diff --git "a/checkpoint-1940/trainer_state.json" "b/checkpoint-1940/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1940/trainer_state.json"
@@ -0,0 +1,13613 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998282082116475,
+  "eval_steps": 500,
+  "global_step": 1940,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005153753650575503,
+      "grad_norm": 7.126006126403809,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 2.5468,
+      "step": 1
+    },
+    {
+      "epoch": 0.0010307507301151005,
+      "grad_norm": 6.614906311035156,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 2.4946,
+      "step": 2
+    },
+    {
+      "epoch": 0.0015461260951726507,
+      "grad_norm": 6.964903831481934,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 2.5099,
+      "step": 3
+    },
+    {
+      "epoch": 0.002061501460230201,
+      "grad_norm": 6.724514484405518,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 2.4418,
+      "step": 4
+    },
+    {
+      "epoch": 0.0025768768252877514,
+      "grad_norm": 7.039041519165039,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 2.4028,
+      "step": 5
+    },
+    {
+      "epoch": 0.0030922521903453013,
+      "grad_norm": 7.229781627655029,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 2.4623,
+      "step": 6
+    },
+    {
+      "epoch": 0.0036076275554028517,
+      "grad_norm": 6.759846210479736,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 2.466,
+      "step": 7
+    },
+    {
+      "epoch": 0.004123002920460402,
+      "grad_norm": 6.571556091308594,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 2.3653,
+      "step": 8
+    },
+    {
+      "epoch": 0.0046383782855179525,
+      "grad_norm": 6.8509840965271,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 2.5106,
+      "step": 9
+    },
+    {
+      "epoch": 0.005153753650575503,
+      "grad_norm": 6.830150604248047,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 2.5043,
+      "step": 10
+    },
+    {
+      "epoch": 0.005669129015633052,
+      "grad_norm": 6.322958469390869,
+      "learning_rate": 5.5e-07,
+      "loss": 2.4498,
+      "step": 11
+    },
+    {
+      "epoch": 0.006184504380690603,
+      "grad_norm": 6.736442565917969,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 2.5081,
+      "step": 12
+    },
+    {
+      "epoch": 0.006699879745748153,
+      "grad_norm": 7.156307697296143,
+      "learning_rate": 6.5e-07,
+      "loss": 2.4825,
+      "step": 13
+    },
+    {
+      "epoch": 0.0072152551108057034,
+      "grad_norm": 7.225257396697998,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 2.5138,
+      "step": 14
+    },
+    {
+      "epoch": 0.007730630475863254,
+      "grad_norm": 7.1376261711120605,
+      "learning_rate": 7.5e-07,
+      "loss": 2.5128,
+      "step": 15
+    },
+    {
+      "epoch": 0.008246005840920804,
+      "grad_norm": 6.821260929107666,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 2.4578,
+      "step": 16
+    },
+    {
+      "epoch": 0.008761381205978355,
+      "grad_norm": 6.766695976257324,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 2.3446,
+      "step": 17
+    },
+    {
+      "epoch": 0.009276756571035905,
+      "grad_norm": 6.862898349761963,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 2.3988,
+      "step": 18
+    },
+    {
+      "epoch": 0.009792131936093455,
+      "grad_norm": 6.1224517822265625,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 2.1776,
+      "step": 19
+    },
+    {
+      "epoch": 0.010307507301151006,
+      "grad_norm": 6.202511787414551,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 2.2496,
+      "step": 20
+    },
+    {
+      "epoch": 0.010822882666208556,
+      "grad_norm": 6.215514659881592,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 2.2555,
+      "step": 21
+    },
+    {
+      "epoch": 0.011338258031266105,
+      "grad_norm": 5.5256524085998535,
+      "learning_rate": 1.1e-06,
+      "loss": 2.1216,
+      "step": 22
+    },
+    {
+      "epoch": 0.011853633396323655,
+      "grad_norm": 5.103215217590332,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 2.1142,
+      "step": 23
+    },
+    {
+      "epoch": 0.012369008761381205,
+      "grad_norm": 4.585116863250732,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 2.1433,
+      "step": 24
+    },
+    {
+      "epoch": 0.012884384126438756,
+      "grad_norm": 4.1703972816467285,
+      "learning_rate": 1.25e-06,
+      "loss": 2.1236,
+      "step": 25
+    },
+    {
+      "epoch": 0.013399759491496306,
+      "grad_norm": 4.232865810394287,
+      "learning_rate": 1.3e-06,
+      "loss": 2.1689,
+      "step": 26
+    },
+    {
+      "epoch": 0.013915134856553856,
+      "grad_norm": 3.503655433654785,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 2.0412,
+      "step": 27
+    },
+    {
+      "epoch": 0.014430510221611407,
+      "grad_norm": 3.439694404602051,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 2.0577,
+      "step": 28
+    },
+    {
+      "epoch": 0.014945885586668957,
+      "grad_norm": 3.234438419342041,
+      "learning_rate": 1.45e-06,
+      "loss": 1.9973,
+      "step": 29
+    },
+    {
+      "epoch": 0.015461260951726508,
+      "grad_norm": 3.018153667449951,
+      "learning_rate": 1.5e-06,
+      "loss": 1.9895,
+      "step": 30
+    },
+    {
+      "epoch": 0.015976636316784056,
+      "grad_norm": 2.669713258743286,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 1.9924,
+      "step": 31
+    },
+    {
+      "epoch": 0.01649201168184161,
+      "grad_norm": 2.5799427032470703,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.9199,
+      "step": 32
+    },
+    {
+      "epoch": 0.017007387046899157,
+      "grad_norm": 2.6986613273620605,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 1.9655,
+      "step": 33
+    },
+    {
+      "epoch": 0.01752276241195671,
+      "grad_norm": 2.537860870361328,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 1.9494,
+      "step": 34
+    },
+    {
+      "epoch": 0.018038137777014258,
+      "grad_norm": 2.5696487426757812,
+      "learning_rate": 1.75e-06,
+      "loss": 1.8842,
+      "step": 35
+    },
+    {
+      "epoch": 0.01855351314207181,
+      "grad_norm": 2.4279472827911377,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 1.8659,
+      "step": 36
+    },
+    {
+      "epoch": 0.01906888850712936,
+      "grad_norm": 2.5312445163726807,
+      "learning_rate": 1.85e-06,
+      "loss": 1.8631,
+      "step": 37
+    },
+    {
+      "epoch": 0.01958426387218691,
+      "grad_norm": 2.5473766326904297,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 1.8355,
+      "step": 38
+    },
+    {
+      "epoch": 0.02009963923724446,
+      "grad_norm": 2.475839138031006,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 1.8697,
+      "step": 39
+    },
+    {
+      "epoch": 0.02061501460230201,
+      "grad_norm": 2.154038190841675,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.7651,
+      "step": 40
+    },
+    {
+      "epoch": 0.02113038996735956,
+      "grad_norm": 2.04150390625,
+      "learning_rate": 2.05e-06,
+      "loss": 1.7075,
+      "step": 41
+    },
+    {
+      "epoch": 0.021645765332417112,
+      "grad_norm": 2.1084539890289307,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 1.7563,
+      "step": 42
+    },
+    {
+      "epoch": 0.02216114069747466,
+      "grad_norm": 2.042217493057251,
+      "learning_rate": 2.15e-06,
+      "loss": 1.6482,
+      "step": 43
+    },
+    {
+      "epoch": 0.02267651606253221,
+      "grad_norm": 1.9986730813980103,
+      "learning_rate": 2.2e-06,
+      "loss": 1.5513,
+      "step": 44
+    },
+    {
+      "epoch": 0.02319189142758976,
+      "grad_norm": 2.1617605686187744,
+      "learning_rate": 2.25e-06,
+      "loss": 1.6998,
+      "step": 45
+    },
+    {
+      "epoch": 0.02370726679264731,
+      "grad_norm": 2.08594012260437,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 1.5833,
+      "step": 46
+    },
+    {
+      "epoch": 0.024222642157704862,
+      "grad_norm": 2.058086633682251,
+      "learning_rate": 2.35e-06,
+      "loss": 1.593,
+      "step": 47
+    },
+    {
+      "epoch": 0.02473801752276241,
+      "grad_norm": 1.754408597946167,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.5644,
+      "step": 48
+    },
+    {
+      "epoch": 0.025253392887819963,
+      "grad_norm": 1.5493403673171997,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 1.5755,
+      "step": 49
+    },
+    {
+      "epoch": 0.02576876825287751,
+      "grad_norm": 1.4086428880691528,
+      "learning_rate": 2.5e-06,
+      "loss": 1.4707,
+      "step": 50
+    },
+    {
+      "epoch": 0.026284143617935064,
+      "grad_norm": 1.4488685131072998,
+      "learning_rate": 2.55e-06,
+      "loss": 1.4902,
+      "step": 51
+    },
+    {
+      "epoch": 0.026799518982992612,
+      "grad_norm": 1.5926673412322998,
+      "learning_rate": 2.6e-06,
+      "loss": 1.4902,
+      "step": 52
+    },
+    {
+      "epoch": 0.027314894348050164,
+      "grad_norm": 1.7660025358200073,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 1.5892,
+      "step": 53
+    },
+    {
+      "epoch": 0.027830269713107713,
+      "grad_norm": 1.5888456106185913,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 1.4604,
+      "step": 54
+    },
+    {
+      "epoch": 0.028345645078165265,
+      "grad_norm": 1.3883376121520996,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 1.4889,
+      "step": 55
+    },
+    {
+      "epoch": 0.028861020443222814,
+      "grad_norm": 1.291335105895996,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 1.4483,
+      "step": 56
+    },
+    {
+      "epoch": 0.029376395808280366,
+      "grad_norm": 1.3788918256759644,
+      "learning_rate": 2.85e-06,
+      "loss": 1.433,
+      "step": 57
+    },
+    {
+      "epoch": 0.029891771173337914,
+      "grad_norm": 1.839239478111267,
+      "learning_rate": 2.9e-06,
+      "loss": 1.5052,
+      "step": 58
+    },
+    {
+      "epoch": 0.030407146538395463,
+      "grad_norm": 1.4417849779129028,
+      "learning_rate": 2.95e-06,
+      "loss": 1.3936,
+      "step": 59
+    },
+    {
+      "epoch": 0.030922521903453015,
+      "grad_norm": 1.3167752027511597,
+      "learning_rate": 3e-06,
+      "loss": 1.4515,
+      "step": 60
+    },
+    {
+      "epoch": 0.03143789726851057,
+      "grad_norm": 1.2653827667236328,
+      "learning_rate": 3.05e-06,
+      "loss": 1.4057,
+      "step": 61
+    },
+    {
+      "epoch": 0.03195327263356811,
+      "grad_norm": 1.1869252920150757,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 1.4312,
+      "step": 62
+    },
+    {
+      "epoch": 0.032468647998625665,
+      "grad_norm": 1.275302767753601,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 1.3919,
+      "step": 63
+    },
+    {
+      "epoch": 0.03298402336368322,
+      "grad_norm": 1.2022939920425415,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.352,
+      "step": 64
+    },
+    {
+      "epoch": 0.03349939872874077,
+      "grad_norm": 1.1764057874679565,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 1.3249,
+      "step": 65
+    },
+    {
+      "epoch": 0.034014774093798314,
+      "grad_norm": 1.1948983669281006,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 1.3397,
+      "step": 66
+    },
+    {
+      "epoch": 0.034530149458855866,
+      "grad_norm": 1.1970183849334717,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 1.4166,
+      "step": 67
+    },
+    {
+      "epoch": 0.03504552482391342,
+      "grad_norm": 1.1630629301071167,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 1.3197,
+      "step": 68
+    },
+    {
+      "epoch": 0.03556090018897097,
+      "grad_norm": 1.1130441427230835,
+      "learning_rate": 3.45e-06,
+      "loss": 1.4192,
+      "step": 69
+    },
+    {
+      "epoch": 0.036076275554028515,
+      "grad_norm": 1.1333914995193481,
+      "learning_rate": 3.5e-06,
+      "loss": 1.402,
+      "step": 70
+    },
+    {
+      "epoch": 0.03659165091908607,
+      "grad_norm": 0.9886772036552429,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 1.3807,
+      "step": 71
+    },
+    {
+      "epoch": 0.03710702628414362,
+      "grad_norm": 1.0266650915145874,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 1.3801,
+      "step": 72
+    },
+    {
+      "epoch": 0.037622401649201165,
+      "grad_norm": 1.1841363906860352,
+      "learning_rate": 3.65e-06,
+      "loss": 1.3611,
+      "step": 73
+    },
+    {
+      "epoch": 0.03813777701425872,
+      "grad_norm": 1.2051784992218018,
+      "learning_rate": 3.7e-06,
+      "loss": 1.3565,
+      "step": 74
+    },
+    {
+      "epoch": 0.03865315237931627,
+      "grad_norm": 1.070709228515625,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 1.3538,
+      "step": 75
+    },
+    {
+      "epoch": 0.03916852774437382,
+      "grad_norm": 0.9480461478233337,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 1.3386,
+      "step": 76
+    },
+    {
+      "epoch": 0.039683903109431366,
+      "grad_norm": 1.0606422424316406,
+      "learning_rate": 3.85e-06,
+      "loss": 1.3701,
+      "step": 77
+    },
+    {
+      "epoch": 0.04019927847448892,
+      "grad_norm": 1.029645323753357,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 1.3649,
+      "step": 78
+    },
+    {
+      "epoch": 0.04071465383954647,
+      "grad_norm": 1.0215805768966675,
+      "learning_rate": 3.95e-06,
+      "loss": 1.3358,
+      "step": 79
+    },
+    {
+      "epoch": 0.04123002920460402,
+      "grad_norm": 1.0088640451431274,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.3529,
+      "step": 80
+    },
+    {
+      "epoch": 0.04174540456966157,
+      "grad_norm": 0.9737629890441895,
+      "learning_rate": 4.05e-06,
+      "loss": 1.3922,
+      "step": 81
+    },
+    {
+      "epoch": 0.04226077993471912,
+      "grad_norm": 0.9443690776824951,
+      "learning_rate": 4.1e-06,
+      "loss": 1.3181,
+      "step": 82
+    },
+    {
+      "epoch": 0.04277615529977667,
+      "grad_norm": 0.9460474252700806,
+      "learning_rate": 4.15e-06,
+      "loss": 1.364,
+      "step": 83
+    },
+    {
+      "epoch": 0.043291530664834224,
+      "grad_norm": 0.9993011355400085,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.3664,
+      "step": 84
+    },
+    {
+      "epoch": 0.04380690602989177,
+      "grad_norm": 0.9136483073234558,
+      "learning_rate": 4.25e-06,
+      "loss": 1.4122,
+      "step": 85
+    },
+    {
+      "epoch": 0.04432228139494932,
+      "grad_norm": 1.0764967203140259,
+      "learning_rate": 4.3e-06,
+      "loss": 1.3409,
+      "step": 86
+    },
+    {
+      "epoch": 0.04483765676000687,
+      "grad_norm": 0.9557569622993469,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 1.3843,
+      "step": 87
+    },
+    {
+      "epoch": 0.04535303212506442,
+      "grad_norm": 0.9669111371040344,
+      "learning_rate": 4.4e-06,
+      "loss": 1.3469,
+      "step": 88
+    },
+    {
+      "epoch": 0.04586840749012197,
+      "grad_norm": 0.9530505537986755,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 1.3295,
+      "step": 89
+    },
+    {
+      "epoch": 0.04638378285517952,
+      "grad_norm": 0.9201849102973938,
+      "learning_rate": 4.5e-06,
+      "loss": 1.3297,
+      "step": 90
+    },
+    {
+      "epoch": 0.046899158220237075,
+      "grad_norm": 0.9111409187316895,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 1.3776,
+      "step": 91
+    },
+    {
+      "epoch": 0.04741453358529462,
+      "grad_norm": 1.0236597061157227,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 1.3248,
+      "step": 92
+    },
+    {
+      "epoch": 0.04792990895035217,
+      "grad_norm": 1.0763428211212158,
+      "learning_rate": 4.65e-06,
+      "loss": 1.3199,
+      "step": 93
+    },
+    {
+      "epoch": 0.048445284315409724,
+      "grad_norm": 1.1314724683761597,
+      "learning_rate": 4.7e-06,
+      "loss": 1.3956,
+      "step": 94
+    },
+    {
+      "epoch": 0.048960659680467276,
+      "grad_norm": 0.9608203768730164,
+      "learning_rate": 4.75e-06,
+      "loss": 1.3685,
+      "step": 95
+    },
+    {
+      "epoch": 0.04947603504552482,
+      "grad_norm": 0.9246897101402283,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.2863,
+      "step": 96
+    },
+    {
+      "epoch": 0.049991410410582374,
+      "grad_norm": 0.870741605758667,
+      "learning_rate": 4.85e-06,
+      "loss": 1.2298,
+      "step": 97
+    },
+    {
+      "epoch": 0.050506785775639926,
+      "grad_norm": 0.9093057513237,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 1.2509,
+      "step": 98
+    },
+    {
+      "epoch": 0.05102216114069748,
+      "grad_norm": 0.9211610555648804,
+      "learning_rate": 4.95e-06,
+      "loss": 1.3473,
+      "step": 99
+    },
+    {
+      "epoch": 0.05153753650575502,
+      "grad_norm": 0.9618102312088013,
+      "learning_rate": 5e-06,
+      "loss": 1.32,
+      "step": 100
+    },
+    {
+      "epoch": 0.052052911870812575,
+      "grad_norm": 0.9188870787620544,
+      "learning_rate": 4.9999999073600874e-06,
+      "loss": 1.2792,
+      "step": 101
+    },
+    {
+      "epoch": 0.05256828723587013,
+      "grad_norm": 0.9936563372612,
+      "learning_rate": 4.999999629440356e-06,
+      "loss": 1.285,
+      "step": 102
+    },
+    {
+      "epoch": 0.05308366260092767,
+      "grad_norm": 0.9464974999427795,
+      "learning_rate": 4.999999166240825e-06,
+      "loss": 1.3415,
+      "step": 103
+    },
+    {
+      "epoch": 0.053599037965985225,
+      "grad_norm": 0.9253501296043396,
+      "learning_rate": 4.999998517761531e-06,
+      "loss": 1.2913,
+      "step": 104
+    },
+    {
+      "epoch": 0.05411441333104278,
+      "grad_norm": 1.006068468093872,
+      "learning_rate": 4.999997684002521e-06,
+      "loss": 1.3463,
+      "step": 105
+    },
+    {
+      "epoch": 0.05462978869610033,
+      "grad_norm": 0.962740957736969,
+      "learning_rate": 4.999996664963857e-06,
+      "loss": 1.3093,
+      "step": 106
+    },
+    {
+      "epoch": 0.055145164061157874,
+      "grad_norm": 0.9607085585594177,
+      "learning_rate": 4.999995460645613e-06,
+      "loss": 1.2853,
+      "step": 107
+    },
+    {
+      "epoch": 0.055660539426215426,
+      "grad_norm": 0.9489694833755493,
+      "learning_rate": 4.999994071047881e-06,
+      "loss": 1.2854,
+      "step": 108
+    },
+    {
+      "epoch": 0.05617591479127298,
+      "grad_norm": 1.1459851264953613,
+      "learning_rate": 4.999992496170762e-06,
+      "loss": 1.2925,
+      "step": 109
+    },
+    {
+      "epoch": 0.05669129015633053,
+      "grad_norm": 0.9199038743972778,
+      "learning_rate": 4.999990736014374e-06,
+      "loss": 1.3164,
+      "step": 110
+    },
+    {
+      "epoch": 0.057206665521388075,
+      "grad_norm": 0.9531081914901733,
+      "learning_rate": 4.999988790578847e-06,
+      "loss": 1.3016,
+      "step": 111
+    },
+    {
+      "epoch": 0.05772204088644563,
+      "grad_norm": 0.9910569190979004,
+      "learning_rate": 4.999986659864325e-06,
+      "loss": 1.2889,
+      "step": 112
+    },
+    {
+      "epoch": 0.05823741625150318,
+      "grad_norm": 0.9495322704315186,
+      "learning_rate": 4.999984343870964e-06,
+      "loss": 1.2881,
+      "step": 113
+    },
+    {
+      "epoch": 0.05875279161656073,
+      "grad_norm": 0.9481725096702576,
+      "learning_rate": 4.999981842598939e-06,
+      "loss": 1.309,
+      "step": 114
+    },
+    {
+      "epoch": 0.05926816698161828,
+      "grad_norm": 0.9376801252365112,
+      "learning_rate": 4.999979156048433e-06,
+      "loss": 1.3022,
+      "step": 115
+    },
+    {
+      "epoch": 0.05978354234667583,
+      "grad_norm": 1.0330339670181274,
+      "learning_rate": 4.999976284219646e-06,
+      "loss": 1.3911,
+      "step": 116
+    },
+    {
+      "epoch": 0.06029891771173338,
+      "grad_norm": 1.1374934911727905,
+      "learning_rate": 4.999973227112792e-06,
+      "loss": 1.2454,
+      "step": 117
+    },
+    {
+      "epoch": 0.060814293076790926,
+      "grad_norm": 0.9146735668182373,
+      "learning_rate": 4.999969984728096e-06,
+      "loss": 1.2508,
+      "step": 118
+    },
+    {
+      "epoch": 0.06132966844184848,
+      "grad_norm": 0.9474186897277832,
+      "learning_rate": 4.9999665570657985e-06,
+      "loss": 1.2772,
+      "step": 119
+    },
+    {
+      "epoch": 0.06184504380690603,
+      "grad_norm": 1.0182085037231445,
+      "learning_rate": 4.9999629441261535e-06,
+      "loss": 1.2795,
+      "step": 120
+    },
+    {
+      "epoch": 0.06236041917196358,
+      "grad_norm": 1.0316343307495117,
+      "learning_rate": 4.999959145909429e-06,
+      "loss": 1.301,
+      "step": 121
+    },
+    {
+      "epoch": 0.06287579453702113,
+      "grad_norm": 0.9423215389251709,
+      "learning_rate": 4.999955162415906e-06,
+      "loss": 1.2808,
+      "step": 122
+    },
+    {
+      "epoch": 0.06339116990207869,
+      "grad_norm": 0.9410876631736755,
+      "learning_rate": 4.999950993645881e-06,
+      "loss": 1.2881,
+      "step": 123
+    },
+    {
+      "epoch": 0.06390654526713622,
+      "grad_norm": 0.96896892786026,
+      "learning_rate": 4.999946639599662e-06,
+      "loss": 1.2718,
+      "step": 124
+    },
+    {
+      "epoch": 0.06442192063219378,
+      "grad_norm": 1.081750512123108,
+      "learning_rate": 4.999942100277572e-06,
+      "loss": 1.2793,
+      "step": 125
+    },
+    {
+      "epoch": 0.06493729599725133,
+      "grad_norm": 0.9917174577713013,
+      "learning_rate": 4.999937375679946e-06,
+      "loss": 1.2755,
+      "step": 126
+    },
+    {
+      "epoch": 0.06545267136230888,
+      "grad_norm": 0.9916597604751587,
+      "learning_rate": 4.9999324658071365e-06,
+      "loss": 1.2558,
+      "step": 127
+    },
+    {
+      "epoch": 0.06596804672736643,
+      "grad_norm": 1.0110090970993042,
+      "learning_rate": 4.999927370659507e-06,
+      "loss": 1.3498,
+      "step": 128
+    },
+    {
+      "epoch": 0.06648342209242399,
+      "grad_norm": 0.9905858039855957,
+      "learning_rate": 4.999922090237433e-06,
+      "loss": 1.2436,
+      "step": 129
+    },
+    {
+      "epoch": 0.06699879745748154,
+      "grad_norm": 1.0680683851242065,
+      "learning_rate": 4.999916624541308e-06,
+      "loss": 1.2331,
+      "step": 130
+    },
+    {
+      "epoch": 0.06751417282253908,
+      "grad_norm": 0.9393003582954407,
+      "learning_rate": 4.9999109735715356e-06,
+      "loss": 1.2809,
+      "step": 131
+    },
+    {
+      "epoch": 0.06802954818759663,
+      "grad_norm": 1.0146749019622803,
+      "learning_rate": 4.999905137328537e-06,
+      "loss": 1.3052,
+      "step": 132
+    },
+    {
+      "epoch": 0.06854492355265418,
+      "grad_norm": 0.8913703560829163,
+      "learning_rate": 4.9998991158127404e-06,
+      "loss": 1.2852,
+      "step": 133
+    },
+    {
+      "epoch": 0.06906029891771173,
+      "grad_norm": 0.8886722326278687,
+      "learning_rate": 4.999892909024597e-06,
+      "loss": 1.2871,
+      "step": 134
+    },
+    {
+      "epoch": 0.06957567428276928,
+      "grad_norm": 0.9775382280349731,
+      "learning_rate": 4.999886516964564e-06,
+      "loss": 1.2081,
+      "step": 135
+    },
+    {
+      "epoch": 0.07009104964782684,
+      "grad_norm": 1.0126523971557617,
+      "learning_rate": 4.999879939633115e-06,
+      "loss": 1.2435,
+      "step": 136
+    },
+    {
+      "epoch": 0.07060642501288439,
+      "grad_norm": 0.9554210305213928,
+      "learning_rate": 4.999873177030739e-06,
+      "loss": 1.2399,
+      "step": 137
+    },
+    {
+      "epoch": 0.07112180037794194,
+      "grad_norm": 0.9512026906013489,
+      "learning_rate": 4.999866229157935e-06,
+      "loss": 1.1091,
+      "step": 138
+    },
+    {
+      "epoch": 0.07163717574299948,
+      "grad_norm": 0.9823938012123108,
+      "learning_rate": 4.9998590960152195e-06,
+      "loss": 1.2531,
+      "step": 139
+    },
+    {
+      "epoch": 0.07215255110805703,
+      "grad_norm": 0.998545229434967,
+      "learning_rate": 4.999851777603122e-06,
+      "loss": 1.258,
+      "step": 140
+    },
+    {
+      "epoch": 0.07266792647311458,
+      "grad_norm": 1.0376691818237305,
+      "learning_rate": 4.999844273922183e-06,
+      "loss": 1.2007,
+      "step": 141
+    },
+    {
+      "epoch": 0.07318330183817214,
+      "grad_norm": 0.9358575344085693,
+      "learning_rate": 4.999836584972959e-06,
+      "loss": 1.2966,
+      "step": 142
+    },
+    {
+      "epoch": 0.07369867720322969,
+      "grad_norm": 1.057644248008728,
+      "learning_rate": 4.99982871075602e-06,
+      "loss": 1.2255,
+      "step": 143
+    },
+    {
+      "epoch": 0.07421405256828724,
+      "grad_norm": 1.049521565437317,
+      "learning_rate": 4.999820651271951e-06,
+      "loss": 1.2814,
+      "step": 144
+    },
+    {
+      "epoch": 0.07472942793334479,
+      "grad_norm": 0.9946297407150269,
+      "learning_rate": 4.999812406521347e-06,
+      "loss": 1.259,
+      "step": 145
+    },
+    {
+      "epoch": 0.07524480329840233,
+      "grad_norm": 0.9518601894378662,
+      "learning_rate": 4.9998039765048215e-06,
+      "loss": 1.2347,
+      "step": 146
+    },
+    {
+      "epoch": 0.07576017866345988,
+      "grad_norm": 0.9874811768531799,
+      "learning_rate": 4.999795361222996e-06,
+      "loss": 1.2508,
+      "step": 147
+    },
+    {
+      "epoch": 0.07627555402851743,
+      "grad_norm": 0.9873331189155579,
+      "learning_rate": 4.999786560676512e-06,
+      "loss": 1.2307,
+      "step": 148
+    },
+    {
+      "epoch": 0.07679092939357499,
+      "grad_norm": 0.9384865760803223,
+      "learning_rate": 4.99977757486602e-06,
+      "loss": 1.2487,
+      "step": 149
+    },
+    {
+      "epoch": 0.07730630475863254,
+      "grad_norm": 1.0104529857635498,
+      "learning_rate": 4.999768403792188e-06,
+      "loss": 1.2755,
+      "step": 150
+    },
+    {
+      "epoch": 0.07782168012369009,
+      "grad_norm": 0.9755220413208008,
+      "learning_rate": 4.999759047455692e-06,
+      "loss": 1.257,
+      "step": 151
+    },
+    {
+      "epoch": 0.07833705548874764,
+      "grad_norm": 0.9542570114135742,
+      "learning_rate": 4.99974950585723e-06,
+      "loss": 1.2804,
+      "step": 152
+    },
+    {
+      "epoch": 0.0788524308538052,
+      "grad_norm": 1.006955862045288,
+      "learning_rate": 4.999739778997505e-06,
+      "loss": 1.3559,
+      "step": 153
+    },
+    {
+      "epoch": 0.07936780621886273,
+      "grad_norm": 0.9349725842475891,
+      "learning_rate": 4.99972986687724e-06,
+      "loss": 1.2523,
+      "step": 154
+    },
+    {
+      "epoch": 0.07988318158392028,
+      "grad_norm": 1.0557712316513062,
+      "learning_rate": 4.99971976949717e-06,
+      "loss": 1.2769,
+      "step": 155
+    },
+    {
+      "epoch": 0.08039855694897784,
+      "grad_norm": 0.9416980743408203,
+      "learning_rate": 4.999709486858042e-06,
+      "loss": 1.2152,
+      "step": 156
+    },
+    {
+      "epoch": 0.08091393231403539,
+      "grad_norm": 0.9875814318656921,
+      "learning_rate": 4.999699018960619e-06,
+      "loss": 1.2225,
+      "step": 157
+    },
+    {
+      "epoch": 0.08142930767909294,
+      "grad_norm": 1.0240778923034668,
+      "learning_rate": 4.999688365805676e-06,
+      "loss": 1.224,
+      "step": 158
+    },
+    {
+      "epoch": 0.0819446830441505,
+      "grad_norm": 1.0420036315917969,
+      "learning_rate": 4.999677527394003e-06,
+      "loss": 1.2509,
+      "step": 159
+    },
+    {
+      "epoch": 0.08246005840920805,
+      "grad_norm": 1.004077434539795,
+      "learning_rate": 4.999666503726404e-06,
+      "loss": 1.2131,
+      "step": 160
+    },
+    {
+      "epoch": 0.08297543377426558,
+      "grad_norm": 1.1523828506469727,
+      "learning_rate": 4.999655294803695e-06,
+      "loss": 1.2318,
+      "step": 161
+    },
+    {
+      "epoch": 0.08349080913932314,
+      "grad_norm": 0.9839702248573303,
+      "learning_rate": 4.999643900626707e-06,
+      "loss": 1.2845,
+      "step": 162
+    },
+    {
+      "epoch": 0.08400618450438069,
+      "grad_norm": 0.9726237058639526,
+      "learning_rate": 4.999632321196284e-06,
+      "loss": 1.2949,
+      "step": 163
+    },
+    {
+      "epoch": 0.08452155986943824,
+      "grad_norm": 0.994744598865509,
+      "learning_rate": 4.999620556513285e-06,
+      "loss": 1.2553,
+      "step": 164
+    },
+    {
+      "epoch": 0.08503693523449579,
+      "grad_norm": 1.0818318128585815,
+      "learning_rate": 4.999608606578581e-06,
+      "loss": 1.2687,
+      "step": 165
+    },
+    {
+      "epoch": 0.08555231059955334,
+      "grad_norm": 1.1112850904464722,
+      "learning_rate": 4.999596471393059e-06,
+      "loss": 1.3266,
+      "step": 166
+    },
+    {
+      "epoch": 0.0860676859646109,
+      "grad_norm": 0.9371221661567688,
+      "learning_rate": 4.999584150957616e-06,
+      "loss": 1.2056,
+      "step": 167
+    },
+    {
+      "epoch": 0.08658306132966845,
+      "grad_norm": 0.9769211411476135,
+      "learning_rate": 4.999571645273168e-06,
+      "loss": 1.2046,
+      "step": 168
+    },
+    {
+      "epoch": 0.08709843669472599,
+      "grad_norm": 1.0210055112838745,
+      "learning_rate": 4.99955895434064e-06,
+      "loss": 1.318,
+      "step": 169
+    },
+    {
+      "epoch": 0.08761381205978354,
+      "grad_norm": 1.0373449325561523,
+      "learning_rate": 4.9995460781609725e-06,
+      "loss": 1.2416,
+      "step": 170
+    },
+    {
+      "epoch": 0.08812918742484109,
+      "grad_norm": 0.9221280813217163,
+      "learning_rate": 4.999533016735121e-06,
+      "loss": 1.1639,
+      "step": 171
+    },
+    {
+      "epoch": 0.08864456278989864,
+      "grad_norm": 0.9534924626350403,
+      "learning_rate": 4.9995197700640515e-06,
+      "loss": 1.2645,
+      "step": 172
+    },
+    {
+      "epoch": 0.0891599381549562,
+      "grad_norm": 1.0295299291610718,
+      "learning_rate": 4.999506338148748e-06,
+      "loss": 1.2781,
+      "step": 173
+    },
+    {
+      "epoch": 0.08967531352001375,
+      "grad_norm": 0.9931419491767883,
+      "learning_rate": 4.999492720990204e-06,
+      "loss": 1.2215,
+      "step": 174
+    },
+    {
+      "epoch": 0.0901906888850713,
+      "grad_norm": 0.9851301908493042,
+      "learning_rate": 4.999478918589431e-06,
+      "loss": 1.2605,
+      "step": 175
+    },
+    {
+      "epoch": 0.09070606425012884,
+      "grad_norm": 0.9377535581588745,
+      "learning_rate": 4.999464930947449e-06,
+      "loss": 1.1951,
+      "step": 176
+    },
+    {
+      "epoch": 0.09122143961518639,
+      "grad_norm": 0.968189001083374,
+      "learning_rate": 4.999450758065297e-06,
+      "loss": 1.2141,
+      "step": 177
+    },
+    {
+      "epoch": 0.09173681498024394,
+      "grad_norm": 1.0255149602890015,
+      "learning_rate": 4.9994363999440245e-06,
+      "loss": 1.2684,
+      "step": 178
+    },
+    {
+      "epoch": 0.0922521903453015,
+      "grad_norm": 1.0423952341079712,
+      "learning_rate": 4.999421856584696e-06,
+      "loss": 1.2221,
+      "step": 179
+    },
+    {
+      "epoch": 0.09276756571035905,
+      "grad_norm": 1.1341553926467896,
+      "learning_rate": 4.999407127988388e-06,
+      "loss": 1.2467,
+      "step": 180
+    },
+    {
+      "epoch": 0.0932829410754166,
+      "grad_norm": 0.9786616563796997,
+      "learning_rate": 4.9993922141561944e-06,
+      "loss": 1.177,
+      "step": 181
+    },
+    {
+      "epoch": 0.09379831644047415,
+      "grad_norm": 1.032413363456726,
+      "learning_rate": 4.999377115089219e-06,
+      "loss": 1.2769,
+      "step": 182
+    },
+    {
+      "epoch": 0.0943136918055317,
+      "grad_norm": 1.0089246034622192,
+      "learning_rate": 4.999361830788581e-06,
+      "loss": 1.2701,
+      "step": 183
+    },
+    {
+      "epoch": 0.09482906717058924,
+      "grad_norm": 1.0086696147918701,
+      "learning_rate": 4.999346361255412e-06,
+      "loss": 1.1965,
+      "step": 184
+    },
+    {
+      "epoch": 0.09534444253564679,
+      "grad_norm": 1.0027146339416504,
+      "learning_rate": 4.999330706490861e-06,
+      "loss": 1.2585,
+      "step": 185
+    },
+    {
+      "epoch": 0.09585981790070434,
+      "grad_norm": 1.0136017799377441,
+      "learning_rate": 4.999314866496085e-06,
+      "loss": 1.2137,
+      "step": 186
+    },
+    {
+      "epoch": 0.0963751932657619,
+      "grad_norm": 1.0555775165557861,
+      "learning_rate": 4.999298841272262e-06,
+      "loss": 1.2377,
+      "step": 187
+    },
+    {
+      "epoch": 0.09689056863081945,
+      "grad_norm": 1.1131389141082764,
+      "learning_rate": 4.999282630820576e-06,
+      "loss": 1.2059,
+      "step": 188
+    },
+    {
+      "epoch": 0.097405943995877,
+      "grad_norm": 1.0880460739135742,
+      "learning_rate": 4.9992662351422305e-06,
+      "loss": 1.2721,
+      "step": 189
+    },
+    {
+      "epoch": 0.09792131936093455,
+      "grad_norm": 1.0187134742736816,
+      "learning_rate": 4.999249654238439e-06,
+      "loss": 1.3137,
+      "step": 190
+    },
+    {
+      "epoch": 0.09843669472599209,
+      "grad_norm": 1.0277626514434814,
+      "learning_rate": 4.999232888110432e-06,
+      "loss": 1.2624,
+      "step": 191
+    },
+    {
+      "epoch": 0.09895207009104964,
+      "grad_norm": 1.046287178993225,
+      "learning_rate": 4.999215936759452e-06,
+      "loss": 1.1828,
+      "step": 192
+    },
+    {
+      "epoch": 0.0994674454561072,
+      "grad_norm": 1.0114896297454834,
+      "learning_rate": 4.999198800186753e-06,
+      "loss": 1.2784,
+      "step": 193
+    },
+    {
+      "epoch": 0.09998282082116475,
+      "grad_norm": 1.0122891664505005,
+      "learning_rate": 4.999181478393608e-06,
+      "loss": 1.2461,
+      "step": 194
+    },
+    {
+      "epoch": 0.1004981961862223,
+      "grad_norm": 0.9655713438987732,
+      "learning_rate": 4.999163971381299e-06,
+      "loss": 1.1965,
+      "step": 195
+    },
+    {
+      "epoch": 0.10101357155127985,
+      "grad_norm": 1.0382893085479736,
+      "learning_rate": 4.999146279151123e-06,
+      "loss": 1.1827,
+      "step": 196
+    },
+    {
+      "epoch": 0.1015289469163374,
+      "grad_norm": 1.0155049562454224,
+      "learning_rate": 4.999128401704394e-06,
+      "loss": 1.2268,
+      "step": 197
+    },
+    {
+      "epoch": 0.10204432228139496,
+      "grad_norm": 1.0959563255310059,
+      "learning_rate": 4.9991103390424334e-06,
+      "loss": 1.2438,
+      "step": 198
+    },
+    {
+      "epoch": 0.1025596976464525,
+      "grad_norm": 1.1797666549682617,
+      "learning_rate": 4.999092091166582e-06,
+      "loss": 1.2539,
+      "step": 199
+    },
+    {
+      "epoch": 0.10307507301151005,
+      "grad_norm": 1.0557608604431152,
+      "learning_rate": 4.999073658078191e-06,
+      "loss": 1.1386,
+      "step": 200
+    },
+    {
+      "epoch": 0.1035904483765676,
+      "grad_norm": 0.9590182304382324,
+      "learning_rate": 4.9990550397786285e-06,
+      "loss": 1.1971,
+      "step": 201
+    },
+    {
+      "epoch": 0.10410582374162515,
+      "grad_norm": 1.0338568687438965,
+      "learning_rate": 4.999036236269272e-06,
+      "loss": 1.2171,
+      "step": 202
+    },
+    {
+      "epoch": 0.1046211991066827,
+      "grad_norm": 0.966436505317688,
+      "learning_rate": 4.999017247551516e-06,
+      "loss": 1.2225,
+      "step": 203
+    },
+    {
+      "epoch": 0.10513657447174025,
+      "grad_norm": 0.985903263092041,
+      "learning_rate": 4.998998073626769e-06,
+      "loss": 1.2416,
+      "step": 204
+    },
+    {
+      "epoch": 0.1056519498367978,
+      "grad_norm": 1.057100772857666,
+      "learning_rate": 4.99897871449645e-06,
+      "loss": 1.2662,
+      "step": 205
+    },
+    {
+      "epoch": 0.10616732520185534,
+      "grad_norm": 0.9916481971740723,
+      "learning_rate": 4.998959170161996e-06,
+      "loss": 1.1879,
+      "step": 206
+    },
+    {
+      "epoch": 0.1066827005669129,
+      "grad_norm": 1.0549089908599854,
+      "learning_rate": 4.9989394406248524e-06,
+      "loss": 1.2343,
+      "step": 207
+    },
+    {
+      "epoch": 0.10719807593197045,
+      "grad_norm": 1.0757921934127808,
+      "learning_rate": 4.998919525886483e-06,
+      "loss": 1.2122,
+      "step": 208
+    },
+    {
+      "epoch": 0.107713451297028,
+      "grad_norm": 1.0240768194198608,
+      "learning_rate": 4.998899425948365e-06,
+      "loss": 1.2599,
+      "step": 209
+    },
+    {
+      "epoch": 0.10822882666208555,
+      "grad_norm": 0.9299185872077942,
+      "learning_rate": 4.998879140811986e-06,
+      "loss": 1.2077,
+      "step": 210
+    },
+    {
+      "epoch": 0.1087442020271431,
+      "grad_norm": 0.9661532640457153,
+      "learning_rate": 4.998858670478851e-06,
+      "loss": 1.1937,
+      "step": 211
+    },
+    {
+      "epoch": 0.10925957739220066,
+      "grad_norm": 1.0395828485488892,
+      "learning_rate": 4.9988380149504755e-06,
+      "loss": 1.2129,
+      "step": 212
+    },
+    {
+      "epoch": 0.10977495275725821,
+      "grad_norm": 1.0316669940948486,
+      "learning_rate": 4.9988171742283916e-06,
+      "loss": 1.2469,
+      "step": 213
+    },
+    {
+      "epoch": 0.11029032812231575,
+      "grad_norm": 1.0263491868972778,
+      "learning_rate": 4.998796148314144e-06,
+      "loss": 1.2414,
+      "step": 214
+    },
+    {
+      "epoch": 0.1108057034873733,
+      "grad_norm": 0.9928942918777466,
+      "learning_rate": 4.998774937209289e-06,
+      "loss": 1.2504,
+      "step": 215
+    },
+    {
+      "epoch": 0.11132107885243085,
+      "grad_norm": 1.1089606285095215,
+      "learning_rate": 4.998753540915401e-06,
+      "loss": 1.2508,
+      "step": 216
+    },
+    {
+      "epoch": 0.1118364542174884,
+      "grad_norm": 1.0831267833709717,
+      "learning_rate": 4.998731959434064e-06,
+      "loss": 1.275,
+      "step": 217
+    },
+    {
+      "epoch": 0.11235182958254596,
+      "grad_norm": 1.089594841003418,
+      "learning_rate": 4.998710192766879e-06,
+      "loss": 1.2942,
+      "step": 218
+    },
+    {
+      "epoch": 0.11286720494760351,
+      "grad_norm": 1.04873788356781,
+      "learning_rate": 4.998688240915457e-06,
+      "loss": 1.2425,
+      "step": 219
+    },
+    {
+      "epoch": 0.11338258031266106,
+      "grad_norm": 1.0630252361297607,
+      "learning_rate": 4.998666103881427e-06,
+      "loss": 1.2628,
+      "step": 220
+    },
+    {
+      "epoch": 0.1138979556777186,
+      "grad_norm": 1.0307626724243164,
+      "learning_rate": 4.998643781666429e-06,
+      "loss": 1.2594,
+      "step": 221
+    },
+    {
+      "epoch": 0.11441333104277615,
+      "grad_norm": 0.9258455634117126,
+      "learning_rate": 4.998621274272117e-06,
+      "loss": 1.1952,
+      "step": 222
+    },
+    {
+      "epoch": 0.1149287064078337,
+      "grad_norm": 0.9914063215255737,
+      "learning_rate": 4.998598581700159e-06,
+      "loss": 1.2426,
+      "step": 223
+    },
+    {
+      "epoch": 0.11544408177289125,
+      "grad_norm": 1.0159790515899658,
+      "learning_rate": 4.998575703952237e-06,
+      "loss": 1.232,
+      "step": 224
+    },
+    {
+      "epoch": 0.11595945713794881,
+      "grad_norm": 1.0941535234451294,
+      "learning_rate": 4.998552641030046e-06,
+      "loss": 1.2329,
+      "step": 225
+    },
+    {
+      "epoch": 0.11647483250300636,
+      "grad_norm": 1.0479755401611328,
+      "learning_rate": 4.9985293929352976e-06,
+      "loss": 1.2104,
+      "step": 226
+    },
+    {
+      "epoch": 0.11699020786806391,
+      "grad_norm": 1.0361521244049072,
+      "learning_rate": 4.9985059596697115e-06,
+      "loss": 1.256,
+      "step": 227
+    },
+    {
+      "epoch": 0.11750558323312146,
+      "grad_norm": 1.0784052610397339,
+      "learning_rate": 4.998482341235026e-06,
+      "loss": 1.2458,
+      "step": 228
+    },
+    {
+      "epoch": 0.118020958598179,
+      "grad_norm": 0.9795831441879272,
+      "learning_rate": 4.9984585376329915e-06,
+      "loss": 1.2419,
+      "step": 229
+    },
+    {
+      "epoch": 0.11853633396323655,
+      "grad_norm": 1.059186577796936,
+      "learning_rate": 4.998434548865372e-06,
+      "loss": 1.1864,
+      "step": 230
+    },
+    {
+      "epoch": 0.1190517093282941,
+      "grad_norm": 1.0938690900802612,
+      "learning_rate": 4.998410374933945e-06,
+      "loss": 1.2137,
+      "step": 231
+    },
+    {
+      "epoch": 0.11956708469335166,
+      "grad_norm": 1.0159286260604858,
+      "learning_rate": 4.998386015840502e-06,
+      "loss": 1.2016,
+      "step": 232
+    },
+    {
+      "epoch": 0.12008246005840921,
+      "grad_norm": 1.1006368398666382,
+      "learning_rate": 4.99836147158685e-06,
+      "loss": 1.1593,
+      "step": 233
+    },
+    {
+      "epoch": 0.12059783542346676,
+      "grad_norm": 0.9765269756317139,
+      "learning_rate": 4.998336742174805e-06,
+      "loss": 1.186,
+      "step": 234
+    },
+    {
+      "epoch": 0.12111321078852431,
+      "grad_norm": 1.060289740562439,
+      "learning_rate": 4.998311827606203e-06,
+      "loss": 1.208,
+      "step": 235
+    },
+    {
+      "epoch": 0.12162858615358185,
+      "grad_norm": 1.0146334171295166,
+      "learning_rate": 4.998286727882888e-06,
+      "loss": 1.2077,
+      "step": 236
+    },
+    {
+      "epoch": 0.1221439615186394,
+      "grad_norm": 1.0203336477279663,
+      "learning_rate": 4.998261443006721e-06,
+      "loss": 1.1844,
+      "step": 237
+    },
+    {
+      "epoch": 0.12265933688369696,
+      "grad_norm": 1.0171735286712646,
+      "learning_rate": 4.998235972979576e-06,
+      "loss": 1.2218,
+      "step": 238
+    },
+    {
+      "epoch": 0.12317471224875451,
+      "grad_norm": 1.0859501361846924,
+      "learning_rate": 4.998210317803341e-06,
+      "loss": 1.1748,
+      "step": 239
+    },
+    {
+      "epoch": 0.12369008761381206,
+      "grad_norm": 1.0058231353759766,
+      "learning_rate": 4.998184477479918e-06,
+      "loss": 1.1903,
+      "step": 240
+    },
+    {
+      "epoch": 0.12420546297886961,
+      "grad_norm": 1.0329095125198364,
+      "learning_rate": 4.998158452011219e-06,
+      "loss": 1.1815,
+      "step": 241
+    },
+    {
+      "epoch": 0.12472083834392717,
+      "grad_norm": 1.0615721940994263,
+      "learning_rate": 4.998132241399176e-06,
+      "loss": 1.2004,
+      "step": 242
+    },
+    {
+      "epoch": 0.12523621370898472,
+      "grad_norm": 1.030116081237793,
+      "learning_rate": 4.998105845645731e-06,
+      "loss": 1.1981,
+      "step": 243
+    },
+    {
+      "epoch": 0.12575158907404227,
+      "grad_norm": 1.0303820371627808,
+      "learning_rate": 4.9980792647528384e-06,
+      "loss": 1.2333,
+      "step": 244
+    },
+    {
+      "epoch": 0.12626696443909982,
+      "grad_norm": 1.1086900234222412,
+      "learning_rate": 4.99805249872247e-06,
+      "loss": 1.1616,
+      "step": 245
+    },
+    {
+      "epoch": 0.12678233980415737,
+      "grad_norm": 1.0880281925201416,
+      "learning_rate": 4.998025547556609e-06,
+      "loss": 1.235,
+      "step": 246
+    },
+    {
+      "epoch": 0.1272977151692149,
