diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12957 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.999729070712544,
+  "eval_steps": 500,
+  "global_step": 1845,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.000541858574911948,
+      "grad_norm": 0.43530920147895813,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.8962,
+      "step": 1
+    },
+    {
+      "epoch": 0.001083717149823896,
+      "grad_norm": 0.6412757635116577,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.9589,
+      "step": 2
+    },
+    {
+      "epoch": 0.001625575724735844,
+      "grad_norm": 0.47238072752952576,
+      "learning_rate": 6e-06,
+      "loss": 1.9824,
+      "step": 3
+    },
+    {
+      "epoch": 0.002167434299647792,
+      "grad_norm": 0.4227930009365082,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.9207,
+      "step": 4
+    },
+    {
+      "epoch": 0.00270929287455974,
+      "grad_norm": 0.46128469705581665,
+      "learning_rate": 1e-05,
+      "loss": 1.9365,
+      "step": 5
+    },
+    {
+      "epoch": 0.003251151449471688,
+      "grad_norm": 0.5934392213821411,
+      "learning_rate": 1.2e-05,
+      "loss": 1.8873,
+      "step": 6
+    },
+    {
+      "epoch": 0.003793010024383636,
+      "grad_norm": 0.39757072925567627,
+      "learning_rate": 1.4000000000000001e-05,
+      "loss": 1.8462,
+      "step": 7
+    },
+    {
+      "epoch": 0.004334868599295584,
+      "grad_norm": 0.3870404064655304,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.9607,
+      "step": 8
+    },
+    {
+      "epoch": 0.004876727174207532,
+      "grad_norm": 0.335467666387558,
+      "learning_rate": 1.8e-05,
+      "loss": 2.0285,
+      "step": 9
+    },
+    {
+      "epoch": 0.00541858574911948,
+      "grad_norm": 0.3384644091129303,
+      "learning_rate": 2e-05,
+      "loss": 1.9302,
+      "step": 10
+    },
+    {
+      "epoch": 0.005960444324031428,
+      "grad_norm": 0.29000476002693176,
+      "learning_rate": 2.2000000000000003e-05,
+      "loss": 1.9069,
+      "step": 11
+    },
+    {
+      "epoch": 0.006502302898943376,
+      "grad_norm": 0.22133825719356537,
+      "learning_rate": 2.4e-05,
+      "loss": 1.8441,
+      "step": 12
+    },
+    {
+      "epoch": 0.007044161473855324,
+      "grad_norm": 0.2738000154495239,
+      "learning_rate": 2.6000000000000002e-05,
+      "loss": 1.8498,
+      "step": 13
+    },
+    {
+      "epoch": 0.007586020048767272,
+      "grad_norm": 0.2651294469833374,
+      "learning_rate": 2.8000000000000003e-05,
+      "loss": 1.8377,
+      "step": 14
+    },
+    {
+      "epoch": 0.00812787862367922,
+      "grad_norm": 0.3199697732925415,
+      "learning_rate": 3e-05,
+      "loss": 1.9074,
+      "step": 15
+    },
+    {
+      "epoch": 0.008669737198591168,
+      "grad_norm": 0.28017792105674744,
+      "learning_rate": 3.2000000000000005e-05,
+      "loss": 1.8188,
+      "step": 16
+    },
+    {
+      "epoch": 0.009211595773503115,
+      "grad_norm": 0.2615055739879608,
+      "learning_rate": 3.4000000000000007e-05,
+      "loss": 1.9161,
+      "step": 17
+    },
+    {
+      "epoch": 0.009753454348415064,
+      "grad_norm": 0.19840335845947266,
+      "learning_rate": 3.6e-05,
+      "loss": 1.8613,
+      "step": 18
+    },
+    {
+      "epoch": 0.010295312923327011,
+      "grad_norm": 0.14807601273059845,
+      "learning_rate": 3.8e-05,
+      "loss": 1.7986,
+      "step": 19
+    },
+    {
+      "epoch": 0.01083717149823896,
+      "grad_norm": 0.15882571041584015,
+      "learning_rate": 4e-05,
+      "loss": 1.8822,
+      "step": 20
+    },
+    {
+      "epoch": 0.011379030073150907,
+      "grad_norm": 0.1381528228521347,
+      "learning_rate": 4.2e-05,
+      "loss": 1.8117,
+      "step": 21
+    },
+    {
+      "epoch": 0.011920888648062856,
+      "grad_norm": 0.17201915383338928,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.8538,
+      "step": 22
+    },
+    {
+      "epoch": 0.012462747222974803,
+      "grad_norm": 0.17138561606407166,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.7731,
+      "step": 23
+    },
+    {
+      "epoch": 0.013004605797886752,
+      "grad_norm": 0.16845813393592834,
+      "learning_rate": 4.8e-05,
+      "loss": 1.8515,
+      "step": 24
+    },
+    {
+      "epoch": 0.0135464643727987,
+      "grad_norm": 0.13196821510791779,
+      "learning_rate": 5e-05,
+      "loss": 1.7643,
+      "step": 25
+    },
+    {
+      "epoch": 0.014088322947710648,
+      "grad_norm": 0.15632478892803192,
+      "learning_rate": 5.2000000000000004e-05,
+      "loss": 1.9114,
+      "step": 26
+    },
+    {
+      "epoch": 0.014630181522622595,
+      "grad_norm": 0.1211211085319519,
+      "learning_rate": 5.4000000000000005e-05,
+      "loss": 1.8366,
+      "step": 27
+    },
+    {
+      "epoch": 0.015172040097534544,
+      "grad_norm": 0.11397989839315414,
+      "learning_rate": 5.6000000000000006e-05,
+      "loss": 1.7707,
+      "step": 28
+    },
+    {
+      "epoch": 0.01571389867244649,
+      "grad_norm": 0.11910925805568695,
+      "learning_rate": 5.8e-05,
+      "loss": 1.8173,
+      "step": 29
+    },
+    {
+      "epoch": 0.01625575724735844,
+      "grad_norm": 0.12323752790689468,
+      "learning_rate": 6e-05,
+      "loss": 1.7767,
+      "step": 30
+    },
+    {
+      "epoch": 0.01679761582227039,
+      "grad_norm": 0.13322019577026367,
+      "learning_rate": 6.2e-05,
+      "loss": 1.8332,
+      "step": 31
+    },
+    {
+      "epoch": 0.017339474397182336,
+      "grad_norm": 0.12622202932834625,
+      "learning_rate": 6.400000000000001e-05,
+      "loss": 1.8144,
+      "step": 32
+    },
+    {
+      "epoch": 0.017881332972094283,
+      "grad_norm": 0.09827224165201187,
+      "learning_rate": 6.6e-05,
+      "loss": 1.8452,
+      "step": 33
+    },
+    {
+      "epoch": 0.01842319154700623,
+      "grad_norm": 0.10827549546957016,
+      "learning_rate": 6.800000000000001e-05,
+      "loss": 1.8212,
+      "step": 34
+    },
+    {
+      "epoch": 0.01896505012191818,
+      "grad_norm": 0.10839946568012238,
+      "learning_rate": 7e-05,
+      "loss": 1.7667,
+      "step": 35
+    },
+    {
+      "epoch": 0.019506908696830128,
+      "grad_norm": 0.09671394526958466,
+      "learning_rate": 7.2e-05,
+      "loss": 1.7415,
+      "step": 36
+    },
+    {
+      "epoch": 0.020048767271742075,
+      "grad_norm": 0.10029514878988266,
+      "learning_rate": 7.4e-05,
+      "loss": 1.748,
+      "step": 37
+    },
+    {
+      "epoch": 0.020590625846654022,
+      "grad_norm": 0.09756534546613693,
+      "learning_rate": 7.6e-05,
+      "loss": 1.8568,
+      "step": 38
+    },
+    {
+      "epoch": 0.02113248442156597,
+      "grad_norm": 0.10066157579421997,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 1.8028,
+      "step": 39
+    },
+    {
+      "epoch": 0.02167434299647792,
+      "grad_norm": 0.08360892534255981,
+      "learning_rate": 8e-05,
+      "loss": 1.7305,
+      "step": 40
+    },
+    {
+      "epoch": 0.022216201571389867,
+      "grad_norm": 0.09223756939172745,
+      "learning_rate": 8.2e-05,
+      "loss": 1.755,
+      "step": 41
+    },
+    {
+      "epoch": 0.022758060146301815,
+      "grad_norm": 0.09345275908708572,
+      "learning_rate": 8.4e-05,
+      "loss": 1.7815,
+      "step": 42
+    },
+    {
+      "epoch": 0.02329991872121376,
+      "grad_norm": 0.09920019656419754,
+      "learning_rate": 8.6e-05,
+      "loss": 1.8139,
+      "step": 43
+    },
+    {
+      "epoch": 0.023841777296125712,
+      "grad_norm": 0.09716581553220749,
+      "learning_rate": 8.800000000000001e-05,
+      "loss": 1.7593,
+      "step": 44
+    },
+    {
+      "epoch": 0.02438363587103766,
+      "grad_norm": 0.09761088341474533,
+      "learning_rate": 9e-05,
+      "loss": 1.8432,
+      "step": 45
+    },
+    {
+      "epoch": 0.024925494445949607,
+      "grad_norm": 0.10412333160638809,
+      "learning_rate": 9.200000000000001e-05,
+      "loss": 1.788,
+      "step": 46
+    },
+    {
+      "epoch": 0.025467353020861554,
+      "grad_norm": 0.09423588216304779,
+      "learning_rate": 9.4e-05,
+      "loss": 1.7802,
+      "step": 47
+    },
+    {
+      "epoch": 0.026009211595773504,
+      "grad_norm": 0.09519781917333603,
+      "learning_rate": 9.6e-05,
+      "loss": 1.7829,
+      "step": 48
+    },
+    {
+      "epoch": 0.02655107017068545,
+      "grad_norm": 0.09376217424869537,
+      "learning_rate": 9.8e-05,
+      "loss": 1.7308,
+      "step": 49
+    },
+    {
+      "epoch": 0.0270929287455974,
+      "grad_norm": 0.09659101814031601,
+      "learning_rate": 0.0001,
+      "loss": 1.7314,
+      "step": 50
+    },
+    {
+      "epoch": 0.027634787320509346,
+      "grad_norm": 0.10163891315460205,
+      "learning_rate": 0.0001,
+      "loss": 1.832,
+      "step": 51
+    },
+    {
+      "epoch": 0.028176645895421296,
+      "grad_norm": 0.10365260392427444,
+      "learning_rate": 0.0001,
+      "loss": 1.7844,
+      "step": 52
+    },
+    {
+      "epoch": 0.028718504470333243,
+      "grad_norm": 0.11085116118192673,
+      "learning_rate": 0.0001,
+      "loss": 1.7985,
+      "step": 53
+    },
+    {
+      "epoch": 0.02926036304524519,
+      "grad_norm": 0.0932428166270256,
+      "learning_rate": 0.0001,
+      "loss": 1.7481,
+      "step": 54
+    },
+    {
+      "epoch": 0.029802221620157138,
+      "grad_norm": 0.0988590344786644,
+      "learning_rate": 0.0001,
+      "loss": 1.7931,
+      "step": 55
+    },
+    {
+      "epoch": 0.03034408019506909,
+      "grad_norm": 0.10976176708936691,
+      "learning_rate": 0.0001,
+      "loss": 1.8151,
+      "step": 56
+    },
+    {
+      "epoch": 0.030885938769981035,
+      "grad_norm": 0.09677836298942566,
+      "learning_rate": 0.0001,
+      "loss": 1.7983,
+      "step": 57
+    },
+    {
+      "epoch": 0.03142779734489298,
+      "grad_norm": 0.11221853643655777,
+      "learning_rate": 0.0001,
+      "loss": 1.7988,
+      "step": 58
+    },
+    {
+      "epoch": 0.03196965591980493,
+      "grad_norm": 0.09712743014097214,
+      "learning_rate": 0.0001,
+      "loss": 1.7875,
+      "step": 59
+    },
+    {
+      "epoch": 0.03251151449471688,
+      "grad_norm": 0.11139461398124695,
+      "learning_rate": 0.0001,
+      "loss": 1.6759,
+      "step": 60
+    },
+    {
+      "epoch": 0.03305337306962883,
+      "grad_norm": 0.10642059892416,
+      "learning_rate": 0.0001,
+      "loss": 1.7627,
+      "step": 61
+    },
+    {
+      "epoch": 0.03359523164454078,
+      "grad_norm": 0.1016763374209404,
+      "learning_rate": 0.0001,
+      "loss": 1.7372,
+      "step": 62
+    },
+    {
+      "epoch": 0.03413709021945272,
+      "grad_norm": 0.10269554704427719,
+      "learning_rate": 0.0001,
+      "loss": 1.7287,
+      "step": 63
+    },
+    {
+      "epoch": 0.03467894879436467,
+      "grad_norm": 0.1037546694278717,
+      "learning_rate": 0.0001,
+      "loss": 1.7356,
+      "step": 64
+    },
+    {
+      "epoch": 0.035220807369276616,
+      "grad_norm": 0.10047534108161926,
+      "learning_rate": 0.0001,
+      "loss": 1.7004,
+      "step": 65
+    },
+    {
+      "epoch": 0.03576266594418857,
+      "grad_norm": 0.09919369220733643,
+      "learning_rate": 0.0001,
+      "loss": 1.6939,
+      "step": 66
+    },
+    {
+      "epoch": 0.03630452451910052,
+      "grad_norm": 0.0938708484172821,
+      "learning_rate": 0.0001,
+      "loss": 1.6959,
+      "step": 67
+    },
+    {
+      "epoch": 0.03684638309401246,
+      "grad_norm": 0.10731608420610428,
+      "learning_rate": 0.0001,
+      "loss": 1.7493,
+      "step": 68
+    },
+    {
+      "epoch": 0.03738824166892441,
+      "grad_norm": 0.09051501005887985,
+      "learning_rate": 0.0001,
+      "loss": 1.6887,
+      "step": 69
+    },
+    {
+      "epoch": 0.03793010024383636,
+      "grad_norm": 0.10202842205762863,
+      "learning_rate": 0.0001,
+      "loss": 1.7613,
+      "step": 70
+    },
+    {
+      "epoch": 0.038471958818748306,
+      "grad_norm": 0.09244200587272644,
+      "learning_rate": 0.0001,
+      "loss": 1.7076,
+      "step": 71
+    },
+    {
+      "epoch": 0.039013817393660256,
+      "grad_norm": 0.09826190024614334,
+      "learning_rate": 0.0001,
+      "loss": 1.7634,
+      "step": 72
+    },
+    {
+      "epoch": 0.0395556759685722,
+      "grad_norm": 0.10156849771738052,
+      "learning_rate": 0.0001,
+      "loss": 1.7772,
+      "step": 73
+    },
+    {
+      "epoch": 0.04009753454348415,
+      "grad_norm": 0.10479609668254852,
+      "learning_rate": 0.0001,
+      "loss": 1.7278,
+      "step": 74
+    },
+    {
+      "epoch": 0.0406393931183961,
+      "grad_norm": 0.09295053780078888,
+      "learning_rate": 0.0001,
+      "loss": 1.7415,
+      "step": 75
+    },
+    {
+      "epoch": 0.041181251693308045,
+      "grad_norm": 0.10030414164066315,
+      "learning_rate": 0.0001,
+      "loss": 1.7147,
+      "step": 76
+    },
+    {
+      "epoch": 0.041723110268219996,
+      "grad_norm": 0.09261474758386612,
+      "learning_rate": 0.0001,
+      "loss": 1.6791,
+      "step": 77
+    },
+    {
+      "epoch": 0.04226496884313194,
+      "grad_norm": 0.09939026832580566,
+      "learning_rate": 0.0001,
+      "loss": 1.7681,
+      "step": 78
+    },
+    {
+      "epoch": 0.04280682741804389,
+      "grad_norm": 0.09494777768850327,
+      "learning_rate": 0.0001,
+      "loss": 1.7222,
+      "step": 79
+    },
+    {
+      "epoch": 0.04334868599295584,
+      "grad_norm": 0.0940719023346901,
+      "learning_rate": 0.0001,
+      "loss": 1.7207,
+      "step": 80
+    },
+    {
+      "epoch": 0.043890544567867784,
+      "grad_norm": 0.11140227317810059,
+      "learning_rate": 0.0001,
+      "loss": 1.7504,
+      "step": 81
+    },
+    {
+      "epoch": 0.044432403142779735,
+      "grad_norm": 0.08920968323945999,
+      "learning_rate": 0.0001,
+      "loss": 1.7073,
+      "step": 82
+    },
+    {
+      "epoch": 0.044974261717691685,
+      "grad_norm": 0.10845290124416351,
+      "learning_rate": 0.0001,
+      "loss": 1.7558,
+      "step": 83
+    },
+    {
+      "epoch": 0.04551612029260363,
+      "grad_norm": 0.08898979425430298,
+      "learning_rate": 0.0001,
+      "loss": 1.7082,
+      "step": 84
+    },
+    {
+      "epoch": 0.04605797886751558,
+      "grad_norm": 0.1022668331861496,
+      "learning_rate": 0.0001,
+      "loss": 1.7659,
+      "step": 85
+    },
+    {
+      "epoch": 0.04659983744242752,
+      "grad_norm": 0.09408645331859589,
+      "learning_rate": 0.0001,
+      "loss": 1.6786,
+      "step": 86
+    },
+    {
+      "epoch": 0.047141696017339474,
+      "grad_norm": 0.0999981164932251,
+      "learning_rate": 0.0001,
+      "loss": 1.7425,
+      "step": 87
+    },
+    {
+      "epoch": 0.047683554592251425,
+      "grad_norm": 0.1010698676109314,
+      "learning_rate": 0.0001,
+      "loss": 1.7288,
+      "step": 88
+    },
+    {
+      "epoch": 0.04822541316716337,
+      "grad_norm": 0.0986516997218132,
+      "learning_rate": 0.0001,
+      "loss": 1.7166,
+      "step": 89
+    },
+    {
+      "epoch": 0.04876727174207532,
+      "grad_norm": 0.11564943194389343,
+      "learning_rate": 0.0001,
+      "loss": 1.7131,
+      "step": 90
+    },
+    {
+      "epoch": 0.04930913031698727,
+      "grad_norm": 0.10187874734401703,
+      "learning_rate": 0.0001,
+      "loss": 1.7864,
+      "step": 91
+    },
+    {
+      "epoch": 0.04985098889189921,
+      "grad_norm": 0.11014661937952042,
+      "learning_rate": 0.0001,
+      "loss": 1.7228,
+      "step": 92
+    },
+    {
+      "epoch": 0.050392847466811164,
+      "grad_norm": 0.09390980005264282,
+      "learning_rate": 0.0001,
+      "loss": 1.6759,
+      "step": 93
+    },
+    {
+      "epoch": 0.05093470604172311,
+      "grad_norm": 0.10052553564310074,
+      "learning_rate": 0.0001,
+      "loss": 1.6606,
+      "step": 94
+    },
+    {
+      "epoch": 0.05147656461663506,
+      "grad_norm": 0.10119052976369858,
+      "learning_rate": 0.0001,
+      "loss": 1.7306,
+      "step": 95
+    },
+    {
+      "epoch": 0.05201842319154701,
+      "grad_norm": 0.10390573740005493,
+      "learning_rate": 0.0001,
+      "loss": 1.7791,
+      "step": 96
+    },
+    {
+      "epoch": 0.05256028176645895,
+      "grad_norm": 0.09522074460983276,
+      "learning_rate": 0.0001,
+      "loss": 1.7072,
+      "step": 97
+    },
+    {
+      "epoch": 0.0531021403413709,
+      "grad_norm": 0.11148355901241302,
+      "learning_rate": 0.0001,
+      "loss": 1.7544,
+      "step": 98
+    },
+    {
+      "epoch": 0.05364399891628285,
+      "grad_norm": 0.09624631702899933,
+      "learning_rate": 0.0001,
+      "loss": 1.7121,
+      "step": 99
+    },
+    {
+      "epoch": 0.0541858574911948,
+      "grad_norm": 0.09778071939945221,
+      "learning_rate": 0.0001,
+      "loss": 1.7695,
+      "step": 100
+    },
+    {
+      "epoch": 0.05472771606610675,
+      "grad_norm": 0.11497867852449417,
+      "learning_rate": 0.0001,
+      "loss": 1.7457,
+      "step": 101
+    },
+    {
+      "epoch": 0.05526957464101869,
+      "grad_norm": 0.09833558648824692,
+      "learning_rate": 0.0001,
+      "loss": 1.7046,
+      "step": 102
+    },
+    {
+      "epoch": 0.05581143321593064,
+      "grad_norm": 0.10873506963253021,
+      "learning_rate": 0.0001,
+      "loss": 1.7898,
+      "step": 103
+    },
+    {
+      "epoch": 0.05635329179084259,
+      "grad_norm": 0.10418610274791718,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 104
+    },
+    {
+      "epoch": 0.056895150365754536,
+      "grad_norm": 0.10068942606449127,
+      "learning_rate": 0.0001,
+      "loss": 1.6845,
+      "step": 105
+    },
+    {
+      "epoch": 0.05743700894066649,
+      "grad_norm": 0.11650339514017105,
+      "learning_rate": 0.0001,
+      "loss": 1.7883,
+      "step": 106
+    },
+    {
+      "epoch": 0.05797886751557844,
+      "grad_norm": 0.09679100662469864,
+      "learning_rate": 0.0001,
+      "loss": 1.7633,
+      "step": 107
+    },
+    {
+      "epoch": 0.05852072609049038,
+      "grad_norm": 0.10021305084228516,
+      "learning_rate": 0.0001,
+      "loss": 1.6978,
+      "step": 108
+    },
+    {
+      "epoch": 0.05906258466540233,
+      "grad_norm": 0.10400475561618805,
+      "learning_rate": 0.0001,
+      "loss": 1.8073,
+      "step": 109
+    },
+    {
+      "epoch": 0.059604443240314275,
+      "grad_norm": 0.11318189650774002,
+      "learning_rate": 0.0001,
+      "loss": 1.7629,
+      "step": 110
+    },
+    {
+      "epoch": 0.060146301815226226,
+      "grad_norm": 0.09900004416704178,
+      "learning_rate": 0.0001,
+      "loss": 1.6763,
+      "step": 111
+    },
+    {
+      "epoch": 0.06068816039013818,
+      "grad_norm": 0.09888526797294617,
+      "learning_rate": 0.0001,
+      "loss": 1.7403,
+      "step": 112
+    },
+    {
+      "epoch": 0.06123001896505012,
+      "grad_norm": 0.09070870280265808,
+      "learning_rate": 0.0001,
+      "loss": 1.6324,
+      "step": 113
+    },
+    {
+      "epoch": 0.06177187753996207,
+      "grad_norm": 0.10428040474653244,
+      "learning_rate": 0.0001,
+      "loss": 1.7713,
+      "step": 114
+    },
+    {
+      "epoch": 0.062313736114874015,
+      "grad_norm": 0.09860703349113464,
+      "learning_rate": 0.0001,
+      "loss": 1.6457,
+      "step": 115
+    },
+    {
+      "epoch": 0.06285559468978597,
+      "grad_norm": 0.09450509399175644,
+      "learning_rate": 0.0001,
+      "loss": 1.7076,
+      "step": 116
+    },
+    {
+      "epoch": 0.06339745326469791,
+      "grad_norm": 0.09789243340492249,
+      "learning_rate": 0.0001,
+      "loss": 1.705,
+      "step": 117
+    },
+    {
+      "epoch": 0.06393931183960987,
+      "grad_norm": 0.1094205304980278,
+      "learning_rate": 0.0001,
+      "loss": 1.7197,
+      "step": 118
+    },
+    {
+      "epoch": 0.06448117041452181,
+      "grad_norm": 0.09620770812034607,
+      "learning_rate": 0.0001,
+      "loss": 1.7808,
+      "step": 119
+    },
+    {
+      "epoch": 0.06502302898943375,
+      "grad_norm": 0.10695544630289078,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 120
+    },
+    {
+      "epoch": 0.06556488756434571,
+      "grad_norm": 0.09900528192520142,
+      "learning_rate": 0.0001,
+      "loss": 1.7043,
+      "step": 121
+    },
+    {
+      "epoch": 0.06610674613925765,
+      "grad_norm": 0.10099681466817856,
+      "learning_rate": 0.0001,
+      "loss": 1.7315,
+      "step": 122
+    },
+    {
+      "epoch": 0.0666486047141696,
+      "grad_norm": 0.09247930347919464,
+      "learning_rate": 0.0001,
+      "loss": 1.6921,
+      "step": 123
+    },
+    {
+      "epoch": 0.06719046328908156,
+      "grad_norm": 0.1142578274011612,
+      "learning_rate": 0.0001,
+      "loss": 1.7487,
+      "step": 124
+    },
+    {
+      "epoch": 0.0677323218639935,
+      "grad_norm": 0.12242250144481659,
+      "learning_rate": 0.0001,
+      "loss": 1.7126,
+      "step": 125
+    },
+    {
+      "epoch": 0.06827418043890544,
+      "grad_norm": 0.10806268453598022,
+      "learning_rate": 0.0001,
+      "loss": 1.732,
+      "step": 126
+    },
+    {
+      "epoch": 0.06881603901381739,
+      "grad_norm": 0.09226547926664352,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 127
+    },
+    {
+      "epoch": 0.06935789758872934,
+      "grad_norm": 0.09697569161653519,
+      "learning_rate": 0.0001,
+      "loss": 1.6693,
+      "step": 128
+    },
+    {
+      "epoch": 0.06989975616364129,
+      "grad_norm": 0.10045196861028671,
+      "learning_rate": 0.0001,
+      "loss": 1.7323,
+      "step": 129
+    },
+    {
+      "epoch": 0.07044161473855323,
+      "grad_norm": 0.1046331375837326,
+      "learning_rate": 0.0001,
+      "loss": 1.7106,
+      "step": 130
+    },
+    {
+      "epoch": 0.07098347331346519,
+      "grad_norm": 0.10865035653114319,
+      "learning_rate": 0.0001,
+      "loss": 1.7459,
+      "step": 131
+    },
+    {
+      "epoch": 0.07152533188837713,
+      "grad_norm": 0.1041274145245552,
+      "learning_rate": 0.0001,
+      "loss": 1.776,
+      "step": 132
+    },
+    {
+      "epoch": 0.07206719046328908,
+      "grad_norm": 0.10414282977581024,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 133
+    },
+    {
+      "epoch": 0.07260904903820103,
+      "grad_norm": 0.12221338599920273,
+      "learning_rate": 0.0001,
+      "loss": 1.6995,
+      "step": 134
+    },
+    {
+      "epoch": 0.07315090761311298,
+      "grad_norm": 0.10498306155204773,
+      "learning_rate": 0.0001,
+      "loss": 1.7143,
+      "step": 135
+    },
+    {
+      "epoch": 0.07369276618802492,
+      "grad_norm": 0.13094539940357208,
+      "learning_rate": 0.0001,
+      "loss": 1.7102,
+      "step": 136
+    },
+    {
+      "epoch": 0.07423462476293688,
+      "grad_norm": 0.120379239320755,
+      "learning_rate": 0.0001,
+      "loss": 1.6664,
+      "step": 137
+    },
+    {
+      "epoch": 0.07477648333784882,
+      "grad_norm": 0.12071491032838821,
+      "learning_rate": 0.0001,
+      "loss": 1.7722,
+      "step": 138
+    },
+    {
+      "epoch": 0.07531834191276077,
+      "grad_norm": 0.1514299064874649,
+      "learning_rate": 0.0001,
+      "loss": 1.6833,
+      "step": 139
+    },
+    {
+      "epoch": 0.07586020048767272,
+      "grad_norm": 0.10760881751775742,
+      "learning_rate": 0.0001,
+      "loss": 1.6733,
+      "step": 140
+    },
+    {
+      "epoch": 0.07640205906258467,
+      "grad_norm": 0.10171575099229813,
+      "learning_rate": 0.0001,
+      "loss": 1.6455,
+      "step": 141
+    },
+    {
+      "epoch": 0.07694391763749661,
+      "grad_norm": 0.11112318933010101,
+      "learning_rate": 0.0001,
+      "loss": 1.6997,
+      "step": 142
+    },
+    {
+      "epoch": 0.07748577621240856,
+      "grad_norm": 0.10578908771276474,
+      "learning_rate": 0.0001,
+      "loss": 1.7001,
+      "step": 143
+    },
+    {
+      "epoch": 0.07802763478732051,
+      "grad_norm": 0.10570411384105682,
+      "learning_rate": 0.0001,
+      "loss": 1.7551,
+      "step": 144
+    },
+    {
+      "epoch": 0.07856949336223246,
+      "grad_norm": 0.09980179369449615,
+      "learning_rate": 0.0001,
+      "loss": 1.6345,
+      "step": 145
+    },
+    {
+      "epoch": 0.0791113519371444,
+      "grad_norm": 0.10519526153802872,
+      "learning_rate": 0.0001,
+      "loss": 1.6707,
+      "step": 146
+    },
+    {
+      "epoch": 0.07965321051205636,
+      "grad_norm": 0.11583776772022247,
+      "learning_rate": 0.0001,
+      "loss": 1.7332,
+      "step": 147
+    },
+    {
+      "epoch": 0.0801950690869683,
+      "grad_norm": 0.10443723201751709,
+      "learning_rate": 0.0001,
+      "loss": 1.7332,
+      "step": 148
+    },
+    {
+      "epoch": 0.08073692766188025,
+      "grad_norm": 0.11888331174850464,
+      "learning_rate": 0.0001,
+      "loss": 1.7212,
+      "step": 149
+    },
+    {
+      "epoch": 0.0812787862367922,
+      "grad_norm": 0.0985705554485321,
+      "learning_rate": 0.0001,
+      "loss": 1.7169,
+      "step": 150
+    },
+    {
+      "epoch": 0.08182064481170415,
+      "grad_norm": 0.11832987517118454,
+      "learning_rate": 0.0001,
+      "loss": 1.6792,
+      "step": 151
+    },
+    {
+      "epoch": 0.08236250338661609,
+      "grad_norm": 0.10486771911382675,
+      "learning_rate": 0.0001,
+      "loss": 1.6463,
+      "step": 152
+    },
+    {
+      "epoch": 0.08290436196152805,
+      "grad_norm": 0.12332990765571594,
+      "learning_rate": 0.0001,
+      "loss": 1.723,
+      "step": 153
+    },
+    {
+      "epoch": 0.08344622053643999,
+      "grad_norm": 0.1274755448102951,
+      "learning_rate": 0.0001,
+      "loss": 1.7229,
+      "step": 154
+    },
+    {
+      "epoch": 0.08398807911135193,
+      "grad_norm": 0.10862415283918381,
+      "learning_rate": 0.0001,
+      "loss": 1.7341,
+      "step": 155
+    },
+    {
+      "epoch": 0.08452993768626388,
+      "grad_norm": 0.12293940037488937,
+      "learning_rate": 0.0001,
+      "loss": 1.7031,
+      "step": 156
+    },
+    {
+      "epoch": 0.08507179626117584,
+      "grad_norm": 0.11512468010187149,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 157
+    },
+    {
+      "epoch": 0.08561365483608778,
+      "grad_norm": 0.10460253059864044,
+      "learning_rate": 0.0001,
+      "loss": 1.6905,
+      "step": 158
+    },
+    {
+      "epoch": 0.08615551341099972,
+      "grad_norm": 0.12816384434700012,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 159
+    },
+    {
+      "epoch": 0.08669737198591168,
+      "grad_norm": 0.09755271673202515,
+      "learning_rate": 0.0001,
+      "loss": 1.6231,
+      "step": 160
+    },
+    {
+      "epoch": 0.08723923056082362,
+      "grad_norm": 0.1255766898393631,
+      "learning_rate": 0.0001,
+      "loss": 1.7978,
+      "step": 161
+    },
+    {
+      "epoch": 0.08778108913573557,
+      "grad_norm": 0.10234647989273071,
+      "learning_rate": 0.0001,
+      "loss": 1.6583,
+      "step": 162
+    },
+    {
+      "epoch": 0.08832294771064753,
+      "grad_norm": 0.1072743609547615,
+      "learning_rate": 0.0001,
+      "loss": 1.7496,
+      "step": 163
+    },
+    {
+      "epoch": 0.08886480628555947,
+      "grad_norm": 0.09563694894313812,
+      "learning_rate": 0.0001,
+      "loss": 1.7028,
+      "step": 164
+    },
+    {
+      "epoch": 0.08940666486047141,
+      "grad_norm": 0.10931570082902908,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 165
+    },
+    {
+      "epoch": 0.08994852343538337,
+      "grad_norm": 0.0974336788058281,
+      "learning_rate": 0.0001,
+      "loss": 1.7089,
+      "step": 166
+    },
+    {
+      "epoch": 0.09049038201029531,
+      "grad_norm": 0.11343063414096832,
+      "learning_rate": 0.0001,
+      "loss": 1.7259,
+      "step": 167
+    },
+    {
+      "epoch": 0.09103224058520726,
+      "grad_norm": 0.09566744416952133,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 168
+    },
+    {
+      "epoch": 0.09157409916011922,
+      "grad_norm": 0.10925479978322983,
+      "learning_rate": 0.0001,
+      "loss": 1.7122,
+      "step": 169
+    },
+    {
+      "epoch": 0.09211595773503116,
+      "grad_norm": 0.1133953258395195,
+      "learning_rate": 0.0001,
+      "loss": 1.731,
+      "step": 170
+    },
+    {
+      "epoch": 0.0926578163099431,
+      "grad_norm": 0.10159795731306076,
+      "learning_rate": 0.0001,
+      "loss": 1.7248,
+      "step": 171
+    },
+    {
+      "epoch": 0.09319967488485505,
+      "grad_norm": 0.09628516435623169,
+      "learning_rate": 0.0001,
+      "loss": 1.6852,
+      "step": 172
+    },
+    {
+      "epoch": 0.093741533459767,
+      "grad_norm": 0.09967010468244553,
+      "learning_rate": 0.0001,
+      "loss": 1.7118,
+      "step": 173
+    },
+    {
+      "epoch": 0.09428339203467895,
+      "grad_norm": 0.09833619743585587,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 174
+    },
+    {
+      "epoch": 0.09482525060959089,
+      "grad_norm": 0.10003640502691269,
+      "learning_rate": 0.0001,
+      "loss": 1.6846,
+      "step": 175
+    },
+    {
+      "epoch": 0.09536710918450285,
+      "grad_norm": 0.09707822650671005,
+      "learning_rate": 0.0001,
+      "loss": 1.6911,
+      "step": 176
+    },
+    {
+      "epoch": 0.09590896775941479,
+      "grad_norm": 0.09149082750082016,
+      "learning_rate": 0.0001,
+      "loss": 1.6732,
+      "step": 177
+    },
+    {
+      "epoch": 0.09645082633432674,
+      "grad_norm": 0.11072653532028198,
+      "learning_rate": 0.0001,
+      "loss": 1.7419,
+      "step": 178
+    },
+    {
+      "epoch": 0.0969926849092387,
+      "grad_norm": 0.09115342795848846,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 179
+    },
+    {
+      "epoch": 0.09753454348415064,
+      "grad_norm": 0.0993184745311737,
+      "learning_rate": 0.0001,
+      "loss": 1.5778,
+      "step": 180
+    },
+    {
+      "epoch": 0.09807640205906258,
+      "grad_norm": 0.09827783703804016,
+      "learning_rate": 0.0001,
+      "loss": 1.6285,
+      "step": 181
+    },
+    {
+      "epoch": 0.09861826063397454,
+      "grad_norm": 0.1036309152841568,
+      "learning_rate": 0.0001,
+      "loss": 1.6966,
+      "step": 182
+    },
+    {
+      "epoch": 0.09916011920888648,
+      "grad_norm": 0.10436933487653732,
+      "learning_rate": 0.0001,
+      "loss": 1.6565,
+      "step": 183
+    },
+    {
+      "epoch": 0.09970197778379843,
+      "grad_norm": 0.10254193842411041,
+      "learning_rate": 0.0001,
+      "loss": 1.776,
+      "step": 184
+    },
+    {
+      "epoch": 0.10024383635871037,
+      "grad_norm": 0.10822900384664536,
+      "learning_rate": 0.0001,
+      "loss": 1.6572,
+      "step": 185
+    },
+    {
+      "epoch": 0.10078569493362233,
+      "grad_norm": 0.1024482473731041,
+      "learning_rate": 0.0001,
+      "loss": 1.7474,
+      "step": 186
+    },
+    {
+      "epoch": 0.10132755350853427,
+      "grad_norm": 0.10238222032785416,
+      "learning_rate": 0.0001,
+      "loss": 1.7035,
+      "step": 187
+    },
+    {
+      "epoch": 0.10186941208344621,
+      "grad_norm": 0.1048850268125534,
+      "learning_rate": 0.0001,
+      "loss": 1.6729,
+      "step": 188
+    },
+    {
+      "epoch": 0.10241127065835817,
+      "grad_norm": 0.10192476212978363,
+      "learning_rate": 0.0001,
+      "loss": 1.6719,
+      "step": 189
+    },
+    {
+      "epoch": 0.10295312923327012,
+      "grad_norm": 0.10868800431489944,
+      "learning_rate": 0.0001,
+      "loss": 1.7563,
+      "step": 190
+    },
+    {
+      "epoch": 0.10349498780818206,
+      "grad_norm": 0.10962677747011185,
+      "learning_rate": 0.0001,
+      "loss": 1.6806,
+      "step": 191
+    },
+    {
+      "epoch": 0.10403684638309402,
+      "grad_norm": 0.09616486728191376,
+      "learning_rate": 0.0001,
+      "loss": 1.7144,
+      "step": 192
+    },
+    {
+      "epoch": 0.10457870495800596,
+      "grad_norm": 0.1281924992799759,
+      "learning_rate": 0.0001,
+      "loss": 1.6827,
+      "step": 193
+    },
+    {
+      "epoch": 0.1051205635329179,
+      "grad_norm": 0.11889411509037018,
+      "learning_rate": 0.0001,
+      "loss": 1.7362,
+      "step": 194
+    },
+    {
+      "epoch": 0.10566242210782986,
+      "grad_norm": 0.1125204935669899,
+      "learning_rate": 0.0001,
+      "loss": 1.7325,
+      "step": 195
+    },
+    {
+      "epoch": 0.1062042806827418,
+      "grad_norm": 0.12667258083820343,
+      "learning_rate": 0.0001,
+      "loss": 1.7177,
+      "step": 196
+    },
+    {
+      "epoch": 0.10674613925765375,
+      "grad_norm": 0.09924928098917007,
+      "learning_rate": 0.0001,
+      "loss": 1.6907,
+      "step": 197
+    },
+    {
+      "epoch": 0.1072879978325657,
+      "grad_norm": 0.11108226329088211,
+      "learning_rate": 0.0001,
+      "loss": 1.6793,
+      "step": 198
+    },
+    {
+      "epoch": 0.10782985640747765,
+      "grad_norm": 0.09894610941410065,
+      "learning_rate": 0.0001,
+      "loss": 1.7,
+      "step": 199
+    },
+    {
+      "epoch": 0.1083717149823896,
+      "grad_norm": 0.12750208377838135,
+      "learning_rate": 0.0001,
+      "loss": 1.649,
+      "step": 200
+    },
+    {
+      "epoch": 0.10891357355730154,
+      "grad_norm": 0.10455714166164398,
+      "learning_rate": 0.0001,
+      "loss": 1.7068,
+      "step": 201
+    },
+    {
+      "epoch": 0.1094554321322135,
+      "grad_norm": 0.1203424334526062,
+      "learning_rate": 0.0001,
+      "loss": 1.7194,
+      "step": 202
+    },
+    {
+      "epoch": 0.10999729070712544,
+      "grad_norm": 0.11161962151527405,
+      "learning_rate": 0.0001,
+      "loss": 1.7124,
+      "step": 203
+    },
+    {
+      "epoch": 0.11053914928203738,
+      "grad_norm": 0.11139895021915436,
+      "learning_rate": 0.0001,
+      "loss": 1.642,
+      "step": 204
+    },
+    {
+      "epoch": 0.11108100785694934,
+      "grad_norm": 0.14790917932987213,
+      "learning_rate": 0.0001,
+      "loss": 1.699,
+      "step": 205
+    },
+    {
+      "epoch": 0.11162286643186128,
+      "grad_norm": 0.1384461224079132,
+      "learning_rate": 0.0001,
+      "loss": 1.719,
+      "step": 206
+    },
+    {
+      "epoch": 0.11216472500677323,
+      "grad_norm": 0.11863648891448975,
+      "learning_rate": 0.0001,
+      "loss": 1.7007,
+      "step": 207
+    },
+    {
+      "epoch": 0.11270658358168519,
+      "grad_norm": 0.1461704522371292,
+      "learning_rate": 0.0001,
+      "loss": 1.6736,
+      "step": 208
+    },
+    {
+      "epoch": 0.11324844215659713,
+      "grad_norm": 0.11320877820253372,
+      "learning_rate": 0.0001,
+      "loss": 1.7092,
+      "step": 209
+    },
+    {
+      "epoch": 0.11379030073150907,
+      "grad_norm": 0.1436343491077423,
+      "learning_rate": 0.0001,
+      "loss": 1.6807,
+      "step": 210
+    },
+    {
+      "epoch": 0.11433215930642103,
+      "grad_norm": 0.12915484607219696,
+      "learning_rate": 0.0001,
+      "loss": 1.7741,
+      "step": 211
+    },
+    {
+      "epoch": 0.11487401788133297,
+      "grad_norm": 0.16453833878040314,
+      "learning_rate": 0.0001,
+      "loss": 1.7416,
+      "step": 212
+    },
+    {
+      "epoch": 0.11541587645624492,
+      "grad_norm": 0.10235466808080673,
+      "learning_rate": 0.0001,
+      "loss": 1.6421,
+      "step": 213
+    },
+    {
+      "epoch": 0.11595773503115687,
+      "grad_norm": 0.13166211545467377,
+      "learning_rate": 0.0001,
+      "loss": 1.7094,
+      "step": 214
+    },
+    {
+      "epoch": 0.11649959360606882,
+      "grad_norm": 0.1139000654220581,
+      "learning_rate": 0.0001,
+      "loss": 1.6567,
+      "step": 215
+    },
+    {
+      "epoch": 0.11704145218098076,
+      "grad_norm": 0.1456412523984909,
+      "learning_rate": 0.0001,
+      "loss": 1.6858,
+      "step": 216
+    },
+    {
+      "epoch": 0.1175833107558927,
+      "grad_norm": 0.10928891599178314,
+      "learning_rate": 0.0001,
+      "loss": 1.7381,
+      "step": 217
+    },
+    {
+      "epoch": 0.11812516933080466,
+      "grad_norm": 0.11825280636548996,
+      "learning_rate": 0.0001,
+      "loss": 1.5774,
+      "step": 218
+    },
+    {
+      "epoch": 0.11866702790571661,
+      "grad_norm": 0.12118617445230484,
+      "learning_rate": 0.0001,
+      "loss": 1.7545,
+      "step": 219
+    },
+    {
+      "epoch": 0.11920888648062855,
+      "grad_norm": 0.13101591169834137,
+      "learning_rate": 0.0001,
+      "loss": 1.7525,
+      "step": 220
+    },
