healed1 / trainer_state.json
novcor's picture
Duplicate from Ffohturk/mistral-7b_ablated_model_layer_1_gate_only_healed
de3d899
Raw
History Blame Contribute Delete
281 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 8715,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0,
"grad_norm": 0.5823844833414168,
"learning_rate": 2.2935779816513765e-08,
"loss": 1.2424,
"step": 1
},
{
"epoch": 0.0,
"grad_norm": 0.4905919347663963,
"learning_rate": 1.1467889908256882e-07,
"loss": 1.1219,
"step": 5
},
{
"epoch": 0.0,
"grad_norm": 0.5016970441782794,
"learning_rate": 2.2935779816513764e-07,
"loss": 1.1942,
"step": 10
},
{
"epoch": 0.0,
"grad_norm": 0.4157563931960181,
"learning_rate": 3.4403669724770646e-07,
"loss": 1.3548,
"step": 15
},
{
"epoch": 0.0,
"grad_norm": 0.4885420154085299,
"learning_rate": 4.587155963302753e-07,
"loss": 1.1494,
"step": 20
},
{
"epoch": 0.0,
"grad_norm": 0.4929741028763733,
"learning_rate": 5.733944954128441e-07,
"loss": 1.1498,
"step": 25
},
{
"epoch": 0.0,
"grad_norm": 0.5715246158149258,
"learning_rate": 6.880733944954129e-07,
"loss": 1.16,
"step": 30
},
{
"epoch": 0.0,
"grad_norm": 0.4626044808381683,
"learning_rate": 8.027522935779817e-07,
"loss": 1.2143,
"step": 35
},
{
"epoch": 0.0,
"grad_norm": 0.5445023189301125,
"learning_rate": 9.174311926605506e-07,
"loss": 1.203,
"step": 40
},
{
"epoch": 0.01,
"grad_norm": 0.4499744513720404,
"learning_rate": 1.0321100917431195e-06,
"loss": 1.1442,
"step": 45
},
{
"epoch": 0.01,
"grad_norm": 0.46179311433884895,
"learning_rate": 1.1467889908256882e-06,
"loss": 1.1865,
"step": 50
},
{
"epoch": 0.01,
"grad_norm": 0.45584819263804105,
"learning_rate": 1.261467889908257e-06,
"loss": 1.2883,
"step": 55
},
{
"epoch": 0.01,
"grad_norm": 1.2262692368041845,
"learning_rate": 1.3761467889908258e-06,
"loss": 1.2787,
"step": 60
},
{
"epoch": 0.01,
"grad_norm": 0.45723705147806554,
"learning_rate": 1.4908256880733945e-06,
"loss": 1.1488,
"step": 65
},
{
"epoch": 0.01,
"grad_norm": 0.4675821894965279,
"learning_rate": 1.6055045871559635e-06,
"loss": 1.1986,
"step": 70
},
{
"epoch": 0.01,
"grad_norm": 0.4621961071654268,
"learning_rate": 1.7201834862385322e-06,
"loss": 1.0978,
"step": 75
},
{
"epoch": 0.01,
"grad_norm": 0.41073179701335816,
"learning_rate": 1.8348623853211011e-06,
"loss": 1.122,
"step": 80
},
{
"epoch": 0.01,
"grad_norm": 0.4057739018674834,
"learning_rate": 1.94954128440367e-06,
"loss": 1.1334,
"step": 85
},
{
"epoch": 0.01,
"grad_norm": 0.35893657829336006,
"learning_rate": 2.064220183486239e-06,
"loss": 1.1138,
"step": 90
},
{
"epoch": 0.01,
"grad_norm": 0.36609879204159684,
"learning_rate": 2.1788990825688075e-06,
"loss": 1.2585,
"step": 95
},
{
"epoch": 0.01,
"grad_norm": 2.431262066302421,
"learning_rate": 2.2935779816513764e-06,
"loss": 1.1167,
"step": 100
},
{
"epoch": 0.01,
"grad_norm": 0.3510941210221594,
"learning_rate": 2.4082568807339453e-06,
"loss": 1.1943,
"step": 105
},
{
"epoch": 0.01,
"grad_norm": 2.515238154461854,
"learning_rate": 2.522935779816514e-06,
"loss": 1.1127,
"step": 110
},
{
"epoch": 0.01,
"grad_norm": 1.1765142638160402,
"learning_rate": 2.6376146788990823e-06,
"loss": 1.2695,
"step": 115
},
{
"epoch": 0.01,
"grad_norm": 0.37902562131840384,
"learning_rate": 2.7522935779816517e-06,
"loss": 1.2287,
"step": 120
},
{
"epoch": 0.01,
"grad_norm": 0.39360152696148626,
"learning_rate": 2.8669724770642206e-06,
"loss": 1.1445,
"step": 125
},
{
"epoch": 0.01,
"grad_norm": 0.38837621749080947,
"learning_rate": 2.981651376146789e-06,
"loss": 1.0888,
"step": 130
},
{
"epoch": 0.02,
"grad_norm": 0.34490774776329997,
"learning_rate": 3.0963302752293576e-06,
"loss": 1.1013,
"step": 135
},
{
"epoch": 0.02,
"grad_norm": 0.3499301672395144,
"learning_rate": 3.211009174311927e-06,
"loss": 1.0732,
"step": 140
},
{
"epoch": 0.02,
"grad_norm": 0.3778069169459394,
"learning_rate": 3.325688073394496e-06,
"loss": 1.2459,
"step": 145
},
{
"epoch": 0.02,
"grad_norm": 0.37166296952335115,
"learning_rate": 3.4403669724770644e-06,
"loss": 1.1317,
"step": 150
},
{
"epoch": 0.02,
"grad_norm": 0.4171715150808447,
"learning_rate": 3.5550458715596333e-06,
"loss": 1.0663,
"step": 155
},
{
"epoch": 0.02,
"grad_norm": 0.3478987264641242,
"learning_rate": 3.6697247706422022e-06,
"loss": 1.1422,
"step": 160
},
{
"epoch": 0.02,
"grad_norm": 0.3704352145519617,
"learning_rate": 3.784403669724771e-06,
"loss": 1.1269,
"step": 165
},
{
"epoch": 0.02,
"grad_norm": 0.33339854072777325,
"learning_rate": 3.89908256880734e-06,
"loss": 1.1184,
"step": 170
},
{
"epoch": 0.02,
"grad_norm": 0.34725527931912664,
"learning_rate": 4.013761467889909e-06,
"loss": 1.1113,
"step": 175
},
{
"epoch": 0.02,
"grad_norm": 0.34247619339724916,
"learning_rate": 4.128440366972478e-06,
"loss": 1.1345,
"step": 180
},
{
"epoch": 0.02,
"grad_norm": 0.3663067107175219,
"learning_rate": 4.2431192660550464e-06,
"loss": 1.0548,
"step": 185
},
{
"epoch": 0.02,
"grad_norm": 0.30711357683756413,
"learning_rate": 4.357798165137615e-06,
"loss": 1.0726,
"step": 190
},
{
"epoch": 0.02,
"grad_norm": 0.3506524482582489,
"learning_rate": 4.4724770642201834e-06,
"loss": 1.0657,
"step": 195
},
{
"epoch": 0.02,
"grad_norm": 0.3351265464149815,
"learning_rate": 4.587155963302753e-06,
"loss": 1.0668,
"step": 200
},
{
"epoch": 0.02,
"grad_norm": 0.45764762247499813,
"learning_rate": 4.701834862385321e-06,
"loss": 1.0592,
"step": 205
},
{
"epoch": 0.02,
"grad_norm": 0.30976405854210276,
"learning_rate": 4.816513761467891e-06,
"loss": 1.0922,
"step": 210
},
{
"epoch": 0.02,
"grad_norm": 0.3337570640123052,
"learning_rate": 4.931192660550459e-06,
"loss": 1.0972,
"step": 215
},
{
"epoch": 0.03,
"grad_norm": 0.2876295636724402,
"learning_rate": 5.045871559633028e-06,
"loss": 1.0736,
"step": 220
},
{
"epoch": 0.03,
"grad_norm": 0.2976854404226244,
"learning_rate": 5.160550458715596e-06,
"loss": 1.0643,
"step": 225
},
{
"epoch": 0.03,
"grad_norm": 0.30511795987184775,
"learning_rate": 5.275229357798165e-06,
"loss": 1.0929,
"step": 230
},
{
"epoch": 0.03,
"grad_norm": 0.2938188871323575,
"learning_rate": 5.389908256880735e-06,
"loss": 1.0741,
"step": 235
},
{
"epoch": 0.03,
"grad_norm": 0.32362675781091854,
"learning_rate": 5.504587155963303e-06,
"loss": 1.0574,
"step": 240
},
{
"epoch": 0.03,
"grad_norm": 0.2927052069097899,
"learning_rate": 5.619266055045872e-06,
"loss": 1.0515,
"step": 245
},
{
"epoch": 0.03,
"grad_norm": 0.2756019196055756,
"learning_rate": 5.733944954128441e-06,
"loss": 1.0483,
"step": 250
},
{
"epoch": 0.03,
"grad_norm": 0.3340921953300538,
"learning_rate": 5.84862385321101e-06,
"loss": 1.1409,
"step": 255
},
{
"epoch": 0.03,
"grad_norm": 0.3669552185167548,
"learning_rate": 5.963302752293578e-06,
"loss": 1.2462,
"step": 260
},
{
"epoch": 0.03,
"grad_norm": 0.296094584558525,
"learning_rate": 6.077981651376147e-06,
"loss": 1.1513,
"step": 265
},
{
"epoch": 0.03,
"grad_norm": 0.2766119528568034,
"learning_rate": 6.192660550458715e-06,
"loss": 1.0475,
"step": 270
},
{
"epoch": 0.03,
"grad_norm": 0.31034489050870046,
"learning_rate": 6.307339449541285e-06,
"loss": 1.1675,
"step": 275
},
{
"epoch": 0.03,
"grad_norm": 0.2887896350092269,
"learning_rate": 6.422018348623854e-06,
"loss": 1.042,
"step": 280
},
{
"epoch": 0.03,
"grad_norm": 0.32196683050088054,
"learning_rate": 6.536697247706422e-06,
"loss": 1.0776,
"step": 285
},
{
"epoch": 0.03,
"grad_norm": 1.2268154680824443,
"learning_rate": 6.651376146788992e-06,
"loss": 1.044,
"step": 290
},
{
"epoch": 0.03,
"grad_norm": 0.28586840530582275,
"learning_rate": 6.76605504587156e-06,
"loss": 1.084,
"step": 295
},
{
"epoch": 0.03,
"grad_norm": 0.2912437500856281,
"learning_rate": 6.880733944954129e-06,
"loss": 1.1166,
"step": 300
},
{
"epoch": 0.03,
"grad_norm": 0.28504744788543623,
"learning_rate": 6.995412844036697e-06,
"loss": 1.0236,
"step": 305
},
{
"epoch": 0.04,
"grad_norm": 0.2967769494512288,
"learning_rate": 7.110091743119267e-06,
"loss": 1.0667,
"step": 310
},
{
"epoch": 0.04,
"grad_norm": 0.3744280687221159,
"learning_rate": 7.224770642201836e-06,
"loss": 1.1752,
"step": 315
},
{
"epoch": 0.04,
"grad_norm": 0.3025012327849207,
"learning_rate": 7.3394495412844045e-06,
"loss": 1.0494,
"step": 320
},
{
"epoch": 0.04,
"grad_norm": 0.27221098018617795,
"learning_rate": 7.454128440366973e-06,
"loss": 1.0065,
"step": 325
},
{
"epoch": 0.04,
"grad_norm": 0.3068147420251171,
"learning_rate": 7.568807339449542e-06,
"loss": 1.063,
"step": 330
},
{
"epoch": 0.04,
"grad_norm": 0.25686019962572354,
"learning_rate": 7.68348623853211e-06,
"loss": 0.9898,
"step": 335
},
{
"epoch": 0.04,
"grad_norm": 0.2819018794662128,
"learning_rate": 7.79816513761468e-06,
"loss": 1.2149,
"step": 340
},
{
"epoch": 0.04,
"grad_norm": 0.29894215443252237,
"learning_rate": 7.912844036697249e-06,
"loss": 1.0261,
"step": 345
},
{
"epoch": 0.04,
"grad_norm": 0.3214524385307281,
"learning_rate": 8.027522935779817e-06,
"loss": 1.0635,
"step": 350
},
{
"epoch": 0.04,
"grad_norm": 0.8998417404618232,
"learning_rate": 8.142201834862386e-06,
"loss": 1.0955,
"step": 355
},
{
"epoch": 0.04,
"grad_norm": 0.29527582468228014,
"learning_rate": 8.256880733944956e-06,
"loss": 1.0522,
"step": 360
},
{
"epoch": 0.04,
"grad_norm": 0.27015121349032645,
"learning_rate": 8.371559633027524e-06,
"loss": 1.0694,
"step": 365
},
{
"epoch": 0.04,
"grad_norm": 0.25063439133314114,
"learning_rate": 8.486238532110093e-06,
"loss": 1.0454,
"step": 370
},
{
"epoch": 0.04,
"grad_norm": 0.29226247463175326,
"learning_rate": 8.600917431192661e-06,
"loss": 1.0576,
"step": 375
},
{
"epoch": 0.04,
"grad_norm": 0.27249487099923464,
"learning_rate": 8.71559633027523e-06,
"loss": 1.0269,
"step": 380
},
{
"epoch": 0.04,
"grad_norm": 0.2907650351232649,
"learning_rate": 8.830275229357798e-06,
"loss": 1.1445,
"step": 385
},
{
"epoch": 0.04,
"grad_norm": 0.28062621481613037,
"learning_rate": 8.944954128440367e-06,
"loss": 1.0276,
"step": 390
},
{
"epoch": 0.05,