+      "grad_norm": 1.0249994993209839,
+      "learning_rate": 4.997998411257252e-06,
+      "loss": 1.2356,
+      "step": 247
+    },
+    {
+      "epoch": 0.12781309053427245,
+      "grad_norm": 1.1765490770339966,
+      "learning_rate": 4.997971089826412e-06,
+      "loss": 1.2925,
+      "step": 248
+    },
+    {
+      "epoch": 0.12832846589933,
+      "grad_norm": 1.064995527267456,
+      "learning_rate": 4.997943583266112e-06,
+      "loss": 1.2629,
+      "step": 249
+    },
+    {
+      "epoch": 0.12884384126438755,
+      "grad_norm": 1.0707592964172363,
+      "learning_rate": 4.997915891578391e-06,
+      "loss": 1.2272,
+      "step": 250
+    },
+    {
+      "epoch": 0.1293592166294451,
+      "grad_norm": 1.0647047758102417,
+      "learning_rate": 4.997888014765301e-06,
+      "loss": 1.1205,
+      "step": 251
+    },
+    {
+      "epoch": 0.12987459199450266,
+      "grad_norm": 1.0850799083709717,
+      "learning_rate": 4.9978599528289094e-06,
+      "loss": 1.1708,
+      "step": 252
+    },
+    {
+      "epoch": 0.1303899673595602,
+      "grad_norm": 1.105841040611267,
+      "learning_rate": 4.997831705771295e-06,
+      "loss": 1.2088,
+      "step": 253
+    },
+    {
+      "epoch": 0.13090534272461776,
+      "grad_norm": 1.0339291095733643,
+      "learning_rate": 4.997803273594551e-06,
+      "loss": 1.1912,
+      "step": 254
+    },
+    {
+      "epoch": 0.13142071808967531,
+      "grad_norm": 0.9818682074546814,
+      "learning_rate": 4.997774656300784e-06,
+      "loss": 1.1551,
+      "step": 255
+    },
+    {
+      "epoch": 0.13193609345473287,
+      "grad_norm": 1.044981598854065,
+      "learning_rate": 4.997745853892118e-06,
+      "loss": 1.1997,
+      "step": 256
+    },
+    {
+      "epoch": 0.13245146881979042,
+      "grad_norm": 1.0077413320541382,
+      "learning_rate": 4.997716866370683e-06,
+      "loss": 1.2071,
+      "step": 257
+    },
+    {
+      "epoch": 0.13296684418484797,
+      "grad_norm": 1.0919009447097778,
+      "learning_rate": 4.99768769373863e-06,
+      "loss": 1.1637,
+      "step": 258
+    },
+    {
+      "epoch": 0.13348221954990552,
+      "grad_norm": 1.122795581817627,
+      "learning_rate": 4.997658335998121e-06,
+      "loss": 1.2381,
+      "step": 259
+    },
+    {
+      "epoch": 0.13399759491496308,
+      "grad_norm": 1.10574471950531,
+      "learning_rate": 4.9976287931513326e-06,
+      "loss": 1.2633,
+      "step": 260
+    },
+    {
+      "epoch": 0.13451297028002063,
+      "grad_norm": 1.1048368215560913,
+      "learning_rate": 4.997599065200452e-06,
+      "loss": 1.1864,
+      "step": 261
+    },
+    {
+      "epoch": 0.13502834564507815,
+      "grad_norm": 0.9917008876800537,
+      "learning_rate": 4.997569152147683e-06,
+      "loss": 1.2127,
+      "step": 262
+    },
+    {
+      "epoch": 0.1355437210101357,
+      "grad_norm": 1.0532373189926147,
+      "learning_rate": 4.997539053995244e-06,
+      "loss": 1.2023,
+      "step": 263
+    },
+    {
+      "epoch": 0.13605909637519326,
+      "grad_norm": 1.0858818292617798,
+      "learning_rate": 4.997508770745364e-06,
+      "loss": 1.2196,
+      "step": 264
+    },
+    {
+      "epoch": 0.1365744717402508,
+      "grad_norm": 1.035705327987671,
+      "learning_rate": 4.9974783024002886e-06,
+      "loss": 1.2244,
+      "step": 265
+    },
+    {
+      "epoch": 0.13708984710530836,
+      "grad_norm": 1.0526639223098755,
+      "learning_rate": 4.997447648962275e-06,
+      "loss": 1.1519,
+      "step": 266
+    },
+    {
+      "epoch": 0.1376052224703659,
+      "grad_norm": 0.9649120569229126,
+      "learning_rate": 4.997416810433595e-06,
+      "loss": 1.1471,
+      "step": 267
+    },
+    {
+      "epoch": 0.13812059783542346,
+      "grad_norm": 1.0110032558441162,
+      "learning_rate": 4.997385786816534e-06,
+      "loss": 1.1555,
+      "step": 268
+    },
+    {
+      "epoch": 0.13863597320048102,
+      "grad_norm": 1.08561110496521,
+      "learning_rate": 4.997354578113392e-06,
+      "loss": 1.2526,
+      "step": 269
+    },
+    {
+      "epoch": 0.13915134856553857,
+      "grad_norm": 1.045015811920166,
+      "learning_rate": 4.997323184326483e-06,
+      "loss": 1.21,
+      "step": 270
+    },
+    {
+      "epoch": 0.13966672393059612,
+      "grad_norm": 1.0490946769714355,
+      "learning_rate": 4.99729160545813e-06,
+      "loss": 1.1412,
+      "step": 271
+    },
+    {
+      "epoch": 0.14018209929565367,
+      "grad_norm": 1.0669465065002441,
+      "learning_rate": 4.997259841510677e-06,
+      "loss": 1.185,
+      "step": 272
+    },
+    {
+      "epoch": 0.14069747466071122,
+      "grad_norm": 1.0484760999679565,
+      "learning_rate": 4.997227892486476e-06,
+      "loss": 1.2348,
+      "step": 273
+    },
+    {
+      "epoch": 0.14121285002576878,
+      "grad_norm": 1.1375056505203247,
+      "learning_rate": 4.997195758387895e-06,
+      "loss": 1.1896,
+      "step": 274
+    },
+    {
+      "epoch": 0.14172822539082633,
+      "grad_norm": 1.055327296257019,
+      "learning_rate": 4.997163439217316e-06,
+      "loss": 1.1287,
+      "step": 275
+    },
+    {
+      "epoch": 0.14224360075588388,
+      "grad_norm": 1.1165558099746704,
+      "learning_rate": 4.997130934977134e-06,
+      "loss": 1.2059,
+      "step": 276
+    },
+    {
+      "epoch": 0.1427589761209414,
+      "grad_norm": 1.208341121673584,
+      "learning_rate": 4.997098245669759e-06,
+      "loss": 1.2388,
+      "step": 277
+    },
+    {
+      "epoch": 0.14327435148599896,
+      "grad_norm": 1.0463298559188843,
+      "learning_rate": 4.997065371297613e-06,
+      "loss": 1.1952,
+      "step": 278
+    },
+    {
+      "epoch": 0.1437897268510565,
+      "grad_norm": 1.0449258089065552,
+      "learning_rate": 4.997032311863131e-06,
+      "loss": 1.172,
+      "step": 279
+    },
+    {
+      "epoch": 0.14430510221611406,
+      "grad_norm": 1.1663317680358887,
+      "learning_rate": 4.996999067368766e-06,
+      "loss": 1.1271,
+      "step": 280
+    },
+    {
+      "epoch": 0.1448204775811716,
+      "grad_norm": 1.086437702178955,
+      "learning_rate": 4.996965637816979e-06,
+      "loss": 1.1885,
+      "step": 281
+    },
+    {
+      "epoch": 0.14533585294622917,
+      "grad_norm": 1.1258007287979126,
+      "learning_rate": 4.996932023210249e-06,
+      "loss": 1.2649,
+      "step": 282
+    },
+    {
+      "epoch": 0.14585122831128672,
+      "grad_norm": 1.0517827272415161,
+      "learning_rate": 4.9968982235510665e-06,
+      "loss": 1.1854,
+      "step": 283
+    },
+    {
+      "epoch": 0.14636660367634427,
+      "grad_norm": 1.1258128881454468,
+      "learning_rate": 4.996864238841936e-06,
+      "loss": 1.196,
+      "step": 284
+    },
+    {
+      "epoch": 0.14688197904140182,
+      "grad_norm": 1.2804971933364868,
+      "learning_rate": 4.996830069085379e-06,
+      "loss": 1.2104,
+      "step": 285
+    },
+    {
+      "epoch": 0.14739735440645937,
+      "grad_norm": 1.0372135639190674,
+      "learning_rate": 4.996795714283924e-06,
+      "loss": 1.2311,
+      "step": 286
+    },
+    {
+      "epoch": 0.14791272977151693,
+      "grad_norm": 1.1211464405059814,
+      "learning_rate": 4.99676117444012e-06,
+      "loss": 1.1966,
+      "step": 287
+    },
+    {
+      "epoch": 0.14842810513657448,
+      "grad_norm": 1.0660741329193115,
+      "learning_rate": 4.996726449556526e-06,
+      "loss": 1.2235,
+      "step": 288
+    },
+    {
+      "epoch": 0.14894348050163203,
+      "grad_norm": 1.0334893465042114,
+      "learning_rate": 4.996691539635715e-06,
+      "loss": 1.222,
+      "step": 289
+    },
+    {
+      "epoch": 0.14945885586668958,
+      "grad_norm": 1.1706844568252563,
+      "learning_rate": 4.996656444680275e-06,
+      "loss": 1.2978,
+      "step": 290
+    },
+    {
+      "epoch": 0.14997423123174713,
+      "grad_norm": 1.1230109930038452,
+      "learning_rate": 4.996621164692805e-06,
+      "loss": 1.1557,
+      "step": 291
+    },
+    {
+      "epoch": 0.15048960659680466,
+      "grad_norm": 1.1593081951141357,
+      "learning_rate": 4.9965856996759234e-06,
+      "loss": 1.214,
+      "step": 292
+    },
+    {
+      "epoch": 0.1510049819618622,
+      "grad_norm": 1.097525954246521,
+      "learning_rate": 4.996550049632255e-06,
+      "loss": 1.0969,
+      "step": 293
+    },
+    {
+      "epoch": 0.15152035732691976,
+      "grad_norm": 1.0412386655807495,
+      "learning_rate": 4.996514214564443e-06,
+      "loss": 1.2315,
+      "step": 294
+    },
+    {
+      "epoch": 0.15203573269197732,
+      "grad_norm": 1.103298306465149,
+      "learning_rate": 4.996478194475145e-06,
+      "loss": 1.1963,
+      "step": 295
+    },
+    {
+      "epoch": 0.15255110805703487,
+      "grad_norm": 1.147619366645813,
+      "learning_rate": 4.996441989367028e-06,
+      "loss": 1.2171,
+      "step": 296
+    },
+    {
+      "epoch": 0.15306648342209242,
+      "grad_norm": 1.0953634977340698,
+      "learning_rate": 4.9964055992427755e-06,
+      "loss": 1.1773,
+      "step": 297
+    },
+    {
+      "epoch": 0.15358185878714997,
+      "grad_norm": 1.1153252124786377,
+      "learning_rate": 4.996369024105086e-06,
+      "loss": 1.257,
+      "step": 298
+    },
+    {
+      "epoch": 0.15409723415220752,
+      "grad_norm": 1.1131457090377808,
+      "learning_rate": 4.996332263956669e-06,
+      "loss": 1.2331,
+      "step": 299
+    },
+    {
+      "epoch": 0.15461260951726508,
+      "grad_norm": 1.1843032836914062,
+      "learning_rate": 4.99629531880025e-06,
+      "loss": 1.2694,
+      "step": 300
+    },
+    {
+      "epoch": 0.15512798488232263,
+      "grad_norm": 1.1492340564727783,
+      "learning_rate": 4.996258188638565e-06,
+      "loss": 1.247,
+      "step": 301
+    },
+    {
+      "epoch": 0.15564336024738018,
+      "grad_norm": 1.1046638488769531,
+      "learning_rate": 4.9962208734743684e-06,
+      "loss": 1.1701,
+      "step": 302
+    },
+    {
+      "epoch": 0.15615873561243773,
+      "grad_norm": 1.108048677444458,
+      "learning_rate": 4.9961833733104234e-06,
+      "loss": 1.1667,
+      "step": 303
+    },
+    {
+      "epoch": 0.15667411097749528,
+      "grad_norm": 1.096451759338379,
+      "learning_rate": 4.996145688149511e-06,
+      "loss": 1.221,
+      "step": 304
+    },
+    {
+      "epoch": 0.15718948634255284,
+      "grad_norm": 1.0377289056777954,
+      "learning_rate": 4.996107817994422e-06,
+      "loss": 1.1303,
+      "step": 305
+    },
+    {
+      "epoch": 0.1577048617076104,
+      "grad_norm": 1.0976608991622925,
+      "learning_rate": 4.996069762847966e-06,
+      "loss": 1.2131,
+      "step": 306
+    },
+    {
+      "epoch": 0.1582202370726679,
+      "grad_norm": 1.0804593563079834,
+      "learning_rate": 4.996031522712961e-06,
+      "loss": 1.1413,
+      "step": 307
+    },
+    {
+      "epoch": 0.15873561243772547,
+      "grad_norm": 1.0988240242004395,
+      "learning_rate": 4.995993097592241e-06,
+      "loss": 1.1787,
+      "step": 308
+    },
+    {
+      "epoch": 0.15925098780278302,
+      "grad_norm": 1.1158981323242188,
+      "learning_rate": 4.995954487488655e-06,
+      "loss": 1.1956,
+      "step": 309
+    },
+    {
+      "epoch": 0.15976636316784057,
+      "grad_norm": 1.06276535987854,
+      "learning_rate": 4.9959156924050635e-06,
+      "loss": 1.1488,
+      "step": 310
+    },
+    {
+      "epoch": 0.16028173853289812,
+      "grad_norm": 1.118338942527771,
+      "learning_rate": 4.995876712344343e-06,
+      "loss": 1.2046,
+      "step": 311
+    },
+    {
+      "epoch": 0.16079711389795567,
+      "grad_norm": 1.1164604425430298,
+      "learning_rate": 4.995837547309381e-06,
+      "loss": 1.1962,
+      "step": 312
+    },
+    {
+      "epoch": 0.16131248926301323,
+      "grad_norm": 1.0719722509384155,
+      "learning_rate": 4.99579819730308e-06,
+      "loss": 1.1474,
+      "step": 313
+    },
+    {
+      "epoch": 0.16182786462807078,
+      "grad_norm": 1.0934160947799683,
+      "learning_rate": 4.995758662328358e-06,
+      "loss": 1.2062,
+      "step": 314
+    },
+    {
+      "epoch": 0.16234323999312833,
+      "grad_norm": 1.1083345413208008,
+      "learning_rate": 4.995718942388143e-06,
+      "loss": 1.181,
+      "step": 315
+    },
+    {
+      "epoch": 0.16285861535818588,
+      "grad_norm": 1.188507080078125,
+      "learning_rate": 4.99567903748538e-06,
+      "loss": 1.2613,
+      "step": 316
+    },
+    {
+      "epoch": 0.16337399072324343,
+      "grad_norm": 1.0850340127944946,
+      "learning_rate": 4.995638947623026e-06,
+      "loss": 1.1736,
+      "step": 317
+    },
+    {
+      "epoch": 0.163889366088301,
+      "grad_norm": 1.1532150506973267,
+      "learning_rate": 4.995598672804053e-06,
+      "loss": 1.2298,
+      "step": 318
+    },
+    {
+      "epoch": 0.16440474145335854,
+      "grad_norm": 1.1475430727005005,
+      "learning_rate": 4.9955582130314436e-06,
+      "loss": 1.2495,
+      "step": 319
+    },
+    {
+      "epoch": 0.1649201168184161,
+      "grad_norm": 1.221464991569519,
+      "learning_rate": 4.995517568308199e-06,
+      "loss": 1.3148,
+      "step": 320
+    },
+    {
+      "epoch": 0.16543549218347364,
+      "grad_norm": 1.0542255640029907,
+      "learning_rate": 4.99547673863733e-06,
+      "loss": 1.1144,
+      "step": 321
+    },
+    {
+      "epoch": 0.16595086754853117,
+      "grad_norm": 1.0566575527191162,
+      "learning_rate": 4.995435724021862e-06,
+      "loss": 1.1591,
+      "step": 322
+    },
+    {
+      "epoch": 0.16646624291358872,
+      "grad_norm": 1.0220612287521362,
+      "learning_rate": 4.995394524464836e-06,
+      "loss": 1.1624,
+      "step": 323
+    },
+    {
+      "epoch": 0.16698161827864627,
+      "grad_norm": 1.1840770244598389,
+      "learning_rate": 4.9953531399693056e-06,
+      "loss": 1.189,
+      "step": 324
+    },
+    {
+      "epoch": 0.16749699364370382,
+      "grad_norm": 1.1456857919692993,
+      "learning_rate": 4.995311570538336e-06,
+      "loss": 1.2096,
+      "step": 325
+    },
+    {
+      "epoch": 0.16801236900876138,
+      "grad_norm": 1.1925227642059326,
+      "learning_rate": 4.99526981617501e-06,
+      "loss": 1.2296,
+      "step": 326
+    },
+    {
+      "epoch": 0.16852774437381893,
+      "grad_norm": 1.1686952114105225,
+      "learning_rate": 4.995227876882421e-06,
+      "loss": 1.2267,
+      "step": 327
+    },
+    {
+      "epoch": 0.16904311973887648,
+      "grad_norm": 1.1459681987762451,
+      "learning_rate": 4.9951857526636775e-06,
+      "loss": 1.1841,
+      "step": 328
+    },
+    {
+      "epoch": 0.16955849510393403,
+      "grad_norm": 1.1983642578125,
+      "learning_rate": 4.995143443521902e-06,
+      "loss": 1.1892,
+      "step": 329
+    },
+    {
+      "epoch": 0.17007387046899158,
+      "grad_norm": 1.1032333374023438,
+      "learning_rate": 4.9951009494602275e-06,
+      "loss": 1.1855,
+      "step": 330
+    },
+    {
+      "epoch": 0.17058924583404914,
+      "grad_norm": 1.1817179918289185,
+      "learning_rate": 4.995058270481807e-06,
+      "loss": 1.1391,
+      "step": 331
+    },
+    {
+      "epoch": 0.1711046211991067,
+      "grad_norm": 1.1281492710113525,
+      "learning_rate": 4.9950154065898015e-06,
+      "loss": 1.2523,
+      "step": 332
+    },
+    {
+      "epoch": 0.17161999656416424,
+      "grad_norm": 1.1411288976669312,
+      "learning_rate": 4.9949723577873875e-06,
+      "loss": 1.2221,
+      "step": 333
+    },
+    {
+      "epoch": 0.1721353719292218,
+      "grad_norm": 1.2243231534957886,
+      "learning_rate": 4.994929124077756e-06,
+      "loss": 1.2469,
+      "step": 334
+    },
+    {
+      "epoch": 0.17265074729427934,
+      "grad_norm": 1.1771996021270752,
+      "learning_rate": 4.994885705464112e-06,
+      "loss": 1.1737,
+      "step": 335
+    },
+    {
+      "epoch": 0.1731661226593369,
+      "grad_norm": 1.1274346113204956,
+      "learning_rate": 4.9948421019496715e-06,
+      "loss": 1.2228,
+      "step": 336
+    },
+    {
+      "epoch": 0.17368149802439442,
+      "grad_norm": 1.0713800191879272,
+      "learning_rate": 4.994798313537668e-06,
+      "loss": 1.2221,
+      "step": 337
+    },
+    {
+      "epoch": 0.17419687338945197,
+      "grad_norm": 1.2393444776535034,
+      "learning_rate": 4.994754340231345e-06,
+      "loss": 1.1559,
+      "step": 338
+    },
+    {
+      "epoch": 0.17471224875450952,
+      "grad_norm": 1.173123836517334,
+      "learning_rate": 4.9947101820339615e-06,
+      "loss": 1.1904,
+      "step": 339
+    },
+    {
+      "epoch": 0.17522762411956708,
+      "grad_norm": 1.181182861328125,
+      "learning_rate": 4.994665838948792e-06,
+      "loss": 1.1822,
+      "step": 340
+    },
+    {
+      "epoch": 0.17574299948462463,
+      "grad_norm": 1.076783537864685,
+      "learning_rate": 4.994621310979121e-06,
+      "loss": 1.2435,
+      "step": 341
+    },
+    {
+      "epoch": 0.17625837484968218,
+      "grad_norm": 1.1784260272979736,
+      "learning_rate": 4.99457659812825e-06,
+      "loss": 1.2318,
+      "step": 342
+    },
+    {
+      "epoch": 0.17677375021473973,
+      "grad_norm": 1.105627179145813,
+      "learning_rate": 4.994531700399491e-06,
+      "loss": 1.238,
+      "step": 343
+    },
+    {
+      "epoch": 0.17728912557979729,
+      "grad_norm": 1.1091389656066895,
+      "learning_rate": 4.994486617796172e-06,
+      "loss": 1.1879,
+      "step": 344
+    },
+    {
+      "epoch": 0.17780450094485484,
+      "grad_norm": 1.0326907634735107,
+      "learning_rate": 4.994441350321636e-06,
+      "loss": 1.1607,
+      "step": 345
+    },
+    {
+      "epoch": 0.1783198763099124,
+      "grad_norm": 1.0360324382781982,
+      "learning_rate": 4.994395897979236e-06,
+      "loss": 1.1719,
+      "step": 346
+    },
+    {
+      "epoch": 0.17883525167496994,
+      "grad_norm": 1.1248440742492676,
+      "learning_rate": 4.994350260772341e-06,
+      "loss": 1.1273,
+      "step": 347
+    },
+    {
+      "epoch": 0.1793506270400275,
+      "grad_norm": 1.0810964107513428,
+      "learning_rate": 4.9943044387043324e-06,
+      "loss": 1.2178,
+      "step": 348
+    },
+    {
+      "epoch": 0.17986600240508505,
+      "grad_norm": 1.1901923418045044,
+      "learning_rate": 4.994258431778608e-06,
+      "loss": 1.2045,
+      "step": 349
+    },
+    {
+      "epoch": 0.1803813777701426,
+      "grad_norm": 1.3648993968963623,
+      "learning_rate": 4.994212239998577e-06,
+      "loss": 1.2434,
+      "step": 350
+    },
+    {
+      "epoch": 0.18089675313520015,
+      "grad_norm": 1.2584412097930908,
+      "learning_rate": 4.994165863367661e-06,
+      "loss": 1.2653,
+      "step": 351
+    },
+    {
+      "epoch": 0.18141212850025767,
+      "grad_norm": 1.1656162738800049,
+      "learning_rate": 4.994119301889299e-06,
+      "loss": 1.2114,
+      "step": 352
+    },
+    {
+      "epoch": 0.18192750386531523,
+      "grad_norm": 1.3355895280838013,
+      "learning_rate": 4.994072555566941e-06,
+      "loss": 1.2333,
+      "step": 353
+    },
+    {
+      "epoch": 0.18244287923037278,
+      "grad_norm": 1.1254146099090576,
+      "learning_rate": 4.994025624404052e-06,
+      "loss": 1.1686,
+      "step": 354
+    },
+    {
+      "epoch": 0.18295825459543033,
+      "grad_norm": 1.1978436708450317,
+      "learning_rate": 4.993978508404109e-06,
+      "loss": 1.121,
+      "step": 355
+    },
+    {
+      "epoch": 0.18347362996048788,
+      "grad_norm": 1.1566013097763062,
+      "learning_rate": 4.9939312075706046e-06,
+      "loss": 1.1487,
+      "step": 356
+    },
+    {
+      "epoch": 0.18398900532554543,
+      "grad_norm": 1.1110261678695679,
+      "learning_rate": 4.993883721907045e-06,
+      "loss": 1.1865,
+      "step": 357
+    },
+    {
+      "epoch": 0.184504380690603,
+      "grad_norm": 1.1373953819274902,
+      "learning_rate": 4.993836051416948e-06,
+      "loss": 1.1603,
+      "step": 358
+    },
+    {
+      "epoch": 0.18501975605566054,
+      "grad_norm": 1.1255364418029785,
+      "learning_rate": 4.993788196103847e-06,
+      "loss": 1.2125,
+      "step": 359
+    },
+    {
+      "epoch": 0.1855351314207181,
+      "grad_norm": 1.1953908205032349,
+      "learning_rate": 4.9937401559712895e-06,
+      "loss": 1.1718,
+      "step": 360
+    },
+    {
+      "epoch": 0.18605050678577564,
+      "grad_norm": 1.1173852682113647,
+      "learning_rate": 4.993691931022836e-06,
+      "loss": 1.1882,
+      "step": 361
+    },
+    {
+      "epoch": 0.1865658821508332,
+      "grad_norm": 1.155800461769104,
+      "learning_rate": 4.993643521262059e-06,
+      "loss": 1.2037,
+      "step": 362
+    },
+    {
+      "epoch": 0.18708125751589075,
+      "grad_norm": 1.1004916429519653,
+      "learning_rate": 4.993594926692547e-06,
+      "loss": 1.2128,
+      "step": 363
+    },
+    {
+      "epoch": 0.1875966328809483,
+      "grad_norm": 1.2067153453826904,
+      "learning_rate": 4.993546147317902e-06,
+      "loss": 1.2047,
+      "step": 364
+    },
+    {
+      "epoch": 0.18811200824600585,
+      "grad_norm": 1.1071127653121948,
+      "learning_rate": 4.993497183141739e-06,
+      "loss": 1.1837,
+      "step": 365
+    },
+    {
+      "epoch": 0.1886273836110634,
+      "grad_norm": 1.2650222778320312,
+      "learning_rate": 4.993448034167685e-06,
+      "loss": 1.1978,
+      "step": 366
+    },
+    {
+      "epoch": 0.18914275897612093,
+      "grad_norm": 1.251857042312622,
+      "learning_rate": 4.993398700399386e-06,
+      "loss": 1.194,
+      "step": 367
+    },
+    {
+      "epoch": 0.18965813434117848,
+      "grad_norm": 1.1337946653366089,
+      "learning_rate": 4.993349181840495e-06,
+      "loss": 1.1191,
+      "step": 368
+    },
+    {
+      "epoch": 0.19017350970623603,
+      "grad_norm": 1.1532639265060425,
+      "learning_rate": 4.9932994784946835e-06,
+      "loss": 1.1845,
+      "step": 369
+    },
+    {
+      "epoch": 0.19068888507129358,
+      "grad_norm": 1.1313118934631348,
+      "learning_rate": 4.993249590365635e-06,
+      "loss": 1.1588,
+      "step": 370
+    },
+    {
+      "epoch": 0.19120426043635114,
+      "grad_norm": 1.1309375762939453,
+      "learning_rate": 4.993199517457046e-06,
+      "loss": 1.1107,
+      "step": 371
+    },
+    {
+      "epoch": 0.1917196358014087,
+      "grad_norm": 1.2140612602233887,
+      "learning_rate": 4.9931492597726285e-06,
+      "loss": 1.1418,
+      "step": 372
+    },
+    {
+      "epoch": 0.19223501116646624,
+      "grad_norm": 1.2233984470367432,
+      "learning_rate": 4.993098817316107e-06,
+      "loss": 1.1627,
+      "step": 373
+    },
+    {
+      "epoch": 0.1927503865315238,
+      "grad_norm": 1.1959316730499268,
+      "learning_rate": 4.993048190091219e-06,
+      "loss": 1.2024,
+      "step": 374
+    },
+    {
+      "epoch": 0.19326576189658135,
+      "grad_norm": 1.1958948373794556,
+      "learning_rate": 4.9929973781017175e-06,
+      "loss": 1.2381,
+      "step": 375
+    },
+    {
+      "epoch": 0.1937811372616389,
+      "grad_norm": 1.071413516998291,
+      "learning_rate": 4.9929463813513676e-06,
+      "loss": 1.1692,
+      "step": 376
+    },
+    {
+      "epoch": 0.19429651262669645,
+      "grad_norm": 1.1295154094696045,
+      "learning_rate": 4.99289519984395e-06,
+      "loss": 1.2469,
+      "step": 377
+    },
+    {
+      "epoch": 0.194811887991754,
+      "grad_norm": 1.059316635131836,
+      "learning_rate": 4.9928438335832575e-06,
+      "loss": 1.1579,
+      "step": 378
+    },
+    {
+      "epoch": 0.19532726335681155,
+      "grad_norm": 1.0744855403900146,
+      "learning_rate": 4.992792282573095e-06,
+      "loss": 1.1765,
+      "step": 379
+    },
+    {
+      "epoch": 0.1958426387218691,
+      "grad_norm": 1.1298083066940308,
+      "learning_rate": 4.992740546817286e-06,
+      "loss": 1.2546,
+      "step": 380
+    },
+    {
+      "epoch": 0.19635801408692666,
+      "grad_norm": 1.0984526872634888,
+      "learning_rate": 4.992688626319662e-06,
+      "loss": 1.1716,
+      "step": 381
+    },
+    {
+      "epoch": 0.19687338945198418,
+      "grad_norm": 1.1361417770385742,
+      "learning_rate": 4.992636521084073e-06,
+      "loss": 1.1532,
+      "step": 382
+    },
+    {
+      "epoch": 0.19738876481704173,
+      "grad_norm": 1.1759017705917358,
+      "learning_rate": 4.9925842311143794e-06,
+      "loss": 1.2112,
+      "step": 383
+    },
+    {
+      "epoch": 0.1979041401820993,
+      "grad_norm": 1.1265839338302612,
+      "learning_rate": 4.992531756414457e-06,
+      "loss": 1.1864,
+      "step": 384
+    },
+    {
+      "epoch": 0.19841951554715684,
+      "grad_norm": 1.319730281829834,
+      "learning_rate": 4.992479096988196e-06,
+      "loss": 1.1995,
+      "step": 385
+    },
+    {
+      "epoch": 0.1989348909122144,
+      "grad_norm": 1.1547197103500366,
+      "learning_rate": 4.992426252839496e-06,
+      "loss": 1.2146,
+      "step": 386
+    },
+    {
+      "epoch": 0.19945026627727194,
+      "grad_norm": 1.1917409896850586,
+      "learning_rate": 4.992373223972276e-06,
+      "loss": 1.1627,
+      "step": 387
+    },
+    {
+      "epoch": 0.1999656416423295,
+      "grad_norm": 1.0941272974014282,
+      "learning_rate": 4.992320010390465e-06,
+      "loss": 1.1789,
+      "step": 388
+    },
+    {
+      "epoch": 0.20048101700738705,
+      "grad_norm": 1.236716389656067,
+      "learning_rate": 4.992266612098008e-06,
+      "loss": 1.2297,
+      "step": 389
+    },
+    {
+      "epoch": 0.2009963923724446,
+      "grad_norm": 1.0751968622207642,
+      "learning_rate": 4.9922130290988606e-06,
+      "loss": 1.1377,
+      "step": 390
+    },
+    {
+      "epoch": 0.20151176773750215,
+      "grad_norm": 1.168311595916748,
+      "learning_rate": 4.992159261396995e-06,
+      "loss": 1.2123,
+      "step": 391
+    },
+    {
+      "epoch": 0.2020271431025597,
+      "grad_norm": 1.1629900932312012,
+      "learning_rate": 4.9921053089963956e-06,
+      "loss": 1.1618,
+      "step": 392
+    },
+    {
+      "epoch": 0.20254251846761726,
+      "grad_norm": 1.1627777814865112,
+      "learning_rate": 4.992051171901062e-06,
+      "loss": 1.2022,
+      "step": 393
+    },
+    {
+      "epoch": 0.2030578938326748,
+      "grad_norm": 1.0521548986434937,
+      "learning_rate": 4.991996850115005e-06,
+      "loss": 1.1376,
+      "step": 394
+    },
+    {
+      "epoch": 0.20357326919773236,
+      "grad_norm": 1.0964367389678955,
+      "learning_rate": 4.991942343642251e-06,
+      "loss": 1.151,
+      "step": 395
+    },
+    {
+      "epoch": 0.2040886445627899,
+      "grad_norm": 1.103646993637085,
+      "learning_rate": 4.99188765248684e-06,
+      "loss": 1.1088,
+      "step": 396
+    },
+    {
+      "epoch": 0.20460401992784744,
+      "grad_norm": 1.2664456367492676,
+      "learning_rate": 4.991832776652824e-06,
+      "loss": 1.1976,
+      "step": 397
+    },
+    {
+      "epoch": 0.205119395292905,
+      "grad_norm": 1.0794459581375122,
+      "learning_rate": 4.991777716144272e-06,
+      "loss": 1.1382,
+      "step": 398
+    },
+    {
+      "epoch": 0.20563477065796254,
+      "grad_norm": 1.1090879440307617,
+      "learning_rate": 4.991722470965263e-06,
+      "loss": 1.1332,
+      "step": 399
+    },
+    {
+      "epoch": 0.2061501460230201,
+      "grad_norm": 1.1466739177703857,
+      "learning_rate": 4.991667041119892e-06,
+      "loss": 1.1976,
+      "step": 400
+    },
+    {
+      "epoch": 0.20666552138807764,
+      "grad_norm": 1.1100345849990845,
+      "learning_rate": 4.991611426612267e-06,
+      "loss": 1.2142,
+      "step": 401
+    },
+    {
+      "epoch": 0.2071808967531352,
+      "grad_norm": 1.1167898178100586,
+      "learning_rate": 4.991555627446509e-06,
+      "loss": 1.1701,
+      "step": 402
+    },
+    {
+      "epoch": 0.20769627211819275,
+      "grad_norm": 1.124314546585083,
+      "learning_rate": 4.9914996436267545e-06,
+      "loss": 1.1326,
+      "step": 403
+    },
+    {
+      "epoch": 0.2082116474832503,
+      "grad_norm": 1.0986205339431763,
+      "learning_rate": 4.991443475157152e-06,
+      "loss": 1.1814,
+      "step": 404
+    },
+    {
+      "epoch": 0.20872702284830785,
+      "grad_norm": 1.1348336935043335,
+      "learning_rate": 4.9913871220418645e-06,
+      "loss": 1.2108,
+      "step": 405
+    },
+    {
+      "epoch": 0.2092423982133654,
+      "grad_norm": 1.117326259613037,
+      "learning_rate": 4.991330584285068e-06,
+      "loss": 1.1665,
+      "step": 406
+    },
+    {
+      "epoch": 0.20975777357842296,
+      "grad_norm": 1.251592755317688,
+      "learning_rate": 4.991273861890952e-06,
+      "loss": 1.1327,
+      "step": 407
+    },
+    {
+      "epoch": 0.2102731489434805,
+      "grad_norm": 1.2220373153686523,
+      "learning_rate": 4.991216954863722e-06,
+      "loss": 1.181,
+      "step": 408
+    },
+    {
+      "epoch": 0.21078852430853806,
+      "grad_norm": 1.265209436416626,
+      "learning_rate": 4.991159863207595e-06,
+      "loss": 1.1855,
+      "step": 409
+    },
+    {
+      "epoch": 0.2113038996735956,
+      "grad_norm": 1.1213948726654053,
+      "learning_rate": 4.991102586926801e-06,
+      "loss": 1.1839,
+      "step": 410
+    },
+    {
+      "epoch": 0.21181927503865317,
+      "grad_norm": 1.1888283491134644,
+      "learning_rate": 4.991045126025585e-06,
+      "loss": 1.1649,
+      "step": 411
+    },
+    {
+      "epoch": 0.2123346504037107,
+      "grad_norm": 1.208165168762207,
+      "learning_rate": 4.990987480508208e-06,
+      "loss": 1.1751,
+      "step": 412
+    },
+    {
+      "epoch": 0.21285002576876824,
+      "grad_norm": 1.254124641418457,
+      "learning_rate": 4.990929650378939e-06,
+      "loss": 1.1697,
+      "step": 413
+    },
+    {
+      "epoch": 0.2133654011338258,
+      "grad_norm": 1.1125601530075073,
+      "learning_rate": 4.990871635642066e-06,
+      "loss": 1.1212,
+      "step": 414
+    },
+    {
+      "epoch": 0.21388077649888335,
+      "grad_norm": 1.1498005390167236,
+      "learning_rate": 4.990813436301888e-06,
+      "loss": 1.1929,
+      "step": 415
+    },
+    {
+      "epoch": 0.2143961518639409,
+      "grad_norm": 1.1422375440597534,
+      "learning_rate": 4.990755052362718e-06,
+      "loss": 1.1379,
+      "step": 416
+    },
+    {
+      "epoch": 0.21491152722899845,
+      "grad_norm": 1.214056134223938,
+      "learning_rate": 4.990696483828883e-06,
+      "loss": 1.1666,
+      "step": 417
+    },
+    {
+      "epoch": 0.215426902594056,
+      "grad_norm": 1.1769565343856812,
+      "learning_rate": 4.990637730704723e-06,
+      "loss": 1.1579,
+      "step": 418
+    },
+    {
+      "epoch": 0.21594227795911355,
+      "grad_norm": 1.1861079931259155,
+      "learning_rate": 4.990578792994594e-06,
+      "loss": 1.162,
+      "step": 419
+    },
+    {
+      "epoch": 0.2164576533241711,
+      "grad_norm": 1.1193722486495972,
+      "learning_rate": 4.990519670702862e-06,
+      "loss": 1.201,
+      "step": 420
+    },
+    {
+      "epoch": 0.21697302868922866,
+      "grad_norm": 1.1538455486297607,
+      "learning_rate": 4.990460363833911e-06,
+      "loss": 1.1193,
+      "step": 421
+    },
+    {
+      "epoch": 0.2174884040542862,
+      "grad_norm": 1.1257574558258057,
+      "learning_rate": 4.990400872392135e-06,
+      "loss": 1.1047,
+      "step": 422
+    },
+    {
+      "epoch": 0.21800377941934376,
+      "grad_norm": 1.1589558124542236,
+      "learning_rate": 4.990341196381943e-06,
+      "loss": 1.203,
+      "step": 423
+    },
+    {
+      "epoch": 0.21851915478440131,
+      "grad_norm": 1.1535980701446533,
+      "learning_rate": 4.990281335807757e-06,
+      "loss": 1.0999,
+      "step": 424
+    },
+    {
+      "epoch": 0.21903453014945887,
+      "grad_norm": 1.198763132095337,
+      "learning_rate": 4.990221290674015e-06,
+      "loss": 1.1161,
+      "step": 425
+    },
+    {
+      "epoch": 0.21954990551451642,
+      "grad_norm": 1.258095383644104,
+      "learning_rate": 4.990161060985164e-06,
+      "loss": 1.2035,
+      "step": 426
+    },
+    {
+      "epoch": 0.22006528087957394,
+      "grad_norm": 1.1130949258804321,
+      "learning_rate": 4.990100646745673e-06,
+      "loss": 1.1441,
+      "step": 427
+    },
+    {
+      "epoch": 0.2205806562446315,
+      "grad_norm": 1.224503517150879,
+      "learning_rate": 4.990040047960015e-06,
+      "loss": 1.1787,
+      "step": 428
+    },
+    {
+      "epoch": 0.22109603160968905,
+      "grad_norm": 1.2775068283081055,
+      "learning_rate": 4.989979264632683e-06,
+      "loss": 1.1818,
+      "step": 429
+    },
+    {
+      "epoch": 0.2216114069747466,
+      "grad_norm": 1.1601276397705078,
+      "learning_rate": 4.989918296768181e-06,
+      "loss": 1.1887,
+      "step": 430
+    },
+    {
+      "epoch": 0.22212678233980415,
+      "grad_norm": 1.1515289545059204,
+      "learning_rate": 4.9898571443710265e-06,
+      "loss": 1.1445,
+      "step": 431
+    },
+    {
+      "epoch": 0.2226421577048617,
+      "grad_norm": 1.141028881072998,
+      "learning_rate": 4.989795807445753e-06,
+      "loss": 1.1901,
+      "step": 432
+    },
+    {
+      "epoch": 0.22315753306991926,
+      "grad_norm": 1.192578673362732,
+      "learning_rate": 4.989734285996907e-06,
+      "loss": 1.1934,
+      "step": 433
+    },
+    {
+      "epoch": 0.2236729084349768,
+      "grad_norm": 1.2101963758468628,
+      "learning_rate": 4.989672580029047e-06,
+      "loss": 1.1493,
+      "step": 434
+    },
+    {
+      "epoch": 0.22418828380003436,
+      "grad_norm": 1.150976538658142,
+      "learning_rate": 4.989610689546747e-06,
+      "loss": 1.2027,
+      "step": 435
+    },
+    {
+      "epoch": 0.2247036591650919,
+      "grad_norm": 1.1535725593566895,
+      "learning_rate": 4.989548614554592e-06,
+      "loss": 1.1637,
+      "step": 436
+    },
+    {
+      "epoch": 0.22521903453014946,
+      "grad_norm": 1.12608802318573,
+      "learning_rate": 4.989486355057184e-06,
+      "loss": 1.1192,
+      "step": 437
+    },
+    {
+      "epoch": 0.22573440989520702,
+      "grad_norm": 1.143556833267212,
+      "learning_rate": 4.989423911059136e-06,
+      "loss": 1.1314,
+      "step": 438
+    },
+    {
+      "epoch": 0.22624978526026457,
+      "grad_norm": 1.2795722484588623,
+      "learning_rate": 4.989361282565077e-06,
+      "loss": 1.1525,
+      "step": 439
+    },
+    {
+      "epoch": 0.22676516062532212,
+      "grad_norm": 1.1248250007629395,
+      "learning_rate": 4.9892984695796484e-06,
+      "loss": 1.1635,
+      "step": 440
+    },
+    {
+      "epoch": 0.22728053599037967,
+      "grad_norm": 1.1849747896194458,
+      "learning_rate": 4.989235472107504e-06,
+      "loss": 1.1905,
+      "step": 441
+    },
+    {
+      "epoch": 0.2277959113554372,
+      "grad_norm": 1.243175983428955,
+      "learning_rate": 4.989172290153314e-06,
+      "loss": 1.1504,
+      "step": 442
+    },
+    {
+      "epoch": 0.22831128672049475,
+      "grad_norm": 1.2722020149230957,
+      "learning_rate": 4.989108923721762e-06,
+      "loss": 1.2047,
+      "step": 443
+    },
+    {
+      "epoch": 0.2288266620855523,
+      "grad_norm": 1.1677055358886719,
+      "learning_rate": 4.989045372817542e-06,
+      "loss": 1.1834,
+      "step": 444
+    },
+    {
+      "epoch": 0.22934203745060985,
+      "grad_norm": 1.2984954118728638,
+      "learning_rate": 4.988981637445365e-06,
+      "loss": 1.2295,
+      "step": 445
+    },
+    {
+      "epoch": 0.2298574128156674,
+      "grad_norm": 1.3156894445419312,
+      "learning_rate": 4.988917717609954e-06,
+      "loss": 1.1091,
+      "step": 446
+    },
+    {
+      "epoch": 0.23037278818072496,
+      "grad_norm": 1.2146261930465698,
+      "learning_rate": 4.988853613316047e-06,
+      "loss": 1.1843,
+      "step": 447
+    },
+    {
+      "epoch": 0.2308881635457825,
+      "grad_norm": 1.1358487606048584,
+      "learning_rate": 4.988789324568394e-06,
+      "loss": 1.1165,
+      "step": 448
+    },
+    {
+      "epoch": 0.23140353891084006,
+      "grad_norm": 1.2033812999725342,
+      "learning_rate": 4.98872485137176e-06,
+      "loss": 1.1391,
+      "step": 449
+    },
+    {
+      "epoch": 0.23191891427589761,
+      "grad_norm": 1.194493293762207,
+      "learning_rate": 4.9886601937309235e-06,
+      "loss": 1.1541,
+      "step": 450
+    },
+    {
+      "epoch": 0.23243428964095517,
+      "grad_norm": 1.121980905532837,
+      "learning_rate": 4.988595351650676e-06,
+      "loss": 1.1393,
+      "step": 451
+    },
+    {
+      "epoch": 0.23294966500601272,
+      "grad_norm": 1.1627380847930908,
+      "learning_rate": 4.988530325135823e-06,
+      "loss": 1.1332,
+      "step": 452
+    },
+    {
+      "epoch": 0.23346504037107027,
+      "grad_norm": 1.3269914388656616,
+      "learning_rate": 4.988465114191183e-06,
+      "loss": 1.1415,
+      "step": 453
+    },
+    {
+      "epoch": 0.23398041573612782,
+      "grad_norm": 1.1122335195541382,
+      "learning_rate": 4.988399718821592e-06,
+      "loss": 1.1451,
+      "step": 454
+    },
+    {
+      "epoch": 0.23449579110118537,
+      "grad_norm": 1.2022422552108765,
+      "learning_rate": 4.988334139031893e-06,
+      "loss": 1.1097,
+      "step": 455
+    },
+    {
+      "epoch": 0.23501116646624293,
+      "grad_norm": 1.1904706954956055,
+      "learning_rate": 4.988268374826948e-06,
+      "loss": 1.1822,
+      "step": 456
+    },
+    {
+      "epoch": 0.23552654183130045,
+      "grad_norm": 1.1483618021011353,
+      "learning_rate": 4.98820242621163e-06,
+      "loss": 1.138,
+      "step": 457
+    },
+    {
+      "epoch": 0.236041917196358,
+      "grad_norm": 1.140466570854187,
+      "learning_rate": 4.988136293190828e-06,
+      "loss": 1.1507,
+      "step": 458
+    },
+    {
+      "epoch": 0.23655729256141556,
+      "grad_norm": 1.12582528591156,
+      "learning_rate": 4.988069975769442e-06,
+      "loss": 1.1049,
+      "step": 459
+    },
+    {
+      "epoch": 0.2370726679264731,
+      "grad_norm": 1.2952275276184082,
+      "learning_rate": 4.9880034739523875e-06,
+      "loss": 1.1716,
+      "step": 460
+    },
+    {
+      "epoch": 0.23758804329153066,
+      "grad_norm": 1.1505956649780273,
+      "learning_rate": 4.9879367877445935e-06,
+      "loss": 1.1647,
+      "step": 461
+    },
+    {
+      "epoch": 0.2381034186565882,
+      "grad_norm": 1.237804889678955,
+      "learning_rate": 4.987869917151001e-06,
+      "loss": 1.2249,
+      "step": 462
+    },
+    {
+      "epoch": 0.23861879402164576,
+      "grad_norm": 1.2095214128494263,
+      "learning_rate": 4.987802862176566e-06,
+      "loss": 1.163,
+      "step": 463
+    },
+    {
+      "epoch": 0.23913416938670332,
+      "grad_norm": 1.2811049222946167,
+      "learning_rate": 4.98773562282626e-06,
+      "loss": 1.1861,
+      "step": 464
+    },
+    {
+      "epoch": 0.23964954475176087,
+      "grad_norm": 1.2019784450531006,
+      "learning_rate": 4.9876681991050635e-06,
+      "loss": 1.1322,
+      "step": 465
+    },
+    {
+      "epoch": 0.24016492011681842,
+      "grad_norm": 1.2116162776947021,
+      "learning_rate": 4.987600591017975e-06,
+      "loss": 1.2162,
+      "step": 466
+    },
+    {
+      "epoch": 0.24068029548187597,
+      "grad_norm": 1.1459062099456787,
+      "learning_rate": 4.987532798570005e-06,
+      "loss": 1.1615,
+      "step": 467
+    },
+    {
+      "epoch": 0.24119567084693352,
+      "grad_norm": 1.2083719968795776,
+      "learning_rate": 4.987464821766178e-06,
+      "loss": 1.1726,
+      "step": 468
+    },
+    {
+      "epoch": 0.24171104621199108,
+      "grad_norm": 1.2292635440826416,
+      "learning_rate": 4.987396660611531e-06,
+      "loss": 1.1212,
+      "step": 469
+    },
+    {
+      "epoch": 0.24222642157704863,
+      "grad_norm": 1.0754214525222778,
+      "learning_rate": 4.987328315111116e-06,
+      "loss": 1.1372,
+      "step": 470
+    },
+    {
+      "epoch": 0.24274179694210618,
+      "grad_norm": 1.205851674079895,
+      "learning_rate": 4.987259785269999e-06,
+      "loss": 1.1463,
+      "step": 471
+    },
+    {
+      "epoch": 0.2432571723071637,
+      "grad_norm": 1.2169773578643799,
+      "learning_rate": 4.987191071093257e-06,
+      "loss": 1.1705,
+      "step": 472
+    },
+    {
+      "epoch": 0.24377254767222126,
+      "grad_norm": 1.1939457654953003,
+      "learning_rate": 4.987122172585984e-06,
+      "loss": 1.1503,
+      "step": 473
+    },
+    {
+      "epoch": 0.2442879230372788,
+      "grad_norm": 1.179269552230835,
+      "learning_rate": 4.987053089753286e-06,
+      "loss": 1.1693,
+      "step": 474
+    },
+    {
+      "epoch": 0.24480329840233636,
+      "grad_norm": 1.1368995904922485,
+      "learning_rate": 4.986983822600283e-06,
+      "loss": 1.1204,
+      "step": 475
+    },
+    {
+      "epoch": 0.2453186737673939,
+      "grad_norm": 1.1833655834197998,
+      "learning_rate": 4.986914371132108e-06,
+      "loss": 1.1512,
+      "step": 476
+    },
+    {
+      "epoch": 0.24583404913245147,
+      "grad_norm": 1.1153130531311035,
+      "learning_rate": 4.986844735353908e-06,
+      "loss": 1.1726,
+      "step": 477
+    },
+    {
+      "epoch": 0.24634942449750902,
+      "grad_norm": 1.1726752519607544,
+      "learning_rate": 4.986774915270845e-06,
+      "loss": 1.1657,
+      "step": 478
+    },
+    {
+      "epoch": 0.24686479986256657,
+      "grad_norm": 1.2259719371795654,
+      "learning_rate": 4.9867049108880915e-06,
+      "loss": 1.1545,
+      "step": 479
+    },
+    {
+      "epoch": 0.24738017522762412,
+      "grad_norm": 1.2755433320999146,
+      "learning_rate": 4.986634722210838e-06,
+      "loss": 1.2373,
+      "step": 480
+    },
+    {
+      "epoch": 0.24789555059268167,
+      "grad_norm": 1.139876365661621,
+      "learning_rate": 4.986564349244285e-06,
+      "loss": 1.0958,
+      "step": 481
+    },
+    {
+      "epoch": 0.24841092595773923,
+      "grad_norm": 1.1974653005599976,
+      "learning_rate": 4.986493791993648e-06,
+      "loss": 1.1678,
+      "step": 482
+    },
+    {
+      "epoch": 0.24892630132279678,
+      "grad_norm": 1.124848484992981,
+      "learning_rate": 4.986423050464156e-06,
+      "loss": 1.1129,
+      "step": 483
+    },
+    {
+      "epoch": 0.24944167668785433,
+      "grad_norm": 1.2005858421325684,
+      "learning_rate": 4.986352124661053e-06,
+      "loss": 1.1665,
+      "step": 484
+    },
+    {
+      "epoch": 0.24995705205291188,
+      "grad_norm": 1.133514642715454,
+      "learning_rate": 4.986281014589594e-06,
+      "loss": 1.0938,
+      "step": 485
+    },
+    {
+      "epoch": 0.25047242741796943,
+      "grad_norm": 1.2003309726715088,
+      "learning_rate": 4.9862097202550496e-06,
+      "loss": 1.1792,
+      "step": 486
+    },
+    {
+      "epoch": 0.25098780278302696,
+      "grad_norm": 1.1737325191497803,
+      "learning_rate": 4.986138241662705e-06,
+      "loss": 1.1225,
+      "step": 487
+    },
+    {
+      "epoch": 0.25150317814808454,
+      "grad_norm": 1.1730244159698486,
+      "learning_rate": 4.9860665788178545e-06,
+      "loss": 1.1572,
+      "step": 488