+    {
+      "epoch": 0.11975074505554051,
+      "grad_norm": 0.09695487469434738,
+      "learning_rate": 0.0001,
+      "loss": 1.6776,
+      "step": 221
+    },
+    {
+      "epoch": 0.12029260363045245,
+      "grad_norm": 0.11916449666023254,
+      "learning_rate": 0.0001,
+      "loss": 1.7396,
+      "step": 222
+    },
+    {
+      "epoch": 0.1208344622053644,
+      "grad_norm": 0.10929539054632187,
+      "learning_rate": 0.0001,
+      "loss": 1.77,
+      "step": 223
+    },
+    {
+      "epoch": 0.12137632078027635,
+      "grad_norm": 0.10067623108625412,
+      "learning_rate": 0.0001,
+      "loss": 1.6743,
+      "step": 224
+    },
+    {
+      "epoch": 0.1219181793551883,
+      "grad_norm": 0.12053167074918747,
+      "learning_rate": 0.0001,
+      "loss": 1.6709,
+      "step": 225
+    },
+    {
+      "epoch": 0.12246003793010024,
+      "grad_norm": 0.12418187409639359,
+      "learning_rate": 0.0001,
+      "loss": 1.7093,
+      "step": 226
+    },
+    {
+      "epoch": 0.1230018965050122,
+      "grad_norm": 0.1266579031944275,
+      "learning_rate": 0.0001,
+      "loss": 1.6896,
+      "step": 227
+    },
+    {
+      "epoch": 0.12354375507992414,
+      "grad_norm": 0.11025568097829819,
+      "learning_rate": 0.0001,
+      "loss": 1.6628,
+      "step": 228
+    },
+    {
+      "epoch": 0.12408561365483609,
+      "grad_norm": 0.11573028564453125,
+      "learning_rate": 0.0001,
+      "loss": 1.6605,
+      "step": 229
+    },
+    {
+      "epoch": 0.12462747222974803,
+      "grad_norm": 0.11125994473695755,
+      "learning_rate": 0.0001,
+      "loss": 1.6752,
+      "step": 230
+    },
+    {
+      "epoch": 0.12516933080465997,
+      "grad_norm": 0.1128525361418724,
+      "learning_rate": 0.0001,
+      "loss": 1.6262,
+      "step": 231
+    },
+    {
+      "epoch": 0.12571118937957193,
+      "grad_norm": 0.09718891233205795,
+      "learning_rate": 0.0001,
+      "loss": 1.6745,
+      "step": 232
+    },
+    {
+      "epoch": 0.1262530479544839,
+      "grad_norm": 0.1529717892408371,
+      "learning_rate": 0.0001,
+      "loss": 1.7352,
+      "step": 233
+    },
+    {
+      "epoch": 0.12679490652939582,
+      "grad_norm": 0.0977558046579361,
+      "learning_rate": 0.0001,
+      "loss": 1.7009,
+      "step": 234
+    },
+    {
+      "epoch": 0.12733676510430778,
+      "grad_norm": 0.14934347569942474,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 235
+    },
+    {
+      "epoch": 0.12787862367921973,
+      "grad_norm": 0.09760584682226181,
+      "learning_rate": 0.0001,
+      "loss": 1.6608,
+      "step": 236
+    },
+    {
+      "epoch": 0.12842048225413166,
+      "grad_norm": 0.11174532771110535,
+      "learning_rate": 0.0001,
+      "loss": 1.739,
+      "step": 237
+    },
+    {
+      "epoch": 0.12896234082904362,
+      "grad_norm": 0.11080338060855865,
+      "learning_rate": 0.0001,
+      "loss": 1.7127,
+      "step": 238
+    },
+    {
+      "epoch": 0.12950419940395558,
+      "grad_norm": 0.10834498703479767,
+      "learning_rate": 0.0001,
+      "loss": 1.6654,
+      "step": 239
+    },
+    {
+      "epoch": 0.1300460579788675,
+      "grad_norm": 0.11970813572406769,
+      "learning_rate": 0.0001,
+      "loss": 1.7051,
+      "step": 240
+    },
+    {
+      "epoch": 0.13058791655377947,
+      "grad_norm": 0.11999791115522385,
+      "learning_rate": 0.0001,
+      "loss": 1.6949,
+      "step": 241
+    },
+    {
+      "epoch": 0.13112977512869142,
+      "grad_norm": 0.10226437449455261,
+      "learning_rate": 0.0001,
+      "loss": 1.6928,
+      "step": 242
+    },
+    {
+      "epoch": 0.13167163370360335,
+      "grad_norm": 0.11835068464279175,
+      "learning_rate": 0.0001,
+      "loss": 1.6934,
+      "step": 243
+    },
+    {
+      "epoch": 0.1322134922785153,
+      "grad_norm": 0.10222174972295761,
+      "learning_rate": 0.0001,
+      "loss": 1.7147,
+      "step": 244
+    },
+    {
+      "epoch": 0.13275535085342727,
+      "grad_norm": 0.09475545585155487,
+      "learning_rate": 0.0001,
+      "loss": 1.6798,
+      "step": 245
+    },
+    {
+      "epoch": 0.1332972094283392,
+      "grad_norm": 0.1266164779663086,
+      "learning_rate": 0.0001,
+      "loss": 1.7424,
+      "step": 246
+    },
+    {
+      "epoch": 0.13383906800325115,
+      "grad_norm": 0.10481266677379608,
+      "learning_rate": 0.0001,
+      "loss": 1.6616,
+      "step": 247
+    },
+    {
+      "epoch": 0.1343809265781631,
+      "grad_norm": 0.10396500676870346,
+      "learning_rate": 0.0001,
+      "loss": 1.7314,
+      "step": 248
+    },
+    {
+      "epoch": 0.13492278515307504,
+      "grad_norm": 0.14904353022575378,
+      "learning_rate": 0.0001,
+      "loss": 1.6938,
+      "step": 249
+    },
+    {
+      "epoch": 0.135464643727987,
+      "grad_norm": 0.0941733717918396,
+      "learning_rate": 0.0001,
+      "loss": 1.6723,
+      "step": 250
+    },
+    {
+      "epoch": 0.13600650230289893,
+      "grad_norm": 0.15529394149780273,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 251
+    },
+    {
+      "epoch": 0.1365483608778109,
+      "grad_norm": 0.12021831423044205,
+      "learning_rate": 0.0001,
+      "loss": 1.721,
+      "step": 252
+    },
+    {
+      "epoch": 0.13709021945272284,
+      "grad_norm": 0.15852448344230652,
+      "learning_rate": 0.0001,
+      "loss": 1.7294,
+      "step": 253
+    },
+    {
+      "epoch": 0.13763207802763477,
+      "grad_norm": 0.11266610771417618,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 254
+    },
+    {
+      "epoch": 0.13817393660254673,
+      "grad_norm": 0.13919475674629211,
+      "learning_rate": 0.0001,
+      "loss": 1.6603,
+      "step": 255
+    },
+    {
+      "epoch": 0.1387157951774587,
+      "grad_norm": 0.11103470623493195,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 256
+    },
+    {
+      "epoch": 0.13925765375237062,
+      "grad_norm": 0.15118271112442017,
+      "learning_rate": 0.0001,
+      "loss": 1.7035,
+      "step": 257
+    },
+    {
+      "epoch": 0.13979951232728258,
+      "grad_norm": 0.11419171839952469,
+      "learning_rate": 0.0001,
+      "loss": 1.6692,
+      "step": 258
+    },
+    {
+      "epoch": 0.14034137090219453,
+      "grad_norm": 0.11274544894695282,
+      "learning_rate": 0.0001,
+      "loss": 1.5942,
+      "step": 259
+    },
+    {
+      "epoch": 0.14088322947710646,
+      "grad_norm": 0.14849093556404114,
+      "learning_rate": 0.0001,
+      "loss": 1.7188,
+      "step": 260
+    },
+    {
+      "epoch": 0.14142508805201842,
+      "grad_norm": 0.10871066898107529,
+      "learning_rate": 0.0001,
+      "loss": 1.7819,
+      "step": 261
+    },
+    {
+      "epoch": 0.14196694662693038,
+      "grad_norm": 0.13676348328590393,
+      "learning_rate": 0.0001,
+      "loss": 1.6057,
+      "step": 262
+    },
+    {
+      "epoch": 0.1425088052018423,
+      "grad_norm": 0.12139032036066055,
+      "learning_rate": 0.0001,
+      "loss": 1.6857,
+      "step": 263
+    },
+    {
+      "epoch": 0.14305066377675427,
+      "grad_norm": 0.11282926797866821,
+      "learning_rate": 0.0001,
+      "loss": 1.6851,
+      "step": 264
+    },
+    {
+      "epoch": 0.14359252235166622,
+      "grad_norm": 0.12527044117450714,
+      "learning_rate": 0.0001,
+      "loss": 1.7553,
+      "step": 265
+    },
+    {
+      "epoch": 0.14413438092657815,
+      "grad_norm": 0.14149242639541626,
+      "learning_rate": 0.0001,
+      "loss": 1.6815,
+      "step": 266
+    },
+    {
+      "epoch": 0.1446762395014901,
+      "grad_norm": 0.10425418615341187,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 267
+    },
+    {
+      "epoch": 0.14521809807640207,
+      "grad_norm": 0.13150091469287872,
+      "learning_rate": 0.0001,
+      "loss": 1.7274,
+      "step": 268
+    },
+    {
+      "epoch": 0.145759956651314,
+      "grad_norm": 0.09711414575576782,
+      "learning_rate": 0.0001,
+      "loss": 1.7004,
+      "step": 269
+    },
+    {
+      "epoch": 0.14630181522622596,
+      "grad_norm": 0.1150781586766243,
+      "learning_rate": 0.0001,
+      "loss": 1.6948,
+      "step": 270
+    },
+    {
+      "epoch": 0.14684367380113791,
+      "grad_norm": 0.11720975488424301,
+      "learning_rate": 0.0001,
+      "loss": 1.6947,
+      "step": 271
+    },
+    {
+      "epoch": 0.14738553237604984,
+      "grad_norm": 0.10064122825860977,
+      "learning_rate": 0.0001,
+      "loss": 1.7101,
+      "step": 272
+    },
+    {
+      "epoch": 0.1479273909509618,
+      "grad_norm": 0.11132777482271194,
+      "learning_rate": 0.0001,
+      "loss": 1.6706,
+      "step": 273
+    },
+    {
+      "epoch": 0.14846924952587376,
+      "grad_norm": 0.10732540488243103,
+      "learning_rate": 0.0001,
+      "loss": 1.655,
+      "step": 274
+    },
+    {
+      "epoch": 0.1490111081007857,
+      "grad_norm": 0.11525800824165344,
+      "learning_rate": 0.0001,
+      "loss": 1.6958,
+      "step": 275
+    },
+    {
+      "epoch": 0.14955296667569765,
+      "grad_norm": 0.1057300716638565,
+      "learning_rate": 0.0001,
+      "loss": 1.6936,
+      "step": 276
+    },
+    {
+      "epoch": 0.1500948252506096,
+      "grad_norm": 0.10139822959899902,
+      "learning_rate": 0.0001,
+      "loss": 1.7037,
+      "step": 277
+    },
+    {
+      "epoch": 0.15063668382552153,
+      "grad_norm": 0.10546170920133591,
+      "learning_rate": 0.0001,
+      "loss": 1.6824,
+      "step": 278
+    },
+    {
+      "epoch": 0.1511785424004335,
+      "grad_norm": 0.0943392738699913,
+      "learning_rate": 0.0001,
+      "loss": 1.6404,
+      "step": 279
+    },
+    {
+      "epoch": 0.15172040097534545,
+      "grad_norm": 0.10627435147762299,
+      "learning_rate": 0.0001,
+      "loss": 1.7194,
+      "step": 280
+    },
+    {
+      "epoch": 0.15226225955025738,
+      "grad_norm": 0.09746234118938446,
+      "learning_rate": 0.0001,
+      "loss": 1.6744,
+      "step": 281
+    },
+    {
+      "epoch": 0.15280411812516934,
+      "grad_norm": 0.09403081238269806,
+      "learning_rate": 0.0001,
+      "loss": 1.6677,
+      "step": 282
+    },
+    {
+      "epoch": 0.15334597670008127,
+      "grad_norm": 0.09880942851305008,
+      "learning_rate": 0.0001,
+      "loss": 1.6597,
+      "step": 283
+    },
+    {
+      "epoch": 0.15388783527499322,
+      "grad_norm": 0.10196152329444885,
+      "learning_rate": 0.0001,
+      "loss": 1.6484,
+      "step": 284
+    },
+    {
+      "epoch": 0.15442969384990518,
+      "grad_norm": 0.09791263937950134,
+      "learning_rate": 0.0001,
+      "loss": 1.7203,
+      "step": 285
+    },
+    {
+      "epoch": 0.1549715524248171,
+      "grad_norm": 0.09796151518821716,
+      "learning_rate": 0.0001,
+      "loss": 1.6459,
+      "step": 286
+    },
+    {
+      "epoch": 0.15551341099972907,
+      "grad_norm": 0.10709752887487411,
+      "learning_rate": 0.0001,
+      "loss": 1.7019,
+      "step": 287
+    },
+    {
+      "epoch": 0.15605526957464103,
+      "grad_norm": 0.09797371178865433,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 288
+    },
+    {
+      "epoch": 0.15659712814955296,
+      "grad_norm": 0.10401292890310287,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 289
+    },
+    {
+      "epoch": 0.1571389867244649,
+      "grad_norm": 0.09455517679452896,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 290
+    },
+    {
+      "epoch": 0.15768084529937687,
+      "grad_norm": 0.09385957568883896,
+      "learning_rate": 0.0001,
+      "loss": 1.6095,
+      "step": 291
+    },
+    {
+      "epoch": 0.1582227038742888,
+      "grad_norm": 0.10476952791213989,
+      "learning_rate": 0.0001,
+      "loss": 1.6574,
+      "step": 292
+    },
+    {
+      "epoch": 0.15876456244920076,
+      "grad_norm": 0.1275656670331955,
+      "learning_rate": 0.0001,
+      "loss": 1.6689,
+      "step": 293
+    },
+    {
+      "epoch": 0.15930642102411272,
+      "grad_norm": 0.0954967811703682,
+      "learning_rate": 0.0001,
+      "loss": 1.708,
+      "step": 294
+    },
+    {
+      "epoch": 0.15984827959902465,
+      "grad_norm": 0.12397397309541702,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 295
+    },
+    {
+      "epoch": 0.1603901381739366,
+      "grad_norm": 0.09896747022867203,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 296
+    },
+    {
+      "epoch": 0.16093199674884856,
+      "grad_norm": 0.13791395723819733,
+      "learning_rate": 0.0001,
+      "loss": 1.7306,
+      "step": 297
+    },
+    {
+      "epoch": 0.1614738553237605,
+      "grad_norm": 0.09542802721261978,
+      "learning_rate": 0.0001,
+      "loss": 1.6463,
+      "step": 298
+    },
+    {
+      "epoch": 0.16201571389867245,
+      "grad_norm": 0.11375903338193893,
+      "learning_rate": 0.0001,
+      "loss": 1.7341,
+      "step": 299
+    },
+    {
+      "epoch": 0.1625575724735844,
+      "grad_norm": 0.10889938473701477,
+      "learning_rate": 0.0001,
+      "loss": 1.731,
+      "step": 300
+    },
+    {
+      "epoch": 0.16309943104849633,
+      "grad_norm": 0.11272832006216049,
+      "learning_rate": 0.0001,
+      "loss": 1.6999,
+      "step": 301
+    },
+    {
+      "epoch": 0.1636412896234083,
+      "grad_norm": 0.11351019144058228,
+      "learning_rate": 0.0001,
+      "loss": 1.6991,
+      "step": 302
+    },
+    {
+      "epoch": 0.16418314819832025,
+      "grad_norm": 0.11626233905553818,
+      "learning_rate": 0.0001,
+      "loss": 1.7394,
+      "step": 303
+    },
+    {
+      "epoch": 0.16472500677323218,
+      "grad_norm": 0.10919834673404694,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 304
+    },
+    {
+      "epoch": 0.16526686534814414,
+      "grad_norm": 0.09492017328739166,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 305
+    },
+    {
+      "epoch": 0.1658087239230561,
+      "grad_norm": 0.11643269658088684,
+      "learning_rate": 0.0001,
+      "loss": 1.6428,
+      "step": 306
+    },
+    {
+      "epoch": 0.16635058249796802,
+      "grad_norm": 0.09301548451185226,
+      "learning_rate": 0.0001,
+      "loss": 1.689,
+      "step": 307
+    },
+    {
+      "epoch": 0.16689244107287998,
+      "grad_norm": 0.099439337849617,
+      "learning_rate": 0.0001,
+      "loss": 1.6594,
+      "step": 308
+    },
+    {
+      "epoch": 0.16743429964779194,
+      "grad_norm": 0.10214002430438995,
+      "learning_rate": 0.0001,
+      "loss": 1.6855,
+      "step": 309
+    },
+    {
+      "epoch": 0.16797615822270387,
+      "grad_norm": 0.0960782915353775,
+      "learning_rate": 0.0001,
+      "loss": 1.6501,
+      "step": 310
+    },
+    {
+      "epoch": 0.16851801679761583,
+      "grad_norm": 0.1130334809422493,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 311
+    },
+    {
+      "epoch": 0.16905987537252776,
+      "grad_norm": 0.09629669040441513,
+      "learning_rate": 0.0001,
+      "loss": 1.6605,
+      "step": 312
+    },
+    {
+      "epoch": 0.16960173394743971,
+      "grad_norm": 0.1021374985575676,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 313
+    },
+    {
+      "epoch": 0.17014359252235167,
+      "grad_norm": 0.09490853548049927,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 314
+    },
+    {
+      "epoch": 0.1706854510972636,
+      "grad_norm": 0.09929566830396652,
+      "learning_rate": 0.0001,
+      "loss": 1.6869,
+      "step": 315
+    },
+    {
+      "epoch": 0.17122730967217556,
+      "grad_norm": 0.09823586791753769,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 316
+    },
+    {
+      "epoch": 0.17176916824708752,
+      "grad_norm": 0.09158893674612045,
+      "learning_rate": 0.0001,
+      "loss": 1.6318,
+      "step": 317
+    },
+    {
+      "epoch": 0.17231102682199945,
+      "grad_norm": 0.09908587485551834,
+      "learning_rate": 0.0001,
+      "loss": 1.6252,
+      "step": 318
+    },
+    {
+      "epoch": 0.1728528853969114,
+      "grad_norm": 0.10510025173425674,
+      "learning_rate": 0.0001,
+      "loss": 1.6959,
+      "step": 319
+    },
+    {
+      "epoch": 0.17339474397182336,
+      "grad_norm": 0.10323290526866913,
+      "learning_rate": 0.0001,
+      "loss": 1.7018,
+      "step": 320
+    },
+    {
+      "epoch": 0.1739366025467353,
+      "grad_norm": 0.10068578273057938,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 321
+    },
+    {
+      "epoch": 0.17447846112164725,
+      "grad_norm": 0.10480581969022751,
+      "learning_rate": 0.0001,
+      "loss": 1.7247,
+      "step": 322
+    },
+    {
+      "epoch": 0.1750203196965592,
+      "grad_norm": 0.10604649782180786,
+      "learning_rate": 0.0001,
+      "loss": 1.7256,
+      "step": 323
+    },
+    {
+      "epoch": 0.17556217827147114,
+      "grad_norm": 0.10042749345302582,
+      "learning_rate": 0.0001,
+      "loss": 1.6702,
+      "step": 324
+    },
+    {
+      "epoch": 0.1761040368463831,
+      "grad_norm": 0.11323428899049759,
+      "learning_rate": 0.0001,
+      "loss": 1.6782,
+      "step": 325
+    },
+    {
+      "epoch": 0.17664589542129505,
+      "grad_norm": 0.10794851928949356,
+      "learning_rate": 0.0001,
+      "loss": 1.6927,
+      "step": 326
+    },
+    {
+      "epoch": 0.17718775399620698,
+      "grad_norm": 0.12508118152618408,
+      "learning_rate": 0.0001,
+      "loss": 1.6474,
+      "step": 327
+    },
+    {
+      "epoch": 0.17772961257111894,
+      "grad_norm": 0.1013966053724289,
+      "learning_rate": 0.0001,
+      "loss": 1.7129,
+      "step": 328
+    },
+    {
+      "epoch": 0.1782714711460309,
+      "grad_norm": 0.13972260057926178,
+      "learning_rate": 0.0001,
+      "loss": 1.7003,
+      "step": 329
+    },
+    {
+      "epoch": 0.17881332972094283,
+      "grad_norm": 0.1019311174750328,
+      "learning_rate": 0.0001,
+      "loss": 1.5998,
+      "step": 330
+    },
+    {
+      "epoch": 0.17935518829585478,
+      "grad_norm": 0.14320091903209686,
+      "learning_rate": 0.0001,
+      "loss": 1.6931,
+      "step": 331
+    },
+    {
+      "epoch": 0.17989704687076674,
+      "grad_norm": 0.10839429497718811,
+      "learning_rate": 0.0001,
+      "loss": 1.6485,
+      "step": 332
+    },
+    {
+      "epoch": 0.18043890544567867,
+      "grad_norm": 0.11457662284374237,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 333
+    },
+    {
+      "epoch": 0.18098076402059063,
+      "grad_norm": 0.11482690274715424,
+      "learning_rate": 0.0001,
+      "loss": 1.6596,
+      "step": 334
+    },
+    {
+      "epoch": 0.1815226225955026,
+      "grad_norm": 0.12048427760601044,
+      "learning_rate": 0.0001,
+      "loss": 1.7077,
+      "step": 335
+    },
+    {
+      "epoch": 0.18206448117041452,
+      "grad_norm": 0.11739542335271835,
+      "learning_rate": 0.0001,
+      "loss": 1.5914,
+      "step": 336
+    },
+    {
+      "epoch": 0.18260633974532647,
+      "grad_norm": 0.11013977229595184,
+      "learning_rate": 0.0001,
+      "loss": 1.6929,
+      "step": 337
+    },
+    {
+      "epoch": 0.18314819832023843,
+      "grad_norm": 0.12094748020172119,
+      "learning_rate": 0.0001,
+      "loss": 1.7103,
+      "step": 338
+    },
+    {
+      "epoch": 0.18369005689515036,
+      "grad_norm": 0.10130317509174347,
+      "learning_rate": 0.0001,
+      "loss": 1.6641,
+      "step": 339
+    },
+    {
+      "epoch": 0.18423191547006232,
+      "grad_norm": 0.10509441047906876,
+      "learning_rate": 0.0001,
+      "loss": 1.6784,
+      "step": 340
+    },
+    {
+      "epoch": 0.18477377404497425,
+      "grad_norm": 0.10123977065086365,
+      "learning_rate": 0.0001,
+      "loss": 1.6677,
+      "step": 341
+    },
+    {
+      "epoch": 0.1853156326198862,
+      "grad_norm": 0.12331061065196991,
+      "learning_rate": 0.0001,
+      "loss": 1.6667,
+      "step": 342
+    },
+    {
+      "epoch": 0.18585749119479816,
+      "grad_norm": 0.09549771249294281,
+      "learning_rate": 0.0001,
+      "loss": 1.6359,
+      "step": 343
+    },
+    {
+      "epoch": 0.1863993497697101,
+      "grad_norm": 0.10492638498544693,
+      "learning_rate": 0.0001,
+      "loss": 1.6516,
+      "step": 344
+    },
+    {
+      "epoch": 0.18694120834462205,
+      "grad_norm": 0.11436952650547028,
+      "learning_rate": 0.0001,
+      "loss": 1.7484,
+      "step": 345
+    },
+    {
+      "epoch": 0.187483066919534,
+      "grad_norm": 0.11754444241523743,
+      "learning_rate": 0.0001,
+      "loss": 1.7375,
+      "step": 346
+    },
+    {
+      "epoch": 0.18802492549444594,
+      "grad_norm": 0.11410810798406601,
+      "learning_rate": 0.0001,
+      "loss": 1.7044,
+      "step": 347
+    },
+    {
+      "epoch": 0.1885667840693579,
+      "grad_norm": 0.10755981504917145,
+      "learning_rate": 0.0001,
+      "loss": 1.7453,
+      "step": 348
+    },
+    {
+      "epoch": 0.18910864264426985,
+      "grad_norm": 0.1117858737707138,
+      "learning_rate": 0.0001,
+      "loss": 1.7144,
+      "step": 349
+    },
+    {
+      "epoch": 0.18965050121918178,
+      "grad_norm": 0.10845154523849487,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 350
+    },
+    {
+      "epoch": 0.19019235979409374,
+      "grad_norm": 0.10038649290800095,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 351
+    },
+    {
+      "epoch": 0.1907342183690057,
+      "grad_norm": 0.10859506577253342,
+      "learning_rate": 0.0001,
+      "loss": 1.5843,
+      "step": 352
+    },
+    {
+      "epoch": 0.19127607694391763,
+      "grad_norm": 0.10127723217010498,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 353
+    },
+    {
+      "epoch": 0.19181793551882959,
+      "grad_norm": 0.11287626624107361,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 354
+    },
+    {
+      "epoch": 0.19235979409374154,
+      "grad_norm": 0.11790165305137634,
+      "learning_rate": 0.0001,
+      "loss": 1.7098,
+      "step": 355
+    },
+    {
+      "epoch": 0.19290165266865347,
+      "grad_norm": 0.10026410967111588,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 356
+    },
+    {
+      "epoch": 0.19344351124356543,
+      "grad_norm": 0.11183635890483856,
+      "learning_rate": 0.0001,
+      "loss": 1.6775,
+      "step": 357
+    },
+    {
+      "epoch": 0.1939853698184774,
+      "grad_norm": 0.10310844331979752,
+      "learning_rate": 0.0001,
+      "loss": 1.6916,
+      "step": 358
+    },
+    {
+      "epoch": 0.19452722839338932,
+      "grad_norm": 0.09812598675489426,
+      "learning_rate": 0.0001,
+      "loss": 1.6442,
+      "step": 359
+    },
+    {
+      "epoch": 0.19506908696830128,
+      "grad_norm": 0.11021661758422852,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 360
+    },
+    {
+      "epoch": 0.19561094554321323,
+      "grad_norm": 0.09344365447759628,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 361
+    },
+    {
+      "epoch": 0.19615280411812516,
+      "grad_norm": 0.10808692872524261,
+      "learning_rate": 0.0001,
+      "loss": 1.6349,
+      "step": 362
+    },
+    {
+      "epoch": 0.19669466269303712,
+      "grad_norm": 0.11180002242326736,
+      "learning_rate": 0.0001,
+      "loss": 1.6901,
+      "step": 363
+    },
+    {
+      "epoch": 0.19723652126794908,
+      "grad_norm": 0.10783065110445023,
+      "learning_rate": 0.0001,
+      "loss": 1.7047,
+      "step": 364
+    },
+    {
+      "epoch": 0.197778379842861,
+      "grad_norm": 0.1020706295967102,
+      "learning_rate": 0.0001,
+      "loss": 1.6763,
+      "step": 365
+    },
+    {
+      "epoch": 0.19832023841777296,
+      "grad_norm": 0.09901352971792221,
+      "learning_rate": 0.0001,
+      "loss": 1.6807,
+      "step": 366
+    },
+    {
+      "epoch": 0.19886209699268492,
+      "grad_norm": 0.10985004901885986,
+      "learning_rate": 0.0001,
+      "loss": 1.6885,
+      "step": 367
+    },
+    {
+      "epoch": 0.19940395556759685,
+      "grad_norm": 0.10345008969306946,
+      "learning_rate": 0.0001,
+      "loss": 1.631,
+      "step": 368
+    },
+    {
+      "epoch": 0.1999458141425088,
+      "grad_norm": 0.09653160721063614,
+      "learning_rate": 0.0001,
+      "loss": 1.6195,
+      "step": 369
+    },
+    {
+      "epoch": 0.20048767271742074,
+      "grad_norm": 0.1158447191119194,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 370
+    },
+    {
+      "epoch": 0.2010295312923327,
+      "grad_norm": 0.1000077947974205,
+      "learning_rate": 0.0001,
+      "loss": 1.6777,
+      "step": 371
+    },
+    {
+      "epoch": 0.20157138986724465,
+      "grad_norm": 0.11747664958238602,
+      "learning_rate": 0.0001,
+      "loss": 1.7121,
+      "step": 372
+    },
+    {
+      "epoch": 0.20211324844215658,
+      "grad_norm": 0.0929274931550026,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 373
+    },
+    {
+      "epoch": 0.20265510701706854,
+      "grad_norm": 0.12577304244041443,
+      "learning_rate": 0.0001,
+      "loss": 1.7982,
+      "step": 374
+    },
+    {
+      "epoch": 0.2031969655919805,
+      "grad_norm": 0.09752815961837769,
+      "learning_rate": 0.0001,
+      "loss": 1.6345,
+      "step": 375
+    },
+    {
+      "epoch": 0.20373882416689243,
+      "grad_norm": 0.11870884150266647,
+      "learning_rate": 0.0001,
+      "loss": 1.6629,
+      "step": 376
+    },
+    {
+      "epoch": 0.2042806827418044,
+      "grad_norm": 0.10201161354780197,
+      "learning_rate": 0.0001,
+      "loss": 1.6312,
+      "step": 377
+    },
+    {
+      "epoch": 0.20482254131671634,
+      "grad_norm": 0.10613151639699936,
+      "learning_rate": 0.0001,
+      "loss": 1.6118,
+      "step": 378
+    },
+    {
+      "epoch": 0.20536439989162827,
+      "grad_norm": 0.10123321413993835,
+      "learning_rate": 0.0001,
+      "loss": 1.6801,
+      "step": 379
+    },
+    {
+      "epoch": 0.20590625846654023,
+      "grad_norm": 0.09528015553951263,
+      "learning_rate": 0.0001,
+      "loss": 1.7344,
+      "step": 380
+    },
+    {
+      "epoch": 0.2064481170414522,
+      "grad_norm": 0.10694505274295807,
+      "learning_rate": 0.0001,
+      "loss": 1.6542,
+      "step": 381
+    },
+    {
+      "epoch": 0.20698997561636412,
+      "grad_norm": 0.10317772626876831,
+      "learning_rate": 0.0001,
+      "loss": 1.6807,
+      "step": 382
+    },
+    {
+      "epoch": 0.20753183419127608,
+      "grad_norm": 0.11525990068912506,
+      "learning_rate": 0.0001,
+      "loss": 1.6948,
+      "step": 383
+    },
+    {
+      "epoch": 0.20807369276618803,
+      "grad_norm": 0.09970982372760773,
+      "learning_rate": 0.0001,
+      "loss": 1.6761,
+      "step": 384
+    },
+    {
+      "epoch": 0.20861555134109996,
+      "grad_norm": 0.11480505764484406,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 385
+    },
+    {
+      "epoch": 0.20915740991601192,
+      "grad_norm": 0.10115650296211243,
+      "learning_rate": 0.0001,
+      "loss": 1.6408,
+      "step": 386
+    },
+    {
+      "epoch": 0.20969926849092388,
+      "grad_norm": 0.13291612267494202,
+      "learning_rate": 0.0001,
+      "loss": 1.7566,
+      "step": 387
+    },
+    {
+      "epoch": 0.2102411270658358,
+      "grad_norm": 0.09564415365457535,
+      "learning_rate": 0.0001,
+      "loss": 1.6884,
+      "step": 388
+    },
+    {
+      "epoch": 0.21078298564074777,
+      "grad_norm": 0.11687403172254562,
+      "learning_rate": 0.0001,
+      "loss": 1.7079,
+      "step": 389
+    },
+    {
+      "epoch": 0.21132484421565972,
+      "grad_norm": 0.09814255684614182,
+      "learning_rate": 0.0001,
+      "loss": 1.7098,
+      "step": 390
+    },
+    {
+      "epoch": 0.21186670279057165,
+      "grad_norm": 0.10159056633710861,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 391
+    },
+    {
+      "epoch": 0.2124085613654836,
+      "grad_norm": 0.10189589112997055,
+      "learning_rate": 0.0001,
+      "loss": 1.7352,
+      "step": 392
+    },
+    {
+      "epoch": 0.21295041994039557,
+      "grad_norm": 0.10089243203401566,
+      "learning_rate": 0.0001,
+      "loss": 1.6998,
+      "step": 393
+    },
+    {
+      "epoch": 0.2134922785153075,
+      "grad_norm": 0.11454922705888748,
+      "learning_rate": 0.0001,
+      "loss": 1.6728,
+      "step": 394
+    },
+    {
+      "epoch": 0.21403413709021946,
+      "grad_norm": 0.101554736495018,
+      "learning_rate": 0.0001,
+      "loss": 1.716,
+      "step": 395
+    },
+    {
+      "epoch": 0.2145759956651314,
+      "grad_norm": 0.1074264720082283,
+      "learning_rate": 0.0001,
+      "loss": 1.6558,
+      "step": 396
+    },
+    {
+      "epoch": 0.21511785424004334,
+      "grad_norm": 0.10516046732664108,
+      "learning_rate": 0.0001,
+      "loss": 1.6887,
+      "step": 397
+    },
+    {
+      "epoch": 0.2156597128149553,
+      "grad_norm": 0.09886284172534943,
+      "learning_rate": 0.0001,
+      "loss": 1.6727,
+      "step": 398
+    },
+    {
+      "epoch": 0.21620157138986723,
+      "grad_norm": 0.09954807907342911,
+      "learning_rate": 0.0001,
+      "loss": 1.6712,
+      "step": 399
+    },
+    {
+      "epoch": 0.2167434299647792,
+      "grad_norm": 0.09766504168510437,
+      "learning_rate": 0.0001,
+      "loss": 1.6929,
+      "step": 400
+    },
+    {
+      "epoch": 0.21728528853969115,
+      "grad_norm": 0.09490859508514404,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 401
+    },
+    {
+      "epoch": 0.21782714711460308,
+      "grad_norm": 0.100528784096241,
+      "learning_rate": 0.0001,
+      "loss": 1.7074,
+      "step": 402
+    },
+    {
+      "epoch": 0.21836900568951503,
+      "grad_norm": 0.10056279599666595,
+      "learning_rate": 0.0001,
+      "loss": 1.6746,
+      "step": 403
+    },
+    {
+      "epoch": 0.218910864264427,
+      "grad_norm": 0.10368312150239944,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 404
+    },
+    {
+      "epoch": 0.21945272283933892,
+      "grad_norm": 0.10767726600170135,
+      "learning_rate": 0.0001,
+      "loss": 1.638,
+      "step": 405
+    },
+    {
+      "epoch": 0.21999458141425088,
+      "grad_norm": 0.09969790279865265,
+      "learning_rate": 0.0001,
+      "loss": 1.6853,
+      "step": 406
+    },
+    {
+      "epoch": 0.22053643998916284,
+      "grad_norm": 0.12781274318695068,
+      "learning_rate": 0.0001,
+      "loss": 1.7639,
+      "step": 407
+    },
+    {
+      "epoch": 0.22107829856407477,
+      "grad_norm": 0.11066941916942596,
+      "learning_rate": 0.0001,
+      "loss": 1.602,
+      "step": 408
+    },
+    {
+      "epoch": 0.22162015713898672,
+      "grad_norm": 0.10894569009542465,
+      "learning_rate": 0.0001,
+      "loss": 1.7199,
+      "step": 409
+    },
+    {
+      "epoch": 0.22216201571389868,
+      "grad_norm": 0.116873599588871,
+      "learning_rate": 0.0001,
+      "loss": 1.7119,
+      "step": 410
+    },
+    {
+      "epoch": 0.2227038742888106,
+      "grad_norm": 0.10727427154779434,
+      "learning_rate": 0.0001,
+      "loss": 1.7022,
+      "step": 411
+    },
+    {
+      "epoch": 0.22324573286372257,
+      "grad_norm": 0.10264337062835693,
+      "learning_rate": 0.0001,
+      "loss": 1.6432,
+      "step": 412
+    },
+    {
+      "epoch": 0.22378759143863453,
+      "grad_norm": 0.11667589843273163,
+      "learning_rate": 0.0001,
+      "loss": 1.6314,
+      "step": 413
+    },
+    {
+      "epoch": 0.22432945001354646,
+      "grad_norm": 0.09983885288238525,
+      "learning_rate": 0.0001,
+      "loss": 1.7386,
+      "step": 414
+    },
+    {
+      "epoch": 0.2248713085884584,
+      "grad_norm": 0.10147198289632797,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 415
+    },
+    {
+      "epoch": 0.22541316716337037,
+      "grad_norm": 0.10878276824951172,
+      "learning_rate": 0.0001,
+      "loss": 1.7047,
+      "step": 416
+    },
+    {
+      "epoch": 0.2259550257382823,
+      "grad_norm": 0.10325547307729721,
+      "learning_rate": 0.0001,
+      "loss": 1.7173,
+      "step": 417
+    },
+    {
+      "epoch": 0.22649688431319426,
+      "grad_norm": 0.09619980305433273,
+      "learning_rate": 0.0001,
+      "loss": 1.6854,
+      "step": 418
+    },
+    {
+      "epoch": 0.22703874288810622,
+      "grad_norm": 0.1018766313791275,
+      "learning_rate": 0.0001,
+      "loss": 1.6268,
+      "step": 419
+    },
+    {
+      "epoch": 0.22758060146301815,
+      "grad_norm": 0.10426010936498642,
+      "learning_rate": 0.0001,
+      "loss": 1.6791,
+      "step": 420
+    },
+    {
+      "epoch": 0.2281224600379301,
+      "grad_norm": 0.09923192858695984,
+      "learning_rate": 0.0001,
+      "loss": 1.7055,
+      "step": 421
+    },
+    {
+      "epoch": 0.22866431861284206,
+      "grad_norm": 0.10499102622270584,
+      "learning_rate": 0.0001,
+      "loss": 1.7197,
+      "step": 422
+    },
+    {
+      "epoch": 0.229206177187754,
+      "grad_norm": 0.10078815370798111,
+      "learning_rate": 0.0001,
+      "loss": 1.6863,
+      "step": 423
+    },
+    {
+      "epoch": 0.22974803576266595,
+      "grad_norm": 0.09503324329853058,
+      "learning_rate": 0.0001,
+      "loss": 1.6076,
+      "step": 424
+    },
+    {
+      "epoch": 0.2302898943375779,
+      "grad_norm": 0.09994073212146759,
+      "learning_rate": 0.0001,
+      "loss": 1.6448,
+      "step": 425
+    },
+    {
+      "epoch": 0.23083175291248983,
+      "grad_norm": 0.1010836511850357,
+      "learning_rate": 0.0001,
+      "loss": 1.6899,
+      "step": 426
+    },
+    {
+      "epoch": 0.2313736114874018,
+      "grad_norm": 0.10668766498565674,
+      "learning_rate": 0.0001,
+      "loss": 1.7398,
+      "step": 427
+    },
+    {
+      "epoch": 0.23191547006231375,
+      "grad_norm": 0.09935601055622101,
+      "learning_rate": 0.0001,
+      "loss": 1.6781,
+      "step": 428
+    },
+    {
+      "epoch": 0.23245732863722568,
+      "grad_norm": 0.10159440338611603,
+      "learning_rate": 0.0001,
+      "loss": 1.6189,
+      "step": 429
+    },
+    {
+      "epoch": 0.23299918721213764,
+      "grad_norm": 0.09990345686674118,
+      "learning_rate": 0.0001,
+      "loss": 1.6952,
+      "step": 430
+    },
+    {
+      "epoch": 0.23354104578704957,
+      "grad_norm": 0.11645045876502991,
+      "learning_rate": 0.0001,
+      "loss": 1.6568,
+      "step": 431
+    },
+    {
+      "epoch": 0.23408290436196152,
+      "grad_norm": 0.1003262847661972,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 432
+    },
+    {
+      "epoch": 0.23462476293687348,
+      "grad_norm": 0.11681778728961945,
+      "learning_rate": 0.0001,
+      "loss": 1.6854,
+      "step": 433
+    },
+    {
+      "epoch": 0.2351666215117854,
+      "grad_norm": 0.09814731776714325,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 434
+    },
+    {
+      "epoch": 0.23570848008669737,
+      "grad_norm": 0.11347492784261703,
+      "learning_rate": 0.0001,
+      "loss": 1.6595,
+      "step": 435
+    },
+    {
+      "epoch": 0.23625033866160933,
+      "grad_norm": 0.09687422215938568,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 436
+    },
+    {
+      "epoch": 0.23679219723652126,
+      "grad_norm": 0.13785980641841888,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 437
+    },
+    {
+      "epoch": 0.23733405581143321,
+      "grad_norm": 0.0989057645201683,
+      "learning_rate": 0.0001,
+      "loss": 1.6681,
+      "step": 438
+    },
+    {
+      "epoch": 0.23787591438634517,
+      "grad_norm": 0.13210071623325348,
+      "learning_rate": 0.0001,
+      "loss": 1.671,
+      "step": 439
+    },
+    {
+      "epoch": 0.2384177729612571,
+      "grad_norm": 0.11280949413776398,
+      "learning_rate": 0.0001,
+      "loss": 1.6713,
+      "step": 440
+    },
+    {
+      "epoch": 0.23895963153616906,
+      "grad_norm": 0.12719376385211945,
+      "learning_rate": 0.0001,
+      "loss": 1.6558,
+      "step": 441
+    },
+    {
+      "epoch": 0.23950149011108102,
+      "grad_norm": 0.1163676455616951,
+      "learning_rate": 0.0001,
+      "loss": 1.6918,
+      "step": 442
+    },
+    {
+      "epoch": 0.24004334868599295,
+      "grad_norm": 0.11804498732089996,
+      "learning_rate": 0.0001,
+      "loss": 1.563,
+      "step": 443
+    },
+    {
+      "epoch": 0.2405852072609049,
+      "grad_norm": 0.12228959053754807,
+      "learning_rate": 0.0001,
+      "loss": 1.6163,
+      "step": 444
+    },
+    {
+      "epoch": 0.24112706583581686,
+      "grad_norm": 0.11664146929979324,
+      "learning_rate": 0.0001,
+      "loss": 1.6865,
+      "step": 445
+    },
+    {
+      "epoch": 0.2416689244107288,
