diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7233 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998780636507743,
+  "eval_steps": 500,
+  "global_step": 1025,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0009754907938056335,
+      "grad_norm": 4.439117513053022,
+      "learning_rate": 9.615384615384617e-08,
+      "loss": 0.3691,
+      "step": 1
+    },
+    {
+      "epoch": 0.001950981587611267,
+      "grad_norm": 4.665976578311697,
+      "learning_rate": 1.9230769230769234e-07,
+      "loss": 0.3804,
+      "step": 2
+    },
+    {
+      "epoch": 0.0029264723814169005,
+      "grad_norm": 4.458573970698933,
+      "learning_rate": 2.884615384615385e-07,
+      "loss": 0.3682,
+      "step": 3
+    },
+    {
+      "epoch": 0.003901963175222534,
+      "grad_norm": 4.21755765446574,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 0.3567,
+      "step": 4
+    },
+    {
+      "epoch": 0.004877453969028167,
+      "grad_norm": 4.1944148540591195,
+      "learning_rate": 4.807692307692308e-07,
+      "loss": 0.3773,
+      "step": 5
+    },
+    {
+      "epoch": 0.005852944762833801,
+      "grad_norm": 3.9774353472382957,
+      "learning_rate": 5.76923076923077e-07,
+      "loss": 0.3502,
+      "step": 6
+    },
+    {
+      "epoch": 0.006828435556639434,
+      "grad_norm": 3.553247347541323,
+      "learning_rate": 6.730769230769231e-07,
+      "loss": 0.3614,
+      "step": 7
+    },
+    {
+      "epoch": 0.007803926350445068,
+      "grad_norm": 2.949302152008964,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 0.3463,
+      "step": 8
+    },
+    {
+      "epoch": 0.008779417144250701,
+      "grad_norm": 2.799468109303019,
+      "learning_rate": 8.653846153846154e-07,
+      "loss": 0.3582,
+      "step": 9
+    },
+    {
+      "epoch": 0.009754907938056334,
+      "grad_norm": 2.6971355767992464,
+      "learning_rate": 9.615384615384617e-07,
+      "loss": 0.3277,
+      "step": 10
+    },
+    {
+      "epoch": 0.010730398731861967,
+      "grad_norm": 2.2435805544189256,
+      "learning_rate": 1.0576923076923078e-06,
+      "loss": 0.3411,
+      "step": 11
+    },
+    {
+      "epoch": 0.011705889525667602,
+      "grad_norm": 2.3858778520973933,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 0.3323,
+      "step": 12
+    },
+    {
+      "epoch": 0.012681380319473235,
+      "grad_norm": 2.325354329209325,
+      "learning_rate": 1.25e-06,
+      "loss": 0.3201,
+      "step": 13
+    },
+    {
+      "epoch": 0.013656871113278868,
+      "grad_norm": 2.268837198744112,
+      "learning_rate": 1.3461538461538462e-06,
+      "loss": 0.3458,
+      "step": 14
+    },
+    {
+      "epoch": 0.014632361907084501,
+      "grad_norm": 2.2354597650236263,
+      "learning_rate": 1.4423076923076922e-06,
+      "loss": 0.3057,
+      "step": 15
+    },
+    {
+      "epoch": 0.015607852700890136,
+      "grad_norm": 2.4354294587165968,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 0.2974,
+      "step": 16
+    },
+    {
+      "epoch": 0.016583343494695767,
+      "grad_norm": 1.9801626818691969,
+      "learning_rate": 1.6346153846153848e-06,
+      "loss": 0.2738,
+      "step": 17
+    },
+    {
+      "epoch": 0.017558834288501402,
+      "grad_norm": 1.7527920071956236,
+      "learning_rate": 1.7307692307692308e-06,
+      "loss": 0.268,
+      "step": 18
+    },
+    {
+      "epoch": 0.018534325082307037,
+      "grad_norm": 1.6175943655326306,
+      "learning_rate": 1.826923076923077e-06,
+      "loss": 0.2851,
+      "step": 19
+    },
+    {
+      "epoch": 0.01950981587611267,
+      "grad_norm": 1.8343780524193107,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 0.2773,
+      "step": 20
+    },
+    {
+      "epoch": 0.020485306669918303,
+      "grad_norm": 2.0687178014460055,
+      "learning_rate": 2.0192307692307692e-06,
+      "loss": 0.279,
+      "step": 21
+    },
+    {
+      "epoch": 0.021460797463723934,
+      "grad_norm": 1.850604452835117,
+      "learning_rate": 2.1153846153846155e-06,
+      "loss": 0.2743,
+      "step": 22
+    },
+    {
+      "epoch": 0.02243628825752957,
+      "grad_norm": 2.149994632477382,
+      "learning_rate": 2.211538461538462e-06,
+      "loss": 0.2694,
+      "step": 23
+    },
+    {
+      "epoch": 0.023411779051335204,
+      "grad_norm": 1.8792389072954687,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 0.2644,
+      "step": 24
+    },
+    {
+      "epoch": 0.024387269845140835,
+      "grad_norm": 1.457358625174777,
+      "learning_rate": 2.403846153846154e-06,
+      "loss": 0.2419,
+      "step": 25
+    },
+    {
+      "epoch": 0.02536276063894647,
+      "grad_norm": 1.7014683746851735,
+      "learning_rate": 2.5e-06,
+      "loss": 0.2544,
+      "step": 26
+    },
+    {
+      "epoch": 0.026338251432752105,
+      "grad_norm": 1.3815851595211681,
+      "learning_rate": 2.5961538461538465e-06,
+      "loss": 0.223,
+      "step": 27
+    },
+    {
+      "epoch": 0.027313742226557736,
+      "grad_norm": 1.3007184555893962,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 0.2456,
+      "step": 28
+    },
+    {
+      "epoch": 0.02828923302036337,
+      "grad_norm": 1.4472785639522512,
+      "learning_rate": 2.7884615384615386e-06,
+      "loss": 0.2516,
+      "step": 29
+    },
+    {
+      "epoch": 0.029264723814169002,
+      "grad_norm": 1.3652941522006306,
+      "learning_rate": 2.8846153846153845e-06,
+      "loss": 0.2416,
+      "step": 30
+    },
+    {
+      "epoch": 0.030240214607974637,
+      "grad_norm": 1.2522050977996877,
+      "learning_rate": 2.980769230769231e-06,
+      "loss": 0.2329,
+      "step": 31
+    },
+    {
+      "epoch": 0.031215705401780272,
+      "grad_norm": 1.1304756888064047,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 0.2217,
+      "step": 32
+    },
+    {
+      "epoch": 0.03219119619558591,
+      "grad_norm": 1.4050382297283845,
+      "learning_rate": 3.1730769230769233e-06,
+      "loss": 0.2309,
+      "step": 33
+    },
+    {
+      "epoch": 0.033166686989391535,
+      "grad_norm": 1.0865416261086687,
+      "learning_rate": 3.2692307692307696e-06,
+      "loss": 0.2217,
+      "step": 34
+    },
+    {
+      "epoch": 0.03414217778319717,
+      "grad_norm": 1.303187815201321,
+      "learning_rate": 3.365384615384616e-06,
+      "loss": 0.2338,
+      "step": 35
+    },
+    {
+      "epoch": 0.035117668577002804,
+      "grad_norm": 1.2021321704753516,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 0.2137,
+      "step": 36
+    },
+    {
+      "epoch": 0.03609315937080844,
+      "grad_norm": 1.1561740561669396,
+      "learning_rate": 3.557692307692308e-06,
+      "loss": 0.2162,
+      "step": 37
+    },
+    {
+      "epoch": 0.037068650164614074,
+      "grad_norm": 1.287983442354845,
+      "learning_rate": 3.653846153846154e-06,
+      "loss": 0.228,
+      "step": 38
+    },
+    {
+      "epoch": 0.0380441409584197,
+      "grad_norm": 1.0789644716188591,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.2106,
+      "step": 39
+    },
+    {
+      "epoch": 0.03901963175222534,
+      "grad_norm": 1.1684419312705403,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 0.218,
+      "step": 40
+    },
+    {
+      "epoch": 0.03999512254603097,
+      "grad_norm": 1.1195832124738565,
+      "learning_rate": 3.942307692307692e-06,
+      "loss": 0.2032,
+      "step": 41
+    },
+    {
+      "epoch": 0.040970613339836606,
+      "grad_norm": 1.0236100570446467,
+      "learning_rate": 4.0384615384615385e-06,
+      "loss": 0.217,
+      "step": 42
+    },
+    {
+      "epoch": 0.04194610413364224,
+      "grad_norm": 1.182743593377574,
+      "learning_rate": 4.134615384615385e-06,
+      "loss": 0.2215,
+      "step": 43
+    },
+    {
+      "epoch": 0.04292159492744787,
+      "grad_norm": 1.1077853276119973,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 0.2138,
+      "step": 44
+    },
+    {
+      "epoch": 0.043897085721253504,
+      "grad_norm": 1.1863700432181472,
+      "learning_rate": 4.326923076923077e-06,
+      "loss": 0.2113,
+      "step": 45
+    },
+    {
+      "epoch": 0.04487257651505914,
+      "grad_norm": 1.1214318092951887,
+      "learning_rate": 4.423076923076924e-06,
+      "loss": 0.2106,
+      "step": 46
+    },
+    {
+      "epoch": 0.04584806730886477,
+      "grad_norm": 1.0805493864146463,
+      "learning_rate": 4.51923076923077e-06,
+      "loss": 0.2104,
+      "step": 47
+    },
+    {
+      "epoch": 0.04682355810267041,
+      "grad_norm": 1.0336295929522208,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 0.1981,
+      "step": 48
+    },
+    {
+      "epoch": 0.04779904889647604,
+      "grad_norm": 1.119869757376067,
+      "learning_rate": 4.711538461538462e-06,
+      "loss": 0.2014,
+      "step": 49
+    },
+    {
+      "epoch": 0.04877453969028167,
+      "grad_norm": 1.030817002409712,
+      "learning_rate": 4.807692307692308e-06,
+      "loss": 0.1996,
+      "step": 50
+    },
+    {
+      "epoch": 0.049750030484087306,
+      "grad_norm": 1.0055854671916695,
+      "learning_rate": 4.903846153846154e-06,
+      "loss": 0.2052,
+      "step": 51
+    },
+    {
+      "epoch": 0.05072552127789294,
+      "grad_norm": 1.107516098502413,
+      "learning_rate": 5e-06,
+      "loss": 0.2168,
+      "step": 52
+    },
+    {
+      "epoch": 0.051701012071698575,
+      "grad_norm": 1.0591849818221413,
+      "learning_rate": 4.9999869688212956e-06,
+      "loss": 0.2099,
+      "step": 53
+    },
+    {
+      "epoch": 0.05267650286550421,
+      "grad_norm": 1.0486177613460919,
+      "learning_rate": 4.999947875421032e-06,
+      "loss": 0.1978,
+      "step": 54
+    },
+    {
+      "epoch": 0.05365199365930984,
+      "grad_norm": 1.0333038397384442,
+      "learning_rate": 4.999882720206755e-06,
+      "loss": 0.1934,
+      "step": 55
+    },
+    {
+      "epoch": 0.05462748445311547,
+      "grad_norm": 1.0394851360953357,
+      "learning_rate": 4.999791503857704e-06,
+      "loss": 0.2019,
+      "step": 56
+    },
+    {
+      "epoch": 0.05560297524692111,
+      "grad_norm": 1.1046443338767855,
+      "learning_rate": 4.999674227324805e-06,
+      "loss": 0.2044,
+      "step": 57
+    },
+    {
+      "epoch": 0.05657846604072674,
+      "grad_norm": 1.2371807932970489,
+      "learning_rate": 4.9995308918306595e-06,
+      "loss": 0.1972,
+      "step": 58
+    },
+    {
+      "epoch": 0.05755395683453238,
+      "grad_norm": 0.9905094942349434,
+      "learning_rate": 4.99936149886953e-06,
+      "loss": 0.19,
+      "step": 59
+    },
+    {
+      "epoch": 0.058529447628338005,
+      "grad_norm": 1.005308785484947,
+      "learning_rate": 4.999166050207331e-06,
+      "loss": 0.1838,
+      "step": 60
+    },
+    {
+      "epoch": 0.05950493842214364,
+      "grad_norm": 1.0197164862121335,
+      "learning_rate": 4.9989445478816e-06,
+      "loss": 0.1906,
+      "step": 61
+    },
+    {
+      "epoch": 0.060480429215949275,
+      "grad_norm": 1.112996579159861,
+      "learning_rate": 4.99869699420149e-06,
+      "loss": 0.1949,
+      "step": 62
+    },
+    {
+      "epoch": 0.06145592000975491,
+      "grad_norm": 0.9210560581443838,
+      "learning_rate": 4.998423391747731e-06,
+      "loss": 0.1869,
+      "step": 63
+    },
+    {
+      "epoch": 0.062431410803560544,
+      "grad_norm": 0.994708396507738,
+      "learning_rate": 4.998123743372615e-06,
+      "loss": 0.1832,
+      "step": 64
+    },
+    {
+      "epoch": 0.06340690159736617,
+      "grad_norm": 1.000252041680365,
+      "learning_rate": 4.997798052199959e-06,
+      "loss": 0.2011,
+      "step": 65
+    },
+    {
+      "epoch": 0.06438239239117181,
+      "grad_norm": 1.001566090006205,
+      "learning_rate": 4.9974463216250735e-06,
+      "loss": 0.1851,
+      "step": 66
+    },
+    {
+      "epoch": 0.06535788318497744,
+      "grad_norm": 1.0035711549242645,
+      "learning_rate": 4.9970685553147316e-06,
+      "loss": 0.1959,
+      "step": 67
+    },
+    {
+      "epoch": 0.06633337397878307,
+      "grad_norm": 1.0321021956412775,
+      "learning_rate": 4.996664757207124e-06,
+      "loss": 0.1948,
+      "step": 68
+    },
+    {
+      "epoch": 0.06730886477258871,
+      "grad_norm": 1.1196394074729383,
+      "learning_rate": 4.996234931511823e-06,
+      "loss": 0.1959,
+      "step": 69
+    },
+    {
+      "epoch": 0.06828435556639434,
+      "grad_norm": 1.204621557285872,
+      "learning_rate": 4.995779082709739e-06,
+      "loss": 0.1962,
+      "step": 70
+    },
+    {
+      "epoch": 0.06925984636019998,
+      "grad_norm": 1.0177518163557318,
+      "learning_rate": 4.995297215553067e-06,
+      "loss": 0.1806,
+      "step": 71
+    },
+    {
+      "epoch": 0.07023533715400561,
+      "grad_norm": 1.0847615883655475,
+      "learning_rate": 4.994789335065245e-06,
+      "loss": 0.2044,
+      "step": 72
+    },
+    {
+      "epoch": 0.07121082794781124,
+      "grad_norm": 1.168873122167307,
+      "learning_rate": 4.9942554465409e-06,
+      "loss": 0.1934,
+      "step": 73
+    },
+    {
+      "epoch": 0.07218631874161688,
+      "grad_norm": 1.019104243672844,
+      "learning_rate": 4.993695555545789e-06,
+      "loss": 0.1905,
+      "step": 74
+    },
+    {
+      "epoch": 0.0731618095354225,
+      "grad_norm": 1.1308032239714556,
+      "learning_rate": 4.993109667916742e-06,
+      "loss": 0.193,
+      "step": 75
+    },
+    {
+      "epoch": 0.07413730032922815,
+      "grad_norm": 1.1401766992269902,
+      "learning_rate": 4.992497789761606e-06,
+      "loss": 0.1793,
+      "step": 76
+    },
+    {
+      "epoch": 0.07511279112303378,
+      "grad_norm": 1.1144999371666817,
+      "learning_rate": 4.991859927459174e-06,
+      "loss": 0.2023,
+      "step": 77
+    },
+    {
+      "epoch": 0.0760882819168394,
+      "grad_norm": 1.191890901059724,
+      "learning_rate": 4.991196087659125e-06,
+      "loss": 0.1858,
+      "step": 78
+    },
+    {
+      "epoch": 0.07706377271064505,
+      "grad_norm": 1.11445392411597,
+      "learning_rate": 4.990506277281952e-06,
+      "loss": 0.1843,
+      "step": 79
+    },
+    {
+      "epoch": 0.07803926350445067,
+      "grad_norm": 1.1630978983896232,
+      "learning_rate": 4.989790503518888e-06,
+      "loss": 0.1775,
+      "step": 80
+    },
+    {
+      "epoch": 0.07901475429825632,
+      "grad_norm": 1.3514121202700597,
+      "learning_rate": 4.989048773831834e-06,
+      "loss": 0.1879,
+      "step": 81
+    },
+    {
+      "epoch": 0.07999024509206194,
+      "grad_norm": 1.0138569119140257,
+      "learning_rate": 4.988281095953279e-06,
+      "loss": 0.1712,
+      "step": 82
+    },
+    {
+      "epoch": 0.08096573588586757,
+      "grad_norm": 1.2876601226086022,
+      "learning_rate": 4.987487477886221e-06,
+      "loss": 0.1806,
+      "step": 83
+    },
+    {
+      "epoch": 0.08194122667967321,
+      "grad_norm": 1.0842905018940538,
+      "learning_rate": 4.986667927904084e-06,
+      "loss": 0.172,
+      "step": 84
+    },
+    {
+      "epoch": 0.08291671747347884,
+      "grad_norm": 1.0011346482797712,
+      "learning_rate": 4.985822454550629e-06,
+      "loss": 0.1775,
+      "step": 85
+    },
+    {
+      "epoch": 0.08389220826728448,
+      "grad_norm": 1.322374107432227,
+      "learning_rate": 4.984951066639869e-06,
+      "loss": 0.1872,
+      "step": 86
+    },
+    {
+      "epoch": 0.08486769906109011,
+      "grad_norm": 1.0531176300373897,
+      "learning_rate": 4.984053773255971e-06,
+      "loss": 0.1848,
+      "step": 87
+    },
+    {
+      "epoch": 0.08584318985489574,
+      "grad_norm": 1.0696103709835127,
+      "learning_rate": 4.9831305837531684e-06,
+      "loss": 0.1789,
+      "step": 88
+    },
+    {
+      "epoch": 0.08681868064870138,
+      "grad_norm": 1.048271406336639,
+      "learning_rate": 4.98218150775566e-06,
+      "loss": 0.1763,
+      "step": 89
+    },
+    {
+      "epoch": 0.08779417144250701,
+      "grad_norm": 0.9981980018547214,
+      "learning_rate": 4.9812065551575075e-06,
+      "loss": 0.1746,
+      "step": 90
+    },
+    {
+      "epoch": 0.08876966223631265,
+      "grad_norm": 1.0986911182229246,
+      "learning_rate": 4.9802057361225375e-06,
+      "loss": 0.2019,
+      "step": 91
+    },
+    {
+      "epoch": 0.08974515303011828,
+      "grad_norm": 1.0482728855361627,
+      "learning_rate": 4.97917906108423e-06,
+      "loss": 0.1725,
+      "step": 92
+    },
+    {
+      "epoch": 0.0907206438239239,
+      "grad_norm": 1.0599473411074543,
+      "learning_rate": 4.978126540745615e-06,
+      "loss": 0.1804,
+      "step": 93
+    },
+    {
+      "epoch": 0.09169613461772955,
+      "grad_norm": 1.078553827467801,
+      "learning_rate": 4.977048186079155e-06,
+      "loss": 0.2011,
+      "step": 94
+    },
+    {
+      "epoch": 0.09267162541153517,
+      "grad_norm": 1.0444452777467914,
+      "learning_rate": 4.975944008326638e-06,
+      "loss": 0.1835,
+      "step": 95
+    },
+    {
+      "epoch": 0.09364711620534082,
+      "grad_norm": 1.0791056951072164,
+      "learning_rate": 4.974814018999053e-06,
+      "loss": 0.1824,
+      "step": 96
+    },
+    {
+      "epoch": 0.09462260699914644,
+      "grad_norm": 1.078717583871865,
+      "learning_rate": 4.973658229876476e-06,
+      "loss": 0.189,
+      "step": 97
+    },
+    {
+      "epoch": 0.09559809779295209,
+      "grad_norm": 1.124139067655016,
+      "learning_rate": 4.97247665300794e-06,
+      "loss": 0.1832,
+      "step": 98
+    },
+    {
+      "epoch": 0.09657358858675771,
+      "grad_norm": 1.0195889033855914,
+      "learning_rate": 4.971269300711318e-06,
+      "loss": 0.1766,
+      "step": 99
+    },
+    {
+      "epoch": 0.09754907938056334,
+      "grad_norm": 1.1122862478289546,
+      "learning_rate": 4.970036185573189e-06,
+      "loss": 0.1805,
+      "step": 100
+    },
+    {
+      "epoch": 0.09852457017436898,
+      "grad_norm": 1.1044026960421833,
+      "learning_rate": 4.968777320448707e-06,
+      "loss": 0.1836,
+      "step": 101
+    },
+    {
+      "epoch": 0.09950006096817461,
+      "grad_norm": 1.204423675193318,
+      "learning_rate": 4.96749271846147e-06,
+      "loss": 0.1767,
+      "step": 102
+    },
+    {
+      "epoch": 0.10047555176198025,
+      "grad_norm": 1.0248442266116176,
+      "learning_rate": 4.96618239300338e-06,
+      "loss": 0.1912,
+      "step": 103
+    },
+    {
+      "epoch": 0.10145104255578588,
+      "grad_norm": 1.1740907364465507,
+      "learning_rate": 4.964846357734505e-06,
+      "loss": 0.1726,
+      "step": 104
+    },
+    {
+      "epoch": 0.10242653334959151,
+      "grad_norm": 0.9983984379050335,
+      "learning_rate": 4.963484626582937e-06,
+      "loss": 0.1787,
+      "step": 105
+    },
+    {
+      "epoch": 0.10340202414339715,
+      "grad_norm": 1.0692395971458715,
+      "learning_rate": 4.9620972137446456e-06,
+      "loss": 0.179,
+      "step": 106
+    },
+    {
+      "epoch": 0.10437751493720278,
+      "grad_norm": 1.0955109652052737,
+      "learning_rate": 4.96068413368333e-06,
+      "loss": 0.1765,
+      "step": 107
+    },
+    {
+      "epoch": 0.10535300573100842,
+      "grad_norm": 1.0684942634997572,
+      "learning_rate": 4.959245401130269e-06,
+      "loss": 0.172,
+      "step": 108
+    },
+    {
+      "epoch": 0.10632849652481405,
+      "grad_norm": 0.9995901623406969,
+      "learning_rate": 4.957781031084169e-06,
+      "loss": 0.1805,
+      "step": 109
+    },
+    {
+      "epoch": 0.10730398731861968,
+      "grad_norm": 0.9323445364396948,
+      "learning_rate": 4.956291038811003e-06,
+      "loss": 0.1853,
+      "step": 110
+    },
+    {
+      "epoch": 0.10827947811242532,
+      "grad_norm": 1.1033013476735185,
+      "learning_rate": 4.954775439843855e-06,
+      "loss": 0.1873,
+      "step": 111
+    },
+    {
+      "epoch": 0.10925496890623095,
+      "grad_norm": 1.1691417741452548,
+      "learning_rate": 4.9532342499827584e-06,
+      "loss": 0.185,
+      "step": 112
+    },
+    {
+      "epoch": 0.11023045970003659,
+      "grad_norm": 1.1918610910255243,
+      "learning_rate": 4.95166748529453e-06,
+      "loss": 0.1892,
+      "step": 113
+    },
+    {
+      "epoch": 0.11120595049384221,
+      "grad_norm": 1.1014111858940245,
+      "learning_rate": 4.950075162112602e-06,
+      "loss": 0.1735,
+      "step": 114
+    },
+    {
+      "epoch": 0.11218144128764784,
+      "grad_norm": 1.216567039997431,
+      "learning_rate": 4.9484572970368516e-06,
+      "loss": 0.1838,
+      "step": 115
+    },
+    {
+      "epoch": 0.11315693208145348,
+      "grad_norm": 1.1097171959187067,
+      "learning_rate": 4.946813906933432e-06,
+      "loss": 0.1874,
+      "step": 116
+    },
+    {
+      "epoch": 0.11413242287525911,
+      "grad_norm": 0.967977115534241,
+      "learning_rate": 4.94514500893459e-06,
+      "loss": 0.1738,
+      "step": 117
+    },
+    {
+      "epoch": 0.11510791366906475,
+      "grad_norm": 0.9657457430010942,
+      "learning_rate": 4.943450620438491e-06,
+      "loss": 0.1752,
+      "step": 118
+    },
+    {
+      "epoch": 0.11608340446287038,
+      "grad_norm": 0.9166596111467156,
+      "learning_rate": 4.941730759109041e-06,
+      "loss": 0.1818,
+      "step": 119
+    },
+    {
+      "epoch": 0.11705889525667601,
+      "grad_norm": 0.9450152131596894,
+      "learning_rate": 4.939985442875695e-06,
+      "loss": 0.1799,
+      "step": 120
+    },
+    {
+      "epoch": 0.11803438605048165,
+      "grad_norm": 0.938177849072337,
+      "learning_rate": 4.938214689933276e-06,
+      "loss": 0.1735,
+      "step": 121
+    },
+    {
+      "epoch": 0.11900987684428728,
+      "grad_norm": 1.0415880049786728,
+      "learning_rate": 4.93641851874178e-06,
+      "loss": 0.1868,
+      "step": 122
+    },
+    {
+      "epoch": 0.11998536763809292,
+      "grad_norm": 0.9746213204544258,
+      "learning_rate": 4.9345969480261925e-06,
+      "loss": 0.1829,
+      "step": 123
+    },
+    {
+      "epoch": 0.12096085843189855,
+      "grad_norm": 0.9444433221181193,
+      "learning_rate": 4.932749996776282e-06,
+      "loss": 0.1731,
+      "step": 124
+    },
+    {
+      "epoch": 0.12193634922570418,
+      "grad_norm": 1.0120487885636702,
+      "learning_rate": 4.93087768424641e-06,
+      "loss": 0.1886,
+      "step": 125
+    },
+    {
+      "epoch": 0.12291184001950982,
+      "grad_norm": 0.9837069724067731,
+      "learning_rate": 4.928980029955329e-06,
+      "loss": 0.1821,
+      "step": 126
+    },
+    {
+      "epoch": 0.12388733081331545,
+      "grad_norm": 0.9425235884710698,
+      "learning_rate": 4.927057053685975e-06,
+      "loss": 0.1801,
+      "step": 127
+    },
+    {
+      "epoch": 0.12486282160712109,
+      "grad_norm": 1.0113092843870408,
+      "learning_rate": 4.925108775485269e-06,
+      "loss": 0.1796,
+      "step": 128
+    },
+    {
+      "epoch": 0.12583831240092672,
+      "grad_norm": 1.0163088829406406,
+      "learning_rate": 4.923135215663897e-06,