+    },
+    {
+      "epoch": 0.25201855351314206,
+      "grad_norm": 1.206847906112671,
+      "learning_rate": 4.985994731725811e-06,
+      "loss": 1.1763,
+      "step": 489
+    },
+    {
+      "epoch": 0.25253392887819964,
+      "grad_norm": 1.1796056032180786,
+      "learning_rate": 4.985922700391901e-06,
+      "loss": 1.1572,
+      "step": 490
+    },
+    {
+      "epoch": 0.25304930424325717,
+      "grad_norm": 1.1734232902526855,
+      "learning_rate": 4.98585048482146e-06,
+      "loss": 1.1403,
+      "step": 491
+    },
+    {
+      "epoch": 0.25356467960831475,
+      "grad_norm": 1.1931697130203247,
+      "learning_rate": 4.98577808501984e-06,
+      "loss": 1.1966,
+      "step": 492
+    },
+    {
+      "epoch": 0.25408005497337227,
+      "grad_norm": 1.2577358484268188,
+      "learning_rate": 4.985705500992408e-06,
+      "loss": 1.1307,
+      "step": 493
+    },
+    {
+      "epoch": 0.2545954303384298,
+      "grad_norm": 1.2262202501296997,
+      "learning_rate": 4.985632732744544e-06,
+      "loss": 1.1662,
+      "step": 494
+    },
+    {
+      "epoch": 0.2551108057034874,
+      "grad_norm": 1.2096452713012695,
+      "learning_rate": 4.9855597802816405e-06,
+      "loss": 1.1953,
+      "step": 495
+    },
+    {
+      "epoch": 0.2556261810685449,
+      "grad_norm": 1.161454200744629,
+      "learning_rate": 4.985486643609103e-06,
+      "loss": 1.1444,
+      "step": 496
+    },
+    {
+      "epoch": 0.2561415564336025,
+      "grad_norm": 1.1236664056777954,
+      "learning_rate": 4.985413322732353e-06,
+      "loss": 1.157,
+      "step": 497
+    },
+    {
+      "epoch": 0.25665693179866,
+      "grad_norm": 1.314698338508606,
+      "learning_rate": 4.9853398176568235e-06,
+      "loss": 1.1376,
+      "step": 498
+    },
+    {
+      "epoch": 0.2571723071637176,
+      "grad_norm": 1.1444615125656128,
+      "learning_rate": 4.985266128387963e-06,
+      "loss": 1.126,
+      "step": 499
+    },
+    {
+      "epoch": 0.2576876825287751,
+      "grad_norm": 1.2295600175857544,
+      "learning_rate": 4.985192254931232e-06,
+      "loss": 1.15,
+      "step": 500
+    },
+    {
+      "epoch": 0.2582030578938327,
+      "grad_norm": 1.197485327720642,
+      "learning_rate": 4.985118197292106e-06,
+      "loss": 1.2171,
+      "step": 501
+    },
+    {
+      "epoch": 0.2587184332588902,
+      "grad_norm": 1.2253940105438232,
+      "learning_rate": 4.985043955476072e-06,
+      "loss": 1.1144,
+      "step": 502
+    },
+    {
+      "epoch": 0.2592338086239478,
+      "grad_norm": 1.2434132099151611,
+      "learning_rate": 4.984969529488636e-06,
+      "loss": 1.1776,
+      "step": 503
+    },
+    {
+      "epoch": 0.2597491839890053,
+      "grad_norm": 1.1865484714508057,
+      "learning_rate": 4.98489491933531e-06,
+      "loss": 1.1125,
+      "step": 504
+    },
+    {
+      "epoch": 0.2602645593540629,
+      "grad_norm": 1.2330963611602783,
+      "learning_rate": 4.984820125021624e-06,
+      "loss": 1.1663,
+      "step": 505
+    },
+    {
+      "epoch": 0.2607799347191204,
+      "grad_norm": 1.1894609928131104,
+      "learning_rate": 4.9847451465531235e-06,
+      "loss": 1.1503,
+      "step": 506
+    },
+    {
+      "epoch": 0.261295310084178,
+      "grad_norm": 1.209188461303711,
+      "learning_rate": 4.984669983935363e-06,
+      "loss": 1.1668,
+      "step": 507
+    },
+    {
+      "epoch": 0.2618106854492355,
+      "grad_norm": 1.1585817337036133,
+      "learning_rate": 4.984594637173915e-06,
+      "loss": 1.1212,
+      "step": 508
+    },
+    {
+      "epoch": 0.26232606081429305,
+      "grad_norm": 1.091931700706482,
+      "learning_rate": 4.9845191062743615e-06,
+      "loss": 1.1311,
+      "step": 509
+    },
+    {
+      "epoch": 0.26284143617935063,
+      "grad_norm": 1.207065463066101,
+      "learning_rate": 4.984443391242301e-06,
+      "loss": 1.18,
+      "step": 510
+    },
+    {
+      "epoch": 0.26335681154440815,
+      "grad_norm": 1.2607285976409912,
+      "learning_rate": 4.984367492083345e-06,
+      "loss": 1.1685,
+      "step": 511
+    },
+    {
+      "epoch": 0.26387218690946573,
+      "grad_norm": 1.1094540357589722,
+      "learning_rate": 4.984291408803118e-06,
+      "loss": 1.0477,
+      "step": 512
+    },
+    {
+      "epoch": 0.26438756227452326,
+      "grad_norm": 1.123847246170044,
+      "learning_rate": 4.9842151414072605e-06,
+      "loss": 1.1365,
+      "step": 513
+    },
+    {
+      "epoch": 0.26490293763958084,
+      "grad_norm": 1.189071536064148,
+      "learning_rate": 4.984138689901423e-06,
+      "loss": 1.1528,
+      "step": 514
+    },
+    {
+      "epoch": 0.26541831300463836,
+      "grad_norm": 1.227555274963379,
+      "learning_rate": 4.984062054291272e-06,
+      "loss": 1.0919,
+      "step": 515
+    },
+    {
+      "epoch": 0.26593368836969594,
+      "grad_norm": 1.148440957069397,
+      "learning_rate": 4.983985234582487e-06,
+      "loss": 1.1267,
+      "step": 516
+    },
+    {
+      "epoch": 0.26644906373475347,
+      "grad_norm": 1.227849006652832,
+      "learning_rate": 4.9839082307807605e-06,
+      "loss": 1.1306,
+      "step": 517
+    },
+    {
+      "epoch": 0.26696443909981105,
+      "grad_norm": 1.255359172821045,
+      "learning_rate": 4.983831042891801e-06,
+      "loss": 1.2317,
+      "step": 518
+    },
+    {
+      "epoch": 0.26747981446486857,
+      "grad_norm": 1.2154213190078735,
+      "learning_rate": 4.983753670921329e-06,
+      "loss": 1.1966,
+      "step": 519
+    },
+    {
+      "epoch": 0.26799518982992615,
+      "grad_norm": 1.2825920581817627,
+      "learning_rate": 4.983676114875078e-06,
+      "loss": 1.2206,
+      "step": 520
+    },
+    {
+      "epoch": 0.2685105651949837,
+      "grad_norm": 1.1783586740493774,
+      "learning_rate": 4.983598374758794e-06,
+      "loss": 1.1799,
+      "step": 521
+    },
+    {
+      "epoch": 0.26902594056004125,
+      "grad_norm": 1.1926697492599487,
+      "learning_rate": 4.983520450578242e-06,
+      "loss": 1.1976,
+      "step": 522
+    },
+    {
+      "epoch": 0.2695413159250988,
+      "grad_norm": 1.291904091835022,
+      "learning_rate": 4.983442342339195e-06,
+      "loss": 1.147,
+      "step": 523
+    },
+    {
+      "epoch": 0.2700566912901563,
+      "grad_norm": 1.2051136493682861,
+      "learning_rate": 4.9833640500474426e-06,
+      "loss": 1.1981,
+      "step": 524
+    },
+    {
+      "epoch": 0.2705720666552139,
+      "grad_norm": 1.1611932516098022,
+      "learning_rate": 4.983285573708786e-06,
+      "loss": 1.1444,
+      "step": 525
+    },
+    {
+      "epoch": 0.2710874420202714,
+      "grad_norm": 1.1475062370300293,
+      "learning_rate": 4.983206913329043e-06,
+      "loss": 1.1773,
+      "step": 526
+    },
+    {
+      "epoch": 0.271602817385329,
+      "grad_norm": 1.2453409433364868,
+      "learning_rate": 4.983128068914041e-06,
+      "loss": 1.1212,
+      "step": 527
+    },
+    {
+      "epoch": 0.2721181927503865,
+      "grad_norm": 1.224652886390686,
+      "learning_rate": 4.983049040469625e-06,
+      "loss": 1.13,
+      "step": 528
+    },
+    {
+      "epoch": 0.2726335681154441,
+      "grad_norm": 1.2433110475540161,
+      "learning_rate": 4.982969828001652e-06,
+      "loss": 1.1605,
+      "step": 529
+    },
+    {
+      "epoch": 0.2731489434805016,
+      "grad_norm": 1.1937365531921387,
+      "learning_rate": 4.982890431515992e-06,
+      "loss": 1.1063,
+      "step": 530
+    },
+    {
+      "epoch": 0.2736643188455592,
+      "grad_norm": 1.1492749452590942,
+      "learning_rate": 4.982810851018529e-06,
+      "loss": 1.1967,
+      "step": 531
+    },
+    {
+      "epoch": 0.2741796942106167,
+      "grad_norm": 1.1265840530395508,
+      "learning_rate": 4.982731086515161e-06,
+      "loss": 1.1016,
+      "step": 532
+    },
+    {
+      "epoch": 0.2746950695756743,
+      "grad_norm": 1.262799620628357,
+      "learning_rate": 4.9826511380118e-06,
+      "loss": 1.1779,
+      "step": 533
+    },
+    {
+      "epoch": 0.2752104449407318,
+      "grad_norm": 1.199607491493225,
+      "learning_rate": 4.982571005514371e-06,
+      "loss": 1.0855,
+      "step": 534
+    },
+    {
+      "epoch": 0.2757258203057894,
+      "grad_norm": 1.227246642112732,
+      "learning_rate": 4.982490689028812e-06,
+      "loss": 1.1195,
+      "step": 535
+    },
+    {
+      "epoch": 0.27624119567084693,
+      "grad_norm": 1.30074942111969,
+      "learning_rate": 4.982410188561077e-06,
+      "loss": 1.1773,
+      "step": 536
+    },
+    {
+      "epoch": 0.2767565710359045,
+      "grad_norm": 1.219909906387329,
+      "learning_rate": 4.982329504117131e-06,
+      "loss": 1.1434,
+      "step": 537
+    },
+    {
+      "epoch": 0.27727194640096203,
+      "grad_norm": 1.3017019033432007,
+      "learning_rate": 4.982248635702953e-06,
+      "loss": 1.2061,
+      "step": 538
+    },
+    {
+      "epoch": 0.27778732176601956,
+      "grad_norm": 1.2198671102523804,
+      "learning_rate": 4.9821675833245375e-06,
+      "loss": 1.1297,
+      "step": 539
+    },
+    {
+      "epoch": 0.27830269713107714,
+      "grad_norm": 1.2024649381637573,
+      "learning_rate": 4.982086346987891e-06,
+      "loss": 1.188,
+      "step": 540
+    },
+    {
+      "epoch": 0.27881807249613466,
+      "grad_norm": 1.2666974067687988,
+      "learning_rate": 4.9820049266990345e-06,
+      "loss": 1.2017,
+      "step": 541
+    },
+    {
+      "epoch": 0.27933344786119224,
+      "grad_norm": 1.197114109992981,
+      "learning_rate": 4.981923322464001e-06,
+      "loss": 1.1401,
+      "step": 542
+    },
+    {
+      "epoch": 0.27984882322624977,
+      "grad_norm": 1.2055046558380127,
+      "learning_rate": 4.9818415342888395e-06,
+      "loss": 1.1649,
+      "step": 543
+    },
+    {
+      "epoch": 0.28036419859130735,
+      "grad_norm": 1.1736465692520142,
+      "learning_rate": 4.981759562179611e-06,
+      "loss": 1.1378,
+      "step": 544
+    },
+    {
+      "epoch": 0.28087957395636487,
+      "grad_norm": 1.273497462272644,
+      "learning_rate": 4.981677406142391e-06,
+      "loss": 1.1159,
+      "step": 545
+    },
+    {
+      "epoch": 0.28139494932142245,
+      "grad_norm": 1.2971827983856201,
+      "learning_rate": 4.981595066183268e-06,
+      "loss": 1.1851,
+      "step": 546
+    },
+    {
+      "epoch": 0.28191032468648,
+      "grad_norm": 1.309173822402954,
+      "learning_rate": 4.981512542308344e-06,
+      "loss": 1.1604,
+      "step": 547
+    },
+    {
+      "epoch": 0.28242570005153755,
+      "grad_norm": 1.2924010753631592,
+      "learning_rate": 4.9814298345237364e-06,
+      "loss": 1.1017,
+      "step": 548
+    },
+    {
+      "epoch": 0.2829410754165951,
+      "grad_norm": 1.2045514583587646,
+      "learning_rate": 4.981346942835573e-06,
+      "loss": 1.1712,
+      "step": 549
+    },
+    {
+      "epoch": 0.28345645078165266,
+      "grad_norm": 1.2449471950531006,
+      "learning_rate": 4.981263867249998e-06,
+      "loss": 1.1483,
+      "step": 550
+    },
+    {
+      "epoch": 0.2839718261467102,
+      "grad_norm": 1.2177159786224365,
+      "learning_rate": 4.981180607773167e-06,
+      "loss": 1.1609,
+      "step": 551
+    },
+    {
+      "epoch": 0.28448720151176776,
+      "grad_norm": 1.238996982574463,
+      "learning_rate": 4.981097164411254e-06,
+      "loss": 1.1247,
+      "step": 552
+    },
+    {
+      "epoch": 0.2850025768768253,
+      "grad_norm": 1.233088731765747,
+      "learning_rate": 4.98101353717044e-06,
+      "loss": 1.1483,
+      "step": 553
+    },
+    {
+      "epoch": 0.2855179522418828,
+      "grad_norm": 1.1626131534576416,
+      "learning_rate": 4.980929726056922e-06,
+      "loss": 1.1444,
+      "step": 554
+    },
+    {
+      "epoch": 0.2860333276069404,
+      "grad_norm": 1.193416714668274,
+      "learning_rate": 4.980845731076915e-06,
+      "loss": 1.1958,
+      "step": 555
+    },
+    {
+      "epoch": 0.2865487029719979,
+      "grad_norm": 1.2389411926269531,
+      "learning_rate": 4.980761552236641e-06,
+      "loss": 1.1193,
+      "step": 556
+    },
+    {
+      "epoch": 0.2870640783370555,
+      "grad_norm": 1.2497589588165283,
+      "learning_rate": 4.98067718954234e-06,
+      "loss": 1.1505,
+      "step": 557
+    },
+    {
+      "epoch": 0.287579453702113,
+      "grad_norm": 1.3262685537338257,
+      "learning_rate": 4.980592643000264e-06,
+      "loss": 1.1793,
+      "step": 558
+    },
+    {
+      "epoch": 0.2880948290671706,
+      "grad_norm": 1.2589205503463745,
+      "learning_rate": 4.980507912616678e-06,
+      "loss": 1.1485,
+      "step": 559
+    },
+    {
+      "epoch": 0.2886102044322281,
+      "grad_norm": 1.1957213878631592,
+      "learning_rate": 4.980422998397864e-06,
+      "loss": 1.1295,
+      "step": 560
+    },
+    {
+      "epoch": 0.2891255797972857,
+      "grad_norm": 1.1982243061065674,
+      "learning_rate": 4.980337900350112e-06,
+      "loss": 1.1534,
+      "step": 561
+    },
+    {
+      "epoch": 0.2896409551623432,
+      "grad_norm": 1.2490700483322144,
+      "learning_rate": 4.980252618479731e-06,
+      "loss": 1.1792,
+      "step": 562
+    },
+    {
+      "epoch": 0.2901563305274008,
+      "grad_norm": 1.2109037637710571,
+      "learning_rate": 4.980167152793041e-06,
+      "loss": 1.0959,
+      "step": 563
+    },
+    {
+      "epoch": 0.29067170589245833,
+      "grad_norm": 1.177827000617981,
+      "learning_rate": 4.9800815032963755e-06,
+      "loss": 1.1518,
+      "step": 564
+    },
+    {
+      "epoch": 0.2911870812575159,
+      "grad_norm": 1.2427681684494019,
+      "learning_rate": 4.979995669996083e-06,
+      "loss": 1.1479,
+      "step": 565
+    },
+    {
+      "epoch": 0.29170245662257344,
+      "grad_norm": 1.17928147315979,
+      "learning_rate": 4.9799096528985235e-06,
+      "loss": 1.1189,
+      "step": 566
+    },
+    {
+      "epoch": 0.292217831987631,
+      "grad_norm": 1.31520676612854,
+      "learning_rate": 4.979823452010073e-06,
+      "loss": 1.1245,
+      "step": 567
+    },
+    {
+      "epoch": 0.29273320735268854,
+      "grad_norm": 1.3166396617889404,
+      "learning_rate": 4.979737067337119e-06,
+      "loss": 1.1499,
+      "step": 568
+    },
+    {
+      "epoch": 0.29324858271774606,
+      "grad_norm": 1.226168155670166,
+      "learning_rate": 4.979650498886065e-06,
+      "loss": 1.1441,
+      "step": 569
+    },
+    {
+      "epoch": 0.29376395808280364,
+      "grad_norm": 1.2373491525650024,
+      "learning_rate": 4.979563746663326e-06,
+      "loss": 1.1841,
+      "step": 570
+    },
+    {
+      "epoch": 0.29427933344786117,
+      "grad_norm": 1.2169537544250488,
+      "learning_rate": 4.979476810675331e-06,
+      "loss": 1.1627,
+      "step": 571
+    },
+    {
+      "epoch": 0.29479470881291875,
+      "grad_norm": 1.20741605758667,
+      "learning_rate": 4.979389690928524e-06,
+      "loss": 1.1586,
+      "step": 572
+    },
+    {
+      "epoch": 0.2953100841779763,
+      "grad_norm": 1.2474502325057983,
+      "learning_rate": 4.97930238742936e-06,
+      "loss": 1.1497,
+      "step": 573
+    },
+    {
+      "epoch": 0.29582545954303385,
+      "grad_norm": 1.2915810346603394,
+      "learning_rate": 4.979214900184311e-06,
+      "loss": 1.1545,
+      "step": 574
+    },
+    {
+      "epoch": 0.2963408349080914,
+      "grad_norm": 1.136528491973877,
+      "learning_rate": 4.97912722919986e-06,
+      "loss": 1.1371,
+      "step": 575
+    },
+    {
+      "epoch": 0.29685621027314896,
+      "grad_norm": 1.2052608728408813,
+      "learning_rate": 4.979039374482504e-06,
+      "loss": 1.1598,
+      "step": 576
+    },
+    {
+      "epoch": 0.2973715856382065,
+      "grad_norm": 1.2963969707489014,
+      "learning_rate": 4.978951336038756e-06,
+      "loss": 1.1278,
+      "step": 577
+    },
+    {
+      "epoch": 0.29788696100326406,
+      "grad_norm": 1.3299596309661865,
+      "learning_rate": 4.978863113875138e-06,
+      "loss": 1.1455,
+      "step": 578
+    },
+    {
+      "epoch": 0.2984023363683216,
+      "grad_norm": 1.178487777709961,
+      "learning_rate": 4.978774707998189e-06,
+      "loss": 1.1572,
+      "step": 579
+    },
+    {
+      "epoch": 0.29891771173337917,
+      "grad_norm": 1.4150831699371338,
+      "learning_rate": 4.978686118414463e-06,
+      "loss": 1.2016,
+      "step": 580
+    },
+    {
+      "epoch": 0.2994330870984367,
+      "grad_norm": 1.1844868659973145,
+      "learning_rate": 4.978597345130523e-06,
+      "loss": 1.1767,
+      "step": 581
+    },
+    {
+      "epoch": 0.29994846246349427,
+      "grad_norm": 1.2888526916503906,
+      "learning_rate": 4.97850838815295e-06,
+      "loss": 1.1512,
+      "step": 582
+    },
+    {
+      "epoch": 0.3004638378285518,
+      "grad_norm": 1.3352088928222656,
+      "learning_rate": 4.978419247488335e-06,
+      "loss": 1.1527,
+      "step": 583
+    },
+    {
+      "epoch": 0.3009792131936093,
+      "grad_norm": 1.26283860206604,
+      "learning_rate": 4.978329923143287e-06,
+      "loss": 1.1724,
+      "step": 584
+    },
+    {
+      "epoch": 0.3014945885586669,
+      "grad_norm": 1.182499885559082,
+      "learning_rate": 4.978240415124424e-06,
+      "loss": 1.0651,
+      "step": 585
+    },
+    {
+      "epoch": 0.3020099639237244,
+      "grad_norm": 1.3811864852905273,
+      "learning_rate": 4.9781507234383795e-06,
+      "loss": 1.0996,
+      "step": 586
+    },
+    {
+      "epoch": 0.302525339288782,
+      "grad_norm": 1.2513461112976074,
+      "learning_rate": 4.9780608480918015e-06,
+      "loss": 1.222,
+      "step": 587
+    },
+    {
+      "epoch": 0.3030407146538395,
+      "grad_norm": 1.345369815826416,
+      "learning_rate": 4.977970789091349e-06,
+      "loss": 1.1866,
+      "step": 588
+    },
+    {
+      "epoch": 0.3035560900188971,
+      "grad_norm": 1.1981523036956787,
+      "learning_rate": 4.977880546443701e-06,
+      "loss": 1.0942,
+      "step": 589
+    },
+    {
+      "epoch": 0.30407146538395463,
+      "grad_norm": 1.2694369554519653,
+      "learning_rate": 4.97779012015554e-06,
+      "loss": 1.1449,
+      "step": 590
+    },
+    {
+      "epoch": 0.3045868407490122,
+      "grad_norm": 1.249166488647461,
+      "learning_rate": 4.977699510233573e-06,
+      "loss": 1.1193,
+      "step": 591
+    },
+    {
+      "epoch": 0.30510221611406974,
+      "grad_norm": 1.222267746925354,
+      "learning_rate": 4.97760871668451e-06,
+      "loss": 1.1391,
+      "step": 592
+    },
+    {
+      "epoch": 0.3056175914791273,
+      "grad_norm": 1.3113726377487183,
+      "learning_rate": 4.977517739515083e-06,
+      "loss": 1.1619,
+      "step": 593
+    },
+    {
+      "epoch": 0.30613296684418484,
+      "grad_norm": 1.276493787765503,
+      "learning_rate": 4.977426578732035e-06,
+      "loss": 1.1728,
+      "step": 594
+    },
+    {
+      "epoch": 0.3066483422092424,
+      "grad_norm": 1.23202383518219,
+      "learning_rate": 4.97733523434212e-06,
+      "loss": 1.0695,
+      "step": 595
+    },
+    {
+      "epoch": 0.30716371757429994,
+      "grad_norm": 1.2622519731521606,
+      "learning_rate": 4.977243706352109e-06,
+      "loss": 1.1807,
+      "step": 596
+    },
+    {
+      "epoch": 0.3076790929393575,
+      "grad_norm": 1.2572592496871948,
+      "learning_rate": 4.977151994768785e-06,
+      "loss": 1.1816,
+      "step": 597
+    },
+    {
+      "epoch": 0.30819446830441505,
+      "grad_norm": 1.232765555381775,
+      "learning_rate": 4.977060099598945e-06,
+      "loss": 1.1402,
+      "step": 598
+    },
+    {
+      "epoch": 0.3087098436694726,
+      "grad_norm": 1.2640727758407593,
+      "learning_rate": 4.9769680208494e-06,
+      "loss": 1.1249,
+      "step": 599
+    },
+    {
+      "epoch": 0.30922521903453015,
+      "grad_norm": 1.138247013092041,
+      "learning_rate": 4.976875758526973e-06,
+      "loss": 1.1192,
+      "step": 600
+    },
+    {
+      "epoch": 0.3097405943995877,
+      "grad_norm": 1.3037362098693848,
+      "learning_rate": 4.976783312638503e-06,
+      "loss": 1.2068,
+      "step": 601
+    },
+    {
+      "epoch": 0.31025596976464526,
+      "grad_norm": 1.187371850013733,
+      "learning_rate": 4.976690683190839e-06,
+      "loss": 1.1223,
+      "step": 602
+    },
+    {
+      "epoch": 0.3107713451297028,
+      "grad_norm": 1.3030028343200684,
+      "learning_rate": 4.976597870190849e-06,
+      "loss": 1.1401,
+      "step": 603
+    },
+    {
+      "epoch": 0.31128672049476036,
+      "grad_norm": 1.1680561304092407,
+      "learning_rate": 4.976504873645411e-06,
+      "loss": 1.1077,
+      "step": 604
+    },
+    {
+      "epoch": 0.3118020958598179,
+      "grad_norm": 1.2486811876296997,
+      "learning_rate": 4.9764116935614146e-06,
+      "loss": 1.1878,
+      "step": 605
+    },
+    {
+      "epoch": 0.31231747122487546,
+      "grad_norm": 1.2635382413864136,
+      "learning_rate": 4.976318329945768e-06,
+      "loss": 1.1504,
+      "step": 606
+    },
+    {
+      "epoch": 0.312832846589933,
+      "grad_norm": 1.1843292713165283,
+      "learning_rate": 4.976224782805391e-06,
+      "loss": 1.2165,
+      "step": 607
+    },
+    {
+      "epoch": 0.31334822195499057,
+      "grad_norm": 1.2353806495666504,
+      "learning_rate": 4.9761310521472136e-06,
+      "loss": 1.0973,
+      "step": 608
+    },
+    {
+      "epoch": 0.3138635973200481,
+      "grad_norm": 1.2707849740982056,
+      "learning_rate": 4.976037137978186e-06,
+      "loss": 1.2043,
+      "step": 609
+    },
+    {
+      "epoch": 0.3143789726851057,
+      "grad_norm": 1.1960488557815552,
+      "learning_rate": 4.975943040305266e-06,
+      "loss": 1.1002,
+      "step": 610
+    },
+    {
+      "epoch": 0.3148943480501632,
+      "grad_norm": 1.2896122932434082,
+      "learning_rate": 4.975848759135427e-06,
+      "loss": 1.1557,
+      "step": 611
+    },
+    {
+      "epoch": 0.3154097234152208,
+      "grad_norm": 1.3175156116485596,
+      "learning_rate": 4.9757542944756576e-06,
+      "loss": 1.1808,
+      "step": 612
+    },
+    {
+      "epoch": 0.3159250987802783,
+      "grad_norm": 1.284606695175171,
+      "learning_rate": 4.9756596463329585e-06,
+      "loss": 1.1451,
+      "step": 613
+    },
+    {
+      "epoch": 0.3164404741453358,
+      "grad_norm": 1.1585352420806885,
+      "learning_rate": 4.975564814714345e-06,
+      "loss": 1.1862,
+      "step": 614
+    },
+    {
+      "epoch": 0.3169558495103934,
+      "grad_norm": 1.1436694860458374,
+      "learning_rate": 4.975469799626844e-06,
+      "loss": 1.0628,
+      "step": 615
+    },
+    {
+      "epoch": 0.31747122487545093,
+      "grad_norm": 1.1853829622268677,
+      "learning_rate": 4.975374601077497e-06,
+      "loss": 1.1256,
+      "step": 616
+    },
+    {
+      "epoch": 0.3179866002405085,
+      "grad_norm": 1.2337476015090942,
+      "learning_rate": 4.97527921907336e-06,
+      "loss": 1.1067,
+      "step": 617
+    },
+    {
+      "epoch": 0.31850197560556603,
+      "grad_norm": 1.2059632539749146,
+      "learning_rate": 4.975183653621503e-06,
+      "loss": 1.1379,
+      "step": 618
+    },
+    {
+      "epoch": 0.3190173509706236,
+      "grad_norm": 1.193901777267456,
+      "learning_rate": 4.975087904729007e-06,
+      "loss": 1.1349,
+      "step": 619
+    },
+    {
+      "epoch": 0.31953272633568114,
+      "grad_norm": 1.285905361175537,
+      "learning_rate": 4.974991972402969e-06,
+      "loss": 1.1368,
+      "step": 620
+    },
+    {
+      "epoch": 0.3200481017007387,
+      "grad_norm": 1.2497121095657349,
+      "learning_rate": 4.974895856650498e-06,
+      "loss": 1.1165,
+      "step": 621
+    },
+    {
+      "epoch": 0.32056347706579624,
+      "grad_norm": 1.208838939666748,
+      "learning_rate": 4.974799557478718e-06,
+      "loss": 1.1446,
+      "step": 622
+    },
+    {
+      "epoch": 0.3210788524308538,
+      "grad_norm": 1.3200178146362305,
+      "learning_rate": 4.974703074894766e-06,
+      "loss": 1.1881,
+      "step": 623
+    },
+    {
+      "epoch": 0.32159422779591135,
+      "grad_norm": 1.2912616729736328,
+      "learning_rate": 4.974606408905791e-06,
+      "loss": 1.1516,
+      "step": 624
+    },
+    {
+      "epoch": 0.3221096031609689,
+      "grad_norm": 1.2733863592147827,
+      "learning_rate": 4.9745095595189585e-06,
+      "loss": 1.1073,
+      "step": 625
+    },
+    {
+      "epoch": 0.32262497852602645,
+      "grad_norm": 1.2568449974060059,
+      "learning_rate": 4.974412526741447e-06,
+      "loss": 1.1035,
+      "step": 626
+    },
+    {
+      "epoch": 0.32314035389108403,
+      "grad_norm": 1.3438385725021362,
+      "learning_rate": 4.974315310580445e-06,
+      "loss": 1.1347,
+      "step": 627
+    },
+    {
+      "epoch": 0.32365572925614156,
+      "grad_norm": 1.2689323425292969,
+      "learning_rate": 4.974217911043161e-06,
+      "loss": 1.16,
+      "step": 628
+    },
+    {
+      "epoch": 0.3241711046211991,
+      "grad_norm": 1.2674888372421265,
+      "learning_rate": 4.97412032813681e-06,
+      "loss": 1.1229,
+      "step": 629
+    },
+    {
+      "epoch": 0.32468647998625666,
+      "grad_norm": 1.2539578676223755,
+      "learning_rate": 4.9740225618686265e-06,
+      "loss": 1.1325,
+      "step": 630
+    },
+    {
+      "epoch": 0.3252018553513142,
+      "grad_norm": 1.2101805210113525,
+      "learning_rate": 4.973924612245855e-06,
+      "loss": 1.126,
+      "step": 631
+    },
+    {
+      "epoch": 0.32571723071637176,
+      "grad_norm": 1.2297779321670532,
+      "learning_rate": 4.9738264792757555e-06,
+      "loss": 1.1918,
+      "step": 632
+    },
+    {
+      "epoch": 0.3262326060814293,
+      "grad_norm": 1.3601030111312866,
+      "learning_rate": 4.9737281629655996e-06,
+      "loss": 1.1677,
+      "step": 633
+    },
+    {
+      "epoch": 0.32674798144648687,
+      "grad_norm": 1.2321043014526367,
+      "learning_rate": 4.973629663322675e-06,
+      "loss": 1.1329,
+      "step": 634
+    },
+    {
+      "epoch": 0.3272633568115444,
+      "grad_norm": 1.2176779508590698,
+      "learning_rate": 4.973530980354281e-06,
+      "loss": 1.1136,
+      "step": 635
+    },
+    {
+      "epoch": 0.327778732176602,
+      "grad_norm": 1.2674428224563599,
+      "learning_rate": 4.973432114067731e-06,
+      "loss": 1.126,
+      "step": 636
+    },
+    {
+      "epoch": 0.3282941075416595,
+      "grad_norm": 1.1574589014053345,
+      "learning_rate": 4.973333064470352e-06,
+      "loss": 1.1402,
+      "step": 637
+    },
+    {
+      "epoch": 0.3288094829067171,
+      "grad_norm": 1.230744481086731,
+      "learning_rate": 4.973233831569487e-06,
+      "loss": 1.1292,
+      "step": 638
+    },
+    {
+      "epoch": 0.3293248582717746,
+      "grad_norm": 1.3396669626235962,
+      "learning_rate": 4.973134415372488e-06,
+      "loss": 1.1737,
+      "step": 639
+    },
+    {
+      "epoch": 0.3298402336368322,
+      "grad_norm": 1.2492640018463135,
+      "learning_rate": 4.973034815886723e-06,
+      "loss": 1.0871,
+      "step": 640
+    },
+    {
+      "epoch": 0.3303556090018897,
+      "grad_norm": 1.348250150680542,
+      "learning_rate": 4.9729350331195746e-06,
+      "loss": 1.1464,
+      "step": 641
+    },
+    {
+      "epoch": 0.3308709843669473,
+      "grad_norm": 1.226974606513977,
+      "learning_rate": 4.972835067078436e-06,
+      "loss": 1.1298,
+      "step": 642
+    },
+    {
+      "epoch": 0.3313863597320048,
+      "grad_norm": 1.2675280570983887,
+      "learning_rate": 4.972734917770719e-06,
+      "loss": 1.0708,
+      "step": 643
+    },
+    {
+      "epoch": 0.33190173509706233,
+      "grad_norm": 1.2955893278121948,
+      "learning_rate": 4.972634585203843e-06,
+      "loss": 1.0924,
+      "step": 644
+    },
+    {
+      "epoch": 0.3324171104621199,
+      "grad_norm": 1.290199875831604,
+      "learning_rate": 4.972534069385245e-06,
+      "loss": 1.1479,
+      "step": 645
+    },
+    {
+      "epoch": 0.33293248582717744,
+      "grad_norm": 1.3217016458511353,
+      "learning_rate": 4.9724333703223746e-06,
+      "loss": 1.1328,
+      "step": 646
+    },
+    {
+      "epoch": 0.333447861192235,
+      "grad_norm": 1.355234980583191,
+      "learning_rate": 4.9723324880226945e-06,
+      "loss": 1.1553,
+      "step": 647
+    },
+    {
+      "epoch": 0.33396323655729254,
+      "grad_norm": 1.3599210977554321,
+      "learning_rate": 4.972231422493681e-06,
+      "loss": 1.1859,
+      "step": 648
+    },
+    {
+      "epoch": 0.3344786119223501,
+      "grad_norm": 1.2493696212768555,
+      "learning_rate": 4.972130173742825e-06,
+      "loss": 1.1226,
+      "step": 649
+    },
+    {
+      "epoch": 0.33499398728740765,
+      "grad_norm": 1.2313398122787476,
+      "learning_rate": 4.9720287417776296e-06,
+      "loss": 1.1063,
+      "step": 650
+    },
+    {
+      "epoch": 0.3355093626524652,
+      "grad_norm": 1.299560546875,
+      "learning_rate": 4.971927126605613e-06,
+      "loss": 1.1341,
+      "step": 651
+    },
+    {
+      "epoch": 0.33602473801752275,
+      "grad_norm": 1.2997859716415405,
+      "learning_rate": 4.9718253282343045e-06,
+      "loss": 1.1745,
+      "step": 652
+    },
+    {
+      "epoch": 0.33654011338258033,
+      "grad_norm": 1.2912580966949463,
+      "learning_rate": 4.971723346671249e-06,
+      "loss": 1.1192,
+      "step": 653
+    },
+    {
+      "epoch": 0.33705548874763785,
+      "grad_norm": 1.2521673440933228,
+      "learning_rate": 4.971621181924006e-06,
+      "loss": 1.1352,
+      "step": 654
+    },
+    {
+      "epoch": 0.33757086411269543,
+      "grad_norm": 1.1946815252304077,
+      "learning_rate": 4.971518834000147e-06,
+      "loss": 1.1158,
+      "step": 655
+    },
+    {
+      "epoch": 0.33808623947775296,
+      "grad_norm": 1.3462064266204834,
+      "learning_rate": 4.971416302907255e-06,
+      "loss": 1.1664,
+      "step": 656
+    },
+    {
+      "epoch": 0.33860161484281054,
+      "grad_norm": 1.317471981048584,
+      "learning_rate": 4.97131358865293e-06,
+      "loss": 1.1294,
+      "step": 657
+    },
+    {
+      "epoch": 0.33911699020786806,
+      "grad_norm": 1.2769732475280762,
+      "learning_rate": 4.971210691244786e-06,
+      "loss": 1.1233,
+      "step": 658
+    },
+    {
+      "epoch": 0.3396323655729256,
+      "grad_norm": 1.2474626302719116,
+      "learning_rate": 4.971107610690445e-06,
+      "loss": 1.1829,
+      "step": 659
+    },
+    {
+      "epoch": 0.34014774093798317,
+      "grad_norm": 1.2211552858352661,
+      "learning_rate": 4.971004346997551e-06,
+      "loss": 1.135,
+      "step": 660
+    },
+    {
+      "epoch": 0.3406631163030407,
+      "grad_norm": 1.2152830362319946,
+      "learning_rate": 4.970900900173755e-06,
+      "loss": 1.1382,
+      "step": 661
+    },
+    {
+      "epoch": 0.34117849166809827,
+      "grad_norm": 1.2274240255355835,
+      "learning_rate": 4.970797270226722e-06,
+      "loss": 1.1046,
+      "step": 662
+    },
+    {
+      "epoch": 0.3416938670331558,
+      "grad_norm": 1.194387435913086,
+      "learning_rate": 4.9706934571641355e-06,
+      "loss": 1.0675,
+      "step": 663
+    },
+    {
+      "epoch": 0.3422092423982134,
+      "grad_norm": 1.2856234312057495,
+      "learning_rate": 4.970589460993687e-06,
+      "loss": 1.1626,
+      "step": 664
+    },
+    {
+      "epoch": 0.3427246177632709,
+      "grad_norm": 1.327237606048584,
+      "learning_rate": 4.970485281723085e-06,
+      "loss": 1.1075,
+      "step": 665
+    },
+    {
+      "epoch": 0.3432399931283285,
+      "grad_norm": 1.2260476350784302,
+      "learning_rate": 4.97038091936005e-06,
+      "loss": 1.1481,
+      "step": 666
+    },
+    {
+      "epoch": 0.343755368493386,
+      "grad_norm": 1.3910338878631592,
+      "learning_rate": 4.970276373912316e-06,
+      "loss": 1.183,
+      "step": 667
+    },
+    {
+      "epoch": 0.3442707438584436,
+      "grad_norm": 1.2876691818237305,
+      "learning_rate": 4.970171645387632e-06,
+      "loss": 1.1146,
+      "step": 668
+    },
+    {
+      "epoch": 0.3447861192235011,
+      "grad_norm": 1.2745333909988403,
+      "learning_rate": 4.970066733793758e-06,
+      "loss": 1.2011,
+      "step": 669
+    },
+    {
+      "epoch": 0.3453014945885587,
+      "grad_norm": 1.3226712942123413,
+      "learning_rate": 4.9699616391384705e-06,
+      "loss": 1.1331,
+      "step": 670
+    },
+    {
+      "epoch": 0.3458168699536162,
+      "grad_norm": 1.2185267210006714,
+      "learning_rate": 4.969856361429559e-06,
+      "loss": 1.168,
+      "step": 671
+    },
+    {
+      "epoch": 0.3463322453186738,
+      "grad_norm": 1.243064045906067,
+      "learning_rate": 4.969750900674825e-06,
+      "loss": 1.126,
+      "step": 672
+    },
+    {
+      "epoch": 0.3468476206837313,
+      "grad_norm": 1.2364131212234497,
+      "learning_rate": 4.969645256882083e-06,
+      "loss": 1.119,
+      "step": 673
+    },
+    {
+      "epoch": 0.34736299604878884,
+      "grad_norm": 1.3395003080368042,
+      "learning_rate": 4.969539430059165e-06,
+      "loss": 1.0998,
+      "step": 674
+    },
+    {
+      "epoch": 0.3478783714138464,
+      "grad_norm": 1.3111714124679565,
+      "learning_rate": 4.969433420213912e-06,
+      "loss": 1.1397,
+      "step": 675
+    },
+    {
+      "epoch": 0.34839374677890395,
+      "grad_norm": 1.2934197187423706,
+      "learning_rate": 4.969327227354182e-06,
+      "loss": 1.1013,
+      "step": 676
+    },
+    {
+      "epoch": 0.3489091221439615,
+      "grad_norm": 1.1807507276535034,
+      "learning_rate": 4.9692208514878445e-06,
+      "loss": 1.0706,
+      "step": 677
+    },
+    {
+      "epoch": 0.34942449750901905,
+      "grad_norm": 1.1788406372070312,
+      "learning_rate": 4.969114292622783e-06,
+      "loss": 1.1055,
+      "step": 678
+    },
+    {
+      "epoch": 0.34993987287407663,
+      "grad_norm": 1.3066070079803467,
+      "learning_rate": 4.969007550766896e-06,
+      "loss": 1.174,
+      "step": 679
+    },
+    {
+      "epoch": 0.35045524823913415,
+      "grad_norm": 1.3041435480117798,
+      "learning_rate": 4.968900625928092e-06,
+      "loss": 1.1639,
+      "step": 680
+    },
+    {
+      "epoch": 0.35097062360419173,
+      "grad_norm": 1.1880706548690796,
+      "learning_rate": 4.968793518114299e-06,
+      "loss": 1.1309,
+      "step": 681
+    },
+    {
+      "epoch": 0.35148599896924926,
+      "grad_norm": 1.2611826658248901,
+      "learning_rate": 4.968686227333451e-06,
+      "loss": 1.1353,
+      "step": 682
+    },
+    {
+      "epoch": 0.35200137433430684,
+      "grad_norm": 1.2815132141113281,
+      "learning_rate": 4.968578753593501e-06,
+      "loss": 1.1029,
+      "step": 683
+    },
+    {
+      "epoch": 0.35251674969936436,
+      "grad_norm": 1.3178620338439941,
+      "learning_rate": 4.9684710969024155e-06,
+      "loss": 1.1045,
+      "step": 684
+    },
+    {
+      "epoch": 0.35303212506442194,
+      "grad_norm": 1.3577544689178467,
+      "learning_rate": 4.968363257268172e-06,
+      "loss": 1.1487,
+      "step": 685
+    },
+    {
+      "epoch": 0.35354750042947947,
+      "grad_norm": 1.3682572841644287,
+      "learning_rate": 4.968255234698762e-06,
+      "loss": 1.1279,
+      "step": 686
+    },
+    {
+      "epoch": 0.35406287579453705,
+      "grad_norm": 1.2696278095245361,
+      "learning_rate": 4.968147029202193e-06,
+      "loss": 1.1658,
+      "step": 687
+    },
+    {
+      "epoch": 0.35457825115959457,
+      "grad_norm": 1.285667896270752,
+      "learning_rate": 4.968038640786484e-06,
+      "loss": 1.1614,
+      "step": 688
+    },
+    {
+      "epoch": 0.3550936265246521,
+      "grad_norm": 1.2449527978897095,
+      "learning_rate": 4.967930069459665e-06,
+      "loss": 1.1607,
+      "step": 689
+    },
+    {
+      "epoch": 0.3556090018897097,
+      "grad_norm": 1.3253920078277588,
+      "learning_rate": 4.967821315229786e-06,
+      "loss": 1.1755,
+      "step": 690
+    },
+    {
+      "epoch": 0.3561243772547672,
+      "grad_norm": 1.2488442659378052,
+      "learning_rate": 4.9677123781049056e-06,
+      "loss": 1.1352,
+      "step": 691
+    },
+    {
+      "epoch": 0.3566397526198248,
+      "grad_norm": 1.1993600130081177,
+      "learning_rate": 4.967603258093097e-06,
+      "loss": 1.1264,
+      "step": 692
+    },
+    {
+      "epoch": 0.3571551279848823,
+      "grad_norm": 1.3461339473724365,
+      "learning_rate": 4.967493955202448e-06,
+      "loss": 1.1764,
+      "step": 693
+    },
+    {
+      "epoch": 0.3576705033499399,
+      "grad_norm": 1.3022563457489014,
+      "learning_rate": 4.967384469441059e-06,
+      "loss": 1.1225,
+      "step": 694
+    },
+    {
+      "epoch": 0.3581858787149974,
+      "grad_norm": 1.2400002479553223,
+      "learning_rate": 4.967274800817044e-06,
+      "loss": 1.1201,
+      "step": 695
+    },
+    {
+      "epoch": 0.358701254080055,
+      "grad_norm": 1.3111238479614258,
+      "learning_rate": 4.9671649493385306e-06,
+      "loss": 1.1748,
+      "step": 696
+    },
+    {
+      "epoch": 0.3592166294451125,
+      "grad_norm": 1.2304924726486206,
+      "learning_rate": 4.96705491501366e-06,
+      "loss": 1.1069,
+      "step": 697
+    },
+    {
+      "epoch": 0.3597320048101701,
+      "grad_norm": 1.2275941371917725,
+      "learning_rate": 4.9669446978505884e-06,
+      "loss": 1.1006,
+      "step": 698
+    },
+    {
+      "epoch": 0.3602473801752276,
+      "grad_norm": 1.2485218048095703,
+      "learning_rate": 4.966834297857483e-06,
+      "loss": 1.2003,
+      "step": 699
+    },
+    {
+      "epoch": 0.3607627555402852,
+      "grad_norm": 1.2599111795425415,
+      "learning_rate": 4.966723715042526e-06,
+      "loss": 1.1177,
+      "step": 700
+    },
+    {
+      "epoch": 0.3612781309053427,
+      "grad_norm": 1.2331079244613647,
+      "learning_rate": 4.966612949413912e-06,
+      "loss": 1.1089,
+      "step": 701
+    },
+    {
+      "epoch": 0.3617935062704003,
+      "grad_norm": 1.2413543462753296,
+      "learning_rate": 4.966502000979852e-06,
+      "loss": 1.1406,
+      "step": 702
+    },
+    {
+      "epoch": 0.3623088816354578,
+      "grad_norm": 1.155291199684143,
+      "learning_rate": 4.966390869748567e-06,
+      "loss": 1.0574,
+      "step": 703
+    },
+    {
+      "epoch": 0.36282425700051535,
+      "grad_norm": 1.1831265687942505,
+      "learning_rate": 4.966279555728295e-06,
+      "loss": 1.1267,
+      "step": 704
+    },
+    {
+      "epoch": 0.36333963236557293,
+      "grad_norm": 1.2939788103103638,
+      "learning_rate": 4.966168058927283e-06,
+      "loss": 1.0658,
+      "step": 705
+    },
+    {
+      "epoch": 0.36385500773063045,
+      "grad_norm": 1.2909181118011475,
+      "learning_rate": 4.966056379353796e-06,
+      "loss": 1.1176,
+      "step": 706
+    },
+    {
+      "epoch": 0.36437038309568803,
+      "grad_norm": 1.2729825973510742,
+      "learning_rate": 4.96594451701611e-06,
+      "loss": 1.1081,
+      "step": 707
+    },
+    {
+      "epoch": 0.36488575846074556,
+      "grad_norm": 1.2541344165802002,
+      "learning_rate": 4.965832471922517e-06,
+      "loss": 1.1312,
+      "step": 708
+    },
+    {
+      "epoch": 0.36540113382580314,
+      "grad_norm": 1.3798422813415527,
+      "learning_rate": 4.965720244081319e-06,
+      "loss": 1.2037,
+      "step": 709
+    },
+    {
+      "epoch": 0.36591650919086066,
+      "grad_norm": 1.2447426319122314,
+      "learning_rate": 4.965607833500834e-06,
+      "loss": 1.1166,
+      "step": 710
+    },
+    {
+      "epoch": 0.36643188455591824,
+      "grad_norm": 1.215897560119629,
+      "learning_rate": 4.965495240189393e-06,
+      "loss": 1.1064,
+      "step": 711
+    },
+    {
+      "epoch": 0.36694725992097577,
+      "grad_norm": 1.2096561193466187,
+      "learning_rate": 4.965382464155341e-06,
+      "loss": 1.0328,
+      "step": 712
+    },
+    {
+      "epoch": 0.36746263528603335,
+      "grad_norm": 1.3168424367904663,
+      "learning_rate": 4.9652695054070344e-06,
+      "loss": 1.12,
+      "step": 713
+    },
+    {
+      "epoch": 0.36797801065109087,
+      "grad_norm": 1.1280862092971802,
+      "learning_rate": 4.9651563639528465e-06,
+      "loss": 1.0804,
+      "step": 714
+    },
+    {
+      "epoch": 0.36849338601614845,
+      "grad_norm": 1.3654303550720215,
+      "learning_rate": 4.9650430398011626e-06,
+      "loss": 1.1334,
+      "step": 715
+    },
+    {
+      "epoch": 0.369008761381206,
+      "grad_norm": 1.2694456577301025,
+      "learning_rate": 4.96492953296038e-06,
+      "loss": 1.1135,
+      "step": 716
+    },
+    {
+      "epoch": 0.36952413674626355,
+      "grad_norm": 1.2896363735198975,
+      "learning_rate": 4.9648158434389124e-06,
+      "loss": 1.152,
+      "step": 717
+    },
+    {
+      "epoch": 0.3700395121113211,
+      "grad_norm": 1.276693344116211,
+      "learning_rate": 4.9647019712451836e-06,
+      "loss": 1.1311,
+      "step": 718
+    },
+    {
+      "epoch": 0.3705548874763786,
+      "grad_norm": 1.407243013381958,
+      "learning_rate": 4.964587916387635e-06,
+      "loss": 1.1585,
+      "step": 719
+    },
+    {
+      "epoch": 0.3710702628414362,
+      "grad_norm": 1.2954100370407104,
+      "learning_rate": 4.964473678874718e-06,
+      "loss": 1.1195,
+      "step": 720
+    },
+    {
+      "epoch": 0.3715856382064937,
+      "grad_norm": 1.2955098152160645,
+      "learning_rate": 4.9643592587149005e-06,
+      "loss": 1.016,
+      "step": 721
+    },
+    {
+      "epoch": 0.3721010135715513,
+      "grad_norm": 1.2117847204208374,
+      "learning_rate": 4.9642446559166605e-06,
+      "loss": 1.1082,
+      "step": 722
+    },
+    {
+      "epoch": 0.3726163889366088,
+      "grad_norm": 1.2178875207901,
+      "learning_rate": 4.964129870488491e-06,
+      "loss": 1.1686,
+      "step": 723
+    },
+    {
+      "epoch": 0.3731317643016664,
+      "grad_norm": 1.27935791015625,
+      "learning_rate": 4.9640149024389014e-06,
+      "loss": 1.1193,
+      "step": 724
+    },
+    {
+      "epoch": 0.3736471396667239,
+      "grad_norm": 1.3828037977218628,
+      "learning_rate": 4.963899751776411e-06,
+      "loss": 1.1167,
+      "step": 725
+    },
+    {
+      "epoch": 0.3741625150317815,
+      "grad_norm": 1.3130462169647217,
+      "learning_rate": 4.9637844185095545e-06,
+      "loss": 1.1457,
+      "step": 726
+    },
+    {
+      "epoch": 0.374677890396839,
+      "grad_norm": 1.2902321815490723,
+      "learning_rate": 4.963668902646879e-06,
+      "loss": 1.2095,
+      "step": 727
+    },
+    {
+      "epoch": 0.3751932657618966,
+      "grad_norm": 1.273491382598877,
+      "learning_rate": 4.963553204196944e-06,
+      "loss": 1.1037,
+      "step": 728
+    },
+    {
+      "epoch": 0.3757086411269541,
+      "grad_norm": 1.422223687171936,
+      "learning_rate": 4.963437323168326e-06,
+      "loss": 1.2536,
+      "step": 729
+    },
+    {
+      "epoch": 0.3762240164920117,
+      "grad_norm": 1.3016620874404907,
+      "learning_rate": 4.963321259569613e-06,
+      "loss": 1.2464,
+      "step": 730
+    },
+    {
+      "epoch": 0.37673939185706923,
