225433526,
+      "grad_norm": 0.19835789501667023,
+      "learning_rate": 1.6970826806708694e-08,
+      "loss": 0.004,
+      "step": 6498
+    },
+    {
+      "epoch": 3.954365682993611,
+      "grad_norm": 0.16036944091320038,
+      "learning_rate": 1.651530510638488e-08,
+      "loss": 0.0027,
+      "step": 6499
+    },
+    {
+      "epoch": 3.9549741405536962,
+      "grad_norm": 0.2660086750984192,
+      "learning_rate": 1.6065978268903215e-08,
+      "loss": 0.0068,
+      "step": 6500
+    },
+    {
+      "epoch": 3.9555825981137813,
+      "grad_norm": 0.18971866369247437,
+      "learning_rate": 1.562284640567735e-08,
+      "loss": 0.0053,
+      "step": 6501
+    },
+    {
+      "epoch": 3.956191055673867,
+      "grad_norm": 0.22135283052921295,
+      "learning_rate": 1.5185909626583284e-08,
+      "loss": 0.0057,
+      "step": 6502
+    },
+    {
+      "epoch": 3.956799513233952,
+      "grad_norm": 0.33586928248405457,
+      "learning_rate": 1.4755168039967682e-08,
+      "loss": 0.0154,
+      "step": 6503
+    },
+    {
+      "epoch": 3.957407970794037,
+      "grad_norm": 0.21343040466308594,
+      "learning_rate": 1.4330621752631224e-08,
+      "loss": 0.0041,
+      "step": 6504
+    },
+    {
+      "epoch": 3.958016428354122,
+      "grad_norm": 0.15883713960647583,
+      "learning_rate": 1.3912270869848032e-08,
+      "loss": 0.0037,
+      "step": 6505
+    },
+    {
+      "epoch": 3.9586248859142072,
+      "grad_norm": 0.33996376395225525,
+      "learning_rate": 1.3500115495351795e-08,
+      "loss": 0.0054,
+      "step": 6506
+    },
+    {
+      "epoch": 3.959233343474293,
+      "grad_norm": 0.22483274340629578,
+      "learning_rate": 1.3094155731335767e-08,
+      "loss": 0.0047,
+      "step": 6507
+    },
+    {
+      "epoch": 3.959841801034378,
+      "grad_norm": 0.3512939214706421,
+      "learning_rate": 1.2694391678463868e-08,
+      "loss": 0.0082,
+      "step": 6508
+    },
+    {
+      "epoch": 3.960450258594463,
+      "grad_norm": 0.3100382685661316,
+      "learning_rate": 1.2300823435862363e-08,
+      "loss": 0.0046,
+      "step": 6509
+    },
+    {
+      "epoch": 3.961058716154548,
+      "grad_norm": 0.15239042043685913,
+      "learning_rate": 1.191345110111708e-08,
+      "loss": 0.0034,
+      "step": 6510
+    },
+    {
+      "epoch": 3.961667173714633,
+      "grad_norm": 0.2930486798286438,
+      "learning_rate": 1.1532274770281737e-08,
+      "loss": 0.0059,
+      "step": 6511
+    },
+    {
+      "epoch": 3.9622756312747187,
+      "grad_norm": 0.18088679015636444,
+      "learning_rate": 1.1157294537869622e-08,
+      "loss": 0.0063,
+      "step": 6512
+    },
+    {
+      "epoch": 3.962884088834804,
+      "grad_norm": 0.2748885452747345,
+      "learning_rate": 1.078851049686469e-08,
+      "loss": 0.0051,
+      "step": 6513
+    },
+    {
+      "epoch": 3.963492546394889,
+      "grad_norm": 0.25067272782325745,
+      "learning_rate": 1.0425922738704908e-08,
+      "loss": 0.0065,
+      "step": 6514
+    },
+    {
+      "epoch": 3.964101003954974,
+      "grad_norm": 0.2923562228679657,
+      "learning_rate": 1.0069531353301687e-08,
+      "loss": 0.0112,
+      "step": 6515
+    },
+    {
+      "epoch": 3.964709461515059,
+      "grad_norm": 0.23911620676517487,
+      "learning_rate": 9.719336429023229e-09,
+      "loss": 0.0069,
+      "step": 6516
+    },
+    {
+      "epoch": 3.9653179190751446,
+      "grad_norm": 0.23181907832622528,
+      "learning_rate": 9.375338052702853e-09,
+      "loss": 0.0054,