+      "loss": 0.1654,
+      "step": 129
+    },
+    {
+      "epoch": 0.12681380319473234,
+      "grad_norm": 1.164966417682963,
+      "learning_rate": 4.921136394796109e-06,
+      "loss": 0.1803,
+      "step": 130
+    },
+    {
+      "epoch": 0.12778929398853797,
+      "grad_norm": 1.1687221264442198,
+      "learning_rate": 4.919112333719498e-06,
+      "loss": 0.1878,
+      "step": 131
+    },
+    {
+      "epoch": 0.12876478478234363,
+      "grad_norm": 0.9888510085952801,
+      "learning_rate": 4.9170630535347866e-06,
+      "loss": 0.1771,
+      "step": 132
+    },
+    {
+      "epoch": 0.12974027557614926,
+      "grad_norm": 1.2032250564065257,
+      "learning_rate": 4.914988575605602e-06,
+      "loss": 0.1893,
+      "step": 133
+    },
+    {
+      "epoch": 0.13071576636995488,
+      "grad_norm": 0.9809911156410793,
+      "learning_rate": 4.91288892155826e-06,
+      "loss": 0.1728,
+      "step": 134
+    },
+    {
+      "epoch": 0.1316912571637605,
+      "grad_norm": 1.2533075523741326,
+      "learning_rate": 4.910764113281533e-06,
+      "loss": 0.1836,
+      "step": 135
+    },
+    {
+      "epoch": 0.13266674795756614,
+      "grad_norm": 1.0029237484325686,
+      "learning_rate": 4.908614172926426e-06,
+      "loss": 0.1757,
+      "step": 136
+    },
+    {
+      "epoch": 0.1336422387513718,
+      "grad_norm": 1.0323575219354129,
+      "learning_rate": 4.906439122905946e-06,
+      "loss": 0.161,
+      "step": 137
+    },
+    {
+      "epoch": 0.13461772954517742,
+      "grad_norm": 0.9974877118091979,
+      "learning_rate": 4.904238985894864e-06,
+      "loss": 0.1644,
+      "step": 138
+    },
+    {
+      "epoch": 0.13559322033898305,
+      "grad_norm": 0.8968544196822601,
+      "learning_rate": 4.902013784829483e-06,
+      "loss": 0.1541,
+      "step": 139
+    },
+    {
+      "epoch": 0.13656871113278868,
+      "grad_norm": 1.067306106970932,
+      "learning_rate": 4.899763542907399e-06,
+      "loss": 0.1689,
+      "step": 140
+    },
+    {
+      "epoch": 0.1375442019265943,
+      "grad_norm": 0.9831550732650544,
+      "learning_rate": 4.897488283587253e-06,
+      "loss": 0.1764,
+      "step": 141
+    },
+    {
+      "epoch": 0.13851969272039996,
+      "grad_norm": 1.098512924207099,
+      "learning_rate": 4.895188030588495e-06,
+      "loss": 0.1671,
+      "step": 142
+    },
+    {
+      "epoch": 0.1394951835142056,
+      "grad_norm": 0.9813597480687253,
+      "learning_rate": 4.892862807891131e-06,
+      "loss": 0.1723,
+      "step": 143
+    },
+    {
+      "epoch": 0.14047067430801122,
+      "grad_norm": 0.9059688793766486,
+      "learning_rate": 4.890512639735475e-06,
+      "loss": 0.1585,
+      "step": 144
+    },
+    {
+      "epoch": 0.14144616510181685,
+      "grad_norm": 0.9421090522610057,
+      "learning_rate": 4.888137550621897e-06,
+      "loss": 0.1667,
+      "step": 145
+    },
+    {
+      "epoch": 0.14242165589562247,
+      "grad_norm": 0.9984203476569828,
+      "learning_rate": 4.885737565310565e-06,
+      "loss": 0.1748,
+      "step": 146
+    },
+    {
+      "epoch": 0.14339714668942813,
+      "grad_norm": 0.9344709716653912,
+      "learning_rate": 4.883312708821188e-06,
+      "loss": 0.172,
+      "step": 147
+    },
+    {
+      "epoch": 0.14437263748323376,
+      "grad_norm": 0.995510093417273,
+      "learning_rate": 4.880863006432758e-06,
+      "loss": 0.1704,
+      "step": 148
+    },
+    {
+      "epoch": 0.14534812827703938,
+      "grad_norm": 0.9658034107481427,
+      "learning_rate": 4.878388483683281e-06,
+      "loss": 0.1779,
+      "step": 149
+    },
+    {
+      "epoch": 0.146323619070845,
+      "grad_norm": 1.0283682187427396,
+      "learning_rate": 4.8758891663695165e-06,
+      "loss": 0.1733,
+      "step": 150
+    },
+    {
+      "epoch": 0.14729910986465064,
+      "grad_norm": 1.0848455109349466,
+      "learning_rate": 4.873365080546706e-06,
+      "loss": 0.1883,
+      "step": 151
+    },
+    {
+      "epoch": 0.1482746006584563,
+      "grad_norm": 1.0034493473399255,
+      "learning_rate": 4.870816252528298e-06,
+      "loss": 0.1725,
+      "step": 152
+    },
+    {
+      "epoch": 0.14925009145226192,
+      "grad_norm": 1.0714206307709666,
+      "learning_rate": 4.868242708885681e-06,
+      "loss": 0.1855,
+      "step": 153
+    },
+    {
+      "epoch": 0.15022558224606755,
+      "grad_norm": 1.0894809817408557,
+      "learning_rate": 4.8656444764479005e-06,
+      "loss": 0.1815,
+      "step": 154
+    },
+    {
+      "epoch": 0.15120107303987318,
+      "grad_norm": 1.0188695855109409,
+      "learning_rate": 4.863021582301381e-06,
+      "loss": 0.171,
+      "step": 155
+    },
+    {
+      "epoch": 0.1521765638336788,
+      "grad_norm": 1.0015265493016945,
+      "learning_rate": 4.860374053789643e-06,
+      "loss": 0.1703,
+      "step": 156
+    },
+    {
+      "epoch": 0.15315205462748446,
+      "grad_norm": 0.9798681079088354,
+      "learning_rate": 4.857701918513023e-06,
+      "loss": 0.1762,
+      "step": 157
+    },
+    {
+      "epoch": 0.1541275454212901,
+      "grad_norm": 0.9770978700089492,
+      "learning_rate": 4.855005204328378e-06,
+      "loss": 0.159,
+      "step": 158
+    },
+    {
+      "epoch": 0.15510303621509572,
+      "grad_norm": 1.0074537647270105,
+      "learning_rate": 4.8522839393487976e-06,
+      "loss": 0.1701,
+      "step": 159
+    },
+    {
+      "epoch": 0.15607852700890135,
+      "grad_norm": 0.983344185894382,
+      "learning_rate": 4.849538151943316e-06,
+      "loss": 0.1742,
+      "step": 160
+    },
+    {
+      "epoch": 0.15705401780270697,
+      "grad_norm": 0.9614400815443347,
+      "learning_rate": 4.84676787073661e-06,
+      "loss": 0.173,
+      "step": 161
+    },
+    {
+      "epoch": 0.15802950859651263,
+      "grad_norm": 0.9526818664956649,
+      "learning_rate": 4.843973124608703e-06,
+      "loss": 0.1732,
+      "step": 162
+    },
+    {
+      "epoch": 0.15900499939031826,
+      "grad_norm": 0.8985619765911587,
+      "learning_rate": 4.841153942694664e-06,
+      "loss": 0.1665,
+      "step": 163
+    },
+    {
+      "epoch": 0.15998049018412389,
+      "grad_norm": 0.9257494791929499,
+      "learning_rate": 4.838310354384304e-06,
+      "loss": 0.1684,
+      "step": 164
+    },
+    {
+      "epoch": 0.1609559809779295,
+      "grad_norm": 0.9783281983993328,
+      "learning_rate": 4.835442389321867e-06,
+      "loss": 0.1692,
+      "step": 165
+    },
+    {
+      "epoch": 0.16193147177173514,
+      "grad_norm": 0.8866263340546827,
+      "learning_rate": 4.832550077405727e-06,
+      "loss": 0.1563,
+      "step": 166
+    },
+    {
+      "epoch": 0.1629069625655408,
+      "grad_norm": 1.0759800513612392,
+      "learning_rate": 4.829633448788072e-06,
+      "loss": 0.1796,
+      "step": 167
+    },
+    {
+      "epoch": 0.16388245335934642,
+      "grad_norm": 1.0415739389017458,
+      "learning_rate": 4.826692533874586e-06,
+      "loss": 0.1736,
+      "step": 168
+    },
+    {
+      "epoch": 0.16485794415315205,
+      "grad_norm": 1.0321014340070922,
+      "learning_rate": 4.823727363324142e-06,
+      "loss": 0.1763,
+      "step": 169
+    },
+    {
+      "epoch": 0.16583343494695768,
+      "grad_norm": 1.149401130099103,
+      "learning_rate": 4.820737968048471e-06,
+      "loss": 0.1735,
+      "step": 170
+    },
+    {
+      "epoch": 0.1668089257407633,
+      "grad_norm": 0.9668062539942579,
+      "learning_rate": 4.8177243792118515e-06,
+      "loss": 0.1843,
+      "step": 171
+    },
+    {
+      "epoch": 0.16778441653456896,
+      "grad_norm": 1.0145067381113042,
+      "learning_rate": 4.8146866282307725e-06,
+      "loss": 0.1719,
+      "step": 172
+    },
+    {
+      "epoch": 0.1687599073283746,
+      "grad_norm": 0.9146830744890082,
+      "learning_rate": 4.811624746773616e-06,
+      "loss": 0.1597,
+      "step": 173
+    },
+    {
+      "epoch": 0.16973539812218022,
+      "grad_norm": 0.9826930086582164,
+      "learning_rate": 4.808538766760321e-06,
+      "loss": 0.1539,
+      "step": 174
+    },
+    {
+      "epoch": 0.17071088891598585,
+      "grad_norm": 0.9622129610913185,
+      "learning_rate": 4.805428720362054e-06,
+      "loss": 0.1682,
+      "step": 175
+    },
+    {
+      "epoch": 0.17168637970979148,
+      "grad_norm": 1.200349435662473,
+      "learning_rate": 4.8022946400008705e-06,
+      "loss": 0.1687,
+      "step": 176
+    },
+    {
+      "epoch": 0.17266187050359713,
+      "grad_norm": 1.0757087822872222,
+      "learning_rate": 4.79913655834938e-06,
+      "loss": 0.1812,
+      "step": 177
+    },
+    {
+      "epoch": 0.17363736129740276,
+      "grad_norm": 1.0368299327714814,
+      "learning_rate": 4.795954508330403e-06,
+      "loss": 0.1721,
+      "step": 178
+    },
+    {
+      "epoch": 0.1746128520912084,
+      "grad_norm": 1.1790584459327953,
+      "learning_rate": 4.79274852311663e-06,
+      "loss": 0.1708,
+      "step": 179
+    },
+    {
+      "epoch": 0.17558834288501401,
+      "grad_norm": 0.9846847616375829,
+      "learning_rate": 4.7895186361302736e-06,
+      "loss": 0.1639,
+      "step": 180
+    },
+    {
+      "epoch": 0.17656383367881964,
+      "grad_norm": 0.9524671440100206,
+      "learning_rate": 4.786264881042722e-06,
+      "loss": 0.1544,
+      "step": 181
+    },
+    {
+      "epoch": 0.1775393244726253,
+      "grad_norm": 1.0337867693599667,
+      "learning_rate": 4.782987291774186e-06,
+      "loss": 0.1586,
+      "step": 182
+    },
+    {
+      "epoch": 0.17851481526643093,
+      "grad_norm": 0.955523189772594,
+      "learning_rate": 4.779685902493346e-06,
+      "loss": 0.1703,
+      "step": 183
+    },
+    {
+      "epoch": 0.17949030606023655,
+      "grad_norm": 0.8949175563809436,
+      "learning_rate": 4.776360747616999e-06,
+      "loss": 0.1536,
+      "step": 184
+    },
+    {
+      "epoch": 0.18046579685404218,
+      "grad_norm": 1.0812607035187014,
+      "learning_rate": 4.773011861809694e-06,
+      "loss": 0.1729,
+      "step": 185
+    },
+    {
+      "epoch": 0.1814412876478478,
+      "grad_norm": 0.9776277191751103,
+      "learning_rate": 4.769639279983372e-06,
+      "loss": 0.1745,
+      "step": 186
+    },
+    {
+      "epoch": 0.18241677844165347,
+      "grad_norm": 1.0274757082533237,
+      "learning_rate": 4.76624303729701e-06,
+      "loss": 0.1651,
+      "step": 187
+    },
+    {
+      "epoch": 0.1833922692354591,
+      "grad_norm": 1.0244957983233014,
+      "learning_rate": 4.762823169156242e-06,
+      "loss": 0.1642,
+      "step": 188
+    },
+    {
+      "epoch": 0.18436776002926472,
+      "grad_norm": 0.9535466834456596,
+      "learning_rate": 4.7593797112129995e-06,
+      "loss": 0.1701,
+      "step": 189
+    },
+    {
+      "epoch": 0.18534325082307035,
+      "grad_norm": 1.0223048067541962,
+      "learning_rate": 4.755912699365135e-06,
+      "loss": 0.1619,
+      "step": 190
+    },
+    {
+      "epoch": 0.18631874161687598,
+      "grad_norm": 0.9985286606007027,
+      "learning_rate": 4.752422169756048e-06,
+      "loss": 0.1541,
+      "step": 191
+    },
+    {
+      "epoch": 0.18729423241068163,
+      "grad_norm": 0.9550222906892099,
+      "learning_rate": 4.748908158774312e-06,
+      "loss": 0.1583,
+      "step": 192
+    },
+    {
+      "epoch": 0.18826972320448726,
+      "grad_norm": 1.0220426764043464,
+      "learning_rate": 4.745370703053291e-06,
+      "loss": 0.1685,
+      "step": 193
+    },
+    {
+      "epoch": 0.1892452139982929,
+      "grad_norm": 0.9903476130292974,
+      "learning_rate": 4.741809839470758e-06,
+      "loss": 0.1624,
+      "step": 194
+    },
+    {
+      "epoch": 0.19022070479209852,
+      "grad_norm": 0.9340244675396043,
+      "learning_rate": 4.738225605148514e-06,
+      "loss": 0.171,
+      "step": 195
+    },
+    {
+      "epoch": 0.19119619558590417,
+      "grad_norm": 0.9933142717509476,
+      "learning_rate": 4.734618037451997e-06,
+      "loss": 0.1677,
+      "step": 196
+    },
+    {
+      "epoch": 0.1921716863797098,
+      "grad_norm": 1.1481162666912414,
+      "learning_rate": 4.730987173989894e-06,
+      "loss": 0.1654,
+      "step": 197
+    },
+    {
+      "epoch": 0.19314717717351543,
+      "grad_norm": 1.0010562488790584,
+      "learning_rate": 4.727333052613749e-06,
+      "loss": 0.1689,
+      "step": 198
+    },
+    {
+      "epoch": 0.19412266796732106,
+      "grad_norm": 0.943204986199299,
+      "learning_rate": 4.7236557114175705e-06,
+      "loss": 0.153,
+      "step": 199
+    },
+    {
+      "epoch": 0.19509815876112668,
+      "grad_norm": 1.004212156065797,
+      "learning_rate": 4.71995518873743e-06,
+      "loss": 0.1738,
+      "step": 200
+    },
+    {
+      "epoch": 0.19607364955493234,
+      "grad_norm": 0.9943567195079659,
+      "learning_rate": 4.716231523151065e-06,
+      "loss": 0.1621,
+      "step": 201
+    },
+    {
+      "epoch": 0.19704914034873797,
+      "grad_norm": 0.9438877027311198,
+      "learning_rate": 4.712484753477478e-06,
+      "loss": 0.1608,
+      "step": 202
+    },
+    {
+      "epoch": 0.1980246311425436,
+      "grad_norm": 1.0575615752386678,
+      "learning_rate": 4.708714918776527e-06,
+      "loss": 0.1712,
+      "step": 203
+    },
+    {
+      "epoch": 0.19900012193634922,
+      "grad_norm": 0.9584987238953627,
+      "learning_rate": 4.704922058348526e-06,
+      "loss": 0.1586,
+      "step": 204
+    },
+    {
+      "epoch": 0.19997561273015485,
+      "grad_norm": 0.9822024229285666,
+      "learning_rate": 4.701106211733827e-06,
+      "loss": 0.1633,
+      "step": 205
+    },
+    {
+      "epoch": 0.2009511035239605,
+      "grad_norm": 1.0567476044736286,
+      "learning_rate": 4.697267418712415e-06,
+      "loss": 0.1682,
+      "step": 206
+    },
+    {
+      "epoch": 0.20192659431776613,
+      "grad_norm": 0.9248879785588529,
+      "learning_rate": 4.693405719303487e-06,
+      "loss": 0.158,
+      "step": 207
+    },
+    {
+      "epoch": 0.20290208511157176,
+      "grad_norm": 1.0157347860608708,
+      "learning_rate": 4.689521153765039e-06,
+      "loss": 0.1585,
+      "step": 208
+    },
+    {
+      "epoch": 0.2038775759053774,
+      "grad_norm": 1.0718965104259908,
+      "learning_rate": 4.685613762593446e-06,
+      "loss": 0.1664,
+      "step": 209
+    },
+    {
+      "epoch": 0.20485306669918302,
+      "grad_norm": 1.1039750845125622,
+      "learning_rate": 4.681683586523037e-06,
+      "loss": 0.1678,
+      "step": 210
+    },
+    {
+      "epoch": 0.20582855749298867,
+      "grad_norm": 1.0549278846097074,
+      "learning_rate": 4.677730666525675e-06,
+      "loss": 0.1661,
+      "step": 211
+    },
+    {
+      "epoch": 0.2068040482867943,
+      "grad_norm": 1.0350210106238715,
+      "learning_rate": 4.673755043810324e-06,
+      "loss": 0.1629,
+      "step": 212
+    },
+    {
+      "epoch": 0.20777953908059993,
+      "grad_norm": 1.0928587210817202,
+      "learning_rate": 4.669756759822625e-06,
+      "loss": 0.1641,
+      "step": 213
+    },
+    {
+      "epoch": 0.20875502987440556,
+      "grad_norm": 1.0139520854082134,
+      "learning_rate": 4.66573585624446e-06,
+      "loss": 0.1689,
+      "step": 214
+    },
+    {
+      "epoch": 0.20973052066821118,
+      "grad_norm": 1.1017739109840923,
+      "learning_rate": 4.661692374993519e-06,
+      "loss": 0.1624,
+      "step": 215
+    },
+    {
+      "epoch": 0.21070601146201684,
+      "grad_norm": 1.2340168628656984,
+      "learning_rate": 4.657626358222864e-06,
+      "loss": 0.1756,
+      "step": 216
+    },
+    {
+      "epoch": 0.21168150225582247,
+      "grad_norm": 0.9230324919898438,
+      "learning_rate": 4.653537848320488e-06,
+      "loss": 0.1602,
+      "step": 217
+    },
+    {
+      "epoch": 0.2126569930496281,
+      "grad_norm": 1.0516263439718125,
+      "learning_rate": 4.6494268879088745e-06,
+      "loss": 0.161,
+      "step": 218
+    },
+    {
+      "epoch": 0.21363248384343372,
+      "grad_norm": 1.0294336700433153,
+      "learning_rate": 4.6452935198445496e-06,
+      "loss": 0.169,
+      "step": 219
+    },
+    {
+      "epoch": 0.21460797463723935,
+      "grad_norm": 0.950242561108817,
+      "learning_rate": 4.64113778721764e-06,
+      "loss": 0.1686,
+      "step": 220
+    },
+    {
+      "epoch": 0.215583465431045,
+      "grad_norm": 1.0972582723890745,
+      "learning_rate": 4.636959733351422e-06,
+      "loss": 0.1581,
+      "step": 221
+    },
+    {
+      "epoch": 0.21655895622485064,
+      "grad_norm": 0.9521794002072761,
+      "learning_rate": 4.632759401801869e-06,
+      "loss": 0.1592,
+      "step": 222
+    },
+    {
+      "epoch": 0.21753444701865626,
+      "grad_norm": 1.1612379311952095,
+      "learning_rate": 4.628536836357196e-06,
+      "loss": 0.166,
+      "step": 223
+    },
+    {
+      "epoch": 0.2185099378124619,
+      "grad_norm": 0.9211219216002743,
+      "learning_rate": 4.6242920810374095e-06,
+      "loss": 0.1542,
+      "step": 224
+    },
+    {
+      "epoch": 0.21948542860626752,
+      "grad_norm": 0.937788934386043,
+      "learning_rate": 4.62002518009384e-06,
+      "loss": 0.1725,
+      "step": 225
+    },
+    {
+      "epoch": 0.22046091940007317,
+      "grad_norm": 0.9773446856643697,
+      "learning_rate": 4.615736178008687e-06,
+      "loss": 0.1566,
+      "step": 226
+    },
+    {
+      "epoch": 0.2214364101938788,
+      "grad_norm": 0.9516125327713063,
+      "learning_rate": 4.611425119494552e-06,
+      "loss": 0.158,
+      "step": 227
+    },
+    {
+      "epoch": 0.22241190098768443,
+      "grad_norm": 1.0157719146604811,
+      "learning_rate": 4.607092049493973e-06,
+      "loss": 0.1633,
+      "step": 228
+    },
+    {
+      "epoch": 0.22338739178149006,
+      "grad_norm": 0.9692342820358121,
+      "learning_rate": 4.6027370131789614e-06,
+      "loss": 0.1622,
+      "step": 229
+    },
+    {
+      "epoch": 0.22436288257529569,
+      "grad_norm": 0.9487979994417061,
+      "learning_rate": 4.5983600559505196e-06,
+      "loss": 0.1596,
+      "step": 230
+    },
+    {
+      "epoch": 0.22533837336910134,
+      "grad_norm": 1.0470848627284035,
+      "learning_rate": 4.593961223438177e-06,
+      "loss": 0.1707,
+      "step": 231
+    },
+    {
+      "epoch": 0.22631386416290697,
+      "grad_norm": 0.9284433450109106,
+      "learning_rate": 4.589540561499512e-06,
+      "loss": 0.149,
+      "step": 232
+    },
+    {
+      "epoch": 0.2272893549567126,
+      "grad_norm": 1.0411956514134435,
+      "learning_rate": 4.585098116219674e-06,
+      "loss": 0.1666,
+      "step": 233
+    },
+    {
+      "epoch": 0.22826484575051822,
+      "grad_norm": 0.9982980604258467,
+      "learning_rate": 4.580633933910901e-06,
+      "loss": 0.1582,
+      "step": 234
+    },
+    {
+      "epoch": 0.22924033654432385,
+      "grad_norm": 1.0353148821593277,
+      "learning_rate": 4.576148061112039e-06,
+      "loss": 0.1798,
+      "step": 235
+    },
+    {
+      "epoch": 0.2302158273381295,
+      "grad_norm": 0.8623535833361177,
+      "learning_rate": 4.571640544588056e-06,
+      "loss": 0.1568,
+      "step": 236
+    },
+    {
+      "epoch": 0.23119131813193514,
+      "grad_norm": 1.141933497627487,
+      "learning_rate": 4.567111431329555e-06,
+      "loss": 0.1556,
+      "step": 237
+    },
+    {
+      "epoch": 0.23216680892574076,
+      "grad_norm": 0.9999053398774652,
+      "learning_rate": 4.562560768552283e-06,
+      "loss": 0.1603,
+      "step": 238
+    },
+    {
+      "epoch": 0.2331422997195464,
+      "grad_norm": 0.9284632452687369,
+      "learning_rate": 4.55798860369664e-06,
+      "loss": 0.1458,
+      "step": 239
+    },
+    {
+      "epoch": 0.23411779051335202,
+      "grad_norm": 0.9872863479156412,
+      "learning_rate": 4.553394984427184e-06,
+      "loss": 0.1621,
+      "step": 240
+    },
+    {
+      "epoch": 0.23509328130715768,
+      "grad_norm": 0.9100468729196777,
+      "learning_rate": 4.548779958632134e-06,
+      "loss": 0.1568,
+      "step": 241
+    },
+    {
+      "epoch": 0.2360687721009633,
+      "grad_norm": 0.8444568550070792,
+      "learning_rate": 4.54414357442287e-06,
+      "loss": 0.1408,
+      "step": 242
+    },
+    {
+      "epoch": 0.23704426289476893,
+      "grad_norm": 1.0223092529639413,
+      "learning_rate": 4.539485880133433e-06,
+      "loss": 0.1609,
+      "step": 243
+    },
+    {
+      "epoch": 0.23801975368857456,
+      "grad_norm": 0.9414329883963978,
+      "learning_rate": 4.534806924320021e-06,
+      "loss": 0.1528,
+      "step": 244
+    },
+    {
+      "epoch": 0.2389952444823802,
+      "grad_norm": 1.0175078568206852,
+      "learning_rate": 4.5301067557604815e-06,
+      "loss": 0.1607,
+      "step": 245
+    },
+    {
+      "epoch": 0.23997073527618584,
+      "grad_norm": 0.9503464111258508,
+      "learning_rate": 4.525385423453803e-06,
+      "loss": 0.1621,
+      "step": 246
+    },
+    {
+      "epoch": 0.24094622606999147,
+      "grad_norm": 1.0317509367040238,
+      "learning_rate": 4.520642976619607e-06,
+      "loss": 0.1562,
+      "step": 247
+    },
+    {
+      "epoch": 0.2419217168637971,
+      "grad_norm": 0.9406630214637611,
+      "learning_rate": 4.515879464697629e-06,
+      "loss": 0.1524,
+      "step": 248
+    },
+    {
+      "epoch": 0.24289720765760273,
+      "grad_norm": 0.9116204738820602,
+      "learning_rate": 4.51109493734721e-06,
+      "loss": 0.1589,
+      "step": 249
+    },
+    {
+      "epoch": 0.24387269845140835,
+      "grad_norm": 0.9686707124955454,
+      "learning_rate": 4.506289444446775e-06,
+      "loss": 0.1661,
+      "step": 250
+    },
+    {
+      "epoch": 0.244848189245214,
+      "grad_norm": 0.8855697170450265,
+      "learning_rate": 4.5014630360933136e-06,
+      "loss": 0.1525,
+      "step": 251
+    },
+    {
+      "epoch": 0.24582368003901964,
+      "grad_norm": 0.8408850368006234,
+      "learning_rate": 4.496615762601857e-06,
+      "loss": 0.1553,
+      "step": 252
+    },
+    {
+      "epoch": 0.24679917083282527,
+      "grad_norm": 0.9791331885043342,
+      "learning_rate": 4.491747674504956e-06,
+      "loss": 0.1613,
+      "step": 253
+    },
+    {
+      "epoch": 0.2477746616266309,
+      "grad_norm": 0.8536640710320617,
+      "learning_rate": 4.48685882255215e-06,
+      "loss": 0.1561,
+      "step": 254
+    },
+    {
+      "epoch": 0.24875015242043652,
+      "grad_norm": 0.9108673457445514,
+      "learning_rate": 4.481949257709442e-06,
+      "loss": 0.1581,
+      "step": 255
+    },
+    {
+      "epoch": 0.24972564321424218,
+      "grad_norm": 0.9929310498573025,
+      "learning_rate": 4.477019031158767e-06,
+      "loss": 0.1635,
+      "step": 256
+    },
+    {
+      "epoch": 0.2507011340080478,
+      "grad_norm": 0.9319735021715781,
+      "learning_rate": 4.472068194297453e-06,
+      "loss": 0.1535,
+      "step": 257
+    },
+    {
+      "epoch": 0.25167662480185343,