"grad_norm": 0.29210192231762516,
"learning_rate": 9.059633027522935e-06,
"loss": 1.0397,
"step": 395
},
{
"epoch": 0.05,
"grad_norm": 0.3014261703373443,
"learning_rate": 9.174311926605506e-06,
"loss": 1.0271,
"step": 400
},
{
"epoch": 0.05,
"grad_norm": 0.26439383876102013,
"learning_rate": 9.288990825688074e-06,
"loss": 1.0158,
"step": 405
},
{
"epoch": 0.05,
"grad_norm": 0.2783203322824112,
"learning_rate": 9.403669724770643e-06,
"loss": 1.1241,
"step": 410
},
{
"epoch": 0.05,
"grad_norm": 22.71875931420255,
"learning_rate": 9.518348623853211e-06,
"loss": 1.0925,
"step": 415
},
{
"epoch": 0.05,
"grad_norm": 0.36202044582768517,
"learning_rate": 9.633027522935781e-06,
"loss": 1.0274,
"step": 420
},
{
"epoch": 0.05,
"grad_norm": 0.2837393785942686,
"learning_rate": 9.74770642201835e-06,
"loss": 1.0395,
"step": 425
},
{
"epoch": 0.05,
"grad_norm": 0.3308199435652666,
"learning_rate": 9.862385321100918e-06,
"loss": 1.1497,
"step": 430
},
{
"epoch": 0.05,
"grad_norm": 0.3360875192695877,
"learning_rate": 9.977064220183487e-06,
"loss": 1.0417,
"step": 435
},
{
"epoch": 0.05,
"grad_norm": 0.37718945722340097,
"learning_rate": 1.0091743119266055e-05,
"loss": 1.0828,
"step": 440
},
{
"epoch": 0.05,
"grad_norm": 0.2842039683090411,
"learning_rate": 1.0206422018348625e-05,
"loss": 1.0834,
"step": 445
},
{
"epoch": 0.05,
"grad_norm": 0.2978065647047479,
"learning_rate": 1.0321100917431192e-05,
"loss": 1.0595,
"step": 450
},
{
"epoch": 0.05,
"grad_norm": 0.2868851656048096,
"learning_rate": 1.0435779816513762e-05,
"loss": 1.174,
"step": 455
},
{
"epoch": 0.05,
"grad_norm": 0.27280759903031054,
"learning_rate": 1.055045871559633e-05,
"loss": 1.0503,
"step": 460
},
{
"epoch": 0.05,
"grad_norm": 0.22621888031538817,
"learning_rate": 1.06651376146789e-05,
"loss": 1.0289,
"step": 465
},
{
"epoch": 0.05,
"grad_norm": 0.3132201115933474,
"learning_rate": 1.077981651376147e-05,
"loss": 1.0522,
"step": 470
},
{
"epoch": 0.05,
"grad_norm": 0.24227118634933778,
"learning_rate": 1.0894495412844036e-05,
"loss": 1.0011,
"step": 475
},
{
"epoch": 0.06,
"grad_norm": 0.31233832149284846,
"learning_rate": 1.1009174311926607e-05,
"loss": 1.0297,
"step": 480
},
{
"epoch": 0.06,
"grad_norm": 0.23616153282769065,
"learning_rate": 1.1123853211009175e-05,
"loss": 1.01,
"step": 485
},
{
"epoch": 0.06,
"grad_norm": 0.24343553653498548,
"learning_rate": 1.1238532110091744e-05,
"loss": 1.0094,
"step": 490
},
{
"epoch": 0.06,
"grad_norm": 0.24987187520611512,
"learning_rate": 1.1353211009174312e-05,
"loss": 0.9974,
"step": 495
},
{
"epoch": 0.06,
"grad_norm": 0.2989523712522986,
"learning_rate": 1.1467889908256882e-05,
"loss": 1.0377,
"step": 500
},
{
"epoch": 0.06,
"grad_norm": 0.363492885311922,
"learning_rate": 1.1582568807339451e-05,
"loss": 1.139,
"step": 505
},
{
"epoch": 0.06,
"grad_norm": 0.2866302064300267,
"learning_rate": 1.169724770642202e-05,
"loss": 1.0654,
"step": 510
},
{
"epoch": 0.06,
"grad_norm": 0.25270381546969,
"learning_rate": 1.181192660550459e-05,
"loss": 1.0209,
"step": 515
},
{
"epoch": 0.06,
"grad_norm": 0.278889216094585,
"learning_rate": 1.1926605504587156e-05,
"loss": 1.0163,
"step": 520
},
{
"epoch": 0.06,
"grad_norm": 0.4128963526992569,
"learning_rate": 1.2041284403669727e-05,
"loss": 1.1798,
"step": 525
},
{
"epoch": 0.06,
"grad_norm": 0.275659511325767,
"learning_rate": 1.2155963302752293e-05,
"loss": 1.0146,
"step": 530
},
{
"epoch": 0.06,
"grad_norm": 0.25468897113519506,
"learning_rate": 1.2270642201834864e-05,
"loss": 1.0226,
"step": 535
},
{
"epoch": 0.06,
"grad_norm": 0.22760580460352303,
"learning_rate": 1.238532110091743e-05,
"loss": 1.0157,
"step": 540
},
{
"epoch": 0.06,
"grad_norm": 0.22679700263879465,
"learning_rate": 1.25e-05,
"loss": 1.0007,
"step": 545
},
{
"epoch": 0.06,
"grad_norm": 0.2581898999455616,
"learning_rate": 1.261467889908257e-05,
"loss": 1.1386,
"step": 550
},
{
"epoch": 0.06,
"grad_norm": 0.2548514198228377,
"learning_rate": 1.2729357798165138e-05,
"loss": 1.1075,
"step": 555
},
{
"epoch": 0.06,
"grad_norm": 0.24634539743634223,
"learning_rate": 1.2844036697247708e-05,
"loss": 1.0112,
"step": 560
},
{
"epoch": 0.06,
"grad_norm": 3.84450131893556,
"learning_rate": 1.2958715596330276e-05,
"loss": 1.1143,
"step": 565
},
{
"epoch": 0.07,
"grad_norm": 0.3641355112272218,
"learning_rate": 1.3073394495412845e-05,
"loss": 1.0637,
"step": 570
},
{
"epoch": 0.07,
"grad_norm": 0.28630058941027514,
"learning_rate": 1.3188073394495413e-05,
"loss": 1.0719,
"step": 575
},
{
"epoch": 0.07,
"grad_norm": 0.2610408147611276,
"learning_rate": 1.3302752293577984e-05,
"loss": 1.0596,
"step": 580
},
{
"epoch": 0.07,
"grad_norm": 0.2263084619157984,
"learning_rate": 1.341743119266055e-05,
"loss": 1.0314,
"step": 585
},
{
"epoch": 0.07,
"grad_norm": 0.20635017773168865,
"learning_rate": 1.353211009174312e-05,
"loss": 1.0072,
"step": 590
},
{
"epoch": 0.07,
"grad_norm": 0.4625199070691555,
"learning_rate": 1.364678899082569e-05,
"loss": 1.0948,
"step": 595
},
{
"epoch": 0.07,
"grad_norm": 0.2582399117426882,
"learning_rate": 1.3761467889908258e-05,
"loss": 1.0443,
"step": 600
},
{
"epoch": 0.07,
"grad_norm": 0.26800152519617865,
"learning_rate": 1.3876146788990828e-05,
"loss": 1.0779,
"step": 605
},
{
"epoch": 0.07,
"grad_norm": 0.25492674805479854,
"learning_rate": 1.3990825688073395e-05,
"loss": 1.0273,
"step": 610
},
{
"epoch": 0.07,
"grad_norm": 0.2328541154069596,
"learning_rate": 1.4105504587155965e-05,
"loss": 1.004,
"step": 615
},
{
"epoch": 0.07,
"grad_norm": 3.9514665444261943,
"learning_rate": 1.4220183486238533e-05,
"loss": 1.0109,
"step": 620
},
{
"epoch": 0.07,
"grad_norm": 0.2279060809107321,
"learning_rate": 1.4334862385321102e-05,
"loss": 1.0377,
"step": 625
},
{
"epoch": 0.07,
"grad_norm": 0.19631783588795365,
"learning_rate": 1.4449541284403672e-05,
"loss": 1.0434,
"step": 630
},
{
"epoch": 0.07,
"grad_norm": 0.20525361280923918,
"learning_rate": 1.4564220183486239e-05,
"loss": 1.0408,
"step": 635
},
{
"epoch": 0.07,
"grad_norm": 0.2052567296965455,
"learning_rate": 1.4678899082568809e-05,
"loss": 1.0536,
"step": 640
},
{
"epoch": 0.07,
"grad_norm": 0.20430007702834524,
"learning_rate": 1.4793577981651377e-05,
"loss": 1.0254,
"step": 645
},
{
"epoch": 0.07,
"grad_norm": 0.21478076774740312,
"learning_rate": 1.4908256880733946e-05,
"loss": 1.1232,
"step": 650
},
{
"epoch": 0.08,
"grad_norm": 0.23236301566014658,
"learning_rate": 1.5022935779816514e-05,
"loss": 1.0043,
"step": 655
},
{
"epoch": 0.08,
"grad_norm": 0.20955433325630513,
"learning_rate": 1.5137614678899085e-05,
"loss": 0.9974,
"step": 660
},
{
"epoch": 0.08,
"grad_norm": 0.2525415841532138,
"learning_rate": 1.5252293577981651e-05,
"loss": 1.0417,
"step": 665
},
{
"epoch": 0.08,
"grad_norm": 0.3485601590014949,
"learning_rate": 1.536697247706422e-05,
"loss": 1.0841,
"step": 670
},
{
"epoch": 0.08,
"grad_norm": 0.22059705113211228,
"learning_rate": 1.548165137614679e-05,
"loss": 1.0499,
"step": 675
},
{
"epoch": 0.08,
"grad_norm": 0.2019519411804577,
"learning_rate": 1.559633027522936e-05,
"loss": 1.0174,
"step": 680
},
{
"epoch": 0.08,
"grad_norm": 0.19704401835099739,
"learning_rate": 1.5711009174311927e-05,
"loss": 1.0317,
"step": 685
},
{
"epoch": 0.08,
"grad_norm": 0.21370545739820304,
"learning_rate": 1.5825688073394497e-05,
"loss": 1.0393,
"step": 690
},
{
"epoch": 0.08,
"grad_norm": 0.20187517836675914,
"learning_rate": 1.5940366972477068e-05,
"loss": 1.0865,
"step": 695
},
{
"epoch": 0.08,
"grad_norm": 0.22984746083095584,
"learning_rate": 1.6055045871559634e-05,
"loss": 1.01,
"step": 700
},
{
"epoch": 0.08,
"grad_norm": 0.21777763381953258,
"learning_rate": 1.6169724770642205e-05,
"loss": 1.1133,
"step": 705
},
{
"epoch": 0.08,
"grad_norm": 0.2085996524082006,
"learning_rate": 1.628440366972477e-05,
"loss": 1.0535,
"step": 710
},
{
"epoch": 0.08,
"grad_norm": 0.24061979451953563,
"learning_rate": 1.639908256880734e-05,
"loss": 1.056,
"step": 715
},
{
"epoch": 0.08,
"grad_norm": 0.25947677155515125,
"learning_rate": 1.6513761467889912e-05,
"loss": 1.0446,
"step": 720
},
{
"epoch": 0.08,
"grad_norm": 0.18099020213781894,
"learning_rate": 1.662844036697248e-05,
"loss": 1.0266,
"step": 725
},
{
"epoch": 0.08,
"grad_norm": 0.2051863393739792,
"learning_rate": 1.674311926605505e-05,
"loss": 0.9989,
"step": 730
},
{
"epoch": 0.08,
"grad_norm": 0.22345382522652854,
"learning_rate": 1.6857798165137616e-05,
"loss": 1.069,
"step": 735
},
{
"epoch": 0.08,
"grad_norm": 0.18684830637572897,
"learning_rate": 1.6972477064220186e-05,
"loss": 1.0968,
"step": 740
},
{
"epoch": 0.09,
"grad_norm": 0.17459442990788546,
"learning_rate": 1.7087155963302753e-05,
"loss": 0.9951,
"step": 745
},
{
"epoch": 0.09,
"grad_norm": 0.17305319608145542,
"learning_rate": 1.7201834862385323e-05,
"loss": 1.0104,
"step": 750
},
{
"epoch": 0.09,
"grad_norm": 0.18022084183510803,
"learning_rate": 1.7316513761467893e-05,
"loss": 1.0849,
"step": 755
},
{
"epoch": 0.09,
"grad_norm": 0.20535429979515105,
"learning_rate": 1.743119266055046e-05,
"loss": 1.0463,
"step": 760
},
{
"epoch": 0.09,
"grad_norm": 0.29006197754037144,
"learning_rate": 1.754587155963303e-05,
"loss": 1.1274,
"step": 765
},
{
"epoch": 0.09,
"grad_norm": 0.2203084039385792,
"learning_rate": 1.7660550458715597e-05,
"loss": 0.981,
"step": 770
},
{
"epoch": 0.09,
"grad_norm": 0.24650891823333457,
"learning_rate": 1.7775229357798167e-05,
"loss": 1.1291,
"step": 775
},
{
"epoch": 0.09,
"grad_norm": 0.23840848054086958,
"learning_rate": 1.7889908256880734e-05,
"loss": 0.9924,
"step": 780
},
{
"epoch": 0.09,
"grad_norm": 0.17531798475030438,
"learning_rate": 1.8004587155963304e-05,
"loss": 1.027,
"step": 785
},
{
"epoch": 0.09,
"grad_norm": 0.17989210938790912,
"learning_rate": 1.811926605504587e-05,
"loss": 0.9668,
"step": 790
},
{
"epoch": 0.09,
"grad_norm": 0.16335254072517272,
"learning_rate": 1.823394495412844e-05,
"loss": 1.0044,
"step": 795