+      "grad_norm": 1.3318872451782227,
+      "learning_rate": 4.963205013409407e-06,
+      "loss": 1.1456,
+      "step": 731
+    },
+    {
+      "epoch": 0.3772547672221268,
+      "grad_norm": 1.3237916231155396,
+      "learning_rate": 4.963088584696322e-06,
+      "loss": 1.0936,
+      "step": 732
+    },
+    {
+      "epoch": 0.37777014258718433,
+      "grad_norm": 1.2826303243637085,
+      "learning_rate": 4.962971973438988e-06,
+      "loss": 1.1412,
+      "step": 733
+    },
+    {
+      "epoch": 0.37828551795224186,
+      "grad_norm": 1.254653811454773,
+      "learning_rate": 4.962855179646046e-06,
+      "loss": 1.1696,
+      "step": 734
+    },
+    {
+      "epoch": 0.37880089331729944,
+      "grad_norm": 1.294478416442871,
+      "learning_rate": 4.962738203326153e-06,
+      "loss": 1.1704,
+      "step": 735
+    },
+    {
+      "epoch": 0.37931626868235696,
+      "grad_norm": 1.2368558645248413,
+      "learning_rate": 4.962621044487978e-06,
+      "loss": 1.1011,
+      "step": 736
+    },
+    {
+      "epoch": 0.37983164404741454,
+      "grad_norm": 1.2399348020553589,
+      "learning_rate": 4.962503703140203e-06,
+      "loss": 1.1748,
+      "step": 737
+    },
+    {
+      "epoch": 0.38034701941247206,
+      "grad_norm": 1.3496227264404297,
+      "learning_rate": 4.962386179291526e-06,
+      "loss": 1.1774,
+      "step": 738
+    },
+    {
+      "epoch": 0.38086239477752964,
+      "grad_norm": 1.2116947174072266,
+      "learning_rate": 4.9622684729506555e-06,
+      "loss": 1.13,
+      "step": 739
+    },
+    {
+      "epoch": 0.38137777014258717,
+      "grad_norm": 1.3135943412780762,
+      "learning_rate": 4.9621505841263155e-06,
+      "loss": 1.172,
+      "step": 740
+    },
+    {
+      "epoch": 0.38189314550764475,
+      "grad_norm": 1.275046467781067,
+      "learning_rate": 4.962032512827243e-06,
+      "loss": 1.1179,
+      "step": 741
+    },
+    {
+      "epoch": 0.3824085208727023,
+      "grad_norm": 1.3962445259094238,
+      "learning_rate": 4.961914259062189e-06,
+      "loss": 1.1301,
+      "step": 742
+    },
+    {
+      "epoch": 0.38292389623775985,
+      "grad_norm": 1.3349716663360596,
+      "learning_rate": 4.961795822839917e-06,
+      "loss": 1.1139,
+      "step": 743
+    },
+    {
+      "epoch": 0.3834392716028174,
+      "grad_norm": 1.3125500679016113,
+      "learning_rate": 4.9616772041692035e-06,
+      "loss": 1.1209,
+      "step": 744
+    },
+    {
+      "epoch": 0.38395464696787496,
+      "grad_norm": 1.3276808261871338,
+      "learning_rate": 4.961558403058841e-06,
+      "loss": 1.115,
+      "step": 745
+    },
+    {
+      "epoch": 0.3844700223329325,
+      "grad_norm": 1.3487229347229004,
+      "learning_rate": 4.961439419517634e-06,
+      "loss": 1.1177,
+      "step": 746
+    },
+    {
+      "epoch": 0.38498539769799006,
+      "grad_norm": 1.262861967086792,
+      "learning_rate": 4.9613202535544e-06,
+      "loss": 1.1196,
+      "step": 747
+    },
+    {
+      "epoch": 0.3855007730630476,
+      "grad_norm": 1.347224473953247,
+      "learning_rate": 4.961200905177971e-06,
+      "loss": 1.1486,
+      "step": 748
+    },
+    {
+      "epoch": 0.3860161484281051,
+      "grad_norm": 1.3654824495315552,
+      "learning_rate": 4.961081374397192e-06,
+      "loss": 1.1499,
+      "step": 749
+    },
+    {
+      "epoch": 0.3865315237931627,
+      "grad_norm": 1.2931691408157349,
+      "learning_rate": 4.960961661220921e-06,
+      "loss": 1.1048,
+      "step": 750
+    },
+    {
+      "epoch": 0.3870468991582202,
+      "grad_norm": 1.3167107105255127,
+      "learning_rate": 4.960841765658032e-06,
+      "loss": 1.1344,
+      "step": 751
+    },
+    {
+      "epoch": 0.3875622745232778,
+      "grad_norm": 1.2615445852279663,
+      "learning_rate": 4.960721687717408e-06,
+      "loss": 1.0626,
+      "step": 752
+    },
+    {
+      "epoch": 0.3880776498883353,
+      "grad_norm": 1.3376338481903076,
+      "learning_rate": 4.960601427407952e-06,
+      "loss": 1.1224,
+      "step": 753
+    },
+    {
+      "epoch": 0.3885930252533929,
+      "grad_norm": 1.3246850967407227,
+      "learning_rate": 4.960480984738573e-06,
+      "loss": 1.1603,
+      "step": 754
+    },
+    {
+      "epoch": 0.3891084006184504,
+      "grad_norm": 1.2683501243591309,
+      "learning_rate": 4.960360359718198e-06,
+      "loss": 1.0908,
+      "step": 755
+    },
+    {
+      "epoch": 0.389623775983508,
+      "grad_norm": 1.259310007095337,
+      "learning_rate": 4.960239552355769e-06,
+      "loss": 1.141,
+      "step": 756
+    },
+    {
+      "epoch": 0.3901391513485655,
+      "grad_norm": 1.2660895586013794,
+      "learning_rate": 4.960118562660236e-06,
+      "loss": 1.1557,
+      "step": 757
+    },
+    {
+      "epoch": 0.3906545267136231,
+      "grad_norm": 1.3115184307098389,
+      "learning_rate": 4.959997390640569e-06,
+      "loss": 1.1707,
+      "step": 758
+    },
+    {
+      "epoch": 0.39116990207868063,
+      "grad_norm": 1.3106032609939575,
+      "learning_rate": 4.959876036305746e-06,
+      "loss": 1.1481,
+      "step": 759
+    },
+    {
+      "epoch": 0.3916852774437382,
+      "grad_norm": 1.3089869022369385,
+      "learning_rate": 4.959754499664762e-06,
+      "loss": 1.1248,
+      "step": 760
+    },
+    {
+      "epoch": 0.39220065280879574,
+      "grad_norm": 1.2966090440750122,
+      "learning_rate": 4.9596327807266245e-06,
+      "loss": 1.1252,
+      "step": 761
+    },
+    {
+      "epoch": 0.3927160281738533,
+      "grad_norm": 1.178955316543579,
+      "learning_rate": 4.959510879500352e-06,
+      "loss": 1.1103,
+      "step": 762
+    },
+    {
+      "epoch": 0.39323140353891084,
+      "grad_norm": 1.309419870376587,
+      "learning_rate": 4.959388795994982e-06,
+      "loss": 1.1676,
+      "step": 763
+    },
+    {
+      "epoch": 0.39374677890396836,
+      "grad_norm": 1.2507402896881104,
+      "learning_rate": 4.959266530219561e-06,
+      "loss": 1.1411,
+      "step": 764
+    },
+    {
+      "epoch": 0.39426215426902594,
+      "grad_norm": 1.3529011011123657,
+      "learning_rate": 4.959144082183149e-06,
+      "loss": 1.1412,
+      "step": 765
+    },
+    {
+      "epoch": 0.39477752963408347,
+      "grad_norm": 1.3554461002349854,
+      "learning_rate": 4.959021451894823e-06,
+      "loss": 1.0517,
+      "step": 766
+    },
+    {
+      "epoch": 0.39529290499914105,
+      "grad_norm": 1.2238177061080933,
+      "learning_rate": 4.958898639363671e-06,
+      "loss": 1.1235,
+      "step": 767
+    },
+    {
+      "epoch": 0.3958082803641986,
+      "grad_norm": 1.3070563077926636,
+      "learning_rate": 4.958775644598793e-06,
+      "loss": 1.1145,
+      "step": 768
+    },
+    {
+      "epoch": 0.39632365572925615,
+      "grad_norm": 1.3427777290344238,
+      "learning_rate": 4.958652467609307e-06,
+      "loss": 1.1272,
+      "step": 769
+    },
+    {
+      "epoch": 0.3968390310943137,
+      "grad_norm": 1.2721115350723267,
+      "learning_rate": 4.95852910840434e-06,
+      "loss": 1.1605,
+      "step": 770
+    },
+    {
+      "epoch": 0.39735440645937126,
+      "grad_norm": 1.2291545867919922,
+      "learning_rate": 4.958405566993035e-06,
+      "loss": 1.1183,
+      "step": 771
+    },
+    {
+      "epoch": 0.3978697818244288,
+      "grad_norm": 1.2380768060684204,
+      "learning_rate": 4.958281843384549e-06,
+      "loss": 1.1272,
+      "step": 772
+    },
+    {
+      "epoch": 0.39838515718948636,
+      "grad_norm": 1.2982110977172852,
+      "learning_rate": 4.958157937588049e-06,
+      "loss": 1.106,
+      "step": 773
+    },
+    {
+      "epoch": 0.3989005325545439,
+      "grad_norm": 1.3379300832748413,
+      "learning_rate": 4.958033849612719e-06,
+      "loss": 1.155,
+      "step": 774
+    },
+    {
+      "epoch": 0.39941590791960147,
+      "grad_norm": 1.3295038938522339,
+      "learning_rate": 4.957909579467757e-06,
+      "loss": 1.1161,
+      "step": 775
+    },
+    {
+      "epoch": 0.399931283284659,
+      "grad_norm": 1.2976667881011963,
+      "learning_rate": 4.9577851271623704e-06,
+      "loss": 1.1668,
+      "step": 776
+    },
+    {
+      "epoch": 0.40044665864971657,
+      "grad_norm": 1.2517850399017334,
+      "learning_rate": 4.9576604927057845e-06,
+      "loss": 1.0963,
+      "step": 777
+    },
+    {
+      "epoch": 0.4009620340147741,
+      "grad_norm": 1.291367530822754,
+      "learning_rate": 4.957535676107235e-06,
+      "loss": 1.1271,
+      "step": 778
+    },
+    {
+      "epoch": 0.4014774093798316,
+      "grad_norm": 1.269709587097168,
+      "learning_rate": 4.957410677375972e-06,
+      "loss": 1.0947,
+      "step": 779
+    },
+    {
+      "epoch": 0.4019927847448892,
+      "grad_norm": 1.3467859029769897,
+      "learning_rate": 4.957285496521261e-06,
+      "loss": 1.1048,
+      "step": 780
+    },
+    {
+      "epoch": 0.4025081601099467,
+      "grad_norm": 1.2900623083114624,
+      "learning_rate": 4.957160133552378e-06,
+      "loss": 1.1626,
+      "step": 781
+    },
+    {
+      "epoch": 0.4030235354750043,
+      "grad_norm": 1.2303043603897095,
+      "learning_rate": 4.957034588478615e-06,
+      "loss": 1.1012,
+      "step": 782
+    },
+    {
+      "epoch": 0.4035389108400618,
+      "grad_norm": 1.327517032623291,
+      "learning_rate": 4.956908861309275e-06,
+      "loss": 1.1346,
+      "step": 783
+    },
+    {
+      "epoch": 0.4040542862051194,
+      "grad_norm": 1.3691375255584717,
+      "learning_rate": 4.956782952053678e-06,
+      "loss": 1.1497,
+      "step": 784
+    },
+    {
+      "epoch": 0.40456966157017693,
+      "grad_norm": 1.3795244693756104,
+      "learning_rate": 4.956656860721153e-06,
+      "loss": 1.1014,
+      "step": 785
+    },
+    {
+      "epoch": 0.4050850369352345,
+      "grad_norm": 1.3545103073120117,
+      "learning_rate": 4.956530587321046e-06,
+      "loss": 1.1772,
+      "step": 786
+    },
+    {
+      "epoch": 0.40560041230029203,
+      "grad_norm": 1.2767109870910645,
+      "learning_rate": 4.956404131862715e-06,
+      "loss": 1.1162,
+      "step": 787
+    },
+    {
+      "epoch": 0.4061157876653496,
+      "grad_norm": 1.332641363143921,
+      "learning_rate": 4.956277494355532e-06,
+      "loss": 1.0951,
+      "step": 788
+    },
+    {
+      "epoch": 0.40663116303040714,
+      "grad_norm": 1.45758056640625,
+      "learning_rate": 4.956150674808883e-06,
+      "loss": 1.1645,
+      "step": 789
+    },
+    {
+      "epoch": 0.4071465383954647,
+      "grad_norm": 1.4294954538345337,
+      "learning_rate": 4.956023673232166e-06,
+      "loss": 1.1215,
+      "step": 790
+    },
+    {
+      "epoch": 0.40766191376052224,
+      "grad_norm": 1.3346449136734009,
+      "learning_rate": 4.955896489634795e-06,
+      "loss": 1.0929,
+      "step": 791
+    },
+    {
+      "epoch": 0.4081772891255798,
+      "grad_norm": 1.3155488967895508,
+      "learning_rate": 4.955769124026193e-06,
+      "loss": 1.1635,
+      "step": 792
+    },
+    {
+      "epoch": 0.40869266449063735,
+      "grad_norm": 1.362632155418396,
+      "learning_rate": 4.955641576415801e-06,
+      "loss": 1.1359,
+      "step": 793
+    },
+    {
+      "epoch": 0.40920803985569487,
+      "grad_norm": 1.3555985689163208,
+      "learning_rate": 4.955513846813072e-06,
+      "loss": 1.1666,
+      "step": 794
+    },
+    {
+      "epoch": 0.40972341522075245,
+      "grad_norm": 1.2590464353561401,
+      "learning_rate": 4.955385935227471e-06,
+      "loss": 1.1201,
+      "step": 795
+    },
+    {
+      "epoch": 0.41023879058581,
+      "grad_norm": 1.331552267074585,
+      "learning_rate": 4.955257841668481e-06,
+      "loss": 1.1116,
+      "step": 796
+    },
+    {
+      "epoch": 0.41075416595086756,
+      "grad_norm": 1.262492299079895,
+      "learning_rate": 4.95512956614559e-06,
+      "loss": 1.0981,
+      "step": 797
+    },
+    {
+      "epoch": 0.4112695413159251,
+      "grad_norm": 1.4217736721038818,
+      "learning_rate": 4.955001108668309e-06,
+      "loss": 1.1947,
+      "step": 798
+    },
+    {
+      "epoch": 0.41178491668098266,
+      "grad_norm": 1.3397902250289917,
+      "learning_rate": 4.954872469246157e-06,
+      "loss": 1.1093,
+      "step": 799
+    },
+    {
+      "epoch": 0.4123002920460402,
+      "grad_norm": 1.4462045431137085,
+      "learning_rate": 4.954743647888667e-06,
+      "loss": 1.1438,
+      "step": 800
+    },
+    {
+      "epoch": 0.41281566741109776,
+      "grad_norm": 1.3599969148635864,
+      "learning_rate": 4.954614644605388e-06,
+      "loss": 1.1726,
+      "step": 801
+    },
+    {
+      "epoch": 0.4133310427761553,
+      "grad_norm": 1.3505985736846924,
+      "learning_rate": 4.954485459405878e-06,
+      "loss": 1.0917,
+      "step": 802
+    },
+    {
+      "epoch": 0.41384641814121287,
+      "grad_norm": 1.262619972229004,
+      "learning_rate": 4.954356092299713e-06,
+      "loss": 1.1617,
+      "step": 803
+    },
+    {
+      "epoch": 0.4143617935062704,
+      "grad_norm": 1.3853216171264648,
+      "learning_rate": 4.954226543296481e-06,
+      "loss": 1.1576,
+      "step": 804
+    },
+    {
+      "epoch": 0.414877168871328,
+      "grad_norm": 1.4354254007339478,
+      "learning_rate": 4.9540968124057826e-06,
+      "loss": 1.1984,
+      "step": 805
+    },
+    {
+      "epoch": 0.4153925442363855,
+      "grad_norm": 1.2844696044921875,
+      "learning_rate": 4.953966899637231e-06,
+      "loss": 1.1313,
+      "step": 806
+    },
+    {
+      "epoch": 0.4159079196014431,
+      "grad_norm": 1.290461778640747,
+      "learning_rate": 4.953836805000457e-06,
+      "loss": 1.1077,
+      "step": 807
+    },
+    {
+      "epoch": 0.4164232949665006,
+      "grad_norm": 1.2542216777801514,
+      "learning_rate": 4.9537065285051e-06,
+      "loss": 1.1277,
+      "step": 808
+    },
+    {
+      "epoch": 0.4169386703315581,
+      "grad_norm": 1.2707545757293701,
+      "learning_rate": 4.953576070160815e-06,
+      "loss": 1.0971,
+      "step": 809
+    },
+    {
+      "epoch": 0.4174540456966157,
+      "grad_norm": 1.3372422456741333,
+      "learning_rate": 4.953445429977272e-06,
+      "loss": 1.1273,
+      "step": 810
+    },
+    {
+      "epoch": 0.41796942106167323,
+      "grad_norm": 1.370647668838501,
+      "learning_rate": 4.953314607964153e-06,
+      "loss": 1.1665,
+      "step": 811
+    },
+    {
+      "epoch": 0.4184847964267308,
+      "grad_norm": 1.2144851684570312,
+      "learning_rate": 4.953183604131151e-06,
+      "loss": 1.0726,
+      "step": 812
+    },
+    {
+      "epoch": 0.41900017179178833,
+      "grad_norm": 1.3138102293014526,
+      "learning_rate": 4.953052418487978e-06,
+      "loss": 1.2001,
+      "step": 813
+    },
+    {
+      "epoch": 0.4195155471568459,
+      "grad_norm": 1.2491734027862549,
+      "learning_rate": 4.952921051044354e-06,
+      "loss": 1.0989,
+      "step": 814
+    },
+    {
+      "epoch": 0.42003092252190344,
+      "grad_norm": 1.2666431665420532,
+      "learning_rate": 4.952789501810017e-06,
+      "loss": 1.1124,
+      "step": 815
+    },
+    {
+      "epoch": 0.420546297886961,
+      "grad_norm": 1.3701388835906982,
+      "learning_rate": 4.952657770794715e-06,
+      "loss": 1.1148,
+      "step": 816
+    },
+    {
+      "epoch": 0.42106167325201854,
+      "grad_norm": 1.3455960750579834,
+      "learning_rate": 4.952525858008211e-06,
+      "loss": 1.1049,
+      "step": 817
+    },
+    {
+      "epoch": 0.4215770486170761,
+      "grad_norm": 1.3758878707885742,
+      "learning_rate": 4.952393763460282e-06,
+      "loss": 1.1845,
+      "step": 818
+    },
+    {
+      "epoch": 0.42209242398213365,
+      "grad_norm": 1.3101781606674194,
+      "learning_rate": 4.952261487160717e-06,
+      "loss": 1.1154,
+      "step": 819
+    },
+    {
+      "epoch": 0.4226077993471912,
+      "grad_norm": 1.3576375246047974,
+      "learning_rate": 4.9521290291193196e-06,
+      "loss": 1.1373,
+      "step": 820
+    },
+    {
+      "epoch": 0.42312317471224875,
+      "grad_norm": 1.429225206375122,
+      "learning_rate": 4.951996389345906e-06,
+      "loss": 1.1445,
+      "step": 821
+    },
+    {
+      "epoch": 0.42363855007730633,
+      "grad_norm": 1.511832594871521,
+      "learning_rate": 4.951863567850307e-06,
+      "loss": 1.1471,
+      "step": 822
+    },
+    {
+      "epoch": 0.42415392544236385,
+      "grad_norm": 1.3234866857528687,
+      "learning_rate": 4.951730564642366e-06,
+      "loss": 1.1579,
+      "step": 823
+    },
+    {
+      "epoch": 0.4246693008074214,
+      "grad_norm": 1.3025860786437988,
+      "learning_rate": 4.951597379731941e-06,
+      "loss": 1.1451,
+      "step": 824
+    },
+    {
+      "epoch": 0.42518467617247896,
+      "grad_norm": 1.3261936902999878,
+      "learning_rate": 4.9514640131289e-06,
+      "loss": 1.0935,
+      "step": 825
+    },
+    {
+      "epoch": 0.4257000515375365,
+      "grad_norm": 1.3229318857192993,
+      "learning_rate": 4.9513304648431306e-06,
+      "loss": 1.1324,
+      "step": 826
+    },
+    {
+      "epoch": 0.42621542690259406,
+      "grad_norm": 1.2966307401657104,
+      "learning_rate": 4.951196734884528e-06,
+      "loss": 1.1034,
+      "step": 827
+    },
+    {
+      "epoch": 0.4267308022676516,
+      "grad_norm": 1.2589709758758545,
+      "learning_rate": 4.951062823263004e-06,
+      "loss": 1.1125,
+      "step": 828
+    },
+    {
+      "epoch": 0.42724617763270917,
+      "grad_norm": 1.2551952600479126,
+      "learning_rate": 4.950928729988483e-06,
+      "loss": 1.134,
+      "step": 829
+    },
+    {
+      "epoch": 0.4277615529977667,
+      "grad_norm": 1.4582545757293701,
+      "learning_rate": 4.950794455070903e-06,
+      "loss": 1.1099,
+      "step": 830
+    },
+    {
+      "epoch": 0.42827692836282427,
+      "grad_norm": 1.2585415840148926,
+      "learning_rate": 4.9506599985202144e-06,
+      "loss": 1.133,
+      "step": 831
+    },
+    {
+      "epoch": 0.4287923037278818,
+      "grad_norm": 1.3387125730514526,
+      "learning_rate": 4.950525360346382e-06,
+      "loss": 1.0834,
+      "step": 832
+    },
+    {
+      "epoch": 0.4293076790929394,
+      "grad_norm": 1.356967806816101,
+      "learning_rate": 4.9503905405593865e-06,
+      "loss": 1.1399,
+      "step": 833
+    },
+    {
+      "epoch": 0.4298230544579969,
+      "grad_norm": 1.212633728981018,
+      "learning_rate": 4.950255539169217e-06,
+      "loss": 1.1188,
+      "step": 834
+    },
+    {
+      "epoch": 0.4303384298230545,
+      "grad_norm": 1.3586931228637695,
+      "learning_rate": 4.9501203561858805e-06,
+      "loss": 1.1225,
+      "step": 835
+    },
+    {
+      "epoch": 0.430853805188112,
+      "grad_norm": 1.323394775390625,
+      "learning_rate": 4.949984991619394e-06,
+      "loss": 1.1006,
+      "step": 836
+    },
+    {
+      "epoch": 0.4313691805531696,
+      "grad_norm": 1.360657811164856,
+      "learning_rate": 4.949849445479791e-06,
+      "loss": 1.1395,
+      "step": 837
+    },
+    {
+      "epoch": 0.4318845559182271,
+      "grad_norm": 1.2871806621551514,
+      "learning_rate": 4.949713717777116e-06,
+      "loss": 1.1298,
+      "step": 838
+    },
+    {
+      "epoch": 0.43239993128328463,
+      "grad_norm": 1.318222999572754,
+      "learning_rate": 4.949577808521429e-06,
+      "loss": 1.0854,
+      "step": 839
+    },
+    {
+      "epoch": 0.4329153066483422,
+      "grad_norm": 1.4249770641326904,
+      "learning_rate": 4.949441717722803e-06,
+      "loss": 1.1626,
+      "step": 840
+    },
+    {
+      "epoch": 0.43343068201339974,
+      "grad_norm": 1.3032678365707397,
+      "learning_rate": 4.949305445391321e-06,
+      "loss": 1.0744,
+      "step": 841
+    },
+    {
+      "epoch": 0.4339460573784573,
+      "grad_norm": 1.4344840049743652,
+      "learning_rate": 4.9491689915370865e-06,
+      "loss": 1.1501,
+      "step": 842
+    },
+    {
+      "epoch": 0.43446143274351484,
+      "grad_norm": 1.396562933921814,
+      "learning_rate": 4.94903235617021e-06,
+      "loss": 1.0927,
+      "step": 843
+    },
+    {
+      "epoch": 0.4349768081085724,
+      "grad_norm": 1.4599568843841553,
+      "learning_rate": 4.948895539300817e-06,
+      "loss": 1.1006,
+      "step": 844
+    },
+    {
+      "epoch": 0.43549218347362995,
+      "grad_norm": 1.3650449514389038,
+      "learning_rate": 4.9487585409390494e-06,
+      "loss": 1.063,
+      "step": 845
+    },
+    {
+      "epoch": 0.4360075588386875,
+      "grad_norm": 1.3364917039871216,
+      "learning_rate": 4.948621361095059e-06,
+      "loss": 1.1678,
+      "step": 846
+    },
+    {
+      "epoch": 0.43652293420374505,
+      "grad_norm": 1.3469786643981934,
+      "learning_rate": 4.948483999779014e-06,
+      "loss": 1.1068,
+      "step": 847
+    },
+    {
+      "epoch": 0.43703830956880263,
+      "grad_norm": 1.3044260740280151,
+      "learning_rate": 4.948346457001092e-06,
+      "loss": 1.1267,
+      "step": 848
+    },
+    {
+      "epoch": 0.43755368493386015,
+      "grad_norm": 1.351914405822754,
+      "learning_rate": 4.948208732771489e-06,
+      "loss": 1.1117,
+      "step": 849
+    },
+    {
+      "epoch": 0.43806906029891773,
+      "grad_norm": 1.3472167253494263,
+      "learning_rate": 4.948070827100411e-06,
+      "loss": 1.1318,
+      "step": 850
+    },
+    {
+      "epoch": 0.43858443566397526,
+      "grad_norm": 1.2645403146743774,
+      "learning_rate": 4.9479327399980775e-06,
+      "loss": 1.0706,
+      "step": 851
+    },
+    {
+      "epoch": 0.43909981102903284,
+      "grad_norm": 1.3215440511703491,
+      "learning_rate": 4.947794471474724e-06,
+      "loss": 1.1504,
+      "step": 852
+    },
+    {
+      "epoch": 0.43961518639409036,
+      "grad_norm": 1.308405876159668,
+      "learning_rate": 4.9476560215405964e-06,
+      "loss": 1.1974,
+      "step": 853
+    },
+    {
+      "epoch": 0.4401305617591479,
+      "grad_norm": 1.222504734992981,
+      "learning_rate": 4.947517390205957e-06,
+      "loss": 1.1557,
+      "step": 854
+    },
+    {
+      "epoch": 0.44064593712420547,
+      "grad_norm": 1.300502896308899,
+      "learning_rate": 4.94737857748108e-06,
+      "loss": 1.1415,
+      "step": 855
+    },
+    {
+      "epoch": 0.441161312489263,
+      "grad_norm": 1.3184834718704224,
+      "learning_rate": 4.947239583376251e-06,
+      "loss": 1.1552,
+      "step": 856
+    },
+    {
+      "epoch": 0.44167668785432057,
+      "grad_norm": 1.3779114484786987,
+      "learning_rate": 4.947100407901773e-06,
+      "loss": 1.1275,
+      "step": 857
+    },
+    {
+      "epoch": 0.4421920632193781,
+      "grad_norm": 1.3452224731445312,
+      "learning_rate": 4.94696105106796e-06,
+      "loss": 1.1038,
+      "step": 858
+    },
+    {
+      "epoch": 0.4427074385844357,
+      "grad_norm": 1.2375786304473877,
+      "learning_rate": 4.946821512885141e-06,
+      "loss": 1.135,
+      "step": 859
+    },
+    {
+      "epoch": 0.4432228139494932,
+      "grad_norm": 1.418386459350586,
+      "learning_rate": 4.946681793363656e-06,
+      "loss": 1.1499,
+      "step": 860
+    },
+    {
+      "epoch": 0.4437381893145508,
+      "grad_norm": 1.3137166500091553,
+      "learning_rate": 4.94654189251386e-06,
+      "loss": 1.1094,
+      "step": 861
+    },
+    {
+      "epoch": 0.4442535646796083,
+      "grad_norm": 1.3592449426651,
+      "learning_rate": 4.9464018103461216e-06,
+      "loss": 1.1588,
+      "step": 862
+    },
+    {
+      "epoch": 0.4447689400446659,
+      "grad_norm": 1.3796391487121582,
+      "learning_rate": 4.946261546870823e-06,
+      "loss": 1.1628,
+      "step": 863
+    },
+    {
+      "epoch": 0.4452843154097234,
+      "grad_norm": 1.268865704536438,
+      "learning_rate": 4.946121102098359e-06,
+      "loss": 1.0974,
+      "step": 864
+    },
+    {
+      "epoch": 0.445799690774781,
+      "grad_norm": 1.3379385471343994,
+      "learning_rate": 4.9459804760391385e-06,
+      "loss": 1.1612,
+      "step": 865
+    },
+    {
+      "epoch": 0.4463150661398385,
+      "grad_norm": 1.4032179117202759,
+      "learning_rate": 4.945839668703583e-06,
+      "loss": 1.102,
+      "step": 866
+    },
+    {
+      "epoch": 0.4468304415048961,
+      "grad_norm": 1.3682565689086914,
+      "learning_rate": 4.945698680102129e-06,
+      "loss": 1.1517,
+      "step": 867
+    },
+    {
+      "epoch": 0.4473458168699536,
+      "grad_norm": 1.2709169387817383,
+      "learning_rate": 4.9455575102452245e-06,
+      "loss": 1.0855,
+      "step": 868
+    },
+    {
+      "epoch": 0.44786119223501114,
+      "grad_norm": 1.3753496408462524,
+      "learning_rate": 4.945416159143332e-06,
+      "loss": 1.0894,
+      "step": 869
+    },
+    {
+      "epoch": 0.4483765676000687,
+      "grad_norm": 1.3432855606079102,
+      "learning_rate": 4.945274626806928e-06,
+      "loss": 1.1431,
+      "step": 870
+    },
+    {
+      "epoch": 0.44889194296512624,
+      "grad_norm": 1.214583158493042,
+      "learning_rate": 4.945132913246501e-06,
+      "loss": 1.0696,
+      "step": 871
+    },
+    {
+      "epoch": 0.4494073183301838,
+      "grad_norm": 1.319673776626587,
+      "learning_rate": 4.944991018472553e-06,
+      "loss": 1.0505,
+      "step": 872
+    },
+    {
+      "epoch": 0.44992269369524135,
+      "grad_norm": 1.3592249155044556,
+      "learning_rate": 4.944848942495603e-06,
+      "loss": 1.133,
+      "step": 873
+    },
+    {
+      "epoch": 0.45043806906029893,
+      "grad_norm": 1.2998868227005005,
+      "learning_rate": 4.944706685326177e-06,
+      "loss": 1.083,
+      "step": 874
+    },
+    {
+      "epoch": 0.45095344442535645,
+      "grad_norm": 1.3852722644805908,
+      "learning_rate": 4.94456424697482e-06,
+      "loss": 1.0628,
+      "step": 875
+    },
+    {
+      "epoch": 0.45146881979041403,
+      "grad_norm": 1.280658483505249,
+      "learning_rate": 4.944421627452088e-06,
+      "loss": 1.0957,
+      "step": 876
+    },
+    {
+      "epoch": 0.45198419515547156,
+      "grad_norm": 1.359499216079712,
+      "learning_rate": 4.9442788267685506e-06,
+      "loss": 1.1592,
+      "step": 877
+    },
+    {
+      "epoch": 0.45249957052052914,
+      "grad_norm": 1.2672083377838135,
+      "learning_rate": 4.944135844934791e-06,
+      "loss": 1.1968,
+      "step": 878
+    },
+    {
+      "epoch": 0.45301494588558666,
+      "grad_norm": 1.3375358581542969,
+      "learning_rate": 4.943992681961406e-06,
+      "loss": 1.1149,
+      "step": 879
+    },
+    {
+      "epoch": 0.45353032125064424,
+      "grad_norm": 1.2706693410873413,
+      "learning_rate": 4.943849337859005e-06,
+      "loss": 1.1546,
+      "step": 880
+    },
+    {
+      "epoch": 0.45404569661570177,
+      "grad_norm": 1.268637776374817,
+      "learning_rate": 4.943705812638212e-06,
+      "loss": 1.1309,
+      "step": 881
+    },
+    {
+      "epoch": 0.45456107198075935,
+      "grad_norm": 1.3276478052139282,
+      "learning_rate": 4.943562106309666e-06,
+      "loss": 1.0721,
+      "step": 882
+    },
+    {
+      "epoch": 0.45507644734581687,
+      "grad_norm": 1.2741299867630005,
+      "learning_rate": 4.943418218884014e-06,
+      "loss": 1.0749,
+      "step": 883
+    },
+    {
+      "epoch": 0.4555918227108744,
+      "grad_norm": 1.3312971591949463,
+      "learning_rate": 4.943274150371921e-06,
+      "loss": 1.115,
+      "step": 884
+    },
+    {
+      "epoch": 0.456107198075932,
+      "grad_norm": 1.270114541053772,
+      "learning_rate": 4.943129900784065e-06,
+      "loss": 1.1562,
+      "step": 885
+    },
+    {
+      "epoch": 0.4566225734409895,
+      "grad_norm": 1.3587193489074707,
+      "learning_rate": 4.942985470131136e-06,
+      "loss": 1.0775,
+      "step": 886
+    },
+    {
+      "epoch": 0.4571379488060471,
+      "grad_norm": 1.4444305896759033,
+      "learning_rate": 4.942840858423838e-06,
+      "loss": 1.1518,
+      "step": 887
+    },
+    {
+      "epoch": 0.4576533241711046,
+      "grad_norm": 1.290061354637146,
+      "learning_rate": 4.942696065672889e-06,
+      "loss": 1.115,
+      "step": 888
+    },
+    {
+      "epoch": 0.4581686995361622,
+      "grad_norm": 1.3777953386306763,
+      "learning_rate": 4.942551091889018e-06,
+      "loss": 1.1741,
+      "step": 889
+    },
+    {
+      "epoch": 0.4586840749012197,
+      "grad_norm": 1.272931694984436,
+      "learning_rate": 4.942405937082971e-06,
+      "loss": 1.0733,
+      "step": 890
+    },
+    {
+      "epoch": 0.4591994502662773,
+      "grad_norm": 1.2352920770645142,
+      "learning_rate": 4.9422606012655065e-06,
+      "loss": 1.1237,
+      "step": 891
+    },
+    {
+      "epoch": 0.4597148256313348,
+      "grad_norm": 1.2935038805007935,
+      "learning_rate": 4.942115084447393e-06,
+      "loss": 1.1502,
+      "step": 892
+    },
+    {
+      "epoch": 0.4602302009963924,
+      "grad_norm": 1.3403886556625366,
+      "learning_rate": 4.941969386639418e-06,
+      "loss": 1.1346,
+      "step": 893
+    },
+    {
+      "epoch": 0.4607455763614499,
+      "grad_norm": 1.307370662689209,
+      "learning_rate": 4.941823507852377e-06,
+      "loss": 1.0788,
+      "step": 894
+    },
+    {
+      "epoch": 0.4612609517265075,
+      "grad_norm": 1.2700450420379639,
+      "learning_rate": 4.941677448097083e-06,
+      "loss": 1.1311,
+      "step": 895
+    },
+    {
+      "epoch": 0.461776327091565,
+      "grad_norm": 1.2455085515975952,
+      "learning_rate": 4.941531207384359e-06,
+      "loss": 1.085,
+      "step": 896
+    },
+    {
+      "epoch": 0.4622917024566226,
+      "grad_norm": 1.3723219633102417,
+      "learning_rate": 4.941384785725046e-06,
+      "loss": 1.0763,
+      "step": 897
+    },
+    {
+      "epoch": 0.4628070778216801,
+      "grad_norm": 1.3822237253189087,
+      "learning_rate": 4.941238183129993e-06,
+      "loss": 1.1476,
+      "step": 898
+    },
+    {
+      "epoch": 0.46332245318673765,
+      "grad_norm": 1.3194243907928467,
+      "learning_rate": 4.941091399610065e-06,
+      "loss": 1.0974,
+      "step": 899
+    },
+    {
+      "epoch": 0.46383782855179523,
+      "grad_norm": 1.3286584615707397,
+      "learning_rate": 4.940944435176143e-06,
+      "loss": 1.1035,
+      "step": 900
+    },
+    {
+      "epoch": 0.46435320391685275,
+      "grad_norm": 1.3928455114364624,
+      "learning_rate": 4.940797289839116e-06,
+      "loss": 1.1511,
+      "step": 901
+    },
+    {
+      "epoch": 0.46486857928191033,
+      "grad_norm": 1.3169835805892944,
+      "learning_rate": 4.940649963609891e-06,
+      "loss": 1.1494,
+      "step": 902
+    },
+    {
+      "epoch": 0.46538395464696786,
+      "grad_norm": 1.3504160642623901,
+      "learning_rate": 4.940502456499384e-06,
+      "loss": 1.1451,
+      "step": 903
+    },
+    {
+      "epoch": 0.46589933001202544,
+      "grad_norm": 1.3698813915252686,
+      "learning_rate": 4.940354768518531e-06,
+      "loss": 1.1354,
+      "step": 904
+    },
+    {
+      "epoch": 0.46641470537708296,
+      "grad_norm": 1.4112703800201416,
+      "learning_rate": 4.940206899678275e-06,
+      "loss": 1.104,
+      "step": 905
+    },
+    {
+      "epoch": 0.46693008074214054,
+      "grad_norm": 1.315726399421692,
+      "learning_rate": 4.9400588499895755e-06,
+      "loss": 1.0902,
+      "step": 906
+    },
+    {
+      "epoch": 0.46744545610719807,
+      "grad_norm": 1.2560619115829468,
+      "learning_rate": 4.939910619463404e-06,
+      "loss": 1.0996,
+      "step": 907
+    },
+    {
+      "epoch": 0.46796083147225565,
+      "grad_norm": 1.3877520561218262,
+      "learning_rate": 4.939762208110747e-06,
+      "loss": 1.152,
+      "step": 908
+    },
+    {
+      "epoch": 0.46847620683731317,
+      "grad_norm": 1.3177300691604614,
+      "learning_rate": 4.939613615942602e-06,
+      "loss": 1.1251,
+      "step": 909
+    },
+    {
+      "epoch": 0.46899158220237075,
+      "grad_norm": 1.262856125831604,
+      "learning_rate": 4.939464842969984e-06,
+      "loss": 1.0812,
+      "step": 910
+    },
+    {
+      "epoch": 0.4695069575674283,
+      "grad_norm": 1.3856077194213867,
+      "learning_rate": 4.939315889203917e-06,
+      "loss": 1.0771,
+      "step": 911
+    },
+    {
+      "epoch": 0.47002233293248585,
+      "grad_norm": 1.4262409210205078,
+      "learning_rate": 4.939166754655441e-06,
+      "loss": 1.0994,
+      "step": 912
+    },
+    {
+      "epoch": 0.4705377082975434,
+      "grad_norm": 1.3417930603027344,
+      "learning_rate": 4.93901743933561e-06,
+      "loss": 1.1559,
+      "step": 913
+    },
+    {
+      "epoch": 0.4710530836626009,
+      "grad_norm": 1.364356279373169,
+      "learning_rate": 4.938867943255486e-06,
+      "loss": 1.1401,
+      "step": 914
+    },
+    {
+      "epoch": 0.4715684590276585,
+      "grad_norm": 1.3301299810409546,
+      "learning_rate": 4.938718266426151e-06,
+      "loss": 1.1299,
+      "step": 915
+    },
+    {
+      "epoch": 0.472083834392716,
+      "grad_norm": 1.3044543266296387,
+      "learning_rate": 4.938568408858699e-06,
+      "loss": 1.1059,
+      "step": 916
+    },
+    {
+      "epoch": 0.4725992097577736,
+      "grad_norm": 1.3932881355285645,
+      "learning_rate": 4.938418370564235e-06,
+      "loss": 1.1629,
+      "step": 917
+    },
+    {
+      "epoch": 0.4731145851228311,
+      "grad_norm": 1.4347057342529297,
+      "learning_rate": 4.938268151553879e-06,
+      "loss": 1.1394,
+      "step": 918
+    },
+    {
+      "epoch": 0.4736299604878887,
+      "grad_norm": 1.3268437385559082,
+      "learning_rate": 4.938117751838762e-06,
+      "loss": 1.1288,
+      "step": 919
+    },
+    {
+      "epoch": 0.4741453358529462,
+      "grad_norm": 1.3698822259902954,
+      "learning_rate": 4.937967171430034e-06,
+      "loss": 1.1079,
+      "step": 920
+    },
+    {
+      "epoch": 0.4746607112180038,
+      "grad_norm": 1.2759448289871216,
+      "learning_rate": 4.937816410338852e-06,
+      "loss": 1.0551,
+      "step": 921
+    },
+    {
+      "epoch": 0.4751760865830613,
+      "grad_norm": 1.2582950592041016,
+      "learning_rate": 4.937665468576389e-06,
+      "loss": 1.1115,
+      "step": 922
+    },
+    {
+      "epoch": 0.4756914619481189,
+      "grad_norm": 1.3074204921722412,
+      "learning_rate": 4.937514346153834e-06,
+      "loss": 1.1403,
+      "step": 923
+    },
+    {
+      "epoch": 0.4762068373131764,
+      "grad_norm": 1.356652021408081,
+      "learning_rate": 4.937363043082385e-06,
+      "loss": 1.1309,
+      "step": 924
+    },
+    {
+      "epoch": 0.476722212678234,
+      "grad_norm": 1.4325493574142456,
+      "learning_rate": 4.937211559373256e-06,
+      "loss": 1.1881,
+      "step": 925
+    },
+    {
+      "epoch": 0.4772375880432915,
+      "grad_norm": 1.3235927820205688,
+      "learning_rate": 4.937059895037674e-06,
+      "loss": 1.0873,
+      "step": 926
+    },
+    {
+      "epoch": 0.4777529634083491,
+      "grad_norm": 1.3582242727279663,
+      "learning_rate": 4.936908050086879e-06,
+      "loss": 1.0874,
+      "step": 927
+    },
+    {
+      "epoch": 0.47826833877340663,
+      "grad_norm": 1.3077458143234253,
+      "learning_rate": 4.9367560245321235e-06,
+      "loss": 1.1308,
+      "step": 928
+    },
+    {
+      "epoch": 0.47878371413846416,
+      "grad_norm": 1.432529091835022,
+      "learning_rate": 4.936603818384676e-06,
+      "loss": 1.0825,
+      "step": 929
+    },
+    {
+      "epoch": 0.47929908950352174,
+      "grad_norm": 1.4369008541107178,
+      "learning_rate": 4.936451431655817e-06,
+      "loss": 1.1137,
+      "step": 930
+    },
+    {
+      "epoch": 0.47981446486857926,
+      "grad_norm": 1.3341528177261353,
+      "learning_rate": 4.936298864356838e-06,
+      "loss": 1.1584,
+      "step": 931
+    },
+    {
+      "epoch": 0.48032984023363684,
+      "grad_norm": 1.3432354927062988,
+      "learning_rate": 4.936146116499047e-06,
+      "loss": 1.083,
+      "step": 932
+    },
+    {
+      "epoch": 0.48084521559869436,
+      "grad_norm": 1.2591673135757446,
+      "learning_rate": 4.935993188093766e-06,
+      "loss": 1.107,
+      "step": 933
+    },
+    {
+      "epoch": 0.48136059096375194,
+      "grad_norm": 1.3461925983428955,
+      "learning_rate": 4.935840079152327e-06,
+      "loss": 1.1574,
+      "step": 934
+    },
+    {
+      "epoch": 0.48187596632880947,
+      "grad_norm": 1.3511323928833008,
+      "learning_rate": 4.9356867896860775e-06,
+      "loss": 1.1066,
+      "step": 935
+    },
+    {
+      "epoch": 0.48239134169386705,
+      "grad_norm": 1.3107941150665283,
+      "learning_rate": 4.935533319706378e-06,
+      "loss": 1.1095,
+      "step": 936
+    },
+    {
+      "epoch": 0.4829067170589246,
+      "grad_norm": 1.4062912464141846,
+      "learning_rate": 4.9353796692246035e-06,
+      "loss": 1.128,
+      "step": 937
+    },
+    {
+      "epoch": 0.48342209242398215,
+      "grad_norm": 1.278143048286438,
+      "learning_rate": 4.93522583825214e-06,
+      "loss": 1.1347,
+      "step": 938
+    },
+    {
+      "epoch": 0.4839374677890397,
+      "grad_norm": 1.3441195487976074,
+      "learning_rate": 4.93507182680039e-06,
+      "loss": 1.0718,
+      "step": 939
+    },
+    {
+      "epoch": 0.48445284315409726,
+      "grad_norm": 1.2608786821365356,
+      "learning_rate": 4.934917634880766e-06,
+      "loss": 1.0523,
+      "step": 940
+    },
+    {
+      "epoch": 0.4849682185191548,
+      "grad_norm": 1.3602216243743896,
+      "learning_rate": 4.934763262504696e-06,
+      "loss": 1.0999,
+      "step": 941
+    },
+    {
+      "epoch": 0.48548359388421236,
+      "grad_norm": 1.2475003004074097,
+      "learning_rate": 4.93460870968362e-06,
+      "loss": 1.1234,
+      "step": 942
+    },
+    {
+      "epoch": 0.4859989692492699,
+      "grad_norm": 1.3348537683486938,
+      "learning_rate": 4.934453976428993e-06,
+      "loss": 1.1116,
+      "step": 943
+    },
+    {
+      "epoch": 0.4865143446143274,
+      "grad_norm": 1.2883623838424683,
+      "learning_rate": 4.9342990627522835e-06,
+      "loss": 1.1644,
+      "step": 944
+    },
+    {
+      "epoch": 0.487029719979385,
+      "grad_norm": 1.4656498432159424,
+      "learning_rate": 4.934143968664971e-06,
+      "loss": 1.1505,
+      "step": 945
+    },
+    {
+      "epoch": 0.4875450953444425,
+      "grad_norm": 1.370727300643921,
+      "learning_rate": 4.93398869417855e-06,
+      "loss": 1.1315,
+      "step": 946
+    },
+    {
+      "epoch": 0.4880604707095001,
+      "grad_norm": 1.3100018501281738,
+      "learning_rate": 4.933833239304529e-06,
+      "loss": 1.13,
+      "step": 947
+    },
+    {
+      "epoch": 0.4885758460745576,
+      "grad_norm": 1.259399175643921,
+      "learning_rate": 4.933677604054428e-06,
+      "loss": 1.0983,
+      "step": 948
+    },
+    {
+      "epoch": 0.4890912214396152,
+      "grad_norm": 1.4250516891479492,
+      "learning_rate": 4.933521788439782e-06,
+      "loss": 1.0165,
+      "step": 949
+    },
+    {
+      "epoch": 0.4896065968046727,
+      "grad_norm": 1.3938947916030884,
+      "learning_rate": 4.933365792472139e-06,
+      "loss": 1.0528,
+      "step": 950
+    },
+    {
+      "epoch": 0.4901219721697303,
+      "grad_norm": 1.431322455406189,
+      "learning_rate": 4.93320961616306e-06,
+      "loss": 1.1415,
+      "step": 951
+    },
+    {
+      "epoch": 0.4906373475347878,
+      "grad_norm": 1.3879321813583374,
+      "learning_rate": 4.93305325952412e-06,
+      "loss": 1.1523,
+      "step": 952
+    },
+    {
+      "epoch": 0.4911527228998454,
+      "grad_norm": 1.3865762948989868,
+      "learning_rate": 4.932896722566905e-06,
+      "loss": 1.1528,
+      "step": 953
+    },
+    {
+      "epoch": 0.49166809826490293,
+      "grad_norm": 1.3409475088119507,
+      "learning_rate": 4.932740005303019e-06,
+      "loss": 1.0749,
+      "step": 954
+    },
+    {
+      "epoch": 0.4921834736299605,
+      "grad_norm": 1.3085192441940308,
+      "learning_rate": 4.932583107744074e-06,
+      "loss": 1.0723,
+      "step": 955
+    },
+    {
+      "epoch": 0.49269884899501803,
+      "grad_norm": 1.5081206560134888,
+      "learning_rate": 4.9324260299017e-06,
+      "loss": 1.124,
+      "step": 956
+    },
+    {
+      "epoch": 0.4932142243600756,
+      "grad_norm": 1.3491485118865967,
+      "learning_rate": 4.932268771787538e-06,
+      "loss": 1.087,
+      "step": 957
+    },
+    {
+      "epoch": 0.49372959972513314,
+      "grad_norm": 1.350962519645691,
+      "learning_rate": 4.932111333413241e-06,
+      "loss": 1.0825,
+      "step": 958
+    },
+    {
+      "epoch": 0.49424497509019066,
+      "grad_norm": 1.2890613079071045,
+      "learning_rate": 4.931953714790479e-06,
+      "loss": 1.1482,
+      "step": 959
+    },
+    {
+      "epoch": 0.49476035045524824,
+      "grad_norm": 1.3699564933776855,
+      "learning_rate": 4.931795915930933e-06,
+      "loss": 1.15,
+      "step": 960
+    },
+    {
+      "epoch": 0.49527572582030577,
+      "grad_norm": 1.3651552200317383,
+      "learning_rate": 4.931637936846298e-06,
+      "loss": 1.1472,
+      "step": 961
+    },
+    {
+      "epoch": 0.49579110118536335,
+      "grad_norm": 1.2937597036361694,
+      "learning_rate": 4.93147977754828e-06,
+      "loss": 1.1043,
+      "step": 962
+    },
+    {
+      "epoch": 0.49630647655042087,
+      "grad_norm": 1.3082481622695923,
+      "learning_rate": 4.931321438048603e-06,
+      "loss": 1.12,
+      "step": 963
+    },
+    {
+      "epoch": 0.49682185191547845,
+      "grad_norm": 1.3514046669006348,
+      "learning_rate": 4.931162918359001e-06,
+      "loss": 1.1209,
+      "step": 964
+    },
+    {
+      "epoch": 0.497337227280536,
+      "grad_norm": 1.3890619277954102,
+      "learning_rate": 4.931004218491223e-06,
+      "loss": 1.1624,
+      "step": 965
+    },
+    {
+      "epoch": 0.49785260264559356,
+      "grad_norm": 1.3216668367385864,
+      "learning_rate": 4.930845338457029e-06,
+      "loss": 1.052,
+      "step": 966
+    },
+    {
+      "epoch": 0.4983679780106511,
+      "grad_norm": 1.391402244567871,
+      "learning_rate": 4.930686278268194e-06,
+      "loss": 1.1068,
+      "step": 967
+    },
+    {
+      "epoch": 0.49888335337570866,
+      "grad_norm": 1.3651992082595825,
+      "learning_rate": 4.930527037936508e-06,
+      "loss": 1.1142,
+      "step": 968
+    },
+    {
+      "epoch": 0.4993987287407662,
+      "grad_norm": 1.415177822113037,
+      "learning_rate": 4.930367617473771e-06,
+      "loss": 1.0975,
+      "step": 969
+    },
+    {
+      "epoch": 0.49991410410582376,
+      "grad_norm": 1.352864146232605,
+      "learning_rate": 4.930208016891799e-06,
+      "loss": 1.1489,
+      "step": 970
+    },
+    {
+      "epoch": 0.5004294794708813,
+      "grad_norm": 1.3537369966506958,
+      "learning_rate": 4.930048236202419e-06,
+      "loss": 1.1184,
+      "step": 971
+    },
+    {
+      "epoch": 0.5009448548359389,
+      "grad_norm": 1.3870673179626465,
+      "learning_rate": 4.929888275417474e-06,
+      "loss": 1.1284,
+      "step": 972
+    },
+    {
+      "epoch": 0.5014602302009964,
+      "grad_norm": 1.3259119987487793,