+      "grad_norm": 0.8234410520072005,
+      "learning_rate": 4.467096798737694e-06,
+      "loss": 0.1419,
+      "step": 258
+    },
+    {
+      "epoch": 0.2526521155956591,
+      "grad_norm": 0.8293157655059741,
+      "learning_rate": 4.462104896306004e-06,
+      "loss": 0.1509,
+      "step": 259
+    },
+    {
+      "epoch": 0.2536276063894647,
+      "grad_norm": 0.981516914054671,
+      "learning_rate": 4.457092539042682e-06,
+      "loss": 0.1503,
+      "step": 260
+    },
+    {
+      "epoch": 0.25460309718327034,
+      "grad_norm": 0.9280314009979089,
+      "learning_rate": 4.452059779201267e-06,
+      "loss": 0.1524,
+      "step": 261
+    },
+    {
+      "epoch": 0.25557858797707594,
+      "grad_norm": 0.9107843402176423,
+      "learning_rate": 4.44700666924799e-06,
+      "loss": 0.1599,
+      "step": 262
+    },
+    {
+      "epoch": 0.2565540787708816,
+      "grad_norm": 0.9504200684566895,
+      "learning_rate": 4.441933261861239e-06,
+      "loss": 0.1603,
+      "step": 263
+    },
+    {
+      "epoch": 0.25752956956468726,
+      "grad_norm": 0.8932484399394331,
+      "learning_rate": 4.436839609930992e-06,
+      "loss": 0.166,
+      "step": 264
+    },
+    {
+      "epoch": 0.25850506035849286,
+      "grad_norm": 0.9657424201285992,
+      "learning_rate": 4.431725766558284e-06,
+      "loss": 0.1592,
+      "step": 265
+    },
+    {
+      "epoch": 0.2594805511522985,
+      "grad_norm": 0.9280800047846618,
+      "learning_rate": 4.426591785054637e-06,
+      "loss": 0.1692,
+      "step": 266
+    },
+    {
+      "epoch": 0.2604560419461041,
+      "grad_norm": 0.909753774216958,
+      "learning_rate": 4.421437718941517e-06,
+      "loss": 0.1546,
+      "step": 267
+    },
+    {
+      "epoch": 0.26143153273990977,
+      "grad_norm": 0.9249375767292966,
+      "learning_rate": 4.416263621949769e-06,
+      "loss": 0.158,
+      "step": 268
+    },
+    {
+      "epoch": 0.2624070235337154,
+      "grad_norm": 0.9638546484727502,
+      "learning_rate": 4.41106954801906e-06,
+      "loss": 0.1549,
+      "step": 269
+    },
+    {
+      "epoch": 0.263382514327521,
+      "grad_norm": 0.9676726584506395,
+      "learning_rate": 4.4058555512973135e-06,
+      "loss": 0.1526,
+      "step": 270
+    },
+    {
+      "epoch": 0.2643580051213267,
+      "grad_norm": 0.9643850490217997,
+      "learning_rate": 4.4006216861401475e-06,
+      "loss": 0.1507,
+      "step": 271
+    },
+    {
+      "epoch": 0.2653334959151323,
+      "grad_norm": 0.9455841602811957,
+      "learning_rate": 4.395368007110307e-06,
+      "loss": 0.1609,
+      "step": 272
+    },
+    {
+      "epoch": 0.26630898670893793,
+      "grad_norm": 1.0032153539367255,
+      "learning_rate": 4.390094568977099e-06,
+      "loss": 0.1517,
+      "step": 273
+    },
+    {
+      "epoch": 0.2672844775027436,
+      "grad_norm": 0.9286676590827331,
+      "learning_rate": 4.384801426715814e-06,
+      "loss": 0.1509,
+      "step": 274
+    },
+    {
+      "epoch": 0.2682599682965492,
+      "grad_norm": 0.8635512942768435,
+      "learning_rate": 4.379488635507157e-06,
+      "loss": 0.1479,
+      "step": 275
+    },
+    {
+      "epoch": 0.26923545909035485,
+      "grad_norm": 0.9730864421049288,
+      "learning_rate": 4.3741562507366754e-06,
+      "loss": 0.1532,
+      "step": 276
+    },
+    {
+      "epoch": 0.27021094988416045,
+      "grad_norm": 0.8738885156811014,
+      "learning_rate": 4.368804327994174e-06,
+      "loss": 0.1579,
+      "step": 277
+    },
+    {
+      "epoch": 0.2711864406779661,
+      "grad_norm": 0.9907365851233481,
+      "learning_rate": 4.363432923073144e-06,
+      "loss": 0.1543,
+      "step": 278
+    },
+    {
+      "epoch": 0.27216193147177176,
+      "grad_norm": 0.9242295443836225,
+      "learning_rate": 4.3580420919701745e-06,
+      "loss": 0.1584,
+      "step": 279
+    },
+    {
+      "epoch": 0.27313742226557736,
+      "grad_norm": 0.9088209928537309,
+      "learning_rate": 4.352631890884373e-06,
+      "loss": 0.1547,
+      "step": 280
+    },
+    {
+      "epoch": 0.274112913059383,
+      "grad_norm": 0.9101013673566347,
+      "learning_rate": 4.347202376216775e-06,
+      "loss": 0.1648,
+      "step": 281
+    },
+    {
+      "epoch": 0.2750884038531886,
+      "grad_norm": 0.9053322800776012,
+      "learning_rate": 4.341753604569764e-06,
+      "loss": 0.1488,
+      "step": 282
+    },
+    {
+      "epoch": 0.27606389464699427,
+      "grad_norm": 0.9622509070577334,
+      "learning_rate": 4.336285632746472e-06,
+      "loss": 0.155,
+      "step": 283
+    },
+    {
+      "epoch": 0.2770393854407999,
+      "grad_norm": 0.9097255445557655,
+      "learning_rate": 4.330798517750194e-06,
+      "loss": 0.138,
+      "step": 284
+    },
+    {
+      "epoch": 0.2780148762346055,
+      "grad_norm": 0.915630210467682,
+      "learning_rate": 4.3252923167837905e-06,
+      "loss": 0.1533,
+      "step": 285
+    },
+    {
+      "epoch": 0.2789903670284112,
+      "grad_norm": 0.9410519479526743,
+      "learning_rate": 4.319767087249094e-06,
+      "loss": 0.1588,
+      "step": 286
+    },
+    {
+      "epoch": 0.2799658578222168,
+      "grad_norm": 0.9225807794360538,
+      "learning_rate": 4.314222886746304e-06,
+      "loss": 0.1546,
+      "step": 287
+    },
+    {
+      "epoch": 0.28094134861602243,
+      "grad_norm": 0.8732635665891079,
+      "learning_rate": 4.308659773073398e-06,
+      "loss": 0.1677,
+      "step": 288
+    },
+    {
+      "epoch": 0.2819168394098281,
+      "grad_norm": 0.979628005682833,
+      "learning_rate": 4.303077804225517e-06,
+      "loss": 0.1626,
+      "step": 289
+    },
+    {
+      "epoch": 0.2828923302036337,
+      "grad_norm": 0.9055375259430347,
+      "learning_rate": 4.297477038394368e-06,
+      "loss": 0.1675,
+      "step": 290
+    },
+    {
+      "epoch": 0.28386782099743935,
+      "grad_norm": 0.9480989580020738,
+      "learning_rate": 4.291857533967616e-06,
+      "loss": 0.1574,
+      "step": 291
+    },
+    {
+      "epoch": 0.28484331179124495,
+      "grad_norm": 0.8985804614382488,
+      "learning_rate": 4.286219349528274e-06,
+      "loss": 0.1557,
+      "step": 292
+    },
+    {
+      "epoch": 0.2858188025850506,
+      "grad_norm": 1.0146034373434893,
+      "learning_rate": 4.280562543854091e-06,
+      "loss": 0.1645,
+      "step": 293
+    },
+    {
+      "epoch": 0.28679429337885626,
+      "grad_norm": 0.890713161023523,
+      "learning_rate": 4.274887175916946e-06,
+      "loss": 0.1567,
+      "step": 294
+    },
+    {
+      "epoch": 0.28776978417266186,
+      "grad_norm": 0.9267996009148974,
+      "learning_rate": 4.269193304882226e-06,
+      "loss": 0.16,
+      "step": 295
+    },
+    {
+      "epoch": 0.2887452749664675,
+      "grad_norm": 0.8938807739528849,
+      "learning_rate": 4.263480990108212e-06,
+      "loss": 0.1578,
+      "step": 296
+    },
+    {
+      "epoch": 0.2897207657602731,
+      "grad_norm": 0.9382622618633856,
+      "learning_rate": 4.257750291145457e-06,
+      "loss": 0.1594,
+      "step": 297
+    },
+    {
+      "epoch": 0.29069625655407877,
+      "grad_norm": 0.9825847935980099,
+      "learning_rate": 4.252001267736174e-06,
+      "loss": 0.1555,
+      "step": 298
+    },
+    {
+      "epoch": 0.2916717473478844,
+      "grad_norm": 0.8990364135002812,
+      "learning_rate": 4.246233979813602e-06,
+      "loss": 0.1603,
+      "step": 299
+    },
+    {
+      "epoch": 0.29264723814169,
+      "grad_norm": 0.8656451974006956,
+      "learning_rate": 4.24044848750139e-06,
+      "loss": 0.1491,
+      "step": 300
+    },
+    {
+      "epoch": 0.2936227289354957,
+      "grad_norm": 0.9253142530395264,
+      "learning_rate": 4.234644851112965e-06,
+      "loss": 0.1607,
+      "step": 301
+    },
+    {
+      "epoch": 0.2945982197293013,
+      "grad_norm": 0.8969951216981602,
+      "learning_rate": 4.228823131150904e-06,
+      "loss": 0.1521,
+      "step": 302
+    },
+    {
+      "epoch": 0.29557371052310694,
+      "grad_norm": 0.9403876224279274,
+      "learning_rate": 4.222983388306308e-06,
+      "loss": 0.1616,
+      "step": 303
+    },
+    {
+      "epoch": 0.2965492013169126,
+      "grad_norm": 0.8925996666326684,
+      "learning_rate": 4.217125683458162e-06,
+      "loss": 0.1581,
+      "step": 304
+    },
+    {
+      "epoch": 0.2975246921107182,
+      "grad_norm": 0.8585766705268447,
+      "learning_rate": 4.211250077672704e-06,
+      "loss": 0.1492,
+      "step": 305
+    },
+    {
+      "epoch": 0.29850018290452385,
+      "grad_norm": 0.9350703216606489,
+      "learning_rate": 4.20535663220279e-06,
+      "loss": 0.1587,
+      "step": 306
+    },
+    {
+      "epoch": 0.29947567369832945,
+      "grad_norm": 0.9188676922420881,
+      "learning_rate": 4.199445408487253e-06,
+      "loss": 0.1593,
+      "step": 307
+    },
+    {
+      "epoch": 0.3004511644921351,
+      "grad_norm": 0.9547113904108987,
+      "learning_rate": 4.1935164681502626e-06,
+      "loss": 0.151,
+      "step": 308
+    },
+    {
+      "epoch": 0.30142665528594076,
+      "grad_norm": 1.0033584349157778,
+      "learning_rate": 4.187569873000684e-06,
+      "loss": 0.1629,
+      "step": 309
+    },
+    {
+      "epoch": 0.30240214607974636,
+      "grad_norm": 0.9859571353144448,
+      "learning_rate": 4.181605685031433e-06,
+      "loss": 0.1734,
+      "step": 310
+    },
+    {
+      "epoch": 0.303377636873552,
+      "grad_norm": 0.8228370416849764,
+      "learning_rate": 4.1756239664188275e-06,
+      "loss": 0.1509,
+      "step": 311
+    },
+    {
+      "epoch": 0.3043531276673576,
+      "grad_norm": 1.0367963117038492,
+      "learning_rate": 4.169624779521944e-06,
+      "loss": 0.1571,
+      "step": 312
+    },
+    {
+      "epoch": 0.30532861846116327,
+      "grad_norm": 0.9669617210813681,
+      "learning_rate": 4.163608186881964e-06,
+      "loss": 0.1628,
+      "step": 313
+    },
+    {
+      "epoch": 0.3063041092549689,
+      "grad_norm": 1.0103654228797525,
+      "learning_rate": 4.157574251221522e-06,
+      "loss": 0.1595,
+      "step": 314
+    },
+    {
+      "epoch": 0.3072796000487745,
+      "grad_norm": 0.9081265589361568,
+      "learning_rate": 4.1515230354440526e-06,
+      "loss": 0.1524,
+      "step": 315
+    },
+    {
+      "epoch": 0.3082550908425802,
+      "grad_norm": 0.924809859907981,
+      "learning_rate": 4.145454602633137e-06,
+      "loss": 0.1619,
+      "step": 316
+    },
+    {
+      "epoch": 0.3092305816363858,
+      "grad_norm": 0.8914505960487598,
+      "learning_rate": 4.139369016051838e-06,
+      "loss": 0.1603,
+      "step": 317
+    },
+    {
+      "epoch": 0.31020607243019144,
+      "grad_norm": 0.8961623985939995,
+      "learning_rate": 4.1332663391420515e-06,
+      "loss": 0.1604,
+      "step": 318
+    },
+    {
+      "epoch": 0.3111815632239971,
+      "grad_norm": 0.8168107498311156,
+      "learning_rate": 4.127146635523835e-06,
+      "loss": 0.1437,
+      "step": 319
+    },
+    {
+      "epoch": 0.3121570540178027,
+      "grad_norm": 0.8664110012815645,
+      "learning_rate": 4.121009968994751e-06,
+      "loss": 0.1506,
+      "step": 320
+    },
+    {
+      "epoch": 0.31313254481160835,
+      "grad_norm": 0.9602948973971756,
+      "learning_rate": 4.114856403529196e-06,
+      "loss": 0.1463,
+      "step": 321
+    },
+    {
+      "epoch": 0.31410803560541395,
+      "grad_norm": 0.9778713250873332,
+      "learning_rate": 4.10868600327774e-06,
+      "loss": 0.1615,
+      "step": 322
+    },
+    {
+      "epoch": 0.3150835263992196,
+      "grad_norm": 1.0278280133622393,
+      "learning_rate": 4.102498832566454e-06,
+      "loss": 0.1537,
+      "step": 323
+    },
+    {
+      "epoch": 0.31605901719302526,
+      "grad_norm": 0.9256019026472626,
+      "learning_rate": 4.096294955896239e-06,
+      "loss": 0.1611,
+      "step": 324
+    },
+    {
+      "epoch": 0.31703450798683086,
+      "grad_norm": 0.8287537359495541,
+      "learning_rate": 4.090074437942155e-06,
+      "loss": 0.1533,
+      "step": 325
+    },
+    {
+      "epoch": 0.3180099987806365,
+      "grad_norm": 0.8835278034235962,
+      "learning_rate": 4.083837343552749e-06,
+      "loss": 0.1493,
+      "step": 326
+    },
+    {
+      "epoch": 0.3189854895744421,
+      "grad_norm": 0.8259069804781106,
+      "learning_rate": 4.077583737749373e-06,
+      "loss": 0.1424,
+      "step": 327
+    },
+    {
+      "epoch": 0.31996098036824777,
+      "grad_norm": 0.941189764723523,
+      "learning_rate": 4.07131368572551e-06,
+      "loss": 0.1523,
+      "step": 328
+    },
+    {
+      "epoch": 0.3209364711620534,
+      "grad_norm": 0.9164961397020092,
+      "learning_rate": 4.0650272528460955e-06,
+      "loss": 0.1539,
+      "step": 329
+    },
+    {
+      "epoch": 0.321911961955859,
+      "grad_norm": 0.8541305146982892,
+      "learning_rate": 4.058724504646834e-06,
+      "loss": 0.1377,
+      "step": 330
+    },
+    {
+      "epoch": 0.3228874527496647,
+      "grad_norm": 0.9562169457452313,
+      "learning_rate": 4.052405506833516e-06,
+      "loss": 0.1534,
+      "step": 331
+    },
+    {
+      "epoch": 0.3238629435434703,
+      "grad_norm": 0.9199148267026572,
+      "learning_rate": 4.046070325281333e-06,
+      "loss": 0.1471,
+      "step": 332
+    },
+    {
+      "epoch": 0.32483843433727594,
+      "grad_norm": 0.9205784248133079,
+      "learning_rate": 4.039719026034191e-06,
+      "loss": 0.1403,
+      "step": 333
+    },
+    {
+      "epoch": 0.3258139251310816,
+      "grad_norm": 0.8621247018342253,
+      "learning_rate": 4.0333516753040225e-06,
+      "loss": 0.1365,
+      "step": 334
+    },
+    {
+      "epoch": 0.3267894159248872,
+      "grad_norm": 0.9318993173542444,
+      "learning_rate": 4.026968339470097e-06,
+      "loss": 0.1446,
+      "step": 335
+    },
+    {
+      "epoch": 0.32776490671869285,
+      "grad_norm": 0.9970504184721904,
+      "learning_rate": 4.020569085078324e-06,
+      "loss": 0.1553,
+      "step": 336
+    },
+    {
+      "epoch": 0.32874039751249845,
+      "grad_norm": 0.9288139122169886,
+      "learning_rate": 4.014153978840568e-06,
+      "loss": 0.1457,
+      "step": 337
+    },
+    {
+      "epoch": 0.3297158883063041,
+      "grad_norm": 0.947326368514792,
+      "learning_rate": 4.007723087633943e-06,
+      "loss": 0.1567,
+      "step": 338
+    },
+    {
+      "epoch": 0.33069137910010976,
+      "grad_norm": 0.933169285653779,
+      "learning_rate": 4.001276478500127e-06,
+      "loss": 0.1564,
+      "step": 339
+    },
+    {
+      "epoch": 0.33166686989391536,
+      "grad_norm": 0.8619195635693533,
+      "learning_rate": 3.994814218644649e-06,
+      "loss": 0.1409,
+      "step": 340
+    },
+    {
+      "epoch": 0.332642360687721,
+      "grad_norm": 0.9533171506313436,
+      "learning_rate": 3.988336375436201e-06,
+      "loss": 0.1472,
+      "step": 341
+    },
+    {
+      "epoch": 0.3336178514815266,
+      "grad_norm": 0.9085852674744761,
+      "learning_rate": 3.981843016405928e-06,
+      "loss": 0.1518,
+      "step": 342
+    },
+    {
+      "epoch": 0.3345933422753323,
+      "grad_norm": 0.8015484531793845,
+      "learning_rate": 3.975334209246727e-06,
+      "loss": 0.1448,
+      "step": 343
+    },
+    {
+      "epoch": 0.33556883306913793,
+      "grad_norm": 0.9409823279258879,
+      "learning_rate": 3.968810021812544e-06,
+      "loss": 0.1555,
+      "step": 344
+    },
+    {
+      "epoch": 0.33654432386294353,
+      "grad_norm": 0.8142592274829353,
+      "learning_rate": 3.962270522117659e-06,
+      "loss": 0.1457,
+      "step": 345
+    },
+    {
+      "epoch": 0.3375198146567492,
+      "grad_norm": 0.9076632142318111,
+      "learning_rate": 3.955715778335984e-06,
+      "loss": 0.1538,
+      "step": 346
+    },
+    {
+      "epoch": 0.3384953054505548,
+      "grad_norm": 0.8974350187857667,
+      "learning_rate": 3.949145858800348e-06,
+      "loss": 0.1481,
+      "step": 347
+    },
+    {
+      "epoch": 0.33947079624436044,
+      "grad_norm": 0.8901297893092709,
+      "learning_rate": 3.942560832001789e-06,
+      "loss": 0.1503,
+      "step": 348
+    },
+    {
+      "epoch": 0.3404462870381661,
+      "grad_norm": 0.8560837002365613,
+      "learning_rate": 3.935960766588835e-06,
+      "loss": 0.1528,
+      "step": 349
+    },
+    {
+      "epoch": 0.3414217778319717,
+      "grad_norm": 0.8443274184309183,
+      "learning_rate": 3.9293457313667905e-06,
+      "loss": 0.1436,
+      "step": 350
+    },
+    {
+      "epoch": 0.34239726862577735,
+      "grad_norm": 1.0282756436782237,
+      "learning_rate": 3.922715795297022e-06,
+      "loss": 0.1482,
+      "step": 351
+    },
+    {
+      "epoch": 0.34337275941958295,
+      "grad_norm": 0.9065229091333961,
+      "learning_rate": 3.916071027496234e-06,
+      "loss": 0.1445,
+      "step": 352
+    },
+    {
+      "epoch": 0.3443482502133886,
+      "grad_norm": 0.9714447847850254,
+      "learning_rate": 3.909411497235752e-06,
+      "loss": 0.1473,
+      "step": 353
+    },
+    {
+      "epoch": 0.34532374100719426,
+      "grad_norm": 0.9561585485417536,
+      "learning_rate": 3.902737273940799e-06,
+      "loss": 0.1471,
+      "step": 354
+    },
+    {
+      "epoch": 0.34629923180099986,
+      "grad_norm": 0.8900543390831095,
+      "learning_rate": 3.8960484271897736e-06,
+      "loss": 0.1527,
+      "step": 355
+    },
+    {
+      "epoch": 0.3472747225948055,
+      "grad_norm": 0.9827409295726128,
+      "learning_rate": 3.889345026713519e-06,
+      "loss": 0.1534,
+      "step": 356
+    },
+    {
+      "epoch": 0.3482502133886111,
+      "grad_norm": 0.8940002734667856,
+      "learning_rate": 3.882627142394605e-06,
+      "loss": 0.1486,
+      "step": 357
+    },
+    {
+      "epoch": 0.3492257041824168,
+      "grad_norm": 0.9944296349312072,
+      "learning_rate": 3.875894844266592e-06,
+      "loss": 0.1531,
+      "step": 358
+    },
+    {
+      "epoch": 0.35020119497622243,
+      "grad_norm": 0.9120571509586781,
+      "learning_rate": 3.869148202513303e-06,
+      "loss": 0.1606,
+      "step": 359
+    },
+    {
+      "epoch": 0.35117668577002803,
+      "grad_norm": 0.9095948987953094,
+      "learning_rate": 3.862387287468095e-06,
+      "loss": 0.137,
+      "step": 360
+    },
+    {
+      "epoch": 0.3521521765638337,
+      "grad_norm": 0.877930473526069,
+      "learning_rate": 3.85561216961312e-06,
+      "loss": 0.1408,
+      "step": 361
+    },
+    {
+      "epoch": 0.3531276673576393,
+      "grad_norm": 0.9699360794407749,
+      "learning_rate": 3.8488229195785965e-06,
+      "loss": 0.147,
+      "step": 362
+    },
+    {
+      "epoch": 0.35410315815144494,
+      "grad_norm": 0.8874366855608934,
+      "learning_rate": 3.842019608142068e-06,
+      "loss": 0.1446,
+      "step": 363
+    },
+    {
+      "epoch": 0.3550786489452506,
+      "grad_norm": 0.8644114874928793,
+      "learning_rate": 3.83520230622767e-06,
+      "loss": 0.1484,
+      "step": 364
+    },
+    {
+      "epoch": 0.3560541397390562,
+      "grad_norm": 0.9393508952046424,
+      "learning_rate": 3.8283710849053835e-06,
+      "loss": 0.1415,
+      "step": 365
+    },
+    {
+      "epoch": 0.35702963053286185,
+      "grad_norm": 0.915863410960202,
+      "learning_rate": 3.821526015390302e-06,
+      "loss": 0.1386,
+      "step": 366
+    },
+    {
+      "epoch": 0.35800512132666745,
+      "grad_norm": 0.9939022499746282,
+      "learning_rate": 3.814667169041887e-06,
+      "loss": 0.156,
+      "step": 367
+    },
+    {
+      "epoch": 0.3589806121204731,
+      "grad_norm": 0.9698635070361811,
+      "learning_rate": 3.8077946173632175e-06,
+      "loss": 0.1351,
+      "step": 368
+    },
+    {
+      "epoch": 0.35995610291427876,
+      "grad_norm": 0.9476385232414736,
+      "learning_rate": 3.800908432000254e-06,
+      "loss": 0.1427,
+      "step": 369
+    },
+    {
+      "epoch": 0.36093159370808436,
+      "grad_norm": 0.856130584200129,
+      "learning_rate": 3.7940086847410875e-06,
+      "loss": 0.1516,
+      "step": 370
+    },
+    {
+      "epoch": 0.36190708450189,
+      "grad_norm": 0.8579511362373262,
+      "learning_rate": 3.7870954475151873e-06,
+      "loss": 0.1453,
+      "step": 371
+    },
+    {
+      "epoch": 0.3628825752956956,
+      "grad_norm": 0.9244210071492551,
+      "learning_rate": 3.780168792392658e-06,
+      "loss": 0.1403,
+      "step": 372
+    },
+    {
+      "epoch": 0.3638580660895013,
+      "grad_norm": 0.8916214840343722,
+      "learning_rate": 3.7732287915834842e-06,
+      "loss": 0.1344,
+      "step": 373
+    },
+    {
+      "epoch": 0.36483355688330693,
+      "grad_norm": 0.9285987965737049,
+      "learning_rate": 3.766275517436779e-06,
+      "loss": 0.1473,
+      "step": 374
+    },
+    {
+      "epoch": 0.36580904767711253,
+      "grad_norm": 1.0202253515003419,
+      "learning_rate": 3.759309042440028e-06,
+      "loss": 0.1626,
+      "step": 375
+    },
+    {
+      "epoch": 0.3667845384709182,
+      "grad_norm": 0.9202482904809027,
+      "learning_rate": 3.752329439218337e-06,
+      "loss": 0.1437,
+      "step": 376
+    },
+    {
+      "epoch": 0.3677600292647238,
+      "grad_norm": 0.868921559794134,
+      "learning_rate": 3.7453367805336697e-06,
+      "loss": 0.1411,
+      "step": 377
+    },
+    {
+      "epoch": 0.36873552005852944,
+      "grad_norm": 0.9278377375229888,
+      "learning_rate": 3.738331139284096e-06,
+      "loss": 0.1462,
+      "step": 378
+    },
+    {
+      "epoch": 0.3697110108523351,
+      "grad_norm": 0.9555741240850121,
+      "learning_rate": 3.731312588503024e-06,
+      "loss": 0.1525,
+      "step": 379
+    },
+    {
+      "epoch": 0.3706865016461407,
+      "grad_norm": 0.9746070816089245,
+      "learning_rate": 3.7242812013584474e-06,
+      "loss": 0.1509,
+      "step": 380
+    },
+    {
+      "epoch": 0.37166199243994635,
+      "grad_norm": 0.8847369605779092,
+      "learning_rate": 3.717237051152175e-06,
+      "loss": 0.1433,
+      "step": 381
+    },
+    {
+      "epoch": 0.37263748323375195,
+      "grad_norm": 1.0235063474249009,
+      "learning_rate": 3.710180211319071e-06,
+      "loss": 0.1486,
+      "step": 382
+    },
+    {
+      "epoch": 0.3736129740275576,
+      "grad_norm": 0.8950016936249444,
+      "learning_rate": 3.703110755426289e-06,
+      "loss": 0.139,
+      "step": 383
+    },
+    {
+      "epoch": 0.37458846482136327,
+      "grad_norm": 0.8847042144920154,
+      "learning_rate": 3.696028757172503e-06,
+      "loss": 0.1521,
+      "step": 384
+    },
+    {
+      "epoch": 0.37556395561516887,
+      "grad_norm": 0.9106515631528226,
+      "learning_rate": 3.68893429038714e-06,
+      "loss": 0.146,
+      "step": 385
+    },
+    {
+      "epoch": 0.3765394464089745,
+      "grad_norm": 0.9546055014669876,
+      "learning_rate": 3.681827429029613e-06,
+      "loss": 0.1431,
+      "step": 386