},
{
"epoch": 0.09,
"grad_norm": 0.1921237635991271,
"learning_rate": 1.834862385321101e-05,
"loss": 1.0431,
"step": 800
},
{
"epoch": 0.09,
"grad_norm": 0.17881822506218778,
"learning_rate": 1.8463302752293578e-05,
"loss": 1.1004,
"step": 805
},
{
"epoch": 0.09,
"grad_norm": 0.1709604777576838,
"learning_rate": 1.8577981651376148e-05,
"loss": 1.043,
"step": 810
},
{
"epoch": 0.09,
"grad_norm": 0.1848298870784547,
"learning_rate": 1.8692660550458715e-05,
"loss": 1.056,
"step": 815
},
{
"epoch": 0.09,
"grad_norm": 0.15474461252109914,
"learning_rate": 1.8807339449541285e-05,
"loss": 0.9946,
"step": 820
},
{
"epoch": 0.09,
"grad_norm": 0.16322025506583887,
"learning_rate": 1.8922018348623855e-05,
"loss": 1.0172,
"step": 825
},
{
"epoch": 0.1,
"grad_norm": 0.18254420688397202,
"learning_rate": 1.9036697247706422e-05,
"loss": 1.0204,
"step": 830
},
{
"epoch": 0.1,
"grad_norm": 0.16281461892628182,
"learning_rate": 1.9151376146788992e-05,
"loss": 1.0147,
"step": 835
},
{
"epoch": 0.1,
"grad_norm": 0.16977834705999914,
"learning_rate": 1.9266055045871563e-05,
"loss": 1.0292,
"step": 840
},
{
"epoch": 0.1,
"grad_norm": 0.16792356043210055,
"learning_rate": 1.938073394495413e-05,
"loss": 1.0216,
"step": 845
},
{
"epoch": 0.1,
"grad_norm": 0.17972157191170515,
"learning_rate": 1.94954128440367e-05,
"loss": 1.0384,
"step": 850
},
{
"epoch": 0.1,
"grad_norm": 0.166599103596908,
"learning_rate": 1.961009174311927e-05,
"loss": 1.0032,
"step": 855
},
{
"epoch": 0.1,
"grad_norm": 0.18126503616347275,
"learning_rate": 1.9724770642201837e-05,
"loss": 1.065,
"step": 860
},
{
"epoch": 0.1,
"grad_norm": 0.28907478494534944,
"learning_rate": 1.9839449541284407e-05,
"loss": 1.0715,
"step": 865
},
{
"epoch": 0.1,
"grad_norm": 0.1940266714567389,
"learning_rate": 1.9954128440366974e-05,
"loss": 1.021,
"step": 870
},
{
"epoch": 0.1,
"grad_norm": 0.1704547576709166,
"learning_rate": 1.999999277982412e-05,
"loss": 1.0329,
"step": 875
},
{
"epoch": 0.1,
"grad_norm": 0.22769274744165113,
"learning_rate": 1.9999948656564835e-05,
"loss": 1.0332,
"step": 880
},
{
"epoch": 0.1,
"grad_norm": 0.1701717718459902,
"learning_rate": 1.999986442143186e-05,
"loss": 1.0386,
"step": 885
},
{
"epoch": 0.1,
"grad_norm": 0.17535479870327772,
"learning_rate": 1.9999740074763082e-05,
"loss": 1.0334,
"step": 890
},
{
"epoch": 0.1,
"grad_norm": 0.14552745937001002,
"learning_rate": 1.9999575617057277e-05,
"loss": 0.9997,
"step": 895
},
{
"epoch": 0.1,
"grad_norm": 0.17064224798244215,
"learning_rate": 1.999937104897412e-05,
"loss": 0.9442,
"step": 900
},
{
"epoch": 0.1,
"grad_norm": 0.18281746237250524,
"learning_rate": 1.999912637133418e-05,
"loss": 0.9834,
"step": 905
},
{
"epoch": 0.1,
"grad_norm": 0.1841506941484732,
"learning_rate": 1.9998841585118906e-05,
"loss": 1.0572,
"step": 910
},
{
"epoch": 0.1,
"grad_norm": 0.19474853136561926,
"learning_rate": 1.9998516691470632e-05,
"loss": 1.0379,
"step": 915
},
{
"epoch": 0.11,
"grad_norm": 0.21059881407319547,
"learning_rate": 1.9998151691692584e-05,
"loss": 1.0387,
"step": 920
},
{
"epoch": 0.11,
"grad_norm": 0.1670258379888007,
"learning_rate": 1.999774658724884e-05,
"loss": 1.0489,
"step": 925
},
{
"epoch": 0.11,
"grad_norm": 0.16905398909703961,
"learning_rate": 1.9997301379764368e-05,
"loss": 0.992,
"step": 930
},
{
"epoch": 0.11,
"grad_norm": 0.13635614170278254,
"learning_rate": 1.9996816071024985e-05,
"loss": 0.9992,
"step": 935
},
{
"epoch": 0.11,
"grad_norm": 2.703260581135581,
"learning_rate": 1.999629066297736e-05,
"loss": 1.0699,
"step": 940
},
{
"epoch": 0.11,
"grad_norm": 0.26182902457172097,
"learning_rate": 1.999572515772902e-05,
"loss": 0.9762,
"step": 945
},
{
"epoch": 0.11,
"grad_norm": 0.17192372945084772,
"learning_rate": 1.9995119557548324e-05,
"loss": 1.0442,
"step": 950
},
{
"epoch": 0.11,
"grad_norm": 0.1865724307378294,
"learning_rate": 1.9994473864864457e-05,
"loss": 1.0088,
"step": 955
},
{
"epoch": 0.11,
"grad_norm": 0.17975701975825453,
"learning_rate": 1.999378808226743e-05,
"loss": 1.0165,
"step": 960
},
{
"epoch": 0.11,
"grad_norm": 0.317797386442062,
"learning_rate": 1.9993062212508054e-05,
"loss": 1.0054,
"step": 965
},
{
"epoch": 0.11,
"grad_norm": 0.19329608403477153,
"learning_rate": 1.999229625849795e-05,
"loss": 1.0394,
"step": 970
},
{
"epoch": 0.11,
"grad_norm": 0.15735842168914,
"learning_rate": 1.9991490223309516e-05,
"loss": 1.025,
"step": 975
},
{
"epoch": 0.11,
"grad_norm": 0.1748362860806812,
"learning_rate": 1.9990644110175925e-05,
"loss": 0.9968,
"step": 980
},
{
"epoch": 0.11,
"grad_norm": 0.17972499772143366,
"learning_rate": 1.9989757922491117e-05,
"loss": 1.0228,
"step": 985
},
{
"epoch": 0.11,
"grad_norm": 0.18104605026678153,
"learning_rate": 1.9988831663809772e-05,
"loss": 1.0362,
"step": 990
},
{
"epoch": 0.11,
"grad_norm": 0.15668822259442713,
"learning_rate": 1.9987865337847307e-05,
"loss": 1.0287,
"step": 995
},
{
"epoch": 0.11,
"grad_norm": 0.1517176879683345,
"learning_rate": 1.9986858948479856e-05,
"loss": 1.0251,
"step": 1000
},
{
"epoch": 0.12,
"grad_norm": 0.18320259716049186,
"learning_rate": 1.9985812499744256e-05,
"loss": 1.0194,
"step": 1005
},
{
"epoch": 0.12,
"grad_norm": 0.16228421127114068,
"learning_rate": 1.9984725995838032e-05,
"loss": 1.0141,
"step": 1010
},
{
"epoch": 0.12,
"grad_norm": 0.1718139362774404,
"learning_rate": 1.9983599441119374e-05,
"loss": 0.9596,
"step": 1015
},
{
"epoch": 0.12,
"grad_norm": 0.1672412854326014,
"learning_rate": 1.998243284010713e-05,
"loss": 1.0155,
"step": 1020
},
{
"epoch": 0.12,
"grad_norm": 0.148581441029014,
"learning_rate": 1.998122619748078e-05,
"loss": 0.9819,
"step": 1025
},
{
"epoch": 0.12,
"grad_norm": 0.18328653714685153,
"learning_rate": 1.9979979518080415e-05,
"loss": 1.0866,
"step": 1030
},
{
"epoch": 0.12,
"grad_norm": 0.16516484177156834,
"learning_rate": 1.9978692806906734e-05,
"loss": 0.9823,
"step": 1035
},
{
"epoch": 0.12,
"grad_norm": 0.15309919421550783,
"learning_rate": 1.997736606912099e-05,
"loss": 1.0508,
"step": 1040
},
{
"epoch": 0.12,
"grad_norm": 0.25099204062845915,
"learning_rate": 1.997599931004501e-05,
"loss": 0.9595,
"step": 1045
},
{
"epoch": 0.12,
"grad_norm": 0.49752501571390245,
"learning_rate": 1.997459253516115e-05,
"loss": 0.9753,
"step": 1050
},
{
"epoch": 0.12,
"grad_norm": 0.19754757268959017,
"learning_rate": 1.997314575011228e-05,
"loss": 0.9879,
"step": 1055
},
{
"epoch": 0.12,
"grad_norm": 0.17382149538888833,
"learning_rate": 1.997165896070175e-05,
"loss": 1.0119,
"step": 1060
},
{
"epoch": 0.12,
"grad_norm": 0.16792496720481118,
"learning_rate": 1.997013217289338e-05,
"loss": 1.0104,
"step": 1065
},
{
"epoch": 0.12,
"grad_norm": 0.15504563627835147,
"learning_rate": 1.996856539281144e-05,
"loss": 0.9855,
"step": 1070
},
{
"epoch": 0.12,
"grad_norm": 0.16341935472061658,
"learning_rate": 1.996695862674061e-05,
"loss": 1.0005,
"step": 1075
},
{
"epoch": 0.12,
"grad_norm": 0.29586292222201216,
"learning_rate": 1.9965311881125956e-05,
"loss": 0.9794,
"step": 1080
},
{
"epoch": 0.12,
"grad_norm": 0.15812114141525616,
"learning_rate": 1.9963625162572926e-05,
"loss": 1.022,
"step": 1085
},
{
"epoch": 0.13,
"grad_norm": 0.1532613131795279,
"learning_rate": 1.9961898477847292e-05,
"loss": 1.0237,
"step": 1090
},
{
"epoch": 0.13,
"grad_norm": 0.17366131896129713,
"learning_rate": 1.9960131833875148e-05,
"loss": 1.0035,
"step": 1095
},
{
"epoch": 0.13,
"grad_norm": 0.14807894096780344,
"learning_rate": 1.995832523774287e-05,
"loss": 1.0186,
"step": 1100
},
{
"epoch": 0.13,
"grad_norm": 0.13930431875920762,
"learning_rate": 1.995647869669709e-05,
"loss": 1.0088,
"step": 1105
},
{
"epoch": 0.13,
"grad_norm": 0.14458634083885646,
"learning_rate": 1.9954592218144674e-05,
"loss": 0.9785,
"step": 1110
},
{
"epoch": 0.13,
"grad_norm": 0.15081716553182822,
"learning_rate": 1.995266580965267e-05,
"loss": 0.9963,
"step": 1115
},
{
"epoch": 0.13,
"grad_norm": 0.36133029712930903,
"learning_rate": 1.995069947894831e-05,
"loss": 1.0193,
"step": 1120
},
{
"epoch": 0.13,
"grad_norm": 0.1699891285865461,
"learning_rate": 1.994869323391895e-05,
"loss": 1.0107,
"step": 1125
},
{
"epoch": 0.13,
"grad_norm": 0.1519273271943623,
"learning_rate": 1.9946647082612064e-05,
"loss": 0.9574,
"step": 1130
},
{
"epoch": 0.13,
"grad_norm": 0.15321248037109864,
"learning_rate": 1.9944561033235186e-05,
"loss": 1.0416,
"step": 1135
},
{
"epoch": 0.13,
"grad_norm": 0.17205314809163413,
"learning_rate": 1.9942435094155894e-05,
"loss": 1.0332,
"step": 1140
},
{
"epoch": 0.13,
"grad_norm": 0.17530609764948807,
"learning_rate": 1.9940269273901773e-05,
"loss": 1.0099,
"step": 1145
},
{
"epoch": 0.13,
"grad_norm": 0.16004122876513682,
"learning_rate": 1.993806358116037e-05,
"loss": 0.9899,
"step": 1150
},
{
"epoch": 0.13,
"grad_norm": 0.17242013279867477,
"learning_rate": 1.9935818024779192e-05,
"loss": 1.0479,
"step": 1155
},
{
"epoch": 0.13,
"grad_norm": 0.18103873207408538,
"learning_rate": 1.9933532613765622e-05,
"loss": 1.0216,
"step": 1160
},
{
"epoch": 0.13,
"grad_norm": 0.13873945080470682,
"learning_rate": 1.9931207357286922e-05,
"loss": 0.9813,
"step": 1165
},
{
"epoch": 0.13,
"grad_norm": 0.1611397140301265,
"learning_rate": 1.9928842264670182e-05,
"loss": 1.0455,
"step": 1170
},
{
"epoch": 0.13,
"grad_norm": 0.18025318650189281,
"learning_rate": 1.9926437345402273e-05,
"loss": 1.0371,
"step": 1175
},
{
"epoch": 0.14,
"grad_norm": 0.15588451317489496,
"learning_rate": 1.992399260912983e-05,
"loss": 0.9489,
"step": 1180
},
{
"epoch": 0.14,
"grad_norm": 0.15553017124682822,
"learning_rate": 1.9921508065659206e-05,
"loss": 0.9928,
"step": 1185
},
{
"epoch": 0.14,
"grad_norm": 0.3017321072545083,
"learning_rate": 1.991898372495641e-05,
"loss": 1.0632,
"step": 1190
},
{
"epoch": 0.14,
"grad_norm": 0.16055353752876278,
"learning_rate": 1.9916419597147104e-05,
"loss": 0.9923,
"step": 1195
},
{
"epoch": 0.14,
"grad_norm": 0.16411490826782263,