+      "grad_norm": 0.13231196999549866,
+      "learning_rate": 0.0001,
+      "loss": 1.6373,
+      "step": 446
+    },
+    {
+      "epoch": 0.24221078298564075,
+      "grad_norm": 0.12525129318237305,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 447
+    },
+    {
+      "epoch": 0.2427526415605527,
+      "grad_norm": 0.11747252941131592,
+      "learning_rate": 0.0001,
+      "loss": 1.6418,
+      "step": 448
+    },
+    {
+      "epoch": 0.24329450013546464,
+      "grad_norm": 0.11415065079927444,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 449
+    },
+    {
+      "epoch": 0.2438363587103766,
+      "grad_norm": 0.12443743646144867,
+      "learning_rate": 0.0001,
+      "loss": 1.5737,
+      "step": 450
+    },
+    {
+      "epoch": 0.24437821728528855,
+      "grad_norm": 0.10312895476818085,
+      "learning_rate": 0.0001,
+      "loss": 1.6198,
+      "step": 451
+    },
+    {
+      "epoch": 0.24492007586020048,
+      "grad_norm": 0.1094118282198906,
+      "learning_rate": 0.0001,
+      "loss": 1.6354,
+      "step": 452
+    },
+    {
+      "epoch": 0.24546193443511244,
+      "grad_norm": 0.10630492120981216,
+      "learning_rate": 0.0001,
+      "loss": 1.6788,
+      "step": 453
+    },
+    {
+      "epoch": 0.2460037930100244,
+      "grad_norm": 0.11164608597755432,
+      "learning_rate": 0.0001,
+      "loss": 1.6228,
+      "step": 454
+    },
+    {
+      "epoch": 0.24654565158493633,
+      "grad_norm": 0.10897690057754517,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 455
+    },
+    {
+      "epoch": 0.24708751015984828,
+      "grad_norm": 0.12230285257101059,
+      "learning_rate": 0.0001,
+      "loss": 1.7062,
+      "step": 456
+    },
+    {
+      "epoch": 0.24762936873476024,
+      "grad_norm": 0.10371015965938568,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 457
+    },
+    {
+      "epoch": 0.24817122730967217,
+      "grad_norm": 0.11259844154119492,
+      "learning_rate": 0.0001,
+      "loss": 1.6497,
+      "step": 458
+    },
+    {
+      "epoch": 0.24871308588458413,
+      "grad_norm": 0.10375142097473145,
+      "learning_rate": 0.0001,
+      "loss": 1.6638,
+      "step": 459
+    },
+    {
+      "epoch": 0.24925494445949606,
+      "grad_norm": 0.09826785326004028,
+      "learning_rate": 0.0001,
+      "loss": 1.6221,
+      "step": 460
+    },
+    {
+      "epoch": 0.24979680303440802,
+      "grad_norm": 0.11986129730939865,
+      "learning_rate": 0.0001,
+      "loss": 1.7278,
+      "step": 461
+    },
+    {
+      "epoch": 0.25033866160931995,
+      "grad_norm": 0.09854941070079803,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 462
+    },
+    {
+      "epoch": 0.2508805201842319,
+      "grad_norm": 0.12960360944271088,
+      "learning_rate": 0.0001,
+      "loss": 1.6793,
+      "step": 463
+    },
+    {
+      "epoch": 0.25142237875914386,
+      "grad_norm": 0.10426454246044159,
+      "learning_rate": 0.0001,
+      "loss": 1.6236,
+      "step": 464
+    },
+    {
+      "epoch": 0.2519642373340558,
+      "grad_norm": 0.10605210810899734,
+      "learning_rate": 0.0001,
+      "loss": 1.6366,
+      "step": 465
+    },
+    {
+      "epoch": 0.2525060959089678,
+      "grad_norm": 0.10613849759101868,
+      "learning_rate": 0.0001,
+      "loss": 1.6782,
+      "step": 466
+    },
+    {
+      "epoch": 0.25304795448387973,
+      "grad_norm": 0.10717356204986572,
+      "learning_rate": 0.0001,
+      "loss": 1.6556,
+      "step": 467
+    },
+    {
+      "epoch": 0.25358981305879164,
+      "grad_norm": 0.10753747075796127,
+      "learning_rate": 0.0001,
+      "loss": 1.6175,
+      "step": 468
+    },
+    {
+      "epoch": 0.2541316716337036,
+      "grad_norm": 0.10362911969423294,
+      "learning_rate": 0.0001,
+      "loss": 1.6179,
+      "step": 469
+    },
+    {
+      "epoch": 0.25467353020861555,
+      "grad_norm": 0.11893575638532639,
+      "learning_rate": 0.0001,
+      "loss": 1.6897,
+      "step": 470
+    },
+    {
+      "epoch": 0.2552153887835275,
+      "grad_norm": 0.10144247114658356,
+      "learning_rate": 0.0001,
+      "loss": 1.6823,
+      "step": 471
+    },
+    {
+      "epoch": 0.25575724735843947,
+      "grad_norm": 0.11344347894191742,
+      "learning_rate": 0.0001,
+      "loss": 1.6524,
+      "step": 472
+    },
+    {
+      "epoch": 0.25629910593335137,
+      "grad_norm": 0.10365281254053116,
+      "learning_rate": 0.0001,
+      "loss": 1.6026,
+      "step": 473
+    },
+    {
+      "epoch": 0.2568409645082633,
+      "grad_norm": 0.13139420747756958,
+      "learning_rate": 0.0001,
+      "loss": 1.7037,
+      "step": 474
+    },
+    {
+      "epoch": 0.2573828230831753,
+      "grad_norm": 0.10608001798391342,
+      "learning_rate": 0.0001,
+      "loss": 1.6697,
+      "step": 475
+    },
+    {
+      "epoch": 0.25792468165808724,
+      "grad_norm": 0.12491496652364731,
+      "learning_rate": 0.0001,
+      "loss": 1.6571,
+      "step": 476
+    },
+    {
+      "epoch": 0.2584665402329992,
+      "grad_norm": 0.10368447750806808,
+      "learning_rate": 0.0001,
+      "loss": 1.774,
+      "step": 477
+    },
+    {
+      "epoch": 0.25900839880791116,
+      "grad_norm": 0.13536019623279572,
+      "learning_rate": 0.0001,
+      "loss": 1.6922,
+      "step": 478
+    },
+    {
+      "epoch": 0.25955025738282306,
+      "grad_norm": 0.10196977853775024,
+      "learning_rate": 0.0001,
+      "loss": 1.588,
+      "step": 479
+    },
+    {
+      "epoch": 0.260092115957735,
+      "grad_norm": 0.09920040518045425,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 480
+    },
+    {
+      "epoch": 0.260633974532647,
+      "grad_norm": 0.12618587911128998,
+      "learning_rate": 0.0001,
+      "loss": 1.7136,
+      "step": 481
+    },
+    {
+      "epoch": 0.26117583310755893,
+      "grad_norm": 0.10382471978664398,
+      "learning_rate": 0.0001,
+      "loss": 1.6486,
+      "step": 482
+    },
+    {
+      "epoch": 0.2617176916824709,
+      "grad_norm": 0.11146610975265503,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 483
+    },
+    {
+      "epoch": 0.26225955025738285,
+      "grad_norm": 0.1008019968867302,
+      "learning_rate": 0.0001,
+      "loss": 1.7193,
+      "step": 484
+    },
+    {
+      "epoch": 0.26280140883229475,
+      "grad_norm": 0.11461275070905685,
+      "learning_rate": 0.0001,
+      "loss": 1.7452,
+      "step": 485
+    },
+    {
+      "epoch": 0.2633432674072067,
+      "grad_norm": 0.10348404943943024,
+      "learning_rate": 0.0001,
+      "loss": 1.7221,
+      "step": 486
+    },
+    {
+      "epoch": 0.26388512598211866,
+      "grad_norm": 0.1001395434141159,
+      "learning_rate": 0.0001,
+      "loss": 1.6268,
+      "step": 487
+    },
+    {
+      "epoch": 0.2644269845570306,
+      "grad_norm": 0.10220689326524734,
+      "learning_rate": 0.0001,
+      "loss": 1.6551,
+      "step": 488
+    },
+    {
+      "epoch": 0.2649688431319426,
+      "grad_norm": 0.10228809714317322,
+      "learning_rate": 0.0001,
+      "loss": 1.6783,
+      "step": 489
+    },
+    {
+      "epoch": 0.26551070170685453,
+      "grad_norm": 0.1036238893866539,
+      "learning_rate": 0.0001,
+      "loss": 1.7073,
+      "step": 490
+    },
+    {
+      "epoch": 0.26605256028176644,
+      "grad_norm": 0.10383958369493484,
+      "learning_rate": 0.0001,
+      "loss": 1.6872,
+      "step": 491
+    },
+    {
+      "epoch": 0.2665944188566784,
+      "grad_norm": 0.10285360366106033,
+      "learning_rate": 0.0001,
+      "loss": 1.6453,
+      "step": 492
+    },
+    {
+      "epoch": 0.26713627743159035,
+      "grad_norm": 0.10352769494056702,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 493
+    },
+    {
+      "epoch": 0.2676781360065023,
+      "grad_norm": 0.10272315889596939,
+      "learning_rate": 0.0001,
+      "loss": 1.6625,
+      "step": 494
+    },
+    {
+      "epoch": 0.26821999458141427,
+      "grad_norm": 0.11963304132223129,
+      "learning_rate": 0.0001,
+      "loss": 1.6603,
+      "step": 495
+    },
+    {
+      "epoch": 0.2687618531563262,
+      "grad_norm": 0.10235189646482468,
+      "learning_rate": 0.0001,
+      "loss": 1.6439,
+      "step": 496
+    },
+    {
+      "epoch": 0.2693037117312381,
+      "grad_norm": 0.13295263051986694,
+      "learning_rate": 0.0001,
+      "loss": 1.6389,
+      "step": 497
+    },
+    {
+      "epoch": 0.2698455703061501,
+      "grad_norm": 0.10163283348083496,
+      "learning_rate": 0.0001,
+      "loss": 1.6521,
+      "step": 498
+    },
+    {
+      "epoch": 0.27038742888106204,
+      "grad_norm": 0.12278732657432556,
+      "learning_rate": 0.0001,
+      "loss": 1.6807,
+      "step": 499
+    },
+    {
+      "epoch": 0.270929287455974,
+      "grad_norm": 0.10807237774133682,
+      "learning_rate": 0.0001,
+      "loss": 1.6579,
+      "step": 500
+    },
+    {
+      "epoch": 0.27147114603088596,
+      "grad_norm": 0.1201099380850792,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 501
+    },
+    {
+      "epoch": 0.27201300460579786,
+      "grad_norm": 0.1056004986166954,
+      "learning_rate": 0.0001,
+      "loss": 1.6274,
+      "step": 502
+    },
+    {
+      "epoch": 0.2725548631807098,
+      "grad_norm": 0.10857392847537994,
+      "learning_rate": 0.0001,
+      "loss": 1.6694,
+      "step": 503
+    },
+    {
+      "epoch": 0.2730967217556218,
+      "grad_norm": 0.11863652616739273,
+      "learning_rate": 0.0001,
+      "loss": 1.681,
+      "step": 504
+    },
+    {
+      "epoch": 0.27363858033053373,
+      "grad_norm": 0.10198169946670532,
+      "learning_rate": 0.0001,
+      "loss": 1.6831,
+      "step": 505
+    },
+    {
+      "epoch": 0.2741804389054457,
+      "grad_norm": 0.1210508868098259,
+      "learning_rate": 0.0001,
+      "loss": 1.6068,
+      "step": 506
+    },
+    {
+      "epoch": 0.27472229748035765,
+      "grad_norm": 0.10578924417495728,
+      "learning_rate": 0.0001,
+      "loss": 1.6544,
+      "step": 507
+    },
+    {
+      "epoch": 0.27526415605526955,
+      "grad_norm": 0.11378353834152222,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 508
+    },
+    {
+      "epoch": 0.2758060146301815,
+      "grad_norm": 0.13088662922382355,
+      "learning_rate": 0.0001,
+      "loss": 1.6272,
+      "step": 509
+    },
+    {
+      "epoch": 0.27634787320509346,
+      "grad_norm": 0.10998623073101044,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 510
+    },
+    {
+      "epoch": 0.2768897317800054,
+      "grad_norm": 0.1617870032787323,
+      "learning_rate": 0.0001,
+      "loss": 1.682,
+      "step": 511
+    },
+    {
+      "epoch": 0.2774315903549174,
+      "grad_norm": 0.13237886130809784,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 512
+    },
+    {
+      "epoch": 0.27797344892982934,
+      "grad_norm": 0.13461074233055115,
+      "learning_rate": 0.0001,
+      "loss": 1.637,
+      "step": 513
+    },
+    {
+      "epoch": 0.27851530750474124,
+      "grad_norm": 0.12076643109321594,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 514
+    },
+    {
+      "epoch": 0.2790571660796532,
+      "grad_norm": 0.14691396057605743,
+      "learning_rate": 0.0001,
+      "loss": 1.6874,
+      "step": 515
+    },
+    {
+      "epoch": 0.27959902465456515,
+      "grad_norm": 0.10137977451086044,
+      "learning_rate": 0.0001,
+      "loss": 1.5602,
+      "step": 516
+    },
+    {
+      "epoch": 0.2801408832294771,
+      "grad_norm": 0.13679617643356323,
+      "learning_rate": 0.0001,
+      "loss": 1.6779,
+      "step": 517
+    },
+    {
+      "epoch": 0.28068274180438907,
+      "grad_norm": 0.10055786371231079,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 518
+    },
+    {
+      "epoch": 0.281224600379301,
+      "grad_norm": 0.11336284130811691,
+      "learning_rate": 0.0001,
+      "loss": 1.633,
+      "step": 519
+    },
+    {
+      "epoch": 0.28176645895421293,
+      "grad_norm": 0.10535544157028198,
+      "learning_rate": 0.0001,
+      "loss": 1.7045,
+      "step": 520
+    },
+    {
+      "epoch": 0.2823083175291249,
+      "grad_norm": 0.10427578538656235,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 521
+    },
+    {
+      "epoch": 0.28285017610403684,
+      "grad_norm": 0.12511974573135376,
+      "learning_rate": 0.0001,
+      "loss": 1.7355,
+      "step": 522
+    },
+    {
+      "epoch": 0.2833920346789488,
+      "grad_norm": 0.105614572763443,
+      "learning_rate": 0.0001,
+      "loss": 1.6429,
+      "step": 523
+    },
+    {
+      "epoch": 0.28393389325386076,
+      "grad_norm": 0.12445893883705139,
+      "learning_rate": 0.0001,
+      "loss": 1.6925,
+      "step": 524
+    },
+    {
+      "epoch": 0.2844757518287727,
+      "grad_norm": 0.12350346893072128,
+      "learning_rate": 0.0001,
+      "loss": 1.7239,
+      "step": 525
+    },
+    {
+      "epoch": 0.2850176104036846,
+      "grad_norm": 0.11333523690700531,
+      "learning_rate": 0.0001,
+      "loss": 1.6035,
+      "step": 526
+    },
+    {
+      "epoch": 0.2855594689785966,
+      "grad_norm": 0.1175926998257637,
+      "learning_rate": 0.0001,
+      "loss": 1.6616,
+      "step": 527
+    },
+    {
+      "epoch": 0.28610132755350853,
+      "grad_norm": 0.10162296891212463,
+      "learning_rate": 0.0001,
+      "loss": 1.6814,
+      "step": 528
+    },
+    {
+      "epoch": 0.2866431861284205,
+      "grad_norm": 0.13615354895591736,
+      "learning_rate": 0.0001,
+      "loss": 1.6623,
+      "step": 529
+    },
+    {
+      "epoch": 0.28718504470333245,
+      "grad_norm": 0.10404025763273239,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 530
+    },
+    {
+      "epoch": 0.2877269032782444,
+      "grad_norm": 0.12436753511428833,
+      "learning_rate": 0.0001,
+      "loss": 1.7225,
+      "step": 531
+    },
+    {
+      "epoch": 0.2882687618531563,
+      "grad_norm": 0.11511912941932678,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 532
+    },
+    {
+      "epoch": 0.28881062042806827,
+      "grad_norm": 0.11416777223348618,
+      "learning_rate": 0.0001,
+      "loss": 1.6608,
+      "step": 533
+    },
+    {
+      "epoch": 0.2893524790029802,
+      "grad_norm": 0.1348169595003128,
+      "learning_rate": 0.0001,
+      "loss": 1.6777,
+      "step": 534
+    },
+    {
+      "epoch": 0.2898943375778922,
+      "grad_norm": 0.09927986562252045,
+      "learning_rate": 0.0001,
+      "loss": 1.5467,
+      "step": 535
+    },
+    {
+      "epoch": 0.29043619615280414,
+      "grad_norm": 0.13574469089508057,
+      "learning_rate": 0.0001,
+      "loss": 1.62,
+      "step": 536
+    },
+    {
+      "epoch": 0.29097805472771604,
+      "grad_norm": 0.10188048332929611,
+      "learning_rate": 0.0001,
+      "loss": 1.6531,
+      "step": 537
+    },
+    {
+      "epoch": 0.291519913302628,
+      "grad_norm": 0.1282220184803009,
+      "learning_rate": 0.0001,
+      "loss": 1.6559,
+      "step": 538
+    },
+    {
+      "epoch": 0.29206177187753996,
+      "grad_norm": 0.1127096638083458,
+      "learning_rate": 0.0001,
+      "loss": 1.6372,
+      "step": 539
+    },
+    {
+      "epoch": 0.2926036304524519,
+      "grad_norm": 0.1312182992696762,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 540
+    },
+    {
+      "epoch": 0.29314548902736387,
+      "grad_norm": 0.10491517931222916,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 541
+    },
+    {
+      "epoch": 0.29368734760227583,
+      "grad_norm": 0.1108413115143776,
+      "learning_rate": 0.0001,
+      "loss": 1.6679,
+      "step": 542
+    },
+    {
+      "epoch": 0.29422920617718773,
+      "grad_norm": 0.1049225851893425,
+      "learning_rate": 0.0001,
+      "loss": 1.6786,
+      "step": 543
+    },
+    {
+      "epoch": 0.2947710647520997,
+      "grad_norm": 0.09961540251970291,
+      "learning_rate": 0.0001,
+      "loss": 1.6962,
+      "step": 544
+    },
+    {
+      "epoch": 0.29531292332701164,
+      "grad_norm": 0.11128535121679306,
+      "learning_rate": 0.0001,
+      "loss": 1.6793,
+      "step": 545
+    },
+    {
+      "epoch": 0.2958547819019236,
+      "grad_norm": 0.1057731881737709,
+      "learning_rate": 0.0001,
+      "loss": 1.6901,
+      "step": 546
+    },
+    {
+      "epoch": 0.29639664047683556,
+      "grad_norm": 0.09957876056432724,
+      "learning_rate": 0.0001,
+      "loss": 1.6725,
+      "step": 547
+    },
+    {
+      "epoch": 0.2969384990517475,
+      "grad_norm": 0.10864951461553574,
+      "learning_rate": 0.0001,
+      "loss": 1.7211,
+      "step": 548
+    },
+    {
+      "epoch": 0.2974803576266594,
+      "grad_norm": 0.10822419822216034,
+      "learning_rate": 0.0001,
+      "loss": 1.759,
+      "step": 549
+    },
+    {
+      "epoch": 0.2980222162015714,
+      "grad_norm": 0.11316762119531631,
+      "learning_rate": 0.0001,
+      "loss": 1.701,
+      "step": 550
+    },
+    {
+      "epoch": 0.29856407477648333,
+      "grad_norm": 0.09931345283985138,
+      "learning_rate": 0.0001,
+      "loss": 1.6934,
+      "step": 551
+    },
+    {
+      "epoch": 0.2991059333513953,
+      "grad_norm": 0.09663783013820648,
+      "learning_rate": 0.0001,
+      "loss": 1.6083,
+      "step": 552
+    },
+    {
+      "epoch": 0.29964779192630725,
+      "grad_norm": 0.11659286171197891,
+      "learning_rate": 0.0001,
+      "loss": 1.6768,
+      "step": 553
+    },
+    {
+      "epoch": 0.3001896505012192,
+      "grad_norm": 0.09947337210178375,
+      "learning_rate": 0.0001,
+      "loss": 1.6738,
+      "step": 554
+    },
+    {
+      "epoch": 0.3007315090761311,
+      "grad_norm": 0.09798821806907654,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 555
+    },
+    {
+      "epoch": 0.30127336765104307,
+      "grad_norm": 0.10035169124603271,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 556
+    },
+    {
+      "epoch": 0.301815226225955,
+      "grad_norm": 0.1058754101395607,
+      "learning_rate": 0.0001,
+      "loss": 1.741,
+      "step": 557
+    },
+    {
+      "epoch": 0.302357084800867,
+      "grad_norm": 0.09941428154706955,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 558
+    },
+    {
+      "epoch": 0.30289894337577894,
+      "grad_norm": 0.09876841306686401,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 559
+    },
+    {
+      "epoch": 0.3034408019506909,
+      "grad_norm": 0.10757549852132797,
+      "learning_rate": 0.0001,
+      "loss": 1.6899,
+      "step": 560
+    },
+    {
+      "epoch": 0.3039826605256028,
+      "grad_norm": 0.1080511137843132,
+      "learning_rate": 0.0001,
+      "loss": 1.6783,
+      "step": 561
+    },
+    {
+      "epoch": 0.30452451910051476,
+      "grad_norm": 0.10125991702079773,
+      "learning_rate": 0.0001,
+      "loss": 1.6049,
+      "step": 562
+    },
+    {
+      "epoch": 0.3050663776754267,
+      "grad_norm": 0.10050351917743683,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 563
+    },
+    {
+      "epoch": 0.30560823625033867,
+      "grad_norm": 0.11019833385944366,
+      "learning_rate": 0.0001,
+      "loss": 1.6247,
+      "step": 564
+    },
+    {
+      "epoch": 0.30615009482525063,
+      "grad_norm": 0.09782030433416367,
+      "learning_rate": 0.0001,
+      "loss": 1.6296,
+      "step": 565
+    },
+    {
+      "epoch": 0.30669195340016253,
+      "grad_norm": 0.11155736446380615,
+      "learning_rate": 0.0001,
+      "loss": 1.6805,
+      "step": 566
+    },
+    {
+      "epoch": 0.3072338119750745,
+      "grad_norm": 0.10135099291801453,
+      "learning_rate": 0.0001,
+      "loss": 1.6303,
+      "step": 567
+    },
+    {
+      "epoch": 0.30777567054998645,
+      "grad_norm": 0.11518092453479767,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 568
+    },
+    {
+      "epoch": 0.3083175291248984,
+      "grad_norm": 0.09859606623649597,
+      "learning_rate": 0.0001,
+      "loss": 1.5585,
+      "step": 569
+    },
+    {
+      "epoch": 0.30885938769981036,
+      "grad_norm": 0.10348886996507645,
+      "learning_rate": 0.0001,
+      "loss": 1.6449,
+      "step": 570
+    },
+    {
+      "epoch": 0.3094012462747223,
+      "grad_norm": 0.11707218736410141,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 571
+    },
+    {
+      "epoch": 0.3099431048496342,
+      "grad_norm": 0.10389493405818939,
+      "learning_rate": 0.0001,
+      "loss": 1.6667,
+      "step": 572
+    },
+    {
+      "epoch": 0.3104849634245462,
+      "grad_norm": 0.11538241803646088,
+      "learning_rate": 0.0001,
+      "loss": 1.6343,
+      "step": 573
+    },
+    {
+      "epoch": 0.31102682199945814,
+      "grad_norm": 0.1050158217549324,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 574
+    },
+    {
+      "epoch": 0.3115686805743701,
+      "grad_norm": 0.11141128093004227,
+      "learning_rate": 0.0001,
+      "loss": 1.7506,
+      "step": 575
+    },
+    {
+      "epoch": 0.31211053914928205,
+      "grad_norm": 0.09710174798965454,
+      "learning_rate": 0.0001,
+      "loss": 1.6177,
+      "step": 576
+    },
+    {
+      "epoch": 0.312652397724194,
+      "grad_norm": 0.10418973118066788,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 577
+    },
+    {
+      "epoch": 0.3131942562991059,
+      "grad_norm": 0.10552489757537842,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 578
+    },
+    {
+      "epoch": 0.31373611487401787,
+      "grad_norm": 0.11075478047132492,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 579
+    },
+    {
+      "epoch": 0.3142779734489298,
+      "grad_norm": 0.09793159365653992,
+      "learning_rate": 0.0001,
+      "loss": 1.6635,
+      "step": 580
+    },
+    {
+      "epoch": 0.3148198320238418,
+      "grad_norm": 0.1039569228887558,
+      "learning_rate": 0.0001,
+      "loss": 1.6711,
+      "step": 581
+    },
+    {
+      "epoch": 0.31536169059875374,
+      "grad_norm": 0.1017010286450386,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 582
+    },
+    {
+      "epoch": 0.3159035491736657,
+      "grad_norm": 0.09644217044115067,
+      "learning_rate": 0.0001,
+      "loss": 1.5719,
+      "step": 583
+    },
+    {
+      "epoch": 0.3164454077485776,
+      "grad_norm": 0.0962899774312973,
+      "learning_rate": 0.0001,
+      "loss": 1.6344,
+      "step": 584
+    },
+    {
+      "epoch": 0.31698726632348956,
+      "grad_norm": 0.10612763464450836,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 585
+    },
+    {
+      "epoch": 0.3175291248984015,
+      "grad_norm": 0.10435491800308228,
+      "learning_rate": 0.0001,
+      "loss": 1.654,
+      "step": 586
+    },
+    {
+      "epoch": 0.3180709834733135,
+      "grad_norm": 0.1048903837800026,
+      "learning_rate": 0.0001,
+      "loss": 1.6114,
+      "step": 587
+    },
+    {
+      "epoch": 0.31861284204822543,
+      "grad_norm": 0.09729161113500595,
+      "learning_rate": 0.0001,
+      "loss": 1.5857,
+      "step": 588
+    },
+    {
+      "epoch": 0.3191547006231374,
+      "grad_norm": 0.1105666309595108,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 589
+    },
+    {
+      "epoch": 0.3196965591980493,
+      "grad_norm": 0.10651809722185135,
+      "learning_rate": 0.0001,
+      "loss": 1.6946,
+      "step": 590
+    },
+    {
+      "epoch": 0.32023841777296125,
+      "grad_norm": 0.09992777556180954,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 591
+    },
+    {
+      "epoch": 0.3207802763478732,
+      "grad_norm": 0.09944906085729599,
+      "learning_rate": 0.0001,
+      "loss": 1.5162,
+      "step": 592
+    },
+    {
+      "epoch": 0.32132213492278516,
+      "grad_norm": 0.10559218376874924,
+      "learning_rate": 0.0001,
+      "loss": 1.6372,
+      "step": 593
+    },
+    {
+      "epoch": 0.3218639934976971,
+      "grad_norm": 0.11096035689115524,
+      "learning_rate": 0.0001,
+      "loss": 1.6527,
+      "step": 594
+    },
+    {
+      "epoch": 0.322405852072609,
+      "grad_norm": 0.10335852205753326,
+      "learning_rate": 0.0001,
+      "loss": 1.6393,
+      "step": 595
+    },
+    {
+      "epoch": 0.322947710647521,
+      "grad_norm": 0.119107186794281,
+      "learning_rate": 0.0001,
+      "loss": 1.7471,
+      "step": 596
+    },
+    {
+      "epoch": 0.32348956922243294,
+      "grad_norm": 0.11370395123958588,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 597
+    },
+    {
+      "epoch": 0.3240314277973449,
+      "grad_norm": 0.11013813316822052,
+      "learning_rate": 0.0001,
+      "loss": 1.6508,
+      "step": 598
+    },
+    {
+      "epoch": 0.32457328637225685,
+      "grad_norm": 0.11540937423706055,
+      "learning_rate": 0.0001,
+      "loss": 1.6585,
+      "step": 599
+    },
+    {
+      "epoch": 0.3251151449471688,
+      "grad_norm": 0.10792079567909241,
+      "learning_rate": 0.0001,
+      "loss": 1.6451,
+      "step": 600
+    },
+    {
+      "epoch": 0.3256570035220807,
+      "grad_norm": 0.12482734024524689,
+      "learning_rate": 0.0001,
+      "loss": 1.6973,
+      "step": 601
+    },
+    {
+      "epoch": 0.32619886209699267,
+      "grad_norm": 0.10253299027681351,
+      "learning_rate": 0.0001,
+      "loss": 1.6808,
+      "step": 602
+    },
+    {
+      "epoch": 0.3267407206719046,
+      "grad_norm": 0.10867742449045181,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 603
+    },
+    {
+      "epoch": 0.3272825792468166,
+      "grad_norm": 0.10623064637184143,
+      "learning_rate": 0.0001,
+      "loss": 1.6831,
+      "step": 604
+    },
+    {
+      "epoch": 0.32782443782172854,
+      "grad_norm": 0.10389222949743271,
+      "learning_rate": 0.0001,
+      "loss": 1.6585,
+      "step": 605
+    },
+    {
+      "epoch": 0.3283662963966405,
+      "grad_norm": 0.10566440224647522,
+      "learning_rate": 0.0001,
+      "loss": 1.6995,
+      "step": 606
+    },
+    {
+      "epoch": 0.3289081549715524,
+      "grad_norm": 0.10380306094884872,
+      "learning_rate": 0.0001,
+      "loss": 1.6832,
+      "step": 607
+    },
+    {
+      "epoch": 0.32945001354646436,
+      "grad_norm": 0.10225769877433777,
+      "learning_rate": 0.0001,
+      "loss": 1.6459,
+      "step": 608
+    },
+    {
+      "epoch": 0.3299918721213763,
+      "grad_norm": 0.10640621185302734,
+      "learning_rate": 0.0001,
+      "loss": 1.6906,
+      "step": 609
+    },
+    {
+      "epoch": 0.3305337306962883,
+      "grad_norm": 0.10927685350179672,
+      "learning_rate": 0.0001,
+      "loss": 1.7528,
+      "step": 610
+    },
+    {
+      "epoch": 0.33107558927120023,
+      "grad_norm": 0.1068321168422699,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 611
+    },
+    {
+      "epoch": 0.3316174478461122,
+      "grad_norm": 0.1025937870144844,
+      "learning_rate": 0.0001,
+      "loss": 1.6462,
+      "step": 612
+    },
+    {
+      "epoch": 0.3321593064210241,
+      "grad_norm": 0.10666365176439285,
+      "learning_rate": 0.0001,
+      "loss": 1.6081,
+      "step": 613
+    },
+    {
+      "epoch": 0.33270116499593605,
+      "grad_norm": 0.10716580599546432,
+      "learning_rate": 0.0001,
+      "loss": 1.6932,
+      "step": 614
+    },
+    {
+      "epoch": 0.333243023570848,
+      "grad_norm": 0.10392767935991287,
+      "learning_rate": 0.0001,
+      "loss": 1.6816,
+      "step": 615
+    },
+    {
+      "epoch": 0.33378488214575996,
+      "grad_norm": 0.10266393423080444,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 616
+    },
+    {
+      "epoch": 0.3343267407206719,
+      "grad_norm": 0.09854996204376221,
+      "learning_rate": 0.0001,
+      "loss": 1.6061,
+      "step": 617
+    },
+    {
+      "epoch": 0.3348685992955839,
+      "grad_norm": 0.10747377574443817,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 618
+    },
+    {
+      "epoch": 0.3354104578704958,
+      "grad_norm": 0.09462318569421768,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 619
+    },
+    {
+      "epoch": 0.33595231644540774,
+      "grad_norm": 0.10060413926839828,
+      "learning_rate": 0.0001,
+      "loss": 1.6282,
+      "step": 620
+    },
+    {
+      "epoch": 0.3364941750203197,
+      "grad_norm": 0.10601072758436203,
+      "learning_rate": 0.0001,
+      "loss": 1.6738,
+      "step": 621
+    },
+    {
+      "epoch": 0.33703603359523165,
+      "grad_norm": 0.1055201068520546,
+      "learning_rate": 0.0001,
+      "loss": 1.6802,
+      "step": 622
+    },
+    {
+      "epoch": 0.3375778921701436,
+      "grad_norm": 0.10256335884332657,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 623
+    },
+    {
+      "epoch": 0.3381197507450555,
+      "grad_norm": 0.10397038608789444,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 624
+    },
+    {
+      "epoch": 0.33866160931996747,
+      "grad_norm": 0.10368067026138306,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 625
+    },
+    {
+      "epoch": 0.33920346789487943,
+      "grad_norm": 0.11170101165771484,
+      "learning_rate": 0.0001,
+      "loss": 1.6358,
+      "step": 626
+    },
+    {
+      "epoch": 0.3397453264697914,
+      "grad_norm": 0.10459307581186295,
+      "learning_rate": 0.0001,
+      "loss": 1.6929,
+      "step": 627
+    },
+    {
+      "epoch": 0.34028718504470334,
+      "grad_norm": 0.1107502356171608,
+      "learning_rate": 0.0001,
+      "loss": 1.6541,
+      "step": 628
+    },
+    {
+      "epoch": 0.3408290436196153,
+      "grad_norm": 0.10105384141206741,
+      "learning_rate": 0.0001,
+      "loss": 1.5745,
+      "step": 629
+    },
+    {
+      "epoch": 0.3413709021945272,
+      "grad_norm": 0.12728677690029144,
+      "learning_rate": 0.0001,
+      "loss": 1.6788,
+      "step": 630
+    },
+    {
+      "epoch": 0.34191276076943916,
+      "grad_norm": 0.11482071131467819,
+      "learning_rate": 0.0001,
+      "loss": 1.6842,
+      "step": 631
+    },
+    {
+      "epoch": 0.3424546193443511,
+      "grad_norm": 0.12361173331737518,
+      "learning_rate": 0.0001,
+      "loss": 1.6981,
+      "step": 632
+    },
+    {
+      "epoch": 0.3429964779192631,
+      "grad_norm": 0.09973514825105667,
+      "learning_rate": 0.0001,
+      "loss": 1.5814,
+      "step": 633
+    },
+    {
+      "epoch": 0.34353833649417503,
+      "grad_norm": 0.10917801409959793,
+      "learning_rate": 0.0001,
+      "loss": 1.6315,
+      "step": 634
+    },
+    {
+      "epoch": 0.344080195069087,
+      "grad_norm": 0.10635889321565628,
+      "learning_rate": 0.0001,
+      "loss": 1.7117,
+      "step": 635
+    },
+    {
+      "epoch": 0.3446220536439989,
+      "grad_norm": 0.10422027111053467,
+      "learning_rate": 0.0001,
+      "loss": 1.642,
+      "step": 636
+    },
+    {
+      "epoch": 0.34516391221891085,
+      "grad_norm": 0.10354586690664291,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 637
+    },
+    {
+      "epoch": 0.3457057707938228,
+      "grad_norm": 0.1029241606593132,
+      "learning_rate": 0.0001,
+      "loss": 1.6797,
+      "step": 638
+    },
+    {
+      "epoch": 0.34624762936873477,
+      "grad_norm": 0.11714927852153778,
+      "learning_rate": 0.0001,
+      "loss": 1.6231,
+      "step": 639
+    },
+    {
+      "epoch": 0.3467894879436467,
+      "grad_norm": 0.10488483309745789,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 640
+    },
+    {
+      "epoch": 0.3473313465185587,
+      "grad_norm": 0.11477353423833847,
+      "learning_rate": 0.0001,
+      "loss": 1.6654,
+      "step": 641
+    },
+    {
+      "epoch": 0.3478732050934706,
+      "grad_norm": 0.10494516044855118,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 642
+    },
+    {
+      "epoch": 0.34841506366838254,
+      "grad_norm": 0.11135809123516083,
+      "learning_rate": 0.0001,
+      "loss": 1.6487,
+      "step": 643
+    },
+    {
+      "epoch": 0.3489569222432945,
+      "grad_norm": 0.10758684575557709,
+      "learning_rate": 0.0001,
+      "loss": 1.6489,
+      "step": 644
+    },
+    {
+      "epoch": 0.34949878081820646,
+      "grad_norm": 0.1039474606513977,
+      "learning_rate": 0.0001,
+      "loss": 1.6933,
+      "step": 645
+    },
+    {
+      "epoch": 0.3500406393931184,
+      "grad_norm": 0.10481411963701248,
+      "learning_rate": 0.0001,
+      "loss": 1.649,
+      "step": 646
+    },
+    {
+      "epoch": 0.35058249796803037,
+      "grad_norm": 0.10531273484230042,
+      "learning_rate": 0.0001,
+      "loss": 1.586,
+      "step": 647
+    },
+    {
+      "epoch": 0.3511243565429423,
+      "grad_norm": 0.11362781375646591,
+      "learning_rate": 0.0001,
+      "loss": 1.6792,
+      "step": 648
+    },
+    {
+      "epoch": 0.35166621511785423,
+      "grad_norm": 0.10828311741352081,
+      "learning_rate": 0.0001,
+      "loss": 1.7087,
+      "step": 649
+    },
+    {
+      "epoch": 0.3522080736927662,
+      "grad_norm": 0.11995867639780045,
+      "learning_rate": 0.0001,
+      "loss": 1.6251,
+      "step": 650
+    },
+    {
+      "epoch": 0.35274993226767815,
+      "grad_norm": 0.10260593146085739,
+      "learning_rate": 0.0001,
+      "loss": 1.617,
+      "step": 651
+    },
+    {
+      "epoch": 0.3532917908425901,
+      "grad_norm": 0.11493167281150818,
+      "learning_rate": 0.0001,
+      "loss": 1.6041,
+      "step": 652
+    },
+    {
+      "epoch": 0.353833649417502,
+      "grad_norm": 0.11218332499265671,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 653
+    },
+    {
+      "epoch": 0.35437550799241396,
+      "grad_norm": 0.12246052175760269,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 654
+    },
+    {
+      "epoch": 0.3549173665673259,
+      "grad_norm": 0.1194012314081192,
+      "learning_rate": 0.0001,
+      "loss": 1.6542,
+      "step": 655
+    },
+    {
+      "epoch": 0.3554592251422379,
+      "grad_norm": 0.172341987490654,
+      "learning_rate": 0.0001,
+      "loss": 1.7179,
+      "step": 656
+    },
+    {
+      "epoch": 0.35600108371714984,
+      "grad_norm": 0.09841404110193253,
+      "learning_rate": 0.0001,
+      "loss": 1.5871,
+      "step": 657
+    },
+    {
+      "epoch": 0.3565429422920618,
+      "grad_norm": 0.1325990855693817,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 658
+    },
+    {
+      "epoch": 0.3570848008669737,
+      "grad_norm": 0.11859223991632462,
+      "learning_rate": 0.0001,
+      "loss": 1.6563,
+      "step": 659
+    },
+    {
+      "epoch": 0.35762665944188565,
+      "grad_norm": 0.1360592097043991,
+      "learning_rate": 0.0001,
+      "loss": 1.6154,
+      "step": 660
+    },
+    {
+      "epoch": 0.3581685180167976,
+      "grad_norm": 0.10509878396987915,
+      "learning_rate": 0.0001,
+      "loss": 1.6394,
+      "step": 661
+    },
+    {
+      "epoch": 0.35871037659170957,
+      "grad_norm": 0.11457318812608719,
+      "learning_rate": 0.0001,
+      "loss": 1.6816,
+      "step": 662
+    },
+    {
+      "epoch": 0.3592522351666215,
+      "grad_norm": 0.10837317258119583,
+      "learning_rate": 0.0001,
+      "loss": 1.714,
+      "step": 663
+    },
+    {
+      "epoch": 0.3597940937415335,
+      "grad_norm": 0.1057453602552414,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 664
+    },
+    {
+      "epoch": 0.3603359523164454,
+      "grad_norm": 0.10376082360744476,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 665
+    },
+    {
+      "epoch": 0.36087781089135734,
+      "grad_norm": 0.11164995282888412,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 666
+    },
+    {
+      "epoch": 0.3614196694662693,
+      "grad_norm": 0.10537731647491455,
+      "learning_rate": 0.0001,
+      "loss": 1.73,
+      "step": 667
+    },
+    {
+      "epoch": 0.36196152804118126,
+      "grad_norm": 0.1094699501991272,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 668
+    },
+    {
+      "epoch": 0.3625033866160932,
+      "grad_norm": 0.09954968839883804,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 669
+    },
+    {
+      "epoch": 0.3630452451910052,
+      "grad_norm": 0.10807305574417114,
+      "learning_rate": 0.0001,
+      "loss": 1.5642,
+      "step": 670
+    },
+    {
+      "epoch": 0.3635871037659171,
+      "grad_norm": 0.11197733134031296,
+      "learning_rate": 0.0001,
+      "loss": 1.6585,
+      "step": 671
+    },
+    {
+      "epoch": 0.36412896234082903,
+      "grad_norm": 0.11459947377443314,
+      "learning_rate": 0.0001,
+      "loss": 1.6864,
+      "step": 672
+    },
+    {
+      "epoch": 0.364670820915741,
+      "grad_norm": 0.11452600359916687,
+      "learning_rate": 0.0001,