+      "step": 6517
+    },
+    {
+      "epoch": 3.9659263766352297,
+      "grad_norm": 0.25822049379348755,
+      "learning_rate": 9.037536309636218e-09,
+      "loss": 0.0059,
+      "step": 6518
+    },
+    {
+      "epoch": 3.966534834195315,
+      "grad_norm": 0.29777318239212036,
+      "learning_rate": 8.705931283586877e-09,
+      "loss": 0.008,
+      "step": 6519
+    },
+    {
+      "epoch": 3.9671432917554,
+      "grad_norm": 0.440320760011673,
+      "learning_rate": 8.380523056777944e-09,
+      "loss": 0.008,
+      "step": 6520
+    },
+    {
+      "epoch": 3.967751749315485,
+      "grad_norm": 0.2474939078092575,
+      "learning_rate": 8.061311709897656e-09,
+      "loss": 0.0038,
+      "step": 6521
+    },
+    {
+      "epoch": 3.9683602068755706,
+      "grad_norm": 0.4338011145591736,
+      "learning_rate": 7.748297322096586e-09,
+      "loss": 0.007,
+      "step": 6522
+    },
+    {
+      "epoch": 3.9689686644356557,
+      "grad_norm": 0.20323798060417175,
+      "learning_rate": 7.441479970990428e-09,
+      "loss": 0.0051,
+      "step": 6523
+    },
+    {
+      "epoch": 3.9695771219957408,
+      "grad_norm": 0.21181131899356842,
+      "learning_rate": 7.140859732654437e-09,
+      "loss": 0.0047,
+      "step": 6524
+    },
+    {
+      "epoch": 3.970185579555826,
+      "grad_norm": 0.25651806592941284,
+      "learning_rate": 6.846436681631763e-09,
+      "loss": 0.006,
+      "step": 6525
+    },
+    {
+      "epoch": 3.970794037115911,
+      "grad_norm": 0.20691846311092377,
+      "learning_rate": 6.558210890927896e-09,
+      "loss": 0.0041,
+      "step": 6526
+    },
+    {
+      "epoch": 3.9714024946759965,
+      "grad_norm": 0.21363694965839386,
+      "learning_rate": 6.2761824320106684e-09,
+      "loss": 0.0032,
+      "step": 6527
+    },
+    {
+      "epoch": 3.9720109522360816,
+      "grad_norm": 0.22716128826141357,
+      "learning_rate": 6.000351374807478e-09,
+      "loss": 0.005,
+      "step": 6528
+    },
+    {
+      "epoch": 3.9726194097961667,
+      "grad_norm": 0.19671691954135895,
+      "learning_rate": 5.730717787716389e-09,
+      "loss": 0.0042,
+      "step": 6529
+    },
+    {
+      "epoch": 3.9732278673562518,
+      "grad_norm": 0.27143892645835876,
+      "learning_rate": 5.467281737597807e-09,
+      "loss": 0.0053,
+      "step": 6530
+    },
+    {
+      "epoch": 3.973836324916337,
+      "grad_norm": 0.20458009839057922,
+      "learning_rate": 5.2100432897661535e-09,
+      "loss": 0.0049,
+      "step": 6531
+    },
+    {
+      "epoch": 3.9744447824764224,
+      "grad_norm": 0.2738465964794159,
+      "learning_rate": 4.959002508012067e-09,
+      "loss": 0.0075,
+      "step": 6532
+    },
+    {
+      "epoch": 3.9750532400365075,
+      "grad_norm": 0.2310987263917923,
+      "learning_rate": 4.714159454580203e-09,
+      "loss": 0.0029,
+      "step": 6533
+    },
+    {
+      "epoch": 3.9756616975965926,
+      "grad_norm": 0.33232930302619934,
+      "learning_rate": 4.47551419018033e-09,
+      "loss": 0.0072,
+      "step": 6534
+    },
+    {
+      "epoch": 3.9762701551566777,
+      "grad_norm": 0.21468819677829742,
+      "learning_rate": 4.243066773990112e-09,
+      "loss": 0.005,
+      "step": 6535
+    },
+    {
+      "epoch": 3.976878612716763,
+      "grad_norm": 0.21161320805549622,
+      "learning_rate": 4.016817263644002e-09,
+      "loss": 0.0057,
+      "step": 6536
+    },
+    {
+      "epoch": 3.9774870702768483,
+      "grad_norm": 0.22753821313381195,