"learning_rate": 1.9913815692516536e-05,
"loss": 1.0081,
"step": 1200
},
{
"epoch": 0.14,
"grad_norm": 0.16981445105849077,
"learning_rate": 1.991117202150951e-05,
"loss": 1.0117,
"step": 1205
},
{
"epoch": 0.14,
"grad_norm": 0.15474347412644857,
"learning_rate": 1.9908488594730338e-05,
"loss": 1.0282,
"step": 1210
},
{
"epoch": 0.14,
"grad_norm": 0.16911514436546357,
"learning_rate": 1.9905765422942808e-05,
"loss": 0.9762,
"step": 1215
},
{
"epoch": 0.14,
"grad_norm": 0.19230875878766188,
"learning_rate": 1.9903002517070122e-05,
"loss": 1.0514,
"step": 1220
},
{
"epoch": 0.14,
"grad_norm": 0.18979647273713313,
"learning_rate": 1.9900199888194876e-05,
"loss": 1.047,
"step": 1225
},
{
"epoch": 0.14,
"grad_norm": 0.16363591400242028,
"learning_rate": 1.989735754755899e-05,
"loss": 1.0483,
"step": 1230
},
{
"epoch": 0.14,
"grad_norm": 0.17857039455058185,
"learning_rate": 1.989447550656369e-05,
"loss": 1.033,
"step": 1235
},
{
"epoch": 0.14,
"grad_norm": 0.16882498969210477,
"learning_rate": 1.989155377676944e-05,
"loss": 1.0346,
"step": 1240
},
{
"epoch": 0.14,
"grad_norm": 0.16162776322327543,
"learning_rate": 1.98885923698959e-05,
"loss": 1.0342,
"step": 1245
},
{
"epoch": 0.14,
"grad_norm": 0.17696332477802224,
"learning_rate": 1.98855912978219e-05,
"loss": 1.0012,
"step": 1250
},
{
"epoch": 0.14,
"grad_norm": 0.18798321244194857,
"learning_rate": 1.9882550572585356e-05,
"loss": 0.9968,
"step": 1255
},
{
"epoch": 0.14,
"grad_norm": 0.18443421514645758,
"learning_rate": 1.987947020638325e-05,
"loss": 1.0335,
"step": 1260
},
{
"epoch": 0.15,
"grad_norm": 0.18265481260663313,
"learning_rate": 1.9876350211571576e-05,
"loss": 0.9881,
"step": 1265
},
{
"epoch": 0.15,
"grad_norm": 0.1648728325099089,
"learning_rate": 1.9873190600665277e-05,
"loss": 1.0304,
"step": 1270
},
{
"epoch": 0.15,
"grad_norm": 0.16683576689593385,
"learning_rate": 1.986999138633821e-05,
"loss": 1.021,
"step": 1275
},
{
"epoch": 0.15,
"grad_norm": 0.181777159003462,
"learning_rate": 1.9866752581423093e-05,
"loss": 1.0262,
"step": 1280
},
{
"epoch": 0.15,
"grad_norm": 0.16301257888715454,
"learning_rate": 1.9863474198911443e-05,
"loss": 1.0393,
"step": 1285
},
{
"epoch": 0.15,
"grad_norm": 0.1722381699184831,
"learning_rate": 1.9860156251953534e-05,
"loss": 1.0454,
"step": 1290
},
{
"epoch": 0.15,
"grad_norm": 0.16074956631460535,
"learning_rate": 1.9856798753858344e-05,
"loss": 1.0082,
"step": 1295
},
{
"epoch": 0.15,
"grad_norm": 0.1705200151971908,
"learning_rate": 1.985340171809349e-05,
"loss": 0.9769,
"step": 1300
},
{
"epoch": 0.15,
"grad_norm": 0.1879218619111294,
"learning_rate": 1.9849965158285195e-05,
"loss": 1.0693,
"step": 1305
},
{
"epoch": 0.15,
"grad_norm": 0.1866441671946385,
"learning_rate": 1.9846489088218214e-05,
"loss": 1.0636,
"step": 1310
},
{
"epoch": 0.15,
"grad_norm": 0.17288166932326893,
"learning_rate": 1.984297352183578e-05,
"loss": 0.9863,
"step": 1315
},
{
"epoch": 0.15,
"grad_norm": 0.20106859629867654,
"learning_rate": 1.9839418473239572e-05,
"loss": 1.0311,
"step": 1320
},
{
"epoch": 0.15,
"grad_norm": 0.1640151040347238,
"learning_rate": 1.9835823956689623e-05,
"loss": 1.0085,
"step": 1325
},
{
"epoch": 0.15,
"grad_norm": 0.1704486396699564,
"learning_rate": 1.9832189986604288e-05,
"loss": 1.0535,
"step": 1330
},
{
"epoch": 0.15,
"grad_norm": 0.18631768048436156,
"learning_rate": 1.9828516577560175e-05,
"loss": 1.0077,
"step": 1335
},
{
"epoch": 0.15,
"grad_norm": 0.15146162186612072,
"learning_rate": 1.9824803744292095e-05,
"loss": 0.9499,
"step": 1340
},
{
"epoch": 0.15,
"grad_norm": 0.18224384077718625,
"learning_rate": 1.9821051501692988e-05,
"loss": 1.0013,
"step": 1345
},
{
"epoch": 0.15,
"grad_norm": 0.15427348075639968,
"learning_rate": 1.9817259864813885e-05,
"loss": 1.0289,
"step": 1350
},
{
"epoch": 0.16,
"grad_norm": 0.15816101251833992,
"learning_rate": 1.9813428848863828e-05,
"loss": 1.0633,
"step": 1355
},
{
"epoch": 0.16,
"grad_norm": 0.16767098786447587,
"learning_rate": 1.980955846920982e-05,
"loss": 0.9969,
"step": 1360
},
{
"epoch": 0.16,
"grad_norm": 0.17580089577052765,
"learning_rate": 1.9805648741376758e-05,
"loss": 0.9727,
"step": 1365
},
{
"epoch": 0.16,
"grad_norm": 0.16262644921000052,
"learning_rate": 1.9801699681047373e-05,
"loss": 1.0056,
"step": 1370
},
{
"epoch": 0.16,
"grad_norm": 0.16318984383721474,
"learning_rate": 1.9797711304062167e-05,
"loss": 0.9981,
"step": 1375
},
{
"epoch": 0.16,
"grad_norm": 0.16043651674590212,
"learning_rate": 1.979368362641935e-05,
"loss": 1.0358,
"step": 1380
},
{
"epoch": 0.16,
"grad_norm": 0.16203539501036138,
"learning_rate": 1.9789616664274775e-05,
"loss": 0.9865,
"step": 1385
},
{
"epoch": 0.16,
"grad_norm": 0.16642870994892164,
"learning_rate": 1.9785510433941873e-05,
"loss": 1.0253,
"step": 1390
},
{
"epoch": 0.16,
"grad_norm": 0.18476932567506452,
"learning_rate": 1.978136495189159e-05,
"loss": 1.0716,
"step": 1395
},
{
"epoch": 0.16,
"grad_norm": 0.16636510089395223,
"learning_rate": 1.9777180234752308e-05,
"loss": 1.0429,
"step": 1400
},
{
"epoch": 0.16,
"grad_norm": 0.18527366416671937,
"learning_rate": 1.977295629930981e-05,
"loss": 1.01,
"step": 1405
},
{
"epoch": 0.16,
"grad_norm": 0.1652333907473212,
"learning_rate": 1.9768693162507167e-05,
"loss": 1.02,
"step": 1410
},
{
"epoch": 0.16,
"grad_norm": 0.1583305403865818,
"learning_rate": 1.9764390841444718e-05,
"loss": 1.0283,
"step": 1415
},
{
"epoch": 0.16,
"grad_norm": 0.17629915987600467,
"learning_rate": 1.976004935337996e-05,
"loss": 1.0091,
"step": 1420
},
{
"epoch": 0.16,
"grad_norm": 0.15846275403375087,
"learning_rate": 1.9755668715727513e-05,
"loss": 1.0134,
"step": 1425
},
{
"epoch": 0.16,
"grad_norm": 0.1666057528517698,
"learning_rate": 1.9751248946059016e-05,
"loss": 0.9956,
"step": 1430
},
{
"epoch": 0.16,
"grad_norm": 0.1648120690081439,
"learning_rate": 1.9746790062103088e-05,
"loss": 0.9876,
"step": 1435
},
{
"epoch": 0.17,
"grad_norm": 0.1492793373601537,
"learning_rate": 1.9742292081745243e-05,
"loss": 1.0194,
"step": 1440
},
{
"epoch": 0.17,
"grad_norm": 0.19313172775251475,
"learning_rate": 1.973775502302781e-05,
"loss": 1.0271,
"step": 1445
},
{
"epoch": 0.17,
"grad_norm": 0.16626743321351461,
"learning_rate": 1.973317890414988e-05,
"loss": 1.0336,
"step": 1450
},
{
"epoch": 0.17,
"grad_norm": 0.17694906609894173,
"learning_rate": 1.9728563743467216e-05,
"loss": 0.9621,
"step": 1455
},
{
"epoch": 0.17,
"grad_norm": 0.17762191011953987,
"learning_rate": 1.9723909559492184e-05,
"loss": 1.0002,
"step": 1460
},
{
"epoch": 0.17,
"grad_norm": 0.16623664041768285,
"learning_rate": 1.971921637089369e-05,
"loss": 0.9959,
"step": 1465
},
{
"epoch": 0.17,
"grad_norm": 0.18391243666760815,
"learning_rate": 1.9714484196497087e-05,
"loss": 1.0118,
"step": 1470
},
{
"epoch": 0.17,
"grad_norm": 0.18968659798223544,
"learning_rate": 1.9709713055284103e-05,
"loss": 1.0326,
"step": 1475
},
{
"epoch": 0.17,
"grad_norm": 0.17316773945383043,
"learning_rate": 1.970490296639279e-05,
"loss": 1.0567,
"step": 1480
},
{
"epoch": 0.17,
"grad_norm": 0.156967480156991,
"learning_rate": 1.9700053949117408e-05,
"loss": 0.9955,
"step": 1485
},
{
"epoch": 0.17,
"grad_norm": 0.17183142790884795,
"learning_rate": 1.9695166022908373e-05,
"loss": 0.9868,
"step": 1490
},
{
"epoch": 0.17,
"grad_norm": 0.18649820756368488,
"learning_rate": 1.9690239207372178e-05,
"loss": 1.0486,
"step": 1495
},
{
"epoch": 0.17,
"grad_norm": 0.14950108706982165,
"learning_rate": 1.9685273522271302e-05,
"loss": 0.9777,
"step": 1500
},
{
"epoch": 0.17,
"grad_norm": 0.179697954696151,
"learning_rate": 1.968026898752414e-05,
"loss": 1.0262,
"step": 1505
},
{
"epoch": 0.17,
"grad_norm": 0.16825240525497248,
"learning_rate": 1.9675225623204924e-05,
"loss": 1.0433,
"step": 1510
},
{
"epoch": 0.17,
"grad_norm": 0.17966992138752028,
"learning_rate": 1.9670143449543632e-05,
"loss": 1.0401,
"step": 1515
},
{
"epoch": 0.17,
"grad_norm": 0.17979837581532043,
"learning_rate": 1.9665022486925925e-05,
"loss": 0.9953,
"step": 1520
},
{
"epoch": 0.17,
"grad_norm": 0.16488882122128387,
"learning_rate": 1.965986275589305e-05,
"loss": 0.9961,
"step": 1525
},
{
"epoch": 0.18,
"grad_norm": 0.1906647429301807,
"learning_rate": 1.9654664277141755e-05,
"loss": 1.0081,
"step": 1530
},
{
"epoch": 0.18,
"grad_norm": 0.4402544866305577,
"learning_rate": 1.9649427071524224e-05,
"loss": 0.9803,
"step": 1535
},
{
"epoch": 0.18,
"grad_norm": 0.1959692615799122,
"learning_rate": 1.9644151160047973e-05,
"loss": 0.9761,
"step": 1540
},
{
"epoch": 0.18,
"grad_norm": 0.16802768865323667,
"learning_rate": 1.9638836563875787e-05,
"loss": 0.9979,
"step": 1545
},
{
"epoch": 0.18,
"grad_norm": 0.1761895322457327,
"learning_rate": 1.9633483304325614e-05,
"loss": 1.0368,
"step": 1550
},
{
"epoch": 0.18,
"grad_norm": 0.1770996720292896,
"learning_rate": 1.9628091402870487e-05,
"loss": 0.976,
"step": 1555
},
{
"epoch": 0.18,
"grad_norm": 0.19204393103893866,
"learning_rate": 1.9622660881138453e-05,
"loss": 0.9966,
"step": 1560
},
{
"epoch": 0.18,
"grad_norm": 0.16944527249439464,
"learning_rate": 1.9617191760912452e-05,
"loss": 0.9678,
"step": 1565
},
{
"epoch": 0.18,
"grad_norm": 0.18890925902960784,
"learning_rate": 1.961168406413027e-05,
"loss": 0.9683,
"step": 1570
},
{
"epoch": 0.18,
"grad_norm": 0.23576102960998133,
"learning_rate": 1.960613781288443e-05,
"loss": 1.0254,
"step": 1575
},
{
"epoch": 0.18,
"grad_norm": 0.1881690601344743,
"learning_rate": 1.9600553029422087e-05,
"loss": 0.9578,
"step": 1580
},
{
"epoch": 0.18,
"grad_norm": 0.15080211694063853,
"learning_rate": 1.9594929736144978e-05,
"loss": 0.9641,
"step": 1585
},
{
"epoch": 0.18,
"grad_norm": 0.1704001648166871,
"learning_rate": 1.958926795560929e-05,
"loss": 1.0022,
"step": 1590
},
{
"epoch": 0.18,
"grad_norm": 0.17344399870636193,
"learning_rate": 1.958356771052562e-05,
"loss": 1.0149,
"step": 1595
},
{
"epoch": 0.18,
"grad_norm": 0.1988678574168928,