+      "loss": 1.6939,
+      "step": 673
+    },
+    {
+      "epoch": 0.36521267949065295,
+      "grad_norm": 0.10195346176624298,
+      "learning_rate": 0.0001,
+      "loss": 1.6413,
+      "step": 674
+    },
+    {
+      "epoch": 0.3657545380655649,
+      "grad_norm": 0.11506009101867676,
+      "learning_rate": 0.0001,
+      "loss": 1.693,
+      "step": 675
+    },
+    {
+      "epoch": 0.36629639664047686,
+      "grad_norm": 0.10550053417682648,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 676
+    },
+    {
+      "epoch": 0.36683825521538876,
+      "grad_norm": 0.11007264256477356,
+      "learning_rate": 0.0001,
+      "loss": 1.6777,
+      "step": 677
+    },
+    {
+      "epoch": 0.3673801137903007,
+      "grad_norm": 0.10047126561403275,
+      "learning_rate": 0.0001,
+      "loss": 1.6317,
+      "step": 678
+    },
+    {
+      "epoch": 0.3679219723652127,
+      "grad_norm": 0.11524063348770142,
+      "learning_rate": 0.0001,
+      "loss": 1.6924,
+      "step": 679
+    },
+    {
+      "epoch": 0.36846383094012464,
+      "grad_norm": 0.10718086361885071,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 680
+    },
+    {
+      "epoch": 0.3690056895150366,
+      "grad_norm": 0.10243549197912216,
+      "learning_rate": 0.0001,
+      "loss": 1.6088,
+      "step": 681
+    },
+    {
+      "epoch": 0.3695475480899485,
+      "grad_norm": 0.1065065935254097,
+      "learning_rate": 0.0001,
+      "loss": 1.6379,
+      "step": 682
+    },
+    {
+      "epoch": 0.37008940666486045,
+      "grad_norm": 0.10459452867507935,
+      "learning_rate": 0.0001,
+      "loss": 1.6894,
+      "step": 683
+    },
+    {
+      "epoch": 0.3706312652397724,
+      "grad_norm": 0.10934063792228699,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 684
+    },
+    {
+      "epoch": 0.37117312381468437,
+      "grad_norm": 0.10245130956172943,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 685
+    },
+    {
+      "epoch": 0.3717149823895963,
+      "grad_norm": 0.10711225867271423,
+      "learning_rate": 0.0001,
+      "loss": 1.6609,
+      "step": 686
+    },
+    {
+      "epoch": 0.3722568409645083,
+      "grad_norm": 0.10337112098932266,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 687
+    },
+    {
+      "epoch": 0.3727986995394202,
+      "grad_norm": 0.11192546039819717,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 688
+    },
+    {
+      "epoch": 0.37334055811433214,
+      "grad_norm": 0.10183734446763992,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 689
+    },
+    {
+      "epoch": 0.3738824166892441,
+      "grad_norm": 0.11607497930526733,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 690
+    },
+    {
+      "epoch": 0.37442427526415606,
+      "grad_norm": 0.10997983813285828,
+      "learning_rate": 0.0001,
+      "loss": 1.6589,
+      "step": 691
+    },
+    {
+      "epoch": 0.374966133839068,
+      "grad_norm": 0.1069745421409607,
+      "learning_rate": 0.0001,
+      "loss": 1.5989,
+      "step": 692
+    },
+    {
+      "epoch": 0.37550799241398,
+      "grad_norm": 0.1032508909702301,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 693
+    },
+    {
+      "epoch": 0.3760498509888919,
+      "grad_norm": 0.10387293994426727,
+      "learning_rate": 0.0001,
+      "loss": 1.6175,
+      "step": 694
+    },
+    {
+      "epoch": 0.37659170956380383,
+      "grad_norm": 0.10035529732704163,
+      "learning_rate": 0.0001,
+      "loss": 1.6991,
+      "step": 695
+    },
+    {
+      "epoch": 0.3771335681387158,
+      "grad_norm": 0.1046760082244873,
+      "learning_rate": 0.0001,
+      "loss": 1.6928,
+      "step": 696
+    },
+    {
+      "epoch": 0.37767542671362775,
+      "grad_norm": 0.10290009528398514,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 697
+    },
+    {
+      "epoch": 0.3782172852885397,
+      "grad_norm": 0.0992322564125061,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 698
+    },
+    {
+      "epoch": 0.37875914386345166,
+      "grad_norm": 0.09865851700305939,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 699
+    },
+    {
+      "epoch": 0.37930100243836357,
+      "grad_norm": 0.09838226437568665,
+      "learning_rate": 0.0001,
+      "loss": 1.5628,
+      "step": 700
+    },
+    {
+      "epoch": 0.3798428610132755,
+      "grad_norm": 0.10080188512802124,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 701
+    },
+    {
+      "epoch": 0.3803847195881875,
+      "grad_norm": 0.10885747522115707,
+      "learning_rate": 0.0001,
+      "loss": 1.6453,
+      "step": 702
+    },
+    {
+      "epoch": 0.38092657816309944,
+      "grad_norm": 0.1085219532251358,
+      "learning_rate": 0.0001,
+      "loss": 1.6163,
+      "step": 703
+    },
+    {
+      "epoch": 0.3814684367380114,
+      "grad_norm": 0.10353940725326538,
+      "learning_rate": 0.0001,
+      "loss": 1.6174,
+      "step": 704
+    },
+    {
+      "epoch": 0.38201029531292335,
+      "grad_norm": 0.10917379707098007,
+      "learning_rate": 0.0001,
+      "loss": 1.663,
+      "step": 705
+    },
+    {
+      "epoch": 0.38255215388783526,
+      "grad_norm": 0.09883071482181549,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 706
+    },
+    {
+      "epoch": 0.3830940124627472,
+      "grad_norm": 0.11153585463762283,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 707
+    },
+    {
+      "epoch": 0.38363587103765917,
+      "grad_norm": 0.11329423636198044,
+      "learning_rate": 0.0001,
+      "loss": 1.6635,
+      "step": 708
+    },
+    {
+      "epoch": 0.38417772961257113,
+      "grad_norm": 0.11220398545265198,
+      "learning_rate": 0.0001,
+      "loss": 1.7422,
+      "step": 709
+    },
+    {
+      "epoch": 0.3847195881874831,
+      "grad_norm": 0.09743146598339081,
+      "learning_rate": 0.0001,
+      "loss": 1.5827,
+      "step": 710
+    },
+    {
+      "epoch": 0.385261446762395,
+      "grad_norm": 0.11069986969232559,
+      "learning_rate": 0.0001,
+      "loss": 1.6659,
+      "step": 711
+    },
+    {
+      "epoch": 0.38580330533730695,
+      "grad_norm": 0.11028224229812622,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 712
+    },
+    {
+      "epoch": 0.3863451639122189,
+      "grad_norm": 0.107399120926857,
+      "learning_rate": 0.0001,
+      "loss": 1.7154,
+      "step": 713
+    },
+    {
+      "epoch": 0.38688702248713086,
+      "grad_norm": 0.10606145858764648,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 714
+    },
+    {
+      "epoch": 0.3874288810620428,
+      "grad_norm": 0.09998262673616409,
+      "learning_rate": 0.0001,
+      "loss": 1.6103,
+      "step": 715
+    },
+    {
+      "epoch": 0.3879707396369548,
+      "grad_norm": 0.10442587733268738,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 716
+    },
+    {
+      "epoch": 0.3885125982118667,
+      "grad_norm": 0.10855881869792938,
+      "learning_rate": 0.0001,
+      "loss": 1.5742,
+      "step": 717
+    },
+    {
+      "epoch": 0.38905445678677864,
+      "grad_norm": 0.10240405052900314,
+      "learning_rate": 0.0001,
+      "loss": 1.6525,
+      "step": 718
+    },
+    {
+      "epoch": 0.3895963153616906,
+      "grad_norm": 0.10467260330915451,
+      "learning_rate": 0.0001,
+      "loss": 1.6496,
+      "step": 719
+    },
+    {
+      "epoch": 0.39013817393660255,
+      "grad_norm": 0.10330167412757874,
+      "learning_rate": 0.0001,
+      "loss": 1.6536,
+      "step": 720
+    },
+    {
+      "epoch": 0.3906800325115145,
+      "grad_norm": 0.10810812562704086,
+      "learning_rate": 0.0001,
+      "loss": 1.659,
+      "step": 721
+    },
+    {
+      "epoch": 0.39122189108642647,
+      "grad_norm": 0.12047295272350311,
+      "learning_rate": 0.0001,
+      "loss": 1.6462,
+      "step": 722
+    },
+    {
+      "epoch": 0.39176374966133837,
+      "grad_norm": 0.09939727187156677,
+      "learning_rate": 0.0001,
+      "loss": 1.6269,
+      "step": 723
+    },
+    {
+      "epoch": 0.3923056082362503,
+      "grad_norm": 0.11785425990819931,
+      "learning_rate": 0.0001,
+      "loss": 1.6571,
+      "step": 724
+    },
+    {
+      "epoch": 0.3928474668111623,
+      "grad_norm": 0.09766566753387451,
+      "learning_rate": 0.0001,
+      "loss": 1.5397,
+      "step": 725
+    },
+    {
+      "epoch": 0.39338932538607424,
+      "grad_norm": 0.10398256778717041,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 726
+    },
+    {
+      "epoch": 0.3939311839609862,
+      "grad_norm": 0.11227762699127197,
+      "learning_rate": 0.0001,
+      "loss": 1.6259,
+      "step": 727
+    },
+    {
+      "epoch": 0.39447304253589816,
+      "grad_norm": 0.10695900768041611,
+      "learning_rate": 0.0001,
+      "loss": 1.581,
+      "step": 728
+    },
+    {
+      "epoch": 0.39501490111081006,
+      "grad_norm": 0.11811427026987076,
+      "learning_rate": 0.0001,
+      "loss": 1.6366,
+      "step": 729
+    },
+    {
+      "epoch": 0.395556759685722,
+      "grad_norm": 0.10070157796144485,
+      "learning_rate": 0.0001,
+      "loss": 1.6585,
+      "step": 730
+    },
+    {
+      "epoch": 0.396098618260634,
+      "grad_norm": 0.13773372769355774,
+      "learning_rate": 0.0001,
+      "loss": 1.685,
+      "step": 731
+    },
+    {
+      "epoch": 0.39664047683554593,
+      "grad_norm": 0.10539493709802628,
+      "learning_rate": 0.0001,
+      "loss": 1.6605,
+      "step": 732
+    },
+    {
+      "epoch": 0.3971823354104579,
+      "grad_norm": 0.13001807034015656,
+      "learning_rate": 0.0001,
+      "loss": 1.6302,
+      "step": 733
+    },
+    {
+      "epoch": 0.39772419398536984,
+      "grad_norm": 0.10862381756305695,
+      "learning_rate": 0.0001,
+      "loss": 1.7,
+      "step": 734
+    },
+    {
+      "epoch": 0.39826605256028175,
+      "grad_norm": 0.14452537894248962,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 735
+    },
+    {
+      "epoch": 0.3988079111351937,
+      "grad_norm": 0.10327973961830139,
+      "learning_rate": 0.0001,
+      "loss": 1.6522,
+      "step": 736
+    },
+    {
+      "epoch": 0.39934976971010566,
+      "grad_norm": 0.13244539499282837,
+      "learning_rate": 0.0001,
+      "loss": 1.6315,
+      "step": 737
+    },
+    {
+      "epoch": 0.3998916282850176,
+      "grad_norm": 0.11024655401706696,
+      "learning_rate": 0.0001,
+      "loss": 1.6563,
+      "step": 738
+    },
+    {
+      "epoch": 0.4004334868599296,
+      "grad_norm": 0.13484923541545868,
+      "learning_rate": 0.0001,
+      "loss": 1.74,
+      "step": 739
+    },
+    {
+      "epoch": 0.4009753454348415,
+      "grad_norm": 0.12410994619131088,
+      "learning_rate": 0.0001,
+      "loss": 1.6311,
+      "step": 740
+    },
+    {
+      "epoch": 0.40151720400975344,
+      "grad_norm": 0.12810753285884857,
+      "learning_rate": 0.0001,
+      "loss": 1.6511,
+      "step": 741
+    },
+    {
+      "epoch": 0.4020590625846654,
+      "grad_norm": 0.12648116052150726,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 742
+    },
+    {
+      "epoch": 0.40260092115957735,
+      "grad_norm": 0.11238245666027069,
+      "learning_rate": 0.0001,
+      "loss": 1.646,
+      "step": 743
+    },
+    {
+      "epoch": 0.4031427797344893,
+      "grad_norm": 0.11832413822412491,
+      "learning_rate": 0.0001,
+      "loss": 1.6314,
+      "step": 744
+    },
+    {
+      "epoch": 0.40368463830940127,
+      "grad_norm": 0.10859525948762894,
+      "learning_rate": 0.0001,
+      "loss": 1.6764,
+      "step": 745
+    },
+    {
+      "epoch": 0.40422649688431317,
+      "grad_norm": 0.11857281625270844,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 746
+    },
+    {
+      "epoch": 0.4047683554592251,
+      "grad_norm": 0.10530444979667664,
+      "learning_rate": 0.0001,
+      "loss": 1.6318,
+      "step": 747
+    },
+    {
+      "epoch": 0.4053102140341371,
+      "grad_norm": 0.11615604907274246,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 748
+    },
+    {
+      "epoch": 0.40585207260904904,
+      "grad_norm": 0.10375487059354782,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 749
+    },
+    {
+      "epoch": 0.406393931183961,
+      "grad_norm": 0.11461639404296875,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 750
+    },
+    {
+      "epoch": 0.40693578975887296,
+      "grad_norm": 0.10505295544862747,
+      "learning_rate": 0.0001,
+      "loss": 1.6822,
+      "step": 751
+    },
+    {
+      "epoch": 0.40747764833378486,
+      "grad_norm": 0.10666048526763916,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 752
+    },
+    {
+      "epoch": 0.4080195069086968,
+      "grad_norm": 0.10326603800058365,
+      "learning_rate": 0.0001,
+      "loss": 1.5992,
+      "step": 753
+    },
+    {
+      "epoch": 0.4085613654836088,
+      "grad_norm": 0.10881656408309937,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 754
+    },
+    {
+      "epoch": 0.40910322405852073,
+      "grad_norm": 0.11879541724920273,
+      "learning_rate": 0.0001,
+      "loss": 1.6566,
+      "step": 755
+    },
+    {
+      "epoch": 0.4096450826334327,
+      "grad_norm": 0.10726792365312576,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 756
+    },
+    {
+      "epoch": 0.41018694120834465,
+      "grad_norm": 0.12982334196567535,
+      "learning_rate": 0.0001,
+      "loss": 1.614,
+      "step": 757
+    },
+    {
+      "epoch": 0.41072879978325655,
+      "grad_norm": 0.10279403626918793,
+      "learning_rate": 0.0001,
+      "loss": 1.6958,
+      "step": 758
+    },
+    {
+      "epoch": 0.4112706583581685,
+      "grad_norm": 0.13861803710460663,
+      "learning_rate": 0.0001,
+      "loss": 1.6631,
+      "step": 759
+    },
+    {
+      "epoch": 0.41181251693308046,
+      "grad_norm": 0.1008225530385971,
+      "learning_rate": 0.0001,
+      "loss": 1.6515,
+      "step": 760
+    },
+    {
+      "epoch": 0.4123543755079924,
+      "grad_norm": 0.11922475695610046,
+      "learning_rate": 0.0001,
+      "loss": 1.6635,
+      "step": 761
+    },
+    {
+      "epoch": 0.4128962340829044,
+      "grad_norm": 0.10964781790971756,
+      "learning_rate": 0.0001,
+      "loss": 1.6617,
+      "step": 762
+    },
+    {
+      "epoch": 0.41343809265781634,
+      "grad_norm": 0.11899418383836746,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 763
+    },
+    {
+      "epoch": 0.41397995123272824,
+      "grad_norm": 0.10138355940580368,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 764
+    },
+    {
+      "epoch": 0.4145218098076402,
+      "grad_norm": 0.10140690952539444,
+      "learning_rate": 0.0001,
+      "loss": 1.6245,
+      "step": 765
+    },
+    {
+      "epoch": 0.41506366838255215,
+      "grad_norm": 0.21386539936065674,
+      "learning_rate": 0.0001,
+      "loss": 1.6831,
+      "step": 766
+    },
+    {
+      "epoch": 0.4156055269574641,
+      "grad_norm": 0.10448230057954788,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 767
+    },
+    {
+      "epoch": 0.41614738553237607,
+      "grad_norm": 0.12657177448272705,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 768
+    },
+    {
+      "epoch": 0.41668924410728797,
+      "grad_norm": 0.10476984828710556,
+      "learning_rate": 0.0001,
+      "loss": 1.6288,
+      "step": 769
+    },
+    {
+      "epoch": 0.41723110268219993,
+      "grad_norm": 0.11148498952388763,
+      "learning_rate": 0.0001,
+      "loss": 1.7032,
+      "step": 770
+    },
+    {
+      "epoch": 0.4177729612571119,
+      "grad_norm": 0.11308016628026962,
+      "learning_rate": 0.0001,
+      "loss": 1.7415,
+      "step": 771
+    },
+    {
+      "epoch": 0.41831481983202384,
+      "grad_norm": 0.11244909465312958,
+      "learning_rate": 0.0001,
+      "loss": 1.6454,
+      "step": 772
+    },
+    {
+      "epoch": 0.4188566784069358,
+      "grad_norm": 0.11405385285615921,
+      "learning_rate": 0.0001,
+      "loss": 1.6767,
+      "step": 773
+    },
+    {
+      "epoch": 0.41939853698184776,
+      "grad_norm": 0.10433017462491989,
+      "learning_rate": 0.0001,
+      "loss": 1.641,
+      "step": 774
+    },
+    {
+      "epoch": 0.41994039555675966,
+      "grad_norm": 0.11652455478906631,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 775
+    },
+    {
+      "epoch": 0.4204822541316716,
+      "grad_norm": 0.1095375195145607,
+      "learning_rate": 0.0001,
+      "loss": 1.6272,
+      "step": 776
+    },
+    {
+      "epoch": 0.4210241127065836,
+      "grad_norm": 0.12977243959903717,
+      "learning_rate": 0.0001,
+      "loss": 1.6794,
+      "step": 777
+    },
+    {
+      "epoch": 0.42156597128149553,
+      "grad_norm": 0.10516102612018585,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 778
+    },
+    {
+      "epoch": 0.4221078298564075,
+      "grad_norm": 0.12051164358854294,
+      "learning_rate": 0.0001,
+      "loss": 1.7294,
+      "step": 779
+    },
+    {
+      "epoch": 0.42264968843131945,
+      "grad_norm": 0.1023736223578453,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 780
+    },
+    {
+      "epoch": 0.42319154700623135,
+      "grad_norm": 0.11457734555006027,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 781
+    },
+    {
+      "epoch": 0.4237334055811433,
+      "grad_norm": 0.11762605607509613,
+      "learning_rate": 0.0001,
+      "loss": 1.7058,
+      "step": 782
+    },
+    {
+      "epoch": 0.42427526415605527,
+      "grad_norm": 0.10679729282855988,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 783
+    },
+    {
+      "epoch": 0.4248171227309672,
+      "grad_norm": 0.11970070749521255,
+      "learning_rate": 0.0001,
+      "loss": 1.6607,
+      "step": 784
+    },
+    {
+      "epoch": 0.4253589813058792,
+      "grad_norm": 0.10416082292795181,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 785
+    },
+    {
+      "epoch": 0.42590083988079114,
+      "grad_norm": 0.12200033664703369,
+      "learning_rate": 0.0001,
+      "loss": 1.5386,
+      "step": 786
+    },
+    {
+      "epoch": 0.42644269845570304,
+      "grad_norm": 0.10045525431632996,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 787
+    },
+    {
+      "epoch": 0.426984557030615,
+      "grad_norm": 0.12497124820947647,
+      "learning_rate": 0.0001,
+      "loss": 1.6139,
+      "step": 788
+    },
+    {
+      "epoch": 0.42752641560552695,
+      "grad_norm": 0.10556544363498688,
+      "learning_rate": 0.0001,
+      "loss": 1.5744,
+      "step": 789
+    },
+    {
+      "epoch": 0.4280682741804389,
+      "grad_norm": 0.12096355110406876,
+      "learning_rate": 0.0001,
+      "loss": 1.6924,
+      "step": 790
+    },
+    {
+      "epoch": 0.42861013275535087,
+      "grad_norm": 0.10694337636232376,
+      "learning_rate": 0.0001,
+      "loss": 1.6428,
+      "step": 791
+    },
+    {
+      "epoch": 0.4291519913302628,
+      "grad_norm": 0.1038854718208313,
+      "learning_rate": 0.0001,
+      "loss": 1.6072,
+      "step": 792
+    },
+    {
+      "epoch": 0.42969384990517473,
+      "grad_norm": 0.11172711104154587,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 793
+    },
+    {
+      "epoch": 0.4302357084800867,
+      "grad_norm": 0.0996694564819336,
+      "learning_rate": 0.0001,
+      "loss": 1.5751,
+      "step": 794
+    },
+    {
+      "epoch": 0.43077756705499864,
+      "grad_norm": 0.11230441927909851,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 795
+    },
+    {
+      "epoch": 0.4313194256299106,
+      "grad_norm": 0.10604969412088394,
+      "learning_rate": 0.0001,
+      "loss": 1.6912,
+      "step": 796
+    },
+    {
+      "epoch": 0.43186128420482256,
+      "grad_norm": 0.10176172107458115,
+      "learning_rate": 0.0001,
+      "loss": 1.6294,
+      "step": 797
+    },
+    {
+      "epoch": 0.43240314277973446,
+      "grad_norm": 0.10164622962474823,
+      "learning_rate": 0.0001,
+      "loss": 1.6278,
+      "step": 798
+    },
+    {
+      "epoch": 0.4329450013546464,
+      "grad_norm": 0.10655150562524796,
+      "learning_rate": 0.0001,
+      "loss": 1.6499,
+      "step": 799
+    },
+    {
+      "epoch": 0.4334868599295584,
+      "grad_norm": 0.10253231972455978,
+      "learning_rate": 0.0001,
+      "loss": 1.7011,
+      "step": 800
+    },
+    {
+      "epoch": 0.43402871850447033,
+      "grad_norm": 0.10146262496709824,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 801
+    },
+    {
+      "epoch": 0.4345705770793823,
+      "grad_norm": 0.11094497889280319,
+      "learning_rate": 0.0001,
+      "loss": 1.6622,
+      "step": 802
+    },
+    {
+      "epoch": 0.43511243565429425,
+      "grad_norm": 0.1083596721291542,
+      "learning_rate": 0.0001,
+      "loss": 1.6403,
+      "step": 803
+    },
+    {
+      "epoch": 0.43565429422920615,
+      "grad_norm": 0.11484643816947937,
+      "learning_rate": 0.0001,
+      "loss": 1.7213,
+      "step": 804
+    },
+    {
+      "epoch": 0.4361961528041181,
+      "grad_norm": 0.10802962630987167,
+      "learning_rate": 0.0001,
+      "loss": 1.6285,
+      "step": 805
+    },
+    {
+      "epoch": 0.43673801137903007,
+      "grad_norm": 0.10771377384662628,
+      "learning_rate": 0.0001,
+      "loss": 1.7367,
+      "step": 806
+    },
+    {
+      "epoch": 0.437279869953942,
+      "grad_norm": 0.10246739536523819,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 807
+    },
+    {
+      "epoch": 0.437821728528854,
+      "grad_norm": 0.09854665398597717,
+      "learning_rate": 0.0001,
+      "loss": 1.5917,
+      "step": 808
+    },
+    {
+      "epoch": 0.43836358710376594,
+      "grad_norm": 0.1048893928527832,
+      "learning_rate": 0.0001,
+      "loss": 1.6388,
+      "step": 809
+    },
+    {
+      "epoch": 0.43890544567867784,
+      "grad_norm": 0.09986301511526108,
+      "learning_rate": 0.0001,
+      "loss": 1.6296,
+      "step": 810
+    },
+    {
+      "epoch": 0.4394473042535898,
+      "grad_norm": 0.10390371829271317,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 811
+    },
+    {
+      "epoch": 0.43998916282850176,
+      "grad_norm": 0.104039765894413,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 812
+    },
+    {
+      "epoch": 0.4405310214034137,
+      "grad_norm": 0.10459230840206146,
+      "learning_rate": 0.0001,
+      "loss": 1.6717,
+      "step": 813
+    },
+    {
+      "epoch": 0.44107287997832567,
+      "grad_norm": 0.10667424649000168,
+      "learning_rate": 0.0001,
+      "loss": 1.6662,
+      "step": 814
+    },
+    {
+      "epoch": 0.44161473855323763,
+      "grad_norm": 0.10749981552362442,
+      "learning_rate": 0.0001,
+      "loss": 1.6664,
+      "step": 815
+    },
+    {
+      "epoch": 0.44215659712814953,
+      "grad_norm": 0.10850109905004501,
+      "learning_rate": 0.0001,
+      "loss": 1.6232,
+      "step": 816
+    },
+    {
+      "epoch": 0.4426984557030615,
+      "grad_norm": 0.10601247102022171,
+      "learning_rate": 0.0001,
+      "loss": 1.6744,
+      "step": 817
+    },
+    {
+      "epoch": 0.44324031427797345,
+      "grad_norm": 0.10685880482196808,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 818
+    },
+    {
+      "epoch": 0.4437821728528854,
+      "grad_norm": 0.10400949418544769,
+      "learning_rate": 0.0001,
+      "loss": 1.624,
+      "step": 819
+    },
+    {
+      "epoch": 0.44432403142779736,
+      "grad_norm": 0.117488332092762,
+      "learning_rate": 0.0001,
+      "loss": 1.7304,
+      "step": 820
+    },
+    {
+      "epoch": 0.4448658900027093,
+      "grad_norm": 0.1075533851981163,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 821
+    },
+    {
+      "epoch": 0.4454077485776212,
+      "grad_norm": 0.1162843331694603,
+      "learning_rate": 0.0001,
+      "loss": 1.679,
+      "step": 822
+    },
+    {
+      "epoch": 0.4459496071525332,
+      "grad_norm": 0.10688474774360657,
+      "learning_rate": 0.0001,
+      "loss": 1.5606,
+      "step": 823
+    },
+    {
+      "epoch": 0.44649146572744514,
+      "grad_norm": 0.1137152761220932,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 824
+    },
+    {
+      "epoch": 0.4470333243023571,
+      "grad_norm": 0.10783565044403076,
+      "learning_rate": 0.0001,
+      "loss": 1.6221,
+      "step": 825
+    },
+    {
+      "epoch": 0.44757518287726905,
+      "grad_norm": 0.11750900000333786,
+      "learning_rate": 0.0001,
+      "loss": 1.6704,
+      "step": 826
+    },
+    {
+      "epoch": 0.44811704145218095,
+      "grad_norm": 0.1140749379992485,
+      "learning_rate": 0.0001,
+      "loss": 1.6463,
+      "step": 827
+    },
+    {
+      "epoch": 0.4486589000270929,
+      "grad_norm": 0.10287487506866455,
+      "learning_rate": 0.0001,
+      "loss": 1.5714,
+      "step": 828
+    },
+    {
+      "epoch": 0.44920075860200487,
+      "grad_norm": 0.123508021235466,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 829
+    },
+    {
+      "epoch": 0.4497426171769168,
+      "grad_norm": 0.10874208062887192,
+      "learning_rate": 0.0001,
+      "loss": 1.5999,
+      "step": 830
+    },
+    {
+      "epoch": 0.4502844757518288,
+      "grad_norm": 0.12272533774375916,
+      "learning_rate": 0.0001,
+      "loss": 1.6757,
+      "step": 831
+    },
+    {
+      "epoch": 0.45082633432674074,
+      "grad_norm": 0.1147417426109314,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 832
+    },
+    {
+      "epoch": 0.45136819290165264,
+      "grad_norm": 0.10647159069776535,
+      "learning_rate": 0.0001,
+      "loss": 1.6185,
+      "step": 833
+    },
+    {
+      "epoch": 0.4519100514765646,
+      "grad_norm": 0.12309867888689041,
+      "learning_rate": 0.0001,
+      "loss": 1.6594,
+      "step": 834
+    },
+    {
+      "epoch": 0.45245191005147656,
+      "grad_norm": 0.11148115247488022,
+      "learning_rate": 0.0001,
+      "loss": 1.6481,
+      "step": 835
+    },
+    {
+      "epoch": 0.4529937686263885,
+      "grad_norm": 0.10988766700029373,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 836
+    },
+    {
+      "epoch": 0.4535356272013005,
+      "grad_norm": 0.1154412180185318,
+      "learning_rate": 0.0001,
+      "loss": 1.6701,
+      "step": 837
+    },
+    {
+      "epoch": 0.45407748577621243,
+      "grad_norm": 0.1137893870472908,
+      "learning_rate": 0.0001,
+      "loss": 1.6679,
+      "step": 838
+    },
+    {
+      "epoch": 0.45461934435112433,
+      "grad_norm": 0.10951977968215942,
+      "learning_rate": 0.0001,
+      "loss": 1.663,
+      "step": 839
+    },
+    {
+      "epoch": 0.4551612029260363,
+      "grad_norm": 0.104254350066185,
+      "learning_rate": 0.0001,
+      "loss": 1.592,
+      "step": 840
+    },
+    {
+      "epoch": 0.45570306150094825,
+      "grad_norm": 0.10429386794567108,
+      "learning_rate": 0.0001,
+      "loss": 1.6041,
+      "step": 841
+    },
+    {
+      "epoch": 0.4562449200758602,
+      "grad_norm": 0.1142311543226242,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 842
+    },
+    {
+      "epoch": 0.45678677865077216,
+      "grad_norm": 0.10919320583343506,
+      "learning_rate": 0.0001,
+      "loss": 1.67,
+      "step": 843
+    },
+    {
+      "epoch": 0.4573286372256841,
+      "grad_norm": 0.11463817954063416,
+      "learning_rate": 0.0001,
+      "loss": 1.6659,
+      "step": 844
+    },
+    {
+      "epoch": 0.457870495800596,
+      "grad_norm": 0.11288406699895859,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 845
+    },
+    {
+      "epoch": 0.458412354375508,
+      "grad_norm": 0.11115078628063202,
+      "learning_rate": 0.0001,
+      "loss": 1.5919,
+      "step": 846
+    },
+    {
+      "epoch": 0.45895421295041994,
+      "grad_norm": 0.12078949809074402,
+      "learning_rate": 0.0001,
+      "loss": 1.6096,
+      "step": 847
+    },
+    {
+      "epoch": 0.4594960715253319,
+      "grad_norm": 0.10990909487009048,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 848
+    },
+    {
+      "epoch": 0.46003793010024385,
+      "grad_norm": 0.1061215028166771,
+      "learning_rate": 0.0001,
+      "loss": 1.5879,
+      "step": 849
+    },
+    {
+      "epoch": 0.4605797886751558,
+      "grad_norm": 0.11332777887582779,
+      "learning_rate": 0.0001,
+      "loss": 1.6653,
+      "step": 850
+    },
+    {
+      "epoch": 0.4611216472500677,
+      "grad_norm": 0.1126977950334549,
+      "learning_rate": 0.0001,
+      "loss": 1.5694,
+      "step": 851
+    },
+    {
+      "epoch": 0.46166350582497967,
+      "grad_norm": 0.10927823930978775,
+      "learning_rate": 0.0001,
+      "loss": 1.6078,
+      "step": 852
+    },
+    {
+      "epoch": 0.4622053643998916,
+      "grad_norm": 0.11738968640565872,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 853
+    },
+    {
+      "epoch": 0.4627472229748036,
+      "grad_norm": 0.10907457768917084,
+      "learning_rate": 0.0001,
+      "loss": 1.585,
+      "step": 854
+    },
+    {
+      "epoch": 0.46328908154971554,
+      "grad_norm": 0.12172666192054749,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 855
+    },
+    {
+      "epoch": 0.4638309401246275,
+      "grad_norm": 0.10580083727836609,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 856
+    },
+    {
+      "epoch": 0.4643727986995394,
+      "grad_norm": 0.13606320321559906,
+      "learning_rate": 0.0001,
+      "loss": 1.6021,
+      "step": 857
+    },
+    {
+      "epoch": 0.46491465727445136,
+      "grad_norm": 0.11144983768463135,
+      "learning_rate": 0.0001,
+      "loss": 1.6698,
+      "step": 858
+    },
+    {
+      "epoch": 0.4654565158493633,
+      "grad_norm": 0.11384718120098114,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 859
+    },
+    {
+      "epoch": 0.4659983744242753,
+      "grad_norm": 0.10568007081747055,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 860
+    },
+    {
+      "epoch": 0.46654023299918723,
+      "grad_norm": 0.12521082162857056,
+      "learning_rate": 0.0001,
+      "loss": 1.6502,
+      "step": 861
+    },
+    {
+      "epoch": 0.46708209157409913,
+      "grad_norm": 0.1050972193479538,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 862
+    },
+    {
+      "epoch": 0.4676239501490111,
+      "grad_norm": 0.11627109348773956,
+      "learning_rate": 0.0001,
+      "loss": 1.6502,
+      "step": 863
+    },
+    {
+      "epoch": 0.46816580872392305,
+      "grad_norm": 0.10397864878177643,
+      "learning_rate": 0.0001,
+      "loss": 1.5968,
+      "step": 864
+    },
+    {
+      "epoch": 0.468707667298835,
+      "grad_norm": 0.12034126371145248,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 865
+    },
+    {
+      "epoch": 0.46924952587374696,
+      "grad_norm": 0.10460563004016876,
+      "learning_rate": 0.0001,
+      "loss": 1.6406,
+      "step": 866
+    },
+    {
+      "epoch": 0.4697913844486589,
+      "grad_norm": 0.10565747320652008,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 867
+    },
+    {
+      "epoch": 0.4703332430235708,
+      "grad_norm": 0.10394787043333054,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 868
+    },
+    {
+      "epoch": 0.4708751015984828,
+      "grad_norm": 0.10103076696395874,
+      "learning_rate": 0.0001,
+      "loss": 1.5833,
+      "step": 869
+    },
+    {
+      "epoch": 0.47141696017339474,
+      "grad_norm": 0.10692822188138962,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 870
+    },
+    {
+      "epoch": 0.4719588187483067,
+      "grad_norm": 0.10232596099376678,
+      "learning_rate": 0.0001,
+      "loss": 1.5858,
+      "step": 871
+    },
+    {
+      "epoch": 0.47250067732321865,
+      "grad_norm": 0.11097247153520584,
+      "learning_rate": 0.0001,
+      "loss": 1.6053,
+      "step": 872
+    },
+    {
+      "epoch": 0.4730425358981306,
+      "grad_norm": 0.10421837866306305,
+      "learning_rate": 0.0001,
+      "loss": 1.6764,
+      "step": 873
+    },
+    {
+      "epoch": 0.4735843944730425,
+      "grad_norm": 0.10425719618797302,
+      "learning_rate": 0.0001,
+      "loss": 1.622,
+      "step": 874
+    },
+    {
+      "epoch": 0.47412625304795447,
+      "grad_norm": 0.10385365039110184,
+      "learning_rate": 0.0001,
+      "loss": 1.629,
+      "step": 875
+    },
+    {
+      "epoch": 0.47466811162286643,
+      "grad_norm": 0.11015225201845169,
+      "learning_rate": 0.0001,
+      "loss": 1.6498,
+      "step": 876
+    },
+    {
+      "epoch": 0.4752099701977784,
+      "grad_norm": 0.10669953376054764,
+      "learning_rate": 0.0001,
+      "loss": 1.6489,
+      "step": 877
+    },
+    {
+      "epoch": 0.47575182877269034,
+      "grad_norm": 0.10671980679035187,
+      "learning_rate": 0.0001,
+      "loss": 1.6483,
+      "step": 878
+    },
+    {
+      "epoch": 0.4762936873476023,
+      "grad_norm": 0.10373824834823608,
+      "learning_rate": 0.0001,
+      "loss": 1.5758,
+      "step": 879
+    },
+    {
+      "epoch": 0.4768355459225142,
+      "grad_norm": 0.10339619219303131,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 880
+    },
+    {
+      "epoch": 0.47737740449742616,
+      "grad_norm": 0.12287923693656921,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 881
+    },
+    {
+      "epoch": 0.4779192630723381,
+      "grad_norm": 0.11370906978845596,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 882
+    },
+    {
+      "epoch": 0.4784611216472501,
+      "grad_norm": 0.10995329171419144,
+      "learning_rate": 0.0001,
+      "loss": 1.6489,
+      "step": 883
+    },
+    {
+      "epoch": 0.47900298022216203,
+      "grad_norm": 0.11137809604406357,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 884
+    },
+    {
+      "epoch": 0.479544838797074,
+      "grad_norm": 0.11190925538539886,
+      "learning_rate": 0.0001,
+      "loss": 1.6146,
+      "step": 885
+    },
+    {
+      "epoch": 0.4800866973719859,
+      "grad_norm": 0.10670820623636246,
+      "learning_rate": 0.0001,
+      "loss": 1.6122,
+      "step": 886
+    },
+    {
+      "epoch": 0.48062855594689785,
+      "grad_norm": 0.12012673169374466,
+      "learning_rate": 0.0001,
+      "loss": 1.6197,
+      "step": 887
+    },
+    {
+      "epoch": 0.4811704145218098,
+      "grad_norm": 0.11470359563827515,
+      "learning_rate": 0.0001,
+      "loss": 1.6321,
+      "step": 888
+    },
+    {
+      "epoch": 0.48171227309672177,
+      "grad_norm": 0.1273769736289978,
+      "learning_rate": 0.0001,
+      "loss": 1.6864,
+      "step": 889
+    },
+    {
+      "epoch": 0.4822541316716337,
+      "grad_norm": 0.11515619605779648,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 890
+    },
+    {
+      "epoch": 0.4827959902465456,
+      "grad_norm": 0.12130913883447647,
+      "learning_rate": 0.0001,
+      "loss": 1.6166,
+      "step": 891
+    },
+    {
+      "epoch": 0.4833378488214576,
+      "grad_norm": 0.11784867197275162,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 892
+    },
+    {
+      "epoch": 0.48387970739636954,
+      "grad_norm": 0.10914217680692673,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 893
+    },
+    {
+      "epoch": 0.4844215659712815,
+      "grad_norm": 0.11672643572092056,
+      "learning_rate": 0.0001,
+      "loss": 1.7317,
+      "step": 894
+    },
+    {
+      "epoch": 0.48496342454619346,
+      "grad_norm": 0.11123480647802353,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 895
+    },
+    {
+      "epoch": 0.4855052831211054,
+      "grad_norm": 0.10239136964082718,
+      "learning_rate": 0.0001,
+      "loss": 1.5774,
+      "step": 896
+    },
+    {
+      "epoch": 0.4860471416960173,
+      "grad_norm": 0.12416405975818634,
+      "learning_rate": 0.0001,
+      "loss": 1.7,
+      "step": 897
+    },
+    {
+      "epoch": 0.4865890002709293,
+      "grad_norm": 0.11070729047060013,
+      "learning_rate": 0.0001,
+      "loss": 1.6748,
+      "step": 898
+    },
+    {
+      "epoch": 0.48713085884584123,
+      "grad_norm": 0.13239996135234833,
+      "learning_rate": 0.0001,
+      "loss": 1.5977,
+      "step": 899
+    },
+    {
+      "epoch": 0.4876727174207532,
+      "grad_norm": 0.10215065628290176,
+      "learning_rate": 0.0001,
+      "loss": 1.5905,
+      "step": 900
+    },
+    {
+      "epoch": 0.48821457599566515,
+      "grad_norm": 0.12313413619995117,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 901
+    },
+    {
+      "epoch": 0.4887564345705771,
+      "grad_norm": 0.11618530005216599,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 902
+    },
+    {
+      "epoch": 0.489298293145489,