+    },
+    {
+      "epoch": 0.3775149372027802,
+      "grad_norm": 0.8713617926689637,
+      "learning_rate": 3.6747082471885454e-06,
+      "loss": 0.1484,
+      "step": 387
+    },
+    {
+      "epoch": 0.3784904279965858,
+      "grad_norm": 0.8981560581623195,
+      "learning_rate": 3.6675768190810023e-06,
+      "loss": 0.1405,
+      "step": 388
+    },
+    {
+      "epoch": 0.37946591879039143,
+      "grad_norm": 0.9576903681690667,
+      "learning_rate": 3.6604332190517144e-06,
+      "loss": 0.1461,
+      "step": 389
+    },
+    {
+      "epoch": 0.38044140958419703,
+      "grad_norm": 0.9982163299550169,
+      "learning_rate": 3.653277521572305e-06,
+      "loss": 0.15,
+      "step": 390
+    },
+    {
+      "epoch": 0.3814169003780027,
+      "grad_norm": 0.8524151991534643,
+      "learning_rate": 3.6461098012405116e-06,
+      "loss": 0.1446,
+      "step": 391
+    },
+    {
+      "epoch": 0.38239239117180834,
+      "grad_norm": 0.9831471719181231,
+      "learning_rate": 3.6389301327794114e-06,
+      "loss": 0.1671,
+      "step": 392
+    },
+    {
+      "epoch": 0.38336788196561394,
+      "grad_norm": 0.9182177457148154,
+      "learning_rate": 3.6317385910366365e-06,
+      "loss": 0.1476,
+      "step": 393
+    },
+    {
+      "epoch": 0.3843433727594196,
+      "grad_norm": 0.9198423171501263,
+      "learning_rate": 3.624535250983601e-06,
+      "loss": 0.153,
+      "step": 394
+    },
+    {
+      "epoch": 0.3853188635532252,
+      "grad_norm": 0.8808600354881101,
+      "learning_rate": 3.6173201877147134e-06,
+      "loss": 0.1456,
+      "step": 395
+    },
+    {
+      "epoch": 0.38629435434703085,
+      "grad_norm": 0.8829324038702485,
+      "learning_rate": 3.6100934764465973e-06,
+      "loss": 0.1413,
+      "step": 396
+    },
+    {
+      "epoch": 0.3872698451408365,
+      "grad_norm": 0.8876955529556401,
+      "learning_rate": 3.6028551925173046e-06,
+      "loss": 0.1512,
+      "step": 397
+    },
+    {
+      "epoch": 0.3882453359346421,
+      "grad_norm": 0.8555692280523547,
+      "learning_rate": 3.595605411385533e-06,
+      "loss": 0.1428,
+      "step": 398
+    },
+    {
+      "epoch": 0.38922082672844777,
+      "grad_norm": 0.8764418318857389,
+      "learning_rate": 3.5883442086298375e-06,
+      "loss": 0.1481,
+      "step": 399
+    },
+    {
+      "epoch": 0.39019631752225337,
+      "grad_norm": 0.926181501103571,
+      "learning_rate": 3.581071659947842e-06,
+      "loss": 0.1394,
+      "step": 400
+    },
+    {
+      "epoch": 0.391171808316059,
+      "grad_norm": 0.9087710728115378,
+      "learning_rate": 3.5737878411554515e-06,
+      "loss": 0.1396,
+      "step": 401
+    },
+    {
+      "epoch": 0.3921472991098647,
+      "grad_norm": 0.9084832908918349,
+      "learning_rate": 3.566492828186063e-06,
+      "loss": 0.1457,
+      "step": 402
+    },
+    {
+      "epoch": 0.3931227899036703,
+      "grad_norm": 0.9198265733188117,
+      "learning_rate": 3.5591866970897695e-06,
+      "loss": 0.1428,
+      "step": 403
+    },
+    {
+      "epoch": 0.39409828069747593,
+      "grad_norm": 0.9028014238279914,
+      "learning_rate": 3.55186952403257e-06,
+      "loss": 0.1459,
+      "step": 404
+    },
+    {
+      "epoch": 0.39507377149128153,
+      "grad_norm": 0.7803451616200923,
+      "learning_rate": 3.5445413852955783e-06,
+      "loss": 0.1405,
+      "step": 405
+    },
+    {
+      "epoch": 0.3960492622850872,
+      "grad_norm": 0.8807693627140049,
+      "learning_rate": 3.5372023572742224e-06,
+      "loss": 0.144,
+      "step": 406
+    },
+    {
+      "epoch": 0.39702475307889284,
+      "grad_norm": 0.9065679616664823,
+      "learning_rate": 3.52985251647745e-06,
+      "loss": 0.152,
+      "step": 407
+    },
+    {
+      "epoch": 0.39800024387269844,
+      "grad_norm": 0.9126785313651608,
+      "learning_rate": 3.522491939526932e-06,
+      "loss": 0.1527,
+      "step": 408
+    },
+    {
+      "epoch": 0.3989757346665041,
+      "grad_norm": 0.9401530910172149,
+      "learning_rate": 3.515120703156264e-06,
+      "loss": 0.1546,
+      "step": 409
+    },
+    {
+      "epoch": 0.3999512254603097,
+      "grad_norm": 0.919541385453652,
+      "learning_rate": 3.507738884210164e-06,
+      "loss": 0.1397,
+      "step": 410
+    },
+    {
+      "epoch": 0.40092671625411536,
+      "grad_norm": 0.9257840949340401,
+      "learning_rate": 3.500346559643675e-06,
+      "loss": 0.1435,
+      "step": 411
+    },
+    {
+      "epoch": 0.401902207047921,
+      "grad_norm": 0.9038785967347172,
+      "learning_rate": 3.4929438065213568e-06,
+      "loss": 0.147,
+      "step": 412
+    },
+    {
+      "epoch": 0.4028776978417266,
+      "grad_norm": 1.0676896384715988,
+      "learning_rate": 3.48553070201649e-06,
+      "loss": 0.1558,
+      "step": 413
+    },
+    {
+      "epoch": 0.40385318863553227,
+      "grad_norm": 0.9754468401487543,
+      "learning_rate": 3.4781073234102665e-06,
+      "loss": 0.1449,
+      "step": 414
+    },
+    {
+      "epoch": 0.40482867942933787,
+      "grad_norm": 0.8939601187260615,
+      "learning_rate": 3.470673748090984e-06,
+      "loss": 0.1396,
+      "step": 415
+    },
+    {
+      "epoch": 0.4058041702231435,
+      "grad_norm": 0.9427428788545429,
+      "learning_rate": 3.4632300535532415e-06,
+      "loss": 0.1427,
+      "step": 416
+    },
+    {
+      "epoch": 0.4067796610169492,
+      "grad_norm": 0.887782687868998,
+      "learning_rate": 3.4557763173971293e-06,
+      "loss": 0.1449,
+      "step": 417
+    },
+    {
+      "epoch": 0.4077551518107548,
+      "grad_norm": 1.0018593992113438,
+      "learning_rate": 3.4483126173274227e-06,
+      "loss": 0.1473,
+      "step": 418
+    },
+    {
+      "epoch": 0.40873064260456043,
+      "grad_norm": 0.9539144339332059,
+      "learning_rate": 3.440839031152769e-06,
+      "loss": 0.1432,
+      "step": 419
+    },
+    {
+      "epoch": 0.40970613339836603,
+      "grad_norm": 0.8796520440270823,
+      "learning_rate": 3.4333556367848773e-06,
+      "loss": 0.1356,
+      "step": 420
+    },
+    {
+      "epoch": 0.4106816241921717,
+      "grad_norm": 0.8930556857980667,
+      "learning_rate": 3.425862512237708e-06,
+      "loss": 0.1414,
+      "step": 421
+    },
+    {
+      "epoch": 0.41165711498597735,
+      "grad_norm": 0.9130386801180627,
+      "learning_rate": 3.418359735626656e-06,
+      "loss": 0.144,
+      "step": 422
+    },
+    {
+      "epoch": 0.41263260577978295,
+      "grad_norm": 0.8591156628779346,
+      "learning_rate": 3.4108473851677408e-06,
+      "loss": 0.1413,
+      "step": 423
+    },
+    {
+      "epoch": 0.4136080965735886,
+      "grad_norm": 0.9654725301304397,
+      "learning_rate": 3.4033255391767865e-06,
+      "loss": 0.1492,
+      "step": 424
+    },
+    {
+      "epoch": 0.4145835873673942,
+      "grad_norm": 0.9083019018961785,
+      "learning_rate": 3.395794276068609e-06,
+      "loss": 0.1452,
+      "step": 425
+    },
+    {
+      "epoch": 0.41555907816119986,
+      "grad_norm": 0.8891185711032482,
+      "learning_rate": 3.388253674356196e-06,
+      "loss": 0.1415,
+      "step": 426
+    },
+    {
+      "epoch": 0.4165345689550055,
+      "grad_norm": 0.9375975455278247,
+      "learning_rate": 3.380703812649891e-06,
+      "loss": 0.1403,
+      "step": 427
+    },
+    {
+      "epoch": 0.4175100597488111,
+      "grad_norm": 0.9020491738743014,
+      "learning_rate": 3.3731447696565713e-06,
+      "loss": 0.1455,
+      "step": 428
+    },
+    {
+      "epoch": 0.41848555054261677,
+      "grad_norm": 0.8574973373936586,
+      "learning_rate": 3.3655766241788285e-06,
+      "loss": 0.1368,
+      "step": 429
+    },
+    {
+      "epoch": 0.41946104133642237,
+      "grad_norm": 0.8757023776531063,
+      "learning_rate": 3.357999455114148e-06,
+      "loss": 0.1446,
+      "step": 430
+    },
+    {
+      "epoch": 0.420436532130228,
+      "grad_norm": 0.9837347529941712,
+      "learning_rate": 3.350413341454086e-06,
+      "loss": 0.1506,
+      "step": 431
+    },
+    {
+      "epoch": 0.4214120229240337,
+      "grad_norm": 0.8472380588550584,
+      "learning_rate": 3.3428183622834442e-06,
+      "loss": 0.1513,
+      "step": 432
+    },
+    {
+      "epoch": 0.4223875137178393,
+      "grad_norm": 0.8387818280951195,
+      "learning_rate": 3.3352145967794463e-06,
+      "loss": 0.1403,
+      "step": 433
+    },
+    {
+      "epoch": 0.42336300451164494,
+      "grad_norm": 0.8972955470900738,
+      "learning_rate": 3.3276021242109152e-06,
+      "loss": 0.1411,
+      "step": 434
+    },
+    {
+      "epoch": 0.42433849530545054,
+      "grad_norm": 0.9676494460221445,
+      "learning_rate": 3.319981023937442e-06,
+      "loss": 0.1508,
+      "step": 435
+    },
+    {
+      "epoch": 0.4253139860992562,
+      "grad_norm": 0.8603463763707745,
+      "learning_rate": 3.3123513754085635e-06,
+      "loss": 0.1412,
+      "step": 436
+    },
+    {
+      "epoch": 0.42628947689306185,
+      "grad_norm": 0.8735162753932564,
+      "learning_rate": 3.3047132581629297e-06,
+      "loss": 0.1356,
+      "step": 437
+    },
+    {
+      "epoch": 0.42726496768686745,
+      "grad_norm": 0.946772891936297,
+      "learning_rate": 3.297066751827478e-06,
+      "loss": 0.1364,
+      "step": 438
+    },
+    {
+      "epoch": 0.4282404584806731,
+      "grad_norm": 0.9312293341123083,
+      "learning_rate": 3.2894119361166e-06,
+      "loss": 0.1384,
+      "step": 439
+    },
+    {
+      "epoch": 0.4292159492744787,
+      "grad_norm": 0.8668697057456637,
+      "learning_rate": 3.2817488908313132e-06,
+      "loss": 0.1407,
+      "step": 440
+    },
+    {
+      "epoch": 0.43019144006828436,
+      "grad_norm": 0.9062795443546382,
+      "learning_rate": 3.274077695858428e-06,
+      "loss": 0.1511,
+      "step": 441
+    },
+    {
+      "epoch": 0.43116693086209,
+      "grad_norm": 0.9197029171414413,
+      "learning_rate": 3.2663984311697137e-06,
+      "loss": 0.1443,
+      "step": 442
+    },
+    {
+      "epoch": 0.4321424216558956,
+      "grad_norm": 0.9874318993389217,
+      "learning_rate": 3.2587111768210677e-06,
+      "loss": 0.1522,
+      "step": 443
+    },
+    {
+      "epoch": 0.43311791244970127,
+      "grad_norm": 0.8997910308707128,
+      "learning_rate": 3.251016012951678e-06,
+      "loss": 0.1376,
+      "step": 444
+    },
+    {
+      "epoch": 0.43409340324350687,
+      "grad_norm": 0.8780597656452647,
+      "learning_rate": 3.2433130197831877e-06,
+      "loss": 0.141,
+      "step": 445
+    },
+    {
+      "epoch": 0.4350688940373125,
+      "grad_norm": 0.8943843664972011,
+      "learning_rate": 3.2356022776188623e-06,
+      "loss": 0.1394,
+      "step": 446
+    },
+    {
+      "epoch": 0.4360443848311182,
+      "grad_norm": 0.9140727392888497,
+      "learning_rate": 3.227883866842749e-06,
+      "loss": 0.1384,
+      "step": 447
+    },
+    {
+      "epoch": 0.4370198756249238,
+      "grad_norm": 0.8988648983999619,
+      "learning_rate": 3.2201578679188396e-06,
+      "loss": 0.1383,
+      "step": 448
+    },
+    {
+      "epoch": 0.43799536641872944,
+      "grad_norm": 0.809856506892555,
+      "learning_rate": 3.2124243613902316e-06,
+      "loss": 0.1404,
+      "step": 449
+    },
+    {
+      "epoch": 0.43897085721253504,
+      "grad_norm": 0.9050899458550025,
+      "learning_rate": 3.204683427878291e-06,
+      "loss": 0.1467,
+      "step": 450
+    },
+    {
+      "epoch": 0.4399463480063407,
+      "grad_norm": 0.8840121082189991,
+      "learning_rate": 3.196935148081808e-06,
+      "loss": 0.1446,
+      "step": 451
+    },
+    {
+      "epoch": 0.44092183880014635,
+      "grad_norm": 0.855404279656821,
+      "learning_rate": 3.189179602776157e-06,
+      "loss": 0.1362,
+      "step": 452
+    },
+    {
+      "epoch": 0.44189732959395195,
+      "grad_norm": 0.8416485853165431,
+      "learning_rate": 3.181416872812455e-06,
+      "loss": 0.1381,
+      "step": 453
+    },
+    {
+      "epoch": 0.4428728203877576,
+      "grad_norm": 0.826136304893414,
+      "learning_rate": 3.1736470391167195e-06,
+      "loss": 0.1439,
+      "step": 454
+    },
+    {
+      "epoch": 0.4438483111815632,
+      "grad_norm": 0.888073097598988,
+      "learning_rate": 3.1658701826890237e-06,
+      "loss": 0.1472,
+      "step": 455
+    },
+    {
+      "epoch": 0.44482380197536886,
+      "grad_norm": 0.8935863712022185,
+      "learning_rate": 3.158086384602652e-06,
+      "loss": 0.1401,
+      "step": 456
+    },
+    {
+      "epoch": 0.4457992927691745,
+      "grad_norm": 0.8809290645701081,
+      "learning_rate": 3.150295726003256e-06,
+      "loss": 0.1527,
+      "step": 457
+    },
+    {
+      "epoch": 0.4467747835629801,
+      "grad_norm": 0.8388623216005323,
+      "learning_rate": 3.142498288108007e-06,
+      "loss": 0.1365,
+      "step": 458
+    },
+    {
+      "epoch": 0.44775027435678577,
+      "grad_norm": 0.9162474538739859,
+      "learning_rate": 3.13469415220475e-06,
+      "loss": 0.1405,
+      "step": 459
+    },
+    {
+      "epoch": 0.44872576515059137,
+      "grad_norm": 0.9225875521669346,
+      "learning_rate": 3.1268833996511584e-06,
+      "loss": 0.1479,
+      "step": 460
+    },
+    {
+      "epoch": 0.449701255944397,
+      "grad_norm": 0.8772214033934059,
+      "learning_rate": 3.119066111873879e-06,
+      "loss": 0.1312,
+      "step": 461
+    },
+    {
+      "epoch": 0.4506767467382027,
+      "grad_norm": 0.8397821561643128,
+      "learning_rate": 3.1112423703676937e-06,
+      "loss": 0.1344,
+      "step": 462
+    },
+    {
+      "epoch": 0.4516522375320083,
+      "grad_norm": 0.9056702372440215,
+      "learning_rate": 3.1034122566946596e-06,
+      "loss": 0.1329,
+      "step": 463
+    },
+    {
+      "epoch": 0.45262772832581394,
+      "grad_norm": 0.9828835323972128,
+      "learning_rate": 3.0955758524832663e-06,
+      "loss": 0.1454,
+      "step": 464
+    },
+    {
+      "epoch": 0.45360321911961954,
+      "grad_norm": 0.949595108992025,
+      "learning_rate": 3.0877332394275806e-06,
+      "loss": 0.1393,
+      "step": 465
+    },
+    {
+      "epoch": 0.4545787099134252,
+      "grad_norm": 0.8820964745475348,
+      "learning_rate": 3.079884499286396e-06,
+      "loss": 0.1303,
+      "step": 466
+    },
+    {
+      "epoch": 0.45555420070723085,
+      "grad_norm": 0.8539218038263852,
+      "learning_rate": 3.0720297138823814e-06,
+      "loss": 0.1369,
+      "step": 467
+    },
+    {
+      "epoch": 0.45652969150103645,
+      "grad_norm": 0.9065920605308372,
+      "learning_rate": 3.0641689651012253e-06,
+      "loss": 0.1418,
+      "step": 468
+    },
+    {
+      "epoch": 0.4575051822948421,
+      "grad_norm": 0.8602281531380704,
+      "learning_rate": 3.056302334890786e-06,
+      "loss": 0.1377,
+      "step": 469
+    },
+    {
+      "epoch": 0.4584806730886477,
+      "grad_norm": 0.8667507800649346,
+      "learning_rate": 3.0484299052602355e-06,
+      "loss": 0.1399,
+      "step": 470
+    },
+    {
+      "epoch": 0.45945616388245336,
+      "grad_norm": 0.913545132678621,
+      "learning_rate": 3.040551758279204e-06,
+      "loss": 0.1369,
+      "step": 471
+    },
+    {
+      "epoch": 0.460431654676259,
+      "grad_norm": 0.8738158833673688,
+      "learning_rate": 3.032667976076923e-06,
+      "loss": 0.1473,
+      "step": 472
+    },
+    {
+      "epoch": 0.4614071454700646,
+      "grad_norm": 0.8801680622206783,
+      "learning_rate": 3.0247786408413725e-06,
+      "loss": 0.1492,
+      "step": 473
+    },
+    {
+      "epoch": 0.4623826362638703,
+      "grad_norm": 0.876084094436277,
+      "learning_rate": 3.0168838348184226e-06,
+      "loss": 0.1429,
+      "step": 474
+    },
+    {
+      "epoch": 0.4633581270576759,
+      "grad_norm": 0.8429300809717286,
+      "learning_rate": 3.008983640310976e-06,
+      "loss": 0.1239,
+      "step": 475
+    },
+    {
+      "epoch": 0.46433361785148153,
+      "grad_norm": 0.8564086119743741,
+      "learning_rate": 3.0010781396781104e-06,
+      "loss": 0.1416,
+      "step": 476
+    },
+    {
+      "epoch": 0.4653091086452872,
+      "grad_norm": 0.9263549241409399,
+      "learning_rate": 2.993167415334218e-06,
+      "loss": 0.147,
+      "step": 477
+    },
+    {
+      "epoch": 0.4662845994390928,
+      "grad_norm": 0.9103066887514243,
+      "learning_rate": 2.9852515497481494e-06,
+      "loss": 0.149,
+      "step": 478
+    },
+    {
+      "epoch": 0.46726009023289844,
+      "grad_norm": 0.8633007227845262,
+      "learning_rate": 2.977330625442352e-06,
+      "loss": 0.1331,
+      "step": 479
+    },
+    {
+      "epoch": 0.46823558102670404,
+      "grad_norm": 0.9029118321491253,
+      "learning_rate": 2.9694047249920095e-06,
+      "loss": 0.1508,
+      "step": 480
+    },
+    {
+      "epoch": 0.4692110718205097,
+      "grad_norm": 0.8011901107322851,
+      "learning_rate": 2.961473931024182e-06,
+      "loss": 0.1347,
+      "step": 481
+    },
+    {
+      "epoch": 0.47018656261431535,
+      "grad_norm": 0.953456056583228,
+      "learning_rate": 2.953538326216944e-06,
+      "loss": 0.1386,
+      "step": 482
+    },
+    {
+      "epoch": 0.47116205340812095,
+      "grad_norm": 0.9087484393028001,
+      "learning_rate": 2.9455979932985237e-06,
+      "loss": 0.1378,
+      "step": 483
+    },
+    {
+      "epoch": 0.4721375442019266,
+      "grad_norm": 0.8453741920558421,
+      "learning_rate": 2.9376530150464388e-06,
+      "loss": 0.1366,
+      "step": 484
+    },
+    {
+      "epoch": 0.4731130349957322,
+      "grad_norm": 0.9941873485158667,
+      "learning_rate": 2.9297034742866336e-06,
+      "loss": 0.1436,
+      "step": 485
+    },
+    {
+      "epoch": 0.47408852578953786,
+      "grad_norm": 0.9362727469002547,
+      "learning_rate": 2.921749453892618e-06,
+      "loss": 0.1335,
+      "step": 486
+    },
+    {
+      "epoch": 0.4750640165833435,
+      "grad_norm": 0.8486907157670049,
+      "learning_rate": 2.913791036784601e-06,
+      "loss": 0.1311,
+      "step": 487
+    },
+    {
+      "epoch": 0.4760395073771491,
+      "grad_norm": 0.941827373025834,
+      "learning_rate": 2.9058283059286275e-06,
+      "loss": 0.1384,
+      "step": 488
+    },
+    {
+      "epoch": 0.4770149981709548,
+      "grad_norm": 1.0578890093377278,
+      "learning_rate": 2.8978613443357107e-06,
+      "loss": 0.1532,
+      "step": 489
+    },
+    {
+      "epoch": 0.4779904889647604,
+      "grad_norm": 0.9196862748457214,
+      "learning_rate": 2.889890235060972e-06,
+      "loss": 0.1451,
+      "step": 490
+    },
+    {
+      "epoch": 0.47896597975856603,
+      "grad_norm": 0.9252094112063749,
+      "learning_rate": 2.88191506120277e-06,
+      "loss": 0.1478,
+      "step": 491
+    },
+    {
+      "epoch": 0.4799414705523717,
+      "grad_norm": 0.9936636709844183,
+      "learning_rate": 2.873935905901839e-06,
+      "loss": 0.1383,
+      "step": 492
+    },
+    {
+      "epoch": 0.4809169613461773,
+      "grad_norm": 0.9220512507450012,
+      "learning_rate": 2.865952852340417e-06,
+      "loss": 0.1414,
+      "step": 493
+    },
+    {
+      "epoch": 0.48189245213998294,
+      "grad_norm": 0.8881328759002474,
+      "learning_rate": 2.8579659837413816e-06,
+      "loss": 0.1381,
+      "step": 494
+    },
+    {
+      "epoch": 0.48286794293378854,
+      "grad_norm": 0.9493132086314534,
+      "learning_rate": 2.8499753833673827e-06,
+      "loss": 0.1356,
+      "step": 495
+    },
+    {
+      "epoch": 0.4838434337275942,
+      "grad_norm": 0.8586849465773233,
+      "learning_rate": 2.8419811345199736e-06,
+      "loss": 0.1288,
+      "step": 496
+    },
+    {
+      "epoch": 0.48481892452139985,
+      "grad_norm": 0.8051743928108915,
+      "learning_rate": 2.8339833205387434e-06,
+      "loss": 0.1272,
+      "step": 497
+    },
+    {
+      "epoch": 0.48579441531520545,
+      "grad_norm": 0.8245960012411253,
+      "learning_rate": 2.825982024800445e-06,
+      "loss": 0.136,
+      "step": 498
+    },
+    {
+      "epoch": 0.4867699061090111,
+      "grad_norm": 0.8849438079565617,
+      "learning_rate": 2.8179773307181307e-06,
+      "loss": 0.1401,
+      "step": 499
+    },
+    {
+      "epoch": 0.4877453969028167,
+      "grad_norm": 1.0134005395077472,
+      "learning_rate": 2.8099693217402807e-06,
+      "loss": 0.1455,
+      "step": 500
+    },
+    {
+      "epoch": 0.4877453969028167,
+      "eval_loss": 0.13670527935028076,
+      "eval_runtime": 121.3231,
+      "eval_samples_per_second": 5.465,
+      "eval_steps_per_second": 0.684,
+      "step": 500
+    },
+    {
+      "epoch": 0.48872088769662236,
+      "grad_norm": 0.9319226416659817,
+      "learning_rate": 2.8019580813499304e-06,
+      "loss": 0.1443,
+      "step": 501
+    },
+    {
+      "epoch": 0.489696378490428,
+      "grad_norm": 0.9042087056844109,
+      "learning_rate": 2.793943693063805e-06,
+      "loss": 0.1382,
+      "step": 502
+    },
+    {
+      "epoch": 0.4906718692842336,
+      "grad_norm": 0.8443370914686048,
+      "learning_rate": 2.7859262404314453e-06,
+      "loss": 0.1226,
+      "step": 503
+    },
+    {
+      "epoch": 0.4916473600780393,
+      "grad_norm": 1.0430324211655913,
+      "learning_rate": 2.7779058070343367e-06,
+      "loss": 0.147,
+      "step": 504
+    },
+    {
+      "epoch": 0.4926228508718449,
+      "grad_norm": 0.9189006680486591,
+      "learning_rate": 2.7698824764850406e-06,
+      "loss": 0.1328,
+      "step": 505
+    },
+    {
+      "epoch": 0.49359834166565053,
+      "grad_norm": 0.8384229297869797,
+      "learning_rate": 2.7618563324263208e-06,
+      "loss": 0.1357,
+      "step": 506
+    },
+    {
+      "epoch": 0.4945738324594562,
+      "grad_norm": 0.9260402086024966,
+      "learning_rate": 2.7538274585302707e-06,
+      "loss": 0.1328,
+      "step": 507
+    },
+    {
+      "epoch": 0.4955493232532618,
+      "grad_norm": 0.8698081640274972,
+      "learning_rate": 2.745795938497443e-06,
+      "loss": 0.1334,
+      "step": 508
+    },
+    {
+      "epoch": 0.49652481404706744,
+      "grad_norm": 0.9155791950153026,
+      "learning_rate": 2.737761856055975e-06,
+      "loss": 0.1355,
+      "step": 509
+    },
+    {
+      "epoch": 0.49750030484087304,
+      "grad_norm": 0.9389141674328999,
+      "learning_rate": 2.7297252949607195e-06,
+      "loss": 0.1338,
+      "step": 510
+    },
+    {
+      "epoch": 0.4984757956346787,
+      "grad_norm": 0.9095624175043169,
+      "learning_rate": 2.721686338992366e-06,
+      "loss": 0.1371,
+      "step": 511
+    },
+    {
+      "epoch": 0.49945128642848435,
+      "grad_norm": 0.8431669151453126,
+      "learning_rate": 2.7136450719565732e-06,
+      "loss": 0.1246,
+      "step": 512
+    },
+    {
+      "epoch": 0.50042677722229,
+      "grad_norm": 0.8523982451840252,
+      "learning_rate": 2.7056015776830907e-06,
+      "loss": 0.1227,
+      "step": 513
+    },
+    {
+      "epoch": 0.5014022680160956,