"learning_rate": 1.9577829023758824e-05,
"loss": 1.0262,
"step": 1600
},
{
"epoch": 0.18,
"grad_norm": 0.18690183074427064,
"learning_rate": 1.957205191832797e-05,
"loss": 0.9753,
"step": 1605
},
{
"epoch": 0.18,
"grad_norm": 0.18745656927992985,
"learning_rate": 1.956623641740623e-05,
"loss": 1.0509,
"step": 1610
},
{
"epoch": 0.19,
"grad_norm": 0.16789860937046228,
"learning_rate": 1.9560382544320788e-05,
"loss": 0.9925,
"step": 1615
},
{
"epoch": 0.19,
"grad_norm": 0.1630548916133644,
"learning_rate": 1.9554490322552745e-05,
"loss": 1.0031,
"step": 1620
},
{
"epoch": 0.19,
"grad_norm": 0.1788079232484033,
"learning_rate": 1.954855977573703e-05,
"loss": 0.9485,
"step": 1625
},
{
"epoch": 0.19,
"grad_norm": 0.17754176497715887,
"learning_rate": 1.95425909276623e-05,
"loss": 1.0117,
"step": 1630
},
{
"epoch": 0.19,
"grad_norm": 0.16400210345357913,
"learning_rate": 1.9536583802270843e-05,
"loss": 1.0005,
"step": 1635
},
{
"epoch": 0.19,
"grad_norm": 0.182196746960171,
"learning_rate": 1.953053842365849e-05,
"loss": 0.9954,
"step": 1640
},
{
"epoch": 0.19,
"grad_norm": 0.1624184353486166,
"learning_rate": 1.9524454816074513e-05,
"loss": 1.0401,
"step": 1645
},
{
"epoch": 0.19,
"grad_norm": 0.2080276508229457,
"learning_rate": 1.9518333003921525e-05,
"loss": 0.9678,
"step": 1650
},
{
"epoch": 0.19,
"grad_norm": 0.17831130847417112,
"learning_rate": 1.951217301175539e-05,
"loss": 0.9716,
"step": 1655
},
{
"epoch": 0.19,
"grad_norm": 0.1794396364087045,
"learning_rate": 1.9505974864285113e-05,
"loss": 1.0202,
"step": 1660
},
{
"epoch": 0.19,
"grad_norm": 0.16983858604812369,
"learning_rate": 1.9499738586372757e-05,
"loss": 0.9313,
"step": 1665
},
{
"epoch": 0.19,
"grad_norm": 0.17062851478617005,
"learning_rate": 1.949346420303333e-05,
"loss": 0.9958,
"step": 1670
},
{
"epoch": 0.19,
"grad_norm": 0.16135946964786269,
"learning_rate": 1.9487151739434685e-05,
"loss": 1.0179,
"step": 1675
},
{
"epoch": 0.19,
"grad_norm": 0.16148753080872155,
"learning_rate": 1.948080122089743e-05,
"loss": 1.0161,
"step": 1680
},
{
"epoch": 0.19,
"grad_norm": 0.1863495955861023,
"learning_rate": 1.947441267289481e-05,
"loss": 1.0121,
"step": 1685
},
{
"epoch": 0.19,
"grad_norm": 0.1568702240498156,
"learning_rate": 1.9467986121052625e-05,
"loss": 1.0135,
"step": 1690
},
{
"epoch": 0.19,
"grad_norm": 0.16939868435805067,
"learning_rate": 1.946152159114911e-05,
"loss": 1.0134,
"step": 1695
},
{
"epoch": 0.2,
"grad_norm": 0.1888153742033124,
"learning_rate": 1.9455019109114833e-05,
"loss": 0.993,
"step": 1700
},
{
"epoch": 0.2,
"grad_norm": 0.1612522324113172,
"learning_rate": 1.944847870103261e-05,
"loss": 0.9789,
"step": 1705
},
{
"epoch": 0.2,
"grad_norm": 0.1696344839941852,
"learning_rate": 1.9441900393137377e-05,
"loss": 0.9886,
"step": 1710
},
{
"epoch": 0.2,
"grad_norm": 0.1706608292781839,
"learning_rate": 1.943528421181609e-05,
"loss": 1.0116,
"step": 1715
},
{
"epoch": 0.2,
"grad_norm": 0.14622565349238575,
"learning_rate": 1.942863018360763e-05,
"loss": 1.0388,
"step": 1720
},
{
"epoch": 0.2,
"grad_norm": 0.18560643944729288,
"learning_rate": 1.942193833520269e-05,
"loss": 1.0001,
"step": 1725
},
{
"epoch": 0.2,
"grad_norm": 0.1516103497323242,
"learning_rate": 1.9415208693443667e-05,
"loss": 1.0192,
"step": 1730
},
{
"epoch": 0.2,
"grad_norm": 0.15495793766788726,
"learning_rate": 1.9408441285324547e-05,
"loss": 0.9681,
"step": 1735
},
{
"epoch": 0.2,
"grad_norm": 0.18655169371813152,
"learning_rate": 1.9401636137990814e-05,
"loss": 1.0073,
"step": 1740
},
{
"epoch": 0.2,
"grad_norm": 0.20015890004640238,
"learning_rate": 1.9394793278739336e-05,
"loss": 0.9897,
"step": 1745
},
{
"epoch": 0.2,
"grad_norm": 0.17011708293249655,
"learning_rate": 1.9387912735018235e-05,
"loss": 1.0214,
"step": 1750
},
{
"epoch": 0.2,
"grad_norm": 0.15667326284470906,
"learning_rate": 1.938099453442681e-05,
"loss": 1.0778,
"step": 1755
},
{
"epoch": 0.2,
"grad_norm": 0.17644565627531678,
"learning_rate": 1.9374038704715398e-05,
"loss": 1.0059,
"step": 1760
},
{
"epoch": 0.2,
"grad_norm": 0.18075237078257975,
"learning_rate": 1.9367045273785273e-05,
"loss": 1.0193,
"step": 1765
},
{
"epoch": 0.2,
"grad_norm": 0.17536235973854006,
"learning_rate": 1.9360014269688557e-05,
"loss": 0.9405,
"step": 1770
},
{
"epoch": 0.2,
"grad_norm": 0.17565082319336725,
"learning_rate": 1.9352945720628054e-05,
"loss": 0.9894,
"step": 1775
},
{
"epoch": 0.2,
"grad_norm": 0.19684546961007965,
"learning_rate": 1.9345839654957195e-05,
"loss": 1.0258,
"step": 1780
},
{
"epoch": 0.2,
"grad_norm": 0.16782416047849205,
"learning_rate": 1.933869610117988e-05,
"loss": 0.9814,
"step": 1785
},
{
"epoch": 0.21,
"grad_norm": 0.18954779187335546,
"learning_rate": 1.9331515087950393e-05,
"loss": 1.0019,
"step": 1790
},
{
"epoch": 0.21,
"grad_norm": 0.17680854715817543,
"learning_rate": 1.9324296644073265e-05,
"loss": 0.9957,
"step": 1795
},
{
"epoch": 0.21,
"grad_norm": 0.22585189987350146,
"learning_rate": 1.9317040798503185e-05,
"loss": 1.0014,
"step": 1800
},
{
"epoch": 0.21,
"grad_norm": 0.1608408827724262,
"learning_rate": 1.9309747580344852e-05,
"loss": 0.9544,
"step": 1805
},
{
"epoch": 0.21,
"grad_norm": 0.15567413432263022,
"learning_rate": 1.9302417018852883e-05,
"loss": 1.0116,
"step": 1810
},
{
"epoch": 0.21,
"grad_norm": 0.16668473049055743,
"learning_rate": 1.9295049143431685e-05,
"loss": 1.0136,
"step": 1815
},
{
"epoch": 0.21,
"grad_norm": 0.16974348065541098,
"learning_rate": 1.928764398363534e-05,
"loss": 1.023,
"step": 1820
},
{
"epoch": 0.21,
"grad_norm": 0.16585078318241206,
"learning_rate": 1.9280201569167477e-05,
"loss": 1.0539,
"step": 1825
},
{
"epoch": 0.21,
"grad_norm": 0.17584872533368381,
"learning_rate": 1.927272192988118e-05,
"loss": 0.9826,
"step": 1830
},
{
"epoch": 0.21,
"grad_norm": 0.15495303750722209,
"learning_rate": 1.9265205095778827e-05,
"loss": 0.9746,
"step": 1835
},
{
"epoch": 0.21,
"grad_norm": 0.17381457992842353,
"learning_rate": 1.9257651097012004e-05,
"loss": 1.0927,
"step": 1840
},
{
"epoch": 0.21,
"grad_norm": 0.1689344548430039,
"learning_rate": 1.9250059963881374e-05,
"loss": 0.9668,
"step": 1845
},
{
"epoch": 0.21,
"grad_norm": 0.1889908242464221,
"learning_rate": 1.924243172683655e-05,
"loss": 0.9855,
"step": 1850
},
{
"epoch": 0.21,
"grad_norm": 0.14602014895442134,
"learning_rate": 1.9234766416475984e-05,
"loss": 1.0022,
"step": 1855
},
{
"epoch": 0.21,
"grad_norm": 0.15905538292384974,
"learning_rate": 1.922706406354682e-05,
"loss": 0.9958,
"step": 1860
},
{
"epoch": 0.21,
"grad_norm": 0.16151025920894505,
"learning_rate": 1.9219324698944798e-05,
"loss": 0.9952,
"step": 1865
},
{
"epoch": 0.21,
"grad_norm": 0.15813751169141482,
"learning_rate": 1.9211548353714123e-05,
"loss": 1.0164,
"step": 1870
},
{
"epoch": 0.22,
"grad_norm": 0.16998318288399103,
"learning_rate": 1.920373505904733e-05,
"loss": 0.9935,
"step": 1875
},
{
"epoch": 0.22,
"grad_norm": 0.17689421597873753,
"learning_rate": 1.9195884846285173e-05,
"loss": 1.0193,
"step": 1880
},
{
"epoch": 0.22,
"grad_norm": 0.16142530130024355,
"learning_rate": 1.9187997746916484e-05,
"loss": 1.031,
"step": 1885
},
{
"epoch": 0.22,
"grad_norm": 0.1536941041734269,
"learning_rate": 1.918007379257806e-05,
"loss": 1.0084,
"step": 1890
},
{
"epoch": 0.22,
"grad_norm": 0.16593197135284424,
"learning_rate": 1.917211301505453e-05,
"loss": 1.0189,
"step": 1895
},
{
"epoch": 0.22,
"grad_norm": 0.17967794093502043,
"learning_rate": 1.916411544627823e-05,
"loss": 1.0166,
"step": 1900
},
{
"epoch": 0.22,
"grad_norm": 0.1708948023534157,
"learning_rate": 1.915608111832907e-05,
"loss": 0.9977,
"step": 1905
},
{
"epoch": 0.22,
"grad_norm": 0.15035831029499927,
"learning_rate": 1.9148010063434407e-05,
"loss": 0.9896,
"step": 1910
},
{
"epoch": 0.22,
"grad_norm": 0.14735227792294917,
"learning_rate": 1.9139902313968925e-05,
"loss": 0.9714,
"step": 1915
},
{
"epoch": 0.22,
"grad_norm": 0.19017803321206933,
"learning_rate": 1.9131757902454494e-05,
"loss": 0.9477,
"step": 1920
},
{
"epoch": 0.22,
"grad_norm": 0.16063726299894956,
"learning_rate": 1.912357686156004e-05,
"loss": 0.9955,
"step": 1925
},
{
"epoch": 0.22,
"grad_norm": 0.17903687187181608,
"learning_rate": 1.9115359224101416e-05,
"loss": 1.013,
"step": 1930
},
{
"epoch": 0.22,
"grad_norm": 0.14071215322148392,
"learning_rate": 1.9107105023041277e-05,
"loss": 0.9695,
"step": 1935
},
{
"epoch": 0.22,
"grad_norm": 0.16713899113608624,
"learning_rate": 1.9098814291488946e-05,
"loss": 1.0394,
"step": 1940
},
{
"epoch": 0.22,
"grad_norm": 0.1644478630668693,
"learning_rate": 1.9090487062700266e-05,
"loss": 0.9696,
"step": 1945
},
{
"epoch": 0.22,
"grad_norm": 0.17602343549022248,
"learning_rate": 1.9082123370077485e-05,
"loss": 0.9823,
"step": 1950
},
{
"epoch": 0.22,
"grad_norm": 0.1797147838933718,
"learning_rate": 1.907372324716911e-05,
"loss": 1.0023,
"step": 1955
},
{
"epoch": 0.22,
"grad_norm": 0.15932969084678994,
"learning_rate": 1.9065286727669783e-05,
"loss": 1.0083,
"step": 1960
},
{
"epoch": 0.23,
"grad_norm": 0.1904108903683644,
"learning_rate": 1.905681384542014e-05,
"loss": 0.9639,
"step": 1965
},
{
"epoch": 0.23,
"grad_norm": 0.15360440299536685,
"learning_rate": 1.9048304634406674e-05,
"loss": 0.9872,
"step": 1970
},
{
"epoch": 0.23,
"grad_norm": 0.16119796002290548,
"learning_rate": 1.9039759128761596e-05,
"loss": 1.0332,
"step": 1975
},
{
"epoch": 0.23,
"grad_norm": 0.1687187914499054,
"learning_rate": 1.9031177362762712e-05,
"loss": 1.0073,
"step": 1980
},
{
"epoch": 0.23,
"grad_norm": 0.1654510620373757,
"learning_rate": 1.9022559370833268e-05,
"loss": 1.0008,
"step": 1985
},
{
"epoch": 0.23,
"grad_norm": 0.18167576186450726,
"learning_rate": 1.9013905187541817e-05,
"loss": 0.9978,
"step": 1990
},
{
"epoch": 0.23,
"grad_norm": 0.1592099865887557,
"learning_rate": 1.9005214847602093e-05,
"loss": 0.9749,
"step": 1995
},
{
"epoch": 0.23,
"grad_norm": 0.16540521212971535,