+      "learning_rate": 4.929728134548818e-06,
+      "loss": 1.1019,
+      "step": 973
+    },
+    {
+      "epoch": 0.5019756055660539,
+      "grad_norm": 1.486378788948059,
+      "learning_rate": 4.92956781360832e-06,
+      "loss": 1.1106,
+      "step": 974
+    },
+    {
+      "epoch": 0.5024909809311114,
+      "grad_norm": 1.4134142398834229,
+      "learning_rate": 4.929407312607861e-06,
+      "loss": 1.0968,
+      "step": 975
+    },
+    {
+      "epoch": 0.5030063562961691,
+      "grad_norm": 1.3311936855316162,
+      "learning_rate": 4.929246631559337e-06,
+      "loss": 1.1293,
+      "step": 976
+    },
+    {
+      "epoch": 0.5035217316612266,
+      "grad_norm": 1.3417316675186157,
+      "learning_rate": 4.929085770474656e-06,
+      "loss": 1.1283,
+      "step": 977
+    },
+    {
+      "epoch": 0.5040371070262841,
+      "grad_norm": 1.2947065830230713,
+      "learning_rate": 4.928924729365739e-06,
+      "loss": 1.1121,
+      "step": 978
+    },
+    {
+      "epoch": 0.5045524823913417,
+      "grad_norm": 1.2538444995880127,
+      "learning_rate": 4.928763508244523e-06,
+      "loss": 1.0937,
+      "step": 979
+    },
+    {
+      "epoch": 0.5050678577563993,
+      "grad_norm": 1.2491954565048218,
+      "learning_rate": 4.928602107122954e-06,
+      "loss": 1.111,
+      "step": 980
+    },
+    {
+      "epoch": 0.5055832331214568,
+      "grad_norm": 1.2861011028289795,
+      "learning_rate": 4.928440526012995e-06,
+      "loss": 1.0607,
+      "step": 981
+    },
+    {
+      "epoch": 0.5060986084865143,
+      "grad_norm": 1.4138412475585938,
+      "learning_rate": 4.928278764926621e-06,
+      "loss": 1.1033,
+      "step": 982
+    },
+    {
+      "epoch": 0.5066139838515719,
+      "grad_norm": 1.4112071990966797,
+      "learning_rate": 4.928116823875821e-06,
+      "loss": 1.1158,
+      "step": 983
+    },
+    {
+      "epoch": 0.5071293592166295,
+      "grad_norm": 1.3451502323150635,
+      "learning_rate": 4.927954702872596e-06,
+      "loss": 1.1566,
+      "step": 984
+    },
+    {
+      "epoch": 0.507644734581687,
+      "grad_norm": 1.423324465751648,
+      "learning_rate": 4.92779240192896e-06,
+      "loss": 1.0742,
+      "step": 985
+    },
+    {
+      "epoch": 0.5081601099467445,
+      "grad_norm": 1.401418685913086,
+      "learning_rate": 4.927629921056944e-06,
+      "loss": 1.0844,
+      "step": 986
+    },
+    {
+      "epoch": 0.5086754853118021,
+      "grad_norm": 1.2989522218704224,
+      "learning_rate": 4.927467260268588e-06,
+      "loss": 1.0857,
+      "step": 987
+    },
+    {
+      "epoch": 0.5091908606768596,
+      "grad_norm": 1.2923847436904907,
+      "learning_rate": 4.927304419575948e-06,
+      "loss": 1.051,
+      "step": 988
+    },
+    {
+      "epoch": 0.5097062360419172,
+      "grad_norm": 1.3584847450256348,
+      "learning_rate": 4.927141398991091e-06,
+      "loss": 1.08,
+      "step": 989
+    },
+    {
+      "epoch": 0.5102216114069748,
+      "grad_norm": 1.3007075786590576,
+      "learning_rate": 4.9269781985261e-06,
+      "loss": 1.0283,
+      "step": 990
+    },
+    {
+      "epoch": 0.5107369867720323,
+      "grad_norm": 1.2998332977294922,
+      "learning_rate": 4.92681481819307e-06,
+      "loss": 1.1638,
+      "step": 991
+    },
+    {
+      "epoch": 0.5112523621370898,
+      "grad_norm": 1.3442524671554565,
+      "learning_rate": 4.92665125800411e-06,
+      "loss": 1.1349,
+      "step": 992
+    },
+    {
+      "epoch": 0.5117677375021474,
+      "grad_norm": 1.3257924318313599,
+      "learning_rate": 4.92648751797134e-06,
+      "loss": 1.1078,
+      "step": 993
+    },
+    {
+      "epoch": 0.512283112867205,
+      "grad_norm": 1.4120101928710938,
+      "learning_rate": 4.926323598106897e-06,
+      "loss": 1.1897,
+      "step": 994
+    },
+    {
+      "epoch": 0.5127984882322625,
+      "grad_norm": 1.3012393712997437,
+      "learning_rate": 4.9261594984229276e-06,
+      "loss": 1.1238,
+      "step": 995
+    },
+    {
+      "epoch": 0.51331386359732,
+      "grad_norm": 1.4377069473266602,
+      "learning_rate": 4.925995218931595e-06,
+      "loss": 1.078,
+      "step": 996
+    },
+    {
+      "epoch": 0.5138292389623776,
+      "grad_norm": 1.389089584350586,
+      "learning_rate": 4.925830759645075e-06,
+      "loss": 1.0876,
+      "step": 997
+    },
+    {
+      "epoch": 0.5143446143274352,
+      "grad_norm": 1.3466755151748657,
+      "learning_rate": 4.925666120575553e-06,
+      "loss": 1.0879,
+      "step": 998
+    },
+    {
+      "epoch": 0.5148599896924927,
+      "grad_norm": 1.488387942314148,
+      "learning_rate": 4.925501301735232e-06,
+      "loss": 1.1214,
+      "step": 999
+    },
+    {
+      "epoch": 0.5153753650575502,
+      "grad_norm": 1.3349742889404297,
+      "learning_rate": 4.92533630313633e-06,
+      "loss": 1.085,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5158907404226079,
+      "grad_norm": 1.4069832563400269,
+      "learning_rate": 4.925171124791071e-06,
+      "loss": 1.0276,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5164061157876654,
+      "grad_norm": 1.3818632364273071,
+      "learning_rate": 4.925005766711699e-06,
+      "loss": 1.112,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5169214911527229,
+      "grad_norm": 1.434308409690857,
+      "learning_rate": 4.9248402289104694e-06,
+      "loss": 1.1009,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5174368665177804,
+      "grad_norm": 1.3247915506362915,
+      "learning_rate": 4.924674511399649e-06,
+      "loss": 1.0939,
+      "step": 1004
+    },
+    {
+      "epoch": 0.517952241882838,
+      "grad_norm": 1.3572499752044678,
+      "learning_rate": 4.92450861419152e-06,
+      "loss": 1.0953,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5184676172478956,
+      "grad_norm": 1.2700880765914917,
+      "learning_rate": 4.9243425372983776e-06,
+      "loss": 1.0553,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5189829926129531,
+      "grad_norm": 1.3506757020950317,
+      "learning_rate": 4.92417628073253e-06,
+      "loss": 1.0627,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5194983679780106,
+      "grad_norm": 1.3553335666656494,
+      "learning_rate": 4.924009844506299e-06,
+      "loss": 1.1347,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5200137433430682,
+      "grad_norm": 1.3869633674621582,
+      "learning_rate": 4.9238432286320194e-06,
+      "loss": 1.1304,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5205291187081258,
+      "grad_norm": 1.4746654033660889,
+      "learning_rate": 4.923676433122039e-06,
+      "loss": 1.1368,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5210444940731833,
+      "grad_norm": 1.3538326025009155,
+      "learning_rate": 4.923509457988719e-06,
+      "loss": 1.1464,
+      "step": 1011
+    },
+    {
+      "epoch": 0.5215598694382408,
+      "grad_norm": 1.3130608797073364,
+      "learning_rate": 4.923342303244436e-06,
+      "loss": 1.0875,
+      "step": 1012
+    },
+    {
+      "epoch": 0.5220752448032984,
+      "grad_norm": 1.4428962469100952,
+      "learning_rate": 4.923174968901576e-06,
+      "loss": 1.1293,
+      "step": 1013
+    },
+    {
+      "epoch": 0.522590620168356,
+      "grad_norm": 1.417741298675537,
+      "learning_rate": 4.923007454972542e-06,
+      "loss": 1.0696,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5231059955334135,
+      "grad_norm": 1.3560295104980469,
+      "learning_rate": 4.9228397614697495e-06,
+      "loss": 1.076,
+      "step": 1015
+    },
+    {
+      "epoch": 0.523621370898471,
+      "grad_norm": 1.4185484647750854,
+      "learning_rate": 4.922671888405625e-06,
+      "loss": 1.1841,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5241367462635286,
+      "grad_norm": 1.3187556266784668,
+      "learning_rate": 4.922503835792609e-06,
+      "loss": 1.0751,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5246521216285861,
+      "grad_norm": 1.3971552848815918,
+      "learning_rate": 4.922335603643159e-06,
+      "loss": 1.051,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5251674969936437,
+      "grad_norm": 1.3550360202789307,
+      "learning_rate": 4.92216719196974e-06,
+      "loss": 1.102,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5256828723587013,
+      "grad_norm": 1.286242961883545,
+      "learning_rate": 4.921998600784837e-06,
+      "loss": 1.0976,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5261982477237588,
+      "grad_norm": 1.5528258085250854,
+      "learning_rate": 4.92182983010094e-06,
+      "loss": 1.1342,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5267136230888163,
+      "grad_norm": 1.4291971921920776,
+      "learning_rate": 4.921660879930561e-06,
+      "loss": 1.1202,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5272289984538739,
+      "grad_norm": 1.4119702577590942,
+      "learning_rate": 4.921491750286219e-06,
+      "loss": 1.128,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5277443738189315,
+      "grad_norm": 1.338379979133606,
+      "learning_rate": 4.9213224411804496e-06,
+      "loss": 1.0659,
+      "step": 1024
+    },
+    {
+      "epoch": 0.528259749183989,
+      "grad_norm": 1.3730744123458862,
+      "learning_rate": 4.9211529526258e-06,
+      "loss": 1.1565,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5287751245490465,
+      "grad_norm": 1.3540492057800293,
+      "learning_rate": 4.920983284634831e-06,
+      "loss": 1.1323,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5292904999141042,
+      "grad_norm": 1.4140055179595947,
+      "learning_rate": 4.9208134372201175e-06,
+      "loss": 1.1136,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5298058752791617,
+      "grad_norm": 1.3417617082595825,
+      "learning_rate": 4.920643410394248e-06,
+      "loss": 1.0707,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5303212506442192,
+      "grad_norm": 1.3641825914382935,
+      "learning_rate": 4.920473204169822e-06,
+      "loss": 1.0844,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5308366260092767,
+      "grad_norm": 1.32991361618042,
+      "learning_rate": 4.920302818559455e-06,
+      "loss": 1.1099,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5313520013743344,
+      "grad_norm": 1.4204286336898804,
+      "learning_rate": 4.920132253575774e-06,
+      "loss": 1.0921,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5318673767393919,
+      "grad_norm": 1.4128758907318115,
+      "learning_rate": 4.91996150923142e-06,
+      "loss": 1.1581,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5323827521044494,
+      "grad_norm": 1.3604763746261597,
+      "learning_rate": 4.9197905855390475e-06,
+      "loss": 1.0801,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5328981274695069,
+      "grad_norm": 1.4139909744262695,
+      "learning_rate": 4.919619482511323e-06,
+      "loss": 1.111,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5334135028345645,
+      "grad_norm": 1.2476108074188232,
+      "learning_rate": 4.919448200160929e-06,
+      "loss": 1.0655,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5339288781996221,
+      "grad_norm": 1.53774893283844,
+      "learning_rate": 4.919276738500558e-06,
+      "loss": 1.1361,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5344442535646796,
+      "grad_norm": 1.355755090713501,
+      "learning_rate": 4.919105097542918e-06,
+      "loss": 1.0765,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5349596289297371,
+      "grad_norm": 1.364965796470642,
+      "learning_rate": 4.918933277300729e-06,
+      "loss": 1.1845,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5354750042947947,
+      "grad_norm": 1.4210840463638306,
+      "learning_rate": 4.918761277786727e-06,
+      "loss": 1.1407,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5359903796598523,
+      "grad_norm": 1.3257887363433838,
+      "learning_rate": 4.918589099013656e-06,
+      "loss": 1.1206,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5365057550249098,
+      "grad_norm": 1.3990665674209595,
+      "learning_rate": 4.918416740994279e-06,
+      "loss": 1.0949,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5370211303899673,
+      "grad_norm": 1.3532612323760986,
+      "learning_rate": 4.918244203741369e-06,
+      "loss": 1.0698,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5375365057550249,
+      "grad_norm": 1.2847880125045776,
+      "learning_rate": 4.918071487267712e-06,
+      "loss": 1.0495,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5380518811200825,
+      "grad_norm": 1.421730399131775,
+      "learning_rate": 4.917898591586111e-06,
+      "loss": 1.1283,
+      "step": 1044
+    },
+    {
+      "epoch": 0.53856725648514,
+      "grad_norm": 1.3442106246948242,
+      "learning_rate": 4.9177255167093776e-06,
+      "loss": 1.1023,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5390826318501976,
+      "grad_norm": 1.4238824844360352,
+      "learning_rate": 4.917552262650339e-06,
+      "loss": 1.0958,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5395980072152551,
+      "grad_norm": 1.3223488330841064,
+      "learning_rate": 4.917378829421836e-06,
+      "loss": 1.1002,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5401133825803126,
+      "grad_norm": 1.3856841325759888,
+      "learning_rate": 4.917205217036721e-06,
+      "loss": 1.0923,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5406287579453702,
+      "grad_norm": 1.363553524017334,
+      "learning_rate": 4.917031425507862e-06,
+      "loss": 1.0757,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5411441333104278,
+      "grad_norm": 1.3707692623138428,
+      "learning_rate": 4.916857454848139e-06,
+      "loss": 1.0935,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5416595086754853,
+      "grad_norm": 1.2749682664871216,
+      "learning_rate": 4.916683305070443e-06,
+      "loss": 1.0831,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5421748840405428,
+      "grad_norm": 1.413495421409607,
+      "learning_rate": 4.916508976187683e-06,
+      "loss": 1.0782,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5426902594056005,
+      "grad_norm": 1.3967243432998657,
+      "learning_rate": 4.916334468212778e-06,
+      "loss": 1.0928,
+      "step": 1053
+    },
+    {
+      "epoch": 0.543205634770658,
+      "grad_norm": 1.3892028331756592,
+      "learning_rate": 4.916159781158661e-06,
+      "loss": 1.0899,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5437210101357155,
+      "grad_norm": 1.2645163536071777,
+      "learning_rate": 4.91598491503828e-06,
+      "loss": 1.0128,
+      "step": 1055
+    },
+    {
+      "epoch": 0.544236385500773,
+      "grad_norm": 1.3261868953704834,
+      "learning_rate": 4.915809869864592e-06,
+      "loss": 1.0605,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5447517608658307,
+      "grad_norm": 1.4390619993209839,
+      "learning_rate": 4.915634645650572e-06,
+      "loss": 1.1214,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5452671362308882,
+      "grad_norm": 1.374504804611206,
+      "learning_rate": 4.915459242409205e-06,
+      "loss": 1.0504,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5457825115959457,
+      "grad_norm": 1.4694688320159912,
+      "learning_rate": 4.915283660153491e-06,
+      "loss": 1.1,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5462978869610032,
+      "grad_norm": 1.3922492265701294,
+      "learning_rate": 4.915107898896443e-06,
+      "loss": 1.0793,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5468132623260609,
+      "grad_norm": 1.3784502744674683,
+      "learning_rate": 4.914931958651086e-06,
+      "loss": 1.141,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5473286376911184,
+      "grad_norm": 1.3735002279281616,
+      "learning_rate": 4.914755839430461e-06,
+      "loss": 1.0931,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5478440130561759,
+      "grad_norm": 1.4746862649917603,
+      "learning_rate": 4.914579541247618e-06,
+      "loss": 1.1115,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5483593884212334,
+      "grad_norm": 1.3234630823135376,
+      "learning_rate": 4.914403064115625e-06,
+      "loss": 1.13,
+      "step": 1064
+    },
+    {
+      "epoch": 0.548874763786291,
+      "grad_norm": 1.326145052909851,
+      "learning_rate": 4.914226408047561e-06,
+      "loss": 1.0677,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5493901391513486,
+      "grad_norm": 1.3029916286468506,
+      "learning_rate": 4.914049573056517e-06,
+      "loss": 1.1147,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5499055145164061,
+      "grad_norm": 1.3349426984786987,
+      "learning_rate": 4.913872559155599e-06,
+      "loss": 1.1048,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5504208898814636,
+      "grad_norm": 1.4011579751968384,
+      "learning_rate": 4.913695366357927e-06,
+      "loss": 1.1096,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5509362652465212,
+      "grad_norm": 1.4701299667358398,
+      "learning_rate": 4.9135179946766306e-06,
+      "loss": 1.0878,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5514516406115788,
+      "grad_norm": 1.2991549968719482,
+      "learning_rate": 4.913340444124858e-06,
+      "loss": 1.1315,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5519670159766363,
+      "grad_norm": 1.4357717037200928,
+      "learning_rate": 4.9131627147157664e-06,
+      "loss": 1.1378,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5524823913416939,
+      "grad_norm": 1.3685011863708496,
+      "learning_rate": 4.912984806462529e-06,
+      "loss": 1.0955,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5529977667067514,
+      "grad_norm": 1.36729097366333,
+      "learning_rate": 4.912806719378328e-06,
+      "loss": 1.1441,
+      "step": 1073
+    },
+    {
+      "epoch": 0.553513142071809,
+      "grad_norm": 1.376643419265747,
+      "learning_rate": 4.912628453476364e-06,
+      "loss": 1.0875,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5540285174368665,
+      "grad_norm": 1.2910010814666748,
+      "learning_rate": 4.912450008769848e-06,
+      "loss": 1.1218,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5545438928019241,
+      "grad_norm": 1.327258586883545,
+      "learning_rate": 4.912271385272005e-06,
+      "loss": 1.0689,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5550592681669816,
+      "grad_norm": 1.2881981134414673,
+      "learning_rate": 4.912092582996074e-06,
+      "loss": 1.0891,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5555746435320391,
+      "grad_norm": 1.3386520147323608,
+      "learning_rate": 4.911913601955305e-06,
+      "loss": 1.1099,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5560900188970967,
+      "grad_norm": 1.4942947626113892,
+      "learning_rate": 4.911734442162963e-06,
+      "loss": 1.1043,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5566053942621543,
+      "grad_norm": 1.4550931453704834,
+      "learning_rate": 4.911555103632326e-06,
+      "loss": 1.1317,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5571207696272118,
+      "grad_norm": 1.3831936120986938,
+      "learning_rate": 4.911375586376685e-06,
+      "loss": 1.0696,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5576361449922693,
+      "grad_norm": 1.3928511142730713,
+      "learning_rate": 4.911195890409345e-06,
+      "loss": 1.0852,
+      "step": 1082
+    },
+    {
+      "epoch": 0.558151520357327,
+      "grad_norm": 1.3147382736206055,
+      "learning_rate": 4.911016015743624e-06,
+      "loss": 1.1356,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5586668957223845,
+      "grad_norm": 1.2969478368759155,
+      "learning_rate": 4.910835962392851e-06,
+      "loss": 1.055,
+      "step": 1084
+    },
+    {
+      "epoch": 0.559182271087442,
+      "grad_norm": 1.2879401445388794,
+      "learning_rate": 4.9106557303703715e-06,
+      "loss": 1.1252,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5596976464524995,
+      "grad_norm": 1.4337031841278076,
+      "learning_rate": 4.910475319689542e-06,
+      "loss": 1.166,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5602130218175572,
+      "grad_norm": 1.4629108905792236,
+      "learning_rate": 4.910294730363734e-06,
+      "loss": 1.0777,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5607283971826147,
+      "grad_norm": 1.3451007604599,
+      "learning_rate": 4.910113962406331e-06,
+      "loss": 1.1065,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5612437725476722,
+      "grad_norm": 1.3503655195236206,
+      "learning_rate": 4.9099330158307295e-06,
+      "loss": 1.0694,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5617591479127297,
+      "grad_norm": 1.2731573581695557,
+      "learning_rate": 4.909751890650341e-06,
+      "loss": 1.0915,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5622745232777874,
+      "grad_norm": 1.3867830038070679,
+      "learning_rate": 4.909570586878587e-06,
+      "loss": 1.133,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5627898986428449,
+      "grad_norm": 1.3512017726898193,
+      "learning_rate": 4.909389104528906e-06,
+      "loss": 1.1475,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5633052740079024,
+      "grad_norm": 1.5640790462493896,
+      "learning_rate": 4.909207443614749e-06,
+      "loss": 1.1382,
+      "step": 1093
+    },
+    {
+      "epoch": 0.56382064937296,
+      "grad_norm": 1.3644344806671143,
+      "learning_rate": 4.909025604149576e-06,
+      "loss": 1.0482,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5643360247380175,
+      "grad_norm": 1.4538713693618774,
+      "learning_rate": 4.908843586146866e-06,
+      "loss": 1.1135,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5648514001030751,
+      "grad_norm": 1.4145160913467407,
+      "learning_rate": 4.908661389620108e-06,
+      "loss": 1.1097,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5653667754681326,
+      "grad_norm": 1.3469972610473633,
+      "learning_rate": 4.908479014582805e-06,
+      "loss": 1.1048,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5658821508331902,
+      "grad_norm": 1.4490288496017456,
+      "learning_rate": 4.908296461048473e-06,
+      "loss": 1.0747,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5663975261982477,
+      "grad_norm": 1.4236392974853516,
+      "learning_rate": 4.908113729030641e-06,
+      "loss": 1.0718,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5669129015633053,
+      "grad_norm": 1.399814248085022,
+      "learning_rate": 4.907930818542853e-06,
+      "loss": 1.1199,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5674282769283628,
+      "grad_norm": 1.3749892711639404,
+      "learning_rate": 4.907747729598663e-06,
+      "loss": 1.15,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5679436522934204,
+      "grad_norm": 1.3988521099090576,
+      "learning_rate": 4.907564462211642e-06,
+      "loss": 1.1193,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5684590276584779,
+      "grad_norm": 1.3416619300842285,
+      "learning_rate": 4.907381016395371e-06,
+      "loss": 1.0519,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5689744030235355,
+      "grad_norm": 1.3166208267211914,
+      "learning_rate": 4.907197392163445e-06,
+      "loss": 1.1237,
+      "step": 1104
+    },
+    {
+      "epoch": 0.569489778388593,
+      "grad_norm": 1.3533228635787964,
+      "learning_rate": 4.907013589529474e-06,
+      "loss": 1.0962,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5700051537536506,
+      "grad_norm": 1.477770447731018,
+      "learning_rate": 4.906829608507081e-06,
+      "loss": 1.1339,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5705205291187081,
+      "grad_norm": 1.417343020439148,
+      "learning_rate": 4.906645449109898e-06,
+      "loss": 1.1963,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5710359044837656,
+      "grad_norm": 1.2893075942993164,
+      "learning_rate": 4.906461111351576e-06,
+      "loss": 1.1646,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5715512798488233,
+      "grad_norm": 1.4533003568649292,
+      "learning_rate": 4.906276595245776e-06,
+      "loss": 1.1308,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5720666552138808,
+      "grad_norm": 1.3315889835357666,
+      "learning_rate": 4.906091900806172e-06,
+      "loss": 1.0804,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5725820305789383,
+      "grad_norm": 1.4494061470031738,
+      "learning_rate": 4.9059070280464525e-06,
+      "loss": 1.145,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5730974059439958,
+      "grad_norm": 1.3500940799713135,
+      "learning_rate": 4.90572197698032e-06,
+      "loss": 1.0763,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5736127813090535,
+      "grad_norm": 1.392949104309082,
+      "learning_rate": 4.905536747621487e-06,
+      "loss": 1.0601,
+      "step": 1113
+    },
+    {
+      "epoch": 0.574128156674111,
+      "grad_norm": 1.416709065437317,
+      "learning_rate": 4.9053513399836825e-06,
+      "loss": 1.1257,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5746435320391685,
+      "grad_norm": 1.3534293174743652,
+      "learning_rate": 4.905165754080648e-06,
+      "loss": 1.0476,
+      "step": 1115
+    },
+    {
+      "epoch": 0.575158907404226,
+      "grad_norm": 1.3441128730773926,
+      "learning_rate": 4.9049799899261354e-06,
+      "loss": 1.1083,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5756742827692837,
+      "grad_norm": 1.3544447422027588,
+      "learning_rate": 4.904794047533914e-06,
+      "loss": 1.117,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5761896581343412,
+      "grad_norm": 1.420417070388794,
+      "learning_rate": 4.904607926917763e-06,
+      "loss": 1.1451,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5767050334993987,
+      "grad_norm": 1.4358339309692383,
+      "learning_rate": 4.904421628091477e-06,
+      "loss": 1.1677,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5772204088644562,
+      "grad_norm": 1.32780122756958,
+      "learning_rate": 4.9042351510688635e-06,
+      "loss": 1.1154,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5777357842295139,
+      "grad_norm": 1.3847447633743286,
+      "learning_rate": 4.904048495863743e-06,
+      "loss": 1.0686,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5782511595945714,
+      "grad_norm": 1.330560326576233,
+      "learning_rate": 4.903861662489946e-06,
+      "loss": 1.1178,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5787665349596289,
+      "grad_norm": 1.2878084182739258,
+      "learning_rate": 4.903674650961321e-06,
+      "loss": 1.0595,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5792819103246865,
+      "grad_norm": 1.4180364608764648,
+      "learning_rate": 4.903487461291728e-06,
+      "loss": 1.1535,
+      "step": 1124
+    },
+    {
+      "epoch": 0.579797285689744,
+      "grad_norm": 1.302341103553772,
+      "learning_rate": 4.90330009349504e-06,
+      "loss": 1.0497,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5803126610548016,
+      "grad_norm": 1.3654823303222656,
+      "learning_rate": 4.9031125475851435e-06,
+      "loss": 1.1243,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5808280364198591,
+      "grad_norm": 1.4400690793991089,
+      "learning_rate": 4.902924823575937e-06,
+      "loss": 1.104,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5813434117849167,
+      "grad_norm": 1.3116410970687866,
+      "learning_rate": 4.902736921481333e-06,
+      "loss": 1.1156,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5818587871499742,
+      "grad_norm": 1.3312976360321045,
+      "learning_rate": 4.902548841315258e-06,
+      "loss": 1.0912,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5823741625150318,
+      "grad_norm": 1.3902885913848877,
+      "learning_rate": 4.902360583091651e-06,
+      "loss": 1.1086,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5828895378800893,
+      "grad_norm": 1.3142887353897095,
+      "learning_rate": 4.9021721468244636e-06,
+      "loss": 1.0117,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5834049132451469,
+      "grad_norm": 1.3276392221450806,
+      "learning_rate": 4.901983532527661e-06,
+      "loss": 1.1226,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5839202886102044,
+      "grad_norm": 1.4024556875228882,
+      "learning_rate": 4.901794740215222e-06,
+      "loss": 1.1554,
+      "step": 1133
+    },
+    {
+      "epoch": 0.584435663975262,
+      "grad_norm": 1.5021957159042358,
+      "learning_rate": 4.90160576990114e-06,
+      "loss": 1.102,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5849510393403196,
+      "grad_norm": 1.3124959468841553,
+      "learning_rate": 4.901416621599417e-06,
+      "loss": 1.0831,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5854664147053771,
+      "grad_norm": 1.2595129013061523,
+      "learning_rate": 4.901227295324073e-06,
+      "loss": 1.0665,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5859817900704346,
+      "grad_norm": 1.3545048236846924,
+      "learning_rate": 4.90103779108914e-06,
+      "loss": 1.0416,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5864971654354921,
+      "grad_norm": 1.398465871810913,
+      "learning_rate": 4.90084810890866e-06,
+      "loss": 1.1606,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5870125408005498,
+      "grad_norm": 1.3547437191009521,
+      "learning_rate": 4.9006582487966935e-06,
+      "loss": 1.0701,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5875279161656073,
+      "grad_norm": 1.3768725395202637,
+      "learning_rate": 4.900468210767309e-06,
+      "loss": 1.095,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5880432915306648,
+      "grad_norm": 1.3751181364059448,
+      "learning_rate": 4.9002779948345925e-06,
+      "loss": 1.1587,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5885586668957223,
+      "grad_norm": 1.3258569240570068,
+      "learning_rate": 4.90008760101264e-06,
+      "loss": 1.0568,
+      "step": 1142
+    },
+    {
+      "epoch": 0.58907404226078,
+      "grad_norm": 1.3601237535476685,
+      "learning_rate": 4.899897029315563e-06,
+      "loss": 1.086,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5895894176258375,
+      "grad_norm": 1.4093011617660522,
+      "learning_rate": 4.899706279757485e-06,
+      "loss": 1.0956,
+      "step": 1144
+    },
+    {
+      "epoch": 0.590104792990895,
+      "grad_norm": 1.3243459463119507,
+      "learning_rate": 4.899515352352541e-06,
+      "loss": 1.0627,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5906201683559525,
+      "grad_norm": 1.3498246669769287,
+      "learning_rate": 4.899324247114883e-06,
+      "loss": 1.069,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5911355437210102,
+      "grad_norm": 1.2826563119888306,
+      "learning_rate": 4.899132964058674e-06,
+      "loss": 1.1011,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5916509190860677,
+      "grad_norm": 1.4750365018844604,
+      "learning_rate": 4.89894150319809e-06,
+      "loss": 1.0783,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5921662944511252,
+      "grad_norm": 1.4334691762924194,
+      "learning_rate": 4.8987498645473195e-06,
+      "loss": 1.0941,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5926816698161828,
+      "grad_norm": 1.42925226688385,
+      "learning_rate": 4.898558048120568e-06,
+      "loss": 1.0639,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5931970451812404,
+      "grad_norm": 1.4867373704910278,
+      "learning_rate": 4.898366053932048e-06,
+      "loss": 1.1295,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5937124205462979,
+      "grad_norm": 1.3330955505371094,
+      "learning_rate": 4.89817388199599e-06,
+      "loss": 1.0712,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5942277959113554,
+      "grad_norm": 1.4610333442687988,
+      "learning_rate": 4.897981532326637e-06,
+      "loss": 1.07,
+      "step": 1153
+    },
+    {
+      "epoch": 0.594743171276413,
+      "grad_norm": 1.3709636926651,
+      "learning_rate": 4.8977890049382434e-06,
+      "loss": 1.1292,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5952585466414705,
+      "grad_norm": 1.3390896320343018,
+      "learning_rate": 4.8975962998450786e-06,
+      "loss": 1.0636,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5957739220065281,
+      "grad_norm": 1.4554226398468018,
+      "learning_rate": 4.897403417061423e-06,
+      "loss": 1.0734,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5962892973715856,
+      "grad_norm": 1.5800076723098755,
+      "learning_rate": 4.897210356601574e-06,
+      "loss": 1.1332,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5968046727366432,
+      "grad_norm": 1.3435883522033691,
+      "learning_rate": 4.897017118479837e-06,
+      "loss": 1.1483,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5973200481017007,
+      "grad_norm": 1.3186991214752197,
+      "learning_rate": 4.8968237027105345e-06,
+      "loss": 1.1018,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5978354234667583,
+      "grad_norm": 1.3472167253494263,
+      "learning_rate": 4.896630109308e-06,
+      "loss": 1.0525,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5983507988318159,
+      "grad_norm": 1.330782413482666,
+      "learning_rate": 4.896436338286583e-06,
+      "loss": 1.0714,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5988661741968734,
+      "grad_norm": 1.3701021671295166,
+      "learning_rate": 4.896242389660643e-06,
+      "loss": 1.1301,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5993815495619309,
+      "grad_norm": 1.3650195598602295,
+      "learning_rate": 4.896048263444553e-06,
+      "loss": 1.1395,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5998969249269885,
+      "grad_norm": 1.413641333580017,
+      "learning_rate": 4.895853959652702e-06,
+      "loss": 1.0819,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6004123002920461,
+      "grad_norm": 1.3155937194824219,
+      "learning_rate": 4.895659478299489e-06,
+      "loss": 1.0741,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6009276756571036,
+      "grad_norm": 1.336728811264038,
+      "learning_rate": 4.895464819399327e-06,
+      "loss": 1.0835,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6014430510221611,
+      "grad_norm": 1.292924165725708,
+      "learning_rate": 4.895269982966644e-06,
+      "loss": 1.1051,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6019584263872186,
+      "grad_norm": 1.3154529333114624,
+      "learning_rate": 4.8950749690158786e-06,
+      "loss": 1.145,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6024738017522763,
+      "grad_norm": 1.3772039413452148,
+      "learning_rate": 4.894879777561484e-06,
+      "loss": 1.0542,
+      "step": 1169
+    },
+    {
+      "epoch": 0.6029891771173338,
+      "grad_norm": 1.3813426494598389,
+      "learning_rate": 4.894684408617927e-06,
+      "loss": 1.1071,
+      "step": 1170
+    },
+    {
+      "epoch": 0.6035045524823913,
+      "grad_norm": 1.3168193101882935,
+      "learning_rate": 4.894488862199685e-06,
+      "loss": 1.0478,
+      "step": 1171
+    },
+    {
+      "epoch": 0.6040199278474488,
+      "grad_norm": 1.3762054443359375,
+      "learning_rate": 4.894293138321252e-06,
+      "loss": 1.1,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6045353032125065,
+      "grad_norm": 1.3713799715042114,
+      "learning_rate": 4.894097236997131e-06,
+      "loss": 1.1282,
+      "step": 1173
+    },
+    {
+      "epoch": 0.605050678577564,
+      "grad_norm": 1.3565059900283813,
+      "learning_rate": 4.893901158241845e-06,
+      "loss": 1.1271,
+      "step": 1174
+    },
+    {
+      "epoch": 0.6055660539426215,
+      "grad_norm": 1.3203566074371338,
+      "learning_rate": 4.893704902069921e-06,
+      "loss": 1.1487,
+      "step": 1175
+    },
+    {
+      "epoch": 0.606081429307679,
+      "grad_norm": 1.4336824417114258,
+      "learning_rate": 4.893508468495908e-06,
+      "loss": 1.0786,
+      "step": 1176
+    },
+    {
+      "epoch": 0.6065968046727367,
+      "grad_norm": 1.4044010639190674,
+      "learning_rate": 4.89331185753436e-06,
+      "loss": 1.0893,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6071121800377942,
+      "grad_norm": 1.4692096710205078,
+      "learning_rate": 4.893115069199852e-06,
+      "loss": 1.0447,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6076275554028517,
+      "grad_norm": 1.3869678974151611,
+      "learning_rate": 4.8929181035069665e-06,
+      "loss": 1.1459,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6081429307679093,
+      "grad_norm": 1.3510606288909912,
+      "learning_rate": 4.892720960470301e-06,
+      "loss": 1.0778,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6086583061329669,
+      "grad_norm": 1.3487348556518555,
+      "learning_rate": 4.892523640104467e-06,
+      "loss": 1.0876,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6091736814980244,
+      "grad_norm": 1.39715576171875,
+      "learning_rate": 4.8923261424240864e-06,
+      "loss": 1.116,
+      "step": 1182
+    },
+    {
+      "epoch": 0.609689056863082,
+      "grad_norm": 1.3872671127319336,
+      "learning_rate": 4.892128467443797e-06,
+      "loss": 1.0469,
+      "step": 1183
+    },
+    {
+      "epoch": 0.6102044322281395,
+      "grad_norm": 1.3705984354019165,
+      "learning_rate": 4.891930615178251e-06,
+      "loss": 1.1037,
+      "step": 1184
+    },
+    {
+      "epoch": 0.610719807593197,
+      "grad_norm": 1.3599284887313843,
+      "learning_rate": 4.8917325856421095e-06,
+      "loss": 1.0936,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6112351829582546,
+      "grad_norm": 1.400292992591858,
+      "learning_rate": 4.891534378850049e-06,
+      "loss": 1.055,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6117505583233122,
+      "grad_norm": 1.4634445905685425,
+      "learning_rate": 4.891335994816759e-06,
+      "loss": 1.1246,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6122659336883697,
+      "grad_norm": 1.3618847131729126,
+      "learning_rate": 4.891137433556943e-06,
+      "loss": 1.0879,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6127813090534272,
+      "grad_norm": 1.3145513534545898,
+      "learning_rate": 4.890938695085315e-06,
+      "loss": 1.1384,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6132966844184848,
+      "grad_norm": 1.292287826538086,
+      "learning_rate": 4.890739779416606e-06,
+      "loss": 1.1067,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6138120597835424,
+      "grad_norm": 1.371002197265625,
+      "learning_rate": 4.8905406865655576e-06,
+      "loss": 1.0729,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6143274351485999,
+      "grad_norm": 1.3559002876281738,
+      "learning_rate": 4.890341416546923e-06,
+      "loss": 1.059,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6148428105136574,
+      "grad_norm": 1.379361867904663,
+      "learning_rate": 4.890141969375473e-06,
+      "loss": 1.0391,
+      "step": 1193
+    },
+    {
+      "epoch": 0.615358185878715,
+      "grad_norm": 1.3302912712097168,
+      "learning_rate": 4.889942345065987e-06,
+      "loss": 1.0825,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6158735612437726,
+      "grad_norm": 1.3227627277374268,
+      "learning_rate": 4.889742543633261e-06,
+      "loss": 1.0492,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6163889366088301,
+      "grad_norm": 1.3230408430099487,
+      "learning_rate": 4.889542565092103e-06,
+      "loss": 1.0268,
+      "step": 1196
+    },
+    {
+      "epoch": 0.6169043119738876,
+      "grad_norm": 1.4277966022491455,
+      "learning_rate": 4.889342409457332e-06,
+      "loss": 1.143,
+      "step": 1197
+    },
+    {
+      "epoch": 0.6174196873389451,
+      "grad_norm": 1.4398748874664307,
+      "learning_rate": 4.889142076743782e-06,
+      "loss": 1.0711,
+      "step": 1198
+    },
+    {
+      "epoch": 0.6179350627040028,
+      "grad_norm": 1.3817853927612305,
+      "learning_rate": 4.888941566966302e-06,
+      "loss": 1.0831,
+      "step": 1199
+    },
+    {
+      "epoch": 0.6184504380690603,
+      "grad_norm": 1.395257830619812,
+      "learning_rate": 4.888740880139751e-06,
+      "loss": 1.1395,
+      "step": 1200
+    },
+    {
+      "epoch": 0.6189658134341178,
+      "grad_norm": 1.3928101062774658,
+      "learning_rate": 4.888540016279002e-06,
+      "loss": 1.1209,
+      "step": 1201
+    },
+    {
+      "epoch": 0.6194811887991754,
+      "grad_norm": 1.3110088109970093,
+      "learning_rate": 4.888338975398941e-06,
+      "loss": 1.0692,
+      "step": 1202
+    },
+    {
+      "epoch": 0.619996564164233,
+      "grad_norm": 1.3726956844329834,
+      "learning_rate": 4.8881377575144695e-06,
+      "loss": 1.156,
+      "step": 1203
+    },
+    {
+      "epoch": 0.6205119395292905,
+      "grad_norm": 1.4747378826141357,
+      "learning_rate": 4.887936362640498e-06,
+      "loss": 1.0864,
+      "step": 1204
+    },
+    {
+      "epoch": 0.621027314894348,
+      "grad_norm": 1.3230063915252686,
+      "learning_rate": 4.887734790791954e-06,
+      "loss": 1.0657,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6215426902594056,
+      "grad_norm": 1.4198203086853027,
+      "learning_rate": 4.8875330419837755e-06,
+      "loss": 1.085,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6220580656244632,
+      "grad_norm": 1.403670072555542,
+      "learning_rate": 4.887331116230914e-06,
+      "loss": 1.137,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6225734409895207,
+      "grad_norm": 1.3896167278289795,
+      "learning_rate": 4.8871290135483355e-06,
+      "loss": 1.1539,
+      "step": 1208
+    },
+    {
+      "epoch": 0.6230888163545782,
+      "grad_norm": 1.440527081489563,
+      "learning_rate": 4.886926733951017e-06,
+      "loss": 1.1683,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6236041917196358,
+      "grad_norm": 1.488156795501709,
+      "learning_rate": 4.886724277453952e-06,
+      "loss": 1.1122,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6241195670846934,
+      "grad_norm": 1.3573834896087646,
+      "learning_rate": 4.886521644072143e-06,
+      "loss": 1.0381,
+      "step": 1211
+    },
+    {
+      "epoch": 0.6246349424497509,
+      "grad_norm": 1.3494553565979004,
+      "learning_rate": 4.886318833820608e-06,
+      "loss": 1.0298,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6251503178148085,
+      "grad_norm": 1.4108291864395142,
+      "learning_rate": 4.886115846714379e-06,
+      "loss": 1.1535,
+      "step": 1213
+    },
+    {
+      "epoch": 0.625665693179866,
+      "grad_norm": 1.3653464317321777,
+      "learning_rate": 4.885912682768498e-06,
+      "loss": 1.0589,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6261810685449235,
+      "grad_norm": 1.397425651550293,
+      "learning_rate": 4.885709341998022e-06,
+      "loss": 1.1063,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6266964439099811,