+      "grad_norm": 0.8800355441963557,
+      "learning_rate": 2.6975559400248876e-06,
+      "loss": 0.1376,
+      "step": 514
+    },
+    {
+      "epoch": 0.5023777588099012,
+      "grad_norm": 0.9058981894611112,
+      "learning_rate": 2.6895082428572777e-06,
+      "loss": 0.1417,
+      "step": 515
+    },
+    {
+      "epoch": 0.5033532496037069,
+      "grad_norm": 0.8780804727101097,
+      "learning_rate": 2.681458570077045e-06,
+      "loss": 0.1346,
+      "step": 516
+    },
+    {
+      "epoch": 0.5043287403975125,
+      "grad_norm": 0.9298727825050748,
+      "learning_rate": 2.673407005601569e-06,
+      "loss": 0.1457,
+      "step": 517
+    },
+    {
+      "epoch": 0.5053042311913182,
+      "grad_norm": 0.8489401556484758,
+      "learning_rate": 2.6653536333679504e-06,
+      "loss": 0.1331,
+      "step": 518
+    },
+    {
+      "epoch": 0.5062797219851237,
+      "grad_norm": 0.8898533627090344,
+      "learning_rate": 2.6572985373321344e-06,
+      "loss": 0.1434,
+      "step": 519
+    },
+    {
+      "epoch": 0.5072552127789294,
+      "grad_norm": 0.9081368712381108,
+      "learning_rate": 2.649241801468039e-06,
+      "loss": 0.1403,
+      "step": 520
+    },
+    {
+      "epoch": 0.508230703572735,
+      "grad_norm": 0.8906355358423613,
+      "learning_rate": 2.641183509766675e-06,
+      "loss": 0.1309,
+      "step": 521
+    },
+    {
+      "epoch": 0.5092061943665407,
+      "grad_norm": 0.9492976151605389,
+      "learning_rate": 2.633123746235274e-06,
+      "loss": 0.1434,
+      "step": 522
+    },
+    {
+      "epoch": 0.5101816851603463,
+      "grad_norm": 0.9210546856991715,
+      "learning_rate": 2.6250625948964125e-06,
+      "loss": 0.1362,
+      "step": 523
+    },
+    {
+      "epoch": 0.5111571759541519,
+      "grad_norm": 0.884606453071947,
+      "learning_rate": 2.617000139787132e-06,
+      "loss": 0.1412,
+      "step": 524
+    },
+    {
+      "epoch": 0.5121326667479575,
+      "grad_norm": 0.9149743385923425,
+      "learning_rate": 2.608936464958068e-06,
+      "loss": 0.1471,
+      "step": 525
+    },
+    {
+      "epoch": 0.5131081575417632,
+      "grad_norm": 0.9748860734433248,
+      "learning_rate": 2.6008716544725692e-06,
+      "loss": 0.1419,
+      "step": 526
+    },
+    {
+      "epoch": 0.5140836483355689,
+      "grad_norm": 0.8273165692055457,
+      "learning_rate": 2.5928057924058264e-06,
+      "loss": 0.1289,
+      "step": 527
+    },
+    {
+      "epoch": 0.5150591391293745,
+      "grad_norm": 0.9005080399868409,
+      "learning_rate": 2.5847389628439905e-06,
+      "loss": 0.1295,
+      "step": 528
+    },
+    {
+      "epoch": 0.51603462992318,
+      "grad_norm": 0.8085104883382979,
+      "learning_rate": 2.576671249883301e-06,
+      "loss": 0.1341,
+      "step": 529
+    },
+    {
+      "epoch": 0.5170101207169857,
+      "grad_norm": 0.9357055967861888,
+      "learning_rate": 2.568602737629204e-06,
+      "loss": 0.1383,
+      "step": 530
+    },
+    {
+      "epoch": 0.5179856115107914,
+      "grad_norm": 0.9860801651928193,
+      "learning_rate": 2.5605335101954796e-06,
+      "loss": 0.1395,
+      "step": 531
+    },
+    {
+      "epoch": 0.518961102304597,
+      "grad_norm": 1.0023841873957609,
+      "learning_rate": 2.552463651703365e-06,
+      "loss": 0.1391,
+      "step": 532
+    },
+    {
+      "epoch": 0.5199365930984027,
+      "grad_norm": 0.9970746647895117,
+      "learning_rate": 2.5443932462806733e-06,
+      "loss": 0.1345,
+      "step": 533
+    },
+    {
+      "epoch": 0.5209120838922082,
+      "grad_norm": 0.8823424259442537,
+      "learning_rate": 2.5363223780609214e-06,
+      "loss": 0.1303,
+      "step": 534
+    },
+    {
+      "epoch": 0.5218875746860139,
+      "grad_norm": 0.8990272295179849,
+      "learning_rate": 2.52825113118245e-06,
+      "loss": 0.138,
+      "step": 535
+    },
+    {
+      "epoch": 0.5228630654798195,
+      "grad_norm": 0.964490190997079,
+      "learning_rate": 2.520179589787547e-06,
+      "loss": 0.1419,
+      "step": 536
+    },
+    {
+      "epoch": 0.5238385562736252,
+      "grad_norm": 0.9850971200909243,
+      "learning_rate": 2.5121078380215713e-06,
+      "loss": 0.1378,
+      "step": 537
+    },
+    {
+      "epoch": 0.5248140470674308,
+      "grad_norm": 0.8473816563308804,
+      "learning_rate": 2.5040359600320747e-06,
+      "loss": 0.1331,
+      "step": 538
+    },
+    {
+      "epoch": 0.5257895378612364,
+      "grad_norm": 0.9063261263061652,
+      "learning_rate": 2.4959640399679253e-06,
+      "loss": 0.143,
+      "step": 539
+    },
+    {
+      "epoch": 0.526765028655042,
+      "grad_norm": 0.8833983728636542,
+      "learning_rate": 2.4878921619784295e-06,
+      "loss": 0.1337,
+      "step": 540
+    },
+    {
+      "epoch": 0.5277405194488477,
+      "grad_norm": 0.8657377624673289,
+      "learning_rate": 2.4798204102124533e-06,
+      "loss": 0.1344,
+      "step": 541
+    },
+    {
+      "epoch": 0.5287160102426534,
+      "grad_norm": 0.9202779408760426,
+      "learning_rate": 2.4717488688175513e-06,
+      "loss": 0.1315,
+      "step": 542
+    },
+    {
+      "epoch": 0.529691501036459,
+      "grad_norm": 0.9088907887564982,
+      "learning_rate": 2.4636776219390794e-06,
+      "loss": 0.1329,
+      "step": 543
+    },
+    {
+      "epoch": 0.5306669918302646,
+      "grad_norm": 0.9310598535059849,
+      "learning_rate": 2.4556067537193276e-06,
+      "loss": 0.1338,
+      "step": 544
+    },
+    {
+      "epoch": 0.5316424826240702,
+      "grad_norm": 0.9003405042179257,
+      "learning_rate": 2.4475363482966356e-06,
+      "loss": 0.1362,
+      "step": 545
+    },
+    {
+      "epoch": 0.5326179734178759,
+      "grad_norm": 0.8206989405119627,
+      "learning_rate": 2.4394664898045208e-06,
+      "loss": 0.133,
+      "step": 546
+    },
+    {
+      "epoch": 0.5335934642116815,
+      "grad_norm": 0.8742977014572649,
+      "learning_rate": 2.4313972623707964e-06,
+      "loss": 0.1343,
+      "step": 547
+    },
+    {
+      "epoch": 0.5345689550054872,
+      "grad_norm": 0.8727361447556689,
+      "learning_rate": 2.4233287501167e-06,
+      "loss": 0.1318,
+      "step": 548
+    },
+    {
+      "epoch": 0.5355444457992927,
+      "grad_norm": 0.9233330686910666,
+      "learning_rate": 2.4152610371560095e-06,
+      "loss": 0.141,
+      "step": 549
+    },
+    {
+      "epoch": 0.5365199365930984,
+      "grad_norm": 0.8918087087553142,
+      "learning_rate": 2.4071942075941744e-06,
+      "loss": 0.1329,
+      "step": 550
+    },
+    {
+      "epoch": 0.537495427386904,
+      "grad_norm": 0.849661983110438,
+      "learning_rate": 2.3991283455274316e-06,
+      "loss": 0.1288,
+      "step": 551
+    },
+    {
+      "epoch": 0.5384709181807097,
+      "grad_norm": 0.8538891952062707,
+      "learning_rate": 2.391063535041933e-06,
+      "loss": 0.1322,
+      "step": 552
+    },
+    {
+      "epoch": 0.5394464089745153,
+      "grad_norm": 0.8801508104924974,
+      "learning_rate": 2.3829998602128685e-06,
+      "loss": 0.1239,
+      "step": 553
+    },
+    {
+      "epoch": 0.5404218997683209,
+      "grad_norm": 0.8524417866635733,
+      "learning_rate": 2.3749374051035883e-06,
+      "loss": 0.1308,
+      "step": 554
+    },
+    {
+      "epoch": 0.5413973905621265,
+      "grad_norm": 0.8827154465529061,
+      "learning_rate": 2.3668762537647254e-06,
+      "loss": 0.1377,
+      "step": 555
+    },
+    {
+      "epoch": 0.5423728813559322,
+      "grad_norm": 0.8050527885767014,
+      "learning_rate": 2.358816490233326e-06,
+      "loss": 0.127,
+      "step": 556
+    },
+    {
+      "epoch": 0.5433483721497379,
+      "grad_norm": 0.8172596997607479,
+      "learning_rate": 2.3507581985319623e-06,
+      "loss": 0.1212,
+      "step": 557
+    },
+    {
+      "epoch": 0.5443238629435435,
+      "grad_norm": 0.8963200330399226,
+      "learning_rate": 2.342701462667866e-06,
+      "loss": 0.1401,
+      "step": 558
+    },
+    {
+      "epoch": 0.5452993537373491,
+      "grad_norm": 0.887052377675435,
+      "learning_rate": 2.3346463666320512e-06,
+      "loss": 0.1323,
+      "step": 559
+    },
+    {
+      "epoch": 0.5462748445311547,
+      "grad_norm": 0.9127806776302525,
+      "learning_rate": 2.3265929943984317e-06,
+      "loss": 0.1361,
+      "step": 560
+    },
+    {
+      "epoch": 0.5472503353249604,
+      "grad_norm": 0.9147589526822345,
+      "learning_rate": 2.318541429922956e-06,
+      "loss": 0.1363,
+      "step": 561
+    },
+    {
+      "epoch": 0.548225826118766,
+      "grad_norm": 0.8944082387887866,
+      "learning_rate": 2.310491757142723e-06,
+      "loss": 0.1414,
+      "step": 562
+    },
+    {
+      "epoch": 0.5492013169125717,
+      "grad_norm": 0.8666818426784915,
+      "learning_rate": 2.3024440599751132e-06,
+      "loss": 0.1338,
+      "step": 563
+    },
+    {
+      "epoch": 0.5501768077063772,
+      "grad_norm": 0.8384112838731508,
+      "learning_rate": 2.29439842231691e-06,
+      "loss": 0.128,
+      "step": 564
+    },
+    {
+      "epoch": 0.5511522985001829,
+      "grad_norm": 0.8598917633955132,
+      "learning_rate": 2.2863549280434285e-06,
+      "loss": 0.1378,
+      "step": 565
+    },
+    {
+      "epoch": 0.5521277892939885,
+      "grad_norm": 0.8490812429791822,
+      "learning_rate": 2.2783136610076345e-06,
+      "loss": 0.1302,
+      "step": 566
+    },
+    {
+      "epoch": 0.5531032800877942,
+      "grad_norm": 0.9671910225128955,
+      "learning_rate": 2.270274705039282e-06,
+      "loss": 0.1371,
+      "step": 567
+    },
+    {
+      "epoch": 0.5540787708815998,
+      "grad_norm": 0.9181284807350794,
+      "learning_rate": 2.2622381439440255e-06,
+      "loss": 0.1435,
+      "step": 568
+    },
+    {
+      "epoch": 0.5550542616754054,
+      "grad_norm": 0.7950987192294108,
+      "learning_rate": 2.2542040615025584e-06,
+      "loss": 0.1219,
+      "step": 569
+    },
+    {
+      "epoch": 0.556029752469211,
+      "grad_norm": 1.0089717587920861,
+      "learning_rate": 2.24617254146973e-06,
+      "loss": 0.1292,
+      "step": 570
+    },
+    {
+      "epoch": 0.5570052432630167,
+      "grad_norm": 0.9858255149161776,
+      "learning_rate": 2.23814366757368e-06,
+      "loss": 0.1393,
+      "step": 571
+    },
+    {
+      "epoch": 0.5579807340568224,
+      "grad_norm": 0.8129514529071972,
+      "learning_rate": 2.23011752351496e-06,
+      "loss": 0.1235,
+      "step": 572
+    },
+    {
+      "epoch": 0.558956224850628,
+      "grad_norm": 0.9188253431247291,
+      "learning_rate": 2.222094192965664e-06,
+      "loss": 0.1275,
+      "step": 573
+    },
+    {
+      "epoch": 0.5599317156444336,
+      "grad_norm": 0.9256786564401078,
+      "learning_rate": 2.214073759568555e-06,
+      "loss": 0.1383,
+      "step": 574
+    },
+    {
+      "epoch": 0.5609072064382392,
+      "grad_norm": 0.810478233219589,
+      "learning_rate": 2.2060563069361955e-06,
+      "loss": 0.1221,
+      "step": 575
+    },
+    {
+      "epoch": 0.5618826972320449,
+      "grad_norm": 0.8139352487642532,
+      "learning_rate": 2.19804191865007e-06,
+      "loss": 0.1203,
+      "step": 576
+    },
+    {
+      "epoch": 0.5628581880258505,
+      "grad_norm": 0.8148607289634421,
+      "learning_rate": 2.19003067825972e-06,
+      "loss": 0.127,
+      "step": 577
+    },
+    {
+      "epoch": 0.5638336788196562,
+      "grad_norm": 0.8730605763594029,
+      "learning_rate": 2.1820226692818693e-06,
+      "loss": 0.1294,
+      "step": 578
+    },
+    {
+      "epoch": 0.5648091696134617,
+      "grad_norm": 0.9081540528860818,
+      "learning_rate": 2.174017975199556e-06,
+      "loss": 0.1418,
+      "step": 579
+    },
+    {
+      "epoch": 0.5657846604072674,
+      "grad_norm": 0.964077636816459,
+      "learning_rate": 2.1660166794612574e-06,
+      "loss": 0.1549,
+      "step": 580
+    },
+    {
+      "epoch": 0.566760151201073,
+      "grad_norm": 0.8682612244819073,
+      "learning_rate": 2.158018865480027e-06,
+      "loss": 0.1339,
+      "step": 581
+    },
+    {
+      "epoch": 0.5677356419948787,
+      "grad_norm": 0.9087722584693885,
+      "learning_rate": 2.1500246166326177e-06,
+      "loss": 0.138,
+      "step": 582
+    },
+    {
+      "epoch": 0.5687111327886843,
+      "grad_norm": 0.8284282518619499,
+      "learning_rate": 2.1420340162586196e-06,
+      "loss": 0.1273,
+      "step": 583
+    },
+    {
+      "epoch": 0.5696866235824899,
+      "grad_norm": 0.899367551533117,
+      "learning_rate": 2.1340471476595836e-06,
+      "loss": 0.1325,
+      "step": 584
+    },
+    {
+      "epoch": 0.5706621143762955,
+      "grad_norm": 0.8721503402061037,
+      "learning_rate": 2.1260640940981616e-06,
+      "loss": 0.1373,
+      "step": 585
+    },
+    {
+      "epoch": 0.5716376051701012,
+      "grad_norm": 0.8007443556277751,
+      "learning_rate": 2.11808493879723e-06,
+      "loss": 0.1314,
+      "step": 586
+    },
+    {
+      "epoch": 0.5726130959639069,
+      "grad_norm": 0.843399287875525,
+      "learning_rate": 2.110109764939029e-06,
+      "loss": 0.1273,
+      "step": 587
+    },
+    {
+      "epoch": 0.5735885867577125,
+      "grad_norm": 0.8694236616188882,
+      "learning_rate": 2.10213865566429e-06,
+      "loss": 0.1292,
+      "step": 588
+    },
+    {
+      "epoch": 0.5745640775515181,
+      "grad_norm": 0.944807048930237,
+      "learning_rate": 2.0941716940713733e-06,
+      "loss": 0.136,
+      "step": 589
+    },
+    {
+      "epoch": 0.5755395683453237,
+      "grad_norm": 0.8012154070123548,
+      "learning_rate": 2.086208963215399e-06,
+      "loss": 0.1219,
+      "step": 590
+    },
+    {
+      "epoch": 0.5765150591391294,
+      "grad_norm": 0.9713169032385847,
+      "learning_rate": 2.0782505461073822e-06,
+      "loss": 0.1272,
+      "step": 591
+    },
+    {
+      "epoch": 0.577490549932935,
+      "grad_norm": 0.833986482831455,
+      "learning_rate": 2.0702965257133664e-06,
+      "loss": 0.1211,
+      "step": 592
+    },
+    {
+      "epoch": 0.5784660407267407,
+      "grad_norm": 0.9223111755392438,
+      "learning_rate": 2.062346984953562e-06,
+      "loss": 0.1229,
+      "step": 593
+    },
+    {
+      "epoch": 0.5794415315205462,
+      "grad_norm": 0.9302281584720075,
+      "learning_rate": 2.0544020067014776e-06,
+      "loss": 0.1383,
+      "step": 594
+    },
+    {
+      "epoch": 0.5804170223143519,
+      "grad_norm": 0.9426149459659149,
+      "learning_rate": 2.0464616737830566e-06,
+      "loss": 0.1337,
+      "step": 595
+    },
+    {
+      "epoch": 0.5813925131081575,
+      "grad_norm": 0.8882577504858508,
+      "learning_rate": 2.03852606897582e-06,
+      "loss": 0.1244,
+      "step": 596
+    },
+    {
+      "epoch": 0.5823680039019632,
+      "grad_norm": 0.8456021234816907,
+      "learning_rate": 2.0305952750079918e-06,
+      "loss": 0.1324,
+      "step": 597
+    },
+    {
+      "epoch": 0.5833434946957688,
+      "grad_norm": 0.933740394118853,
+      "learning_rate": 2.0226693745576494e-06,
+      "loss": 0.136,
+      "step": 598
+    },
+    {
+      "epoch": 0.5843189854895744,
+      "grad_norm": 0.9462976911157949,
+      "learning_rate": 2.0147484502518514e-06,
+      "loss": 0.1286,
+      "step": 599
+    },
+    {
+      "epoch": 0.58529447628338,
+      "grad_norm": 0.9098208372409959,
+      "learning_rate": 2.006832584665783e-06,
+      "loss": 0.132,
+      "step": 600
+    },
+    {
+      "epoch": 0.5862699670771857,
+      "grad_norm": 0.795331526345968,
+      "learning_rate": 1.99892186032189e-06,
+      "loss": 0.1151,
+      "step": 601
+    },
+    {
+      "epoch": 0.5872454578709914,
+      "grad_norm": 0.8664446011660321,
+      "learning_rate": 1.9910163596890247e-06,
+      "loss": 0.1288,
+      "step": 602
+    },
+    {
+      "epoch": 0.588220948664797,
+      "grad_norm": 0.8610758137851285,
+      "learning_rate": 1.983116165181578e-06,
+      "loss": 0.1239,
+      "step": 603
+    },
+    {
+      "epoch": 0.5891964394586026,
+      "grad_norm": 0.8560179576447229,
+      "learning_rate": 1.9752213591586288e-06,
+      "loss": 0.1276,
+      "step": 604
+    },
+    {
+      "epoch": 0.5901719302524082,
+      "grad_norm": 0.8603930368509891,
+      "learning_rate": 1.9673320239230783e-06,
+      "loss": 0.1305,
+      "step": 605
+    },
+    {
+      "epoch": 0.5911474210462139,
+      "grad_norm": 0.8443701642736344,
+      "learning_rate": 1.9594482417207973e-06,
+      "loss": 0.1261,
+      "step": 606
+    },
+    {
+      "epoch": 0.5921229118400195,
+      "grad_norm": 0.7909939129569269,
+      "learning_rate": 1.951570094739765e-06,
+      "loss": 0.1211,
+      "step": 607
+    },
+    {
+      "epoch": 0.5930984026338252,
+      "grad_norm": 0.931299792718923,
+      "learning_rate": 1.9436976651092143e-06,
+      "loss": 0.1382,
+      "step": 608
+    },
+    {
+      "epoch": 0.5940738934276307,
+      "grad_norm": 0.8781253875372456,
+      "learning_rate": 1.9358310348987755e-06,
+      "loss": 0.1273,
+      "step": 609
+    },
+    {
+      "epoch": 0.5950493842214364,
+      "grad_norm": 0.8570421027558416,
+      "learning_rate": 1.9279702861176203e-06,
+      "loss": 0.1342,
+      "step": 610
+    },
+    {
+      "epoch": 0.596024875015242,
+      "grad_norm": 0.8451871775967923,
+      "learning_rate": 1.9201155007136045e-06,
+      "loss": 0.1365,
+      "step": 611
+    },
+    {
+      "epoch": 0.5970003658090477,
+      "grad_norm": 0.9384520823084104,
+      "learning_rate": 1.9122667605724202e-06,
+      "loss": 0.1348,
+      "step": 612
+    },
+    {
+      "epoch": 0.5979758566028534,
+      "grad_norm": 0.8922350956011054,
+      "learning_rate": 1.9044241475167339e-06,
+      "loss": 0.1348,
+      "step": 613
+    },
+    {
+      "epoch": 0.5989513473966589,
+      "grad_norm": 0.8385907676608247,
+      "learning_rate": 1.8965877433053409e-06,
+      "loss": 0.1287,
+      "step": 614
+    },
+    {
+      "epoch": 0.5999268381904646,
+      "grad_norm": 0.8438875643530197,
+      "learning_rate": 1.8887576296323069e-06,
+      "loss": 0.1234,
+      "step": 615
+    },
+    {
+      "epoch": 0.6009023289842702,
+      "grad_norm": 0.8782492476271099,
+      "learning_rate": 1.8809338881261213e-06,
+      "loss": 0.1312,
+      "step": 616
+    },
+    {
+      "epoch": 0.6018778197780759,
+      "grad_norm": 0.8210821361205313,
+      "learning_rate": 1.8731166003488427e-06,
+      "loss": 0.1256,
+      "step": 617
+    },
+    {
+      "epoch": 0.6028533105718815,
+      "grad_norm": 0.8313931745166832,
+      "learning_rate": 1.86530584779525e-06,
+      "loss": 0.1269,
+      "step": 618
+    },
+    {
+      "epoch": 0.6038288013656871,
+      "grad_norm": 0.8248814372114766,
+      "learning_rate": 1.857501711891993e-06,
+      "loss": 0.1277,
+      "step": 619
+    },
+    {
+      "epoch": 0.6048042921594927,
+      "grad_norm": 0.8300683641876326,
+      "learning_rate": 1.8497042739967447e-06,
+      "loss": 0.1321,
+      "step": 620
+    },
+    {
+      "epoch": 0.6057797829532984,
+      "grad_norm": 0.8967700470622615,
+      "learning_rate": 1.841913615397348e-06,
+      "loss": 0.1273,
+      "step": 621
+    },
+    {
+      "epoch": 0.606755273747104,
+      "grad_norm": 0.937831066737166,
+      "learning_rate": 1.834129817310977e-06,
+      "loss": 0.1293,
+      "step": 622
+    },
+    {
+      "epoch": 0.6077307645409097,
+      "grad_norm": 0.8608866113112006,
+      "learning_rate": 1.826352960883281e-06,
+      "loss": 0.1276,
+      "step": 623
+    },
+    {
+      "epoch": 0.6087062553347152,
+      "grad_norm": 0.8562897857077051,
+      "learning_rate": 1.8185831271875457e-06,
+      "loss": 0.1204,
+      "step": 624
+    },
+    {
+      "epoch": 0.6096817461285209,
+      "grad_norm": 0.8326388599515545,
+      "learning_rate": 1.8108203972238436e-06,
+      "loss": 0.124,
+      "step": 625
+    },
+    {
+      "epoch": 0.6106572369223265,
+      "grad_norm": 0.9146045731524001,
+      "learning_rate": 1.8030648519181926e-06,
+      "loss": 0.1327,
+      "step": 626
+    },
+    {
+      "epoch": 0.6116327277161322,
+      "grad_norm": 0.8125344568360147,
+      "learning_rate": 1.7953165721217086e-06,
+      "loss": 0.1206,
+      "step": 627
+    },
+    {
+      "epoch": 0.6126082185099379,
+      "grad_norm": 0.8788372999011178,
+      "learning_rate": 1.7875756386097686e-06,
+      "loss": 0.1376,
+      "step": 628
+    },
+    {
+      "epoch": 0.6135837093037434,
+      "grad_norm": 0.8647255719378499,
+      "learning_rate": 1.779842132081162e-06,
+      "loss": 0.1243,
+      "step": 629
+    },
+    {
+      "epoch": 0.614559200097549,
+      "grad_norm": 0.8756476987196234,
+      "learning_rate": 1.7721161331572522e-06,
+      "loss": 0.132,
+      "step": 630
+    },
+    {
+      "epoch": 0.6155346908913547,
+      "grad_norm": 0.9563185708098122,
+      "learning_rate": 1.7643977223811392e-06,
+      "loss": 0.1364,
+      "step": 631
+    },
+    {
+      "epoch": 0.6165101816851604,
+      "grad_norm": 0.8320678225276503,
+      "learning_rate": 1.7566869802168132e-06,
+      "loss": 0.1189,
+      "step": 632
+    },
+    {
+      "epoch": 0.617485672478966,
+      "grad_norm": 0.9093829860333068,
+      "learning_rate": 1.7489839870483236e-06,
+      "loss": 0.1277,
+      "step": 633
+    },
+    {
+      "epoch": 0.6184611632727716,
+      "grad_norm": 0.9011561073082258,
+      "learning_rate": 1.7412888231789327e-06,
+      "loss": 0.1358,
+      "step": 634
+    },
+    {
+      "epoch": 0.6194366540665772,
+      "grad_norm": 1.0084984996858963,
+      "learning_rate": 1.7336015688302869e-06,
+      "loss": 0.1268,
+      "step": 635
+    },
+    {
+      "epoch": 0.6204121448603829,
+      "grad_norm": 0.9152283340289553,
+      "learning_rate": 1.725922304141573e-06,
+      "loss": 0.1295,
+      "step": 636
+    },
+    {
+      "epoch": 0.6213876356541885,
+      "grad_norm": 0.9238351863571306,
+      "learning_rate": 1.718251109168688e-06,
+      "loss": 0.1318,
+      "step": 637
+    },
+    {
+      "epoch": 0.6223631264479942,
+      "grad_norm": 0.866304353620812,
+      "learning_rate": 1.7105880638834007e-06,
+      "loss": 0.1212,
+      "step": 638
+    },
+    {
+      "epoch": 0.6233386172417997,
+      "grad_norm": 0.95501299054228,
+      "learning_rate": 1.7029332481725234e-06,
+      "loss": 0.1364,
+      "step": 639
+    },
+    {
+      "epoch": 0.6243141080356054,
+      "grad_norm": 0.8839661191980624,
+      "learning_rate": 1.6952867418370707e-06,
+      "loss": 0.127,
+      "step": 640
+    },
+    {
+      "epoch": 0.625289598829411,
+      "grad_norm": 0.8559388882214561,
+      "learning_rate": 1.6876486245914375e-06,
+      "loss": 0.1249,
+      "step": 641
+    },
+    {
+      "epoch": 0.6262650896232167,
+      "grad_norm": 0.9364435799205211,
+      "learning_rate": 1.6800189760625585e-06,
+      "loss": 0.1329,
+      "step": 642
+    },