"learning_rate": 1.8996488385872856e-05,
"loss": 0.9732,
"step": 2000
},
{
"epoch": 0.23,
"grad_norm": 0.16728562456497673,
"learning_rate": 1.898772583735776e-05,
"loss": 0.9458,
"step": 2005
},
{
"epoch": 0.23,
"grad_norm": 0.19351786177060903,
"learning_rate": 1.8978927237205203e-05,
"loss": 1.0002,
"step": 2010
},
{
"epoch": 0.23,
"grad_norm": 0.1734367064503199,
"learning_rate": 1.8970092620708208e-05,
"loss": 0.9745,
"step": 2015
},
{
"epoch": 0.23,
"grad_norm": 0.1662127252766583,
"learning_rate": 1.8961222023304256e-05,
"loss": 1.0338,
"step": 2020
},
{
"epoch": 0.23,
"grad_norm": 0.14746644761400565,
"learning_rate": 1.895231548057516e-05,
"loss": 0.9622,
"step": 2025
},
{
"epoch": 0.23,
"grad_norm": 0.2088364797368587,
"learning_rate": 1.894337302824691e-05,
"loss": 0.9672,
"step": 2030
},
{
"epoch": 0.23,
"grad_norm": 0.1769166332830946,
"learning_rate": 1.8934394702189544e-05,
"loss": 0.9946,
"step": 2035
},
{
"epoch": 0.23,
"grad_norm": 0.16604025013115048,
"learning_rate": 1.8925380538416996e-05,
"loss": 0.9994,
"step": 2040
},
{
"epoch": 0.23,
"grad_norm": 0.15917943943186041,
"learning_rate": 1.8916330573086953e-05,
"loss": 1.024,
"step": 2045
},
{
"epoch": 0.24,
"grad_norm": 0.20835021096712633,
"learning_rate": 1.8907244842500706e-05,
"loss": 1.0195,
"step": 2050
},
{
"epoch": 0.24,
"grad_norm": 0.2662809216468397,
"learning_rate": 1.8898123383103008e-05,
"loss": 0.9843,
"step": 2055
},
{
"epoch": 0.24,
"grad_norm": 0.17636330174565598,
"learning_rate": 1.8888966231481936e-05,
"loss": 0.9821,
"step": 2060
},
{
"epoch": 0.24,
"grad_norm": 0.1755120251861156,
"learning_rate": 1.8879773424368722e-05,
"loss": 1.0434,
"step": 2065
},
{
"epoch": 0.24,
"grad_norm": 0.17513015353824543,
"learning_rate": 1.887054499863763e-05,
"loss": 1.1095,
"step": 2070
},
{
"epoch": 0.24,
"grad_norm": 0.18151329873305128,
"learning_rate": 1.8861280991305796e-05,
"loss": 0.9658,
"step": 2075
},
{
"epoch": 0.24,
"grad_norm": 0.15725722749970544,
"learning_rate": 1.8851981439533084e-05,
"loss": 1.0239,
"step": 2080
},
{
"epoch": 0.24,
"grad_norm": 0.1520577506588838,
"learning_rate": 1.884264638062192e-05,
"loss": 0.9926,
"step": 2085
},
{
"epoch": 0.24,
"grad_norm": 0.1439001394194517,
"learning_rate": 1.8833275852017173e-05,
"loss": 0.9688,
"step": 2090
},
{
"epoch": 0.24,
"grad_norm": 0.21724787993488534,
"learning_rate": 1.882386989130598e-05,
"loss": 1.0776,
"step": 2095
},
{
"epoch": 0.24,
"grad_norm": 0.2702981749067786,
"learning_rate": 1.881442853621761e-05,
"loss": 1.1017,
"step": 2100
},
{
"epoch": 0.24,
"grad_norm": 0.20815680578120352,
"learning_rate": 1.8804951824623298e-05,
"loss": 0.9604,
"step": 2105
},
{
"epoch": 0.24,
"grad_norm": 0.16681978045265497,
"learning_rate": 1.8795439794536098e-05,
"loss": 1.015,
"step": 2110
},
{
"epoch": 0.24,
"grad_norm": 3.7700056050726727,
"learning_rate": 1.8785892484110758e-05,
"loss": 1.0759,
"step": 2115
},
{
"epoch": 0.24,
"grad_norm": 0.29730878825264534,
"learning_rate": 1.877630993164351e-05,
"loss": 0.9703,
"step": 2120
},
{
"epoch": 0.24,
"grad_norm": 0.22038976858581708,
"learning_rate": 1.876669217557197e-05,
"loss": 1.0473,
"step": 2125
},
{
"epoch": 0.24,
"grad_norm": 0.15753484513053292,
"learning_rate": 1.875703925447496e-05,
"loss": 0.9629,
"step": 2130
},
{
"epoch": 0.24,
"grad_norm": 0.17063863931268383,
"learning_rate": 1.874735120707235e-05,
"loss": 0.9813,
"step": 2135
},
{
"epoch": 0.25,
"grad_norm": 0.1493578241435406,
"learning_rate": 1.873762807222492e-05,
"loss": 1.0504,
"step": 2140
},
{
"epoch": 0.25,
"grad_norm": 0.14616344046469326,
"learning_rate": 1.8727869888934176e-05,
"loss": 0.95,
"step": 2145
},
{
"epoch": 0.25,
"grad_norm": 0.15313870950138356,
"learning_rate": 1.8718076696342226e-05,
"loss": 1.0336,
"step": 2150
},
{
"epoch": 0.25,
"grad_norm": 0.16012483360509658,
"learning_rate": 1.8708248533731604e-05,
"loss": 0.9477,
"step": 2155
},
{
"epoch": 0.25,
"grad_norm": 0.16206539468493553,
"learning_rate": 1.869838544052511e-05,
"loss": 1.0145,
"step": 2160
},
{
"epoch": 0.25,
"grad_norm": 0.1589176587228814,
"learning_rate": 1.8688487456285665e-05,
"loss": 1.0335,
"step": 2165
},
{
"epoch": 0.25,
"grad_norm": 0.1467702341656832,
"learning_rate": 1.867855462071614e-05,
"loss": 1.0253,
"step": 2170
},
{
"epoch": 0.25,
"grad_norm": 0.14022879328380106,
"learning_rate": 1.8668586973659207e-05,
"loss": 0.9936,
"step": 2175
},
{
"epoch": 0.25,
"grad_norm": 0.15508696907259692,
"learning_rate": 1.8658584555097168e-05,
"loss": 1.0398,
"step": 2180
},
{
"epoch": 0.25,
"grad_norm": 0.15340886678035695,
"learning_rate": 1.8648547405151805e-05,
"loss": 1.0312,
"step": 2185
},
{
"epoch": 0.25,
"grad_norm": 0.14492676159851775,
"learning_rate": 1.8638475564084216e-05,
"loss": 1.0033,
"step": 2190
},
{
"epoch": 0.25,
"grad_norm": 0.13574059414067258,
"learning_rate": 1.8628369072294646e-05,
"loss": 1.0456,
"step": 2195
},
{
"epoch": 0.25,
"grad_norm": 0.1472970746123335,
"learning_rate": 1.8618227970322338e-05,
"loss": 0.9593,
"step": 2200
},
{
"epoch": 0.25,
"grad_norm": 0.1497908002451653,
"learning_rate": 1.860805229884536e-05,
"loss": 1.0123,
"step": 2205
},
{
"epoch": 0.25,
"grad_norm": 0.18700201040820522,
"learning_rate": 1.859784209868045e-05,
"loss": 1.0189,
"step": 2210
},
{
"epoch": 0.25,
"grad_norm": 0.16807382207588892,
"learning_rate": 1.8587597410782837e-05,
"loss": 0.967,
"step": 2215
},
{
"epoch": 0.25,
"grad_norm": 0.13487095774113284,
"learning_rate": 1.8577318276246106e-05,
"loss": 1.0151,
"step": 2220
},
{
"epoch": 0.26,
"grad_norm": 0.15391839381218714,
"learning_rate": 1.856700473630199e-05,
"loss": 1.0165,
"step": 2225
},
{
"epoch": 0.26,
"grad_norm": 0.1448298429179419,
"learning_rate": 1.855665683232025e-05,
"loss": 1.0194,
"step": 2230
},
{
"epoch": 0.26,
"grad_norm": 0.13849531546582777,
"learning_rate": 1.8546274605808482e-05,
"loss": 1.0048,
"step": 2235
},
{
"epoch": 0.26,
"grad_norm": 0.13627822702338083,
"learning_rate": 1.8535858098411952e-05,
"loss": 0.9697,
"step": 2240
},
{
"epoch": 0.26,
"grad_norm": 0.14500985335088185,
"learning_rate": 1.8525407351913438e-05,
"loss": 1.0266,
"step": 2245
},
{
"epoch": 0.26,
"grad_norm": 0.16805732683664462,
"learning_rate": 1.851492240823306e-05,
"loss": 0.9972,
"step": 2250
},
{
"epoch": 0.26,
"grad_norm": 0.21962882391008803,
"learning_rate": 1.8504403309428105e-05,
"loss": 0.9821,
"step": 2255
},
{
"epoch": 0.26,
"grad_norm": 0.15336463173741993,
"learning_rate": 1.8493850097692866e-05,
"loss": 1.0142,
"step": 2260
},
{
"epoch": 0.26,
"grad_norm": 0.1557315228388552,
"learning_rate": 1.8483262815358475e-05,
"loss": 1.001,
"step": 2265
},
{
"epoch": 0.26,
"grad_norm": 0.14973073479538218,
"learning_rate": 1.8472641504892713e-05,
"loss": 1.0481,
"step": 2270
},
{
"epoch": 0.26,
"grad_norm": 0.14182369202692782,
"learning_rate": 1.846198620889988e-05,
"loss": 1.014,
"step": 2275
},
{
"epoch": 0.26,
"grad_norm": 0.1433656379177142,
"learning_rate": 1.8451296970120566e-05,
"loss": 0.9606,
"step": 2280
},
{
"epoch": 0.26,
"grad_norm": 0.13449600524470437,
"learning_rate": 1.8440573831431546e-05,
"loss": 1.0031,
"step": 2285
},
{
"epoch": 0.26,
"grad_norm": 0.15786224200700844,
"learning_rate": 1.8429816835845548e-05,
"loss": 1.0427,
"step": 2290
},
{
"epoch": 0.26,
"grad_norm": 0.16764324137065814,
"learning_rate": 1.8419026026511127e-05,
"loss": 1.0153,
"step": 2295
},
{
"epoch": 0.26,
"grad_norm": 0.14131702632138432,
"learning_rate": 1.840820144671246e-05,
"loss": 1.0049,
"step": 2300
},
{
"epoch": 0.26,
"grad_norm": 0.16984520487366736,
"learning_rate": 1.839734313986919e-05,
"loss": 0.981,
"step": 2305
},
{
"epoch": 0.27,
"grad_norm": 0.16132711501382002,
"learning_rate": 1.838645114953624e-05,
"loss": 1.0401,
"step": 2310
},
{
"epoch": 0.27,
"grad_norm": 0.15466015692881285,
"learning_rate": 1.8375525519403648e-05,
"loss": 0.9965,
"step": 2315
},
{
"epoch": 0.27,
"grad_norm": 0.14443067813391347,
"learning_rate": 1.8364566293296394e-05,
"loss": 0.9728,
"step": 2320
},
{
"epoch": 0.27,
"grad_norm": 0.15015559927713984,
"learning_rate": 1.8353573515174206e-05,
"loss": 0.9728,
"step": 2325
},
{
"epoch": 0.27,
"grad_norm": 0.14264736039920997,
"learning_rate": 1.834254722913141e-05,
"loss": 0.9943,
"step": 2330
},
{
"epoch": 0.27,
"grad_norm": 0.13900257699383117,
"learning_rate": 1.8331487479396728e-05,
"loss": 0.9766,
"step": 2335
},
{
"epoch": 0.27,
"grad_norm": 0.16604489637827716,
"learning_rate": 1.8320394310333118e-05,
"loss": 0.9996,
"step": 2340
},
{
"epoch": 0.27,
"grad_norm": 0.15883475377408934,
"learning_rate": 1.8309267766437587e-05,
"loss": 0.9691,
"step": 2345
},
{
"epoch": 0.27,
"grad_norm": 0.14141248532354636,
"learning_rate": 1.8298107892341013e-05,
"loss": 1.0342,
"step": 2350
},
{
"epoch": 0.27,
"grad_norm": 0.18856588895931942,
"learning_rate": 1.8286914732807984e-05,
"loss": 0.9888,
"step": 2355
},
{
"epoch": 0.27,
"grad_norm": 0.16256523314678267,
"learning_rate": 1.8275688332736577e-05,
"loss": 1.0087,
"step": 2360
},
{
"epoch": 0.27,
"grad_norm": 0.18512409188553108,
"learning_rate": 1.8264428737158223e-05,
"loss": 1.0282,
"step": 2365
},
{
"epoch": 0.27,
"grad_norm": 0.134192286162916,
"learning_rate": 1.8253135991237503e-05,
"loss": 0.9647,
"step": 2370
},
{
"epoch": 0.27,
"grad_norm": 0.14345412325796655,
"learning_rate": 1.8241810140271966e-05,
"loss": 1.0358,
"step": 2375
},
{
"epoch": 0.27,
"grad_norm": 0.15071290527219164,
"learning_rate": 1.8230451229691958e-05,
"loss": 1.0166,
"step": 2380
},
{
"epoch": 0.27,
"grad_norm": 0.13882843358984506,
"learning_rate": 1.821905930506043e-05,
"loss": 0.9774,
"step": 2385
},
{
"epoch": 0.27,
"grad_norm": 0.15593692820696595,
"learning_rate": 1.8207634412072765e-05,
"loss": 0.993,
"step": 2390
},
{
"epoch": 0.27,
"grad_norm": 0.16918306422857426,
"learning_rate": 1.819617659655658e-05,
"loss": 0.9739,
"step": 2395
},
{
"epoch": 0.28,