+      "grad_norm": 0.11647929251194,
+      "learning_rate": 0.0001,
+      "loss": 1.6813,
+      "step": 903
+    },
+    {
+      "epoch": 0.48984015172040096,
+      "grad_norm": 0.12761937081813812,
+      "learning_rate": 0.0001,
+      "loss": 1.6652,
+      "step": 904
+    },
+    {
+      "epoch": 0.4903820102953129,
+      "grad_norm": 0.13200801610946655,
+      "learning_rate": 0.0001,
+      "loss": 1.7227,
+      "step": 905
+    },
+    {
+      "epoch": 0.4909238688702249,
+      "grad_norm": 0.11139468103647232,
+      "learning_rate": 0.0001,
+      "loss": 1.6142,
+      "step": 906
+    },
+    {
+      "epoch": 0.49146572744513684,
+      "grad_norm": 0.11942821741104126,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 907
+    },
+    {
+      "epoch": 0.4920075860200488,
+      "grad_norm": 0.11830473691225052,
+      "learning_rate": 0.0001,
+      "loss": 1.6957,
+      "step": 908
+    },
+    {
+      "epoch": 0.4925494445949607,
+      "grad_norm": 0.12805293500423431,
+      "learning_rate": 0.0001,
+      "loss": 1.7355,
+      "step": 909
+    },
+    {
+      "epoch": 0.49309130316987265,
+      "grad_norm": 0.1151934266090393,
+      "learning_rate": 0.0001,
+      "loss": 1.5538,
+      "step": 910
+    },
+    {
+      "epoch": 0.4936331617447846,
+      "grad_norm": 0.11872893571853638,
+      "learning_rate": 0.0001,
+      "loss": 1.6597,
+      "step": 911
+    },
+    {
+      "epoch": 0.49417502031969657,
+      "grad_norm": 0.11597079783678055,
+      "learning_rate": 0.0001,
+      "loss": 1.683,
+      "step": 912
+    },
+    {
+      "epoch": 0.4947168788946085,
+      "grad_norm": 0.10476577281951904,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 913
+    },
+    {
+      "epoch": 0.4952587374695205,
+      "grad_norm": 0.12531496584415436,
+      "learning_rate": 0.0001,
+      "loss": 1.6894,
+      "step": 914
+    },
+    {
+      "epoch": 0.4958005960444324,
+      "grad_norm": 0.10528986901044846,
+      "learning_rate": 0.0001,
+      "loss": 1.6081,
+      "step": 915
+    },
+    {
+      "epoch": 0.49634245461934434,
+      "grad_norm": 0.12118838727474213,
+      "learning_rate": 0.0001,
+      "loss": 1.6453,
+      "step": 916
+    },
+    {
+      "epoch": 0.4968843131942563,
+      "grad_norm": 0.11182791739702225,
+      "learning_rate": 0.0001,
+      "loss": 1.6216,
+      "step": 917
+    },
+    {
+      "epoch": 0.49742617176916826,
+      "grad_norm": 0.11458185315132141,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 918
+    },
+    {
+      "epoch": 0.4979680303440802,
+      "grad_norm": 0.10385309159755707,
+      "learning_rate": 0.0001,
+      "loss": 1.5791,
+      "step": 919
+    },
+    {
+      "epoch": 0.4985098889189921,
+      "grad_norm": 0.112492136657238,
+      "learning_rate": 0.0001,
+      "loss": 1.5542,
+      "step": 920
+    },
+    {
+      "epoch": 0.4990517474939041,
+      "grad_norm": 0.10807826370000839,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 921
+    },
+    {
+      "epoch": 0.49959360606881603,
+      "grad_norm": 0.10996340960264206,
+      "learning_rate": 0.0001,
+      "loss": 1.5915,
+      "step": 922
+    },
+    {
+      "epoch": 0.5001354646437279,
+      "grad_norm": 0.12523028254508972,
+      "learning_rate": 0.0001,
+      "loss": 1.6721,
+      "step": 923
+    },
+    {
+      "epoch": 0.5006773232186399,
+      "grad_norm": 0.10851467400789261,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 924
+    },
+    {
+      "epoch": 0.5012191817935518,
+      "grad_norm": 0.12981802225112915,
+      "learning_rate": 0.0001,
+      "loss": 1.6243,
+      "step": 925
+    },
+    {
+      "epoch": 0.5017610403684638,
+      "grad_norm": 0.1098252385854721,
+      "learning_rate": 0.0001,
+      "loss": 1.6871,
+      "step": 926
+    },
+    {
+      "epoch": 0.5023028989433758,
+      "grad_norm": 0.1102382093667984,
+      "learning_rate": 0.0001,
+      "loss": 1.6117,
+      "step": 927
+    },
+    {
+      "epoch": 0.5028447575182877,
+      "grad_norm": 0.11334949731826782,
+      "learning_rate": 0.0001,
+      "loss": 1.6496,
+      "step": 928
+    },
+    {
+      "epoch": 0.5033866160931997,
+      "grad_norm": 0.12077392637729645,
+      "learning_rate": 0.0001,
+      "loss": 1.534,
+      "step": 929
+    },
+    {
+      "epoch": 0.5039284746681116,
+      "grad_norm": 0.12375643104314804,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 930
+    },
+    {
+      "epoch": 0.5044703332430236,
+      "grad_norm": 0.11033709347248077,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 931
+    },
+    {
+      "epoch": 0.5050121918179356,
+      "grad_norm": 0.14901308715343475,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 932
+    },
+    {
+      "epoch": 0.5055540503928475,
+      "grad_norm": 0.10846343636512756,
+      "learning_rate": 0.0001,
+      "loss": 1.6386,
+      "step": 933
+    },
+    {
+      "epoch": 0.5060959089677595,
+      "grad_norm": 0.13782711327075958,
+      "learning_rate": 0.0001,
+      "loss": 1.6564,
+      "step": 934
+    },
+    {
+      "epoch": 0.5066377675426713,
+      "grad_norm": 0.11779604107141495,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 935
+    },
+    {
+      "epoch": 0.5071796261175833,
+      "grad_norm": 0.11643415689468384,
+      "learning_rate": 0.0001,
+      "loss": 1.6694,
+      "step": 936
+    },
+    {
+      "epoch": 0.5077214846924952,
+      "grad_norm": 0.13498874008655548,
+      "learning_rate": 0.0001,
+      "loss": 1.5645,
+      "step": 937
+    },
+    {
+      "epoch": 0.5082633432674072,
+      "grad_norm": 0.11232909560203552,
+      "learning_rate": 0.0001,
+      "loss": 1.6849,
+      "step": 938
+    },
+    {
+      "epoch": 0.5088052018423191,
+      "grad_norm": 0.1520954966545105,
+      "learning_rate": 0.0001,
+      "loss": 1.7269,
+      "step": 939
+    },
+    {
+      "epoch": 0.5093470604172311,
+      "grad_norm": 0.10932721942663193,
+      "learning_rate": 0.0001,
+      "loss": 1.6174,
+      "step": 940
+    },
+    {
+      "epoch": 0.5098889189921431,
+      "grad_norm": 0.11859098821878433,
+      "learning_rate": 0.0001,
+      "loss": 1.631,
+      "step": 941
+    },
+    {
+      "epoch": 0.510430777567055,
+      "grad_norm": 0.12220406532287598,
+      "learning_rate": 0.0001,
+      "loss": 1.6904,
+      "step": 942
+    },
+    {
+      "epoch": 0.510972636141967,
+      "grad_norm": 0.11712736636400223,
+      "learning_rate": 0.0001,
+      "loss": 1.6838,
+      "step": 943
+    },
+    {
+      "epoch": 0.5115144947168789,
+      "grad_norm": 0.12665888667106628,
+      "learning_rate": 0.0001,
+      "loss": 1.663,
+      "step": 944
+    },
+    {
+      "epoch": 0.5120563532917909,
+      "grad_norm": 0.109413743019104,
+      "learning_rate": 0.0001,
+      "loss": 1.6294,
+      "step": 945
+    },
+    {
+      "epoch": 0.5125982118667027,
+      "grad_norm": 0.11188017576932907,
+      "learning_rate": 0.0001,
+      "loss": 1.6092,
+      "step": 946
+    },
+    {
+      "epoch": 0.5131400704416147,
+      "grad_norm": 0.10553425550460815,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 947
+    },
+    {
+      "epoch": 0.5136819290165267,
+      "grad_norm": 0.11408665776252747,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 948
+    },
+    {
+      "epoch": 0.5142237875914386,
+      "grad_norm": 0.11511809378862381,
+      "learning_rate": 0.0001,
+      "loss": 1.6262,
+      "step": 949
+    },
+    {
+      "epoch": 0.5147656461663506,
+      "grad_norm": 0.12249071151018143,
+      "learning_rate": 0.0001,
+      "loss": 1.651,
+      "step": 950
+    },
+    {
+      "epoch": 0.5153075047412625,
+      "grad_norm": 0.11768020689487457,
+      "learning_rate": 0.0001,
+      "loss": 1.6488,
+      "step": 951
+    },
+    {
+      "epoch": 0.5158493633161745,
+      "grad_norm": 0.12663574516773224,
+      "learning_rate": 0.0001,
+      "loss": 1.5593,
+      "step": 952
+    },
+    {
+      "epoch": 0.5163912218910864,
+      "grad_norm": 0.11186866462230682,
+      "learning_rate": 0.0001,
+      "loss": 1.5988,
+      "step": 953
+    },
+    {
+      "epoch": 0.5169330804659984,
+      "grad_norm": 0.10449830442667007,
+      "learning_rate": 0.0001,
+      "loss": 1.6612,
+      "step": 954
+    },
+    {
+      "epoch": 0.5174749390409104,
+      "grad_norm": 0.11975737661123276,
+      "learning_rate": 0.0001,
+      "loss": 1.6603,
+      "step": 955
+    },
+    {
+      "epoch": 0.5180167976158223,
+      "grad_norm": 0.11905168741941452,
+      "learning_rate": 0.0001,
+      "loss": 1.6822,
+      "step": 956
+    },
+    {
+      "epoch": 0.5185586561907343,
+      "grad_norm": 0.10545483231544495,
+      "learning_rate": 0.0001,
+      "loss": 1.5699,
+      "step": 957
+    },
+    {
+      "epoch": 0.5191005147656461,
+      "grad_norm": 0.10853522270917892,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 958
+    },
+    {
+      "epoch": 0.5196423733405581,
+      "grad_norm": 0.12581409513950348,
+      "learning_rate": 0.0001,
+      "loss": 1.6801,
+      "step": 959
+    },
+    {
+      "epoch": 0.52018423191547,
+      "grad_norm": 0.10773199796676636,
+      "learning_rate": 0.0001,
+      "loss": 1.558,
+      "step": 960
+    },
+    {
+      "epoch": 0.520726090490382,
+      "grad_norm": 0.11740659922361374,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 961
+    },
+    {
+      "epoch": 0.521267949065294,
+      "grad_norm": 0.113717220723629,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 962
+    },
+    {
+      "epoch": 0.5218098076402059,
+      "grad_norm": 0.11157488822937012,
+      "learning_rate": 0.0001,
+      "loss": 1.6061,
+      "step": 963
+    },
+    {
+      "epoch": 0.5223516662151179,
+      "grad_norm": 0.11375845968723297,
+      "learning_rate": 0.0001,
+      "loss": 1.603,
+      "step": 964
+    },
+    {
+      "epoch": 0.5228935247900298,
+      "grad_norm": 0.10816894471645355,
+      "learning_rate": 0.0001,
+      "loss": 1.6872,
+      "step": 965
+    },
+    {
+      "epoch": 0.5234353833649418,
+      "grad_norm": 0.10593447834253311,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 966
+    },
+    {
+      "epoch": 0.5239772419398537,
+      "grad_norm": 0.1219218447804451,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 967
+    },
+    {
+      "epoch": 0.5245191005147657,
+      "grad_norm": 0.11132703721523285,
+      "learning_rate": 0.0001,
+      "loss": 1.6937,
+      "step": 968
+    },
+    {
+      "epoch": 0.5250609590896776,
+      "grad_norm": 0.12354977428913116,
+      "learning_rate": 0.0001,
+      "loss": 1.5715,
+      "step": 969
+    },
+    {
+      "epoch": 0.5256028176645895,
+      "grad_norm": 0.1068456694483757,
+      "learning_rate": 0.0001,
+      "loss": 1.5655,
+      "step": 970
+    },
+    {
+      "epoch": 0.5261446762395015,
+      "grad_norm": 0.12704645097255707,
+      "learning_rate": 0.0001,
+      "loss": 1.6729,
+      "step": 971
+    },
+    {
+      "epoch": 0.5266865348144134,
+      "grad_norm": 0.11148631572723389,
+      "learning_rate": 0.0001,
+      "loss": 1.6056,
+      "step": 972
+    },
+    {
+      "epoch": 0.5272283933893254,
+      "grad_norm": 0.13174760341644287,
+      "learning_rate": 0.0001,
+      "loss": 1.7158,
+      "step": 973
+    },
+    {
+      "epoch": 0.5277702519642373,
+      "grad_norm": 0.10785822570323944,
+      "learning_rate": 0.0001,
+      "loss": 1.6236,
+      "step": 974
+    },
+    {
+      "epoch": 0.5283121105391493,
+      "grad_norm": 0.13404607772827148,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 975
+    },
+    {
+      "epoch": 0.5288539691140612,
+      "grad_norm": 0.10916657000780106,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 976
+    },
+    {
+      "epoch": 0.5293958276889732,
+      "grad_norm": 0.11268014460802078,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 977
+    },
+    {
+      "epoch": 0.5299376862638852,
+      "grad_norm": 0.13467144966125488,
+      "learning_rate": 0.0001,
+      "loss": 1.6437,
+      "step": 978
+    },
+    {
+      "epoch": 0.5304795448387971,
+      "grad_norm": 0.12466847151517868,
+      "learning_rate": 0.0001,
+      "loss": 1.6565,
+      "step": 979
+    },
+    {
+      "epoch": 0.5310214034137091,
+      "grad_norm": 0.1344923973083496,
+      "learning_rate": 0.0001,
+      "loss": 1.61,
+      "step": 980
+    },
+    {
+      "epoch": 0.5315632619886209,
+      "grad_norm": 0.10806672275066376,
+      "learning_rate": 0.0001,
+      "loss": 1.5325,
+      "step": 981
+    },
+    {
+      "epoch": 0.5321051205635329,
+      "grad_norm": 0.1430986523628235,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 982
+    },
+    {
+      "epoch": 0.5326469791384448,
+      "grad_norm": 0.10808341205120087,
+      "learning_rate": 0.0001,
+      "loss": 1.6631,
+      "step": 983
+    },
+    {
+      "epoch": 0.5331888377133568,
+      "grad_norm": 0.11897061765193939,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 984
+    },
+    {
+      "epoch": 0.5337306962882687,
+      "grad_norm": 0.11438190191984177,
+      "learning_rate": 0.0001,
+      "loss": 1.525,
+      "step": 985
+    },
+    {
+      "epoch": 0.5342725548631807,
+      "grad_norm": 0.12540669739246368,
+      "learning_rate": 0.0001,
+      "loss": 1.6432,
+      "step": 986
+    },
+    {
+      "epoch": 0.5348144134380927,
+      "grad_norm": 0.12103486806154251,
+      "learning_rate": 0.0001,
+      "loss": 1.5627,
+      "step": 987
+    },
+    {
+      "epoch": 0.5353562720130046,
+      "grad_norm": 0.10832776874303818,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 988
+    },
+    {
+      "epoch": 0.5358981305879166,
+      "grad_norm": 0.13159573078155518,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 989
+    },
+    {
+      "epoch": 0.5364399891628285,
+      "grad_norm": 0.101260244846344,
+      "learning_rate": 0.0001,
+      "loss": 1.549,
+      "step": 990
+    },
+    {
+      "epoch": 0.5369818477377405,
+      "grad_norm": 0.12760812044143677,
+      "learning_rate": 0.0001,
+      "loss": 1.6317,
+      "step": 991
+    },
+    {
+      "epoch": 0.5375237063126524,
+      "grad_norm": 0.1126088872551918,
+      "learning_rate": 0.0001,
+      "loss": 1.5693,
+      "step": 992
+    },
+    {
+      "epoch": 0.5380655648875643,
+      "grad_norm": 0.13307423889636993,
+      "learning_rate": 0.0001,
+      "loss": 1.6957,
+      "step": 993
+    },
+    {
+      "epoch": 0.5386074234624763,
+      "grad_norm": 0.11326015740633011,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 994
+    },
+    {
+      "epoch": 0.5391492820373882,
+      "grad_norm": 0.11083459854125977,
+      "learning_rate": 0.0001,
+      "loss": 1.5401,
+      "step": 995
+    },
+    {
+      "epoch": 0.5396911406123002,
+      "grad_norm": 0.12092125415802002,
+      "learning_rate": 0.0001,
+      "loss": 1.6074,
+      "step": 996
+    },
+    {
+      "epoch": 0.5402329991872121,
+      "grad_norm": 0.11150885373353958,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 997
+    },
+    {
+      "epoch": 0.5407748577621241,
+      "grad_norm": 0.12295088917016983,
+      "learning_rate": 0.0001,
+      "loss": 1.64,
+      "step": 998
+    },
+    {
+      "epoch": 0.541316716337036,
+      "grad_norm": 0.1229565292596817,
+      "learning_rate": 0.0001,
+      "loss": 1.6828,
+      "step": 999
+    },
+    {
+      "epoch": 0.541858574911948,
+      "grad_norm": 0.13467469811439514,
+      "learning_rate": 0.0001,
+      "loss": 1.5319,
+      "step": 1000
+    },
+    {
+      "epoch": 0.54240043348686,
+      "grad_norm": 0.11792799085378647,
+      "learning_rate": 0.0001,
+      "loss": 1.6261,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5429422920617719,
+      "grad_norm": 0.11050891876220703,
+      "learning_rate": 0.0001,
+      "loss": 1.512,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5434841506366839,
+      "grad_norm": 0.11616943776607513,
+      "learning_rate": 0.0001,
+      "loss": 1.6496,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5440260092115957,
+      "grad_norm": 0.10420012474060059,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5445678677865077,
+      "grad_norm": 0.1318942904472351,
+      "learning_rate": 0.0001,
+      "loss": 1.6199,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5451097263614196,
+      "grad_norm": 0.1039443239569664,
+      "learning_rate": 0.0001,
+      "loss": 1.6361,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5456515849363316,
+      "grad_norm": 0.11699001491069794,
+      "learning_rate": 0.0001,
+      "loss": 1.6031,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5461934435112435,
+      "grad_norm": 0.11182098835706711,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5467353020861555,
+      "grad_norm": 0.13418853282928467,
+      "learning_rate": 0.0001,
+      "loss": 1.6667,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5472771606610675,
+      "grad_norm": 0.11194122582674026,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5478190192359794,
+      "grad_norm": 0.11652825772762299,
+      "learning_rate": 0.0001,
+      "loss": 1.6053,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5483608778108914,
+      "grad_norm": 0.11751655489206314,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5489027363858033,
+      "grad_norm": 0.1132623553276062,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5494445949607153,
+      "grad_norm": 0.1154278889298439,
+      "learning_rate": 0.0001,
+      "loss": 1.612,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5499864535356273,
+      "grad_norm": 0.11338574439287186,
+      "learning_rate": 0.0001,
+      "loss": 1.5906,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5505283121105391,
+      "grad_norm": 0.13384795188903809,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 1016
+    },
+    {
+      "epoch": 0.551070170685451,
+      "grad_norm": 0.1129109263420105,
+      "learning_rate": 0.0001,
+      "loss": 1.6028,
+      "step": 1017
+    },
+    {
+      "epoch": 0.551612029260363,
+      "grad_norm": 0.10652612894773483,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 1018
+    },
+    {
+      "epoch": 0.552153887835275,
+      "grad_norm": 0.12745411694049835,
+      "learning_rate": 0.0001,
+      "loss": 1.6464,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5526957464101869,
+      "grad_norm": 0.11083333939313889,
+      "learning_rate": 0.0001,
+      "loss": 1.5846,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5532376049850989,
+      "grad_norm": 0.11409562081098557,
+      "learning_rate": 0.0001,
+      "loss": 1.5939,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5537794635600108,
+      "grad_norm": 0.1340702772140503,
+      "learning_rate": 0.0001,
+      "loss": 1.6235,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5543213221349228,
+      "grad_norm": 0.11746427416801453,
+      "learning_rate": 0.0001,
+      "loss": 1.6509,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5548631807098348,
+      "grad_norm": 0.11569317430257797,
+      "learning_rate": 0.0001,
+      "loss": 1.6787,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5554050392847467,
+      "grad_norm": 0.10323930531740189,
+      "learning_rate": 0.0001,
+      "loss": 1.6268,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5559468978596587,
+      "grad_norm": 0.1139453575015068,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5564887564345706,
+      "grad_norm": 0.11864421516656876,
+      "learning_rate": 0.0001,
+      "loss": 1.6012,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5570306150094825,
+      "grad_norm": 0.11720296740531921,
+      "learning_rate": 0.0001,
+      "loss": 1.688,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5575724735843944,
+      "grad_norm": 0.11221726983785629,
+      "learning_rate": 0.0001,
+      "loss": 1.664,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5581143321593064,
+      "grad_norm": 0.1101880744099617,
+      "learning_rate": 0.0001,
+      "loss": 1.5422,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5586561907342183,
+      "grad_norm": 0.1299498975276947,
+      "learning_rate": 0.0001,
+      "loss": 1.598,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5591980493091303,
+      "grad_norm": 0.10887090861797333,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5597399078840423,
+      "grad_norm": 0.11657392233610153,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5602817664589542,
+      "grad_norm": 0.11606595665216446,
+      "learning_rate": 0.0001,
+      "loss": 1.6024,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5608236250338662,
+      "grad_norm": 0.10020414739847183,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5613654836087781,
+      "grad_norm": 0.11371590197086334,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5619073421836901,
+      "grad_norm": 0.10955478996038437,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 1037
+    },
+    {
+      "epoch": 0.562449200758602,
+      "grad_norm": 0.10717011988162994,
+      "learning_rate": 0.0001,
+      "loss": 1.5618,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5629910593335139,
+      "grad_norm": 0.1138799712061882,
+      "learning_rate": 0.0001,
+      "loss": 1.6216,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5635329179084259,
+      "grad_norm": 0.10846222937107086,
+      "learning_rate": 0.0001,
+      "loss": 1.7128,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5640747764833378,
+      "grad_norm": 0.11552631109952927,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5646166350582498,
+      "grad_norm": 0.10366268455982208,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5651584936331617,
+      "grad_norm": 0.10912511497735977,
+      "learning_rate": 0.0001,
+      "loss": 1.577,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5657003522080737,
+      "grad_norm": 0.11409895122051239,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5662422107829856,
+      "grad_norm": 0.1048092395067215,
+      "learning_rate": 0.0001,
+      "loss": 1.6501,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5667840693578976,
+      "grad_norm": 0.11702732741832733,
+      "learning_rate": 0.0001,
+      "loss": 1.6636,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5673259279328096,
+      "grad_norm": 0.11176422983407974,
+      "learning_rate": 0.0001,
+      "loss": 1.6681,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5678677865077215,
+      "grad_norm": 0.11163422465324402,
+      "learning_rate": 0.0001,
+      "loss": 1.5839,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5684096450826335,
+      "grad_norm": 0.10756085067987442,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5689515036575454,
+      "grad_norm": 0.10789232701063156,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5694933622324573,
+      "grad_norm": 0.10716675966978073,
+      "learning_rate": 0.0001,
+      "loss": 1.5847,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5700352208073692,
+      "grad_norm": 0.11019588261842728,
+      "learning_rate": 0.0001,
+      "loss": 1.6325,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5705770793822812,
+      "grad_norm": 0.10888416320085526,
+      "learning_rate": 0.0001,
+      "loss": 1.612,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5711189379571932,
+      "grad_norm": 0.10670910775661469,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5716607965321051,
+      "grad_norm": 0.105767622590065,
+      "learning_rate": 0.0001,
+      "loss": 1.6084,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5722026551070171,
+      "grad_norm": 0.10374841094017029,
+      "learning_rate": 0.0001,
+      "loss": 1.6413,
+      "step": 1056
+    },
+    {
+      "epoch": 0.572744513681929,
+      "grad_norm": 0.1148417666554451,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 1057
+    },
+    {
+      "epoch": 0.573286372256841,
+      "grad_norm": 0.10581746697425842,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5738282308317529,
+      "grad_norm": 0.11768446117639542,
+      "learning_rate": 0.0001,
+      "loss": 1.587,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5743700894066649,
+      "grad_norm": 0.10721245408058167,
+      "learning_rate": 0.0001,
+      "loss": 1.6631,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5749119479815769,
+      "grad_norm": 0.11277879774570465,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5754538065564888,
+      "grad_norm": 0.11104092746973038,
+      "learning_rate": 0.0001,
+      "loss": 1.6965,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5759956651314007,
+      "grad_norm": 0.11404809355735779,
+      "learning_rate": 0.0001,
+      "loss": 1.538,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5765375237063126,
+      "grad_norm": 0.10597414523363113,
+      "learning_rate": 0.0001,
+      "loss": 1.6545,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5770793822812246,
+      "grad_norm": 0.11097580939531326,
+      "learning_rate": 0.0001,
+      "loss": 1.6236,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5776212408561365,
+      "grad_norm": 0.1147984117269516,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5781630994310485,
+      "grad_norm": 0.132141575217247,
+      "learning_rate": 0.0001,
+      "loss": 1.689,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5787049580059604,
+      "grad_norm": 0.1056254580616951,
+      "learning_rate": 0.0001,
+      "loss": 1.6023,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5792468165808724,
+      "grad_norm": 0.11850099265575409,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5797886751557844,
+      "grad_norm": 0.1169082298874855,
+      "learning_rate": 0.0001,
+      "loss": 1.5843,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5803305337306963,
+      "grad_norm": 0.11147496849298477,
+      "learning_rate": 0.0001,
+      "loss": 1.5971,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5808723923056083,
+      "grad_norm": 0.11038046330213547,
+      "learning_rate": 0.0001,
+      "loss": 1.6245,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5814142508805202,
+      "grad_norm": 0.10539822280406952,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5819561094554321,
+      "grad_norm": 0.12041206657886505,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 1074
+    },
+    {
+      "epoch": 0.582497968030344,
+      "grad_norm": 0.11082960665225983,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 1075
+    },
+    {
+      "epoch": 0.583039826605256,
+      "grad_norm": 0.12492961436510086,
+      "learning_rate": 0.0001,
+      "loss": 1.5566,
+      "step": 1076
+    },
+    {
+      "epoch": 0.583581685180168,
+      "grad_norm": 0.11056699603796005,
+      "learning_rate": 0.0001,
+      "loss": 1.5725,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5841235437550799,
+      "grad_norm": 0.11277845501899719,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5846654023299919,
+      "grad_norm": 0.1074712872505188,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5852072609049038,
+      "grad_norm": 0.10993622243404388,
+      "learning_rate": 0.0001,
+      "loss": 1.5977,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5857491194798158,
+      "grad_norm": 0.11170591413974762,
+      "learning_rate": 0.0001,
+      "loss": 1.6254,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5862909780547277,
+      "grad_norm": 0.1051788404583931,
+      "learning_rate": 0.0001,
+      "loss": 1.6192,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5868328366296397,
+      "grad_norm": 0.11397905647754669,
+      "learning_rate": 0.0001,
+      "loss": 1.6512,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5873746952045517,
+      "grad_norm": 0.11181167513132095,
+      "learning_rate": 0.0001,
+      "loss": 1.6593,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5879165537794636,
+      "grad_norm": 0.1070471853017807,
+      "learning_rate": 0.0001,
+      "loss": 1.6699,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5884584123543755,
+      "grad_norm": 0.11638469249010086,
+      "learning_rate": 0.0001,
+      "loss": 1.5877,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5890002709292874,
+      "grad_norm": 0.10291191935539246,
+      "learning_rate": 0.0001,
+      "loss": 1.6169,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5895421295041994,
+      "grad_norm": 0.10486762970685959,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5900839880791113,
+      "grad_norm": 0.12095001339912415,
+      "learning_rate": 0.0001,
+      "loss": 1.6791,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5906258466540233,
+      "grad_norm": 0.1037125512957573,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5911677052289352,
+      "grad_norm": 0.11337342858314514,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5917095638038472,
+      "grad_norm": 0.11184096336364746,
+      "learning_rate": 0.0001,
+      "loss": 1.6075,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5922514223787592,
+      "grad_norm": 0.1075851172208786,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5927932809536711,
+      "grad_norm": 0.11272459477186203,
+      "learning_rate": 0.0001,
+      "loss": 1.5801,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5933351395285831,
+      "grad_norm": 0.1100969910621643,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 1095
+    },
+    {
+      "epoch": 0.593876998103495,
+      "grad_norm": 0.10581720620393753,
+      "learning_rate": 0.0001,
+      "loss": 1.5741,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5944188566784069,
+      "grad_norm": 0.11307034641504288,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5949607152533188,
+      "grad_norm": 0.1128588542342186,
+      "learning_rate": 0.0001,
+      "loss": 1.6952,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5955025738282308,
+      "grad_norm": 0.11140985786914825,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5960444324031428,
+      "grad_norm": 0.11232408136129379,
+      "learning_rate": 0.0001,
+      "loss": 1.6637,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5965862909780547,
+      "grad_norm": 0.12389969825744629,
+      "learning_rate": 0.0001,
+      "loss": 1.6926,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5971281495529667,
+      "grad_norm": 0.11053042858839035,
+      "learning_rate": 0.0001,
+      "loss": 1.5399,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5976700081278786,
+      "grad_norm": 0.11876285076141357,
+      "learning_rate": 0.0001,
+      "loss": 1.6231,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5982118667027906,
+      "grad_norm": 0.11597663164138794,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5987537252777025,
+      "grad_norm": 0.11426155269145966,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5992955838526145,
+      "grad_norm": 0.13617175817489624,
+      "learning_rate": 0.0001,
+      "loss": 1.6879,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5998374424275265,
+      "grad_norm": 0.12879601120948792,
+      "learning_rate": 0.0001,
+      "loss": 1.6476,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6003793010024384,
+      "grad_norm": 0.11163545399904251,
+      "learning_rate": 0.0001,
+      "loss": 1.6301,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6009211595773503,
+      "grad_norm": 0.10743013024330139,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6014630181522622,
+      "grad_norm": 0.1111200824379921,
+      "learning_rate": 0.0001,
+      "loss": 1.629,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6020048767271742,
+      "grad_norm": 0.11115365475416183,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6025467353020861,
+      "grad_norm": 0.10575967282056808,
+      "learning_rate": 0.0001,
+      "loss": 1.5795,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6030885938769981,
+      "grad_norm": 0.11594882607460022,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 1113
+    },
+    {
+      "epoch": 0.60363045245191,
+      "grad_norm": 0.1093570813536644,
+      "learning_rate": 0.0001,
+      "loss": 1.5788,
+      "step": 1114
+    },
+    {
+      "epoch": 0.604172311026822,
+      "grad_norm": 0.10965988039970398,
+      "learning_rate": 0.0001,
+      "loss": 1.6068,
+      "step": 1115
+    },
+    {
+      "epoch": 0.604714169601734,
+      "grad_norm": 0.12208772450685501,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6052560281766459,
+      "grad_norm": 0.10870245844125748,
+      "learning_rate": 0.0001,
+      "loss": 1.602,
+      "step": 1117
+    },
+    {
+      "epoch": 0.6057978867515579,
+      "grad_norm": 0.12144634127616882,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6063397453264698,
+      "grad_norm": 0.11317051947116852,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6068816039013818,
+      "grad_norm": 0.12039986252784729,
+      "learning_rate": 0.0001,
+      "loss": 1.6377,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6074234624762936,
+      "grad_norm": 0.1101534515619278,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6079653210512056,
+      "grad_norm": 0.11682349443435669,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6085071796261176,
+      "grad_norm": 0.10640917718410492,
+      "learning_rate": 0.0001,
+      "loss": 1.5877,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6090490382010295,
+      "grad_norm": 0.10861624032258987,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6095908967759415,
+      "grad_norm": 0.10645128786563873,
+      "learning_rate": 0.0001,
+      "loss": 1.637,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6101327553508534,
+      "grad_norm": 0.10756590217351913,
+      "learning_rate": 0.0001,
+      "loss": 1.6351,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6106746139257654,
+      "grad_norm": 0.10730548202991486,
+      "learning_rate": 0.0001,
+      "loss": 1.5956,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6112164725006773,
+      "grad_norm": 0.10790326446294785,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6117583310755893,
+      "grad_norm": 0.10814284533262253,
+      "learning_rate": 0.0001,
+      "loss": 1.6249,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6123001896505013,
+      "grad_norm": 0.11095184832811356,
+      "learning_rate": 0.0001,
+      "loss": 1.6556,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6128420482254132,
+      "grad_norm": 0.11202125251293182,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6133839068003251,
+      "grad_norm": 0.1061285063624382,
+      "learning_rate": 0.0001,
+      "loss": 1.5906,
+      "step": 1132
+    },
+    {
+      "epoch": 0.613925765375237,
+      "grad_norm": 0.11121700704097748,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 1133
+    },
+    {
+      "epoch": 0.614467623950149,
+      "grad_norm": 0.11035364866256714,
+      "learning_rate": 0.0001,
+      "loss": 1.6682,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6150094825250609,
+      "grad_norm": 0.11122874170541763,
+      "learning_rate": 0.0001,
+      "loss": 1.6271,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6155513410999729,
+      "grad_norm": 0.10672580450773239,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6160931996748849,
+      "grad_norm": 0.11283120512962341,
+      "learning_rate": 0.0001,
+      "loss": 1.6331,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6166350582497968,
+      "grad_norm": 0.10613071173429489,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 1138
+    },
+    {
+      "epoch": 0.6171769168247088,
+      "grad_norm": 0.11445633322000504,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6177187753996207,
+      "grad_norm": 0.10982243716716766,
+      "learning_rate": 0.0001,
+      "loss": 1.599,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6182606339745327,
+      "grad_norm": 0.11763114482164383,
+      "learning_rate": 0.0001,
+      "loss": 1.7074,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6188024925494446,
+      "grad_norm": 0.11409911513328552,
+      "learning_rate": 0.0001,
+      "loss": 1.609,
+      "step": 1142
+    },
+    {
+      "epoch": 0.6193443511243566,
+      "grad_norm": 0.12272528558969498,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6198862096992684,