+      "learning_rate": 3.796765715244344e-09,
+      "loss": 0.0048,
+      "step": 6537
+    },
+    {
+      "epoch": 3.9780955278369334,
+      "grad_norm": 0.229328915476799,
+      "learning_rate": 3.5829121833530488e-09,
+      "loss": 0.0063,
+      "step": 6538
+    },
+    {
+      "epoch": 3.9787039853970185,
+      "grad_norm": 0.25419679284095764,
+      "learning_rate": 3.3752567209971444e-09,
+      "loss": 0.0048,
+      "step": 6539
+    },
+    {
+      "epoch": 3.9793124429571036,
+      "grad_norm": 0.2747920751571655,
+      "learning_rate": 3.173799379665998e-09,
+      "loss": 0.0061,
+      "step": 6540
+    },
+    {
+      "epoch": 3.9799209005171887,
+      "grad_norm": 0.24330097436904907,
+      "learning_rate": 2.978540209314096e-09,
+      "loss": 0.0035,
+      "step": 6541
+    },
+    {
+      "epoch": 3.9805293580772743,
+      "grad_norm": 0.20687943696975708,
+      "learning_rate": 2.789479258358263e-09,
+      "loss": 0.0051,
+      "step": 6542
+    },
+    {
+      "epoch": 3.9811378156373594,
+      "grad_norm": 0.22026467323303223,
+      "learning_rate": 2.6066165736748914e-09,
+      "loss": 0.0068,
+      "step": 6543
+    },
+    {
+      "epoch": 3.9817462731974445,
+      "grad_norm": 0.22411870956420898,
+      "learning_rate": 2.429952200611041e-09,
+      "loss": 0.0042,
+      "step": 6544
+    },
+    {
+      "epoch": 3.9823547307575295,
+      "grad_norm": 0.250068336725235,
+      "learning_rate": 2.2594861829650094e-09,
+      "loss": 0.0051,
+      "step": 6545
+    },
+    {
+      "epoch": 3.9829631883176146,
+      "grad_norm": 0.2509680688381195,
+      "learning_rate": 2.095218563011314e-09,
+      "loss": 0.0048,
+      "step": 6546
+    },
+    {
+      "epoch": 3.9835716458777,
+      "grad_norm": 0.20843267440795898,
+      "learning_rate": 1.9371493814784867e-09,
+      "loss": 0.0041,
+      "step": 6547
+    },
+    {
+      "epoch": 3.9841801034377853,
+      "grad_norm": 0.12238210439682007,
+      "learning_rate": 1.7852786775629516e-09,
+      "loss": 0.0014,
+      "step": 6548
+    },
+    {
+      "epoch": 3.9847885609978704,
+      "grad_norm": 0.22475853562355042,
+      "learning_rate": 1.6396064889206974e-09,
+      "loss": 0.0046,
+      "step": 6549
+    },
+    {
+      "epoch": 3.9853970185579555,
+      "grad_norm": 0.1801108866930008,
+      "learning_rate": 1.5001328516728309e-09,
+      "loss": 0.0037,
+      "step": 6550
+    },
+    {
+      "epoch": 3.9860054761180406,
+      "grad_norm": 0.2936258912086487,
+      "learning_rate": 1.3668578004027988e-09,
+      "loss": 0.0074,
+      "step": 6551
+    },
+    {
+      "epoch": 3.986613933678126,
+      "grad_norm": 0.27133995294570923,
+      "learning_rate": 1.2397813681591653e-09,
+      "loss": 0.0061,
+      "step": 6552
+    },
+    {
+      "epoch": 3.987222391238211,
+      "grad_norm": 0.29440489411354065,
+      "learning_rate": 1.1189035864500597e-09,
+      "loss": 0.006,
+      "step": 6553
+    },
+    {
+      "epoch": 3.9878308487982963,
+      "grad_norm": 0.24190589785575867,
+      "learning_rate": 1.0042244852487281e-09,
+      "loss": 0.0042,
+      "step": 6554
+    },
+    {
+      "epoch": 3.9884393063583814,
+      "grad_norm": 0.2079675942659378,
+      "learning_rate": 8.957440929879823e-10,
+      "loss": 0.0042,
+      "step": 6555
+    },
+    {
+      "epoch": 3.9890477639184665,
+      "grad_norm": 0.2172488123178482,
+      "learning_rate": 7.934624365685261e-10,
+      "loss": 0.0041,
+      "step": 6556
+    },
+    {
+      "epoch": 3.989656221478552,