"grad_norm": 0.17204199770796053,
"learning_rate": 1.8184685904471554e-05,
"loss": 1.0257,
"step": 2400
},
{
"epoch": 0.28,
"grad_norm": 0.15330654679567915,
"learning_rate": 1.817316238190925e-05,
"loss": 0.9718,
"step": 2405
},
{
"epoch": 0.28,
"grad_norm": 0.14810232525452924,
"learning_rate": 1.816160607509291e-05,
"loss": 0.9776,
"step": 2410
},
{
"epoch": 0.28,
"grad_norm": 0.14997829130451903,
"learning_rate": 1.815001703037728e-05,
"loss": 1.0882,
"step": 2415
},
{
"epoch": 0.28,
"grad_norm": 0.13847095743272989,
"learning_rate": 1.8138395294248433e-05,
"loss": 1.003,
"step": 2420
},
{
"epoch": 0.28,
"grad_norm": 0.1516529053624639,
"learning_rate": 1.812674091332357e-05,
"loss": 0.9675,
"step": 2425
},
{
"epoch": 0.28,
"grad_norm": 0.14250514492931923,
"learning_rate": 1.811505393435083e-05,
"loss": 0.9944,
"step": 2430
},
{
"epoch": 0.28,
"grad_norm": 0.1465574890961,
"learning_rate": 1.8103334404209123e-05,
"loss": 0.9779,
"step": 2435
},
{
"epoch": 0.28,
"grad_norm": 0.15933541368122822,
"learning_rate": 1.809158236990792e-05,
"loss": 0.9963,
"step": 2440
},
{
"epoch": 0.28,
"grad_norm": 0.13417887512830484,
"learning_rate": 1.8079797878587077e-05,
"loss": 1.0652,
"step": 2445
},
{
"epoch": 0.28,
"grad_norm": 0.13282909562536516,
"learning_rate": 1.8067980977516634e-05,
"loss": 1.0192,
"step": 2450
},
{
"epoch": 0.28,
"grad_norm": 0.21656508172127717,
"learning_rate": 1.805613171409664e-05,
"loss": 1.0007,
"step": 2455
},
{
"epoch": 0.28,
"grad_norm": 0.14364784302979358,
"learning_rate": 1.804425013585696e-05,
"loss": 1.0033,
"step": 2460
},
{
"epoch": 0.28,
"grad_norm": 0.1811135259820456,
"learning_rate": 1.803233629045707e-05,
"loss": 1.0095,
"step": 2465
},
{
"epoch": 0.28,
"grad_norm": 0.16244857458347148,
"learning_rate": 1.802039022568588e-05,
"loss": 0.9793,
"step": 2470
},
{
"epoch": 0.28,
"grad_norm": 0.14301605378866003,
"learning_rate": 1.8008411989461546e-05,
"loss": 0.996,
"step": 2475
},
{
"epoch": 0.28,
"grad_norm": 0.15285820395617733,
"learning_rate": 1.799640162983126e-05,
"loss": 1.0131,
"step": 2480
},
{
"epoch": 0.29,
"grad_norm": 0.14598983844847963,
"learning_rate": 1.7984359194971075e-05,
"loss": 0.9893,
"step": 2485
},
{
"epoch": 0.29,
"grad_norm": 0.15218860579087728,
"learning_rate": 1.79722847331857e-05,
"loss": 0.977,
"step": 2490
},
{
"epoch": 0.29,
"grad_norm": 0.15348094825183217,
"learning_rate": 1.796017829290831e-05,
"loss": 1.0048,
"step": 2495
},
{
"epoch": 0.29,
"grad_norm": 0.14711580001121471,
"learning_rate": 1.794803992270036e-05,
"loss": 0.9969,
"step": 2500
},
{
"epoch": 0.29,
"grad_norm": 0.14630037297330156,
"learning_rate": 1.793586967125138e-05,
"loss": 1.0276,
"step": 2505
},
{
"epoch": 0.29,
"grad_norm": 0.1520787054740122,
"learning_rate": 1.7923667587378767e-05,
"loss": 1.0173,
"step": 2510
},
{
"epoch": 0.29,
"grad_norm": 0.14013195796987,
"learning_rate": 1.7911433720027624e-05,
"loss": 0.9897,
"step": 2515
},
{
"epoch": 0.29,
"grad_norm": 0.1405846571037524,
"learning_rate": 1.789916811827054e-05,
"loss": 1.0277,
"step": 2520
},
{
"epoch": 0.29,
"grad_norm": 0.1624624162901583,
"learning_rate": 1.788687083130739e-05,
"loss": 1.0036,
"step": 2525
},
{
"epoch": 0.29,
"grad_norm": 0.1468801347543722,
"learning_rate": 1.7874541908465152e-05,
"loss": 0.9605,
"step": 2530
},
{
"epoch": 0.29,
"grad_norm": 0.15623906032544085,
"learning_rate": 1.7862181399197698e-05,
"loss": 1.0088,
"step": 2535
},
{
"epoch": 0.29,
"grad_norm": 0.12500556930667633,
"learning_rate": 1.7849789353085602e-05,
"loss": 0.9704,
"step": 2540
},
{
"epoch": 0.29,
"grad_norm": 0.13386647270461752,
"learning_rate": 1.7837365819835936e-05,
"loss": 0.9891,
"step": 2545
},
{
"epoch": 0.29,
"grad_norm": 0.15262508397839739,
"learning_rate": 1.782491084928208e-05,
"loss": 0.9897,
"step": 2550
},
{
"epoch": 0.29,
"grad_norm": 0.15962085164448628,
"learning_rate": 1.781242449138351e-05,
"loss": 0.9882,
"step": 2555
},
{
"epoch": 0.29,
"grad_norm": 0.157756811152772,
"learning_rate": 1.77999067962256e-05,
"loss": 1.037,
"step": 2560
},
{
"epoch": 0.29,
"grad_norm": 0.16950004434506324,
"learning_rate": 1.7787357814019432e-05,
"loss": 1.0039,
"step": 2565
},
{
"epoch": 0.29,
"grad_norm": 0.14643778315681977,
"learning_rate": 1.7774777595101583e-05,
"loss": 1.0147,
"step": 2570
},
{
"epoch": 0.3,
"grad_norm": 0.16589496235241818,
"learning_rate": 1.7762166189933926e-05,
"loss": 1.009,
"step": 2575
},
{
"epoch": 0.3,
"grad_norm": 0.1496648231562305,
"learning_rate": 1.7749523649103434e-05,
"loss": 0.9344,
"step": 2580
},
{
"epoch": 0.3,
"grad_norm": 0.16324447759814598,
"learning_rate": 1.7736850023321957e-05,
"loss": 1.0266,
"step": 2585
},
{
"epoch": 0.3,
"grad_norm": 0.14654326919089034,
"learning_rate": 1.7724145363426054e-05,
"loss": 0.989,
"step": 2590
},
{
"epoch": 0.3,
"grad_norm": 0.17734774369418213,
"learning_rate": 1.7711409720376757e-05,
"loss": 1.0178,
"step": 2595
},
{
"epoch": 0.3,
"grad_norm": 0.1627843359993115,
"learning_rate": 1.769864314525938e-05,
"loss": 1.0477,
"step": 2600
},
{
"epoch": 0.3,
"grad_norm": 0.14646262361993048,
"learning_rate": 1.7685845689283302e-05,
"loss": 0.9397,
"step": 2605
},
{
"epoch": 0.3,
"grad_norm": 0.1725541569419232,
"learning_rate": 1.7673017403781783e-05,
"loss": 1.0165,
"step": 2610
},
{
"epoch": 0.3,
"grad_norm": 0.1519176826962997,
"learning_rate": 1.766015834021175e-05,
"loss": 1.0049,
"step": 2615
},
{
"epoch": 0.3,
"grad_norm": 0.1486957013947854,
"learning_rate": 1.7647268550153583e-05,
"loss": 0.9792,
"step": 2620
},
{
"epoch": 0.3,
"grad_norm": 0.1456711653426539,
"learning_rate": 1.7634348085310898e-05,
"loss": 1.001,
"step": 2625
},
{
"epoch": 0.3,
"grad_norm": 0.15576128755669955,
"learning_rate": 1.762139699751038e-05,
"loss": 0.9921,
"step": 2630
},
{
"epoch": 0.3,
"grad_norm": 0.14913570675719823,
"learning_rate": 1.7608415338701523e-05,
"loss": 0.9902,
"step": 2635
},
{
"epoch": 0.3,
"grad_norm": 0.14167236488997018,
"learning_rate": 1.759540316095647e-05,
"loss": 1.0002,
"step": 2640
},
{
"epoch": 0.3,
"grad_norm": 0.17125413249355748,
"learning_rate": 1.758236051646977e-05,
"loss": 0.9635,
"step": 2645
},
{
"epoch": 0.3,
"grad_norm": 0.14911892684494912,
"learning_rate": 1.7569287457558178e-05,
"loss": 1.0009,
"step": 2650
},
{
"epoch": 0.3,
"grad_norm": 0.1409916304530217,
"learning_rate": 1.7556184036660456e-05,
"loss": 1.0098,
"step": 2655
},
{
"epoch": 0.31,
"grad_norm": 0.14971473836864732,
"learning_rate": 1.754305030633715e-05,
"loss": 1.0185,
"step": 2660
},
{
"epoch": 0.31,
"grad_norm": 0.1349425138428904,
"learning_rate": 1.7529886319270385e-05,
"loss": 0.9945,
"step": 2665
},
{
"epoch": 0.31,
"grad_norm": 0.14079028656503348,
"learning_rate": 1.7516692128263647e-05,
"loss": 0.9481,
"step": 2670
},
{
"epoch": 0.31,
"grad_norm": 0.15083202179118058,
"learning_rate": 1.750346778624158e-05,
"loss": 0.9954,
"step": 2675
},
{
"epoch": 0.31,
"grad_norm": 0.13874295399420766,
"learning_rate": 1.7490213346249775e-05,
"loss": 0.9791,
"step": 2680
},
{
"epoch": 0.31,
"grad_norm": 0.1545784530442434,
"learning_rate": 1.7476928861454546e-05,
"loss": 0.9728,
"step": 2685
},
{
"epoch": 0.31,
"grad_norm": 0.15231654326658317,
"learning_rate": 1.7463614385142718e-05,
"loss": 0.9555,
"step": 2690
},
{
"epoch": 0.31,
"grad_norm": 0.14801634637395344,
"learning_rate": 1.745026997072143e-05,
"loss": 0.9709,
"step": 2695
},
{
"epoch": 0.31,
"grad_norm": 0.21814340940337537,
"learning_rate": 1.7436895671717902e-05,
"loss": 1.0107,
"step": 2700
},
{
"epoch": 0.31,
"grad_norm": 0.17498401787967957,
"learning_rate": 1.742349154177923e-05,
"loss": 1.0287,
"step": 2705
},
{
"epoch": 0.31,
"grad_norm": 0.14168176204404248,
"learning_rate": 1.7410057634672165e-05,
"loss": 1.0606,
"step": 2710
},
{
"epoch": 0.31,
"grad_norm": 0.1661042928335187,
"learning_rate": 1.73965940042829e-05,
"loss": 1.025,
"step": 2715
},
{
"epoch": 0.31,
"grad_norm": 0.16158918884822218,
"learning_rate": 1.7383100704616862e-05,
"loss": 0.9995,
"step": 2720
},
{
"epoch": 0.31,
"grad_norm": 0.17015289646010429,
"learning_rate": 1.7369577789798478e-05,
"loss": 1.0401,
"step": 2725
},
{
"epoch": 0.31,
"grad_norm": 0.16167800569979143,
"learning_rate": 1.7356025314070973e-05,
"loss": 1.0016,
"step": 2730
},
{
"epoch": 0.31,
"grad_norm": 0.1431956324760432,
"learning_rate": 1.7342443331796147e-05,
"loss": 1.0073,
"step": 2735
},
{
"epoch": 0.31,
"grad_norm": 0.14818400917144067,
"learning_rate": 1.7328831897454153e-05,
"loss": 0.9925,
"step": 2740
},
{
"epoch": 0.31,
"grad_norm": 0.15139814427830187,
"learning_rate": 1.7315191065643287e-05,
"loss": 0.957,
"step": 2745
},
{
"epoch": 0.32,
"grad_norm": 0.1707134683456167,
"learning_rate": 1.730152089107976e-05,
"loss": 0.9992,
"step": 2750
},
{
"epoch": 0.32,
"grad_norm": 0.14664409787351823,
"learning_rate": 1.7287821428597487e-05,
"loss": 1.0353,
"step": 2755
},
{
"epoch": 0.32,
"grad_norm": 0.15286513514052796,
"learning_rate": 1.727409273314786e-05,
"loss": 0.9849,
"step": 2760
},
{
"epoch": 0.32,
"grad_norm": 0.14839234320713257,
"learning_rate": 1.7260334859799527e-05,
"loss": 0.9796,
"step": 2765
},
{
"epoch": 0.32,
"grad_norm": 0.12443858058243129,
"learning_rate": 1.724654786373819e-05,
"loss": 0.9663,
"step": 2770
},
{
"epoch": 0.32,
"grad_norm": 0.18239963079082802,
"learning_rate": 1.723273180026635e-05,
"loss": 0.9887,
"step": 2775
},
{
"epoch": 0.32,
"grad_norm": 0.13722868119976805,
"learning_rate": 1.7218886724803105e-05,
"loss": 0.919,
"step": 2780
},
{
"epoch": 0.32,
"grad_norm": 0.1451078878475434,
"learning_rate": 1.7205012692883947e-05,
"loss": 0.9876,
"step": 2785
},
{
"epoch": 0.32,
"grad_norm": 0.15195572431833393,
"learning_rate": 1.7191109760160493e-05,
"loss": 0.9627,
"step": 2790
},
{
"epoch": 0.32,
"grad_norm": 0.14694144259228795,
"learning_rate": 1.7177177982400297e-05,
"loss": 0.9518,
"step": 2795
},
{
"epoch": 0.32,
"grad_norm": 0.15418266761800634,