+    {
+      "epoch": 0.6272405804170224,
+      "grad_norm": 0.8651014555168739,
+      "learning_rate": 1.672397875789086e-06,
+      "loss": 0.125,
+      "step": 643
+    },
+    {
+      "epoch": 0.6282160712108279,
+      "grad_norm": 0.8892108367186486,
+      "learning_rate": 1.6647854032205547e-06,
+      "loss": 0.1304,
+      "step": 644
+    },
+    {
+      "epoch": 0.6291915620046336,
+      "grad_norm": 0.8503679132174746,
+      "learning_rate": 1.6571816377165568e-06,
+      "loss": 0.1249,
+      "step": 645
+    },
+    {
+      "epoch": 0.6301670527984392,
+      "grad_norm": 0.9314057258730163,
+      "learning_rate": 1.6495866585459142e-06,
+      "loss": 0.131,
+      "step": 646
+    },
+    {
+      "epoch": 0.6311425435922449,
+      "grad_norm": 0.9002310278187038,
+      "learning_rate": 1.6420005448858522e-06,
+      "loss": 0.1329,
+      "step": 647
+    },
+    {
+      "epoch": 0.6321180343860505,
+      "grad_norm": 0.8263610327236112,
+      "learning_rate": 1.6344233758211717e-06,
+      "loss": 0.1217,
+      "step": 648
+    },
+    {
+      "epoch": 0.6330935251798561,
+      "grad_norm": 0.8099996895450952,
+      "learning_rate": 1.6268552303434298e-06,
+      "loss": 0.1211,
+      "step": 649
+    },
+    {
+      "epoch": 0.6340690159736617,
+      "grad_norm": 0.8528126491211017,
+      "learning_rate": 1.6192961873501096e-06,
+      "loss": 0.1297,
+      "step": 650
+    },
+    {
+      "epoch": 0.6350445067674674,
+      "grad_norm": 0.8585959373381494,
+      "learning_rate": 1.611746325643805e-06,
+      "loss": 0.1295,
+      "step": 651
+    },
+    {
+      "epoch": 0.636019997561273,
+      "grad_norm": 0.8655236759185483,
+      "learning_rate": 1.6042057239313919e-06,
+      "loss": 0.1246,
+      "step": 652
+    },
+    {
+      "epoch": 0.6369954883550787,
+      "grad_norm": 0.8207113182539624,
+      "learning_rate": 1.5966744608232137e-06,
+      "loss": 0.12,
+      "step": 653
+    },
+    {
+      "epoch": 0.6379709791488842,
+      "grad_norm": 0.8614663092183132,
+      "learning_rate": 1.5891526148322594e-06,
+      "loss": 0.1187,
+      "step": 654
+    },
+    {
+      "epoch": 0.6389464699426899,
+      "grad_norm": 0.9001233935199348,
+      "learning_rate": 1.5816402643733441e-06,
+      "loss": 0.1276,
+      "step": 655
+    },
+    {
+      "epoch": 0.6399219607364955,
+      "grad_norm": 0.8508200436128184,
+      "learning_rate": 1.5741374877622922e-06,
+      "loss": 0.1267,
+      "step": 656
+    },
+    {
+      "epoch": 0.6408974515303012,
+      "grad_norm": 0.9227739705381223,
+      "learning_rate": 1.566644363215123e-06,
+      "loss": 0.1344,
+      "step": 657
+    },
+    {
+      "epoch": 0.6418729423241069,
+      "grad_norm": 0.853646215929862,
+      "learning_rate": 1.5591609688472313e-06,
+      "loss": 0.1363,
+      "step": 658
+    },
+    {
+      "epoch": 0.6428484331179124,
+      "grad_norm": 0.9878682768114257,
+      "learning_rate": 1.551687382672578e-06,
+      "loss": 0.1354,
+      "step": 659
+    },
+    {
+      "epoch": 0.643823923911718,
+      "grad_norm": 0.8622608892200061,
+      "learning_rate": 1.5442236826028705e-06,
+      "loss": 0.1281,
+      "step": 660
+    },
+    {
+      "epoch": 0.6447994147055237,
+      "grad_norm": 0.8868098887146023,
+      "learning_rate": 1.5367699464467596e-06,
+      "loss": 0.1316,
+      "step": 661
+    },
+    {
+      "epoch": 0.6457749054993294,
+      "grad_norm": 0.8568792305689042,
+      "learning_rate": 1.5293262519090169e-06,
+      "loss": 0.1289,
+      "step": 662
+    },
+    {
+      "epoch": 0.646750396293135,
+      "grad_norm": 0.8345103045814822,
+      "learning_rate": 1.5218926765897345e-06,
+      "loss": 0.1238,
+      "step": 663
+    },
+    {
+      "epoch": 0.6477258870869406,
+      "grad_norm": 0.8296168710839235,
+      "learning_rate": 1.5144692979835103e-06,
+      "loss": 0.1279,
+      "step": 664
+    },
+    {
+      "epoch": 0.6487013778807462,
+      "grad_norm": 0.8651995871451914,
+      "learning_rate": 1.5070561934786437e-06,
+      "loss": 0.131,
+      "step": 665
+    },
+    {
+      "epoch": 0.6496768686745519,
+      "grad_norm": 0.9169910575714909,
+      "learning_rate": 1.4996534403563267e-06,
+      "loss": 0.1334,
+      "step": 666
+    },
+    {
+      "epoch": 0.6506523594683575,
+      "grad_norm": 0.7693306954167114,
+      "learning_rate": 1.4922611157898364e-06,
+      "loss": 0.1168,
+      "step": 667
+    },
+    {
+      "epoch": 0.6516278502621632,
+      "grad_norm": 0.863479055519848,
+      "learning_rate": 1.4848792968437376e-06,
+      "loss": 0.1248,
+      "step": 668
+    },
+    {
+      "epoch": 0.6526033410559687,
+      "grad_norm": 0.8668901068891359,
+      "learning_rate": 1.477508060473069e-06,
+      "loss": 0.1321,
+      "step": 669
+    },
+    {
+      "epoch": 0.6535788318497744,
+      "grad_norm": 0.8378480519694856,
+      "learning_rate": 1.4701474835225515e-06,
+      "loss": 0.1312,
+      "step": 670
+    },
+    {
+      "epoch": 0.65455432264358,
+      "grad_norm": 0.811001196208897,
+      "learning_rate": 1.4627976427257784e-06,
+      "loss": 0.1201,
+      "step": 671
+    },
+    {
+      "epoch": 0.6555298134373857,
+      "grad_norm": 0.8636188619772824,
+      "learning_rate": 1.4554586147044225e-06,
+      "loss": 0.1365,
+      "step": 672
+    },
+    {
+      "epoch": 0.6565053042311914,
+      "grad_norm": 0.85744996390658,
+      "learning_rate": 1.4481304759674303e-06,
+      "loss": 0.1249,
+      "step": 673
+    },
+    {
+      "epoch": 0.6574807950249969,
+      "grad_norm": 0.8435636506247961,
+      "learning_rate": 1.4408133029102322e-06,
+      "loss": 0.1195,
+      "step": 674
+    },
+    {
+      "epoch": 0.6584562858188026,
+      "grad_norm": 0.8509255085231769,
+      "learning_rate": 1.4335071718139379e-06,
+      "loss": 0.1183,
+      "step": 675
+    },
+    {
+      "epoch": 0.6594317766126082,
+      "grad_norm": 0.8793960273956329,
+      "learning_rate": 1.4262121588445493e-06,
+      "loss": 0.1245,
+      "step": 676
+    },
+    {
+      "epoch": 0.6604072674064139,
+      "grad_norm": 0.8970156523355322,
+      "learning_rate": 1.418928340052159e-06,
+      "loss": 0.1286,
+      "step": 677
+    },
+    {
+      "epoch": 0.6613827582002195,
+      "grad_norm": 0.8644162321708901,
+      "learning_rate": 1.411655791370164e-06,
+      "loss": 0.1252,
+      "step": 678
+    },
+    {
+      "epoch": 0.6623582489940251,
+      "grad_norm": 0.8396437394115862,
+      "learning_rate": 1.4043945886144673e-06,
+      "loss": 0.1179,
+      "step": 679
+    },
+    {
+      "epoch": 0.6633337397878307,
+      "grad_norm": 0.8448586465328236,
+      "learning_rate": 1.397144807482696e-06,
+      "loss": 0.1252,
+      "step": 680
+    },
+    {
+      "epoch": 0.6643092305816364,
+      "grad_norm": 0.8718372351836149,
+      "learning_rate": 1.3899065235534031e-06,
+      "loss": 0.1216,
+      "step": 681
+    },
+    {
+      "epoch": 0.665284721375442,
+      "grad_norm": 0.8407139122291305,
+      "learning_rate": 1.382679812285287e-06,
+      "loss": 0.1283,
+      "step": 682
+    },
+    {
+      "epoch": 0.6662602121692477,
+      "grad_norm": 0.8404593100461625,
+      "learning_rate": 1.375464749016399e-06,
+      "loss": 0.1188,
+      "step": 683
+    },
+    {
+      "epoch": 0.6672357029630532,
+      "grad_norm": 0.940858079210062,
+      "learning_rate": 1.3682614089633637e-06,
+      "loss": 0.129,
+      "step": 684
+    },
+    {
+      "epoch": 0.6682111937568589,
+      "grad_norm": 0.8569084229583452,
+      "learning_rate": 1.3610698672205897e-06,
+      "loss": 0.1281,
+      "step": 685
+    },
+    {
+      "epoch": 0.6691866845506645,
+      "grad_norm": 0.8797960805331662,
+      "learning_rate": 1.3538901987594882e-06,
+      "loss": 0.1292,
+      "step": 686
+    },
+    {
+      "epoch": 0.6701621753444702,
+      "grad_norm": 0.8374942493553367,
+      "learning_rate": 1.3467224784276961e-06,
+      "loss": 0.1309,
+      "step": 687
+    },
+    {
+      "epoch": 0.6711376661382759,
+      "grad_norm": 0.7710018857731183,
+      "learning_rate": 1.339566780948287e-06,
+      "loss": 0.1139,
+      "step": 688
+    },
+    {
+      "epoch": 0.6721131569320814,
+      "grad_norm": 0.8657391409702617,
+      "learning_rate": 1.3324231809189985e-06,
+      "loss": 0.1309,
+      "step": 689
+    },
+    {
+      "epoch": 0.6730886477258871,
+      "grad_norm": 0.8127948325482396,
+      "learning_rate": 1.3252917528114556e-06,
+      "loss": 0.1267,
+      "step": 690
+    },
+    {
+      "epoch": 0.6740641385196927,
+      "grad_norm": 0.8122952852781401,
+      "learning_rate": 1.3181725709703874e-06,
+      "loss": 0.1213,
+      "step": 691
+    },
+    {
+      "epoch": 0.6750396293134984,
+      "grad_norm": 0.851528828559218,
+      "learning_rate": 1.3110657096128606e-06,
+      "loss": 0.12,
+      "step": 692
+    },
+    {
+      "epoch": 0.676015120107304,
+      "grad_norm": 0.86862592870714,
+      "learning_rate": 1.3039712428274976e-06,
+      "loss": 0.1231,
+      "step": 693
+    },
+    {
+      "epoch": 0.6769906109011096,
+      "grad_norm": 0.8696639730293406,
+      "learning_rate": 1.2968892445737117e-06,
+      "loss": 0.1294,
+      "step": 694
+    },
+    {
+      "epoch": 0.6779661016949152,
+      "grad_norm": 0.9002780776036385,
+      "learning_rate": 1.2898197886809289e-06,
+      "loss": 0.1256,
+      "step": 695
+    },
+    {
+      "epoch": 0.6789415924887209,
+      "grad_norm": 0.8530790957831813,
+      "learning_rate": 1.2827629488478254e-06,
+      "loss": 0.13,
+      "step": 696
+    },
+    {
+      "epoch": 0.6799170832825265,
+      "grad_norm": 0.8450717054989292,
+      "learning_rate": 1.2757187986415528e-06,
+      "loss": 0.1306,
+      "step": 697
+    },
+    {
+      "epoch": 0.6808925740763322,
+      "grad_norm": 0.8714024729066125,
+      "learning_rate": 1.2686874114969767e-06,
+      "loss": 0.1323,
+      "step": 698
+    },
+    {
+      "epoch": 0.6818680648701377,
+      "grad_norm": 0.8591419822215062,
+      "learning_rate": 1.2616688607159045e-06,
+      "loss": 0.1396,
+      "step": 699
+    },
+    {
+      "epoch": 0.6828435556639434,
+      "grad_norm": 0.8361114781431095,
+      "learning_rate": 1.2546632194663305e-06,
+      "loss": 0.1254,
+      "step": 700
+    },
+    {
+      "epoch": 0.683819046457749,
+      "grad_norm": 0.8242742327238427,
+      "learning_rate": 1.2476705607816639e-06,
+      "loss": 0.1234,
+      "step": 701
+    },
+    {
+      "epoch": 0.6847945372515547,
+      "grad_norm": 0.8637116393284514,
+      "learning_rate": 1.2406909575599717e-06,
+      "loss": 0.125,
+      "step": 702
+    },
+    {
+      "epoch": 0.6857700280453604,
+      "grad_norm": 0.8884418622966431,
+      "learning_rate": 1.2337244825632217e-06,
+      "loss": 0.135,
+      "step": 703
+    },
+    {
+      "epoch": 0.6867455188391659,
+      "grad_norm": 0.8275460859037888,
+      "learning_rate": 1.2267712084165168e-06,
+      "loss": 0.1158,
+      "step": 704
+    },
+    {
+      "epoch": 0.6877210096329716,
+      "grad_norm": 0.8553642751195779,
+      "learning_rate": 1.2198312076073427e-06,
+      "loss": 0.128,
+      "step": 705
+    },
+    {
+      "epoch": 0.6886965004267772,
+      "grad_norm": 0.8837324796776882,
+      "learning_rate": 1.2129045524848138e-06,
+      "loss": 0.128,
+      "step": 706
+    },
+    {
+      "epoch": 0.6896719912205829,
+      "grad_norm": 0.8270144975513911,
+      "learning_rate": 1.205991315258914e-06,
+      "loss": 0.1185,
+      "step": 707
+    },
+    {
+      "epoch": 0.6906474820143885,
+      "grad_norm": 0.8738272053373212,
+      "learning_rate": 1.199091567999746e-06,
+      "loss": 0.1264,
+      "step": 708
+    },
+    {
+      "epoch": 0.6916229728081941,
+      "grad_norm": 0.8344314699090399,
+      "learning_rate": 1.1922053826367833e-06,
+      "loss": 0.1152,
+      "step": 709
+    },
+    {
+      "epoch": 0.6925984636019997,
+      "grad_norm": 0.8650165946821683,
+      "learning_rate": 1.1853328309581139e-06,
+      "loss": 0.1283,
+      "step": 710
+    },
+    {
+      "epoch": 0.6935739543958054,
+      "grad_norm": 0.8374700294304306,
+      "learning_rate": 1.1784739846096982e-06,
+      "loss": 0.1238,
+      "step": 711
+    },
+    {
+      "epoch": 0.694549445189611,
+      "grad_norm": 0.8380502969131793,
+      "learning_rate": 1.1716289150946173e-06,
+      "loss": 0.1234,
+      "step": 712
+    },
+    {
+      "epoch": 0.6955249359834167,
+      "grad_norm": 0.9064794381405805,
+      "learning_rate": 1.1647976937723315e-06,
+      "loss": 0.1306,
+      "step": 713
+    },
+    {
+      "epoch": 0.6965004267772222,
+      "grad_norm": 0.832398941447385,
+      "learning_rate": 1.1579803918579322e-06,
+      "loss": 0.125,
+      "step": 714
+    },
+    {
+      "epoch": 0.6974759175710279,
+      "grad_norm": 0.8174727802602623,
+      "learning_rate": 1.1511770804214046e-06,
+      "loss": 0.1248,
+      "step": 715
+    },
+    {
+      "epoch": 0.6984514083648335,
+      "grad_norm": 0.8474018969337508,
+      "learning_rate": 1.1443878303868805e-06,
+      "loss": 0.1338,
+      "step": 716
+    },
+    {
+      "epoch": 0.6994268991586392,
+      "grad_norm": 0.8466582401794491,
+      "learning_rate": 1.1376127125319065e-06,
+      "loss": 0.1148,
+      "step": 717
+    },
+    {
+      "epoch": 0.7004023899524449,
+      "grad_norm": 0.8912296637806607,
+      "learning_rate": 1.1308517974866973e-06,
+      "loss": 0.1274,
+      "step": 718
+    },
+    {
+      "epoch": 0.7013778807462504,
+      "grad_norm": 0.9491089244563283,
+      "learning_rate": 1.1241051557334087e-06,
+      "loss": 0.124,
+      "step": 719
+    },
+    {
+      "epoch": 0.7023533715400561,
+      "grad_norm": 0.9380181290695212,
+      "learning_rate": 1.117372857605395e-06,
+      "loss": 0.1354,
+      "step": 720
+    },
+    {
+      "epoch": 0.7033288623338617,
+      "grad_norm": 0.8177994093605291,
+      "learning_rate": 1.110654973286481e-06,
+      "loss": 0.1164,
+      "step": 721
+    },
+    {
+      "epoch": 0.7043043531276674,
+      "grad_norm": 0.8347927318132022,
+      "learning_rate": 1.1039515728102273e-06,
+      "loss": 0.1275,
+      "step": 722
+    },
+    {
+      "epoch": 0.705279843921473,
+      "grad_norm": 0.8970474242898054,
+      "learning_rate": 1.0972627260592014e-06,
+      "loss": 0.1231,
+      "step": 723
+    },
+    {
+      "epoch": 0.7062553347152786,
+      "grad_norm": 0.795847982137367,
+      "learning_rate": 1.0905885027642484e-06,
+      "loss": 0.1177,
+      "step": 724
+    },
+    {
+      "epoch": 0.7072308255090842,
+      "grad_norm": 0.8814347768470174,
+      "learning_rate": 1.0839289725037669e-06,
+      "loss": 0.1334,
+      "step": 725
+    },
+    {
+      "epoch": 0.7082063163028899,
+      "grad_norm": 0.828253666509173,
+      "learning_rate": 1.0772842047029786e-06,
+      "loss": 0.1218,
+      "step": 726
+    },
+    {
+      "epoch": 0.7091818070966955,
+      "grad_norm": 0.812866815136845,
+      "learning_rate": 1.0706542686332102e-06,
+      "loss": 0.1143,
+      "step": 727
+    },
+    {
+      "epoch": 0.7101572978905012,
+      "grad_norm": 0.8556204829905426,
+      "learning_rate": 1.064039233411166e-06,
+      "loss": 0.1264,
+      "step": 728
+    },
+    {
+      "epoch": 0.7111327886843067,
+      "grad_norm": 0.8841697516339617,
+      "learning_rate": 1.057439167998212e-06,
+      "loss": 0.1278,
+      "step": 729
+    },
+    {
+      "epoch": 0.7121082794781124,
+      "grad_norm": 0.8636362569900907,
+      "learning_rate": 1.050854141199652e-06,
+      "loss": 0.1235,
+      "step": 730
+    },
+    {
+      "epoch": 0.713083770271918,
+      "grad_norm": 0.8283746245515586,
+      "learning_rate": 1.0442842216640168e-06,
+      "loss": 0.1208,
+      "step": 731
+    },
+    {
+      "epoch": 0.7140592610657237,
+      "grad_norm": 0.8728491657133771,
+      "learning_rate": 1.037729477882341e-06,
+      "loss": 0.117,
+      "step": 732
+    },
+    {
+      "epoch": 0.7150347518595294,
+      "grad_norm": 0.8915555008915115,
+      "learning_rate": 1.0311899781874563e-06,
+      "loss": 0.1293,
+      "step": 733
+    },
+    {
+      "epoch": 0.7160102426533349,
+      "grad_norm": 0.8380362900075755,
+      "learning_rate": 1.0246657907532727e-06,
+      "loss": 0.1216,
+      "step": 734
+    },
+    {
+      "epoch": 0.7169857334471406,
+      "grad_norm": 0.8565497720933896,
+      "learning_rate": 1.018156983594073e-06,
+      "loss": 0.1149,
+      "step": 735
+    },
+    {
+      "epoch": 0.7179612242409462,
+      "grad_norm": 0.8358087384760822,
+      "learning_rate": 1.0116636245637997e-06,
+      "loss": 0.115,
+      "step": 736
+    },
+    {
+      "epoch": 0.7189367150347519,
+      "grad_norm": 0.8847500849421209,
+      "learning_rate": 1.0051857813553516e-06,
+      "loss": 0.123,
+      "step": 737
+    },
+    {
+      "epoch": 0.7199122058285575,
+      "grad_norm": 0.8794173909505707,
+      "learning_rate": 9.987235214998741e-07,
+      "loss": 0.1277,
+      "step": 738
+    },
+    {
+      "epoch": 0.7208876966223631,
+      "grad_norm": 0.8665043928636729,
+      "learning_rate": 9.922769123660564e-07,
+      "loss": 0.1197,
+      "step": 739
+    },
+    {
+      "epoch": 0.7218631874161687,
+      "grad_norm": 0.8361820544782747,
+      "learning_rate": 9.858460211594331e-07,
+      "loss": 0.124,
+      "step": 740
+    },
+    {
+      "epoch": 0.7228386782099744,
+      "grad_norm": 0.846259877698448,
+      "learning_rate": 9.79430914921677e-07,
+      "loss": 0.1207,
+      "step": 741
+    },
+    {
+      "epoch": 0.72381416900378,
+      "grad_norm": 0.8468390031221599,
+      "learning_rate": 9.730316605299041e-07,
+      "loss": 0.1226,
+      "step": 742
+    },
+    {
+      "epoch": 0.7247896597975857,
+      "grad_norm": 0.9014801026605848,
+      "learning_rate": 9.666483246959785e-07,
+      "loss": 0.119,
+      "step": 743
+    },
+    {
+      "epoch": 0.7257651505913912,
+      "grad_norm": 0.9115482145040746,
+      "learning_rate": 9.602809739658105e-07,
+      "loss": 0.1245,
+      "step": 744
+    },
+    {
+      "epoch": 0.7267406413851969,
+      "grad_norm": 0.8643571047269124,
+      "learning_rate": 9.53929674718668e-07,
+      "loss": 0.1216,
+      "step": 745
+    },
+    {
+      "epoch": 0.7277161321790026,
+      "grad_norm": 0.8382233183752549,
+      "learning_rate": 9.475944931664851e-07,
+      "loss": 0.125,
+      "step": 746
+    },
+    {
+      "epoch": 0.7286916229728082,
+      "grad_norm": 0.8536460797398011,
+      "learning_rate": 9.412754953531664e-07,
+      "loss": 0.1223,
+      "step": 747
+    },
+    {
+      "epoch": 0.7296671137666139,
+      "grad_norm": 0.8756760541580888,
+      "learning_rate": 9.349727471539052e-07,
+      "loss": 0.1215,
+      "step": 748
+    },
+    {
+      "epoch": 0.7306426045604194,
+      "grad_norm": 0.884349162839785,
+      "learning_rate": 9.286863142744907e-07,
+      "loss": 0.1237,
+      "step": 749
+    },
+    {
+      "epoch": 0.7316180953542251,
+      "grad_norm": 0.8232259080657391,
+      "learning_rate": 9.224162622506283e-07,
+      "loss": 0.118,
+      "step": 750
+    },
+    {
+      "epoch": 0.7325935861480307,
+      "grad_norm": 0.8865909364907526,
+      "learning_rate": 9.161626564472511e-07,
+      "loss": 0.1271,
+      "step": 751
+    },
+    {
+      "epoch": 0.7335690769418364,
+      "grad_norm": 0.8299409590937745,
+      "learning_rate": 9.099255620578451e-07,
+      "loss": 0.1212,
+      "step": 752
+    },
+    {
+      "epoch": 0.734544567735642,
+      "grad_norm": 0.841743660255562,
+      "learning_rate": 9.037050441037614e-07,
+      "loss": 0.1209,
+      "step": 753
+    },
+    {
+      "epoch": 0.7355200585294476,
+      "grad_norm": 0.9030416307912014,
+      "learning_rate": 8.975011674335468e-07,
+      "loss": 0.1271,
+      "step": 754
+    },
+    {
+      "epoch": 0.7364955493232532,
+      "grad_norm": 0.8090547787848019,
+      "learning_rate": 8.9131399672226e-07,
+      "loss": 0.1174,
+      "step": 755
+    },
+    {
+      "epoch": 0.7374710401170589,
+      "grad_norm": 0.8525347836022709,
+      "learning_rate": 8.851435964708044e-07,
+      "loss": 0.1138,
+      "step": 756
+    },
+    {
+      "epoch": 0.7384465309108645,
+      "grad_norm": 0.8419910445268589,
+      "learning_rate": 8.789900310052491e-07,
+      "loss": 0.1215,
+      "step": 757
+    },
+    {
+      "epoch": 0.7394220217046702,
+      "grad_norm": 0.8490121207415494,
+      "learning_rate": 8.728533644761647e-07,
+      "loss": 0.1235,
+      "step": 758
+    },
+    {
+      "epoch": 0.7403975124984757,
+      "grad_norm": 0.8982804136005714,
+      "learning_rate": 8.667336608579488e-07,
+      "loss": 0.1226,
+      "step": 759
+    },
+    {
+      "epoch": 0.7413730032922814,
+      "grad_norm": 0.8467362174197229,
+      "learning_rate": 8.606309839481628e-07,
+      "loss": 0.123,
+      "step": 760
+    },
+    {
+      "epoch": 0.742348494086087,
+      "grad_norm": 0.8380377962583168,
+      "learning_rate": 8.545453973668643e-07,
+      "loss": 0.1182,
+      "step": 761
+    },
+    {
+      "epoch": 0.7433239848798927,
+      "grad_norm": 0.8630957150176094,
+      "learning_rate": 8.484769645559482e-07,
+      "loss": 0.1195,
+      "step": 762
+    },
+    {
+      "epoch": 0.7442994756736984,
+      "grad_norm": 0.8473968774046894,
+      "learning_rate": 8.424257487784787e-07,
+      "loss": 0.1221,
+      "step": 763
+    },
+    {
+      "epoch": 0.7452749664675039,
+      "grad_norm": 0.9113778016570447,
+      "learning_rate": 8.363918131180371e-07,
+      "loss": 0.1256,
+      "step": 764
+    },
+    {
+      "epoch": 0.7462504572613096,
+      "grad_norm": 0.9276069699799473,
+      "learning_rate": 8.303752204780563e-07,
+      "loss": 0.1248,
+      "step": 765
+    },
+    {
+      "epoch": 0.7472259480551152,
+      "grad_norm": 0.8167762715238328,
+      "learning_rate": 8.243760335811734e-07,
+      "loss": 0.1246,
+      "step": 766
+    },
+    {
+      "epoch": 0.7482014388489209,
+      "grad_norm": 0.8357245883738734,
+      "learning_rate": 8.183943149685678e-07,
+      "loss": 0.1173,
+      "step": 767
+    },
+    {
+      "epoch": 0.7491769296427265,
+      "grad_norm": 0.8026503286373335,
+      "learning_rate": 8.124301269993168e-07,
+      "loss": 0.1151,
+      "step": 768
+    },
+    {
+      "epoch": 0.7501524204365322,
+      "grad_norm": 0.9315071378112674,
+      "learning_rate": 8.06483531849738e-07,
+      "loss": 0.1235,
+      "step": 769
+    },
+    {
+      "epoch": 0.7511279112303377,
+      "grad_norm": 0.9034958615476297,
+      "learning_rate": 8.00554591512748e-07,
+      "loss": 0.1262,
+      "step": 770
+    },
+    {
+      "epoch": 0.7521034020241434,
+      "grad_norm": 0.8669626589106916,
+      "learning_rate": 7.946433677972104e-07,
+      "loss": 0.1243,