+      "grad_norm": 0.10473363101482391,
+      "learning_rate": 0.0001,
+      "loss": 1.6001,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6204280682741804,
+      "grad_norm": 0.10619332641363144,
+      "learning_rate": 0.0001,
+      "loss": 1.5967,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6209699268490924,
+      "grad_norm": 0.11708606034517288,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6215117854240043,
+      "grad_norm": 0.11772305518388748,
+      "learning_rate": 0.0001,
+      "loss": 1.6009,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6220536439989163,
+      "grad_norm": 0.109929159283638,
+      "learning_rate": 0.0001,
+      "loss": 1.6116,
+      "step": 1148
+    },
+    {
+      "epoch": 0.6225955025738282,
+      "grad_norm": 0.11920684576034546,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6231373611487402,
+      "grad_norm": 0.1033051609992981,
+      "learning_rate": 0.0001,
+      "loss": 1.5432,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6236792197236521,
+      "grad_norm": 0.10906940698623657,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 1151
+    },
+    {
+      "epoch": 0.6242210782985641,
+      "grad_norm": 0.10953330248594284,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6247629368734761,
+      "grad_norm": 0.11275649815797806,
+      "learning_rate": 0.0001,
+      "loss": 1.6157,
+      "step": 1153
+    },
+    {
+      "epoch": 0.625304795448388,
+      "grad_norm": 0.1148492619395256,
+      "learning_rate": 0.0001,
+      "loss": 1.6836,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6258466540232999,
+      "grad_norm": 0.10278470069169998,
+      "learning_rate": 0.0001,
+      "loss": 1.5747,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6263885125982118,
+      "grad_norm": 0.12111508101224899,
+      "learning_rate": 0.0001,
+      "loss": 1.5517,
+      "step": 1156
+    },
+    {
+      "epoch": 0.6269303711731238,
+      "grad_norm": 0.10696611553430557,
+      "learning_rate": 0.0001,
+      "loss": 1.5089,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6274722297480357,
+      "grad_norm": 0.1212284117937088,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6280140883229477,
+      "grad_norm": 0.10886389762163162,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6285559468978597,
+      "grad_norm": 0.11113671213388443,
+      "learning_rate": 0.0001,
+      "loss": 1.5473,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6290978054727716,
+      "grad_norm": 0.10953107476234436,
+      "learning_rate": 0.0001,
+      "loss": 1.603,
+      "step": 1161
+    },
+    {
+      "epoch": 0.6296396640476836,
+      "grad_norm": 0.1181061714887619,
+      "learning_rate": 0.0001,
+      "loss": 1.6897,
+      "step": 1162
+    },
+    {
+      "epoch": 0.6301815226225955,
+      "grad_norm": 0.12265321612358093,
+      "learning_rate": 0.0001,
+      "loss": 1.6021,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6307233811975075,
+      "grad_norm": 0.11034665256738663,
+      "learning_rate": 0.0001,
+      "loss": 1.681,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6312652397724194,
+      "grad_norm": 0.10766472667455673,
+      "learning_rate": 0.0001,
+      "loss": 1.5982,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6318070983473314,
+      "grad_norm": 0.12280579656362534,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6323489569222432,
+      "grad_norm": 0.10680815577507019,
+      "learning_rate": 0.0001,
+      "loss": 1.6087,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6328908154971552,
+      "grad_norm": 0.11566875129938126,
+      "learning_rate": 0.0001,
+      "loss": 1.536,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6334326740720672,
+      "grad_norm": 0.12584605813026428,
+      "learning_rate": 0.0001,
+      "loss": 1.6341,
+      "step": 1169
+    },
+    {
+      "epoch": 0.6339745326469791,
+      "grad_norm": 0.12214954942464828,
+      "learning_rate": 0.0001,
+      "loss": 1.6041,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6345163912218911,
+      "grad_norm": 0.12063132226467133,
+      "learning_rate": 0.0001,
+      "loss": 1.6083,
+      "step": 1171
+    },
+    {
+      "epoch": 0.635058249796803,
+      "grad_norm": 0.11362355202436447,
+      "learning_rate": 0.0001,
+      "loss": 1.5818,
+      "step": 1172
+    },
+    {
+      "epoch": 0.635600108371715,
+      "grad_norm": 0.11989773064851761,
+      "learning_rate": 0.0001,
+      "loss": 1.6392,
+      "step": 1173
+    },
+    {
+      "epoch": 0.636141966946627,
+      "grad_norm": 0.1103493869304657,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 1174
+    },
+    {
+      "epoch": 0.6366838255215389,
+      "grad_norm": 0.11394510418176651,
+      "learning_rate": 0.0001,
+      "loss": 1.7143,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6372256840964509,
+      "grad_norm": 0.11421728134155273,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 1176
+    },
+    {
+      "epoch": 0.6377675426713628,
+      "grad_norm": 0.11668463051319122,
+      "learning_rate": 0.0001,
+      "loss": 1.6418,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6383094012462748,
+      "grad_norm": 0.10839111357927322,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6388512598211866,
+      "grad_norm": 0.11521162837743759,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6393931183960986,
+      "grad_norm": 0.11129699647426605,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6399349769710105,
+      "grad_norm": 0.11090409010648727,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6404768355459225,
+      "grad_norm": 0.11323232203722,
+      "learning_rate": 0.0001,
+      "loss": 1.642,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6410186941208345,
+      "grad_norm": 0.10628697276115417,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 1183
+    },
+    {
+      "epoch": 0.6415605526957464,
+      "grad_norm": 0.11381271481513977,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 1184
+    },
+    {
+      "epoch": 0.6421024112706584,
+      "grad_norm": 0.114079050719738,
+      "learning_rate": 0.0001,
+      "loss": 1.6762,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6426442698455703,
+      "grad_norm": 0.10727323591709137,
+      "learning_rate": 0.0001,
+      "loss": 1.5488,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6431861284204823,
+      "grad_norm": 0.11506172269582748,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6437279869953942,
+      "grad_norm": 0.11387591063976288,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6442698455703062,
+      "grad_norm": 0.11320862174034119,
+      "learning_rate": 0.0001,
+      "loss": 1.6238,
+      "step": 1189
+    },
+    {
+      "epoch": 0.644811704145218,
+      "grad_norm": 0.12005679309368134,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 1190
+    },
+    {
+      "epoch": 0.64535356272013,
+      "grad_norm": 0.11804357171058655,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 1191
+    },
+    {
+      "epoch": 0.645895421295042,
+      "grad_norm": 0.13115650415420532,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6464372798699539,
+      "grad_norm": 0.12141333520412445,
+      "learning_rate": 0.0001,
+      "loss": 1.5315,
+      "step": 1193
+    },
+    {
+      "epoch": 0.6469791384448659,
+      "grad_norm": 0.11301975697278976,
+      "learning_rate": 0.0001,
+      "loss": 1.5603,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6475209970197778,
+      "grad_norm": 0.10810407996177673,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6480628555946898,
+      "grad_norm": 0.1332971453666687,
+      "learning_rate": 0.0001,
+      "loss": 1.5945,
+      "step": 1196
+    },
+    {
+      "epoch": 0.6486047141696017,
+      "grad_norm": 0.11277639120817184,
+      "learning_rate": 0.0001,
+      "loss": 1.5758,
+      "step": 1197
+    },
+    {
+      "epoch": 0.6491465727445137,
+      "grad_norm": 0.12171521037817001,
+      "learning_rate": 0.0001,
+      "loss": 1.6254,
+      "step": 1198
+    },
+    {
+      "epoch": 0.6496884313194257,
+      "grad_norm": 0.10795322060585022,
+      "learning_rate": 0.0001,
+      "loss": 1.6728,
+      "step": 1199
+    },
+    {
+      "epoch": 0.6502302898943376,
+      "grad_norm": 0.10944177210330963,
+      "learning_rate": 0.0001,
+      "loss": 1.6788,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6507721484692496,
+      "grad_norm": 0.11161891371011734,
+      "learning_rate": 0.0001,
+      "loss": 1.595,
+      "step": 1201
+    },
+    {
+      "epoch": 0.6513140070441614,
+      "grad_norm": 0.11086007207632065,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 1202
+    },
+    {
+      "epoch": 0.6518558656190734,
+      "grad_norm": 0.10901938378810883,
+      "learning_rate": 0.0001,
+      "loss": 1.6363,
+      "step": 1203
+    },
+    {
+      "epoch": 0.6523977241939853,
+      "grad_norm": 0.11241175979375839,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 1204
+    },
+    {
+      "epoch": 0.6529395827688973,
+      "grad_norm": 0.11044015735387802,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6534814413438093,
+      "grad_norm": 0.10838108509778976,
+      "learning_rate": 0.0001,
+      "loss": 1.6266,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6540232999187212,
+      "grad_norm": 0.11160946637392044,
+      "learning_rate": 0.0001,
+      "loss": 1.5725,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6545651584936332,
+      "grad_norm": 0.11063013225793839,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 1208
+    },
+    {
+      "epoch": 0.6551070170685451,
+      "grad_norm": 0.11617591977119446,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6556488756434571,
+      "grad_norm": 0.11063467711210251,
+      "learning_rate": 0.0001,
+      "loss": 1.6289,
+      "step": 1210
+    },
+    {
+      "epoch": 0.656190734218369,
+      "grad_norm": 0.11112099140882492,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 1211
+    },
+    {
+      "epoch": 0.656732592793281,
+      "grad_norm": 0.11596956104040146,
+      "learning_rate": 0.0001,
+      "loss": 1.6406,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6572744513681928,
+      "grad_norm": 0.11139282584190369,
+      "learning_rate": 0.0001,
+      "loss": 1.6783,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6578163099431048,
+      "grad_norm": 0.11790062487125397,
+      "learning_rate": 0.0001,
+      "loss": 1.6357,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6583581685180168,
+      "grad_norm": 0.11024824529886246,
+      "learning_rate": 0.0001,
+      "loss": 1.5921,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6589000270929287,
+      "grad_norm": 0.10710059851408005,
+      "learning_rate": 0.0001,
+      "loss": 1.6154,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6594418856678407,
+      "grad_norm": 0.1107814610004425,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6599837442427526,
+      "grad_norm": 0.10909678041934967,
+      "learning_rate": 0.0001,
+      "loss": 1.6034,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6605256028176646,
+      "grad_norm": 0.12020471692085266,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6610674613925765,
+      "grad_norm": 0.10776503384113312,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6616093199674885,
+      "grad_norm": 0.10916958749294281,
+      "learning_rate": 0.0001,
+      "loss": 1.6218,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6621511785424005,
+      "grad_norm": 0.13428278267383575,
+      "learning_rate": 0.0001,
+      "loss": 1.5782,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6626930371173124,
+      "grad_norm": 0.11061003804206848,
+      "learning_rate": 0.0001,
+      "loss": 1.5575,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6632348956922244,
+      "grad_norm": 0.13390862941741943,
+      "learning_rate": 0.0001,
+      "loss": 1.6252,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6637767542671362,
+      "grad_norm": 0.10983669757843018,
+      "learning_rate": 0.0001,
+      "loss": 1.6359,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6643186128420482,
+      "grad_norm": 0.11912653595209122,
+      "learning_rate": 0.0001,
+      "loss": 1.5967,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6648604714169601,
+      "grad_norm": 0.11113286763429642,
+      "learning_rate": 0.0001,
+      "loss": 1.5528,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6654023299918721,
+      "grad_norm": 0.11705653369426727,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6659441885667841,
+      "grad_norm": 0.11865860223770142,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 1229
+    },
+    {
+      "epoch": 0.666486047141696,
+      "grad_norm": 0.11989482492208481,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 1230
+    },
+    {
+      "epoch": 0.667027905716608,
+      "grad_norm": 0.11474844813346863,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6675697642915199,
+      "grad_norm": 0.10648138076066971,
+      "learning_rate": 0.0001,
+      "loss": 1.6084,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6681116228664319,
+      "grad_norm": 0.11240139603614807,
+      "learning_rate": 0.0001,
+      "loss": 1.6245,
+      "step": 1233
+    },
+    {
+      "epoch": 0.6686534814413438,
+      "grad_norm": 0.1064700037240982,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6691953400162558,
+      "grad_norm": 0.1053568571805954,
+      "learning_rate": 0.0001,
+      "loss": 1.6178,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6697371985911678,
+      "grad_norm": 0.11085887253284454,
+      "learning_rate": 0.0001,
+      "loss": 1.606,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6702790571660796,
+      "grad_norm": 0.10840357840061188,
+      "learning_rate": 0.0001,
+      "loss": 1.6147,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6708209157409916,
+      "grad_norm": 0.11242581158876419,
+      "learning_rate": 0.0001,
+      "loss": 1.6199,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6713627743159035,
+      "grad_norm": 0.10839978605508804,
+      "learning_rate": 0.0001,
+      "loss": 1.5522,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6719046328908155,
+      "grad_norm": 0.11424274742603302,
+      "learning_rate": 0.0001,
+      "loss": 1.637,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6724464914657274,
+      "grad_norm": 0.10760857164859772,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 1241
+    },
+    {
+      "epoch": 0.6729883500406394,
+      "grad_norm": 0.11048325896263123,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6735302086155514,
+      "grad_norm": 0.11890499293804169,
+      "learning_rate": 0.0001,
+      "loss": 1.6474,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6740720671904633,
+      "grad_norm": 0.12253390997648239,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6746139257653753,
+      "grad_norm": 0.10284169018268585,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6751557843402872,
+      "grad_norm": 0.1186986118555069,
+      "learning_rate": 0.0001,
+      "loss": 1.6631,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6756976429151992,
+      "grad_norm": 0.11720992624759674,
+      "learning_rate": 0.0001,
+      "loss": 1.6543,
+      "step": 1247
+    },
+    {
+      "epoch": 0.676239501490111,
+      "grad_norm": 0.1224815845489502,
+      "learning_rate": 0.0001,
+      "loss": 1.6462,
+      "step": 1248
+    },
+    {
+      "epoch": 0.676781360065023,
+      "grad_norm": 0.13205626606941223,
+      "learning_rate": 0.0001,
+      "loss": 1.6532,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6773232186399349,
+      "grad_norm": 0.11484634131193161,
+      "learning_rate": 0.0001,
+      "loss": 1.6192,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6778650772148469,
+      "grad_norm": 0.10785496979951859,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6784069357897589,
+      "grad_norm": 0.10727088153362274,
+      "learning_rate": 0.0001,
+      "loss": 1.6261,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6789487943646708,
+      "grad_norm": 0.10857659578323364,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6794906529395828,
+      "grad_norm": 0.10896406322717667,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6800325115144947,
+      "grad_norm": 0.11162500828504562,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6805743700894067,
+      "grad_norm": 0.1097574532032013,
+      "learning_rate": 0.0001,
+      "loss": 1.6019,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6811162286643186,
+      "grad_norm": 0.10825852304697037,
+      "learning_rate": 0.0001,
+      "loss": 1.6086,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6816580872392306,
+      "grad_norm": 0.11132719367742538,
+      "learning_rate": 0.0001,
+      "loss": 1.6945,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6821999458141426,
+      "grad_norm": 0.10512889176607132,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6827418043890544,
+      "grad_norm": 0.11441686749458313,
+      "learning_rate": 0.0001,
+      "loss": 1.5839,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6832836629639664,
+      "grad_norm": 0.10713682323694229,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6838255215388783,
+      "grad_norm": 0.10768163204193115,
+      "learning_rate": 0.0001,
+      "loss": 1.574,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6843673801137903,
+      "grad_norm": 0.12209385633468628,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6849092386887022,
+      "grad_norm": 0.10845624655485153,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6854510972636142,
+      "grad_norm": 0.12702186405658722,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6859929558385262,
+      "grad_norm": 0.10910164564847946,
+      "learning_rate": 0.0001,
+      "loss": 1.6058,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6865348144134381,
+      "grad_norm": 0.12984834611415863,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6870766729883501,
+      "grad_norm": 0.10938838124275208,
+      "learning_rate": 0.0001,
+      "loss": 1.6494,
+      "step": 1268
+    },
+    {
+      "epoch": 0.687618531563262,
+      "grad_norm": 0.12864528596401215,
+      "learning_rate": 0.0001,
+      "loss": 1.6437,
+      "step": 1269
+    },
+    {
+      "epoch": 0.688160390138174,
+      "grad_norm": 0.11228509992361069,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6887022487130858,
+      "grad_norm": 0.14109857380390167,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6892441072879978,
+      "grad_norm": 0.11296511441469193,
+      "learning_rate": 0.0001,
+      "loss": 1.5076,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6897859658629097,
+      "grad_norm": 0.13155023753643036,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6903278244378217,
+      "grad_norm": 0.12987349927425385,
+      "learning_rate": 0.0001,
+      "loss": 1.581,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6908696830127337,
+      "grad_norm": 0.1197320967912674,
+      "learning_rate": 0.0001,
+      "loss": 1.6488,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6914115415876456,
+      "grad_norm": 0.13939501345157623,
+      "learning_rate": 0.0001,
+      "loss": 1.6226,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6919534001625576,
+      "grad_norm": 0.11400146782398224,
+      "learning_rate": 0.0001,
+      "loss": 1.6771,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6924952587374695,
+      "grad_norm": 0.1326090395450592,
+      "learning_rate": 0.0001,
+      "loss": 1.6103,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6930371173123815,
+      "grad_norm": 0.11246149986982346,
+      "learning_rate": 0.0001,
+      "loss": 1.6278,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6935789758872934,
+      "grad_norm": 0.129455104470253,
+      "learning_rate": 0.0001,
+      "loss": 1.6432,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6941208344622054,
+      "grad_norm": 0.10411140322685242,
+      "learning_rate": 0.0001,
+      "loss": 1.5131,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6946626930371174,
+      "grad_norm": 0.11562889814376831,
+      "learning_rate": 0.0001,
+      "loss": 1.5684,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6952045516120292,
+      "grad_norm": 0.12287727743387222,
+      "learning_rate": 0.0001,
+      "loss": 1.5769,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6957464101869412,
+      "grad_norm": 0.11982248723506927,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6962882687618531,
+      "grad_norm": 0.11934027075767517,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6968301273367651,
+      "grad_norm": 0.10960278660058975,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 1286
+    },
+    {
+      "epoch": 0.697371985911677,
+      "grad_norm": 0.1165471076965332,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 1287
+    },
+    {
+      "epoch": 0.697913844486589,
+      "grad_norm": 0.11723136901855469,
+      "learning_rate": 0.0001,
+      "loss": 1.6482,
+      "step": 1288
+    },
+    {
+      "epoch": 0.698455703061501,
+      "grad_norm": 0.12743420898914337,
+      "learning_rate": 0.0001,
+      "loss": 1.6297,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6989975616364129,
+      "grad_norm": 0.1108272448182106,
+      "learning_rate": 0.0001,
+      "loss": 1.645,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6995394202113249,
+      "grad_norm": 0.1132800430059433,
+      "learning_rate": 0.0001,
+      "loss": 1.5948,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7000812787862368,
+      "grad_norm": 0.10868791490793228,
+      "learning_rate": 0.0001,
+      "loss": 1.6254,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7006231373611488,
+      "grad_norm": 0.10428892821073532,
+      "learning_rate": 0.0001,
+      "loss": 1.5417,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7011649959360607,
+      "grad_norm": 0.10921305418014526,
+      "learning_rate": 0.0001,
+      "loss": 1.6507,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7017068545109726,
+      "grad_norm": 0.1108875572681427,
+      "learning_rate": 0.0001,
+      "loss": 1.5401,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7022487130858845,
+      "grad_norm": 0.11594570428133011,
+      "learning_rate": 0.0001,
+      "loss": 1.6711,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7027905716607965,
+      "grad_norm": 0.10920985043048859,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7033324302357085,
+      "grad_norm": 0.11153151839971542,
+      "learning_rate": 0.0001,
+      "loss": 1.5176,
+      "step": 1298
+    },
+    {
+      "epoch": 0.7038742888106204,
+      "grad_norm": 0.12206792831420898,
+      "learning_rate": 0.0001,
+      "loss": 1.5998,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7044161473855324,
+      "grad_norm": 0.11783460527658463,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7049580059604443,
+      "grad_norm": 0.11225185543298721,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7054998645353563,
+      "grad_norm": 0.10486938804388046,
+      "learning_rate": 0.0001,
+      "loss": 1.5367,
+      "step": 1302
+    },
+    {
+      "epoch": 0.7060417231102682,
+      "grad_norm": 0.10571251064538956,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7065835816851802,
+      "grad_norm": 0.10847926884889603,
+      "learning_rate": 0.0001,
+      "loss": 1.5934,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7071254402600922,
+      "grad_norm": 0.10501236468553543,
+      "learning_rate": 0.0001,
+      "loss": 1.5665,
+      "step": 1305
+    },
+    {
+      "epoch": 0.707667298835004,
+      "grad_norm": 0.10688818991184235,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 1306
+    },
+    {
+      "epoch": 0.708209157409916,
+      "grad_norm": 0.11243797838687897,
+      "learning_rate": 0.0001,
+      "loss": 1.5612,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7087510159848279,
+      "grad_norm": 0.11504892259836197,
+      "learning_rate": 0.0001,
+      "loss": 1.6366,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7092928745597399,
+      "grad_norm": 0.10843952745199203,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7098347331346518,
+      "grad_norm": 0.12078722566366196,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7103765917095638,
+      "grad_norm": 0.1083657294511795,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7109184502844758,
+      "grad_norm": 0.124849334359169,
+      "learning_rate": 0.0001,
+      "loss": 1.6311,
+      "step": 1312
+    },
+    {
+      "epoch": 0.7114603088593877,
+      "grad_norm": 0.11586251109838486,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 1313
+    },
+    {
+      "epoch": 0.7120021674342997,
+      "grad_norm": 0.12780465185642242,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 1314
+    },
+    {
+      "epoch": 0.7125440260092116,
+      "grad_norm": 0.1156454011797905,
+      "learning_rate": 0.0001,
+      "loss": 1.6577,
+      "step": 1315
+    },
+    {
+      "epoch": 0.7130858845841236,
+      "grad_norm": 0.10399503260850906,
+      "learning_rate": 0.0001,
+      "loss": 1.5726,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7136277431590355,
+      "grad_norm": 0.13720951974391937,
+      "learning_rate": 0.0001,
+      "loss": 1.6312,
+      "step": 1317
+    },
+    {
+      "epoch": 0.7141696017339474,
+      "grad_norm": 0.1113177239894867,
+      "learning_rate": 0.0001,
+      "loss": 1.6302,
+      "step": 1318
+    },
+    {
+      "epoch": 0.7147114603088593,
+      "grad_norm": 0.11841485649347305,
+      "learning_rate": 0.0001,
+      "loss": 1.6748,
+      "step": 1319
+    },
+    {
+      "epoch": 0.7152533188837713,
+      "grad_norm": 0.10778098553419113,
+      "learning_rate": 0.0001,
+      "loss": 1.6085,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7157951774586833,
+      "grad_norm": 0.12273528426885605,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 1321
+    },
+    {
+      "epoch": 0.7163370360335952,
+      "grad_norm": 0.11129626631736755,
+      "learning_rate": 0.0001,
+      "loss": 1.645,
+      "step": 1322
+    },
+    {
+      "epoch": 0.7168788946085072,
+      "grad_norm": 0.11495073139667511,
+      "learning_rate": 0.0001,
+      "loss": 1.594,
+      "step": 1323
+    },
+    {
+      "epoch": 0.7174207531834191,
+      "grad_norm": 0.13620105385780334,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7179626117583311,
+      "grad_norm": 0.11839475482702255,
+      "learning_rate": 0.0001,
+      "loss": 1.619,
+      "step": 1325
+    },
+    {
+      "epoch": 0.718504470333243,
+      "grad_norm": 0.1258990615606308,
+      "learning_rate": 0.0001,
+      "loss": 1.6376,
+      "step": 1326
+    },
+    {
+      "epoch": 0.719046328908155,
+      "grad_norm": 0.11513817310333252,
+      "learning_rate": 0.0001,
+      "loss": 1.7049,
+      "step": 1327
+    },
+    {
+      "epoch": 0.719588187483067,
+      "grad_norm": 0.11587271839380264,
+      "learning_rate": 0.0001,
+      "loss": 1.6574,
+      "step": 1328
+    },
+    {
+      "epoch": 0.7201300460579788,
+      "grad_norm": 0.10889148712158203,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 1329
+    },
+    {
+      "epoch": 0.7206719046328908,
+      "grad_norm": 0.12195302546024323,
+      "learning_rate": 0.0001,
+      "loss": 1.6679,
+      "step": 1330
+    },
+    {
+      "epoch": 0.7212137632078027,
+      "grad_norm": 0.1033078134059906,
+      "learning_rate": 0.0001,
+      "loss": 1.5089,
+      "step": 1331
+    },
+    {
+      "epoch": 0.7217556217827147,
+      "grad_norm": 0.11243455857038498,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7222974803576266,
+      "grad_norm": 0.11422085016965866,
+      "learning_rate": 0.0001,
+      "loss": 1.6454,
+      "step": 1333
+    },
+    {
+      "epoch": 0.7228393389325386,
+      "grad_norm": 0.1344115287065506,
+      "learning_rate": 0.0001,
+      "loss": 1.624,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7233811975074506,
+      "grad_norm": 0.10726635903120041,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 1335
+    },
+    {
+      "epoch": 0.7239230560823625,
+      "grad_norm": 0.11307653784751892,
+      "learning_rate": 0.0001,
+      "loss": 1.6068,
+      "step": 1336
+    },
+    {
+      "epoch": 0.7244649146572745,
+      "grad_norm": 0.10623101890087128,
+      "learning_rate": 0.0001,
+      "loss": 1.5428,
+      "step": 1337
+    },
+    {
+      "epoch": 0.7250067732321864,
+      "grad_norm": 0.11460467427968979,
+      "learning_rate": 0.0001,
+      "loss": 1.6064,
+      "step": 1338
+    },
+    {
+      "epoch": 0.7255486318070984,
+      "grad_norm": 0.10942284762859344,
+      "learning_rate": 0.0001,
+      "loss": 1.5735,
+      "step": 1339
+    },
+    {
+      "epoch": 0.7260904903820103,
+      "grad_norm": 0.11083227396011353,
+      "learning_rate": 0.0001,
+      "loss": 1.679,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7266323489569222,
+      "grad_norm": 0.11292646080255508,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 1341
+    },
+    {
+      "epoch": 0.7271742075318341,
+      "grad_norm": 0.1074613481760025,
+      "learning_rate": 0.0001,
+      "loss": 1.5761,
+      "step": 1342
+    },
+    {
+      "epoch": 0.7277160661067461,
+      "grad_norm": 0.11060116440057755,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 1343
+    },
+    {
+      "epoch": 0.7282579246816581,
+      "grad_norm": 0.1097818911075592,
+      "learning_rate": 0.0001,
+      "loss": 1.5837,
+      "step": 1344
+    },
+    {
+      "epoch": 0.72879978325657,
+      "grad_norm": 0.10460485517978668,
+      "learning_rate": 0.0001,
+      "loss": 1.5402,
+      "step": 1345
+    },
+    {
+      "epoch": 0.729341641831482,
+      "grad_norm": 0.10759514570236206,
+      "learning_rate": 0.0001,
+      "loss": 1.6523,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7298835004063939,
+      "grad_norm": 0.11438702791929245,
+      "learning_rate": 0.0001,
+      "loss": 1.6547,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7304253589813059,
+      "grad_norm": 0.11222861707210541,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 1348
+    },
+    {
+      "epoch": 0.7309672175562179,
+      "grad_norm": 0.11586595326662064,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 1349
+    },
+    {
+      "epoch": 0.7315090761311298,
+      "grad_norm": 0.115098737180233,
+      "learning_rate": 0.0001,
+      "loss": 1.5906,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7320509347060418,
+      "grad_norm": 0.1137763112783432,
+      "learning_rate": 0.0001,
+      "loss": 1.5883,
+      "step": 1351
+    },
+    {
+      "epoch": 0.7325927932809537,
+      "grad_norm": 0.12027715146541595,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 1352
+    },
+    {
+      "epoch": 0.7331346518558656,
+      "grad_norm": 0.11682108789682388,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 1353
+    },
+    {
+      "epoch": 0.7336765104307775,
+      "grad_norm": 0.11551779508590698,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 1354
+    },
+    {
+      "epoch": 0.7342183690056895,
+      "grad_norm": 0.11855588108301163,
+      "learning_rate": 0.0001,
+      "loss": 1.6419,
+      "step": 1355
+    },
+    {
+      "epoch": 0.7347602275806014,
+      "grad_norm": 0.10961764305830002,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 1356
+    },
+    {
+      "epoch": 0.7353020861555134,
+      "grad_norm": 0.11965698003768921,
+      "learning_rate": 0.0001,
+      "loss": 1.5622,
+      "step": 1357
+    },
+    {
+      "epoch": 0.7358439447304254,
+      "grad_norm": 0.11887028068304062,
+      "learning_rate": 0.0001,
+      "loss": 1.5606,
+      "step": 1358
+    },
+    {
+      "epoch": 0.7363858033053373,
+      "grad_norm": 0.12262041866779327,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 1359
+    },
+    {
+      "epoch": 0.7369276618802493,
+      "grad_norm": 0.12121155858039856,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7374695204551612,
+      "grad_norm": 0.1178460344672203,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 1361
+    },
+    {
+      "epoch": 0.7380113790300732,
+      "grad_norm": 0.1300288289785385,
+      "learning_rate": 0.0001,
+      "loss": 1.5982,
+      "step": 1362
+    },
+    {
+      "epoch": 0.7385532376049851,
+      "grad_norm": 0.10765120387077332,
+      "learning_rate": 0.0001,
+      "loss": 1.5586,
+      "step": 1363
+    },
+    {
+      "epoch": 0.739095096179897,
+      "grad_norm": 0.12704487144947052,
+      "learning_rate": 0.0001,
+      "loss": 1.6121,
+      "step": 1364
+    },
+    {
+      "epoch": 0.739636954754809,
+      "grad_norm": 0.11278389394283295,
+      "learning_rate": 0.0001,
+      "loss": 1.6363,
+      "step": 1365
+    },
+    {
+      "epoch": 0.7401788133297209,
+      "grad_norm": 0.11722677946090698,
+      "learning_rate": 0.0001,
+      "loss": 1.6207,
+      "step": 1366
+    },
+    {
+      "epoch": 0.7407206719046329,
+      "grad_norm": 0.11791490763425827,
+      "learning_rate": 0.0001,
+      "loss": 1.5942,
+      "step": 1367
+    },
+    {
+      "epoch": 0.7412625304795448,
+      "grad_norm": 0.10806185752153397,
+      "learning_rate": 0.0001,
+      "loss": 1.5503,
+      "step": 1368
+    },
+    {
+      "epoch": 0.7418043890544568,
+      "grad_norm": 0.11138779670000076,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 1369
+    },
+    {
+      "epoch": 0.7423462476293687,
+      "grad_norm": 0.12193464487791061,
+      "learning_rate": 0.0001,
+      "loss": 1.6423,
+      "step": 1370
+    },
+    {
+      "epoch": 0.7428881062042807,
+      "grad_norm": 0.12534891068935394,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 1371
+    },
+    {
+      "epoch": 0.7434299647791927,
+      "grad_norm": 0.12328426539897919,
+      "learning_rate": 0.0001,
+      "loss": 1.6351,
+      "step": 1372
+    },
+    {
+      "epoch": 0.7439718233541046,
+      "grad_norm": 0.11469186097383499,
+      "learning_rate": 0.0001,
+      "loss": 1.638,
+      "step": 1373
+    },
+    {
+      "epoch": 0.7445136819290166,
+      "grad_norm": 0.13587550818920135,
+      "learning_rate": 0.0001,
+      "loss": 1.5469,
+      "step": 1374
+    },
+    {
+      "epoch": 0.7450555405039285,
+      "grad_norm": 0.125528484582901,
+      "learning_rate": 0.0001,
+      "loss": 1.6619,
+      "step": 1375
+    },
+    {
+      "epoch": 0.7455973990788404,
+      "grad_norm": 0.12690255045890808,
+      "learning_rate": 0.0001,
+      "loss": 1.6139,
+      "step": 1376
+    },
+    {
+      "epoch": 0.7461392576537523,
+      "grad_norm": 0.12415812164545059,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 1377
+    },
+    {
+      "epoch": 0.7466811162286643,
+      "grad_norm": 0.12645591795444489,
+      "learning_rate": 0.0001,
+      "loss": 1.5801,
+      "step": 1378
+    },
+    {
+      "epoch": 0.7472229748035762,
+      "grad_norm": 0.13845498859882355,
+      "learning_rate": 0.0001,
+      "loss": 1.6025,
+      "step": 1379
+    },
+    {
+      "epoch": 0.7477648333784882,
+      "grad_norm": 0.11183203011751175,
+      "learning_rate": 0.0001,
+      "loss": 1.6569,
+      "step": 1380
+    },
+    {
+      "epoch": 0.7483066919534002,
+      "grad_norm": 0.12703092396259308,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 1381
+    },
+    {
+      "epoch": 0.7488485505283121,
+      "grad_norm": 0.11635295301675797,
+      "learning_rate": 0.0001,
+      "loss": 1.6565,
+      "step": 1382
+    },
+    {
+      "epoch": 0.7493904091032241,
+      "grad_norm": 0.12178429961204529,
+      "learning_rate": 0.0001,
+      "loss": 1.6406,
+      "step": 1383
+    },
+    {
+      "epoch": 0.749932267678136,
+      "grad_norm": 0.11872105300426483,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 1384
+    },
+    {
+      "epoch": 0.750474126253048,
+      "grad_norm": 0.11461754888296127,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 1385
+    },
+    {
+      "epoch": 0.75101598482796,
+      "grad_norm": 0.11746447533369064,
+      "learning_rate": 0.0001,