+      "grad_norm": 0.23664279282093048,
+      "learning_rate": 6.973795413534045e-10,
+      "loss": 0.0086,
+      "step": 6557
+    },
+    {
+      "epoch": 3.990264679038637,
+      "grad_norm": 0.12852723896503448,
+      "learning_rate": 6.074954311652281e-10,
+      "loss": 0.0013,
+      "step": 6558
+    },
+    {
+      "epoch": 3.9908731365987222,
+      "grad_norm": 0.24331684410572052,
+      "learning_rate": 5.238101282944996e-10,
+      "loss": 0.0056,
+      "step": 6559
+    },
+    {
+      "epoch": 3.9914815941588073,
+      "grad_norm": 0.49298709630966187,
+      "learning_rate": 4.4632365348573625e-10,
+      "loss": 0.0224,
+      "step": 6560
+    },
+    {
+      "epoch": 3.9920900517188924,
+      "grad_norm": 0.31049343943595886,
+      "learning_rate": 3.750360259596741e-10,
+      "loss": 0.0094,
+      "step": 6561
+    },
+    {
+      "epoch": 3.992698509278978,
+      "grad_norm": 0.2852802276611328,
+      "learning_rate": 3.099472633855127e-10,
+      "loss": 0.0075,
+      "step": 6562
+    },
+    {
+      "epoch": 3.993306966839063,
+      "grad_norm": 0.20569032430648804,
+      "learning_rate": 2.5105738190867033e-10,
+      "loss": 0.0051,
+      "step": 6563
+    },
+    {
+      "epoch": 3.993915424399148,
+      "grad_norm": 0.23459145426750183,
+      "learning_rate": 1.9836639612580422e-10,
+      "loss": 0.0032,
+      "step": 6564
+    },
+    {
+      "epoch": 3.9945238819592332,
+      "grad_norm": 0.24636982381343842,
+      "learning_rate": 1.518743191070149e-10,
+      "loss": 0.0067,
+      "step": 6565
+    },
+    {
+      "epoch": 3.9951323395193183,
+      "grad_norm": 0.21381396055221558,
+      "learning_rate": 1.1158116237919292e-10,
+      "loss": 0.0054,
+      "step": 6566
+    },
+    {
+      "epoch": 3.995740797079404,
+      "grad_norm": 0.2777492105960846,
+      "learning_rate": 7.748693593156997e-11,
+      "loss": 0.0209,
+      "step": 6567
+    },
+    {
+      "epoch": 3.996349254639489,
+      "grad_norm": 0.229709655046463,
+      "learning_rate": 4.959164821849438e-11,
+      "loss": 0.0044,
+      "step": 6568
+    },
+    {
+      "epoch": 3.996957712199574,
+      "grad_norm": 0.21392680704593658,
+      "learning_rate": 2.7895306156655587e-11,
+      "loss": 0.0033,
+      "step": 6569
+    },
+    {
+      "epoch": 3.997566169759659,
+      "grad_norm": 0.23358656466007233,
+      "learning_rate": 1.2397915127859705e-11,
+      "loss": 0.0094,
+      "step": 6570
+    },
+    {
+      "epoch": 3.9981746273197443,
+      "grad_norm": 0.23164047300815582,
+      "learning_rate": 3.099478973478398e-12,
+      "loss": 0.0039,
+      "step": 6571
+    },
+    {
+      "epoch": 3.99878308487983,
+      "grad_norm": 0.16882868111133575,
+      "learning_rate": 0.0,
+      "loss": 0.002,
+      "step": 6572
+    },
+    {
+      "epoch": 3.99878308487983,
+      "eval_loss": 1.711881399154663,
+      "eval_runtime": 104.2004,
+      "eval_samples_per_second": 7.313,
+      "eval_steps_per_second": 0.461,
+      "step": 6572
+    },
+    {
+      "epoch": 3.99878308487983,
+      "step": 6572,
+      "total_flos": 9.69717152698545e+18,
+      "train_loss": 0.1857661268678992,
+      "train_runtime": 120983.7455,
+      "train_samples_per_second": 1.739,
+      "train_steps_per_second": 0.054
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 6572,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "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": 9.69717152698545e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}