+      "step": 771
+    },
+    {
+      "epoch": 0.753078892817949,
+      "grad_norm": 0.8864921293794893,
+      "learning_rate": 7.887499223272968e-07,
+      "loss": 0.1298,
+      "step": 772
+    },
+    {
+      "epoch": 0.7540543836117547,
+      "grad_norm": 0.8392771141328131,
+      "learning_rate": 7.828743165418393e-07,
+      "loss": 0.111,
+      "step": 773
+    },
+    {
+      "epoch": 0.7550298744055604,
+      "grad_norm": 0.821355233876034,
+      "learning_rate": 7.770166116936923e-07,
+      "loss": 0.1139,
+      "step": 774
+    },
+    {
+      "epoch": 0.7560053651993659,
+      "grad_norm": 0.8688766276433155,
+      "learning_rate": 7.711768688490962e-07,
+      "loss": 0.1219,
+      "step": 775
+    },
+    {
+      "epoch": 0.7569808559931716,
+      "grad_norm": 0.9962941039117293,
+      "learning_rate": 7.653551488870356e-07,
+      "loss": 0.1283,
+      "step": 776
+    },
+    {
+      "epoch": 0.7579563467869772,
+      "grad_norm": 0.8627221635187925,
+      "learning_rate": 7.595515124986106e-07,
+      "loss": 0.1237,
+      "step": 777
+    },
+    {
+      "epoch": 0.7589318375807829,
+      "grad_norm": 0.8592174071340959,
+      "learning_rate": 7.537660201863989e-07,
+      "loss": 0.1236,
+      "step": 778
+    },
+    {
+      "epoch": 0.7599073283745885,
+      "grad_norm": 0.8914444059496847,
+      "learning_rate": 7.479987322638274e-07,
+      "loss": 0.126,
+      "step": 779
+    },
+    {
+      "epoch": 0.7608828191683941,
+      "grad_norm": 0.838972249342774,
+      "learning_rate": 7.422497088545436e-07,
+      "loss": 0.1225,
+      "step": 780
+    },
+    {
+      "epoch": 0.7618583099621997,
+      "grad_norm": 0.8350154394486518,
+      "learning_rate": 7.365190098917896e-07,
+      "loss": 0.1172,
+      "step": 781
+    },
+    {
+      "epoch": 0.7628338007560054,
+      "grad_norm": 0.8527764567297885,
+      "learning_rate": 7.308066951177742e-07,
+      "loss": 0.1194,
+      "step": 782
+    },
+    {
+      "epoch": 0.763809291549811,
+      "grad_norm": 0.7880927395631528,
+      "learning_rate": 7.251128240830543e-07,
+      "loss": 0.1122,
+      "step": 783
+    },
+    {
+      "epoch": 0.7647847823436167,
+      "grad_norm": 0.8570625979073652,
+      "learning_rate": 7.194374561459094e-07,
+      "loss": 0.1275,
+      "step": 784
+    },
+    {
+      "epoch": 0.7657602731374222,
+      "grad_norm": 0.9193780613269857,
+      "learning_rate": 7.137806504717276e-07,
+      "loss": 0.136,
+      "step": 785
+    },
+    {
+      "epoch": 0.7667357639312279,
+      "grad_norm": 0.8788105843538093,
+      "learning_rate": 7.081424660323846e-07,
+      "loss": 0.1213,
+      "step": 786
+    },
+    {
+      "epoch": 0.7677112547250335,
+      "grad_norm": 0.8629850615115177,
+      "learning_rate": 7.025229616056326e-07,
+      "loss": 0.1223,
+      "step": 787
+    },
+    {
+      "epoch": 0.7686867455188392,
+      "grad_norm": 0.8509382827108314,
+      "learning_rate": 6.969221957744832e-07,
+      "loss": 0.1325,
+      "step": 788
+    },
+    {
+      "epoch": 0.7696622363126449,
+      "grad_norm": 0.8104302202106697,
+      "learning_rate": 6.913402269266026e-07,
+      "loss": 0.1193,
+      "step": 789
+    },
+    {
+      "epoch": 0.7706377271064504,
+      "grad_norm": 0.7876537221330931,
+      "learning_rate": 6.857771132536958e-07,
+      "loss": 0.1177,
+      "step": 790
+    },
+    {
+      "epoch": 0.771613217900256,
+      "grad_norm": 0.8452359406579485,
+      "learning_rate": 6.802329127509072e-07,
+      "loss": 0.1236,
+      "step": 791
+    },
+    {
+      "epoch": 0.7725887086940617,
+      "grad_norm": 0.9041467869529476,
+      "learning_rate": 6.747076832162094e-07,
+      "loss": 0.1331,
+      "step": 792
+    },
+    {
+      "epoch": 0.7735641994878674,
+      "grad_norm": 0.8750663033303995,
+      "learning_rate": 6.692014822498066e-07,
+      "loss": 0.1179,
+      "step": 793
+    },
+    {
+      "epoch": 0.774539690281673,
+      "grad_norm": 0.7709557191713534,
+      "learning_rate": 6.637143672535282e-07,
+      "loss": 0.1123,
+      "step": 794
+    },
+    {
+      "epoch": 0.7755151810754786,
+      "grad_norm": 0.8489536750773737,
+      "learning_rate": 6.582463954302368e-07,
+      "loss": 0.1189,
+      "step": 795
+    },
+    {
+      "epoch": 0.7764906718692842,
+      "grad_norm": 0.7986353115358458,
+      "learning_rate": 6.527976237832256e-07,
+      "loss": 0.1145,
+      "step": 796
+    },
+    {
+      "epoch": 0.7774661626630899,
+      "grad_norm": 0.8636955041864243,
+      "learning_rate": 6.473681091156289e-07,
+      "loss": 0.1235,
+      "step": 797
+    },
+    {
+      "epoch": 0.7784416534568955,
+      "grad_norm": 0.8485394380202599,
+      "learning_rate": 6.419579080298263e-07,
+      "loss": 0.1218,
+      "step": 798
+    },
+    {
+      "epoch": 0.7794171442507012,
+      "grad_norm": 0.8626133292687657,
+      "learning_rate": 6.36567076926857e-07,
+      "loss": 0.1227,
+      "step": 799
+    },
+    {
+      "epoch": 0.7803926350445067,
+      "grad_norm": 0.8270229140897313,
+      "learning_rate": 6.311956720058263e-07,
+      "loss": 0.1278,
+      "step": 800
+    },
+    {
+      "epoch": 0.7813681258383124,
+      "grad_norm": 0.8821105926200152,
+      "learning_rate": 6.258437492633254e-07,
+      "loss": 0.1251,
+      "step": 801
+    },
+    {
+      "epoch": 0.782343616632118,
+      "grad_norm": 0.7683315982295926,
+      "learning_rate": 6.205113644928429e-07,
+      "loss": 0.1139,
+      "step": 802
+    },
+    {
+      "epoch": 0.7833191074259237,
+      "grad_norm": 0.8270618740008568,
+      "learning_rate": 6.151985732841867e-07,
+      "loss": 0.1117,
+      "step": 803
+    },
+    {
+      "epoch": 0.7842945982197294,
+      "grad_norm": 0.8155271658159645,
+      "learning_rate": 6.099054310229008e-07,
+      "loss": 0.1308,
+      "step": 804
+    },
+    {
+      "epoch": 0.7852700890135349,
+      "grad_norm": 0.7877985298756666,
+      "learning_rate": 6.046319928896926e-07,
+      "loss": 0.1157,
+      "step": 805
+    },
+    {
+      "epoch": 0.7862455798073406,
+      "grad_norm": 0.8527729286111828,
+      "learning_rate": 5.993783138598532e-07,
+      "loss": 0.1254,
+      "step": 806
+    },
+    {
+      "epoch": 0.7872210706011462,
+      "grad_norm": 0.8162522251396951,
+      "learning_rate": 5.941444487026877e-07,
+      "loss": 0.1193,
+      "step": 807
+    },
+    {
+      "epoch": 0.7881965613949519,
+      "grad_norm": 0.8198785573921169,
+      "learning_rate": 5.889304519809402e-07,
+      "loss": 0.1225,
+      "step": 808
+    },
+    {
+      "epoch": 0.7891720521887575,
+      "grad_norm": 0.8649744547672364,
+      "learning_rate": 5.837363780502309e-07,
+      "loss": 0.1301,
+      "step": 809
+    },
+    {
+      "epoch": 0.7901475429825631,
+      "grad_norm": 0.8253170535592403,
+      "learning_rate": 5.785622810584834e-07,
+      "loss": 0.1173,
+      "step": 810
+    },
+    {
+      "epoch": 0.7911230337763687,
+      "grad_norm": 0.7949976111986673,
+      "learning_rate": 5.734082149453634e-07,
+      "loss": 0.1134,
+      "step": 811
+    },
+    {
+      "epoch": 0.7920985245701744,
+      "grad_norm": 0.8803222417671256,
+      "learning_rate": 5.682742334417171e-07,
+      "loss": 0.125,
+      "step": 812
+    },
+    {
+      "epoch": 0.79307401536398,
+      "grad_norm": 0.8262641815619376,
+      "learning_rate": 5.631603900690077e-07,
+      "loss": 0.1214,
+      "step": 813
+    },
+    {
+      "epoch": 0.7940495061577857,
+      "grad_norm": 0.849372433500513,
+      "learning_rate": 5.580667381387623e-07,
+      "loss": 0.1244,
+      "step": 814
+    },
+    {
+      "epoch": 0.7950249969515912,
+      "grad_norm": 0.7908963105671865,
+      "learning_rate": 5.529933307520102e-07,
+      "loss": 0.1087,
+      "step": 815
+    },
+    {
+      "epoch": 0.7960004877453969,
+      "grad_norm": 0.8908242498302418,
+      "learning_rate": 5.47940220798735e-07,
+      "loss": 0.1171,
+      "step": 816
+    },
+    {
+      "epoch": 0.7969759785392025,
+      "grad_norm": 0.8393797963498955,
+      "learning_rate": 5.429074609573184e-07,
+      "loss": 0.1226,
+      "step": 817
+    },
+    {
+      "epoch": 0.7979514693330082,
+      "grad_norm": 0.8535887142626548,
+      "learning_rate": 5.378951036939966e-07,
+      "loss": 0.1238,
+      "step": 818
+    },
+    {
+      "epoch": 0.7989269601268139,
+      "grad_norm": 0.8500240894184807,
+      "learning_rate": 5.329032012623064e-07,
+      "loss": 0.1244,
+      "step": 819
+    },
+    {
+      "epoch": 0.7999024509206194,
+      "grad_norm": 0.8632462798593744,
+      "learning_rate": 5.279318057025476e-07,
+      "loss": 0.1131,
+      "step": 820
+    },
+    {
+      "epoch": 0.8008779417144251,
+      "grad_norm": 0.8502544027437935,
+      "learning_rate": 5.229809688412337e-07,
+      "loss": 0.125,
+      "step": 821
+    },
+    {
+      "epoch": 0.8018534325082307,
+      "grad_norm": 0.8275247834448751,
+      "learning_rate": 5.180507422905585e-07,
+      "loss": 0.1182,
+      "step": 822
+    },
+    {
+      "epoch": 0.8028289233020364,
+      "grad_norm": 0.8976360407929419,
+      "learning_rate": 5.131411774478503e-07,
+      "loss": 0.1308,
+      "step": 823
+    },
+    {
+      "epoch": 0.803804414095842,
+      "grad_norm": 0.8925880179518361,
+      "learning_rate": 5.082523254950447e-07,
+      "loss": 0.1194,
+      "step": 824
+    },
+    {
+      "epoch": 0.8047799048896476,
+      "grad_norm": 0.7964366205604348,
+      "learning_rate": 5.033842373981429e-07,
+      "loss": 0.1157,
+      "step": 825
+    },
+    {
+      "epoch": 0.8057553956834532,
+      "grad_norm": 0.8629927042044921,
+      "learning_rate": 4.985369639066867e-07,
+      "loss": 0.1231,
+      "step": 826
+    },
+    {
+      "epoch": 0.8067308864772589,
+      "grad_norm": 0.8130144409152532,
+      "learning_rate": 4.937105555532252e-07,
+      "loss": 0.1205,
+      "step": 827
+    },
+    {
+      "epoch": 0.8077063772710645,
+      "grad_norm": 0.7968686512185147,
+      "learning_rate": 4.889050626527905e-07,
+      "loss": 0.1222,
+      "step": 828
+    },
+    {
+      "epoch": 0.8086818680648702,
+      "grad_norm": 0.8084128367948819,
+      "learning_rate": 4.841205353023715e-07,
+      "loss": 0.1176,
+      "step": 829
+    },
+    {
+      "epoch": 0.8096573588586757,
+      "grad_norm": 0.8260965745640515,
+      "learning_rate": 4.79357023380394e-07,
+      "loss": 0.1168,
+      "step": 830
+    },
+    {
+      "epoch": 0.8106328496524814,
+      "grad_norm": 0.8136591394426701,
+      "learning_rate": 4.746145765461965e-07,
+      "loss": 0.1246,
+      "step": 831
+    },
+    {
+      "epoch": 0.811608340446287,
+      "grad_norm": 0.84880591266123,
+      "learning_rate": 4.6989324423951847e-07,
+      "loss": 0.1207,
+      "step": 832
+    },
+    {
+      "epoch": 0.8125838312400927,
+      "grad_norm": 0.8381317559160933,
+      "learning_rate": 4.651930756799794e-07,
+      "loss": 0.1214,
+      "step": 833
+    },
+    {
+      "epoch": 0.8135593220338984,
+      "grad_norm": 0.8225053300538132,
+      "learning_rate": 4.6051411986656775e-07,
+      "loss": 0.1222,
+      "step": 834
+    },
+    {
+      "epoch": 0.8145348128277039,
+      "grad_norm": 0.7758201411310363,
+      "learning_rate": 4.5585642557713083e-07,
+      "loss": 0.1108,
+      "step": 835
+    },
+    {
+      "epoch": 0.8155103036215096,
+      "grad_norm": 0.7841422014242995,
+      "learning_rate": 4.512200413678672e-07,
+      "loss": 0.1149,
+      "step": 836
+    },
+    {
+      "epoch": 0.8164857944153152,
+      "grad_norm": 0.8433260125398965,
+      "learning_rate": 4.466050155728166e-07,
+      "loss": 0.1246,
+      "step": 837
+    },
+    {
+      "epoch": 0.8174612852091209,
+      "grad_norm": 0.8004993062409861,
+      "learning_rate": 4.420113963033607e-07,
+      "loss": 0.1139,
+      "step": 838
+    },
+    {
+      "epoch": 0.8184367760029265,
+      "grad_norm": 0.8544261665582314,
+      "learning_rate": 4.374392314477174e-07,
+      "loss": 0.1442,
+      "step": 839
+    },
+    {
+      "epoch": 0.8194122667967321,
+      "grad_norm": 0.8269193033610706,
+      "learning_rate": 4.3288856867044564e-07,
+      "loss": 0.117,
+      "step": 840
+    },
+    {
+      "epoch": 0.8203877575905377,
+      "grad_norm": 0.8056654304417562,
+      "learning_rate": 4.2835945541194414e-07,
+      "loss": 0.1106,
+      "step": 841
+    },
+    {
+      "epoch": 0.8213632483843434,
+      "grad_norm": 0.8225949604333195,
+      "learning_rate": 4.238519388879614e-07,
+      "loss": 0.118,
+      "step": 842
+    },
+    {
+      "epoch": 0.822338739178149,
+      "grad_norm": 0.8057418784933338,
+      "learning_rate": 4.1936606608909887e-07,
+      "loss": 0.1155,
+      "step": 843
+    },
+    {
+      "epoch": 0.8233142299719547,
+      "grad_norm": 0.8424817751997794,
+      "learning_rate": 4.149018837803262e-07,
+      "loss": 0.124,
+      "step": 844
+    },
+    {
+      "epoch": 0.8242897207657602,
+      "grad_norm": 0.7785385373841229,
+      "learning_rate": 4.1045943850048843e-07,
+      "loss": 0.1229,
+      "step": 845
+    },
+    {
+      "epoch": 0.8252652115595659,
+      "grad_norm": 0.8139230651458013,
+      "learning_rate": 4.0603877656182373e-07,
+      "loss": 0.1201,
+      "step": 846
+    },
+    {
+      "epoch": 0.8262407023533715,
+      "grad_norm": 0.8136291299903247,
+      "learning_rate": 4.0163994404948144e-07,
+      "loss": 0.1147,
+      "step": 847
+    },
+    {
+      "epoch": 0.8272161931471772,
+      "grad_norm": 0.7604644453749246,
+      "learning_rate": 3.9726298682103923e-07,
+      "loss": 0.1136,
+      "step": 848
+    },
+    {
+      "epoch": 0.8281916839409829,
+      "grad_norm": 0.8003636956695743,
+      "learning_rate": 3.929079505060271e-07,
+      "loss": 0.1104,
+      "step": 849
+    },
+    {
+      "epoch": 0.8291671747347884,
+      "grad_norm": 0.8051074779072083,
+      "learning_rate": 3.8857488050544903e-07,
+      "loss": 0.1196,
+      "step": 850
+    },
+    {
+      "epoch": 0.8301426655285941,
+      "grad_norm": 0.8417299086430947,
+      "learning_rate": 3.842638219913142e-07,
+      "loss": 0.1187,
+      "step": 851
+    },
+    {
+      "epoch": 0.8311181563223997,
+      "grad_norm": 0.8473564595322601,
+      "learning_rate": 3.799748199061612e-07,
+      "loss": 0.1238,
+      "step": 852
+    },
+    {
+      "epoch": 0.8320936471162054,
+      "grad_norm": 0.8530209378089045,
+      "learning_rate": 3.7570791896259147e-07,
+      "loss": 0.1195,
+      "step": 853
+    },
+    {
+      "epoch": 0.833069137910011,
+      "grad_norm": 0.8683334198893611,
+      "learning_rate": 3.7146316364280426e-07,
+      "loss": 0.1202,
+      "step": 854
+    },
+    {
+      "epoch": 0.8340446287038166,
+      "grad_norm": 0.8185200067165666,
+      "learning_rate": 3.6724059819813223e-07,
+      "loss": 0.1244,
+      "step": 855
+    },
+    {
+      "epoch": 0.8350201194976222,
+      "grad_norm": 0.8871001344843794,
+      "learning_rate": 3.630402666485783e-07,
+      "loss": 0.1257,
+      "step": 856
+    },
+    {
+      "epoch": 0.8359956102914279,
+      "grad_norm": 0.8335712456953084,
+      "learning_rate": 3.5886221278236045e-07,
+      "loss": 0.117,
+      "step": 857
+    },
+    {
+      "epoch": 0.8369711010852335,
+      "grad_norm": 0.8724817716079213,
+      "learning_rate": 3.547064801554509e-07,
+      "loss": 0.1223,
+      "step": 858
+    },
+    {
+      "epoch": 0.8379465918790392,
+      "grad_norm": 0.8327646410251778,
+      "learning_rate": 3.5057311209112625e-07,
+      "loss": 0.1273,
+      "step": 859
+    },
+    {
+      "epoch": 0.8389220826728447,
+      "grad_norm": 0.8553723242208955,
+      "learning_rate": 3.4646215167951166e-07,
+      "loss": 0.1237,
+      "step": 860
+    },
+    {
+      "epoch": 0.8398975734666504,
+      "grad_norm": 0.8667258133361339,
+      "learning_rate": 3.4237364177713615e-07,
+      "loss": 0.1194,
+      "step": 861
+    },
+    {
+      "epoch": 0.840873064260456,
+      "grad_norm": 0.775526945884146,
+      "learning_rate": 3.383076250064815e-07,
+      "loss": 0.111,
+      "step": 862
+    },
+    {
+      "epoch": 0.8418485550542617,
+      "grad_norm": 0.8350772764316703,
+      "learning_rate": 3.342641437555411e-07,
+      "loss": 0.1227,
+      "step": 863
+    },
+    {
+      "epoch": 0.8428240458480674,
+      "grad_norm": 0.8737189518152014,
+      "learning_rate": 3.3024324017737555e-07,
+      "loss": 0.1219,
+      "step": 864
+    },
+    {
+      "epoch": 0.8437995366418729,
+      "grad_norm": 0.8525338624818053,
+      "learning_rate": 3.262449561896766e-07,
+      "loss": 0.1208,
+      "step": 865
+    },
+    {
+      "epoch": 0.8447750274356786,
+      "grad_norm": 0.7808415133904315,
+      "learning_rate": 3.2226933347432516e-07,
+      "loss": 0.1115,
+      "step": 866
+    },
+    {
+      "epoch": 0.8457505182294842,
+      "grad_norm": 0.8212690915534371,
+      "learning_rate": 3.183164134769631e-07,
+      "loss": 0.1192,
+      "step": 867
+    },
+    {
+      "epoch": 0.8467260090232899,
+      "grad_norm": 0.8432223045748855,
+      "learning_rate": 3.143862374065548e-07,
+      "loss": 0.1271,
+      "step": 868
+    },
+    {
+      "epoch": 0.8477014998170955,
+      "grad_norm": 0.809073364351975,
+      "learning_rate": 3.104788462349612e-07,
+      "loss": 0.1164,
+      "step": 869
+    },
+    {
+      "epoch": 0.8486769906109011,
+      "grad_norm": 0.8122295362266347,
+      "learning_rate": 3.065942806965139e-07,
+      "loss": 0.1166,
+      "step": 870
+    },
+    {
+      "epoch": 0.8496524814047067,
+      "grad_norm": 0.8628135926038828,
+      "learning_rate": 3.0273258128758585e-07,
+      "loss": 0.1231,
+      "step": 871
+    },
+    {
+      "epoch": 0.8506279721985124,
+      "grad_norm": 0.9038737225256949,
+      "learning_rate": 2.98893788266173e-07,
+      "loss": 0.1269,
+      "step": 872
+    },
+    {
+      "epoch": 0.851603462992318,
+      "grad_norm": 0.8415109669107367,
+      "learning_rate": 2.9507794165147487e-07,
+      "loss": 0.1185,
+      "step": 873
+    },
+    {
+      "epoch": 0.8525789537861237,
+      "grad_norm": 0.8260696094654174,
+      "learning_rate": 2.9128508122347324e-07,
+      "loss": 0.1209,
+      "step": 874
+    },
+    {
+      "epoch": 0.8535544445799292,
+      "grad_norm": 0.7950123942247705,
+      "learning_rate": 2.875152465225234e-07,
+      "loss": 0.1107,
+      "step": 875
+    },
+    {
+      "epoch": 0.8545299353737349,
+      "grad_norm": 0.8036333920809307,
+      "learning_rate": 2.837684768489354e-07,
+      "loss": 0.1159,
+      "step": 876
+    },
+    {
+      "epoch": 0.8555054261675406,
+      "grad_norm": 0.8543568873961171,
+      "learning_rate": 2.800448112625709e-07,
+      "loss": 0.1278,
+      "step": 877
+    },
+    {
+      "epoch": 0.8564809169613462,
+      "grad_norm": 0.8705052724557507,
+      "learning_rate": 2.7634428858242995e-07,
+      "loss": 0.1244,
+      "step": 878
+    },
+    {
+      "epoch": 0.8574564077551519,
+      "grad_norm": 0.8648954628473264,
+      "learning_rate": 2.7266694738625143e-07,
+      "loss": 0.1134,
+      "step": 879
+    },
+    {
+      "epoch": 0.8584318985489574,
+      "grad_norm": 0.8569598558609475,
+      "learning_rate": 2.690128260101069e-07,
+      "loss": 0.1216,
+      "step": 880
+    },
+    {
+      "epoch": 0.8594073893427631,
+      "grad_norm": 0.8615706510023747,
+      "learning_rate": 2.6538196254800393e-07,
+      "loss": 0.1282,
+      "step": 881
+    },
+    {
+      "epoch": 0.8603828801365687,
+      "grad_norm": 0.8654838887587664,
+      "learning_rate": 2.617743948514867e-07,
+      "loss": 0.1279,
+      "step": 882
+    },
+    {
+      "epoch": 0.8613583709303744,
+      "grad_norm": 0.8150117399404916,
+      "learning_rate": 2.5819016052924224e-07,
+      "loss": 0.1186,
+      "step": 883
+    },
+    {
+      "epoch": 0.86233386172418,
+      "grad_norm": 0.8042159227280485,
+      "learning_rate": 2.5462929694670986e-07,
+      "loss": 0.1162,
+      "step": 884
+    },
+    {
+      "epoch": 0.8633093525179856,
+      "grad_norm": 0.8098200477108216,
+      "learning_rate": 2.5109184122568797e-07,
+      "loss": 0.1079,
+      "step": 885
+    },
+    {
+      "epoch": 0.8642848433117912,
+      "grad_norm": 0.8037754437911059,
+      "learning_rate": 2.4757783024395244e-07,
+      "loss": 0.1168,
+      "step": 886
+    },
+    {
+      "epoch": 0.8652603341055969,
+      "grad_norm": 0.8174214190622696,
+      "learning_rate": 2.44087300634866e-07,
+      "loss": 0.1143,
+      "step": 887
+    },
+    {
+      "epoch": 0.8662358248994025,
+      "grad_norm": 0.9250267238278224,
+      "learning_rate": 2.4062028878700074e-07,
+      "loss": 0.1242,
+      "step": 888
+    },
+    {
+      "epoch": 0.8672113156932082,
+      "grad_norm": 0.843085805747338,
+      "learning_rate": 2.3717683084375832e-07,
+      "loss": 0.1179,
+      "step": 889
+    },
+    {
+      "epoch": 0.8681868064870137,
+      "grad_norm": 0.8349662619877875,
+      "learning_rate": 2.3375696270299093e-07,
+      "loss": 0.1193,
+      "step": 890
+    },
+    {
+      "epoch": 0.8691622972808194,
+      "grad_norm": 0.7931357279276846,
+      "learning_rate": 2.3036072001662829e-07,
+      "loss": 0.1112,
+      "step": 891
+    },
+    {
+      "epoch": 0.870137788074625,
+      "grad_norm": 0.8058217864023394,
+      "learning_rate": 2.2698813819030802e-07,
+      "loss": 0.1128,
+      "step": 892
+    },
+    {
+      "epoch": 0.8711132788684307,
+      "grad_norm": 0.8105881503397407,
+      "learning_rate": 2.2363925238300167e-07,
+      "loss": 0.1186,
+      "step": 893
+    },
+    {
+      "epoch": 0.8720887696622364,
+      "grad_norm": 0.793047905194702,
+      "learning_rate": 2.2031409750665422e-07,
+      "loss": 0.1136,
+      "step": 894
+    },
+    {
+      "epoch": 0.8730642604560419,
+      "grad_norm": 0.887128059993288,
+      "learning_rate": 2.170127082258147e-07,
+      "loss": 0.1179,
+      "step": 895
+    },
+    {
+      "epoch": 0.8740397512498476,
+      "grad_norm": 0.7729380382439517,
+      "learning_rate": 2.1373511895727866e-07,
+      "loss": 0.1093,
+      "step": 896
+    },
+    {
+      "epoch": 0.8750152420436532,
+      "grad_norm": 0.8334653630393726,
+      "learning_rate": 2.1048136386972645e-07,
+      "loss": 0.1201,
+      "step": 897
+    },
+    {
+      "epoch": 0.8759907328374589,
+      "grad_norm": 0.8109427974591241,
+      "learning_rate": 2.0725147688337054e-07,
+      "loss": 0.1209,
+      "step": 898
+    },
+    {
+      "epoch": 0.8769662236312645,
+      "grad_norm": 0.8075287607010287,
+      "learning_rate": 2.040454916695972e-07,
+      "loss": 0.1104,
+      "step": 899
+    },
+    {
+      "epoch": 0.8779417144250701,
+      "grad_norm": 0.7907021456077222,
+      "learning_rate": 2.0086344165062054e-07,