+      "grad_norm": 1.334436297416687,
+      "learning_rate": 4.885505824418022e-06,
+      "loss": 1.04,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6272118192750387,
+      "grad_norm": 1.3866463899612427,
+      "learning_rate": 4.88530213004358e-06,
+      "loss": 1.0938,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6277271946400962,
+      "grad_norm": 1.3958983421325684,
+      "learning_rate": 4.885098258889793e-06,
+      "loss": 1.1031,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6282425700051537,
+      "grad_norm": 1.3861192464828491,
+      "learning_rate": 4.88489421097177e-06,
+      "loss": 1.0862,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6287579453702113,
+      "grad_norm": 1.3057092428207397,
+      "learning_rate": 4.884689986304634e-06,
+      "loss": 1.0929,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6292733207352689,
+      "grad_norm": 1.4449186325073242,
+      "learning_rate": 4.88448558490352e-06,
+      "loss": 1.1231,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6297886961003264,
+      "grad_norm": 1.2825870513916016,
+      "learning_rate": 4.884281006783576e-06,
+      "loss": 1.0594,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6303040714653839,
+      "grad_norm": 1.310887336730957,
+      "learning_rate": 4.8840762519599635e-06,
+      "loss": 1.1037,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6308194468304416,
+      "grad_norm": 1.440755844116211,
+      "learning_rate": 4.883871320447859e-06,
+      "loss": 1.1015,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6313348221954991,
+      "grad_norm": 1.355614185333252,
+      "learning_rate": 4.883666212262448e-06,
+      "loss": 1.0333,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6318501975605566,
+      "grad_norm": 1.4518755674362183,
+      "learning_rate": 4.883460927418933e-06,
+      "loss": 1.0318,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6323655729256141,
+      "grad_norm": 1.4447261095046997,
+      "learning_rate": 4.883255465932529e-06,
+      "loss": 1.1,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6328809482906717,
+      "grad_norm": 1.4496663808822632,
+      "learning_rate": 4.883049827818461e-06,
+      "loss": 1.0935,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6333963236557293,
+      "grad_norm": 1.4826685190200806,
+      "learning_rate": 4.88284401309197e-06,
+      "loss": 1.1498,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6339116990207868,
+      "grad_norm": 1.4179977178573608,
+      "learning_rate": 4.882638021768311e-06,
+      "loss": 1.0997,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6344270743858443,
+      "grad_norm": 1.2896610498428345,
+      "learning_rate": 4.882431853862748e-06,
+      "loss": 1.0368,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6349424497509019,
+      "grad_norm": 1.4019243717193604,
+      "learning_rate": 4.8822255093905615e-06,
+      "loss": 1.0817,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6354578251159595,
+      "grad_norm": 1.3491712808609009,
+      "learning_rate": 4.882018988367045e-06,
+      "loss": 1.0498,
+      "step": 1233
+    },
+    {
+      "epoch": 0.635973200481017,
+      "grad_norm": 1.3700180053710938,
+      "learning_rate": 4.881812290807502e-06,
+      "loss": 1.0581,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6364885758460745,
+      "grad_norm": 1.420124888420105,
+      "learning_rate": 4.881605416727252e-06,
+      "loss": 1.1296,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6370039512111321,
+      "grad_norm": 1.5151277780532837,
+      "learning_rate": 4.881398366141628e-06,
+      "loss": 1.0867,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6375193265761897,
+      "grad_norm": 1.4695781469345093,
+      "learning_rate": 4.881191139065975e-06,
+      "loss": 1.0429,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6380347019412472,
+      "grad_norm": 1.4124763011932373,
+      "learning_rate": 4.880983735515649e-06,
+      "loss": 1.0662,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6385500773063048,
+      "grad_norm": 1.4462312459945679,
+      "learning_rate": 4.8807761555060225e-06,
+      "loss": 1.1051,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6390654526713623,
+      "grad_norm": 1.4130510091781616,
+      "learning_rate": 4.88056839905248e-06,
+      "loss": 1.0944,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6395808280364199,
+      "grad_norm": 1.4933778047561646,
+      "learning_rate": 4.880360466170417e-06,
+      "loss": 1.0561,
+      "step": 1241
+    },
+    {
+      "epoch": 0.6400962034014774,
+      "grad_norm": 1.3993109464645386,
+      "learning_rate": 4.880152356875247e-06,
+      "loss": 1.084,
+      "step": 1242
+    },
+    {
+      "epoch": 0.640611578766535,
+      "grad_norm": 1.3925676345825195,
+      "learning_rate": 4.87994407118239e-06,
+      "loss": 1.058,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6411269541315925,
+      "grad_norm": 1.4072788953781128,
+      "learning_rate": 4.8797356091072834e-06,
+      "loss": 1.1107,
+      "step": 1244
+    },
+    {
+      "epoch": 0.64164232949665,
+      "grad_norm": 1.4816721677780151,
+      "learning_rate": 4.879526970665378e-06,
+      "loss": 1.1502,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6421577048617076,
+      "grad_norm": 1.3737413883209229,
+      "learning_rate": 4.8793181558721355e-06,
+      "loss": 1.0962,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6426730802267652,
+      "grad_norm": 1.4381495714187622,
+      "learning_rate": 4.879109164743031e-06,
+      "loss": 1.1229,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6431884555918227,
+      "grad_norm": 1.4116510152816772,
+      "learning_rate": 4.878899997293554e-06,
+      "loss": 1.1239,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6437038309568802,
+      "grad_norm": 1.3602337837219238,
+      "learning_rate": 4.878690653539207e-06,
+      "loss": 1.0246,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6442192063219379,
+      "grad_norm": 1.4238361120224,
+      "learning_rate": 4.878481133495503e-06,
+      "loss": 1.1035,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6447345816869954,
+      "grad_norm": 1.3942928314208984,
+      "learning_rate": 4.878271437177972e-06,
+      "loss": 1.0803,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6452499570520529,
+      "grad_norm": 1.4026683568954468,
+      "learning_rate": 4.878061564602153e-06,
+      "loss": 1.1248,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6457653324171104,
+      "grad_norm": 1.600054144859314,
+      "learning_rate": 4.877851515783601e-06,
+      "loss": 1.0738,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6462807077821681,
+      "grad_norm": 1.4209798574447632,
+      "learning_rate": 4.8776412907378845e-06,
+      "loss": 1.0954,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6467960831472256,
+      "grad_norm": 1.3372230529785156,
+      "learning_rate": 4.877430889480581e-06,
+      "loss": 1.0453,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6473114585122831,
+      "grad_norm": 1.3959600925445557,
+      "learning_rate": 4.877220312027285e-06,
+      "loss": 1.0663,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6478268338773406,
+      "grad_norm": 1.451054334640503,
+      "learning_rate": 4.877009558393604e-06,
+      "loss": 1.1363,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6483422092423982,
+      "grad_norm": 1.4474360942840576,
+      "learning_rate": 4.876798628595156e-06,
+      "loss": 1.1172,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6488575846074558,
+      "grad_norm": 1.4710158109664917,
+      "learning_rate": 4.876587522647573e-06,
+      "loss": 1.1213,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6493729599725133,
+      "grad_norm": 1.3668112754821777,
+      "learning_rate": 4.876376240566501e-06,
+      "loss": 1.1113,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6498883353375708,
+      "grad_norm": 1.3233379125595093,
+      "learning_rate": 4.876164782367599e-06,
+      "loss": 1.1201,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6504037107026284,
+      "grad_norm": 1.504072904586792,
+      "learning_rate": 4.875953148066538e-06,
+      "loss": 1.075,
+      "step": 1262
+    },
+    {
+      "epoch": 0.650919086067686,
+      "grad_norm": 1.4190407991409302,
+      "learning_rate": 4.875741337679004e-06,
+      "loss": 1.0805,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6514344614327435,
+      "grad_norm": 1.443258285522461,
+      "learning_rate": 4.875529351220692e-06,
+      "loss": 1.1517,
+      "step": 1264
+    },
+    {
+      "epoch": 0.651949836797801,
+      "grad_norm": 1.3786828517913818,
+      "learning_rate": 4.875317188707315e-06,
+      "loss": 1.1067,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6524652121628586,
+      "grad_norm": 1.3303650617599487,
+      "learning_rate": 4.875104850154596e-06,
+      "loss": 1.0173,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6529805875279162,
+      "grad_norm": 1.3950949907302856,
+      "learning_rate": 4.8748923355782715e-06,
+      "loss": 1.0967,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6534959628929737,
+      "grad_norm": 1.3890631198883057,
+      "learning_rate": 4.874679644994092e-06,
+      "loss": 1.1126,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6540113382580313,
+      "grad_norm": 1.3618621826171875,
+      "learning_rate": 4.874466778417819e-06,
+      "loss": 1.115,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6545267136230888,
+      "grad_norm": 1.344192385673523,
+      "learning_rate": 4.874253735865231e-06,
+      "loss": 1.0854,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6550420889881464,
+      "grad_norm": 1.4194031953811646,
+      "learning_rate": 4.874040517352116e-06,
+      "loss": 1.1088,
+      "step": 1271
+    },
+    {
+      "epoch": 0.655557464353204,
+      "grad_norm": 1.3026530742645264,
+      "learning_rate": 4.873827122894274e-06,
+      "loss": 1.1148,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6560728397182615,
+      "grad_norm": 1.3492462635040283,
+      "learning_rate": 4.873613552507522e-06,
+      "loss": 1.0606,
+      "step": 1273
+    },
+    {
+      "epoch": 0.656588215083319,
+      "grad_norm": 1.3427256345748901,
+      "learning_rate": 4.873399806207688e-06,
+      "loss": 1.0943,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6571035904483765,
+      "grad_norm": 1.3541367053985596,
+      "learning_rate": 4.8731858840106136e-06,
+      "loss": 1.1692,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6576189658134342,
+      "grad_norm": 1.361036777496338,
+      "learning_rate": 4.8729717859321515e-06,
+      "loss": 1.173,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6581343411784917,
+      "grad_norm": 1.497694492340088,
+      "learning_rate": 4.8727575119881705e-06,
+      "loss": 1.1123,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6586497165435492,
+      "grad_norm": 1.354684591293335,
+      "learning_rate": 4.872543062194549e-06,
+      "loss": 1.0228,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6591650919086067,
+      "grad_norm": 1.4002902507781982,
+      "learning_rate": 4.872328436567183e-06,
+      "loss": 1.0547,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6596804672736644,
+      "grad_norm": 1.4142106771469116,
+      "learning_rate": 4.872113635121975e-06,
+      "loss": 1.0316,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6601958426387219,
+      "grad_norm": 1.3479201793670654,
+      "learning_rate": 4.871898657874848e-06,
+      "loss": 1.0617,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6607112180037794,
+      "grad_norm": 1.5031355619430542,
+      "learning_rate": 4.871683504841734e-06,
+      "loss": 1.1113,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6612265933688369,
+      "grad_norm": 1.39547598361969,
+      "learning_rate": 4.871468176038575e-06,
+      "loss": 1.0954,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6617419687338946,
+      "grad_norm": 1.3589881658554077,
+      "learning_rate": 4.8712526714813335e-06,
+      "loss": 1.0636,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6622573440989521,
+      "grad_norm": 1.3259470462799072,
+      "learning_rate": 4.8710369911859784e-06,
+      "loss": 1.0264,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6627727194640096,
+      "grad_norm": 1.4223607778549194,
+      "learning_rate": 4.870821135168495e-06,
+      "loss": 1.0917,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6632880948290671,
+      "grad_norm": 1.3349062204360962,
+      "learning_rate": 4.870605103444881e-06,
+      "loss": 1.0898,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6638034701941247,
+      "grad_norm": 1.3264777660369873,
+      "learning_rate": 4.870388896031147e-06,
+      "loss": 1.0527,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6643188455591823,
+      "grad_norm": 1.4408807754516602,
+      "learning_rate": 4.870172512943316e-06,
+      "loss": 1.1317,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6648342209242398,
+      "grad_norm": 1.5054023265838623,
+      "learning_rate": 4.869955954197425e-06,
+      "loss": 1.0686,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6653495962892974,
+      "grad_norm": 1.5121653079986572,
+      "learning_rate": 4.869739219809523e-06,
+      "loss": 1.0969,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6658649716543549,
+      "grad_norm": 1.376417636871338,
+      "learning_rate": 4.869522309795673e-06,
+      "loss": 1.0612,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6663803470194125,
+      "grad_norm": 1.3991574048995972,
+      "learning_rate": 4.8693052241719515e-06,
+      "loss": 1.0951,
+      "step": 1293
+    },
+    {
+      "epoch": 0.66689572238447,
+      "grad_norm": 1.3785655498504639,
+      "learning_rate": 4.869087962954446e-06,
+      "loss": 1.071,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6674110977495276,
+      "grad_norm": 1.4480599164962769,
+      "learning_rate": 4.868870526159258e-06,
+      "loss": 1.1226,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6679264731145851,
+      "grad_norm": 1.4805289506912231,
+      "learning_rate": 4.868652913802504e-06,
+      "loss": 1.094,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6684418484796427,
+      "grad_norm": 1.3418303728103638,
+      "learning_rate": 4.86843512590031e-06,
+      "loss": 1.0665,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6689572238447002,
+      "grad_norm": 1.3156306743621826,
+      "learning_rate": 4.868217162468815e-06,
+      "loss": 1.0971,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6694725992097578,
+      "grad_norm": 1.429788589477539,
+      "learning_rate": 4.867999023524176e-06,
+      "loss": 1.1008,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6699879745748153,
+      "grad_norm": 1.3980108499526978,
+      "learning_rate": 4.867780709082559e-06,
+      "loss": 1.1053,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6705033499398729,
+      "grad_norm": 1.454076886177063,
+      "learning_rate": 4.867562219160143e-06,
+      "loss": 1.1428,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6710187253049305,
+      "grad_norm": 1.5541733503341675,
+      "learning_rate": 4.86734355377312e-06,
+      "loss": 1.0759,
+      "step": 1302
+    },
+    {
+      "epoch": 0.671534100669988,
+      "grad_norm": 1.4767528772354126,
+      "learning_rate": 4.867124712937696e-06,
+      "loss": 1.0519,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6720494760350455,
+      "grad_norm": 1.4581927061080933,
+      "learning_rate": 4.8669056966700914e-06,
+      "loss": 1.0816,
+      "step": 1304
+    },
+    {
+      "epoch": 0.672564851400103,
+      "grad_norm": 1.389357089996338,
+      "learning_rate": 4.8666865049865356e-06,
+      "loss": 1.1189,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6730802267651607,
+      "grad_norm": 1.36599600315094,
+      "learning_rate": 4.8664671379032755e-06,
+      "loss": 1.1449,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6735956021302182,
+      "grad_norm": 1.364134430885315,
+      "learning_rate": 4.866247595436568e-06,
+      "loss": 1.1239,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6741109774952757,
+      "grad_norm": 1.476054310798645,
+      "learning_rate": 4.866027877602682e-06,
+      "loss": 1.1238,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6746263528603332,
+      "grad_norm": 1.428935170173645,
+      "learning_rate": 4.865807984417903e-06,
+      "loss": 1.073,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6751417282253909,
+      "grad_norm": 1.4435595273971558,
+      "learning_rate": 4.865587915898528e-06,
+      "loss": 1.1536,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6756571035904484,
+      "grad_norm": 1.5261706113815308,
+      "learning_rate": 4.8653676720608674e-06,
+      "loss": 1.0984,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6761724789555059,
+      "grad_norm": 1.401324987411499,
+      "learning_rate": 4.865147252921242e-06,
+      "loss": 1.0563,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6766878543205634,
+      "grad_norm": 1.2494412660598755,
+      "learning_rate": 4.8649266584959875e-06,
+      "loss": 1.0598,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6772032296856211,
+      "grad_norm": 1.3429361581802368,
+      "learning_rate": 4.864705888801454e-06,
+      "loss": 1.1142,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6777186050506786,
+      "grad_norm": 1.407556176185608,
+      "learning_rate": 4.864484943854002e-06,
+      "loss": 1.0681,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6782339804157361,
+      "grad_norm": 1.3672014474868774,
+      "learning_rate": 4.864263823670006e-06,
+      "loss": 1.1239,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6787493557807937,
+      "grad_norm": 1.4719024896621704,
+      "learning_rate": 4.864042528265855e-06,
+      "loss": 1.1257,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6792647311458512,
+      "grad_norm": 1.3469836711883545,
+      "learning_rate": 4.863821057657949e-06,
+      "loss": 1.0655,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6797801065109088,
+      "grad_norm": 1.4195460081100464,
+      "learning_rate": 4.863599411862701e-06,
+      "loss": 1.0797,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6802954818759663,
+      "grad_norm": 1.378151297569275,
+      "learning_rate": 4.863377590896539e-06,
+      "loss": 1.0966,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6808108572410239,
+      "grad_norm": 1.3662402629852295,
+      "learning_rate": 4.863155594775901e-06,
+      "loss": 1.1436,
+      "step": 1321
+    },
+    {
+      "epoch": 0.6813262326060814,
+      "grad_norm": 1.300870418548584,
+      "learning_rate": 4.86293342351724e-06,
+      "loss": 1.1113,
+      "step": 1322
+    },
+    {
+      "epoch": 0.681841607971139,
+      "grad_norm": 1.332433819770813,
+      "learning_rate": 4.862711077137022e-06,
+      "loss": 1.0535,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6823569833361965,
+      "grad_norm": 1.4568887948989868,
+      "learning_rate": 4.8624885556517255e-06,
+      "loss": 1.096,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6828723587012541,
+      "grad_norm": 1.3251585960388184,
+      "learning_rate": 4.8622658590778414e-06,
+      "loss": 1.0773,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6833877340663116,
+      "grad_norm": 1.4032732248306274,
+      "learning_rate": 4.8620429874318745e-06,
+      "loss": 1.1054,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6839031094313692,
+      "grad_norm": 1.4974113702774048,
+      "learning_rate": 4.861819940730342e-06,
+      "loss": 1.1095,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6844184847964268,
+      "grad_norm": 1.4674794673919678,
+      "learning_rate": 4.861596718989775e-06,
+      "loss": 1.0566,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6849338601614843,
+      "grad_norm": 1.4152774810791016,
+      "learning_rate": 4.861373322226717e-06,
+      "loss": 1.1515,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6854492355265418,
+      "grad_norm": 1.5060356855392456,
+      "learning_rate": 4.861149750457724e-06,
+      "loss": 1.1229,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6859646108915994,
+      "grad_norm": 1.3054554462432861,
+      "learning_rate": 4.8609260036993645e-06,
+      "loss": 1.0185,
+      "step": 1331
+    },
+    {
+      "epoch": 0.686479986256657,
+      "grad_norm": 1.4473235607147217,
+      "learning_rate": 4.8607020819682215e-06,
+      "loss": 1.086,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6869953616217145,
+      "grad_norm": 1.3452242612838745,
+      "learning_rate": 4.860477985280891e-06,
+      "loss": 1.1032,
+      "step": 1333
+    },
+    {
+      "epoch": 0.687510736986772,
+      "grad_norm": 1.418095588684082,
+      "learning_rate": 4.86025371365398e-06,
+      "loss": 1.0907,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6880261123518295,
+      "grad_norm": 1.3405357599258423,
+      "learning_rate": 4.86002926710411e-06,
+      "loss": 1.1112,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6885414877168872,
+      "grad_norm": 1.4815081357955933,
+      "learning_rate": 4.859804645647916e-06,
+      "loss": 1.1308,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6890568630819447,
+      "grad_norm": 1.3682807683944702,
+      "learning_rate": 4.859579849302044e-06,
+      "loss": 1.1069,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6895722384470022,
+      "grad_norm": 1.3925576210021973,
+      "learning_rate": 4.8593548780831545e-06,
+      "loss": 1.1286,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6900876138120597,
+      "grad_norm": 1.3944441080093384,
+      "learning_rate": 4.859129732007921e-06,
+      "loss": 1.115,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6906029891771174,
+      "grad_norm": 1.5272127389907837,
+      "learning_rate": 4.85890441109303e-06,
+      "loss": 1.0707,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6911183645421749,
+      "grad_norm": 1.3506841659545898,
+      "learning_rate": 4.858678915355178e-06,
+      "loss": 1.098,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6916337399072324,
+      "grad_norm": 1.3435107469558716,
+      "learning_rate": 4.85845324481108e-06,
+      "loss": 1.0728,
+      "step": 1342
+    },
+    {
+      "epoch": 0.69214911527229,
+      "grad_norm": 1.4491636753082275,
+      "learning_rate": 4.858227399477458e-06,
+      "loss": 1.1069,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6926644906373476,
+      "grad_norm": 1.4442695379257202,
+      "learning_rate": 4.858001379371052e-06,
+      "loss": 1.1027,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6931798660024051,
+      "grad_norm": 1.373400092124939,
+      "learning_rate": 4.857775184508613e-06,
+      "loss": 1.1197,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6936952413674626,
+      "grad_norm": 1.4126453399658203,
+      "learning_rate": 4.857548814906903e-06,
+      "loss": 1.0845,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6942106167325202,
+      "grad_norm": 1.4741835594177246,
+      "learning_rate": 4.857322270582699e-06,
+      "loss": 1.0952,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6947259920975777,
+      "grad_norm": 1.427833914756775,
+      "learning_rate": 4.857095551552792e-06,
+      "loss": 1.0763,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6952413674626353,
+      "grad_norm": 1.4624282121658325,
+      "learning_rate": 4.856868657833983e-06,
+      "loss": 1.093,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6957567428276928,
+      "grad_norm": 1.4597184658050537,
+      "learning_rate": 4.856641589443089e-06,
+      "loss": 1.0819,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6962721181927504,
+      "grad_norm": 1.4595695734024048,
+      "learning_rate": 4.856414346396937e-06,
+      "loss": 1.1036,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6967874935578079,
+      "grad_norm": 1.4645100831985474,
+      "learning_rate": 4.85618692871237e-06,
+      "loss": 1.0967,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6973028689228655,
+      "grad_norm": 1.4653542041778564,
+      "learning_rate": 4.855959336406241e-06,
+      "loss": 1.0282,
+      "step": 1353
+    },
+    {
+      "epoch": 0.697818244287923,
+      "grad_norm": 1.3708640336990356,
+      "learning_rate": 4.855731569495417e-06,
+      "loss": 1.1137,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6983336196529806,
+      "grad_norm": 1.3453445434570312,
+      "learning_rate": 4.855503627996781e-06,
+      "loss": 1.0917,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6988489950180381,
+      "grad_norm": 1.4275254011154175,
+      "learning_rate": 4.855275511927223e-06,
+      "loss": 1.1415,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6993643703830957,
+      "grad_norm": 1.384270191192627,
+      "learning_rate": 4.8550472213036505e-06,
+      "loss": 1.1024,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6998797457481533,
+      "grad_norm": 1.4288511276245117,
+      "learning_rate": 4.854818756142982e-06,
+      "loss": 1.0809,
+      "step": 1358
+    },
+    {
+      "epoch": 0.7003951211132108,
+      "grad_norm": 1.397618293762207,
+      "learning_rate": 4.85459011646215e-06,
+      "loss": 1.1454,
+      "step": 1359
+    },
+    {
+      "epoch": 0.7009104964782683,
+      "grad_norm": 1.3595001697540283,
+      "learning_rate": 4.8543613022781e-06,
+      "loss": 1.1873,
+      "step": 1360
+    },
+    {
+      "epoch": 0.7014258718433259,
+      "grad_norm": 1.2345764636993408,
+      "learning_rate": 4.854132313607789e-06,
+      "loss": 1.0707,
+      "step": 1361
+    },
+    {
+      "epoch": 0.7019412472083835,
+      "grad_norm": 1.4134441614151,
+      "learning_rate": 4.853903150468187e-06,
+      "loss": 1.1097,
+      "step": 1362
+    },
+    {
+      "epoch": 0.702456622573441,
+      "grad_norm": 1.4602175951004028,
+      "learning_rate": 4.85367381287628e-06,
+      "loss": 1.1177,
+      "step": 1363
+    },
+    {
+      "epoch": 0.7029719979384985,
+      "grad_norm": 1.4819403886795044,
+      "learning_rate": 4.853444300849064e-06,
+      "loss": 1.1511,
+      "step": 1364
+    },
+    {
+      "epoch": 0.703487373303556,
+      "grad_norm": 1.4806100130081177,
+      "learning_rate": 4.853214614403546e-06,
+      "loss": 1.0773,
+      "step": 1365
+    },
+    {
+      "epoch": 0.7040027486686137,
+      "grad_norm": 1.4940937757492065,
+      "learning_rate": 4.852984753556752e-06,
+      "loss": 1.0999,
+      "step": 1366
+    },
+    {
+      "epoch": 0.7045181240336712,
+      "grad_norm": 1.3197970390319824,
+      "learning_rate": 4.852754718325715e-06,
+      "loss": 1.0316,
+      "step": 1367
+    },
+    {
+      "epoch": 0.7050334993987287,
+      "grad_norm": 1.3321506977081299,
+      "learning_rate": 4.852524508727485e-06,
+      "loss": 1.0527,
+      "step": 1368
+    },
+    {
+      "epoch": 0.7055488747637862,
+      "grad_norm": 1.4333229064941406,
+      "learning_rate": 4.8522941247791225e-06,
+      "loss": 1.0176,
+      "step": 1369
+    },
+    {
+      "epoch": 0.7060642501288439,
+      "grad_norm": 1.4624855518341064,
+      "learning_rate": 4.852063566497701e-06,
+      "loss": 1.1335,
+      "step": 1370
+    },
+    {
+      "epoch": 0.7065796254939014,
+      "grad_norm": 1.4005401134490967,
+      "learning_rate": 4.851832833900309e-06,
+      "loss": 1.0841,
+      "step": 1371
+    },
+    {
+      "epoch": 0.7070950008589589,
+      "grad_norm": 1.5284771919250488,
+      "learning_rate": 4.851601927004045e-06,
+      "loss": 1.1569,
+      "step": 1372
+    },
+    {
+      "epoch": 0.7076103762240165,
+      "grad_norm": 1.4555931091308594,
+      "learning_rate": 4.851370845826023e-06,
+      "loss": 1.0373,
+      "step": 1373
+    },
+    {
+      "epoch": 0.7081257515890741,
+      "grad_norm": 1.3138989210128784,
+      "learning_rate": 4.8511395903833695e-06,
+      "loss": 1.0617,
+      "step": 1374
+    },
+    {
+      "epoch": 0.7086411269541316,
+      "grad_norm": 1.4344639778137207,
+      "learning_rate": 4.850908160693222e-06,
+      "loss": 1.0887,
+      "step": 1375
+    },
+    {
+      "epoch": 0.7091565023191891,
+      "grad_norm": 1.4773739576339722,
+      "learning_rate": 4.850676556772733e-06,
+      "loss": 1.1262,
+      "step": 1376
+    },
+    {
+      "epoch": 0.7096718776842467,
+      "grad_norm": 1.3824067115783691,
+      "learning_rate": 4.850444778639067e-06,
+      "loss": 1.1232,
+      "step": 1377
+    },
+    {
+      "epoch": 0.7101872530493042,
+      "grad_norm": 1.4284874200820923,
+      "learning_rate": 4.8502128263094e-06,
+      "loss": 1.0868,
+      "step": 1378
+    },
+    {
+      "epoch": 0.7107026284143618,
+      "grad_norm": 1.380581021308899,
+      "learning_rate": 4.849980699800926e-06,
+      "loss": 1.0432,
+      "step": 1379
+    },
+    {
+      "epoch": 0.7112180037794193,
+      "grad_norm": 1.4592028856277466,
+      "learning_rate": 4.849748399130845e-06,
+      "loss": 1.054,
+      "step": 1380
+    },
+    {
+      "epoch": 0.7117333791444769,
+      "grad_norm": 1.3985048532485962,
+      "learning_rate": 4.849515924316373e-06,
+      "loss": 1.0763,
+      "step": 1381
+    },
+    {
+      "epoch": 0.7122487545095344,
+      "grad_norm": 1.3862051963806152,
+      "learning_rate": 4.849283275374742e-06,
+      "loss": 1.0463,
+      "step": 1382
+    },
+    {
+      "epoch": 0.712764129874592,
+      "grad_norm": 1.4439160823822021,
+      "learning_rate": 4.849050452323193e-06,
+      "loss": 1.0953,
+      "step": 1383
+    },
+    {
+      "epoch": 0.7132795052396496,
+      "grad_norm": 1.487596035003662,
+      "learning_rate": 4.848817455178979e-06,
+      "loss": 1.0759,
+      "step": 1384
+    },
+    {
+      "epoch": 0.7137948806047071,
+      "grad_norm": 1.5926995277404785,
+      "learning_rate": 4.848584283959371e-06,
+      "loss": 1.0944,
+      "step": 1385
+    },
+    {
+      "epoch": 0.7143102559697646,
+      "grad_norm": 1.4446570873260498,
+      "learning_rate": 4.848350938681648e-06,
+      "loss": 1.0602,
+      "step": 1386
+    },
+    {
+      "epoch": 0.7148256313348222,
+      "grad_norm": 1.3776057958602905,
+      "learning_rate": 4.848117419363103e-06,
+      "loss": 1.0768,
+      "step": 1387
+    },
+    {
+      "epoch": 0.7153410066998798,
+      "grad_norm": 1.321755051612854,
+      "learning_rate": 4.847883726021044e-06,
+      "loss": 1.0443,
+      "step": 1388
+    },
+    {
+      "epoch": 0.7158563820649373,
+      "grad_norm": 1.3724958896636963,
+      "learning_rate": 4.847649858672789e-06,
+      "loss": 1.0562,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7163717574299948,
+      "grad_norm": 1.5155479907989502,
+      "learning_rate": 4.847415817335672e-06,
+      "loss": 1.1133,
+      "step": 1390
+    },
+    {
+      "epoch": 0.7168871327950525,
+      "grad_norm": 1.4411449432373047,
+      "learning_rate": 4.847181602027038e-06,
+      "loss": 1.105,
+      "step": 1391
+    },
+    {
+      "epoch": 0.71740250816011,
+      "grad_norm": 1.482319951057434,
+      "learning_rate": 4.846947212764244e-06,
+      "loss": 1.096,
+      "step": 1392
+    },
+    {
+      "epoch": 0.7179178835251675,
+      "grad_norm": 1.427704095840454,
+      "learning_rate": 4.846712649564662e-06,
+      "loss": 1.0962,
+      "step": 1393
+    },
+    {
+      "epoch": 0.718433258890225,
+      "grad_norm": 1.517991542816162,
+      "learning_rate": 4.846477912445675e-06,
+      "loss": 1.1519,
+      "step": 1394
+    },
+    {
+      "epoch": 0.7189486342552825,
+      "grad_norm": 1.4028595685958862,
+      "learning_rate": 4.846243001424681e-06,
+      "loss": 1.0597,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7194640096203402,
+      "grad_norm": 1.4641873836517334,
+      "learning_rate": 4.846007916519089e-06,
+      "loss": 1.1123,
+      "step": 1396
+    },
+    {
+      "epoch": 0.7199793849853977,
+      "grad_norm": 1.4246588945388794,
+      "learning_rate": 4.845772657746321e-06,
+      "loss": 1.1262,
+      "step": 1397
+    },
+    {
+      "epoch": 0.7204947603504552,
+      "grad_norm": 1.4003269672393799,
+      "learning_rate": 4.8455372251238145e-06,
+      "loss": 1.1096,
+      "step": 1398
+    },
+    {
+      "epoch": 0.7210101357155128,
+      "grad_norm": 1.3927829265594482,
+      "learning_rate": 4.845301618669017e-06,
+      "loss": 1.0745,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7215255110805704,
+      "grad_norm": 1.4084463119506836,
+      "learning_rate": 4.845065838399388e-06,
+      "loss": 1.0459,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7220408864456279,
+      "grad_norm": 1.4316151142120361,
+      "learning_rate": 4.844829884332404e-06,
+      "loss": 1.1075,
+      "step": 1401
+    },
+    {
+      "epoch": 0.7225562618106854,
+      "grad_norm": 1.3410732746124268,
+      "learning_rate": 4.84459375648555e-06,
+      "loss": 1.1076,
+      "step": 1402
+    },
+    {
+      "epoch": 0.723071637175743,
+      "grad_norm": 1.3912278413772583,
+      "learning_rate": 4.844357454876327e-06,
+      "loss": 1.0403,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7235870125408006,
+      "grad_norm": 1.3922302722930908,
+      "learning_rate": 4.844120979522248e-06,
+      "loss": 1.1087,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7241023879058581,
+      "grad_norm": 1.449554204940796,
+      "learning_rate": 4.843884330440839e-06,
+      "loss": 1.0056,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7246177632709156,
+      "grad_norm": 1.482605218887329,
+      "learning_rate": 4.843647507649637e-06,
+      "loss": 1.1308,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7251331386359732,
+      "grad_norm": 1.3908926248550415,
+      "learning_rate": 4.843410511166194e-06,
+      "loss": 1.0587,
+      "step": 1407
+    },
+    {
+      "epoch": 0.7256485140010307,
+      "grad_norm": 1.4425791501998901,
+      "learning_rate": 4.843173341008075e-06,
+      "loss": 1.0651,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7261638893660883,
+      "grad_norm": 1.3466830253601074,
+      "learning_rate": 4.8429359971928566e-06,
+      "loss": 1.0018,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7266792647311459,
+      "grad_norm": 1.416321873664856,
+      "learning_rate": 4.8426984797381286e-06,
+      "loss": 1.0695,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7271946400962034,
+      "grad_norm": 1.4909610748291016,
+      "learning_rate": 4.842460788661494e-06,
+      "loss": 1.0843,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7277100154612609,
+      "grad_norm": 1.4182647466659546,
+      "learning_rate": 4.842222923980569e-06,
+      "loss": 1.1163,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7282253908263185,
+      "grad_norm": 1.4103379249572754,
+      "learning_rate": 4.841984885712981e-06,
+      "loss": 1.0438,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7287407661913761,
+      "grad_norm": 1.3778151273727417,
+      "learning_rate": 4.841746673876373e-06,
+      "loss": 1.0951,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7292561415564336,
+      "grad_norm": 1.4006578922271729,
+      "learning_rate": 4.841508288488399e-06,
+      "loss": 1.0421,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7297715169214911,
+      "grad_norm": 1.3837188482284546,
+      "learning_rate": 4.8412697295667255e-06,
+      "loss": 1.0583,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7302868922865487,
+      "grad_norm": 1.331186294555664,
+      "learning_rate": 4.841030997129033e-06,
+      "loss": 1.0416,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7308022676516063,
+      "grad_norm": 1.5883468389511108,
+      "learning_rate": 4.840792091193014e-06,
+      "loss": 1.159,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7313176430166638,
+      "grad_norm": 1.3742607831954956,
+      "learning_rate": 4.840553011776376e-06,
+      "loss": 1.0589,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7318330183817213,
+      "grad_norm": 1.5150750875473022,
+      "learning_rate": 4.8403137588968345e-06,
+      "loss": 1.1202,
+      "step": 1420
+    },
+    {
+      "epoch": 0.732348393746779,
+      "grad_norm": 1.5002319812774658,
+      "learning_rate": 4.8400743325721234e-06,
+      "loss": 1.0265,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7328637691118365,
+      "grad_norm": 1.386014461517334,
+      "learning_rate": 4.839834732819987e-06,
+      "loss": 1.1345,
+      "step": 1422
+    },
+    {
+      "epoch": 0.733379144476894,
+      "grad_norm": 1.4680486917495728,
+      "learning_rate": 4.839594959658181e-06,
+      "loss": 1.1227,
+      "step": 1423
+    },
+    {
+      "epoch": 0.7338945198419515,
+      "grad_norm": 1.4484657049179077,
+      "learning_rate": 4.839355013104477e-06,
+      "loss": 1.0728,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7344098952070091,
+      "grad_norm": 1.460443139076233,
+      "learning_rate": 4.839114893176658e-06,
+      "loss": 1.1153,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7349252705720667,
+      "grad_norm": 1.4195655584335327,
+      "learning_rate": 4.838874599892517e-06,
+      "loss": 1.04,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7354406459371242,
+      "grad_norm": 1.4055482149124146,
+      "learning_rate": 4.838634133269866e-06,
+      "loss": 1.0857,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7359560213021817,
+      "grad_norm": 1.3839490413665771,
+      "learning_rate": 4.8383934933265246e-06,
+      "loss": 1.0826,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7364713966672393,
+      "grad_norm": 1.41427743434906,
+      "learning_rate": 4.838152680080328e-06,
+      "loss": 1.0739,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7369867720322969,
+      "grad_norm": 1.4746673107147217,
+      "learning_rate": 4.837911693549122e-06,
+      "loss": 1.1009,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7375021473973544,
+      "grad_norm": 1.4033443927764893,
+      "learning_rate": 4.837670533750769e-06,
+      "loss": 1.101,
+      "step": 1431
+    },
+    {
+      "epoch": 0.738017522762412,
+      "grad_norm": 1.5204118490219116,
+      "learning_rate": 4.837429200703139e-06,
+      "loss": 1.1172,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7385328981274695,
+      "grad_norm": 1.424597978591919,
+      "learning_rate": 4.837187694424119e-06,
+      "loss": 1.1488,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7390482734925271,
+      "grad_norm": 1.4700113534927368,
+      "learning_rate": 4.836946014931608e-06,
+      "loss": 1.1524,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7395636488575846,
+      "grad_norm": 1.5409232378005981,
+      "learning_rate": 4.836704162243517e-06,
+      "loss": 1.1443,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7400790242226422,
+      "grad_norm": 1.4130326509475708,
+      "learning_rate": 4.836462136377769e-06,
+      "loss": 1.0748,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7405943995876997,
+      "grad_norm": 1.3824483156204224,
+      "learning_rate": 4.836219937352302e-06,
+      "loss": 1.0329,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7411097749527572,
+      "grad_norm": 1.4915496110916138,
+      "learning_rate": 4.835977565185067e-06,
+      "loss": 1.1141,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7416251503178148,
+      "grad_norm": 1.395885944366455,
+      "learning_rate": 4.835735019894025e-06,
+      "loss": 1.0768,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7421405256828724,
+      "grad_norm": 1.3711423873901367,
+      "learning_rate": 4.835492301497151e-06,
+      "loss": 1.0872,
+      "step": 1440
+    },
+    {
+      "epoch": 0.7426559010479299,
+      "grad_norm": 1.3521068096160889,
+      "learning_rate": 4.835249410012435e-06,
+      "loss": 1.0662,
+      "step": 1441
+    },
+    {
+      "epoch": 0.7431712764129874,
+      "grad_norm": 1.5525802373886108,
+      "learning_rate": 4.835006345457878e-06,
+      "loss": 1.0982,
+      "step": 1442
+    },
+    {
+      "epoch": 0.743686651778045,
+      "grad_norm": 1.4747167825698853,
+      "learning_rate": 4.8347631078514925e-06,
+      "loss": 1.0946,
+      "step": 1443
+    },
+    {
+      "epoch": 0.7442020271431026,
+      "grad_norm": 1.5162389278411865,
+      "learning_rate": 4.834519697211307e-06,
+      "loss": 1.0752,
+      "step": 1444
+    },
+    {
+      "epoch": 0.7447174025081601,
+      "grad_norm": 1.4368155002593994,
+      "learning_rate": 4.834276113555359e-06,
+      "loss": 1.0341,
+      "step": 1445
+    },
+    {
+      "epoch": 0.7452327778732176,
+      "grad_norm": 1.3954429626464844,
+      "learning_rate": 4.834032356901704e-06,
+      "loss": 1.086,
+      "step": 1446
+    },
+    {
+      "epoch": 0.7457481532382753,
+      "grad_norm": 1.3964476585388184,
+      "learning_rate": 4.833788427268404e-06,
+      "loss": 1.0735,
+      "step": 1447
+    },
+    {
+      "epoch": 0.7462635286033328,
+      "grad_norm": 1.421526551246643,
+      "learning_rate": 4.833544324673539e-06,
+      "loss": 1.0678,
+      "step": 1448
+    },
+    {
+      "epoch": 0.7467789039683903,
+      "grad_norm": 1.5207988023757935,
+      "learning_rate": 4.8333000491352005e-06,
+      "loss": 1.1284,
+      "step": 1449
+    },
+    {
+      "epoch": 0.7472942793334478,
+      "grad_norm": 1.478915810585022,
+      "learning_rate": 4.8330556006714915e-06,
+      "loss": 1.0822,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7478096546985055,
+      "grad_norm": 1.372940182685852,
+      "learning_rate": 4.832810979300527e-06,
+      "loss": 1.0774,
+      "step": 1451
+    },
+    {
+      "epoch": 0.748325030063563,
+      "grad_norm": 1.3675280809402466,
+      "learning_rate": 4.83256618504044e-06,
+      "loss": 1.1038,
+      "step": 1452
+    },
+    {
+      "epoch": 0.7488404054286205,
+      "grad_norm": 1.4146181344985962,
+      "learning_rate": 4.832321217909368e-06,
+      "loss": 1.1329,
+      "step": 1453
+    },
+    {
+      "epoch": 0.749355780793678,
+      "grad_norm": 1.511335015296936,
+      "learning_rate": 4.832076077925469e-06,
+      "loss": 1.0523,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7498711561587356,
+      "grad_norm": 1.5131245851516724,
+      "learning_rate": 4.8318307651069105e-06,
+      "loss": 1.0709,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7503865315237932,
+      "grad_norm": 1.3569769859313965,
+      "learning_rate": 4.831585279471873e-06,
+      "loss": 1.0382,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7509019068888507,
+      "grad_norm": 1.4782321453094482,
+      "learning_rate": 4.83133962103855e-06,
+      "loss": 1.0544,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7514172822539082,
+      "grad_norm": 1.3547550439834595,
+      "learning_rate": 4.831093789825147e-06,