+      "loss": 1.5677,
+      "step": 1386
+    },
+    {
+      "epoch": 0.7515578434028719,
+      "grad_norm": 0.11214593797922134,
+      "learning_rate": 0.0001,
+      "loss": 1.5812,
+      "step": 1387
+    },
+    {
+      "epoch": 0.7520997019777838,
+      "grad_norm": 0.12819403409957886,
+      "learning_rate": 0.0001,
+      "loss": 1.5942,
+      "step": 1388
+    },
+    {
+      "epoch": 0.7526415605526957,
+      "grad_norm": 0.11914908140897751,
+      "learning_rate": 0.0001,
+      "loss": 1.5917,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7531834191276077,
+      "grad_norm": 0.1351640224456787,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 1390
+    },
+    {
+      "epoch": 0.7537252777025196,
+      "grad_norm": 0.10753028094768524,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 1391
+    },
+    {
+      "epoch": 0.7542671362774316,
+      "grad_norm": 0.12622836232185364,
+      "learning_rate": 0.0001,
+      "loss": 1.5256,
+      "step": 1392
+    },
+    {
+      "epoch": 0.7548089948523435,
+      "grad_norm": 0.12361244857311249,
+      "learning_rate": 0.0001,
+      "loss": 1.5888,
+      "step": 1393
+    },
+    {
+      "epoch": 0.7553508534272555,
+      "grad_norm": 0.12058836221694946,
+      "learning_rate": 0.0001,
+      "loss": 1.5289,
+      "step": 1394
+    },
+    {
+      "epoch": 0.7558927120021675,
+      "grad_norm": 0.12233013659715652,
+      "learning_rate": 0.0001,
+      "loss": 1.6512,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7564345705770794,
+      "grad_norm": 0.1099410429596901,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 1396
+    },
+    {
+      "epoch": 0.7569764291519914,
+      "grad_norm": 0.1292077898979187,
+      "learning_rate": 0.0001,
+      "loss": 1.6571,
+      "step": 1397
+    },
+    {
+      "epoch": 0.7575182877269033,
+      "grad_norm": 0.11043606698513031,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 1398
+    },
+    {
+      "epoch": 0.7580601463018152,
+      "grad_norm": 0.12485775351524353,
+      "learning_rate": 0.0001,
+      "loss": 1.581,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7586020048767271,
+      "grad_norm": 0.12309932708740234,
+      "learning_rate": 0.0001,
+      "loss": 1.5904,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7591438634516391,
+      "grad_norm": 0.1205844134092331,
+      "learning_rate": 0.0001,
+      "loss": 1.6394,
+      "step": 1401
+    },
+    {
+      "epoch": 0.759685722026551,
+      "grad_norm": 0.1165509968996048,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 1402
+    },
+    {
+      "epoch": 0.760227580601463,
+      "grad_norm": 0.10734628140926361,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 1403
+    },
+    {
+      "epoch": 0.760769439176375,
+      "grad_norm": 0.11562097817659378,
+      "learning_rate": 0.0001,
+      "loss": 1.5174,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7613112977512869,
+      "grad_norm": 0.11254249513149261,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7618531563261989,
+      "grad_norm": 0.12835754454135895,
+      "learning_rate": 0.0001,
+      "loss": 1.6195,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7623950149011108,
+      "grad_norm": 0.11308079212903976,
+      "learning_rate": 0.0001,
+      "loss": 1.6227,
+      "step": 1407
+    },
+    {
+      "epoch": 0.7629368734760228,
+      "grad_norm": 0.10926039516925812,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7634787320509347,
+      "grad_norm": 0.11723046004772186,
+      "learning_rate": 0.0001,
+      "loss": 1.6109,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7640205906258467,
+      "grad_norm": 0.11550256609916687,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7645624492007586,
+      "grad_norm": 0.11019278317689896,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7651043077756705,
+      "grad_norm": 0.10726244002580643,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7656461663505825,
+      "grad_norm": 0.11210119724273682,
+      "learning_rate": 0.0001,
+      "loss": 1.6542,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7661880249254944,
+      "grad_norm": 0.12100458890199661,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7667298835004064,
+      "grad_norm": 0.1142323911190033,
+      "learning_rate": 0.0001,
+      "loss": 1.6707,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7672717420753183,
+      "grad_norm": 0.11224039644002914,
+      "learning_rate": 0.0001,
+      "loss": 1.5877,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7678136006502303,
+      "grad_norm": 0.10693727433681488,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7683554592251423,
+      "grad_norm": 0.10863177478313446,
+      "learning_rate": 0.0001,
+      "loss": 1.6159,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7688973178000542,
+      "grad_norm": 0.11036339402198792,
+      "learning_rate": 0.0001,
+      "loss": 1.6452,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7694391763749662,
+      "grad_norm": 0.11554139107465744,
+      "learning_rate": 0.0001,
+      "loss": 1.7011,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7699810349498781,
+      "grad_norm": 0.10645373910665512,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 1421
+    },
+    {
+      "epoch": 0.77052289352479,
+      "grad_norm": 0.1134442389011383,
+      "learning_rate": 0.0001,
+      "loss": 1.5695,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7710647520997019,
+      "grad_norm": 0.11070393770933151,
+      "learning_rate": 0.0001,
+      "loss": 1.5552,
+      "step": 1423
+    },
+    {
+      "epoch": 0.7716066106746139,
+      "grad_norm": 0.10615087300539017,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7721484692495258,
+      "grad_norm": 0.11273098737001419,
+      "learning_rate": 0.0001,
+      "loss": 1.5949,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7726903278244378,
+      "grad_norm": 0.11117110401391983,
+      "learning_rate": 0.0001,
+      "loss": 1.6057,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7732321863993498,
+      "grad_norm": 0.11692970991134644,
+      "learning_rate": 0.0001,
+      "loss": 1.608,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7737740449742617,
+      "grad_norm": 0.11009499430656433,
+      "learning_rate": 0.0001,
+      "loss": 1.5598,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7743159035491737,
+      "grad_norm": 0.11022921651601791,
+      "learning_rate": 0.0001,
+      "loss": 1.5573,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7748577621240856,
+      "grad_norm": 0.11034803092479706,
+      "learning_rate": 0.0001,
+      "loss": 1.6008,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7753996206989976,
+      "grad_norm": 0.10987105965614319,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7759414792739096,
+      "grad_norm": 0.10811598598957062,
+      "learning_rate": 0.0001,
+      "loss": 1.5902,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7764833378488215,
+      "grad_norm": 0.11513319611549377,
+      "learning_rate": 0.0001,
+      "loss": 1.6208,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7770251964237334,
+      "grad_norm": 0.10890763998031616,
+      "learning_rate": 0.0001,
+      "loss": 1.6296,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7775670549986453,
+      "grad_norm": 0.11387001723051071,
+      "learning_rate": 0.0001,
+      "loss": 1.6355,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7781089135735573,
+      "grad_norm": 0.11095337569713593,
+      "learning_rate": 0.0001,
+      "loss": 1.5434,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7786507721484692,
+      "grad_norm": 0.11978809535503387,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7791926307233812,
+      "grad_norm": 0.10958437621593475,
+      "learning_rate": 0.0001,
+      "loss": 1.6057,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7797344892982931,
+      "grad_norm": 0.11390339583158493,
+      "learning_rate": 0.0001,
+      "loss": 1.599,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7802763478732051,
+      "grad_norm": 0.11856385320425034,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 1440
+    },
+    {
+      "epoch": 0.7808182064481171,
+      "grad_norm": 0.11179933696985245,
+      "learning_rate": 0.0001,
+      "loss": 1.589,
+      "step": 1441
+    },
+    {
+      "epoch": 0.781360065023029,
+      "grad_norm": 0.11821099370718002,
+      "learning_rate": 0.0001,
+      "loss": 1.6782,
+      "step": 1442
+    },
+    {
+      "epoch": 0.781901923597941,
+      "grad_norm": 0.11423931270837784,
+      "learning_rate": 0.0001,
+      "loss": 1.6673,
+      "step": 1443
+    },
+    {
+      "epoch": 0.7824437821728529,
+      "grad_norm": 0.11144139617681503,
+      "learning_rate": 0.0001,
+      "loss": 1.5897,
+      "step": 1444
+    },
+    {
+      "epoch": 0.7829856407477649,
+      "grad_norm": 0.11281657218933105,
+      "learning_rate": 0.0001,
+      "loss": 1.6016,
+      "step": 1445
+    },
+    {
+      "epoch": 0.7835274993226767,
+      "grad_norm": 0.12321026623249054,
+      "learning_rate": 0.0001,
+      "loss": 1.6521,
+      "step": 1446
+    },
+    {
+      "epoch": 0.7840693578975887,
+      "grad_norm": 0.11253192275762558,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 1447
+    },
+    {
+      "epoch": 0.7846112164725006,
+      "grad_norm": 0.12315592169761658,
+      "learning_rate": 0.0001,
+      "loss": 1.6874,
+      "step": 1448
+    },
+    {
+      "epoch": 0.7851530750474126,
+      "grad_norm": 0.11318863183259964,
+      "learning_rate": 0.0001,
+      "loss": 1.4853,
+      "step": 1449
+    },
+    {
+      "epoch": 0.7856949336223246,
+      "grad_norm": 0.11038617789745331,
+      "learning_rate": 0.0001,
+      "loss": 1.5986,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7862367921972365,
+      "grad_norm": 0.12725314497947693,
+      "learning_rate": 0.0001,
+      "loss": 1.5371,
+      "step": 1451
+    },
+    {
+      "epoch": 0.7867786507721485,
+      "grad_norm": 0.11323723942041397,
+      "learning_rate": 0.0001,
+      "loss": 1.6198,
+      "step": 1452
+    },
+    {
+      "epoch": 0.7873205093470604,
+      "grad_norm": 0.13219019770622253,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 1453
+    },
+    {
+      "epoch": 0.7878623679219724,
+      "grad_norm": 0.11126276105642319,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7884042264968844,
+      "grad_norm": 0.12477418780326843,
+      "learning_rate": 0.0001,
+      "loss": 1.633,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7889460850717963,
+      "grad_norm": 0.11030017584562302,
+      "learning_rate": 0.0001,
+      "loss": 1.5997,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7894879436467082,
+      "grad_norm": 0.11762640625238419,
+      "learning_rate": 0.0001,
+      "loss": 1.595,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7900298022216201,
+      "grad_norm": 0.11468858271837234,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7905716607965321,
+      "grad_norm": 0.11451311409473419,
+      "learning_rate": 0.0001,
+      "loss": 1.5491,
+      "step": 1459
+    },
+    {
+      "epoch": 0.791113519371444,
+      "grad_norm": 0.12499954551458359,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 1460
+    },
+    {
+      "epoch": 0.791655377946356,
+      "grad_norm": 0.11596235632896423,
+      "learning_rate": 0.0001,
+      "loss": 1.5977,
+      "step": 1461
+    },
+    {
+      "epoch": 0.792197236521268,
+      "grad_norm": 0.12140510976314545,
+      "learning_rate": 0.0001,
+      "loss": 1.6266,
+      "step": 1462
+    },
+    {
+      "epoch": 0.7927390950961799,
+      "grad_norm": 0.11294886469841003,
+      "learning_rate": 0.0001,
+      "loss": 1.5746,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7932809536710919,
+      "grad_norm": 0.12936176359653473,
+      "learning_rate": 0.0001,
+      "loss": 1.5847,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7938228122460038,
+      "grad_norm": 0.10991284996271133,
+      "learning_rate": 0.0001,
+      "loss": 1.5691,
+      "step": 1465
+    },
+    {
+      "epoch": 0.7943646708209158,
+      "grad_norm": 0.12364400923252106,
+      "learning_rate": 0.0001,
+      "loss": 1.5882,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7949065293958277,
+      "grad_norm": 0.12108241766691208,
+      "learning_rate": 0.0001,
+      "loss": 1.6564,
+      "step": 1467
+    },
+    {
+      "epoch": 0.7954483879707397,
+      "grad_norm": 0.12121594697237015,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7959902465456515,
+      "grad_norm": 0.11251336336135864,
+      "learning_rate": 0.0001,
+      "loss": 1.5009,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7965321051205635,
+      "grad_norm": 0.10613922774791718,
+      "learning_rate": 0.0001,
+      "loss": 1.4703,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7970739636954755,
+      "grad_norm": 0.11324452608823776,
+      "learning_rate": 0.0001,
+      "loss": 1.588,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7976158222703874,
+      "grad_norm": 0.12203339487314224,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7981576808452994,
+      "grad_norm": 0.11607290059328079,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 1473
+    },
+    {
+      "epoch": 0.7986995394202113,
+      "grad_norm": 0.11903396248817444,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7992413979951233,
+      "grad_norm": 0.11452293395996094,
+      "learning_rate": 0.0001,
+      "loss": 1.608,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7997832565700352,
+      "grad_norm": 0.1214212104678154,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 1476
+    },
+    {
+      "epoch": 0.8003251151449472,
+      "grad_norm": 0.11239127069711685,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 1477
+    },
+    {
+      "epoch": 0.8008669737198592,
+      "grad_norm": 0.12349693477153778,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 1478
+    },
+    {
+      "epoch": 0.8014088322947711,
+      "grad_norm": 0.11215246468782425,
+      "learning_rate": 0.0001,
+      "loss": 1.5435,
+      "step": 1479
+    },
+    {
+      "epoch": 0.801950690869683,
+      "grad_norm": 0.12210462242364883,
+      "learning_rate": 0.0001,
+      "loss": 1.6145,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8024925494445949,
+      "grad_norm": 0.1136668398976326,
+      "learning_rate": 0.0001,
+      "loss": 1.6317,
+      "step": 1481
+    },
+    {
+      "epoch": 0.8030344080195069,
+      "grad_norm": 0.1275959461927414,
+      "learning_rate": 0.0001,
+      "loss": 1.6444,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8035762665944188,
+      "grad_norm": 0.11355528235435486,
+      "learning_rate": 0.0001,
+      "loss": 1.6268,
+      "step": 1483
+    },
+    {
+      "epoch": 0.8041181251693308,
+      "grad_norm": 0.11473184078931808,
+      "learning_rate": 0.0001,
+      "loss": 1.5067,
+      "step": 1484
+    },
+    {
+      "epoch": 0.8046599837442427,
+      "grad_norm": 0.12829527258872986,
+      "learning_rate": 0.0001,
+      "loss": 1.6146,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8052018423191547,
+      "grad_norm": 0.1119534894824028,
+      "learning_rate": 0.0001,
+      "loss": 1.5945,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8057437008940667,
+      "grad_norm": 0.12512047588825226,
+      "learning_rate": 0.0001,
+      "loss": 1.6249,
+      "step": 1487
+    },
+    {
+      "epoch": 0.8062855594689786,
+      "grad_norm": 0.11971007287502289,
+      "learning_rate": 0.0001,
+      "loss": 1.6031,
+      "step": 1488
+    },
+    {
+      "epoch": 0.8068274180438906,
+      "grad_norm": 0.11718935519456863,
+      "learning_rate": 0.0001,
+      "loss": 1.4837,
+      "step": 1489
+    },
+    {
+      "epoch": 0.8073692766188025,
+      "grad_norm": 0.121201291680336,
+      "learning_rate": 0.0001,
+      "loss": 1.6057,
+      "step": 1490
+    },
+    {
+      "epoch": 0.8079111351937145,
+      "grad_norm": 0.12238092720508575,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 1491
+    },
+    {
+      "epoch": 0.8084529937686263,
+      "grad_norm": 0.1160135269165039,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 1492
+    },
+    {
+      "epoch": 0.8089948523435383,
+      "grad_norm": 0.12266741693019867,
+      "learning_rate": 0.0001,
+      "loss": 1.578,
+      "step": 1493
+    },
+    {
+      "epoch": 0.8095367109184503,
+      "grad_norm": 0.12016115337610245,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 1494
+    },
+    {
+      "epoch": 0.8100785694933622,
+      "grad_norm": 0.12188202142715454,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 1495
+    },
+    {
+      "epoch": 0.8106204280682742,
+      "grad_norm": 0.12086134403944016,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 1496
+    },
+    {
+      "epoch": 0.8111622866431861,
+      "grad_norm": 0.11443319916725159,
+      "learning_rate": 0.0001,
+      "loss": 1.6446,
+      "step": 1497
+    },
+    {
+      "epoch": 0.8117041452180981,
+      "grad_norm": 0.11497814953327179,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 1498
+    },
+    {
+      "epoch": 0.81224600379301,
+      "grad_norm": 0.11970335990190506,
+      "learning_rate": 0.0001,
+      "loss": 1.6244,
+      "step": 1499
+    },
+    {
+      "epoch": 0.812787862367922,
+      "grad_norm": 0.11911267787218094,
+      "learning_rate": 0.0001,
+      "loss": 1.6128,
+      "step": 1500
+    },
+    {
+      "epoch": 0.813329720942834,
+      "grad_norm": 0.11438138782978058,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 1501
+    },
+    {
+      "epoch": 0.8138715795177459,
+      "grad_norm": 0.1144600659608841,
+      "learning_rate": 0.0001,
+      "loss": 1.5604,
+      "step": 1502
+    },
+    {
+      "epoch": 0.8144134380926579,
+      "grad_norm": 0.11835461854934692,
+      "learning_rate": 0.0001,
+      "loss": 1.5793,
+      "step": 1503
+    },
+    {
+      "epoch": 0.8149552966675697,
+      "grad_norm": 0.11068934202194214,
+      "learning_rate": 0.0001,
+      "loss": 1.5806,
+      "step": 1504
+    },
+    {
+      "epoch": 0.8154971552424817,
+      "grad_norm": 0.11289520561695099,
+      "learning_rate": 0.0001,
+      "loss": 1.6141,
+      "step": 1505
+    },
+    {
+      "epoch": 0.8160390138173936,
+      "grad_norm": 0.11913792043924332,
+      "learning_rate": 0.0001,
+      "loss": 1.5351,
+      "step": 1506
+    },
+    {
+      "epoch": 0.8165808723923056,
+      "grad_norm": 0.11238417029380798,
+      "learning_rate": 0.0001,
+      "loss": 1.5889,
+      "step": 1507
+    },
+    {
+      "epoch": 0.8171227309672175,
+      "grad_norm": 0.14196346700191498,
+      "learning_rate": 0.0001,
+      "loss": 1.6738,
+      "step": 1508
+    },
+    {
+      "epoch": 0.8176645895421295,
+      "grad_norm": 0.1079757958650589,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 1509
+    },
+    {
+      "epoch": 0.8182064481170415,
+      "grad_norm": 0.12480804324150085,
+      "learning_rate": 0.0001,
+      "loss": 1.6725,
+      "step": 1510
+    },
+    {
+      "epoch": 0.8187483066919534,
+      "grad_norm": 0.12916377186775208,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 1511
+    },
+    {
+      "epoch": 0.8192901652668654,
+      "grad_norm": 0.11713727563619614,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 1512
+    },
+    {
+      "epoch": 0.8198320238417773,
+      "grad_norm": 0.13380920886993408,
+      "learning_rate": 0.0001,
+      "loss": 1.5758,
+      "step": 1513
+    },
+    {
+      "epoch": 0.8203738824166893,
+      "grad_norm": 0.11877034604549408,
+      "learning_rate": 0.0001,
+      "loss": 1.5613,
+      "step": 1514
+    },
+    {
+      "epoch": 0.8209157409916011,
+      "grad_norm": 0.12588542699813843,
+      "learning_rate": 0.0001,
+      "loss": 1.6197,
+      "step": 1515
+    },
+    {
+      "epoch": 0.8214575995665131,
+      "grad_norm": 0.1106722354888916,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 1516
+    },
+    {
+      "epoch": 0.821999458141425,
+      "grad_norm": 0.1203051209449768,
+      "learning_rate": 0.0001,
+      "loss": 1.6268,
+      "step": 1517
+    },
+    {
+      "epoch": 0.822541316716337,
+      "grad_norm": 0.10928182303905487,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 1518
+    },
+    {
+      "epoch": 0.823083175291249,
+      "grad_norm": 0.12915323674678802,
+      "learning_rate": 0.0001,
+      "loss": 1.625,
+      "step": 1519
+    },
+    {
+      "epoch": 0.8236250338661609,
+      "grad_norm": 0.11285202205181122,
+      "learning_rate": 0.0001,
+      "loss": 1.5965,
+      "step": 1520
+    },
+    {
+      "epoch": 0.8241668924410729,
+      "grad_norm": 0.1192052811384201,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 1521
+    },
+    {
+      "epoch": 0.8247087510159848,
+      "grad_norm": 0.10996171832084656,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 1522
+    },
+    {
+      "epoch": 0.8252506095908968,
+      "grad_norm": 0.13144908845424652,
+      "learning_rate": 0.0001,
+      "loss": 1.6359,
+      "step": 1523
+    },
+    {
+      "epoch": 0.8257924681658088,
+      "grad_norm": 0.13091161847114563,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 1524
+    },
+    {
+      "epoch": 0.8263343267407207,
+      "grad_norm": 0.1159205511212349,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 1525
+    },
+    {
+      "epoch": 0.8268761853156327,
+      "grad_norm": 0.14918328821659088,
+      "learning_rate": 0.0001,
+      "loss": 1.6605,
+      "step": 1526
+    },
+    {
+      "epoch": 0.8274180438905445,
+      "grad_norm": 0.11102715879678726,
+      "learning_rate": 0.0001,
+      "loss": 1.62,
+      "step": 1527
+    },
+    {
+      "epoch": 0.8279599024654565,
+      "grad_norm": 0.13266460597515106,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 1528
+    },
+    {
+      "epoch": 0.8285017610403684,
+      "grad_norm": 0.12381194531917572,
+      "learning_rate": 0.0001,
+      "loss": 1.6163,
+      "step": 1529
+    },
+    {
+      "epoch": 0.8290436196152804,
+      "grad_norm": 0.1211717426776886,
+      "learning_rate": 0.0001,
+      "loss": 1.5247,
+      "step": 1530
+    },
+    {
+      "epoch": 0.8295854781901923,
+      "grad_norm": 0.13359831273555756,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 1531
+    },
+    {
+      "epoch": 0.8301273367651043,
+      "grad_norm": 0.11108848452568054,
+      "learning_rate": 0.0001,
+      "loss": 1.5986,
+      "step": 1532
+    },
+    {
+      "epoch": 0.8306691953400163,
+      "grad_norm": 0.149922713637352,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 1533
+    },
+    {
+      "epoch": 0.8312110539149282,
+      "grad_norm": 0.11118970066308975,
+      "learning_rate": 0.0001,
+      "loss": 1.6018,
+      "step": 1534
+    },
+    {
+      "epoch": 0.8317529124898402,
+      "grad_norm": 0.13041344285011292,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 1535
+    },
+    {
+      "epoch": 0.8322947710647521,
+      "grad_norm": 0.11675264686346054,
+      "learning_rate": 0.0001,
+      "loss": 1.4878,
+      "step": 1536
+    },
+    {
+      "epoch": 0.8328366296396641,
+      "grad_norm": 0.13788774609565735,
+      "learning_rate": 0.0001,
+      "loss": 1.6123,
+      "step": 1537
+    },
+    {
+      "epoch": 0.8333784882145759,
+      "grad_norm": 0.12169042974710464,
+      "learning_rate": 0.0001,
+      "loss": 1.6448,
+      "step": 1538
+    },
+    {
+      "epoch": 0.8339203467894879,
+      "grad_norm": 0.11629584431648254,
+      "learning_rate": 0.0001,
+      "loss": 1.5966,
+      "step": 1539
+    },
+    {
+      "epoch": 0.8344622053643999,
+      "grad_norm": 0.12246973067522049,
+      "learning_rate": 0.0001,
+      "loss": 1.6119,
+      "step": 1540
+    },
+    {
+      "epoch": 0.8350040639393118,
+      "grad_norm": 0.11287442594766617,
+      "learning_rate": 0.0001,
+      "loss": 1.5607,
+      "step": 1541
+    },
+    {
+      "epoch": 0.8355459225142238,
+      "grad_norm": 0.11604714393615723,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 1542
+    },
+    {
+      "epoch": 0.8360877810891357,
+      "grad_norm": 0.11228446662425995,
+      "learning_rate": 0.0001,
+      "loss": 1.5365,
+      "step": 1543
+    },
+    {
+      "epoch": 0.8366296396640477,
+      "grad_norm": 0.12683261930942535,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 1544
+    },
+    {
+      "epoch": 0.8371714982389596,
+      "grad_norm": 0.11033685505390167,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 1545
+    },
+    {
+      "epoch": 0.8377133568138716,
+      "grad_norm": 0.1107892394065857,
+      "learning_rate": 0.0001,
+      "loss": 1.5048,
+      "step": 1546
+    },
+    {
+      "epoch": 0.8382552153887836,
+      "grad_norm": 0.11377444863319397,
+      "learning_rate": 0.0001,
+      "loss": 1.6175,
+      "step": 1547
+    },
+    {
+      "epoch": 0.8387970739636955,
+      "grad_norm": 0.10976476967334747,
+      "learning_rate": 0.0001,
+      "loss": 1.5409,
+      "step": 1548
+    },
+    {
+      "epoch": 0.8393389325386075,
+      "grad_norm": 0.11617878079414368,
+      "learning_rate": 0.0001,
+      "loss": 1.6056,
+      "step": 1549
+    },
+    {
+      "epoch": 0.8398807911135193,
+      "grad_norm": 0.11169909685850143,
+      "learning_rate": 0.0001,
+      "loss": 1.6166,
+      "step": 1550
+    },
+    {
+      "epoch": 0.8404226496884313,
+      "grad_norm": 0.1158488541841507,
+      "learning_rate": 0.0001,
+      "loss": 1.6312,
+      "step": 1551
+    },
+    {
+      "epoch": 0.8409645082633432,
+      "grad_norm": 0.11124692112207413,
+      "learning_rate": 0.0001,
+      "loss": 1.6399,
+      "step": 1552
+    },
+    {
+      "epoch": 0.8415063668382552,
+      "grad_norm": 0.11569403111934662,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 1553
+    },
+    {
+      "epoch": 0.8420482254131672,
+      "grad_norm": 0.1057923287153244,
+      "learning_rate": 0.0001,
+      "loss": 1.5125,
+      "step": 1554
+    },
+    {
+      "epoch": 0.8425900839880791,
+      "grad_norm": 0.11266665160655975,
+      "learning_rate": 0.0001,
+      "loss": 1.5761,
+      "step": 1555
+    },
+    {
+      "epoch": 0.8431319425629911,
+      "grad_norm": 0.1118483692407608,
+      "learning_rate": 0.0001,
+      "loss": 1.6092,
+      "step": 1556
+    },
+    {
+      "epoch": 0.843673801137903,
+      "grad_norm": 0.12413822114467621,
+      "learning_rate": 0.0001,
+      "loss": 1.7189,
+      "step": 1557
+    },
+    {
+      "epoch": 0.844215659712815,
+      "grad_norm": 0.11185813695192337,
+      "learning_rate": 0.0001,
+      "loss": 1.5966,
+      "step": 1558
+    },
+    {
+      "epoch": 0.8447575182877269,
+      "grad_norm": 0.11938433349132538,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 1559
+    },
+    {
+      "epoch": 0.8452993768626389,
+      "grad_norm": 0.1135723888874054,
+      "learning_rate": 0.0001,
+      "loss": 1.5854,
+      "step": 1560
+    },
+    {
+      "epoch": 0.8458412354375509,
+      "grad_norm": 0.1225573718547821,
+      "learning_rate": 0.0001,
+      "loss": 1.6488,
+      "step": 1561
+    },
+    {
+      "epoch": 0.8463830940124627,
+      "grad_norm": 0.11226433515548706,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 1562
+    },
+    {
+      "epoch": 0.8469249525873747,
+      "grad_norm": 0.13214121758937836,
+      "learning_rate": 0.0001,
+      "loss": 1.7078,
+      "step": 1563
+    },
+    {
+      "epoch": 0.8474668111622866,
+      "grad_norm": 0.1112339049577713,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 1564
+    },
+    {
+      "epoch": 0.8480086697371986,
+      "grad_norm": 0.12047260254621506,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 1565
+    },
+    {
+      "epoch": 0.8485505283121105,
+      "grad_norm": 0.1157664805650711,
+      "learning_rate": 0.0001,
+      "loss": 1.5619,
+      "step": 1566
+    },
+    {
+      "epoch": 0.8490923868870225,
+      "grad_norm": 0.11576071381568909,
+      "learning_rate": 0.0001,
+      "loss": 1.6143,
+      "step": 1567
+    },
+    {
+      "epoch": 0.8496342454619344,
+      "grad_norm": 0.11805963516235352,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 1568
+    },
+    {
+      "epoch": 0.8501761040368464,
+      "grad_norm": 0.11272211372852325,
+      "learning_rate": 0.0001,
+      "loss": 1.6067,
+      "step": 1569
+    },
+    {
+      "epoch": 0.8507179626117584,
+      "grad_norm": 0.12440693378448486,
+      "learning_rate": 0.0001,
+      "loss": 1.5523,
+      "step": 1570
+    },
+    {
+      "epoch": 0.8512598211866703,
+      "grad_norm": 0.1087651476264,
+      "learning_rate": 0.0001,
+      "loss": 1.5478,
+      "step": 1571
+    },
+    {
+      "epoch": 0.8518016797615823,
+      "grad_norm": 0.130729079246521,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 1572
+    },
+    {
+      "epoch": 0.8523435383364941,
+      "grad_norm": 0.11392576992511749,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 1573
+    },
+    {
+      "epoch": 0.8528853969114061,
+      "grad_norm": 0.12628479301929474,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 1574
+    },
+    {
+      "epoch": 0.853427255486318,
+      "grad_norm": 0.11537759751081467,
+      "learning_rate": 0.0001,
+      "loss": 1.5839,
+      "step": 1575
+    },
+    {
+      "epoch": 0.85396911406123,
+      "grad_norm": 0.12094999104738235,
+      "learning_rate": 0.0001,
+      "loss": 1.5822,
+      "step": 1576
+    },
+    {
+      "epoch": 0.854510972636142,
+      "grad_norm": 0.12219975888729095,
+      "learning_rate": 0.0001,
+      "loss": 1.6344,
+      "step": 1577
+    },
+    {
+      "epoch": 0.8550528312110539,
+      "grad_norm": 0.11441099643707275,
+      "learning_rate": 0.0001,
+      "loss": 1.6,
+      "step": 1578
+    },
+    {
+      "epoch": 0.8555946897859659,
+      "grad_norm": 0.11610981822013855,
+      "learning_rate": 0.0001,
+      "loss": 1.4884,
+      "step": 1579
+    },
+    {
+      "epoch": 0.8561365483608778,
+      "grad_norm": 0.11038801074028015,
+      "learning_rate": 0.0001,
+      "loss": 1.5624,
+      "step": 1580
+    },
+    {
+      "epoch": 0.8566784069357898,
+      "grad_norm": 0.1118963435292244,
+      "learning_rate": 0.0001,
+      "loss": 1.6111,
+      "step": 1581
+    },
+    {
+      "epoch": 0.8572202655107017,
+      "grad_norm": 0.10892907530069351,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 1582
+    },
+    {
+      "epoch": 0.8577621240856137,
+      "grad_norm": 0.11459089815616608,
+      "learning_rate": 0.0001,
+      "loss": 1.6161,
+      "step": 1583
+    },
+    {
+      "epoch": 0.8583039826605257,
+      "grad_norm": 0.11651195585727692,
+      "learning_rate": 0.0001,
+      "loss": 1.6321,
+      "step": 1584
+    },
+    {
+      "epoch": 0.8588458412354375,
+      "grad_norm": 0.11360273510217667,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 1585
+    },
+    {
+      "epoch": 0.8593876998103495,
+      "grad_norm": 0.1199052482843399,
+      "learning_rate": 0.0001,
+      "loss": 1.6462,
+      "step": 1586
+    },
+    {
+      "epoch": 0.8599295583852614,
+      "grad_norm": 0.10889112204313278,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 1587
+    },
+    {
+      "epoch": 0.8604714169601734,
+      "grad_norm": 0.11999368667602539,
+      "learning_rate": 0.0001,
+      "loss": 1.5743,
+      "step": 1588
+    },
+    {
+      "epoch": 0.8610132755350853,
+      "grad_norm": 0.11200165003538132,
+      "learning_rate": 0.0001,
+      "loss": 1.56,
+      "step": 1589
+    },
+    {
+      "epoch": 0.8615551341099973,
+      "grad_norm": 0.12450709939002991,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 1590
+    },
+    {
+      "epoch": 0.8620969926849092,
+      "grad_norm": 0.1191956177353859,
+      "learning_rate": 0.0001,
+      "loss": 1.5778,
+      "step": 1591
+    },
+    {
+      "epoch": 0.8626388512598212,
+      "grad_norm": 0.11459418386220932,
+      "learning_rate": 0.0001,
+      "loss": 1.6384,
+      "step": 1592
+    },
+    {
+      "epoch": 0.8631807098347332,
+      "grad_norm": 0.1097928062081337,
+      "learning_rate": 0.0001,
+      "loss": 1.5656,
+      "step": 1593
+    },
+    {
+      "epoch": 0.8637225684096451,
+      "grad_norm": 0.1171281486749649,
+      "learning_rate": 0.0001,
+      "loss": 1.6466,
+      "step": 1594
+    },
+    {
+      "epoch": 0.8642644269845571,
+      "grad_norm": 0.1114104762673378,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 1595
+    },
+    {
+      "epoch": 0.8648062855594689,
+      "grad_norm": 0.11036089062690735,
+      "learning_rate": 0.0001,
+      "loss": 1.6135,
+      "step": 1596
+    },
+    {
+      "epoch": 0.8653481441343809,
+      "grad_norm": 0.11589822173118591,
+      "learning_rate": 0.0001,
+      "loss": 1.6394,
+      "step": 1597
+    },
+    {
+      "epoch": 0.8658900027092928,
+      "grad_norm": 0.11399046331644058,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 1598
+    },
+    {
+      "epoch": 0.8664318612842048,
+      "grad_norm": 0.11825186759233475,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 1599
+    },
+    {
+      "epoch": 0.8669737198591168,
+      "grad_norm": 0.10993462055921555,
+      "learning_rate": 0.0001,
+      "loss": 1.5817,
+      "step": 1600
+    },
+    {
+      "epoch": 0.8675155784340287,
+      "grad_norm": 0.11864310503005981,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 1601
+    },
+    {
+      "epoch": 0.8680574370089407,
+      "grad_norm": 0.11332954466342926,
+      "learning_rate": 0.0001,
+      "loss": 1.6195,
+      "step": 1602
+    },
+    {
+      "epoch": 0.8685992955838526,
+      "grad_norm": 0.12260433286428452,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 1603
+    },
+    {
+      "epoch": 0.8691411541587646,
+      "grad_norm": 0.11385665833950043,
+      "learning_rate": 0.0001,
+      "loss": 1.6244,
+      "step": 1604
+    },
+    {
+      "epoch": 0.8696830127336765,
+      "grad_norm": 0.12241163849830627,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 1605
+    },
+    {
+      "epoch": 0.8702248713085885,
+      "grad_norm": 0.12055647373199463,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 1606
+    },
+    {
+      "epoch": 0.8707667298835005,
+      "grad_norm": 0.11905807256698608,
+      "learning_rate": 0.0001,
+      "loss": 1.5488,
+      "step": 1607
+    },
+    {
+      "epoch": 0.8713085884584123,
+      "grad_norm": 0.11884074658155441,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 1608
+    },
+    {
+      "epoch": 0.8718504470333243,
+      "grad_norm": 0.11740755289793015,
+      "learning_rate": 0.0001,
+      "loss": 1.5892,
+      "step": 1609
+    },
+    {
+      "epoch": 0.8723923056082362,
+      "grad_norm": 0.11741320043802261,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 1610
+    },
+    {
+      "epoch": 0.8729341641831482,
+      "grad_norm": 0.1233908161520958,
+      "learning_rate": 0.0001,
+      "loss": 1.6241,
+      "step": 1611
+    },
+    {
+      "epoch": 0.8734760227580601,
+      "grad_norm": 0.12131594866514206,
+      "learning_rate": 0.0001,
+      "loss": 1.5793,
+      "step": 1612
+    },
+    {
+      "epoch": 0.8740178813329721,
+      "grad_norm": 0.119773268699646,
+      "learning_rate": 0.0001,
+      "loss": 1.5435,
+      "step": 1613
+    },
+    {
+      "epoch": 0.874559739907884,
+      "grad_norm": 0.12392249703407288,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 1614
+    },
+    {
+      "epoch": 0.875101598482796,
+      "grad_norm": 0.11792998760938644,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 1615
+    },
+    {
+      "epoch": 0.875643457057708,
+      "grad_norm": 0.12786734104156494,
+      "learning_rate": 0.0001,
+      "loss": 1.6538,
+      "step": 1616
+    },
+    {
+      "epoch": 0.8761853156326199,
+      "grad_norm": 0.11728377640247345,
+      "learning_rate": 0.0001,
+      "loss": 1.6365,
+      "step": 1617
+    },
+    {
+      "epoch": 0.8767271742075319,
+      "grad_norm": 0.12557296454906464,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 1618
+    },
+    {
+      "epoch": 0.8772690327824438,
+      "grad_norm": 0.12002141028642654,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 1619