+      "loss": 0.1186,
+      "step": 900
+    },
+    {
+      "epoch": 0.8789172052188757,
+      "grad_norm": 0.8321531164794602,
+      "learning_rate": 1.9770535999912967e-07,
+      "loss": 0.1242,
+      "step": 901
+    },
+    {
+      "epoch": 0.8798926960126814,
+      "grad_norm": 0.836504402336581,
+      "learning_rate": 1.945712796379462e-07,
+      "loss": 0.1195,
+      "step": 902
+    },
+    {
+      "epoch": 0.880868186806487,
+      "grad_norm": 0.8414921499203574,
+      "learning_rate": 1.914612332396787e-07,
+      "loss": 0.1191,
+      "step": 903
+    },
+    {
+      "epoch": 0.8818436776002927,
+      "grad_norm": 0.8268050582099995,
+      "learning_rate": 1.883752532263844e-07,
+      "loss": 0.1212,
+      "step": 904
+    },
+    {
+      "epoch": 0.8828191683940982,
+      "grad_norm": 0.7804225339880899,
+      "learning_rate": 1.8531337176922792e-07,
+      "loss": 0.109,
+      "step": 905
+    },
+    {
+      "epoch": 0.8837946591879039,
+      "grad_norm": 0.8089563734380512,
+      "learning_rate": 1.8227562078814903e-07,
+      "loss": 0.1185,
+      "step": 906
+    },
+    {
+      "epoch": 0.8847701499817096,
+      "grad_norm": 0.8524343318848043,
+      "learning_rate": 1.7926203195152898e-07,
+      "loss": 0.1236,
+      "step": 907
+    },
+    {
+      "epoch": 0.8857456407755152,
+      "grad_norm": 0.8146203036325108,
+      "learning_rate": 1.7627263667585882e-07,
+      "loss": 0.1118,
+      "step": 908
+    },
+    {
+      "epoch": 0.8867211315693209,
+      "grad_norm": 0.8438170384778132,
+      "learning_rate": 1.7330746612541387e-07,
+      "loss": 0.1153,
+      "step": 909
+    },
+    {
+      "epoch": 0.8876966223631264,
+      "grad_norm": 0.816511650231362,
+      "learning_rate": 1.7036655121192875e-07,
+      "loss": 0.1262,
+      "step": 910
+    },
+    {
+      "epoch": 0.8886721131569321,
+      "grad_norm": 0.8384874737504223,
+      "learning_rate": 1.6744992259427272e-07,
+      "loss": 0.1241,
+      "step": 911
+    },
+    {
+      "epoch": 0.8896476039507377,
+      "grad_norm": 0.8111140618017998,
+      "learning_rate": 1.6455761067813352e-07,
+      "loss": 0.1137,
+      "step": 912
+    },
+    {
+      "epoch": 0.8906230947445434,
+      "grad_norm": 0.8137411483354814,
+      "learning_rate": 1.6168964561569716e-07,
+      "loss": 0.1177,
+      "step": 913
+    },
+    {
+      "epoch": 0.891598585538349,
+      "grad_norm": 0.7826271884075486,
+      "learning_rate": 1.5884605730533686e-07,
+      "loss": 0.1173,
+      "step": 914
+    },
+    {
+      "epoch": 0.8925740763321546,
+      "grad_norm": 0.8815949613672592,
+      "learning_rate": 1.5602687539129745e-07,
+      "loss": 0.1224,
+      "step": 915
+    },
+    {
+      "epoch": 0.8935495671259602,
+      "grad_norm": 0.8365843017681764,
+      "learning_rate": 1.5323212926339037e-07,
+      "loss": 0.119,
+      "step": 916
+    },
+    {
+      "epoch": 0.8945250579197659,
+      "grad_norm": 0.7838035275941735,
+      "learning_rate": 1.504618480566844e-07,
+      "loss": 0.1137,
+      "step": 917
+    },
+    {
+      "epoch": 0.8955005487135715,
+      "grad_norm": 0.8571652908859096,
+      "learning_rate": 1.4771606065120293e-07,
+      "loss": 0.1176,
+      "step": 918
+    },
+    {
+      "epoch": 0.8964760395073772,
+      "grad_norm": 0.8042632494131905,
+      "learning_rate": 1.4499479567162328e-07,
+      "loss": 0.1193,
+      "step": 919
+    },
+    {
+      "epoch": 0.8974515303011827,
+      "grad_norm": 0.8620451657826478,
+      "learning_rate": 1.4229808148697732e-07,
+      "loss": 0.1245,
+      "step": 920
+    },
+    {
+      "epoch": 0.8984270210949884,
+      "grad_norm": 0.8376113680024069,
+      "learning_rate": 1.396259462103572e-07,
+      "loss": 0.1101,
+      "step": 921
+    },
+    {
+      "epoch": 0.899402511888794,
+      "grad_norm": 0.8311869479243955,
+      "learning_rate": 1.3697841769861996e-07,
+      "loss": 0.1198,
+      "step": 922
+    },
+    {
+      "epoch": 0.9003780026825997,
+      "grad_norm": 0.869133372824549,
+      "learning_rate": 1.3435552355210018e-07,
+      "loss": 0.1172,
+      "step": 923
+    },
+    {
+      "epoch": 0.9013534934764054,
+      "grad_norm": 0.8549023606786766,
+      "learning_rate": 1.3175729111431946e-07,
+      "loss": 0.1248,
+      "step": 924
+    },
+    {
+      "epoch": 0.9023289842702109,
+      "grad_norm": 0.8266104807259295,
+      "learning_rate": 1.2918374747170225e-07,
+      "loss": 0.1157,
+      "step": 925
+    },
+    {
+      "epoch": 0.9033044750640166,
+      "grad_norm": 0.8725974407693242,
+      "learning_rate": 1.266349194532951e-07,
+      "loss": 0.1212,
+      "step": 926
+    },
+    {
+      "epoch": 0.9042799658578222,
+      "grad_norm": 0.8869251827136297,
+      "learning_rate": 1.2411083363048386e-07,
+      "loss": 0.1168,
+      "step": 927
+    },
+    {
+      "epoch": 0.9052554566516279,
+      "grad_norm": 0.838978233063295,
+      "learning_rate": 1.2161151631671974e-07,
+      "loss": 0.1193,
+      "step": 928
+    },
+    {
+      "epoch": 0.9062309474454335,
+      "grad_norm": 0.8158933020010286,
+      "learning_rate": 1.1913699356724317e-07,
+      "loss": 0.1139,
+      "step": 929
+    },
+    {
+      "epoch": 0.9072064382392391,
+      "grad_norm": 0.8302262653149984,
+      "learning_rate": 1.166872911788125e-07,
+      "loss": 0.1193,
+      "step": 930
+    },
+    {
+      "epoch": 0.9081819290330447,
+      "grad_norm": 0.8281059144615461,
+      "learning_rate": 1.1426243468943582e-07,
+      "loss": 0.1218,
+      "step": 931
+    },
+    {
+      "epoch": 0.9091574198268504,
+      "grad_norm": 0.8150387396832428,
+      "learning_rate": 1.1186244937810315e-07,
+      "loss": 0.1133,
+      "step": 932
+    },
+    {
+      "epoch": 0.910132910620656,
+      "grad_norm": 0.8174814241523217,
+      "learning_rate": 1.0948736026452495e-07,
+      "loss": 0.1111,
+      "step": 933
+    },
+    {
+      "epoch": 0.9111084014144617,
+      "grad_norm": 0.8072685450600614,
+      "learning_rate": 1.071371921088693e-07,
+      "loss": 0.1219,
+      "step": 934
+    },
+    {
+      "epoch": 0.9120838922082672,
+      "grad_norm": 0.8462154845847925,
+      "learning_rate": 1.0481196941150574e-07,
+      "loss": 0.1131,
+      "step": 935
+    },
+    {
+      "epoch": 0.9130593830020729,
+      "grad_norm": 0.8194674338836323,
+      "learning_rate": 1.0251171641274788e-07,
+      "loss": 0.1208,
+      "step": 936
+    },
+    {
+      "epoch": 0.9140348737958786,
+      "grad_norm": 0.8051460380945417,
+      "learning_rate": 1.0023645709260233e-07,
+      "loss": 0.111,
+      "step": 937
+    },
+    {
+      "epoch": 0.9150103645896842,
+      "grad_norm": 0.7953737847156309,
+      "learning_rate": 9.798621517051716e-08,
+      "loss": 0.1134,
+      "step": 938
+    },
+    {
+      "epoch": 0.9159858553834899,
+      "grad_norm": 0.8137082279917823,
+      "learning_rate": 9.576101410513655e-08,
+      "loss": 0.1184,
+      "step": 939
+    },
+    {
+      "epoch": 0.9169613461772954,
+      "grad_norm": 0.81376889042028,
+      "learning_rate": 9.356087709405465e-08,
+      "loss": 0.1141,
+      "step": 940
+    },
+    {
+      "epoch": 0.9179368369711011,
+      "grad_norm": 0.8493873934604731,
+      "learning_rate": 9.138582707357429e-08,
+      "loss": 0.1178,
+      "step": 941
+    },
+    {
+      "epoch": 0.9189123277649067,
+      "grad_norm": 0.8480580946913866,
+      "learning_rate": 8.923588671846784e-08,
+      "loss": 0.119,
+      "step": 942
+    },
+    {
+      "epoch": 0.9198878185587124,
+      "grad_norm": 0.8259767599447636,
+      "learning_rate": 8.711107844174089e-08,
+      "loss": 0.1224,
+      "step": 943
+    },
+    {
+      "epoch": 0.920863309352518,
+      "grad_norm": 0.8414744578683574,
+      "learning_rate": 8.501142439439808e-08,
+      "loss": 0.1184,
+      "step": 944
+    },
+    {
+      "epoch": 0.9218388001463236,
+      "grad_norm": 0.7500528197265454,
+      "learning_rate": 8.29369464652141e-08,
+      "loss": 0.0985,
+      "step": 945
+    },
+    {
+      "epoch": 0.9228142909401292,
+      "grad_norm": 0.8122867841326354,
+      "learning_rate": 8.088766628050193e-08,
+      "loss": 0.1102,
+      "step": 946
+    },
+    {
+      "epoch": 0.9237897817339349,
+      "grad_norm": 0.8378630528612887,
+      "learning_rate": 7.886360520389158e-08,
+      "loss": 0.1187,
+      "step": 947
+    },
+    {
+      "epoch": 0.9247652725277405,
+      "grad_norm": 0.7993318052605702,
+      "learning_rate": 7.686478433610339e-08,
+      "loss": 0.113,
+      "step": 948
+    },
+    {
+      "epoch": 0.9257407633215462,
+      "grad_norm": 0.8282335985098225,
+      "learning_rate": 7.48912245147318e-08,
+      "loss": 0.1183,
+      "step": 949
+    },
+    {
+      "epoch": 0.9267162541153517,
+      "grad_norm": 0.8238348823155386,
+      "learning_rate": 7.294294631402493e-08,
+      "loss": 0.1202,
+      "step": 950
+    },
+    {
+      "epoch": 0.9276917449091574,
+      "grad_norm": 0.8024852399758585,
+      "learning_rate": 7.101997004467203e-08,
+      "loss": 0.122,
+      "step": 951
+    },
+    {
+      "epoch": 0.9286672357029631,
+      "grad_norm": 0.7847553156886314,
+      "learning_rate": 6.912231575359057e-08,
+      "loss": 0.1104,
+      "step": 952
+    },
+    {
+      "epoch": 0.9296427264967687,
+      "grad_norm": 0.8975888719583766,
+      "learning_rate": 6.725000322371916e-08,
+      "loss": 0.1186,
+      "step": 953
+    },
+    {
+      "epoch": 0.9306182172905744,
+      "grad_norm": 0.8212833223497472,
+      "learning_rate": 6.540305197380859e-08,
+      "loss": 0.122,
+      "step": 954
+    },
+    {
+      "epoch": 0.9315937080843799,
+      "grad_norm": 0.7982479941876005,
+      "learning_rate": 6.358148125822e-08,
+      "loss": 0.1136,
+      "step": 955
+    },
+    {
+      "epoch": 0.9325691988781856,
+      "grad_norm": 0.8592832767848003,
+      "learning_rate": 6.178531006672484e-08,
+      "loss": 0.1276,
+      "step": 956
+    },
+    {
+      "epoch": 0.9335446896719912,
+      "grad_norm": 0.8873811445927061,
+      "learning_rate": 6.001455712430492e-08,
+      "loss": 0.126,
+      "step": 957
+    },
+    {
+      "epoch": 0.9345201804657969,
+      "grad_norm": 0.8556443661624066,
+      "learning_rate": 5.826924089095881e-08,
+      "loss": 0.1176,
+      "step": 958
+    },
+    {
+      "epoch": 0.9354956712596025,
+      "grad_norm": 0.7924528652097824,
+      "learning_rate": 5.654937956150852e-08,
+      "loss": 0.1208,
+      "step": 959
+    },
+    {
+      "epoch": 0.9364711620534081,
+      "grad_norm": 0.8207355407913433,
+      "learning_rate": 5.4854991065410866e-08,
+      "loss": 0.1236,
+      "step": 960
+    },
+    {
+      "epoch": 0.9374466528472137,
+      "grad_norm": 0.823166152063427,
+      "learning_rate": 5.3186093066568965e-08,
+      "loss": 0.1195,
+      "step": 961
+    },
+    {
+      "epoch": 0.9384221436410194,
+      "grad_norm": 0.9014815011677116,
+      "learning_rate": 5.154270296314878e-08,
+      "loss": 0.1317,
+      "step": 962
+    },
+    {
+      "epoch": 0.939397634434825,
+      "grad_norm": 0.8362981567362682,
+      "learning_rate": 4.992483788739927e-08,
+      "loss": 0.1208,
+      "step": 963
+    },
+    {
+      "epoch": 0.9403731252286307,
+      "grad_norm": 0.840511469516904,
+      "learning_rate": 4.833251470547084e-08,
+      "loss": 0.1293,
+      "step": 964
+    },
+    {
+      "epoch": 0.9413486160224362,
+      "grad_norm": 0.831812913196967,
+      "learning_rate": 4.6765750017242206e-08,
+      "loss": 0.1217,
+      "step": 965
+    },
+    {
+      "epoch": 0.9423241068162419,
+      "grad_norm": 0.8367591325240412,
+      "learning_rate": 4.5224560156145734e-08,
+      "loss": 0.118,
+      "step": 966
+    },
+    {
+      "epoch": 0.9432995976100476,
+      "grad_norm": 0.879426458908269,
+      "learning_rate": 4.370896118899792e-08,
+      "loss": 0.1145,
+      "step": 967
+    },
+    {
+      "epoch": 0.9442750884038532,
+      "grad_norm": 0.8098437070862734,
+      "learning_rate": 4.221896891583144e-08,
+      "loss": 0.1144,
+      "step": 968
+    },
+    {
+      "epoch": 0.9452505791976589,
+      "grad_norm": 0.7667200836569177,
+      "learning_rate": 4.0754598869730824e-08,
+      "loss": 0.11,
+      "step": 969
+    },
+    {
+      "epoch": 0.9462260699914644,
+      "grad_norm": 0.8556949255820121,
+      "learning_rate": 3.931586631667039e-08,
+      "loss": 0.1246,
+      "step": 970
+    },
+    {
+      "epoch": 0.9472015607852701,
+      "grad_norm": 0.8278433518221362,
+      "learning_rate": 3.7902786255354927e-08,
+      "loss": 0.1244,
+      "step": 971
+    },
+    {
+      "epoch": 0.9481770515790757,
+      "grad_norm": 0.8279334840526833,
+      "learning_rate": 3.65153734170634e-08,
+      "loss": 0.1167,
+      "step": 972
+    },
+    {
+      "epoch": 0.9491525423728814,
+      "grad_norm": 0.8237905685355518,
+      "learning_rate": 3.515364226549523e-08,
+      "loss": 0.1205,
+      "step": 973
+    },
+    {
+      "epoch": 0.950128033166687,
+      "grad_norm": 0.8493136167563948,
+      "learning_rate": 3.381760699662062e-08,
+      "loss": 0.1223,
+      "step": 974
+    },
+    {
+      "epoch": 0.9511035239604926,
+      "grad_norm": 0.8361911343136549,
+      "learning_rate": 3.25072815385305e-08,
+      "loss": 0.1161,
+      "step": 975
+    },
+    {
+      "epoch": 0.9520790147542982,
+      "grad_norm": 0.80954482668594,
+      "learning_rate": 3.1222679551293486e-08,
+      "loss": 0.1129,
+      "step": 976
+    },
+    {
+      "epoch": 0.9530545055481039,
+      "grad_norm": 0.82063025704283,
+      "learning_rate": 2.996381442681162e-08,
+      "loss": 0.1191,
+      "step": 977
+    },
+    {
+      "epoch": 0.9540299963419095,
+      "grad_norm": 0.828098268541136,
+      "learning_rate": 2.8730699288682107e-08,
+      "loss": 0.1262,
+      "step": 978
+    },
+    {
+      "epoch": 0.9550054871357152,
+      "grad_norm": 0.8093239279254736,
+      "learning_rate": 2.7523346992060217e-08,
+      "loss": 0.1159,
+      "step": 979
+    },
+    {
+      "epoch": 0.9559809779295207,
+      "grad_norm": 0.8448131471914473,
+      "learning_rate": 2.634177012352468e-08,
+      "loss": 0.1181,
+      "step": 980
+    },
+    {
+      "epoch": 0.9569564687233264,
+      "grad_norm": 0.850456037528909,
+      "learning_rate": 2.5185981000946657e-08,
+      "loss": 0.1232,
+      "step": 981
+    },
+    {
+      "epoch": 0.9579319595171321,
+      "grad_norm": 0.7998292649044384,
+      "learning_rate": 2.405599167336209e-08,
+      "loss": 0.1152,
+      "step": 982
+    },
+    {
+      "epoch": 0.9589074503109377,
+      "grad_norm": 0.8711285830989083,
+      "learning_rate": 2.295181392084511e-08,
+      "loss": 0.1192,
+      "step": 983
+    },
+    {
+      "epoch": 0.9598829411047434,
+      "grad_norm": 0.831799017458,
+      "learning_rate": 2.187345925438594e-08,
+      "loss": 0.1143,
+      "step": 984
+    },
+    {
+      "epoch": 0.9608584318985489,
+      "grad_norm": 0.8032074365365065,
+      "learning_rate": 2.0820938915770417e-08,
+      "loss": 0.1105,
+      "step": 985
+    },
+    {
+      "epoch": 0.9618339226923546,
+      "grad_norm": 0.7978427999769494,
+      "learning_rate": 1.9794263877463138e-08,
+      "loss": 0.1061,
+      "step": 986
+    },
+    {
+      "epoch": 0.9628094134861602,
+      "grad_norm": 0.8519627713839724,
+      "learning_rate": 1.879344484249257e-08,
+      "loss": 0.1229,
+      "step": 987
+    },
+    {
+      "epoch": 0.9637849042799659,
+      "grad_norm": 0.8131083664763187,
+      "learning_rate": 1.781849224434029e-08,
+      "loss": 0.1178,
+      "step": 988
+    },
+    {
+      "epoch": 0.9647603950737715,
+      "grad_norm": 0.8711093922420086,
+      "learning_rate": 1.686941624683164e-08,
+      "loss": 0.1268,
+      "step": 989
+    },
+    {
+      "epoch": 0.9657358858675771,
+      "grad_norm": 0.8153485749863514,
+      "learning_rate": 1.5946226744029402e-08,
+      "loss": 0.1139,
+      "step": 990
+    },
+    {
+      "epoch": 0.9667113766613827,
+      "grad_norm": 0.8470338609149345,
+      "learning_rate": 1.504893336013169e-08,
+      "loss": 0.117,
+      "step": 991
+    },
+    {
+      "epoch": 0.9676868674551884,
+      "grad_norm": 0.8952280575519709,
+      "learning_rate": 1.4177545449370889e-08,
+      "loss": 0.1193,
+      "step": 992
+    },
+    {
+      "epoch": 0.968662358248994,
+      "grad_norm": 0.8393629890478395,
+      "learning_rate": 1.3332072095916527e-08,
+      "loss": 0.1148,
+      "step": 993
+    },
+    {
+      "epoch": 0.9696378490427997,
+      "grad_norm": 0.8484999365301013,
+      "learning_rate": 1.2512522113779235e-08,
+      "loss": 0.1212,
+      "step": 994
+    },
+    {
+      "epoch": 0.9706133398366052,
+      "grad_norm": 0.8047573717380992,
+      "learning_rate": 1.1718904046721657e-08,
+      "loss": 0.1169,
+      "step": 995
+    },
+    {
+      "epoch": 0.9715888306304109,
+      "grad_norm": 0.7801955720110736,
+      "learning_rate": 1.0951226168166574e-08,
+      "loss": 0.1158,
+      "step": 996
+    },
+    {
+      "epoch": 0.9725643214242166,
+      "grad_norm": 0.8469472102608405,
+      "learning_rate": 1.0209496481112247e-08,
+      "loss": 0.119,
+      "step": 997
+    },
+    {
+      "epoch": 0.9735398122180222,
+      "grad_norm": 0.7899772186862825,
+      "learning_rate": 9.493722718048326e-09,
+      "loss": 0.108,
+      "step": 998
+    },
+    {
+      "epoch": 0.9745153030118279,
+      "grad_norm": 0.8122633623112819,
+      "learning_rate": 8.803912340875076e-09,
+      "loss": 0.1149,
+      "step": 999
+    },
+    {
+      "epoch": 0.9754907938056334,
+      "grad_norm": 0.861388140402043,
+      "learning_rate": 8.140072540826772e-09,
+      "loss": 0.1229,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9754907938056334,
+      "eval_loss": 0.11711173504590988,
+      "eval_runtime": 127.3978,
+      "eval_samples_per_second": 5.204,
+      "eval_steps_per_second": 0.652,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9764662845994391,
+      "grad_norm": 0.795759380098638,
+      "learning_rate": 7.502210238395091e-09,
+      "loss": 0.1202,
+      "step": 1001
+    },
+    {
+      "epoch": 0.9774417753932447,
+      "grad_norm": 0.7911131054538534,
+      "learning_rate": 6.890332083258622e-09,
+      "loss": 0.1206,
+      "step": 1002
+    },
+    {
+      "epoch": 0.9784172661870504,
+      "grad_norm": 0.8225435058710419,
+      "learning_rate": 6.304444454212077e-09,
+      "loss": 0.1253,
+      "step": 1003
+    },
+    {
+      "epoch": 0.979392756980856,
+      "grad_norm": 0.7897526541638892,
+      "learning_rate": 5.7445534591002435e-09,
+      "loss": 0.115,
+      "step": 1004
+    },
+    {
+      "epoch": 0.9803682477746616,
+      "grad_norm": 0.8241723887349869,
+      "learning_rate": 5.210664934754972e-09,
+      "loss": 0.1248,
+      "step": 1005
+    },
+    {
+      "epoch": 0.9813437385684672,
+      "grad_norm": 0.8404643550865318,
+      "learning_rate": 4.702784446934116e-09,
+      "loss": 0.1185,
+      "step": 1006
+    },
+    {
+      "epoch": 0.9823192293622729,
+      "grad_norm": 0.8329025507816016,
+      "learning_rate": 4.220917290262139e-09,
+      "loss": 0.1221,
+      "step": 1007
+    },
+    {
+      "epoch": 0.9832947201560785,
+      "grad_norm": 0.8272251709257826,
+      "learning_rate": 3.765068488177093e-09,
+      "loss": 0.1243,
+      "step": 1008
+    },
+    {
+      "epoch": 0.9842702109498842,
+      "grad_norm": 0.8289663001334425,
+      "learning_rate": 3.335242792876503e-09,
+      "loss": 0.1121,
+      "step": 1009
+    },
+    {
+      "epoch": 0.9852457017436898,
+      "grad_norm": 0.8935773554975892,
+      "learning_rate": 2.931444685269069e-09,
+      "loss": 0.1296,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9862211925374954,
+      "grad_norm": 0.8147353591204097,
+      "learning_rate": 2.553678374926649e-09,
+      "loss": 0.1197,
+      "step": 1011
+    },
+    {
+      "epoch": 0.9871966833313011,
+      "grad_norm": 0.8356734189663587,
+      "learning_rate": 2.201947800041515e-09,
+      "loss": 0.1233,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9881721741251067,
+      "grad_norm": 0.8277826045315564,
+      "learning_rate": 1.8762566273852754e-09,
+      "loss": 0.1102,
+      "step": 1013
+    },
+    {
+      "epoch": 0.9891476649189124,
+      "grad_norm": 0.837735922536016,
+      "learning_rate": 1.576608252269185e-09,
+      "loss": 0.1282,
+      "step": 1014
+    },
+    {
+      "epoch": 0.9901231557127179,
+      "grad_norm": 0.8185965403465947,
+      "learning_rate": 1.3030057985108368e-09,
+      "loss": 0.1184,
+      "step": 1015
+    },
+    {
+      "epoch": 0.9910986465065236,
+      "grad_norm": 0.7801036446264883,
+      "learning_rate": 1.055452118400302e-09,
+      "loss": 0.111,
+      "step": 1016
+    },
+    {
+      "epoch": 0.9920741373003292,
+      "grad_norm": 0.7975320236568699,
+      "learning_rate": 8.339497926704299e-10,
+      "loss": 0.114,
+      "step": 1017
+    },
+    {
+      "epoch": 0.9930496280941349,
+      "grad_norm": 0.8525955644349905,
+      "learning_rate": 6.385011304704814e-10,
+      "loss": 0.1257,
+      "step": 1018
+    },
+    {
+      "epoch": 0.9940251188879405,
+      "grad_norm": 0.8091050888964465,
+      "learning_rate": 4.691081693411481e-10,
+      "loss": 0.1147,
+      "step": 1019
+    },
+    {
+      "epoch": 0.9950006096817461,
+      "grad_norm": 0.9205192964480091,
+      "learning_rate": 3.2577267519484645e-10,
+      "loss": 0.1257,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9959761004755517,
+      "grad_norm": 0.816877602174351,
+      "learning_rate": 2.0849614229601078e-10,
+      "loss": 0.1145,
+      "step": 1021
+    },
+    {
+      "epoch": 0.9969515912693574,
+      "grad_norm": 0.8091963243679661,
+      "learning_rate": 1.1727979324527294e-10,
+      "loss": 0.1185,
+      "step": 1022
+    },
+    {
+      "epoch": 0.997927082063163,
+      "grad_norm": 0.8922096606338227,
+      "learning_rate": 5.212457896835998e-11,
+      "loss": 0.118,
+      "step": 1023
+    },
+    {
+      "epoch": 0.9989025728569687,
+      "grad_norm": 0.7963658344295353,
+      "learning_rate": 1.303117870443682e-11,
+      "loss": 0.118,
+      "step": 1024
+    },
+    {
+      "epoch": 0.9998780636507743,
+      "grad_norm": 0.832964772750828,
+      "learning_rate": 0.0,
+      "loss": 0.1208,
+      "step": 1025
+    },
+    {
+      "epoch": 0.9998780636507743,
+      "step": 1025,
+      "total_flos": 241738000957440.0,
+      "train_loss": 0.1467944400630346,
+      "train_runtime": 47673.2991,
+      "train_samples_per_second": 1.376,
+      "train_steps_per_second": 0.022
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1025,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 241738000957440.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}