+      "loss": 1.0903,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7519326576189658,
+      "grad_norm": 1.5269999504089355,
+      "learning_rate": 4.830847785849884e-06,
+      "loss": 1.1192,
+      "step": 1459
+    },
+    {
+      "epoch": 0.7524480329840234,
+      "grad_norm": 1.4924741983413696,
+      "learning_rate": 4.830601609130991e-06,
+      "loss": 1.0927,
+      "step": 1460
+    },
+    {
+      "epoch": 0.7529634083490809,
+      "grad_norm": 1.3824459314346313,
+      "learning_rate": 4.830355259686715e-06,
+      "loss": 1.0975,
+      "step": 1461
+    },
+    {
+      "epoch": 0.7534787837141385,
+      "grad_norm": 1.3687019348144531,
+      "learning_rate": 4.8301087375353125e-06,
+      "loss": 1.0942,
+      "step": 1462
+    },
+    {
+      "epoch": 0.753994159079196,
+      "grad_norm": 1.36841881275177,
+      "learning_rate": 4.829862042695053e-06,
+      "loss": 1.1201,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7545095344442536,
+      "grad_norm": 1.3933489322662354,
+      "learning_rate": 4.82961517518422e-06,
+      "loss": 1.0588,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7550249098093111,
+      "grad_norm": 1.4302362203598022,
+      "learning_rate": 4.829368135021111e-06,
+      "loss": 1.0806,
+      "step": 1465
+    },
+    {
+      "epoch": 0.7555402851743687,
+      "grad_norm": 1.4782177209854126,
+      "learning_rate": 4.829120922224031e-06,
+      "loss": 1.1152,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7560556605394262,
+      "grad_norm": 1.412581205368042,
+      "learning_rate": 4.828873536811305e-06,
+      "loss": 1.0746,
+      "step": 1467
+    },
+    {
+      "epoch": 0.7565710359044837,
+      "grad_norm": 1.4294135570526123,
+      "learning_rate": 4.828625978801264e-06,
+      "loss": 1.045,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7570864112695413,
+      "grad_norm": 1.5693269968032837,
+      "learning_rate": 4.828378248212259e-06,
+      "loss": 1.102,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7576017866345989,
+      "grad_norm": 1.4787834882736206,
+      "learning_rate": 4.828130345062646e-06,
+      "loss": 1.0813,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7581171619996564,
+      "grad_norm": 1.4717901945114136,
+      "learning_rate": 4.8278822693708e-06,
+      "loss": 1.0525,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7586325373647139,
+      "grad_norm": 1.3925107717514038,
+      "learning_rate": 4.827634021155104e-06,
+      "loss": 1.0394,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7591479127297716,
+      "grad_norm": 1.4415829181671143,
+      "learning_rate": 4.827385600433959e-06,
+      "loss": 1.046,
+      "step": 1473
+    },
+    {
+      "epoch": 0.7596632880948291,
+      "grad_norm": 1.4588543176651,
+      "learning_rate": 4.827137007225774e-06,
+      "loss": 1.062,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7601786634598866,
+      "grad_norm": 1.485809564590454,
+      "learning_rate": 4.826888241548973e-06,
+      "loss": 1.1431,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7606940388249441,
+      "grad_norm": 1.4579215049743652,
+      "learning_rate": 4.826639303421993e-06,
+      "loss": 1.0524,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7612094141900018,
+      "grad_norm": 1.5078346729278564,
+      "learning_rate": 4.826390192863283e-06,
+      "loss": 1.0672,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7617247895550593,
+      "grad_norm": 1.4479871988296509,
+      "learning_rate": 4.826140909891306e-06,
+      "loss": 1.1443,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7622401649201168,
+      "grad_norm": 1.5223277807235718,
+      "learning_rate": 4.825891454524535e-06,
+      "loss": 1.1458,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7627555402851743,
+      "grad_norm": 1.4447087049484253,
+      "learning_rate": 4.825641826781459e-06,
+      "loss": 1.1803,
+      "step": 1480
+    },
+    {
+      "epoch": 0.763270915650232,
+      "grad_norm": 1.4929414987564087,
+      "learning_rate": 4.8253920266805775e-06,
+      "loss": 1.1128,
+      "step": 1481
+    },
+    {
+      "epoch": 0.7637862910152895,
+      "grad_norm": 1.4813355207443237,
+      "learning_rate": 4.825142054240405e-06,
+      "loss": 1.0554,
+      "step": 1482
+    },
+    {
+      "epoch": 0.764301666380347,
+      "grad_norm": 1.3813947439193726,
+      "learning_rate": 4.824891909479466e-06,
+      "loss": 1.0799,
+      "step": 1483
+    },
+    {
+      "epoch": 0.7648170417454045,
+      "grad_norm": 1.278801679611206,
+      "learning_rate": 4.824641592416299e-06,
+      "loss": 1.0465,
+      "step": 1484
+    },
+    {
+      "epoch": 0.7653324171104621,
+      "grad_norm": 1.435333490371704,
+      "learning_rate": 4.824391103069457e-06,
+      "loss": 1.1328,
+      "step": 1485
+    },
+    {
+      "epoch": 0.7658477924755197,
+      "grad_norm": 1.3309084177017212,
+      "learning_rate": 4.824140441457504e-06,
+      "loss": 1.0975,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7663631678405772,
+      "grad_norm": 1.4422417879104614,
+      "learning_rate": 4.823889607599016e-06,
+      "loss": 1.0134,
+      "step": 1487
+    },
+    {
+      "epoch": 0.7668785432056348,
+      "grad_norm": 1.4464055299758911,
+      "learning_rate": 4.823638601512583e-06,
+      "loss": 1.1186,
+      "step": 1488
+    },
+    {
+      "epoch": 0.7673939185706923,
+      "grad_norm": 1.434095859527588,
+      "learning_rate": 4.823387423216809e-06,
+      "loss": 1.1073,
+      "step": 1489
+    },
+    {
+      "epoch": 0.7679092939357499,
+      "grad_norm": 1.3983433246612549,
+      "learning_rate": 4.823136072730308e-06,
+      "loss": 1.1088,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7684246693008074,
+      "grad_norm": 1.3363049030303955,
+      "learning_rate": 4.822884550071707e-06,
+      "loss": 1.0794,
+      "step": 1491
+    },
+    {
+      "epoch": 0.768940044665865,
+      "grad_norm": 1.4010666608810425,
+      "learning_rate": 4.822632855259649e-06,
+      "loss": 1.0241,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7694554200309225,
+      "grad_norm": 1.3678712844848633,
+      "learning_rate": 4.822380988312787e-06,
+      "loss": 1.0472,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7699707953959801,
+      "grad_norm": 1.4276469945907593,
+      "learning_rate": 4.8221289492497865e-06,
+      "loss": 1.0723,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7704861707610376,
+      "grad_norm": 1.6227951049804688,
+      "learning_rate": 4.821876738089327e-06,
+      "loss": 1.0553,
+      "step": 1495
+    },
+    {
+      "epoch": 0.7710015461260952,
+      "grad_norm": 1.3535913228988647,
+      "learning_rate": 4.821624354850101e-06,
+      "loss": 1.0506,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7715169214911527,
+      "grad_norm": 1.3324251174926758,
+      "learning_rate": 4.821371799550812e-06,
+      "loss": 1.0405,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7720322968562102,
+      "grad_norm": 1.3958076238632202,
+      "learning_rate": 4.821119072210178e-06,
+      "loss": 1.0549,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7725476722212679,
+      "grad_norm": 1.536328911781311,
+      "learning_rate": 4.820866172846929e-06,
+      "loss": 1.1138,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7730630475863254,
+      "grad_norm": 1.4695860147476196,
+      "learning_rate": 4.820613101479809e-06,
+      "loss": 1.1443,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7735784229513829,
+      "grad_norm": 1.4227354526519775,
+      "learning_rate": 4.820359858127571e-06,
+      "loss": 1.048,
+      "step": 1501
+    },
+    {
+      "epoch": 0.7740937983164404,
+      "grad_norm": 1.4998966455459595,
+      "learning_rate": 4.820106442808985e-06,
+      "loss": 1.1246,
+      "step": 1502
+    },
+    {
+      "epoch": 0.7746091736814981,
+      "grad_norm": 1.337685465812683,
+      "learning_rate": 4.819852855542833e-06,
+      "loss": 1.0523,
+      "step": 1503
+    },
+    {
+      "epoch": 0.7751245490465556,
+      "grad_norm": 1.4763025045394897,
+      "learning_rate": 4.8195990963479074e-06,
+      "loss": 1.0989,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7756399244116131,
+      "grad_norm": 1.3346480131149292,
+      "learning_rate": 4.819345165243015e-06,
+      "loss": 1.0289,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7761552997766706,
+      "grad_norm": 1.4553526639938354,
+      "learning_rate": 4.8190910622469755e-06,
+      "loss": 1.1123,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7766706751417283,
+      "grad_norm": 1.4475520849227905,
+      "learning_rate": 4.81883678737862e-06,
+      "loss": 1.0793,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7771860505067858,
+      "grad_norm": 1.3199334144592285,
+      "learning_rate": 4.818582340656796e-06,
+      "loss": 1.0676,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7777014258718433,
+      "grad_norm": 1.5657682418823242,
+      "learning_rate": 4.818327722100357e-06,
+      "loss": 1.0967,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7782168012369008,
+      "grad_norm": 1.5058013200759888,
+      "learning_rate": 4.818072931728176e-06,
+      "loss": 1.1101,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7787321766019585,
+      "grad_norm": 1.399551272392273,
+      "learning_rate": 4.817817969559137e-06,
+      "loss": 1.0789,
+      "step": 1511
+    },
+    {
+      "epoch": 0.779247551967016,
+      "grad_norm": 1.3868203163146973,
+      "learning_rate": 4.8175628356121315e-06,
+      "loss": 1.0851,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7797629273320735,
+      "grad_norm": 1.445791244506836,
+      "learning_rate": 4.817307529906072e-06,
+      "loss": 1.0933,
+      "step": 1513
+    },
+    {
+      "epoch": 0.780278302697131,
+      "grad_norm": 1.4244306087493896,
+      "learning_rate": 4.817052052459879e-06,
+      "loss": 1.0602,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7807936780621886,
+      "grad_norm": 1.3899831771850586,
+      "learning_rate": 4.816796403292485e-06,
+      "loss": 1.1358,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7813090534272462,
+      "grad_norm": 1.469082236289978,
+      "learning_rate": 4.816540582422838e-06,
+      "loss": 1.0909,
+      "step": 1516
+    },
+    {
+      "epoch": 0.7818244287923037,
+      "grad_norm": 1.4377481937408447,
+      "learning_rate": 4.816284589869895e-06,
+      "loss": 1.0468,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7823398041573613,
+      "grad_norm": 1.5348589420318604,
+      "learning_rate": 4.8160284256526315e-06,
+      "loss": 1.0998,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7828551795224188,
+      "grad_norm": 1.4987494945526123,
+      "learning_rate": 4.81577208979003e-06,
+      "loss": 1.1229,
+      "step": 1519
+    },
+    {
+      "epoch": 0.7833705548874764,
+      "grad_norm": 1.4241549968719482,
+      "learning_rate": 4.815515582301089e-06,
+      "loss": 1.0814,
+      "step": 1520
+    },
+    {
+      "epoch": 0.783885930252534,
+      "grad_norm": 1.3505808115005493,
+      "learning_rate": 4.815258903204818e-06,
+      "loss": 1.0634,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7844013056175915,
+      "grad_norm": 1.5867079496383667,
+      "learning_rate": 4.815002052520242e-06,
+      "loss": 1.0748,
+      "step": 1522
+    },
+    {
+      "epoch": 0.784916680982649,
+      "grad_norm": 1.5646687746047974,
+      "learning_rate": 4.8147450302663935e-06,
+      "loss": 1.0922,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7854320563477066,
+      "grad_norm": 1.3784677982330322,
+      "learning_rate": 4.814487836462323e-06,
+      "loss": 1.1321,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7859474317127642,
+      "grad_norm": 1.3318524360656738,
+      "learning_rate": 4.814230471127092e-06,
+      "loss": 1.0415,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7864628070778217,
+      "grad_norm": 1.3590197563171387,
+      "learning_rate": 4.813972934279773e-06,
+      "loss": 1.1184,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7869781824428792,
+      "grad_norm": 1.5370324850082397,
+      "learning_rate": 4.8137152259394525e-06,
+      "loss": 1.0759,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7874935578079367,
+      "grad_norm": 1.4619084596633911,
+      "learning_rate": 4.813457346125231e-06,
+      "loss": 1.1344,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7880089331729944,
+      "grad_norm": 1.4561723470687866,
+      "learning_rate": 4.81319929485622e-06,
+      "loss": 1.0823,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7885243085380519,
+      "grad_norm": 1.4673832654953003,
+      "learning_rate": 4.8129410721515435e-06,
+      "loss": 1.119,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7890396839031094,
+      "grad_norm": 1.5363526344299316,
+      "learning_rate": 4.8126826780303405e-06,
+      "loss": 1.1164,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7895550592681669,
+      "grad_norm": 1.4449119567871094,
+      "learning_rate": 4.812424112511759e-06,
+      "loss": 1.0176,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7900704346332246,
+      "grad_norm": 1.3174402713775635,
+      "learning_rate": 4.8121653756149634e-06,
+      "loss": 1.069,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7905858099982821,
+      "grad_norm": 1.4082295894622803,
+      "learning_rate": 4.811906467359128e-06,
+      "loss": 1.0688,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7911011853633396,
+      "grad_norm": 1.4260579347610474,
+      "learning_rate": 4.811647387763442e-06,
+      "loss": 1.1151,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7916165607283971,
+      "grad_norm": 1.3793349266052246,
+      "learning_rate": 4.811388136847106e-06,
+      "loss": 1.0744,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7921319360934548,
+      "grad_norm": 1.3310152292251587,
+      "learning_rate": 4.811128714629333e-06,
+      "loss": 1.103,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7926473114585123,
+      "grad_norm": 1.4189320802688599,
+      "learning_rate": 4.8108691211293505e-06,
+      "loss": 1.0989,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7931626868235698,
+      "grad_norm": 1.5097674131393433,
+      "learning_rate": 4.810609356366396e-06,
+      "loss": 1.1754,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7936780621886274,
+      "grad_norm": 1.4355213642120361,
+      "learning_rate": 4.810349420359722e-06,
+      "loss": 1.097,
+      "step": 1540
+    },
+    {
+      "epoch": 0.794193437553685,
+      "grad_norm": 1.5523383617401123,
+      "learning_rate": 4.810089313128593e-06,
+      "loss": 1.1276,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7947088129187425,
+      "grad_norm": 1.373019814491272,
+      "learning_rate": 4.809829034692285e-06,
+      "loss": 1.0699,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7952241882838,
+      "grad_norm": 1.4765411615371704,
+      "learning_rate": 4.809568585070089e-06,
+      "loss": 1.0658,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7957395636488576,
+      "grad_norm": 1.4936842918395996,
+      "learning_rate": 4.809307964281308e-06,
+      "loss": 1.0711,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7962549390139151,
+      "grad_norm": 1.3967301845550537,
+      "learning_rate": 4.8090471723452545e-06,
+      "loss": 1.1055,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7967703143789727,
+      "grad_norm": 1.496781826019287,
+      "learning_rate": 4.808786209281259e-06,
+      "loss": 1.1327,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7972856897440302,
+      "grad_norm": 1.3763643503189087,
+      "learning_rate": 4.80852507510866e-06,
+      "loss": 1.0183,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7978010651090878,
+      "grad_norm": 1.4482959508895874,
+      "learning_rate": 4.808263769846811e-06,
+      "loss": 1.0594,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7983164404741453,
+      "grad_norm": 1.4853146076202393,
+      "learning_rate": 4.808002293515079e-06,
+      "loss": 1.1297,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7988318158392029,
+      "grad_norm": 1.4585527181625366,
+      "learning_rate": 4.807740646132843e-06,
+      "loss": 1.1218,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7993471912042605,
+      "grad_norm": 1.5468567609786987,
+      "learning_rate": 4.8074788277194915e-06,
+      "loss": 1.0809,
+      "step": 1551
+    },
+    {
+      "epoch": 0.799862566569318,
+      "grad_norm": 1.6054692268371582,
+      "learning_rate": 4.807216838294431e-06,
+      "loss": 1.1209,
+      "step": 1552
+    },
+    {
+      "epoch": 0.8003779419343755,
+      "grad_norm": 1.4000954627990723,
+      "learning_rate": 4.806954677877076e-06,
+      "loss": 1.0575,
+      "step": 1553
+    },
+    {
+      "epoch": 0.8008933172994331,
+      "grad_norm": 1.486785650253296,
+      "learning_rate": 4.806692346486857e-06,
+      "loss": 1.0333,
+      "step": 1554
+    },
+    {
+      "epoch": 0.8014086926644907,
+      "grad_norm": 1.4417297840118408,
+      "learning_rate": 4.806429844143215e-06,
+      "loss": 1.1086,
+      "step": 1555
+    },
+    {
+      "epoch": 0.8019240680295482,
+      "grad_norm": 1.4330593347549438,
+      "learning_rate": 4.806167170865606e-06,
+      "loss": 1.0953,
+      "step": 1556
+    },
+    {
+      "epoch": 0.8024394433946057,
+      "grad_norm": 1.5242398977279663,
+      "learning_rate": 4.805904326673496e-06,
+      "loss": 1.081,
+      "step": 1557
+    },
+    {
+      "epoch": 0.8029548187596632,
+      "grad_norm": 1.4185315370559692,
+      "learning_rate": 4.805641311586365e-06,
+      "loss": 1.0364,
+      "step": 1558
+    },
+    {
+      "epoch": 0.8034701941247209,
+      "grad_norm": 1.4745197296142578,
+      "learning_rate": 4.805378125623707e-06,
+      "loss": 1.1042,
+      "step": 1559
+    },
+    {
+      "epoch": 0.8039855694897784,
+      "grad_norm": 1.4390727281570435,
+      "learning_rate": 4.805114768805024e-06,
+      "loss": 1.1048,
+      "step": 1560
+    },
+    {
+      "epoch": 0.8045009448548359,
+      "grad_norm": 1.3734118938446045,
+      "learning_rate": 4.804851241149837e-06,
+      "loss": 1.0185,
+      "step": 1561
+    },
+    {
+      "epoch": 0.8050163202198934,
+      "grad_norm": 1.424613356590271,
+      "learning_rate": 4.804587542677675e-06,
+      "loss": 1.1485,
+      "step": 1562
+    },
+    {
+      "epoch": 0.8055316955849511,
+      "grad_norm": 1.3926194906234741,
+      "learning_rate": 4.8043236734080815e-06,
+      "loss": 1.0809,
+      "step": 1563
+    },
+    {
+      "epoch": 0.8060470709500086,
+      "grad_norm": 1.406784176826477,
+      "learning_rate": 4.804059633360612e-06,
+      "loss": 1.0953,
+      "step": 1564
+    },
+    {
+      "epoch": 0.8065624463150661,
+      "grad_norm": 1.4212043285369873,
+      "learning_rate": 4.803795422554837e-06,
+      "loss": 1.1341,
+      "step": 1565
+    },
+    {
+      "epoch": 0.8070778216801237,
+      "grad_norm": 1.4541586637496948,
+      "learning_rate": 4.803531041010336e-06,
+      "loss": 1.0635,
+      "step": 1566
+    },
+    {
+      "epoch": 0.8075931970451813,
+      "grad_norm": 1.3861113786697388,
+      "learning_rate": 4.803266488746702e-06,
+      "loss": 1.0407,
+      "step": 1567
+    },
+    {
+      "epoch": 0.8081085724102388,
+      "grad_norm": 1.4320156574249268,
+      "learning_rate": 4.8030017657835435e-06,
+      "loss": 1.1112,
+      "step": 1568
+    },
+    {
+      "epoch": 0.8086239477752963,
+      "grad_norm": 1.3842825889587402,
+      "learning_rate": 4.8027368721404776e-06,
+      "loss": 1.0811,
+      "step": 1569
+    },
+    {
+      "epoch": 0.8091393231403539,
+      "grad_norm": 1.461434245109558,
+      "learning_rate": 4.802471807837137e-06,
+      "loss": 1.1129,
+      "step": 1570
+    },
+    {
+      "epoch": 0.8096546985054115,
+      "grad_norm": 1.4281977415084839,
+      "learning_rate": 4.8022065728931675e-06,
+      "loss": 1.1268,
+      "step": 1571
+    },
+    {
+      "epoch": 0.810170073870469,
+      "grad_norm": 1.3471062183380127,
+      "learning_rate": 4.801941167328223e-06,
+      "loss": 1.0179,
+      "step": 1572
+    },
+    {
+      "epoch": 0.8106854492355265,
+      "grad_norm": 1.4150283336639404,
+      "learning_rate": 4.8016755911619775e-06,
+      "loss": 1.0745,
+      "step": 1573
+    },
+    {
+      "epoch": 0.8112008246005841,
+      "grad_norm": 1.344197392463684,
+      "learning_rate": 4.8014098444141085e-06,
+      "loss": 1.1219,
+      "step": 1574
+    },
+    {
+      "epoch": 0.8117161999656416,
+      "grad_norm": 1.3955225944519043,
+      "learning_rate": 4.801143927104315e-06,
+      "loss": 1.0709,
+      "step": 1575
+    },
+    {
+      "epoch": 0.8122315753306992,
+      "grad_norm": 1.495924472808838,
+      "learning_rate": 4.800877839252302e-06,
+      "loss": 1.0954,
+      "step": 1576
+    },
+    {
+      "epoch": 0.8127469506957568,
+      "grad_norm": 1.537859320640564,
+      "learning_rate": 4.8006115808777924e-06,
+      "loss": 1.0299,
+      "step": 1577
+    },
+    {
+      "epoch": 0.8132623260608143,
+      "grad_norm": 1.3771979808807373,
+      "learning_rate": 4.800345152000517e-06,
+      "loss": 1.1131,
+      "step": 1578
+    },
+    {
+      "epoch": 0.8137777014258718,
+      "grad_norm": 1.3602322340011597,
+      "learning_rate": 4.8000785526402215e-06,
+      "loss": 1.0479,
+      "step": 1579
+    },
+    {
+      "epoch": 0.8142930767909294,
+      "grad_norm": 1.5317022800445557,
+      "learning_rate": 4.799811782816665e-06,
+      "loss": 1.0893,
+      "step": 1580
+    },
+    {
+      "epoch": 0.814808452155987,
+      "grad_norm": 1.4737638235092163,
+      "learning_rate": 4.799544842549618e-06,
+      "loss": 1.0801,
+      "step": 1581
+    },
+    {
+      "epoch": 0.8153238275210445,
+      "grad_norm": 1.4101364612579346,
+      "learning_rate": 4.799277731858863e-06,
+      "loss": 1.0891,
+      "step": 1582
+    },
+    {
+      "epoch": 0.815839202886102,
+      "grad_norm": 1.4352830648422241,
+      "learning_rate": 4.799010450764198e-06,
+      "loss": 1.0226,
+      "step": 1583
+    },
+    {
+      "epoch": 0.8163545782511596,
+      "grad_norm": 1.4002106189727783,
+      "learning_rate": 4.798742999285431e-06,
+      "loss": 1.0875,
+      "step": 1584
+    },
+    {
+      "epoch": 0.8168699536162172,
+      "grad_norm": 1.5443047285079956,
+      "learning_rate": 4.798475377442381e-06,
+      "loss": 1.1324,
+      "step": 1585
+    },
+    {
+      "epoch": 0.8173853289812747,
+      "grad_norm": 1.4441038370132446,
+      "learning_rate": 4.798207585254886e-06,
+      "loss": 1.0994,
+      "step": 1586
+    },
+    {
+      "epoch": 0.8179007043463322,
+      "grad_norm": 1.542736530303955,
+      "learning_rate": 4.79793962274279e-06,
+      "loss": 1.1252,
+      "step": 1587
+    },
+    {
+      "epoch": 0.8184160797113897,
+      "grad_norm": 1.426893711090088,
+      "learning_rate": 4.797671489925953e-06,
+      "loss": 1.1229,
+      "step": 1588
+    },
+    {
+      "epoch": 0.8189314550764474,
+      "grad_norm": 1.498072624206543,
+      "learning_rate": 4.7974031868242456e-06,
+      "loss": 1.0909,
+      "step": 1589
+    },
+    {
+      "epoch": 0.8194468304415049,
+      "grad_norm": 1.4439469575881958,
+      "learning_rate": 4.797134713457554e-06,
+      "loss": 1.0871,
+      "step": 1590
+    },
+    {
+      "epoch": 0.8199622058065624,
+      "grad_norm": 1.5463494062423706,
+      "learning_rate": 4.7968660698457745e-06,
+      "loss": 1.1358,
+      "step": 1591
+    },
+    {
+      "epoch": 0.82047758117162,
+      "grad_norm": 1.5635348558425903,
+      "learning_rate": 4.796597256008816e-06,
+      "loss": 1.0658,
+      "step": 1592
+    },
+    {
+      "epoch": 0.8209929565366776,
+      "grad_norm": 1.3745373487472534,
+      "learning_rate": 4.796328271966603e-06,
+      "loss": 1.0555,
+      "step": 1593
+    },
+    {
+      "epoch": 0.8215083319017351,
+      "grad_norm": 1.3798346519470215,
+      "learning_rate": 4.796059117739069e-06,
+      "loss": 1.1016,
+      "step": 1594
+    },
+    {
+      "epoch": 0.8220237072667926,
+      "grad_norm": 1.4389573335647583,
+      "learning_rate": 4.795789793346161e-06,
+      "loss": 1.1273,
+      "step": 1595
+    },
+    {
+      "epoch": 0.8225390826318502,
+      "grad_norm": 1.5225825309753418,
+      "learning_rate": 4.7955202988078405e-06,
+      "loss": 1.1082,
+      "step": 1596
+    },
+    {
+      "epoch": 0.8230544579969078,
+      "grad_norm": 1.3396738767623901,
+      "learning_rate": 4.7952506341440785e-06,
+      "loss": 1.0616,
+      "step": 1597
+    },
+    {
+      "epoch": 0.8235698333619653,
+      "grad_norm": 1.537752389907837,
+      "learning_rate": 4.794980799374862e-06,
+      "loss": 1.128,
+      "step": 1598
+    },
+    {
+      "epoch": 0.8240852087270228,
+      "grad_norm": 1.5089632272720337,
+      "learning_rate": 4.794710794520188e-06,
+      "loss": 1.1146,
+      "step": 1599
+    },
+    {
+      "epoch": 0.8246005840920804,
+      "grad_norm": 1.4799798727035522,
+      "learning_rate": 4.794440619600069e-06,
+      "loss": 1.063,
+      "step": 1600
+    },
+    {
+      "epoch": 0.825115959457138,
+      "grad_norm": 1.39742910861969,
+      "learning_rate": 4.794170274634525e-06,
+      "loss": 1.0339,
+      "step": 1601
+    },
+    {
+      "epoch": 0.8256313348221955,
+      "grad_norm": 1.3962360620498657,
+      "learning_rate": 4.793899759643595e-06,
+      "loss": 1.088,
+      "step": 1602
+    },
+    {
+      "epoch": 0.826146710187253,
+      "grad_norm": 1.3832945823669434,
+      "learning_rate": 4.793629074647325e-06,
+      "loss": 1.1044,
+      "step": 1603
+    },
+    {
+      "epoch": 0.8266620855523106,
+      "grad_norm": 1.3732894659042358,
+      "learning_rate": 4.793358219665777e-06,
+      "loss": 1.0417,
+      "step": 1604
+    },
+    {
+      "epoch": 0.8271774609173681,
+      "grad_norm": 1.4401576519012451,
+      "learning_rate": 4.793087194719024e-06,
+      "loss": 1.0932,
+      "step": 1605
+    },
+    {
+      "epoch": 0.8276928362824257,
+      "grad_norm": 1.4844118356704712,
+      "learning_rate": 4.7928159998271524e-06,
+      "loss": 1.0269,
+      "step": 1606
+    },
+    {
+      "epoch": 0.8282082116474833,
+      "grad_norm": 1.522064447402954,
+      "learning_rate": 4.792544635010262e-06,
+      "loss": 1.0943,
+      "step": 1607
+    },
+    {
+      "epoch": 0.8287235870125408,
+      "grad_norm": 1.42251718044281,
+      "learning_rate": 4.792273100288462e-06,
+      "loss": 1.0776,
+      "step": 1608
+    },
+    {
+      "epoch": 0.8292389623775983,
+      "grad_norm": 1.3952099084854126,
+      "learning_rate": 4.792001395681879e-06,
+      "loss": 1.0534,
+      "step": 1609
+    },
+    {
+      "epoch": 0.829754337742656,
+      "grad_norm": 1.3191391229629517,
+      "learning_rate": 4.791729521210649e-06,
+      "loss": 1.0532,
+      "step": 1610
+    },
+    {
+      "epoch": 0.8302697131077135,
+      "grad_norm": 1.4547145366668701,
+      "learning_rate": 4.791457476894918e-06,
+      "loss": 1.0893,
+      "step": 1611
+    },
+    {
+      "epoch": 0.830785088472771,
+      "grad_norm": 1.436697006225586,
+      "learning_rate": 4.791185262754852e-06,
+      "loss": 1.0766,
+      "step": 1612
+    },
+    {
+      "epoch": 0.8313004638378285,
+      "grad_norm": 1.414806842803955,
+      "learning_rate": 4.790912878810623e-06,
+      "loss": 1.0948,
+      "step": 1613
+    },
+    {
+      "epoch": 0.8318158392028862,
+      "grad_norm": 1.393926739692688,
+      "learning_rate": 4.790640325082419e-06,
+      "loss": 1.0085,
+      "step": 1614
+    },
+    {
+      "epoch": 0.8323312145679437,
+      "grad_norm": 1.4000269174575806,
+      "learning_rate": 4.7903676015904375e-06,
+      "loss": 0.9886,
+      "step": 1615
+    },
+    {
+      "epoch": 0.8328465899330012,
+      "grad_norm": 1.448590636253357,
+      "learning_rate": 4.790094708354892e-06,
+      "loss": 1.0801,
+      "step": 1616
+    },
+    {
+      "epoch": 0.8333619652980587,
+      "grad_norm": 1.398142695426941,
+      "learning_rate": 4.789821645396008e-06,
+      "loss": 1.1023,
+      "step": 1617
+    },
+    {
+      "epoch": 0.8338773406631163,
+      "grad_norm": 1.4157236814498901,
+      "learning_rate": 4.789548412734021e-06,
+      "loss": 1.0309,
+      "step": 1618
+    },
+    {
+      "epoch": 0.8343927160281739,
+      "grad_norm": 1.4708627462387085,
+      "learning_rate": 4.789275010389182e-06,
+      "loss": 1.0756,
+      "step": 1619
+    },
+    {
+      "epoch": 0.8349080913932314,
+      "grad_norm": 1.467354655265808,
+      "learning_rate": 4.789001438381753e-06,
+      "loss": 1.0863,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8354234667582889,
+      "grad_norm": 1.4737015962600708,
+      "learning_rate": 4.7887276967320085e-06,
+      "loss": 1.114,
+      "step": 1621
+    },
+    {
+      "epoch": 0.8359388421233465,
+      "grad_norm": 1.3399767875671387,
+      "learning_rate": 4.7884537854602366e-06,
+      "loss": 1.0471,
+      "step": 1622
+    },
+    {
+      "epoch": 0.8364542174884041,
+      "grad_norm": 1.4572374820709229,
+      "learning_rate": 4.788179704586737e-06,
+      "loss": 1.1347,
+      "step": 1623
+    },
+    {
+      "epoch": 0.8369695928534616,
+      "grad_norm": 1.3600082397460938,
+      "learning_rate": 4.787905454131824e-06,
+      "loss": 1.0226,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8374849682185191,
+      "grad_norm": 1.410685420036316,
+      "learning_rate": 4.787631034115819e-06,
+      "loss": 1.04,
+      "step": 1625
+    },
+    {
+      "epoch": 0.8380003435835767,
+      "grad_norm": 1.3994461297988892,
+      "learning_rate": 4.787356444559064e-06,
+      "loss": 1.0858,
+      "step": 1626
+    },
+    {
+      "epoch": 0.8385157189486343,
+      "grad_norm": 1.431602120399475,
+      "learning_rate": 4.7870816854819065e-06,
+      "loss": 1.1271,
+      "step": 1627
+    },
+    {
+      "epoch": 0.8390310943136918,
+      "grad_norm": 1.6482824087142944,
+      "learning_rate": 4.786806756904712e-06,
+      "loss": 1.1201,
+      "step": 1628
+    },
+    {
+      "epoch": 0.8395464696787494,
+      "grad_norm": 1.5163174867630005,
+      "learning_rate": 4.786531658847853e-06,
+      "loss": 1.0999,
+      "step": 1629
+    },
+    {
+      "epoch": 0.8400618450438069,
+      "grad_norm": 1.414298415184021,
+      "learning_rate": 4.786256391331719e-06,
+      "loss": 1.1207,
+      "step": 1630
+    },
+    {
+      "epoch": 0.8405772204088645,
+      "grad_norm": 1.457314372062683,
+      "learning_rate": 4.785980954376711e-06,
+      "loss": 1.0699,
+      "step": 1631
+    },
+    {
+      "epoch": 0.841092595773922,
+      "grad_norm": 1.3886388540267944,
+      "learning_rate": 4.7857053480032426e-06,
+      "loss": 1.0496,
+      "step": 1632
+    },
+    {
+      "epoch": 0.8416079711389796,
+      "grad_norm": 1.5067790746688843,
+      "learning_rate": 4.785429572231738e-06,
+      "loss": 1.0683,
+      "step": 1633
+    },
+    {
+      "epoch": 0.8421233465040371,
+      "grad_norm": 1.4911561012268066,
+      "learning_rate": 4.785153627082636e-06,
+      "loss": 1.0831,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8426387218690946,
+      "grad_norm": 1.4155385494232178,
+      "learning_rate": 4.784877512576388e-06,
+      "loss": 1.0909,
+      "step": 1635
+    },
+    {
+      "epoch": 0.8431540972341522,
+      "grad_norm": 1.4064384698867798,
+      "learning_rate": 4.7846012287334565e-06,
+      "loss": 1.0895,
+      "step": 1636
+    },
+    {
+      "epoch": 0.8436694725992098,
+      "grad_norm": 1.4730241298675537,
+      "learning_rate": 4.784324775574318e-06,
+      "loss": 1.0602,
+      "step": 1637
+    },
+    {
+      "epoch": 0.8441848479642673,
+      "grad_norm": 1.4730726480484009,
+      "learning_rate": 4.78404815311946e-06,
+      "loss": 1.0693,
+      "step": 1638
+    },
+    {
+      "epoch": 0.8447002233293248,
+      "grad_norm": 1.4139877557754517,
+      "learning_rate": 4.783771361389387e-06,
+      "loss": 1.0198,
+      "step": 1639
+    },
+    {
+      "epoch": 0.8452155986943825,
+      "grad_norm": 1.3953264951705933,
+      "learning_rate": 4.783494400404608e-06,
+      "loss": 1.0857,
+      "step": 1640
+    },
+    {
+      "epoch": 0.84573097405944,
+      "grad_norm": 1.3684098720550537,
+      "learning_rate": 4.783217270185651e-06,
+      "loss": 0.9952,
+      "step": 1641
+    },
+    {
+      "epoch": 0.8462463494244975,
+      "grad_norm": 1.4131075143814087,
+      "learning_rate": 4.782939970753056e-06,
+      "loss": 1.0336,
+      "step": 1642
+    },
+    {
+      "epoch": 0.846761724789555,
+      "grad_norm": 1.4862349033355713,
+      "learning_rate": 4.782662502127372e-06,
+      "loss": 1.0847,
+      "step": 1643
+    },
+    {
+      "epoch": 0.8472771001546127,
+      "grad_norm": 1.4583141803741455,
+      "learning_rate": 4.782384864329164e-06,
+      "loss": 1.0507,
+      "step": 1644
+    },
+    {
+      "epoch": 0.8477924755196702,
+      "grad_norm": 1.4953395128250122,
+      "learning_rate": 4.782107057379008e-06,
+      "loss": 1.1075,
+      "step": 1645
+    },
+    {
+      "epoch": 0.8483078508847277,
+      "grad_norm": 1.4994615316390991,
+      "learning_rate": 4.781829081297493e-06,
+      "loss": 1.1372,
+      "step": 1646
+    },
+    {
+      "epoch": 0.8488232262497852,
+      "grad_norm": 1.4864728450775146,
+      "learning_rate": 4.781550936105219e-06,
+      "loss": 1.055,
+      "step": 1647
+    },
+    {
+      "epoch": 0.8493386016148428,
+      "grad_norm": 1.4858756065368652,
+      "learning_rate": 4.781272621822803e-06,
+      "loss": 1.0852,
+      "step": 1648
+    },
+    {
+      "epoch": 0.8498539769799004,
+      "grad_norm": 1.4812358617782593,
+      "learning_rate": 4.780994138470868e-06,
+      "loss": 1.0363,
+      "step": 1649
+    },
+    {
+      "epoch": 0.8503693523449579,
+      "grad_norm": 1.4463202953338623,
+      "learning_rate": 4.780715486070054e-06,
+      "loss": 1.0709,
+      "step": 1650
+    },
+    {
+      "epoch": 0.8508847277100154,
+      "grad_norm": 1.392630934715271,
+      "learning_rate": 4.780436664641014e-06,
+      "loss": 1.0162,
+      "step": 1651
+    },
+    {
+      "epoch": 0.851400103075073,
+      "grad_norm": 1.4439646005630493,
+      "learning_rate": 4.78015767420441e-06,
+      "loss": 1.0674,
+      "step": 1652
+    },
+    {
+      "epoch": 0.8519154784401306,
+      "grad_norm": 1.5503273010253906,
+      "learning_rate": 4.779878514780919e-06,
+      "loss": 1.0988,
+      "step": 1653
+    },
+    {
+      "epoch": 0.8524308538051881,
+      "grad_norm": 1.379058837890625,
+      "learning_rate": 4.779599186391232e-06,
+      "loss": 1.0799,
+      "step": 1654
+    },
+    {
+      "epoch": 0.8529462291702457,
+      "grad_norm": 1.4692100286483765,
+      "learning_rate": 4.779319689056047e-06,
+      "loss": 1.0233,
+      "step": 1655
+    },
+    {
+      "epoch": 0.8534616045353032,
+      "grad_norm": 1.5064520835876465,
+      "learning_rate": 4.7790400227960806e-06,
+      "loss": 1.0766,
+      "step": 1656
+    },
+    {
+      "epoch": 0.8539769799003608,
+      "grad_norm": 1.5419743061065674,
+      "learning_rate": 4.77876018763206e-06,
+      "loss": 1.0705,
+      "step": 1657
+    },
+    {
+      "epoch": 0.8544923552654183,
+      "grad_norm": 1.3527169227600098,
+      "learning_rate": 4.778480183584722e-06,
+      "loss": 1.0732,
+      "step": 1658
+    },
+    {
+      "epoch": 0.8550077306304759,
+      "grad_norm": 1.447250247001648,
+      "learning_rate": 4.778200010674819e-06,
+      "loss": 1.1307,
+      "step": 1659
+    },
+    {
+      "epoch": 0.8555231059955334,
+      "grad_norm": 1.3942347764968872,
+      "learning_rate": 4.777919668923117e-06,
+      "loss": 0.9875,
+      "step": 1660
+    },
+    {
+      "epoch": 0.856038481360591,
+      "grad_norm": 1.4627864360809326,
+      "learning_rate": 4.77763915835039e-06,
+      "loss": 1.0418,
+      "step": 1661
+    },
+    {
+      "epoch": 0.8565538567256485,
+      "grad_norm": 1.4622726440429688,
+      "learning_rate": 4.777358478977428e-06,
+      "loss": 1.0525,
+      "step": 1662
+    },
+    {
+      "epoch": 0.8570692320907061,
+      "grad_norm": 1.4471572637557983,
+      "learning_rate": 4.777077630825035e-06,
+      "loss": 1.0698,
+      "step": 1663
+    },
+    {
+      "epoch": 0.8575846074557636,
+      "grad_norm": 1.3427664041519165,
+      "learning_rate": 4.776796613914022e-06,
+      "loss": 1.0346,
+      "step": 1664
+    },
+    {
+      "epoch": 0.8580999828208211,
+      "grad_norm": 1.359036922454834,
+      "learning_rate": 4.7765154282652174e-06,
+      "loss": 1.106,
+      "step": 1665
+    },
+    {
+      "epoch": 0.8586153581858788,
+      "grad_norm": 1.426246166229248,
+      "learning_rate": 4.7762340738994596e-06,
+      "loss": 1.0604,
+      "step": 1666
+    },
+    {
+      "epoch": 0.8591307335509363,
+      "grad_norm": 1.5278077125549316,
+      "learning_rate": 4.775952550837601e-06,
+      "loss": 1.1139,
+      "step": 1667
+    },
+    {
+      "epoch": 0.8596461089159938,
+      "grad_norm": 1.4093070030212402,
+      "learning_rate": 4.775670859100505e-06,
+      "loss": 1.1157,
+      "step": 1668
+    },
+    {
+      "epoch": 0.8601614842810513,
+      "grad_norm": 1.4400460720062256,
+      "learning_rate": 4.77538899870905e-06,
+      "loss": 1.0485,
+      "step": 1669
+    },
+    {
+      "epoch": 0.860676859646109,
+      "grad_norm": 1.4249861240386963,
+      "learning_rate": 4.775106969684123e-06,
+      "loss": 1.046,
+      "step": 1670
+    },
+    {
+      "epoch": 0.8611922350111665,
+      "grad_norm": 1.6405653953552246,
+      "learning_rate": 4.774824772046627e-06,
+      "loss": 1.1025,
+      "step": 1671
+    },
+    {
+      "epoch": 0.861707610376224,
+      "grad_norm": 1.469918966293335,
+      "learning_rate": 4.774542405817476e-06,
+      "loss": 1.0906,
+      "step": 1672
+    },
+    {
+      "epoch": 0.8622229857412815,
+      "grad_norm": 1.4990814924240112,
+      "learning_rate": 4.774259871017597e-06,
+      "loss": 1.1187,
+      "step": 1673
+    },
+    {
+      "epoch": 0.8627383611063392,
+      "grad_norm": 1.4857546091079712,
+      "learning_rate": 4.773977167667929e-06,
+      "loss": 1.0131,
+      "step": 1674
+    },
+    {
+      "epoch": 0.8632537364713967,
+      "grad_norm": 1.5064942836761475,
+      "learning_rate": 4.773694295789423e-06,
+      "loss": 1.1068,
+      "step": 1675
+    },
+    {
+      "epoch": 0.8637691118364542,
+      "grad_norm": 1.4220467805862427,
+      "learning_rate": 4.773411255403044e-06,
+      "loss": 1.1122,
+      "step": 1676
+    },
+    {
+      "epoch": 0.8642844872015117,
+      "grad_norm": 1.460793375968933,
+      "learning_rate": 4.773128046529769e-06,
+      "loss": 1.0739,
+      "step": 1677
+    },
+    {
+      "epoch": 0.8647998625665693,
+      "grad_norm": 1.4518553018569946,
+      "learning_rate": 4.772844669190586e-06,
+      "loss": 1.0477,
+      "step": 1678
+    },
+    {
+      "epoch": 0.8653152379316269,
+      "grad_norm": 1.4068254232406616,
+      "learning_rate": 4.772561123406497e-06,
+      "loss": 1.11,
+      "step": 1679
+    },
+    {
+      "epoch": 0.8658306132966844,
+      "grad_norm": 1.4043121337890625,
+      "learning_rate": 4.772277409198517e-06,
+      "loss": 1.0531,
+      "step": 1680
+    },
+    {
+      "epoch": 0.866345988661742,
+      "grad_norm": 1.501981258392334,
+      "learning_rate": 4.771993526587671e-06,
+      "loss": 1.0452,
+      "step": 1681
+    },
+    {
+      "epoch": 0.8668613640267995,
+      "grad_norm": 1.3823513984680176,
+      "learning_rate": 4.771709475595e-06,
+      "loss": 1.1162,
+      "step": 1682
+    },
+    {
+      "epoch": 0.8673767393918571,
+      "grad_norm": 1.4772850275039673,
+      "learning_rate": 4.771425256241555e-06,
+      "loss": 1.0594,
+      "step": 1683
+    },
+    {
+      "epoch": 0.8678921147569146,
+      "grad_norm": 1.4475082159042358,
+      "learning_rate": 4.771140868548399e-06,
+      "loss": 1.102,
+      "step": 1684
+    },
+    {
+      "epoch": 0.8684074901219722,
+      "grad_norm": 1.3387587070465088,
+      "learning_rate": 4.770856312536609e-06,
+      "loss": 1.0464,
+      "step": 1685
+    },
+    {
+      "epoch": 0.8689228654870297,
+      "grad_norm": 1.4844287633895874,
+      "learning_rate": 4.770571588227275e-06,
+      "loss": 1.0448,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8694382408520873,
+      "grad_norm": 1.4855479001998901,
+      "learning_rate": 4.7702866956414974e-06,
+      "loss": 1.0607,
+      "step": 1687
+    },
+    {
+      "epoch": 0.8699536162171448,
+      "grad_norm": 1.3221168518066406,
+      "learning_rate": 4.77000163480039e-06,
+      "loss": 1.0648,
+      "step": 1688
+    },
+    {
+      "epoch": 0.8704689915822024,
+      "grad_norm": 1.4294415712356567,
+      "learning_rate": 4.76971640572508e-06,
+      "loss": 0.9887,
+      "step": 1689
+    },
+    {
+      "epoch": 0.8709843669472599,
+      "grad_norm": 1.4959137439727783,
+      "learning_rate": 4.769431008436705e-06,
+      "loss": 1.1318,
+      "step": 1690
+    },
+    {
+      "epoch": 0.8714997423123174,
+      "grad_norm": 1.5279430150985718,
+      "learning_rate": 4.769145442956418e-06,
+      "loss": 1.1354,
+      "step": 1691
+    },
+    {
+      "epoch": 0.872015117677375,
+      "grad_norm": 1.42916738986969,
+      "learning_rate": 4.768859709305382e-06,
+      "loss": 1.1105,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8725304930424326,
+      "grad_norm": 1.361976146697998,
+      "learning_rate": 4.768573807504773e-06,
+      "loss": 1.038,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8730458684074901,
+      "grad_norm": 1.5706747770309448,
+      "learning_rate": 4.76828773757578e-06,
+      "loss": 1.0973,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8735612437725476,
+      "grad_norm": 1.5899977684020996,
+      "learning_rate": 4.768001499539605e-06,
+      "loss": 1.0994,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8740766191376053,
+      "grad_norm": 1.4544364213943481,
+      "learning_rate": 4.767715093417461e-06,
+      "loss": 1.0912,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8745919945026628,
+      "grad_norm": 1.4816296100616455,
+      "learning_rate": 4.767428519230573e-06,
+      "loss": 1.1046,
+      "step": 1697
+    },
+    {
+      "epoch": 0.8751073698677203,
+      "grad_norm": 1.491078495979309,
+      "learning_rate": 4.767141777000181e-06,
+      "loss": 1.0394,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8756227452327778,
+      "grad_norm": 1.393487572669983,
+      "learning_rate": 4.766854866747536e-06,
+      "loss": 1.1015,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8761381205978355,
+      "grad_norm": 1.5540200471878052,
+      "learning_rate": 4.7665677884939005e-06,