+    },
+    {
+      "epoch": 0.8778108913573557,
+      "grad_norm": 0.12448980659246445,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8783527499322676,
+      "grad_norm": 0.12446106970310211,
+      "learning_rate": 0.0001,
+      "loss": 1.6156,
+      "step": 1621
+    },
+    {
+      "epoch": 0.8788946085071796,
+      "grad_norm": 0.1173003539443016,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 1622
+    },
+    {
+      "epoch": 0.8794364670820916,
+      "grad_norm": 0.1369483768939972,
+      "learning_rate": 0.0001,
+      "loss": 1.6625,
+      "step": 1623
+    },
+    {
+      "epoch": 0.8799783256570035,
+      "grad_norm": 0.11268099397420883,
+      "learning_rate": 0.0001,
+      "loss": 1.5504,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8805201842319155,
+      "grad_norm": 0.11444398015737534,
+      "learning_rate": 0.0001,
+      "loss": 1.5284,
+      "step": 1625
+    },
+    {
+      "epoch": 0.8810620428068274,
+      "grad_norm": 0.116224005818367,
+      "learning_rate": 0.0001,
+      "loss": 1.6135,
+      "step": 1626
+    },
+    {
+      "epoch": 0.8816039013817394,
+      "grad_norm": 0.1233014464378357,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 1627
+    },
+    {
+      "epoch": 0.8821457599566513,
+      "grad_norm": 0.11984910815954208,
+      "learning_rate": 0.0001,
+      "loss": 1.5448,
+      "step": 1628
+    },
+    {
+      "epoch": 0.8826876185315633,
+      "grad_norm": 0.13135050237178802,
+      "learning_rate": 0.0001,
+      "loss": 1.6433,
+      "step": 1629
+    },
+    {
+      "epoch": 0.8832294771064753,
+      "grad_norm": 0.120527483522892,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 1630
+    },
+    {
+      "epoch": 0.8837713356813871,
+      "grad_norm": 0.13846857845783234,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 1631
+    },
+    {
+      "epoch": 0.8843131942562991,
+      "grad_norm": 0.11715717613697052,
+      "learning_rate": 0.0001,
+      "loss": 1.5607,
+      "step": 1632
+    },
+    {
+      "epoch": 0.884855052831211,
+      "grad_norm": 0.11982149630784988,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 1633
+    },
+    {
+      "epoch": 0.885396911406123,
+      "grad_norm": 0.11110392957925797,
+      "learning_rate": 0.0001,
+      "loss": 1.5525,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8859387699810349,
+      "grad_norm": 0.12085683643817902,
+      "learning_rate": 0.0001,
+      "loss": 1.5901,
+      "step": 1635
+    },
+    {
+      "epoch": 0.8864806285559469,
+      "grad_norm": 0.12121672928333282,
+      "learning_rate": 0.0001,
+      "loss": 1.5523,
+      "step": 1636
+    },
+    {
+      "epoch": 0.8870224871308588,
+      "grad_norm": 0.12281519919633865,
+      "learning_rate": 0.0001,
+      "loss": 1.6853,
+      "step": 1637
+    },
+    {
+      "epoch": 0.8875643457057708,
+      "grad_norm": 0.10896273702383041,
+      "learning_rate": 0.0001,
+      "loss": 1.516,
+      "step": 1638
+    },
+    {
+      "epoch": 0.8881062042806828,
+      "grad_norm": 0.11349642276763916,
+      "learning_rate": 0.0001,
+      "loss": 1.5398,
+      "step": 1639
+    },
+    {
+      "epoch": 0.8886480628555947,
+      "grad_norm": 0.11644469946622849,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 1640
+    },
+    {
+      "epoch": 0.8891899214305067,
+      "grad_norm": 0.11534909158945084,
+      "learning_rate": 0.0001,
+      "loss": 1.6435,
+      "step": 1641
+    },
+    {
+      "epoch": 0.8897317800054186,
+      "grad_norm": 0.11221238225698471,
+      "learning_rate": 0.0001,
+      "loss": 1.575,
+      "step": 1642
+    },
+    {
+      "epoch": 0.8902736385803305,
+      "grad_norm": 0.10964131355285645,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 1643
+    },
+    {
+      "epoch": 0.8908154971552424,
+      "grad_norm": 0.1157451868057251,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 1644
+    },
+    {
+      "epoch": 0.8913573557301544,
+      "grad_norm": 0.1303514987230301,
+      "learning_rate": 0.0001,
+      "loss": 1.5881,
+      "step": 1645
+    },
+    {
+      "epoch": 0.8918992143050664,
+      "grad_norm": 0.12164656072854996,
+      "learning_rate": 0.0001,
+      "loss": 1.6291,
+      "step": 1646
+    },
+    {
+      "epoch": 0.8924410728799783,
+      "grad_norm": 0.12777622044086456,
+      "learning_rate": 0.0001,
+      "loss": 1.5742,
+      "step": 1647
+    },
+    {
+      "epoch": 0.8929829314548903,
+      "grad_norm": 0.12283733487129211,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 1648
+    },
+    {
+      "epoch": 0.8935247900298022,
+      "grad_norm": 0.12476850301027298,
+      "learning_rate": 0.0001,
+      "loss": 1.6376,
+      "step": 1649
+    },
+    {
+      "epoch": 0.8940666486047142,
+      "grad_norm": 0.11374060809612274,
+      "learning_rate": 0.0001,
+      "loss": 1.5208,
+      "step": 1650
+    },
+    {
+      "epoch": 0.8946085071796261,
+      "grad_norm": 0.12660513818264008,
+      "learning_rate": 0.0001,
+      "loss": 1.5475,
+      "step": 1651
+    },
+    {
+      "epoch": 0.8951503657545381,
+      "grad_norm": 0.12348679453134537,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 1652
+    },
+    {
+      "epoch": 0.8956922243294501,
+      "grad_norm": 0.1383376121520996,
+      "learning_rate": 0.0001,
+      "loss": 1.5918,
+      "step": 1653
+    },
+    {
+      "epoch": 0.8962340829043619,
+      "grad_norm": 0.11630037426948547,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 1654
+    },
+    {
+      "epoch": 0.8967759414792739,
+      "grad_norm": 0.12958136200904846,
+      "learning_rate": 0.0001,
+      "loss": 1.5429,
+      "step": 1655
+    },
+    {
+      "epoch": 0.8973178000541858,
+      "grad_norm": 0.11349201202392578,
+      "learning_rate": 0.0001,
+      "loss": 1.5671,
+      "step": 1656
+    },
+    {
+      "epoch": 0.8978596586290978,
+      "grad_norm": 0.11807017773389816,
+      "learning_rate": 0.0001,
+      "loss": 1.5202,
+      "step": 1657
+    },
+    {
+      "epoch": 0.8984015172040097,
+      "grad_norm": 0.1251143217086792,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 1658
+    },
+    {
+      "epoch": 0.8989433757789217,
+      "grad_norm": 0.11176281422376633,
+      "learning_rate": 0.0001,
+      "loss": 1.6278,
+      "step": 1659
+    },
+    {
+      "epoch": 0.8994852343538337,
+      "grad_norm": 0.12428291141986847,
+      "learning_rate": 0.0001,
+      "loss": 1.5621,
+      "step": 1660
+    },
+    {
+      "epoch": 0.9000270929287456,
+      "grad_norm": 0.11493722349405289,
+      "learning_rate": 0.0001,
+      "loss": 1.5835,
+      "step": 1661
+    },
+    {
+      "epoch": 0.9005689515036576,
+      "grad_norm": 0.11907187104225159,
+      "learning_rate": 0.0001,
+      "loss": 1.5335,
+      "step": 1662
+    },
+    {
+      "epoch": 0.9011108100785695,
+      "grad_norm": 0.1341247409582138,
+      "learning_rate": 0.0001,
+      "loss": 1.6146,
+      "step": 1663
+    },
+    {
+      "epoch": 0.9016526686534815,
+      "grad_norm": 0.12102586776018143,
+      "learning_rate": 0.0001,
+      "loss": 1.6316,
+      "step": 1664
+    },
+    {
+      "epoch": 0.9021945272283934,
+      "grad_norm": 0.12353098392486572,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 1665
+    },
+    {
+      "epoch": 0.9027363858033053,
+      "grad_norm": 0.11463508754968643,
+      "learning_rate": 0.0001,
+      "loss": 1.5422,
+      "step": 1666
+    },
+    {
+      "epoch": 0.9032782443782172,
+      "grad_norm": 0.14872030913829803,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 1667
+    },
+    {
+      "epoch": 0.9038201029531292,
+      "grad_norm": 0.12321729958057404,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 1668
+    },
+    {
+      "epoch": 0.9043619615280412,
+      "grad_norm": 0.1364787518978119,
+      "learning_rate": 0.0001,
+      "loss": 1.5988,
+      "step": 1669
+    },
+    {
+      "epoch": 0.9049038201029531,
+      "grad_norm": 0.145307719707489,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 1670
+    },
+    {
+      "epoch": 0.9054456786778651,
+      "grad_norm": 0.12828241288661957,
+      "learning_rate": 0.0001,
+      "loss": 1.5977,
+      "step": 1671
+    },
+    {
+      "epoch": 0.905987537252777,
+      "grad_norm": 0.1327933818101883,
+      "learning_rate": 0.0001,
+      "loss": 1.5433,
+      "step": 1672
+    },
+    {
+      "epoch": 0.906529395827689,
+      "grad_norm": 0.11078044027090073,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 1673
+    },
+    {
+      "epoch": 0.907071254402601,
+      "grad_norm": 0.1254003942012787,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 1674
+    },
+    {
+      "epoch": 0.9076131129775129,
+      "grad_norm": 0.11752931028604507,
+      "learning_rate": 0.0001,
+      "loss": 1.589,
+      "step": 1675
+    },
+    {
+      "epoch": 0.9081549715524249,
+      "grad_norm": 0.12012698501348495,
+      "learning_rate": 0.0001,
+      "loss": 1.6072,
+      "step": 1676
+    },
+    {
+      "epoch": 0.9086968301273368,
+      "grad_norm": 0.12614655494689941,
+      "learning_rate": 0.0001,
+      "loss": 1.5748,
+      "step": 1677
+    },
+    {
+      "epoch": 0.9092386887022487,
+      "grad_norm": 0.1087908074259758,
+      "learning_rate": 0.0001,
+      "loss": 1.6286,
+      "step": 1678
+    },
+    {
+      "epoch": 0.9097805472771606,
+      "grad_norm": 0.1440020352602005,
+      "learning_rate": 0.0001,
+      "loss": 1.634,
+      "step": 1679
+    },
+    {
+      "epoch": 0.9103224058520726,
+      "grad_norm": 0.11258941143751144,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 1680
+    },
+    {
+      "epoch": 0.9108642644269845,
+      "grad_norm": 0.12051312625408173,
+      "learning_rate": 0.0001,
+      "loss": 1.5715,
+      "step": 1681
+    },
+    {
+      "epoch": 0.9114061230018965,
+      "grad_norm": 0.11230110377073288,
+      "learning_rate": 0.0001,
+      "loss": 1.5831,
+      "step": 1682
+    },
+    {
+      "epoch": 0.9119479815768085,
+      "grad_norm": 0.1212247908115387,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 1683
+    },
+    {
+      "epoch": 0.9124898401517204,
+      "grad_norm": 0.11118316650390625,
+      "learning_rate": 0.0001,
+      "loss": 1.5785,
+      "step": 1684
+    },
+    {
+      "epoch": 0.9130316987266324,
+      "grad_norm": 0.11062508076429367,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 1685
+    },
+    {
+      "epoch": 0.9135735573015443,
+      "grad_norm": 0.13190412521362305,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 1686
+    },
+    {
+      "epoch": 0.9141154158764563,
+      "grad_norm": 0.10946314036846161,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 1687
+    },
+    {
+      "epoch": 0.9146572744513682,
+      "grad_norm": 0.11208927631378174,
+      "learning_rate": 0.0001,
+      "loss": 1.5419,
+      "step": 1688
+    },
+    {
+      "epoch": 0.9151991330262801,
+      "grad_norm": 0.12088596820831299,
+      "learning_rate": 0.0001,
+      "loss": 1.5833,
+      "step": 1689
+    },
+    {
+      "epoch": 0.915740991601192,
+      "grad_norm": 0.1129712462425232,
+      "learning_rate": 0.0001,
+      "loss": 1.5547,
+      "step": 1690
+    },
+    {
+      "epoch": 0.916282850176104,
+      "grad_norm": 0.12110590934753418,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 1691
+    },
+    {
+      "epoch": 0.916824708751016,
+      "grad_norm": 0.11715506762266159,
+      "learning_rate": 0.0001,
+      "loss": 1.5834,
+      "step": 1692
+    },
+    {
+      "epoch": 0.9173665673259279,
+      "grad_norm": 0.11173558980226517,
+      "learning_rate": 0.0001,
+      "loss": 1.5223,
+      "step": 1693
+    },
+    {
+      "epoch": 0.9179084259008399,
+      "grad_norm": 0.12259289622306824,
+      "learning_rate": 0.0001,
+      "loss": 1.5992,
+      "step": 1694
+    },
+    {
+      "epoch": 0.9184502844757518,
+      "grad_norm": 0.11358395963907242,
+      "learning_rate": 0.0001,
+      "loss": 1.6259,
+      "step": 1695
+    },
+    {
+      "epoch": 0.9189921430506638,
+      "grad_norm": 0.13116388022899628,
+      "learning_rate": 0.0001,
+      "loss": 1.5542,
+      "step": 1696
+    },
+    {
+      "epoch": 0.9195340016255757,
+      "grad_norm": 0.10953519493341446,
+      "learning_rate": 0.0001,
+      "loss": 1.486,
+      "step": 1697
+    },
+    {
+      "epoch": 0.9200758602004877,
+      "grad_norm": 0.11799793690443039,
+      "learning_rate": 0.0001,
+      "loss": 1.6026,
+      "step": 1698
+    },
+    {
+      "epoch": 0.9206177187753997,
+      "grad_norm": 0.1310616433620453,
+      "learning_rate": 0.0001,
+      "loss": 1.5784,
+      "step": 1699
+    },
+    {
+      "epoch": 0.9211595773503116,
+      "grad_norm": 0.11110175400972366,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 1700
+    },
+    {
+      "epoch": 0.9217014359252235,
+      "grad_norm": 0.12602512538433075,
+      "learning_rate": 0.0001,
+      "loss": 1.5275,
+      "step": 1701
+    },
+    {
+      "epoch": 0.9222432945001354,
+      "grad_norm": 0.11771464347839355,
+      "learning_rate": 0.0001,
+      "loss": 1.6476,
+      "step": 1702
+    },
+    {
+      "epoch": 0.9227851530750474,
+      "grad_norm": 0.11211036890745163,
+      "learning_rate": 0.0001,
+      "loss": 1.5361,
+      "step": 1703
+    },
+    {
+      "epoch": 0.9233270116499593,
+      "grad_norm": 0.1293649971485138,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 1704
+    },
+    {
+      "epoch": 0.9238688702248713,
+      "grad_norm": 0.11830790340900421,
+      "learning_rate": 0.0001,
+      "loss": 1.6019,
+      "step": 1705
+    },
+    {
+      "epoch": 0.9244107287997833,
+      "grad_norm": 0.11535318195819855,
+      "learning_rate": 0.0001,
+      "loss": 1.5832,
+      "step": 1706
+    },
+    {
+      "epoch": 0.9249525873746952,
+      "grad_norm": 0.11281119287014008,
+      "learning_rate": 0.0001,
+      "loss": 1.5781,
+      "step": 1707
+    },
+    {
+      "epoch": 0.9254944459496072,
+      "grad_norm": 0.11184824258089066,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 1708
+    },
+    {
+      "epoch": 0.9260363045245191,
+      "grad_norm": 0.11843252182006836,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 1709
+    },
+    {
+      "epoch": 0.9265781630994311,
+      "grad_norm": 0.11637989431619644,
+      "learning_rate": 0.0001,
+      "loss": 1.5784,
+      "step": 1710
+    },
+    {
+      "epoch": 0.927120021674343,
+      "grad_norm": 0.11847338825464249,
+      "learning_rate": 0.0001,
+      "loss": 1.6241,
+      "step": 1711
+    },
+    {
+      "epoch": 0.927661880249255,
+      "grad_norm": 0.1154191642999649,
+      "learning_rate": 0.0001,
+      "loss": 1.5312,
+      "step": 1712
+    },
+    {
+      "epoch": 0.9282037388241668,
+      "grad_norm": 0.11177432537078857,
+      "learning_rate": 0.0001,
+      "loss": 1.5722,
+      "step": 1713
+    },
+    {
+      "epoch": 0.9287455973990788,
+      "grad_norm": 0.11820822209119797,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 1714
+    },
+    {
+      "epoch": 0.9292874559739908,
+      "grad_norm": 0.1316254734992981,
+      "learning_rate": 0.0001,
+      "loss": 1.6296,
+      "step": 1715
+    },
+    {
+      "epoch": 0.9298293145489027,
+      "grad_norm": 0.12370502203702927,
+      "learning_rate": 0.0001,
+      "loss": 1.6166,
+      "step": 1716
+    },
+    {
+      "epoch": 0.9303711731238147,
+      "grad_norm": 0.13926295936107635,
+      "learning_rate": 0.0001,
+      "loss": 1.5673,
+      "step": 1717
+    },
+    {
+      "epoch": 0.9309130316987266,
+      "grad_norm": 0.11580811440944672,
+      "learning_rate": 0.0001,
+      "loss": 1.5704,
+      "step": 1718
+    },
+    {
+      "epoch": 0.9314548902736386,
+      "grad_norm": 0.14968527853488922,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 1719
+    },
+    {
+      "epoch": 0.9319967488485505,
+      "grad_norm": 0.11721023172140121,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 1720
+    },
+    {
+      "epoch": 0.9325386074234625,
+      "grad_norm": 0.13286975026130676,
+      "learning_rate": 0.0001,
+      "loss": 1.5621,
+      "step": 1721
+    },
+    {
+      "epoch": 0.9330804659983745,
+      "grad_norm": 0.1223980113863945,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 1722
+    },
+    {
+      "epoch": 0.9336223245732864,
+      "grad_norm": 0.13075362145900726,
+      "learning_rate": 0.0001,
+      "loss": 1.5676,
+      "step": 1723
+    },
+    {
+      "epoch": 0.9341641831481983,
+      "grad_norm": 0.13382086157798767,
+      "learning_rate": 0.0001,
+      "loss": 1.5355,
+      "step": 1724
+    },
+    {
+      "epoch": 0.9347060417231102,
+      "grad_norm": 0.11884968727827072,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 1725
+    },
+    {
+      "epoch": 0.9352479002980222,
+      "grad_norm": 0.12873336672782898,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 1726
+    },
+    {
+      "epoch": 0.9357897588729341,
+      "grad_norm": 0.11676906049251556,
+      "learning_rate": 0.0001,
+      "loss": 1.62,
+      "step": 1727
+    },
+    {
+      "epoch": 0.9363316174478461,
+      "grad_norm": 0.1319124698638916,
+      "learning_rate": 0.0001,
+      "loss": 1.6633,
+      "step": 1728
+    },
+    {
+      "epoch": 0.9368734760227581,
+      "grad_norm": 0.11038797348737717,
+      "learning_rate": 0.0001,
+      "loss": 1.45,
+      "step": 1729
+    },
+    {
+      "epoch": 0.93741533459767,
+      "grad_norm": 0.12079478055238724,
+      "learning_rate": 0.0001,
+      "loss": 1.5593,
+      "step": 1730
+    },
+    {
+      "epoch": 0.937957193172582,
+      "grad_norm": 0.1391608864068985,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 1731
+    },
+    {
+      "epoch": 0.9384990517474939,
+      "grad_norm": 0.11178398132324219,
+      "learning_rate": 0.0001,
+      "loss": 1.5399,
+      "step": 1732
+    },
+    {
+      "epoch": 0.9390409103224059,
+      "grad_norm": 0.14346595108509064,
+      "learning_rate": 0.0001,
+      "loss": 1.6717,
+      "step": 1733
+    },
+    {
+      "epoch": 0.9395827688973178,
+      "grad_norm": 0.11750081181526184,
+      "learning_rate": 0.0001,
+      "loss": 1.6113,
+      "step": 1734
+    },
+    {
+      "epoch": 0.9401246274722298,
+      "grad_norm": 0.1212116926908493,
+      "learning_rate": 0.0001,
+      "loss": 1.5555,
+      "step": 1735
+    },
+    {
+      "epoch": 0.9406664860471416,
+      "grad_norm": 0.13526900112628937,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 1736
+    },
+    {
+      "epoch": 0.9412083446220536,
+      "grad_norm": 0.11274862289428711,
+      "learning_rate": 0.0001,
+      "loss": 1.5397,
+      "step": 1737
+    },
+    {
+      "epoch": 0.9417502031969656,
+      "grad_norm": 0.1345253586769104,
+      "learning_rate": 0.0001,
+      "loss": 1.5766,
+      "step": 1738
+    },
+    {
+      "epoch": 0.9422920617718775,
+      "grad_norm": 0.1218753382563591,
+      "learning_rate": 0.0001,
+      "loss": 1.5997,
+      "step": 1739
+    },
+    {
+      "epoch": 0.9428339203467895,
+      "grad_norm": 0.1240580603480339,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 1740
+    },
+    {
+      "epoch": 0.9433757789217014,
+      "grad_norm": 0.13162104785442352,
+      "learning_rate": 0.0001,
+      "loss": 1.5921,
+      "step": 1741
+    },
+    {
+      "epoch": 0.9439176374966134,
+      "grad_norm": 0.1223171278834343,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 1742
+    },
+    {
+      "epoch": 0.9444594960715254,
+      "grad_norm": 0.1349639743566513,
+      "learning_rate": 0.0001,
+      "loss": 1.6118,
+      "step": 1743
+    },
+    {
+      "epoch": 0.9450013546464373,
+      "grad_norm": 0.12058819830417633,
+      "learning_rate": 0.0001,
+      "loss": 1.6638,
+      "step": 1744
+    },
+    {
+      "epoch": 0.9455432132213493,
+      "grad_norm": 0.12474081665277481,
+      "learning_rate": 0.0001,
+      "loss": 1.5619,
+      "step": 1745
+    },
+    {
+      "epoch": 0.9460850717962612,
+      "grad_norm": 0.1322060078382492,
+      "learning_rate": 0.0001,
+      "loss": 1.5174,
+      "step": 1746
+    },
+    {
+      "epoch": 0.9466269303711731,
+      "grad_norm": 0.11868065595626831,
+      "learning_rate": 0.0001,
+      "loss": 1.5275,
+      "step": 1747
+    },
+    {
+      "epoch": 0.947168788946085,
+      "grad_norm": 0.13948048651218414,
+      "learning_rate": 0.0001,
+      "loss": 1.5391,
+      "step": 1748
+    },
+    {
+      "epoch": 0.947710647520997,
+      "grad_norm": 0.11156269907951355,
+      "learning_rate": 0.0001,
+      "loss": 1.5508,
+      "step": 1749
+    },
+    {
+      "epoch": 0.9482525060959089,
+      "grad_norm": 0.14344418048858643,
+      "learning_rate": 0.0001,
+      "loss": 1.5386,
+      "step": 1750
+    },
+    {
+      "epoch": 0.9487943646708209,
+      "grad_norm": 0.1274086982011795,
+      "learning_rate": 0.0001,
+      "loss": 1.5993,
+      "step": 1751
+    },
+    {
+      "epoch": 0.9493362232457329,
+      "grad_norm": 0.1266939342021942,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 1752
+    },
+    {
+      "epoch": 0.9498780818206448,
+      "grad_norm": 0.14160290360450745,
+      "learning_rate": 0.0001,
+      "loss": 1.568,
+      "step": 1753
+    },
+    {
+      "epoch": 0.9504199403955568,
+      "grad_norm": 0.11376998573541641,
+      "learning_rate": 0.0001,
+      "loss": 1.5564,
+      "step": 1754
+    },
+    {
+      "epoch": 0.9509617989704687,
+      "grad_norm": 0.14705996215343475,
+      "learning_rate": 0.0001,
+      "loss": 1.6263,
+      "step": 1755
+    },
+    {
+      "epoch": 0.9515036575453807,
+      "grad_norm": 0.11236948519945145,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 1756
+    },
+    {
+      "epoch": 0.9520455161202926,
+      "grad_norm": 0.12842684984207153,
+      "learning_rate": 0.0001,
+      "loss": 1.6045,
+      "step": 1757
+    },
+    {
+      "epoch": 0.9525873746952046,
+      "grad_norm": 0.12077702581882477,
+      "learning_rate": 0.0001,
+      "loss": 1.5668,
+      "step": 1758
+    },
+    {
+      "epoch": 0.9531292332701164,
+      "grad_norm": 0.11721716821193695,
+      "learning_rate": 0.0001,
+      "loss": 1.5491,
+      "step": 1759
+    },
+    {
+      "epoch": 0.9536710918450284,
+      "grad_norm": 0.12973424792289734,
+      "learning_rate": 0.0001,
+      "loss": 1.6049,
+      "step": 1760
+    },
+    {
+      "epoch": 0.9542129504199404,
+      "grad_norm": 0.11184488981962204,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 1761
+    },
+    {
+      "epoch": 0.9547548089948523,
+      "grad_norm": 0.12777452170848846,
+      "learning_rate": 0.0001,
+      "loss": 1.5778,
+      "step": 1762
+    },
+    {
+      "epoch": 0.9552966675697643,
+      "grad_norm": 0.12511278688907623,
+      "learning_rate": 0.0001,
+      "loss": 1.5786,
+      "step": 1763
+    },
+    {
+      "epoch": 0.9558385261446762,
+      "grad_norm": 0.1258372664451599,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 1764
+    },
+    {
+      "epoch": 0.9563803847195882,
+      "grad_norm": 0.1296481341123581,
+      "learning_rate": 0.0001,
+      "loss": 1.5824,
+      "step": 1765
+    },
+    {
+      "epoch": 0.9569222432945002,
+      "grad_norm": 0.11928050220012665,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 1766
+    },
+    {
+      "epoch": 0.9574641018694121,
+      "grad_norm": 0.11857929080724716,
+      "learning_rate": 0.0001,
+      "loss": 1.5373,
+      "step": 1767
+    },
+    {
+      "epoch": 0.9580059604443241,
+      "grad_norm": 0.12479015439748764,
+      "learning_rate": 0.0001,
+      "loss": 1.6207,
+      "step": 1768
+    },
+    {
+      "epoch": 0.958547819019236,
+      "grad_norm": 0.11598216742277145,
+      "learning_rate": 0.0001,
+      "loss": 1.5384,
+      "step": 1769
+    },
+    {
+      "epoch": 0.959089677594148,
+      "grad_norm": 0.11646448820829391,
+      "learning_rate": 0.0001,
+      "loss": 1.5858,
+      "step": 1770
+    },
+    {
+      "epoch": 0.9596315361690598,
+      "grad_norm": 0.12080811709165573,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 1771
+    },
+    {
+      "epoch": 0.9601733947439718,
+      "grad_norm": 0.11604882776737213,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 1772
+    },
+    {
+      "epoch": 0.9607152533188837,
+      "grad_norm": 0.1229996308684349,
+      "learning_rate": 0.0001,
+      "loss": 1.5694,
+      "step": 1773
+    },
+    {
+      "epoch": 0.9612571118937957,
+      "grad_norm": 0.13511215150356293,
+      "learning_rate": 0.0001,
+      "loss": 1.6532,
+      "step": 1774
+    },
+    {
+      "epoch": 0.9617989704687077,
+      "grad_norm": 0.12474493682384491,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 1775
+    },
+    {
+      "epoch": 0.9623408290436196,
+      "grad_norm": 0.11818260699510574,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 1776
+    },
+    {
+      "epoch": 0.9628826876185316,
+      "grad_norm": 0.11510884016752243,
+      "learning_rate": 0.0001,
+      "loss": 1.5688,
+      "step": 1777
+    },
+    {
+      "epoch": 0.9634245461934435,
+      "grad_norm": 0.11183018237352371,
+      "learning_rate": 0.0001,
+      "loss": 1.585,
+      "step": 1778
+    },
+    {
+      "epoch": 0.9639664047683555,
+      "grad_norm": 0.12041230499744415,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 1779
+    },
+    {
+      "epoch": 0.9645082633432674,
+      "grad_norm": 0.11413277685642242,
+      "learning_rate": 0.0001,
+      "loss": 1.6082,
+      "step": 1780
+    },
+    {
+      "epoch": 0.9650501219181794,
+      "grad_norm": 0.11770571023225784,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 1781
+    },
+    {
+      "epoch": 0.9655919804930913,
+      "grad_norm": 0.11431930214166641,
+      "learning_rate": 0.0001,
+      "loss": 1.5738,
+      "step": 1782
+    },
+    {
+      "epoch": 0.9661338390680032,
+      "grad_norm": 0.12779121100902557,
+      "learning_rate": 0.0001,
+      "loss": 1.5681,
+      "step": 1783
+    },
+    {
+      "epoch": 0.9666756976429152,
+      "grad_norm": 0.1112782433629036,
+      "learning_rate": 0.0001,
+      "loss": 1.5436,
+      "step": 1784
+    },
+    {
+      "epoch": 0.9672175562178271,
+      "grad_norm": 0.11836463958024979,
+      "learning_rate": 0.0001,
+      "loss": 1.5875,
+      "step": 1785
+    },
+    {
+      "epoch": 0.9677594147927391,
+      "grad_norm": 0.11033179610967636,
+      "learning_rate": 0.0001,
+      "loss": 1.5581,
+      "step": 1786
+    },
+    {
+      "epoch": 0.968301273367651,
+      "grad_norm": 0.12732066214084625,
+      "learning_rate": 0.0001,
+      "loss": 1.5556,
+      "step": 1787
+    },
+    {
+      "epoch": 0.968843131942563,
+      "grad_norm": 0.11075282841920853,
+      "learning_rate": 0.0001,
+      "loss": 1.5906,
+      "step": 1788
+    },
+    {
+      "epoch": 0.969384990517475,
+      "grad_norm": 0.11174791306257248,
+      "learning_rate": 0.0001,
+      "loss": 1.4853,
+      "step": 1789
+    },
+    {
+      "epoch": 0.9699268490923869,
+      "grad_norm": 0.10969841480255127,
+      "learning_rate": 0.0001,
+      "loss": 1.6018,
+      "step": 1790
+    },
+    {
+      "epoch": 0.9704687076672989,
+      "grad_norm": 0.12162282317876816,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 1791
+    },
+    {
+      "epoch": 0.9710105662422108,
+      "grad_norm": 0.11701533943414688,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 1792
+    },
+    {
+      "epoch": 0.9715524248171228,
+      "grad_norm": 0.12743960320949554,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 1793
+    },
+    {
+      "epoch": 0.9720942833920346,
+      "grad_norm": 0.1182185560464859,
+      "learning_rate": 0.0001,
+      "loss": 1.5864,
+      "step": 1794
+    },
+    {
+      "epoch": 0.9726361419669466,
+      "grad_norm": 0.12218203395605087,
+      "learning_rate": 0.0001,
+      "loss": 1.6351,
+      "step": 1795
+    },
+    {
+      "epoch": 0.9731780005418585,
+      "grad_norm": 0.11644001305103302,
+      "learning_rate": 0.0001,
+      "loss": 1.5516,
+      "step": 1796
+    },
+    {
+      "epoch": 0.9737198591167705,
+      "grad_norm": 0.1234685555100441,
+      "learning_rate": 0.0001,
+      "loss": 1.631,
+      "step": 1797
+    },
+    {
+      "epoch": 0.9742617176916825,
+      "grad_norm": 0.11179257184267044,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 1798
+    },
+    {
+      "epoch": 0.9748035762665944,
+      "grad_norm": 0.11813155561685562,
+      "learning_rate": 0.0001,
+      "loss": 1.6825,
+      "step": 1799
+    },
+    {
+      "epoch": 0.9753454348415064,
+      "grad_norm": 0.11452137678861618,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 1800
+    },
+    {
+      "epoch": 0.9758872934164183,
+      "grad_norm": 0.1203099936246872,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 1801
+    },
+    {
+      "epoch": 0.9764291519913303,
+      "grad_norm": 0.11747494339942932,
+      "learning_rate": 0.0001,
+      "loss": 1.5647,
+      "step": 1802
+    },
+    {
+      "epoch": 0.9769710105662422,
+      "grad_norm": 0.11051321774721146,
+      "learning_rate": 0.0001,
+      "loss": 1.5415,
+      "step": 1803
+    },
+    {
+      "epoch": 0.9775128691411542,
+      "grad_norm": 0.11837146431207657,
+      "learning_rate": 0.0001,
+      "loss": 1.5948,
+      "step": 1804
+    },
+    {
+      "epoch": 0.978054727716066,
+      "grad_norm": 0.11796199530363083,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 1805
+    },
+    {
+      "epoch": 0.978596586290978,
+      "grad_norm": 0.11934317648410797,
+      "learning_rate": 0.0001,
+      "loss": 1.5606,
+      "step": 1806
+    },
+    {
+      "epoch": 0.97913844486589,
+      "grad_norm": 0.11518701165914536,
+      "learning_rate": 0.0001,
+      "loss": 1.5921,
+      "step": 1807
+    },
+    {
+      "epoch": 0.9796803034408019,
+      "grad_norm": 0.12122237682342529,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 1808
+    },
+    {
+      "epoch": 0.9802221620157139,
+      "grad_norm": 0.1308448165655136,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 1809
+    },
+    {
+      "epoch": 0.9807640205906258,
+      "grad_norm": 0.12254752963781357,
+      "learning_rate": 0.0001,
+      "loss": 1.6409,
+      "step": 1810
+    },
+    {
+      "epoch": 0.9813058791655378,
+      "grad_norm": 0.12072083353996277,
+      "learning_rate": 0.0001,
+      "loss": 1.5395,
+      "step": 1811
+    },
+    {
+      "epoch": 0.9818477377404498,
+      "grad_norm": 0.12756118178367615,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 1812
+    },
+    {
+      "epoch": 0.9823895963153617,
+      "grad_norm": 0.11563752591609955,
+      "learning_rate": 0.0001,
+      "loss": 1.6181,
+      "step": 1813
+    },
+    {
+      "epoch": 0.9829314548902737,
+      "grad_norm": 0.12231991440057755,
+      "learning_rate": 0.0001,
+      "loss": 1.6162,
+      "step": 1814
+    },
+    {
+      "epoch": 0.9834733134651856,
+      "grad_norm": 0.11458272486925125,
+      "learning_rate": 0.0001,
+      "loss": 1.5106,
+      "step": 1815
+    },
+    {
+      "epoch": 0.9840151720400976,
+      "grad_norm": 0.12017235159873962,
+      "learning_rate": 0.0001,
+      "loss": 1.5772,
+      "step": 1816
+    },
+    {
+      "epoch": 0.9845570306150094,
+      "grad_norm": 0.12292500585317612,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 1817
+    },
+    {
+      "epoch": 0.9850988891899214,
+      "grad_norm": 0.11395881325006485,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 1818
+    },
+    {
+      "epoch": 0.9856407477648333,
+      "grad_norm": 0.11829330772161484,
+      "learning_rate": 0.0001,
+      "loss": 1.5448,
+      "step": 1819
+    },
+    {
+      "epoch": 0.9861826063397453,
+      "grad_norm": 0.12044976651668549,
+      "learning_rate": 0.0001,
+      "loss": 1.5628,
+      "step": 1820
+    },
+    {
+      "epoch": 0.9867244649146573,
+      "grad_norm": 0.13126184046268463,
+      "learning_rate": 0.0001,
+      "loss": 1.5905,
+      "step": 1821
+    },
+    {
+      "epoch": 0.9872663234895692,
+      "grad_norm": 0.11481881886720657,
+      "learning_rate": 0.0001,
+      "loss": 1.5681,
+      "step": 1822
+    },
+    {
+      "epoch": 0.9878081820644812,
+      "grad_norm": 0.11921236664056778,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 1823
+    },
+    {
+      "epoch": 0.9883500406393931,
+      "grad_norm": 0.13419100642204285,
+      "learning_rate": 0.0001,
+      "loss": 1.6291,
+      "step": 1824
+    },
+    {
+      "epoch": 0.9888918992143051,
+      "grad_norm": 0.11893222481012344,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 1825
+    },
+    {
+      "epoch": 0.989433757789217,
+      "grad_norm": 0.12331171333789825,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 1826
+    },
+    {
+      "epoch": 0.989975616364129,
+      "grad_norm": 0.12092404067516327,
+      "learning_rate": 0.0001,
+      "loss": 1.6025,
+      "step": 1827
+    },
+    {
+      "epoch": 0.990517474939041,
+      "grad_norm": 0.13040265440940857,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 1828
+    },
+    {
+      "epoch": 0.9910593335139528,
+      "grad_norm": 0.11708499491214752,
+      "learning_rate": 0.0001,
+      "loss": 1.558,
+      "step": 1829
+    },
+    {
+      "epoch": 0.9916011920888648,
+      "grad_norm": 0.11232500523328781,
+      "learning_rate": 0.0001,
+      "loss": 1.4854,
+      "step": 1830
+    },
+    {
+      "epoch": 0.9921430506637767,
+      "grad_norm": 0.12061132490634918,
+      "learning_rate": 0.0001,
+      "loss": 1.5493,
+      "step": 1831
+    },
+    {
+      "epoch": 0.9926849092386887,
+      "grad_norm": 0.11637751758098602,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 1832
+    },
+    {
+      "epoch": 0.9932267678136006,
+      "grad_norm": 0.12868823111057281,
+      "learning_rate": 0.0001,
+      "loss": 1.6286,
+      "step": 1833
+    },
+    {
+      "epoch": 0.9937686263885126,
+      "grad_norm": 0.11612503975629807,
+      "learning_rate": 0.0001,
+      "loss": 1.5745,
+      "step": 1834
+    },
+    {
+      "epoch": 0.9943104849634246,
+      "grad_norm": 0.11189901828765869,
+      "learning_rate": 0.0001,
+      "loss": 1.5413,
+      "step": 1835
+    },
+    {
+      "epoch": 0.9948523435383365,
+      "grad_norm": 0.11703436076641083,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 1836
+    },
+    {
+      "epoch": 0.9953942021132485,
+      "grad_norm": 0.12222783267498016,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 1837
+    },
+    {
+      "epoch": 0.9959360606881604,
+      "grad_norm": 0.11635846644639969,
+      "learning_rate": 0.0001,
+      "loss": 1.5942,
+      "step": 1838
+    },
+    {
+      "epoch": 0.9964779192630724,
+      "grad_norm": 0.12245723605155945,
+      "learning_rate": 0.0001,
+      "loss": 1.5653,
+      "step": 1839
+    },
+    {
+      "epoch": 0.9970197778379842,
+      "grad_norm": 0.1175963282585144,
+      "learning_rate": 0.0001,
+      "loss": 1.6186,
+      "step": 1840
+    },
+    {
+      "epoch": 0.9975616364128962,
+      "grad_norm": 0.13284112513065338,
+      "learning_rate": 0.0001,
+      "loss": 1.5772,
+      "step": 1841
+    },
+    {
+      "epoch": 0.9981034949878081,
+      "grad_norm": 0.11542721837759018,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 1842
+    },
+    {
+      "epoch": 0.9986453535627201,
+      "grad_norm": 0.11047124862670898,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 1843
+    },
+    {
+      "epoch": 0.9991872121376321,
+      "grad_norm": 0.11980320513248444,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 1844
+    },
+    {
+      "epoch": 0.999729070712544,
+      "grad_norm": 0.11151415854692459,
+      "learning_rate": 0.0001,
+      "loss": 1.5566,
+      "step": 1845
+    },
+    {
+      "epoch": 0.999729070712544,
+      "step": 1845,
+      "total_flos": 2.0947907107325215e+19,
+      "train_loss": 1.6380996824926155,
+      "train_runtime": 59531.8826,
+      "train_samples_per_second": 0.496,
+      "train_steps_per_second": 0.031
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1845,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.0947907107325215e+19,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}