+      "loss": 1.1323,
+      "step": 1700
+    },
+    {
+      "epoch": 0.876653495962893,
+      "grad_norm": 1.4467482566833496,
+      "learning_rate": 4.766280542260551e-06,
+      "loss": 1.1179,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8771688713279505,
+      "grad_norm": 1.3639161586761475,
+      "learning_rate": 4.765993128068776e-06,
+      "loss": 1.0567,
+      "step": 1702
+    },
+    {
+      "epoch": 0.877684246693008,
+      "grad_norm": 1.432669997215271,
+      "learning_rate": 4.765705545939876e-06,
+      "loss": 1.0641,
+      "step": 1703
+    },
+    {
+      "epoch": 0.8781996220580657,
+      "grad_norm": 1.479915976524353,
+      "learning_rate": 4.765417795895164e-06,
+      "loss": 1.0978,
+      "step": 1704
+    },
+    {
+      "epoch": 0.8787149974231232,
+      "grad_norm": 1.4626078605651855,
+      "learning_rate": 4.7651298779559675e-06,
+      "loss": 1.0788,
+      "step": 1705
+    },
+    {
+      "epoch": 0.8792303727881807,
+      "grad_norm": 1.5150612592697144,
+      "learning_rate": 4.7648417921436226e-06,
+      "loss": 1.1402,
+      "step": 1706
+    },
+    {
+      "epoch": 0.8797457481532382,
+      "grad_norm": 1.407073736190796,
+      "learning_rate": 4.764553538479481e-06,
+      "loss": 1.1194,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8802611235182958,
+      "grad_norm": 1.4320027828216553,
+      "learning_rate": 4.764265116984905e-06,
+      "loss": 1.1377,
+      "step": 1708
+    },
+    {
+      "epoch": 0.8807764988833534,
+      "grad_norm": 1.5515000820159912,
+      "learning_rate": 4.76397652768127e-06,
+      "loss": 1.0789,
+      "step": 1709
+    },
+    {
+      "epoch": 0.8812918742484109,
+      "grad_norm": 1.3981959819793701,
+      "learning_rate": 4.7636877705899645e-06,
+      "loss": 1.0753,
+      "step": 1710
+    },
+    {
+      "epoch": 0.8818072496134685,
+      "grad_norm": 1.3862941265106201,
+      "learning_rate": 4.7633988457323895e-06,
+      "loss": 1.1064,
+      "step": 1711
+    },
+    {
+      "epoch": 0.882322624978526,
+      "grad_norm": 1.468210220336914,
+      "learning_rate": 4.763109753129956e-06,
+      "loss": 1.062,
+      "step": 1712
+    },
+    {
+      "epoch": 0.8828380003435836,
+      "grad_norm": 1.433226227760315,
+      "learning_rate": 4.762820492804091e-06,
+      "loss": 1.0428,
+      "step": 1713
+    },
+    {
+      "epoch": 0.8833533757086411,
+      "grad_norm": 1.4374988079071045,
+      "learning_rate": 4.76253106477623e-06,
+      "loss": 1.0639,
+      "step": 1714
+    },
+    {
+      "epoch": 0.8838687510736987,
+      "grad_norm": 1.3978921175003052,
+      "learning_rate": 4.762241469067826e-06,
+      "loss": 1.0801,
+      "step": 1715
+    },
+    {
+      "epoch": 0.8843841264387562,
+      "grad_norm": 1.521583914756775,
+      "learning_rate": 4.761951705700339e-06,
+      "loss": 1.0827,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8848995018038138,
+      "grad_norm": 1.511265516281128,
+      "learning_rate": 4.761661774695244e-06,
+      "loss": 1.0729,
+      "step": 1717
+    },
+    {
+      "epoch": 0.8854148771688714,
+      "grad_norm": 1.364372968673706,
+      "learning_rate": 4.761371676074031e-06,
+      "loss": 1.0126,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8859302525339289,
+      "grad_norm": 1.4365894794464111,
+      "learning_rate": 4.761081409858197e-06,
+      "loss": 1.0979,
+      "step": 1719
+    },
+    {
+      "epoch": 0.8864456278989864,
+      "grad_norm": 1.5464420318603516,
+      "learning_rate": 4.760790976069255e-06,
+      "loss": 1.0434,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8869610032640439,
+      "grad_norm": 1.3821452856063843,
+      "learning_rate": 4.760500374728729e-06,
+      "loss": 1.0707,
+      "step": 1721
+    },
+    {
+      "epoch": 0.8874763786291016,
+      "grad_norm": 1.4921995401382446,
+      "learning_rate": 4.760209605858158e-06,
+      "loss": 1.0479,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8879917539941591,
+      "grad_norm": 1.4335943460464478,
+      "learning_rate": 4.759918669479089e-06,
+      "loss": 1.0297,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8885071293592166,
+      "grad_norm": 1.6402289867401123,
+      "learning_rate": 4.759627565613086e-06,
+      "loss": 1.0822,
+      "step": 1724
+    },
+    {
+      "epoch": 0.8890225047242741,
+      "grad_norm": 1.4306977987289429,
+      "learning_rate": 4.7593362942817215e-06,
+      "loss": 1.102,
+      "step": 1725
+    },
+    {
+      "epoch": 0.8895378800893318,
+      "grad_norm": 1.355398178100586,
+      "learning_rate": 4.759044855506584e-06,
+      "loss": 1.1356,
+      "step": 1726
+    },
+    {
+      "epoch": 0.8900532554543893,
+      "grad_norm": 1.5569396018981934,
+      "learning_rate": 4.7587532493092706e-06,
+      "loss": 1.0976,
+      "step": 1727
+    },
+    {
+      "epoch": 0.8905686308194468,
+      "grad_norm": 1.453972578048706,
+      "learning_rate": 4.7584614757113946e-06,
+      "loss": 1.0623,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8910840061845043,
+      "grad_norm": 1.3104867935180664,
+      "learning_rate": 4.758169534734579e-06,
+      "loss": 1.0573,
+      "step": 1729
+    },
+    {
+      "epoch": 0.891599381549562,
+      "grad_norm": 1.5181728601455688,
+      "learning_rate": 4.757877426400459e-06,
+      "loss": 1.1099,
+      "step": 1730
+    },
+    {
+      "epoch": 0.8921147569146195,
+      "grad_norm": 1.472532033920288,
+      "learning_rate": 4.757585150730686e-06,
+      "loss": 1.1179,
+      "step": 1731
+    },
+    {
+      "epoch": 0.892630132279677,
+      "grad_norm": 1.4582340717315674,
+      "learning_rate": 4.757292707746919e-06,
+      "loss": 1.0528,
+      "step": 1732
+    },
+    {
+      "epoch": 0.8931455076447345,
+      "grad_norm": 1.2936973571777344,
+      "learning_rate": 4.757000097470832e-06,
+      "loss": 1.0641,
+      "step": 1733
+    },
+    {
+      "epoch": 0.8936608830097922,
+      "grad_norm": 1.4027740955352783,
+      "learning_rate": 4.756707319924111e-06,
+      "loss": 1.0126,
+      "step": 1734
+    },
+    {
+      "epoch": 0.8941762583748497,
+      "grad_norm": 1.448681354522705,
+      "learning_rate": 4.756414375128454e-06,
+      "loss": 1.0498,
+      "step": 1735
+    },
+    {
+      "epoch": 0.8946916337399072,
+      "grad_norm": 1.3526110649108887,
+      "learning_rate": 4.756121263105573e-06,
+      "loss": 1.0887,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8952070091049648,
+      "grad_norm": 1.6526248455047607,
+      "learning_rate": 4.755827983877189e-06,
+      "loss": 1.0404,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8957223844700223,
+      "grad_norm": 1.5655080080032349,
+      "learning_rate": 4.75553453746504e-06,
+      "loss": 1.0874,
+      "step": 1738
+    },
+    {
+      "epoch": 0.8962377598350799,
+      "grad_norm": 1.5430996417999268,
+      "learning_rate": 4.755240923890871e-06,
+      "loss": 1.0711,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8967531352001374,
+      "grad_norm": 1.5913987159729004,
+      "learning_rate": 4.754947143176445e-06,
+      "loss": 1.146,
+      "step": 1740
+    },
+    {
+      "epoch": 0.897268510565195,
+      "grad_norm": 1.5870795249938965,
+      "learning_rate": 4.754653195343533e-06,
+      "loss": 1.0612,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8977838859302525,
+      "grad_norm": 1.4600235223770142,
+      "learning_rate": 4.754359080413921e-06,
+      "loss": 1.0459,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8982992612953101,
+      "grad_norm": 1.448747992515564,
+      "learning_rate": 4.754064798409406e-06,
+      "loss": 1.0574,
+      "step": 1743
+    },
+    {
+      "epoch": 0.8988146366603676,
+      "grad_norm": 1.4169598817825317,
+      "learning_rate": 4.7537703493517965e-06,
+      "loss": 1.0326,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8993300120254252,
+      "grad_norm": 1.4494582414627075,
+      "learning_rate": 4.753475733262917e-06,
+      "loss": 1.0229,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8998453873904827,
+      "grad_norm": 1.4525147676467896,
+      "learning_rate": 4.7531809501646e-06,
+      "loss": 1.047,
+      "step": 1746
+    },
+    {
+      "epoch": 0.9003607627555403,
+      "grad_norm": 1.3870949745178223,
+      "learning_rate": 4.752886000078694e-06,
+      "loss": 1.1006,
+      "step": 1747
+    },
+    {
+      "epoch": 0.9008761381205979,
+      "grad_norm": 1.5141352415084839,
+      "learning_rate": 4.752590883027058e-06,
+      "loss": 1.0681,
+      "step": 1748
+    },
+    {
+      "epoch": 0.9013915134856554,
+      "grad_norm": 1.396710991859436,
+      "learning_rate": 4.752295599031563e-06,
+      "loss": 1.1347,
+      "step": 1749
+    },
+    {
+      "epoch": 0.9019068888507129,
+      "grad_norm": 1.4481197595596313,
+      "learning_rate": 4.752000148114093e-06,
+      "loss": 1.0563,
+      "step": 1750
+    },
+    {
+      "epoch": 0.9024222642157704,
+      "grad_norm": 1.4166992902755737,
+      "learning_rate": 4.751704530296546e-06,
+      "loss": 1.0358,
+      "step": 1751
+    },
+    {
+      "epoch": 0.9029376395808281,
+      "grad_norm": 1.5104217529296875,
+      "learning_rate": 4.75140874560083e-06,
+      "loss": 1.066,
+      "step": 1752
+    },
+    {
+      "epoch": 0.9034530149458856,
+      "grad_norm": 1.3824926614761353,
+      "learning_rate": 4.751112794048864e-06,
+      "loss": 1.162,
+      "step": 1753
+    },
+    {
+      "epoch": 0.9039683903109431,
+      "grad_norm": 1.457301378250122,
+      "learning_rate": 4.750816675662585e-06,
+      "loss": 1.0805,
+      "step": 1754
+    },
+    {
+      "epoch": 0.9044837656760006,
+      "grad_norm": 1.503834843635559,
+      "learning_rate": 4.750520390463937e-06,
+      "loss": 1.052,
+      "step": 1755
+    },
+    {
+      "epoch": 0.9049991410410583,
+      "grad_norm": 1.5135822296142578,
+      "learning_rate": 4.75022393847488e-06,
+      "loss": 1.1013,
+      "step": 1756
+    },
+    {
+      "epoch": 0.9055145164061158,
+      "grad_norm": 1.4753880500793457,
+      "learning_rate": 4.7499273197173816e-06,
+      "loss": 1.0819,
+      "step": 1757
+    },
+    {
+      "epoch": 0.9060298917711733,
+      "grad_norm": 1.4338304996490479,
+      "learning_rate": 4.749630534213426e-06,
+      "loss": 1.0369,
+      "step": 1758
+    },
+    {
+      "epoch": 0.9065452671362308,
+      "grad_norm": 1.4980969429016113,
+      "learning_rate": 4.74933358198501e-06,
+      "loss": 1.0672,
+      "step": 1759
+    },
+    {
+      "epoch": 0.9070606425012885,
+      "grad_norm": 1.4337161779403687,
+      "learning_rate": 4.74903646305414e-06,
+      "loss": 1.0707,
+      "step": 1760
+    },
+    {
+      "epoch": 0.907576017866346,
+      "grad_norm": 1.5216634273529053,
+      "learning_rate": 4.748739177442837e-06,
+      "loss": 1.0928,
+      "step": 1761
+    },
+    {
+      "epoch": 0.9080913932314035,
+      "grad_norm": 1.4658452272415161,
+      "learning_rate": 4.748441725173132e-06,
+      "loss": 1.0317,
+      "step": 1762
+    },
+    {
+      "epoch": 0.9086067685964611,
+      "grad_norm": 1.3911515474319458,
+      "learning_rate": 4.748144106267071e-06,
+      "loss": 1.1078,
+      "step": 1763
+    },
+    {
+      "epoch": 0.9091221439615187,
+      "grad_norm": 1.3948339223861694,
+      "learning_rate": 4.74784632074671e-06,
+      "loss": 1.0886,
+      "step": 1764
+    },
+    {
+      "epoch": 0.9096375193265762,
+      "grad_norm": 1.3779579401016235,
+      "learning_rate": 4.747548368634119e-06,
+      "loss": 1.1183,
+      "step": 1765
+    },
+    {
+      "epoch": 0.9101528946916337,
+      "grad_norm": 1.5271098613739014,
+      "learning_rate": 4.747250249951381e-06,
+      "loss": 1.0797,
+      "step": 1766
+    },
+    {
+      "epoch": 0.9106682700566913,
+      "grad_norm": 1.4378695487976074,
+      "learning_rate": 4.746951964720589e-06,
+      "loss": 1.0775,
+      "step": 1767
+    },
+    {
+      "epoch": 0.9111836454217488,
+      "grad_norm": 1.438913345336914,
+      "learning_rate": 4.746653512963849e-06,
+      "loss": 1.0841,
+      "step": 1768
+    },
+    {
+      "epoch": 0.9116990207868064,
+      "grad_norm": 1.488240122795105,
+      "learning_rate": 4.74635489470328e-06,
+      "loss": 1.1125,
+      "step": 1769
+    },
+    {
+      "epoch": 0.912214396151864,
+      "grad_norm": 1.5185062885284424,
+      "learning_rate": 4.746056109961014e-06,
+      "loss": 1.0561,
+      "step": 1770
+    },
+    {
+      "epoch": 0.9127297715169215,
+      "grad_norm": 1.3826409578323364,
+      "learning_rate": 4.745757158759194e-06,
+      "loss": 1.1072,
+      "step": 1771
+    },
+    {
+      "epoch": 0.913245146881979,
+      "grad_norm": 1.4739314317703247,
+      "learning_rate": 4.745458041119976e-06,
+      "loss": 1.0743,
+      "step": 1772
+    },
+    {
+      "epoch": 0.9137605222470366,
+      "grad_norm": 1.4196566343307495,
+      "learning_rate": 4.7451587570655274e-06,
+      "loss": 1.0947,
+      "step": 1773
+    },
+    {
+      "epoch": 0.9142758976120942,
+      "grad_norm": 1.4553617238998413,
+      "learning_rate": 4.7448593066180305e-06,
+      "loss": 1.1062,
+      "step": 1774
+    },
+    {
+      "epoch": 0.9147912729771517,
+      "grad_norm": 1.4958308935165405,
+      "learning_rate": 4.744559689799677e-06,
+      "loss": 1.1129,
+      "step": 1775
+    },
+    {
+      "epoch": 0.9153066483422092,
+      "grad_norm": 1.470471739768982,
+      "learning_rate": 4.744259906632672e-06,
+      "loss": 1.0458,
+      "step": 1776
+    },
+    {
+      "epoch": 0.9158220237072668,
+      "grad_norm": 1.401932716369629,
+      "learning_rate": 4.743959957139233e-06,
+      "loss": 1.0178,
+      "step": 1777
+    },
+    {
+      "epoch": 0.9163373990723244,
+      "grad_norm": 1.4637651443481445,
+      "learning_rate": 4.74365984134159e-06,
+      "loss": 1.0887,
+      "step": 1778
+    },
+    {
+      "epoch": 0.9168527744373819,
+      "grad_norm": 1.4661166667938232,
+      "learning_rate": 4.743359559261985e-06,
+      "loss": 1.0175,
+      "step": 1779
+    },
+    {
+      "epoch": 0.9173681498024394,
+      "grad_norm": 1.4048773050308228,
+      "learning_rate": 4.7430591109226735e-06,
+      "loss": 1.0619,
+      "step": 1780
+    },
+    {
+      "epoch": 0.9178835251674969,
+      "grad_norm": 1.4620808362960815,
+      "learning_rate": 4.7427584963459206e-06,
+      "loss": 1.0759,
+      "step": 1781
+    },
+    {
+      "epoch": 0.9183989005325546,
+      "grad_norm": 1.4824326038360596,
+      "learning_rate": 4.742457715554006e-06,
+      "loss": 1.0613,
+      "step": 1782
+    },
+    {
+      "epoch": 0.9189142758976121,
+      "grad_norm": 1.4724524021148682,
+      "learning_rate": 4.742156768569223e-06,
+      "loss": 1.0845,
+      "step": 1783
+    },
+    {
+      "epoch": 0.9194296512626696,
+      "grad_norm": 1.4528087377548218,
+      "learning_rate": 4.741855655413871e-06,
+      "loss": 1.0805,
+      "step": 1784
+    },
+    {
+      "epoch": 0.9199450266277271,
+      "grad_norm": 1.5265142917633057,
+      "learning_rate": 4.741554376110272e-06,
+      "loss": 1.073,
+      "step": 1785
+    },
+    {
+      "epoch": 0.9204604019927848,
+      "grad_norm": 1.4664232730865479,
+      "learning_rate": 4.74125293068075e-06,
+      "loss": 1.052,
+      "step": 1786
+    },
+    {
+      "epoch": 0.9209757773578423,
+      "grad_norm": 1.5155519247055054,
+      "learning_rate": 4.7409513191476464e-06,
+      "loss": 1.0568,
+      "step": 1787
+    },
+    {
+      "epoch": 0.9214911527228998,
+      "grad_norm": 1.5170753002166748,
+      "learning_rate": 4.740649541533316e-06,
+      "loss": 1.1125,
+      "step": 1788
+    },
+    {
+      "epoch": 0.9220065280879574,
+      "grad_norm": 1.379411220550537,
+      "learning_rate": 4.740347597860121e-06,
+      "loss": 1.0134,
+      "step": 1789
+    },
+    {
+      "epoch": 0.922521903453015,
+      "grad_norm": 1.4462095499038696,
+      "learning_rate": 4.7400454881504426e-06,
+      "loss": 1.1216,
+      "step": 1790
+    },
+    {
+      "epoch": 0.9230372788180725,
+      "grad_norm": 1.4491463899612427,
+      "learning_rate": 4.7397432124266685e-06,
+      "loss": 1.0911,
+      "step": 1791
+    },
+    {
+      "epoch": 0.92355265418313,
+      "grad_norm": 1.4337164163589478,
+      "learning_rate": 4.739440770711202e-06,
+      "loss": 1.1264,
+      "step": 1792
+    },
+    {
+      "epoch": 0.9240680295481876,
+      "grad_norm": 1.3294486999511719,
+      "learning_rate": 4.739138163026457e-06,
+      "loss": 1.0702,
+      "step": 1793
+    },
+    {
+      "epoch": 0.9245834049132452,
+      "grad_norm": 1.4707309007644653,
+      "learning_rate": 4.7388353893948594e-06,
+      "loss": 1.0553,
+      "step": 1794
+    },
+    {
+      "epoch": 0.9250987802783027,
+      "grad_norm": 1.3893729448318481,
+      "learning_rate": 4.7385324498388505e-06,
+      "loss": 1.0428,
+      "step": 1795
+    },
+    {
+      "epoch": 0.9256141556433602,
+      "grad_norm": 1.4164751768112183,
+      "learning_rate": 4.738229344380881e-06,
+      "loss": 1.0856,
+      "step": 1796
+    },
+    {
+      "epoch": 0.9261295310084178,
+      "grad_norm": 1.6173774003982544,
+      "learning_rate": 4.737926073043413e-06,
+      "loss": 1.0157,
+      "step": 1797
+    },
+    {
+      "epoch": 0.9266449063734753,
+      "grad_norm": 1.326172113418579,
+      "learning_rate": 4.737622635848924e-06,
+      "loss": 1.0672,
+      "step": 1798
+    },
+    {
+      "epoch": 0.9271602817385329,
+      "grad_norm": 1.457281470298767,
+      "learning_rate": 4.7373190328199024e-06,
+      "loss": 1.1605,
+      "step": 1799
+    },
+    {
+      "epoch": 0.9276756571035905,
+      "grad_norm": 1.3793095350265503,
+      "learning_rate": 4.737015263978849e-06,
+      "loss": 1.0886,
+      "step": 1800
+    },
+    {
+      "epoch": 0.928191032468648,
+      "grad_norm": 1.5238492488861084,
+      "learning_rate": 4.736711329348276e-06,
+      "loss": 1.1237,
+      "step": 1801
+    },
+    {
+      "epoch": 0.9287064078337055,
+      "grad_norm": 1.3719145059585571,
+      "learning_rate": 4.7364072289507086e-06,
+      "loss": 1.0802,
+      "step": 1802
+    },
+    {
+      "epoch": 0.9292217831987631,
+      "grad_norm": 1.419771671295166,
+      "learning_rate": 4.736102962808685e-06,
+      "loss": 1.0647,
+      "step": 1803
+    },
+    {
+      "epoch": 0.9297371585638207,
+      "grad_norm": 1.4599947929382324,
+      "learning_rate": 4.7357985309447534e-06,
+      "loss": 1.0956,
+      "step": 1804
+    },
+    {
+      "epoch": 0.9302525339288782,
+      "grad_norm": 1.4967395067214966,
+      "learning_rate": 4.735493933381478e-06,
+      "loss": 1.0781,
+      "step": 1805
+    },
+    {
+      "epoch": 0.9307679092939357,
+      "grad_norm": 1.4744763374328613,
+      "learning_rate": 4.735189170141431e-06,
+      "loss": 1.0775,
+      "step": 1806
+    },
+    {
+      "epoch": 0.9312832846589933,
+      "grad_norm": 1.377068042755127,
+      "learning_rate": 4.734884241247201e-06,
+      "loss": 1.0297,
+      "step": 1807
+    },
+    {
+      "epoch": 0.9317986600240509,
+      "grad_norm": 1.4732213020324707,
+      "learning_rate": 4.734579146721385e-06,
+      "loss": 1.1569,
+      "step": 1808
+    },
+    {
+      "epoch": 0.9323140353891084,
+      "grad_norm": 1.48055100440979,
+      "learning_rate": 4.734273886586595e-06,
+      "loss": 1.0906,
+      "step": 1809
+    },
+    {
+      "epoch": 0.9328294107541659,
+      "grad_norm": 1.4947986602783203,
+      "learning_rate": 4.7339684608654545e-06,
+      "loss": 1.0595,
+      "step": 1810
+    },
+    {
+      "epoch": 0.9333447861192234,
+      "grad_norm": 1.4390573501586914,
+      "learning_rate": 4.7336628695805996e-06,
+      "loss": 1.005,
+      "step": 1811
+    },
+    {
+      "epoch": 0.9338601614842811,
+      "grad_norm": 1.4130473136901855,
+      "learning_rate": 4.733357112754677e-06,
+      "loss": 1.0602,
+      "step": 1812
+    },
+    {
+      "epoch": 0.9343755368493386,
+      "grad_norm": 1.4222996234893799,
+      "learning_rate": 4.733051190410348e-06,
+      "loss": 1.0702,
+      "step": 1813
+    },
+    {
+      "epoch": 0.9348909122143961,
+      "grad_norm": 1.394693374633789,
+      "learning_rate": 4.732745102570284e-06,
+      "loss": 1.0983,
+      "step": 1814
+    },
+    {
+      "epoch": 0.9354062875794537,
+      "grad_norm": 1.4363412857055664,
+      "learning_rate": 4.732438849257172e-06,
+      "loss": 1.1394,
+      "step": 1815
+    },
+    {
+      "epoch": 0.9359216629445113,
+      "grad_norm": 1.5439742803573608,
+      "learning_rate": 4.732132430493707e-06,
+      "loss": 1.144,
+      "step": 1816
+    },
+    {
+      "epoch": 0.9364370383095688,
+      "grad_norm": 1.386092185974121,
+      "learning_rate": 4.731825846302599e-06,
+      "loss": 0.9641,
+      "step": 1817
+    },
+    {
+      "epoch": 0.9369524136746263,
+      "grad_norm": 1.4146597385406494,
+      "learning_rate": 4.731519096706569e-06,
+      "loss": 1.0827,
+      "step": 1818
+    },
+    {
+      "epoch": 0.9374677890396839,
+      "grad_norm": 1.3727401494979858,
+      "learning_rate": 4.7312121817283515e-06,
+      "loss": 1.0087,
+      "step": 1819
+    },
+    {
+      "epoch": 0.9379831644047415,
+      "grad_norm": 1.4614678621292114,
+      "learning_rate": 4.730905101390692e-06,
+      "loss": 1.091,
+      "step": 1820
+    },
+    {
+      "epoch": 0.938498539769799,
+      "grad_norm": 1.3576834201812744,
+      "learning_rate": 4.730597855716349e-06,
+      "loss": 0.9969,
+      "step": 1821
+    },
+    {
+      "epoch": 0.9390139151348565,
+      "grad_norm": 1.4184889793395996,
+      "learning_rate": 4.7302904447280925e-06,
+      "loss": 1.084,
+      "step": 1822
+    },
+    {
+      "epoch": 0.9395292904999141,
+      "grad_norm": 1.560845971107483,
+      "learning_rate": 4.729982868448707e-06,
+      "loss": 1.0897,
+      "step": 1823
+    },
+    {
+      "epoch": 0.9400446658649717,
+      "grad_norm": 1.4771478176116943,
+      "learning_rate": 4.729675126900987e-06,
+      "loss": 1.0167,
+      "step": 1824
+    },
+    {
+      "epoch": 0.9405600412300292,
+      "grad_norm": 1.4902037382125854,
+      "learning_rate": 4.729367220107738e-06,
+      "loss": 1.058,
+      "step": 1825
+    },
+    {
+      "epoch": 0.9410754165950868,
+      "grad_norm": 1.3810451030731201,
+      "learning_rate": 4.729059148091782e-06,
+      "loss": 1.0945,
+      "step": 1826
+    },
+    {
+      "epoch": 0.9415907919601443,
+      "grad_norm": 1.4192668199539185,
+      "learning_rate": 4.7287509108759485e-06,
+      "loss": 1.106,
+      "step": 1827
+    },
+    {
+      "epoch": 0.9421061673252018,
+      "grad_norm": 1.463755488395691,
+      "learning_rate": 4.728442508483084e-06,
+      "loss": 1.0567,
+      "step": 1828
+    },
+    {
+      "epoch": 0.9426215426902594,
+      "grad_norm": 1.4863239526748657,
+      "learning_rate": 4.728133940936043e-06,
+      "loss": 1.076,
+      "step": 1829
+    },
+    {
+      "epoch": 0.943136918055317,
+      "grad_norm": 1.5006253719329834,
+      "learning_rate": 4.727825208257694e-06,
+      "loss": 1.06,
+      "step": 1830
+    },
+    {
+      "epoch": 0.9436522934203745,
+      "grad_norm": 1.5008097887039185,
+      "learning_rate": 4.72751631047092e-06,
+      "loss": 1.0828,
+      "step": 1831
+    },
+    {
+      "epoch": 0.944167668785432,
+      "grad_norm": 1.5320098400115967,
+      "learning_rate": 4.727207247598612e-06,
+      "loss": 1.0552,
+      "step": 1832
+    },
+    {
+      "epoch": 0.9446830441504896,
+      "grad_norm": 1.4904122352600098,
+      "learning_rate": 4.726898019663675e-06,
+      "loss": 1.104,
+      "step": 1833
+    },
+    {
+      "epoch": 0.9451984195155472,
+      "grad_norm": 1.4500375986099243,
+      "learning_rate": 4.726588626689027e-06,
+      "loss": 1.0646,
+      "step": 1834
+    },
+    {
+      "epoch": 0.9457137948806047,
+      "grad_norm": 1.4086084365844727,
+      "learning_rate": 4.726279068697598e-06,
+      "loss": 1.0219,
+      "step": 1835
+    },
+    {
+      "epoch": 0.9462291702456622,
+      "grad_norm": 1.483607530593872,
+      "learning_rate": 4.72596934571233e-06,
+      "loss": 1.0645,
+      "step": 1836
+    },
+    {
+      "epoch": 0.9467445456107199,
+      "grad_norm": 1.4772380590438843,
+      "learning_rate": 4.725659457756177e-06,
+      "loss": 1.0689,
+      "step": 1837
+    },
+    {
+      "epoch": 0.9472599209757774,
+      "grad_norm": 1.5030094385147095,
+      "learning_rate": 4.725349404852105e-06,
+      "loss": 1.02,
+      "step": 1838
+    },
+    {
+      "epoch": 0.9477752963408349,
+      "grad_norm": 1.500354528427124,
+      "learning_rate": 4.725039187023094e-06,
+      "loss": 1.054,
+      "step": 1839
+    },
+    {
+      "epoch": 0.9482906717058924,
+      "grad_norm": 1.4762042760849,
+      "learning_rate": 4.724728804292132e-06,
+      "loss": 1.1067,
+      "step": 1840
+    },
+    {
+      "epoch": 0.94880604707095,
+      "grad_norm": 1.3946943283081055,
+      "learning_rate": 4.724418256682226e-06,
+      "loss": 1.0548,
+      "step": 1841
+    },
+    {
+      "epoch": 0.9493214224360076,
+      "grad_norm": 1.484532117843628,
+      "learning_rate": 4.724107544216388e-06,
+      "loss": 1.1011,
+      "step": 1842
+    },
+    {
+      "epoch": 0.9498367978010651,
+      "grad_norm": 1.4259350299835205,
+      "learning_rate": 4.723796666917646e-06,
+      "loss": 1.1398,
+      "step": 1843
+    },
+    {
+      "epoch": 0.9503521731661226,
+      "grad_norm": 1.4509137868881226,
+      "learning_rate": 4.723485624809042e-06,
+      "loss": 1.0666,
+      "step": 1844
+    },
+    {
+      "epoch": 0.9508675485311802,
+      "grad_norm": 1.504083275794983,
+      "learning_rate": 4.723174417913625e-06,
+      "loss": 1.0515,
+      "step": 1845
+    },
+    {
+      "epoch": 0.9513829238962378,
+      "grad_norm": 1.4487136602401733,
+      "learning_rate": 4.722863046254461e-06,
+      "loss": 1.047,
+      "step": 1846
+    },
+    {
+      "epoch": 0.9518982992612953,
+      "grad_norm": 1.3897000551223755,
+      "learning_rate": 4.7225515098546246e-06,
+      "loss": 1.0678,
+      "step": 1847
+    },
+    {
+      "epoch": 0.9524136746263528,
+      "grad_norm": 1.6226310729980469,
+      "learning_rate": 4.722239808737207e-06,
+      "loss": 1.0792,
+      "step": 1848
+    },
+    {
+      "epoch": 0.9529290499914104,
+      "grad_norm": 1.4630800485610962,
+      "learning_rate": 4.721927942925308e-06,
+      "loss": 1.0854,
+      "step": 1849
+    },
+    {
+      "epoch": 0.953444425356468,
+      "grad_norm": 1.4378085136413574,
+      "learning_rate": 4.7216159124420385e-06,
+      "loss": 1.1144,
+      "step": 1850
+    },
+    {
+      "epoch": 0.9539598007215255,
+      "grad_norm": 1.448045253753662,
+      "learning_rate": 4.7213037173105255e-06,
+      "loss": 1.0519,
+      "step": 1851
+    },
+    {
+      "epoch": 0.954475176086583,
+      "grad_norm": 1.4926683902740479,
+      "learning_rate": 4.720991357553907e-06,
+      "loss": 1.0104,
+      "step": 1852
+    },
+    {
+      "epoch": 0.9549905514516406,
+      "grad_norm": 1.468258261680603,
+      "learning_rate": 4.720678833195332e-06,
+      "loss": 1.0357,
+      "step": 1853
+    },
+    {
+      "epoch": 0.9555059268166982,
+      "grad_norm": 1.4844988584518433,
+      "learning_rate": 4.720366144257961e-06,
+      "loss": 1.1176,
+      "step": 1854
+    },
+    {
+      "epoch": 0.9560213021817557,
+      "grad_norm": 1.3619471788406372,
+      "learning_rate": 4.7200532907649695e-06,
+      "loss": 1.0835,
+      "step": 1855
+    },
+    {
+      "epoch": 0.9565366775468133,
+      "grad_norm": 1.4447712898254395,
+      "learning_rate": 4.719740272739544e-06,
+      "loss": 1.1339,
+      "step": 1856
+    },
+    {
+      "epoch": 0.9570520529118708,
+      "grad_norm": 1.3479328155517578,
+      "learning_rate": 4.719427090204881e-06,
+      "loss": 1.0829,
+      "step": 1857
+    },
+    {
+      "epoch": 0.9575674282769283,
+      "grad_norm": 1.4314838647842407,
+      "learning_rate": 4.719113743184193e-06,
+      "loss": 1.0286,
+      "step": 1858
+    },
+    {
+      "epoch": 0.958082803641986,
+      "grad_norm": 1.3909810781478882,
+      "learning_rate": 4.7188002317007005e-06,
+      "loss": 1.0825,
+      "step": 1859
+    },
+    {
+      "epoch": 0.9585981790070435,
+      "grad_norm": 1.47553288936615,
+      "learning_rate": 4.718486555777641e-06,
+      "loss": 1.0356,
+      "step": 1860
+    },
+    {
+      "epoch": 0.959113554372101,
+      "grad_norm": 1.4584418535232544,
+      "learning_rate": 4.71817271543826e-06,
+      "loss": 1.0442,
+      "step": 1861
+    },
+    {
+      "epoch": 0.9596289297371585,
+      "grad_norm": 1.3760011196136475,
+      "learning_rate": 4.717858710705817e-06,
+      "loss": 1.0484,
+      "step": 1862
+    },
+    {
+      "epoch": 0.9601443051022162,
+      "grad_norm": 1.4110684394836426,
+      "learning_rate": 4.7175445416035845e-06,
+      "loss": 1.141,
+      "step": 1863
+    },
+    {
+      "epoch": 0.9606596804672737,
+      "grad_norm": 1.3495771884918213,
+      "learning_rate": 4.717230208154845e-06,
+      "loss": 1.0851,
+      "step": 1864
+    },
+    {
+      "epoch": 0.9611750558323312,
+      "grad_norm": 1.4929475784301758,
+      "learning_rate": 4.716915710382895e-06,
+      "loss": 1.0608,
+      "step": 1865
+    },
+    {
+      "epoch": 0.9616904311973887,
+      "grad_norm": 1.4913374185562134,
+      "learning_rate": 4.716601048311043e-06,
+      "loss": 1.0582,
+      "step": 1866
+    },
+    {
+      "epoch": 0.9622058065624464,
+      "grad_norm": 1.5947993993759155,
+      "learning_rate": 4.716286221962607e-06,
+      "loss": 1.1079,
+      "step": 1867
+    },
+    {
+      "epoch": 0.9627211819275039,
+      "grad_norm": 1.397476077079773,
+      "learning_rate": 4.715971231360922e-06,
+      "loss": 1.0905,
+      "step": 1868
+    },
+    {
+      "epoch": 0.9632365572925614,
+      "grad_norm": 1.3617136478424072,
+      "learning_rate": 4.715656076529331e-06,
+      "loss": 1.0949,
+      "step": 1869
+    },
+    {
+      "epoch": 0.9637519326576189,
+      "grad_norm": 1.4512900114059448,
+      "learning_rate": 4.715340757491191e-06,
+      "loss": 1.0835,
+      "step": 1870
+    },
+    {
+      "epoch": 0.9642673080226765,
+      "grad_norm": 1.3670510053634644,
+      "learning_rate": 4.715025274269872e-06,
+      "loss": 1.024,
+      "step": 1871
+    },
+    {
+      "epoch": 0.9647826833877341,
+      "grad_norm": 1.524481177330017,
+      "learning_rate": 4.7147096268887535e-06,
+      "loss": 1.0453,
+      "step": 1872
+    },
+    {
+      "epoch": 0.9652980587527916,
+      "grad_norm": 1.3488364219665527,
+      "learning_rate": 4.71439381537123e-06,
+      "loss": 1.0241,
+      "step": 1873
+    },
+    {
+      "epoch": 0.9658134341178491,
+      "grad_norm": 1.4507735967636108,
+      "learning_rate": 4.714077839740706e-06,
+      "loss": 1.0571,
+      "step": 1874
+    },
+    {
+      "epoch": 0.9663288094829067,
+      "grad_norm": 1.4156478643417358,
+      "learning_rate": 4.7137617000206e-06,
+      "loss": 1.0305,
+      "step": 1875
+    },
+    {
+      "epoch": 0.9668441848479643,
+      "grad_norm": 1.511183500289917,
+      "learning_rate": 4.7134453962343405e-06,
+      "loss": 1.0363,
+      "step": 1876
+    },
+    {
+      "epoch": 0.9673595602130218,
+      "grad_norm": 1.5476479530334473,
+      "learning_rate": 4.71312892840537e-06,
+      "loss": 1.0561,
+      "step": 1877
+    },
+    {
+      "epoch": 0.9678749355780794,
+      "grad_norm": 1.5617843866348267,
+      "learning_rate": 4.712812296557143e-06,
+      "loss": 1.0723,
+      "step": 1878
+    },
+    {
+      "epoch": 0.9683903109431369,
+      "grad_norm": 1.3265645503997803,
+      "learning_rate": 4.712495500713125e-06,
+      "loss": 1.0373,
+      "step": 1879
+    },
+    {
+      "epoch": 0.9689056863081945,
+      "grad_norm": 1.4959145784378052,
+      "learning_rate": 4.7121785408967945e-06,
+      "loss": 1.0563,
+      "step": 1880
+    },
+    {
+      "epoch": 0.969421061673252,
+      "grad_norm": 1.4450513124465942,
+      "learning_rate": 4.711861417131642e-06,
+      "loss": 1.0584,
+      "step": 1881
+    },
+    {
+      "epoch": 0.9699364370383096,
+      "grad_norm": 1.4012947082519531,
+      "learning_rate": 4.711544129441171e-06,
+      "loss": 1.0402,
+      "step": 1882
+    },
+    {
+      "epoch": 0.9704518124033671,
+      "grad_norm": 1.5120168924331665,
+      "learning_rate": 4.7112266778488956e-06,
+      "loss": 1.1354,
+      "step": 1883
+    },
+    {
+      "epoch": 0.9709671877684247,
+      "grad_norm": 1.4814488887786865,
+      "learning_rate": 4.710909062378342e-06,
+      "loss": 1.0379,
+      "step": 1884
+    },
+    {
+      "epoch": 0.9714825631334822,
+      "grad_norm": 1.4443869590759277,
+      "learning_rate": 4.71059128305305e-06,
+      "loss": 1.049,
+      "step": 1885
+    },
+    {
+      "epoch": 0.9719979384985398,
+      "grad_norm": 1.6024912595748901,
+      "learning_rate": 4.710273339896571e-06,
+      "loss": 1.0218,
+      "step": 1886
+    },
+    {
+      "epoch": 0.9725133138635973,
+      "grad_norm": 1.5310415029525757,
+      "learning_rate": 4.709955232932468e-06,
+      "loss": 1.1243,
+      "step": 1887
+    },
+    {
+      "epoch": 0.9730286892286548,
+      "grad_norm": 1.4426299333572388,
+      "learning_rate": 4.7096369621843176e-06,
+      "loss": 1.0082,
+      "step": 1888
+    },
+    {
+      "epoch": 0.9735440645937125,
+      "grad_norm": 1.407894253730774,
+      "learning_rate": 4.709318527675706e-06,
+      "loss": 1.0448,
+      "step": 1889
+    },
+    {
+      "epoch": 0.97405943995877,
+      "grad_norm": 1.3813974857330322,
+      "learning_rate": 4.708999929430234e-06,
+      "loss": 1.0825,
+      "step": 1890
+    },
+    {
+      "epoch": 0.9745748153238275,
+      "grad_norm": 1.442893385887146,
+      "learning_rate": 4.708681167471513e-06,
+      "loss": 1.053,
+      "step": 1891
+    },
+    {
+      "epoch": 0.975090190688885,
+      "grad_norm": 1.4274073839187622,
+      "learning_rate": 4.708362241823166e-06,
+      "loss": 1.0981,
+      "step": 1892
+    },
+    {
+      "epoch": 0.9756055660539427,
+      "grad_norm": 1.453469157218933,
+      "learning_rate": 4.7080431525088325e-06,
+      "loss": 1.0822,
+      "step": 1893
+    },
+    {
+      "epoch": 0.9761209414190002,
+      "grad_norm": 1.330512285232544,
+      "learning_rate": 4.707723899552158e-06,
+      "loss": 1.0873,
+      "step": 1894
+    },
+    {
+      "epoch": 0.9766363167840577,
+      "grad_norm": 1.3868836164474487,
+      "learning_rate": 4.707404482976803e-06,
+      "loss": 1.0318,
+      "step": 1895
+    },
+    {
+      "epoch": 0.9771516921491152,
+      "grad_norm": 1.4036799669265747,
+      "learning_rate": 4.707084902806441e-06,
+      "loss": 1.0617,
+      "step": 1896
+    },
+    {
+      "epoch": 0.9776670675141729,
+      "grad_norm": 1.4641084671020508,
+      "learning_rate": 4.7067651590647564e-06,
+      "loss": 1.0892,
+      "step": 1897
+    },
+    {
+      "epoch": 0.9781824428792304,
+      "grad_norm": 1.45862877368927,
+      "learning_rate": 4.706445251775446e-06,
+      "loss": 1.0931,
+      "step": 1898
+    },
+    {
+      "epoch": 0.9786978182442879,
+      "grad_norm": 1.4056271314620972,
+      "learning_rate": 4.70612518096222e-06,
+      "loss": 1.0556,
+      "step": 1899
+    },
+    {
+      "epoch": 0.9792131936093454,
+      "grad_norm": 1.386319875717163,
+      "learning_rate": 4.7058049466487974e-06,
+      "loss": 1.0281,
+      "step": 1900
+    },
+    {
+      "epoch": 0.979728568974403,
+      "grad_norm": 1.4791563749313354,
+      "learning_rate": 4.705484548858912e-06,
+      "loss": 1.0581,
+      "step": 1901
+    },
+    {
+      "epoch": 0.9802439443394606,
+      "grad_norm": 1.5182712078094482,
+      "learning_rate": 4.70516398761631e-06,
+      "loss": 1.0856,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9807593197045181,
+      "grad_norm": 1.4754759073257446,
+      "learning_rate": 4.704843262944748e-06,
+      "loss": 1.0979,
+      "step": 1903
+    },
+    {
+      "epoch": 0.9812746950695757,
+      "grad_norm": 1.4757529497146606,
+      "learning_rate": 4.704522374867996e-06,
+      "loss": 1.0508,
+      "step": 1904
+    },
+    {
+      "epoch": 0.9817900704346332,
+      "grad_norm": 1.5718193054199219,
+      "learning_rate": 4.704201323409835e-06,
+      "loss": 1.1101,
+      "step": 1905
+    },
+    {
+      "epoch": 0.9823054457996908,
+      "grad_norm": 1.4598045349121094,
+      "learning_rate": 4.703880108594059e-06,
+      "loss": 1.1212,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9828208211647483,
+      "grad_norm": 1.3712278604507446,
+      "learning_rate": 4.703558730444474e-06,
+      "loss": 1.0175,
+      "step": 1907
+    },
+    {
+      "epoch": 0.9833361965298059,
+      "grad_norm": 1.3985803127288818,
+      "learning_rate": 4.7032371889848985e-06,
+      "loss": 1.0278,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9838515718948634,
+      "grad_norm": 1.5293681621551514,
+      "learning_rate": 4.702915484239161e-06,
+      "loss": 1.1274,
+      "step": 1909
+    },
+    {
+      "epoch": 0.984366947259921,
+      "grad_norm": 1.5148296356201172,
+      "learning_rate": 4.7025936162311055e-06,
+      "loss": 1.0513,
+      "step": 1910
+    },
+    {
+      "epoch": 0.9848823226249785,
+      "grad_norm": 1.4407379627227783,
+      "learning_rate": 4.7022715849845844e-06,
+      "loss": 1.041,
+      "step": 1911
+    },
+    {
+      "epoch": 0.9853976979900361,
+      "grad_norm": 1.6437877416610718,
+      "learning_rate": 4.701949390523466e-06,
+      "loss": 1.0867,
+      "step": 1912
+    },
+    {
+      "epoch": 0.9859130733550936,
+      "grad_norm": 1.4243892431259155,
+      "learning_rate": 4.701627032871627e-06,
+      "loss": 1.0369,
+      "step": 1913
+    },
+    {
+      "epoch": 0.9864284487201512,
+      "grad_norm": 1.4947564601898193,
+      "learning_rate": 4.7013045120529595e-06,
+      "loss": 1.0783,
+      "step": 1914
+    },
+    {
+      "epoch": 0.9869438240852088,
+      "grad_norm": 1.5142713785171509,
+      "learning_rate": 4.700981828091365e-06,
+      "loss": 1.082,
+      "step": 1915
+    },
+    {
+      "epoch": 0.9874591994502663,
+      "grad_norm": 1.457655668258667,
+      "learning_rate": 4.70065898101076e-06,
+      "loss": 1.0564,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9879745748153238,
+      "grad_norm": 1.4288212060928345,
+      "learning_rate": 4.700335970835067e-06,
+      "loss": 1.0861,
+      "step": 1917
+    },
+    {
+      "epoch": 0.9884899501803813,
+      "grad_norm": 1.4429469108581543,
+      "learning_rate": 4.70001279758823e-06,
+      "loss": 1.0327,
+      "step": 1918
+    },
+    {
+      "epoch": 0.989005325545439,
+      "grad_norm": 1.3988040685653687,
+      "learning_rate": 4.699689461294197e-06,
+      "loss": 1.0651,
+      "step": 1919
+    },
+    {
+      "epoch": 0.9895207009104965,
+      "grad_norm": 1.574169397354126,
+      "learning_rate": 4.699365961976933e-06,
+      "loss": 1.0939,
+      "step": 1920
+    },
+    {
+      "epoch": 0.990036076275554,
+      "grad_norm": 1.3809009790420532,
+      "learning_rate": 4.699042299660411e-06,
+      "loss": 1.0724,
+      "step": 1921
+    },
+    {
+      "epoch": 0.9905514516406115,
+      "grad_norm": 1.5404826402664185,
+      "learning_rate": 4.69871847436862e-06,
+      "loss": 1.0632,
+      "step": 1922
+    },
+    {
+      "epoch": 0.9910668270056692,
+      "grad_norm": 1.4825998544692993,
+      "learning_rate": 4.698394486125558e-06,
+      "loss": 1.0113,
+      "step": 1923
+    },
+    {
+      "epoch": 0.9915822023707267,
+      "grad_norm": 1.4356918334960938,
+      "learning_rate": 4.698070334955237e-06,
+      "loss": 1.091,
+      "step": 1924
+    },
+    {
+      "epoch": 0.9920975777357842,
+      "grad_norm": 1.4427188634872437,
+      "learning_rate": 4.69774602088168e-06,
+      "loss": 1.0901,
+      "step": 1925
+    },
+    {
+      "epoch": 0.9926129531008417,
+      "grad_norm": 1.4783024787902832,
+      "learning_rate": 4.697421543928924e-06,
+      "loss": 1.0303,
+      "step": 1926
+    },
+    {
+      "epoch": 0.9931283284658994,
+      "grad_norm": 1.478934407234192,
+      "learning_rate": 4.697096904121016e-06,
+      "loss": 1.1187,
+      "step": 1927
+    },
+    {
+      "epoch": 0.9936437038309569,
+      "grad_norm": 1.5085822343826294,
+      "learning_rate": 4.696772101482014e-06,
+      "loss": 1.0884,
+      "step": 1928
+    },
+    {
+      "epoch": 0.9941590791960144,
+      "grad_norm": 1.4221833944320679,
+      "learning_rate": 4.6964471360359925e-06,
+      "loss": 1.0532,
+      "step": 1929
+    },
+    {
+      "epoch": 0.994674454561072,
+      "grad_norm": 1.5141397714614868,
+      "learning_rate": 4.696122007807033e-06,
+      "loss": 1.1066,
+      "step": 1930
+    },
+    {
+      "epoch": 0.9951898299261295,
+      "grad_norm": 1.4299890995025635,
+      "learning_rate": 4.695796716819233e-06,
+      "loss": 0.9984,
+      "step": 1931
+    },
+    {
+      "epoch": 0.9957052052911871,
+      "grad_norm": 1.5260546207427979,
+      "learning_rate": 4.695471263096699e-06,
+      "loss": 1.1296,
+      "step": 1932
+    },
+    {
+      "epoch": 0.9962205806562446,
+      "grad_norm": 1.358848214149475,
+      "learning_rate": 4.695145646663552e-06,
+      "loss": 1.0145,
+      "step": 1933
+    },
+    {
+      "epoch": 0.9967359560213022,
+      "grad_norm": 1.4124057292938232,
+      "learning_rate": 4.694819867543924e-06,
+      "loss": 1.0814,
+      "step": 1934
+    },
+    {
+      "epoch": 0.9972513313863597,
+      "grad_norm": 1.4105831384658813,
+      "learning_rate": 4.694493925761958e-06,
+      "loss": 1.0178,
+      "step": 1935
+    },
+    {
+      "epoch": 0.9977667067514173,
+      "grad_norm": 1.4723960161209106,
+      "learning_rate": 4.694167821341812e-06,
+      "loss": 1.1139,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9982820821164748,
+      "grad_norm": 1.416146993637085,
+      "learning_rate": 4.693841554307653e-06,
+      "loss": 1.0567,
+      "step": 1937
+    },
+    {
+      "epoch": 0.9987974574815324,
+      "grad_norm": 1.4232710599899292,
+      "learning_rate": 4.693515124683662e-06,
+      "loss": 1.0747,
+      "step": 1938
+    },
+    {
+      "epoch": 0.9993128328465899,
+      "grad_norm": 1.5446377992630005,
+      "learning_rate": 4.69318853249403e-06,
+      "loss": 1.0057,
+      "step": 1939
+    },
+    {
+      "epoch": 0.9998282082116475,
+      "grad_norm": 1.505684494972229,
+      "learning_rate": 4.692861777762963e-06,
+      "loss": 1.0992,
+      "step": 1940
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 11640,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 1940,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.345877427776389e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}