"learning_rate": 1.7163217415486623e-05,
"loss": 0.9699,
"step": 2800
},
{
"epoch": 0.32,
"grad_norm": 0.14194030695379925,
"learning_rate": 1.7149228115418202e-05,
"loss": 1.0031,
"step": 2805
},
{
"epoch": 0.32,
"grad_norm": 0.14123700518485774,
"learning_rate": 1.7135210138309034e-05,
"loss": 0.9945,
"step": 2810
},
{
"epoch": 0.32,
"grad_norm": 0.15617977647538556,
"learning_rate": 1.7121163540388128e-05,
"loss": 1.0096,
"step": 2815
},
{
"epoch": 0.32,
"grad_norm": 0.16091752335112725,
"learning_rate": 1.710708837799932e-05,
"loss": 1.0103,
"step": 2820
},
{
"epoch": 0.32,
"grad_norm": 0.15056334011117056,
"learning_rate": 1.7092984707601007e-05,
"loss": 1.0175,
"step": 2825
},
{
"epoch": 0.32,
"grad_norm": 0.13869823062322806,
"learning_rate": 1.7078852585765952e-05,
"loss": 1.0016,
"step": 2830
},
{
"epoch": 0.33,
"grad_norm": 0.15282778985992726,
"learning_rate": 1.7064692069181027e-05,
"loss": 0.9893,
"step": 2835
},
{
"epoch": 0.33,
"grad_norm": 0.18229062579598063,
"learning_rate": 1.7050503214647013e-05,
"loss": 1.0122,
"step": 2840
},
{
"epoch": 0.33,
"grad_norm": 0.1579288907575329,
"learning_rate": 1.703628607907835e-05,
"loss": 1.0067,
"step": 2845
},
{
"epoch": 0.33,
"grad_norm": 0.1659046846425838,
"learning_rate": 1.7022040719502935e-05,
"loss": 0.9777,
"step": 2850
},
{
"epoch": 0.33,
"grad_norm": 0.14594278412200198,
"learning_rate": 1.700776719306186e-05,
"loss": 1.0337,
"step": 2855
},
{
"epoch": 0.33,
"grad_norm": 0.15285341425112814,
"learning_rate": 1.6993465557009215e-05,
"loss": 0.9314,
"step": 2860
},
{
"epoch": 0.33,
"grad_norm": 0.16884110389294024,
"learning_rate": 1.697913586871183e-05,
"loss": 0.9419,
"step": 2865
},
{
"epoch": 0.33,
"grad_norm": 0.16967711155515589,
"learning_rate": 1.696477818564907e-05,
"loss": 0.9776,
"step": 2870
},
{
"epoch": 0.33,
"grad_norm": 0.1581725597707753,
"learning_rate": 1.6950392565412583e-05,
"loss": 0.9874,
"step": 2875
},
{
"epoch": 0.33,
"grad_norm": 0.20901785893274283,
"learning_rate": 1.693597906570608e-05,
"loss": 1.0035,
"step": 2880
},
{
"epoch": 0.33,
"grad_norm": 0.16183249364221827,
"learning_rate": 1.6921537744345115e-05,
"loss": 0.9843,
"step": 2885
},
{
"epoch": 0.33,
"grad_norm": 0.13382507381261877,
"learning_rate": 1.6907068659256827e-05,
"loss": 1.0178,
"step": 2890
},
{
"epoch": 0.33,
"grad_norm": 0.30185514910448424,
"learning_rate": 1.689257186847973e-05,
"loss": 0.9764,
"step": 2895
},
{
"epoch": 0.33,
"grad_norm": 0.15672812936563213,
"learning_rate": 1.6878047430163458e-05,
"loss": 0.9393,
"step": 2900
},
{
"epoch": 0.33,
"grad_norm": 0.15161834573664232,
"learning_rate": 1.6863495402568558e-05,
"loss": 0.9774,
"step": 2905
},
{
"epoch": 0.33,
"grad_norm": 0.15121191184404867,
"learning_rate": 1.684891584406624e-05,
"loss": 1.0474,
"step": 2910
},
{
"epoch": 0.33,
"grad_norm": 0.15139017850449124,
"learning_rate": 1.6834308813138146e-05,
"loss": 0.9971,
"step": 2915
},
{
"epoch": 0.34,
"grad_norm": 0.14044156232770924,
"learning_rate": 1.6819674368376113e-05,
"loss": 0.9979,
"step": 2920
},
{
"epoch": 0.34,
"grad_norm": 0.17353031957666357,
"learning_rate": 1.680501256848194e-05,
"loss": 1.013,
"step": 2925
},
{
"epoch": 0.34,
"grad_norm": 0.14055982436656086,
"learning_rate": 1.6790323472267157e-05,
"loss": 0.9844,
"step": 2930
},
{
"epoch": 0.34,
"grad_norm": 0.14874877922452998,
"learning_rate": 1.6775607138652783e-05,
"loss": 1.0008,
"step": 2935
},
{
"epoch": 0.34,
"grad_norm": 0.14624585621100494,
"learning_rate": 1.6760863626669096e-05,
"loss": 0.9727,
"step": 2940
},
{
"epoch": 0.34,
"grad_norm": 0.15985911326897653,
"learning_rate": 1.674609299545539e-05,
"loss": 0.9999,
"step": 2945
},
{
"epoch": 0.34,
"grad_norm": 0.15430179692726206,
"learning_rate": 1.673129530425973e-05,
"loss": 0.9467,
"step": 2950
},
{
"epoch": 0.34,
"grad_norm": 0.14419417291695694,
"learning_rate": 1.6716470612438742e-05,
"loss": 0.9621,
"step": 2955
},
{
"epoch": 0.34,
"grad_norm": 0.14851493689076908,
"learning_rate": 1.6701618979457345e-05,
"loss": 1.0134,
"step": 2960
},
{
"epoch": 0.34,
"grad_norm": 0.15002562297728478,
"learning_rate": 1.668674046488852e-05,
"loss": 1.0595,
"step": 2965
},
{
"epoch": 0.34,
"grad_norm": 0.14648618473936664,
"learning_rate": 1.66718351284131e-05,
"loss": 0.9901,
"step": 2970
},
{
"epoch": 0.34,
"grad_norm": 0.1699561407565958,
"learning_rate": 1.6656903029819468e-05,
"loss": 0.9943,
"step": 2975
},
{
"epoch": 0.34,
"grad_norm": 0.1353756893860777,
"learning_rate": 1.6641944229003395e-05,
"loss": 0.9655,
"step": 2980
},
{
"epoch": 0.34,
"grad_norm": 0.15577056581889925,
"learning_rate": 1.6626958785967734e-05,
"loss": 1.0255,
"step": 2985
},
{
"epoch": 0.34,
"grad_norm": 0.14081314863125893,
"learning_rate": 1.6611946760822214e-05,
"loss": 1.0082,
"step": 2990
},
{
"epoch": 0.34,
"grad_norm": 0.15801572066868766,
"learning_rate": 1.6596908213783186e-05,
"loss": 0.9885,
"step": 2995
},
{
"epoch": 0.34,
"grad_norm": 0.16952916841722168,
"learning_rate": 1.6581843205173398e-05,
"loss": 1.0476,
"step": 3000
},
{
"epoch": 0.34,
"grad_norm": 0.1422275375103095,
"learning_rate": 1.6566751795421727e-05,
"loss": 1.0409,
"step": 3005
},
{
"epoch": 0.35,
"grad_norm": 0.15041028883336063,
"learning_rate": 1.6551634045062958e-05,
"loss": 0.946,
"step": 3010
},
{
"epoch": 0.35,
"grad_norm": 0.14606094256630023,
"learning_rate": 1.653649001473753e-05,
"loss": 0.9895,
"step": 3015
},
{
"epoch": 0.35,
"grad_norm": 0.15867444540973671,
"learning_rate": 1.6521319765191295e-05,
"loss": 1.0383,
"step": 3020
},
{
"epoch": 0.35,
"grad_norm": 0.14577222738242976,
"learning_rate": 1.650612335727529e-05,
"loss": 0.9897,
"step": 3025
},
{
"epoch": 0.35,
"grad_norm": 0.16327142441684675,
"learning_rate": 1.6490900851945455e-05,
"loss": 0.9757,
"step": 3030
},
{
"epoch": 0.35,
"grad_norm": 0.16319441310094543,
"learning_rate": 1.647565231026243e-05,
"loss": 1.0549,
"step": 3035
},
{
"epoch": 0.35,
"grad_norm": 0.16268419106475082,
"learning_rate": 1.6460377793391285e-05,
"loss": 0.9535,
"step": 3040
},
{
"epoch": 0.35,
"grad_norm": 0.16246030819163074,
"learning_rate": 1.644507736260129e-05,
"loss": 1.0343,
"step": 3045
},
{
"epoch": 0.35,
"grad_norm": 0.156690439691343,
"learning_rate": 1.642975107926565e-05,
"loss": 1.0007,
"step": 3050
},
{
"epoch": 0.35,
"grad_norm": 0.16009415193410287,
"learning_rate": 1.641439900486127e-05,
"loss": 1.0045,
"step": 3055
},
{
"epoch": 0.35,
"grad_norm": 0.1414180174877536,
"learning_rate": 1.6399021200968522e-05,
"loss": 1.0066,
"step": 3060
},
{
"epoch": 0.35,
"grad_norm": 0.14415889171482144,
"learning_rate": 1.638361772927097e-05,
"loss": 0.9562,
"step": 3065
},
{
"epoch": 0.35,
"grad_norm": 0.18822206606734404,
"learning_rate": 1.6368188651555145e-05,
"loss": 1.002,
"step": 3070
},
{
"epoch": 0.35,
"grad_norm": 0.15168553198825174,
"learning_rate": 1.6352734029710284e-05,
"loss": 0.9753,
"step": 3075
},
{
"epoch": 0.35,
"grad_norm": 0.16539116920420904,
"learning_rate": 1.633725392572809e-05,
"loss": 1.0022,
"step": 3080
},
{
"epoch": 0.35,
"grad_norm": 0.16011477011931916,
"learning_rate": 1.6321748401702483e-05,
"loss": 1.0308,
"step": 3085
},
{
"epoch": 0.35,
"grad_norm": 0.18024081067238523,
"learning_rate": 1.6306217519829337e-05,
"loss": 1.0306,
"step": 3090
},
{
"epoch": 0.36,
"grad_norm": 0.1628109071521388,
"learning_rate": 1.629066134240626e-05,
"loss": 0.9811,
"step": 3095
},
{
"epoch": 0.36,
"grad_norm": 0.1422522255078145,
"learning_rate": 1.62750799318323e-05,
"loss": 0.9223,
"step": 3100
},
{
"epoch": 0.36,
"grad_norm": 0.15857512788986095,
"learning_rate": 1.6259473350607745e-05,
"loss": 1.0091,
"step": 3105
},
{
"epoch": 0.36,
"grad_norm": 0.1708568940660707,
"learning_rate": 1.6243841661333836e-05,
"loss": 0.9815,
"step": 3110
},
{
"epoch": 0.36,
"grad_norm": 0.1740003351944693,
"learning_rate": 1.6228184926712535e-05,
"loss": 1.022,
"step": 3115
},
{
"epoch": 0.36,
"grad_norm": 0.15369528246309377,
"learning_rate": 1.621250320954625e-05,
"loss": 0.9734,
"step": 3120
},
{
"epoch": 0.36,
"grad_norm": 0.17273160831145878,
"learning_rate": 1.6196796572737617e-05,
"loss": 0.9924,
"step": 3125
},
{
"epoch": 0.36,
"grad_norm": 0.14607491789927318,
"learning_rate": 1.6181065079289217e-05,
"loss": 1.0251,
"step": 3130
},
{
"epoch": 0.36,
"grad_norm": 0.13003429826387689,
"learning_rate": 1.616530879230335e-05,
"loss": 0.9919,
"step": 3135
},
{
"epoch": 0.36,
"grad_norm": 0.1551941865112347,
"learning_rate": 1.6149527774981756e-05,
"loss": 0.9765,
"step": 3140
},
{
"epoch": 0.36,
"grad_norm": 0.14091390735778414,
"learning_rate": 1.6133722090625378e-05,
"loss": 1.0026,
"step": 3145
},
{
"epoch": 0.36,
"grad_norm": 0.15631726189041886,
"learning_rate": 1.6117891802634095e-05,
"loss": 0.949,
"step": 3150
},
{
"epoch": 0.36,
"grad_norm": 0.14948519292056064,
"learning_rate": 1.6102036974506494e-05,
"loss": 0.9667,
"step": 3155
},
{
"epoch": 0.36,
"grad_norm": 0.13969500661386375,
"learning_rate": 1.6086157669839585e-05,
"loss": 0.9864,
"step": 3160
},
{
"epoch": 0.36,
"grad_norm": 0.1586162099931364,
"learning_rate": 1.6070253952328564e-05,
"loss": 1.0117,
"step": 3165
},
{
"epoch": 0.36,
"grad_norm": 0.16901669387951887,
"learning_rate": 1.6054325885766542e-05,
"loss": 0.9652,
"step": 3170
},
{
"epoch": 0.36,
"grad_norm": 0.15558787236108415,
"learning_rate": 1.603837353404431e-05,
"loss": 0.9622,
"step": 3175
},
{
"epoch": 0.36,
"grad_norm": 0.17484838574718448,
"learning_rate": 1.6022396961150074e-05,
"loss": 0.9632,
"step": 3180
},
{
"epoch": 0.37,
"grad_norm": 0.15639125922317623,
"learning_rate": 1.6006396231169185e-05,
"loss": 0.9875,
"step": 3185
},
{
"epoch": 0.37,
"grad_norm": 0.16840050627008415,
"learning_rate": 1.5990371408283897e-05,
"loss": 1.0382,
"step": 3190
},
{
"epoch": 0.37,
"grad_norm": 0.15300706212693554,
"learning_rate": 1.597432255677311e-05,
"loss": 1.0137,
"step": 3195
},
{
"epoch": 0.37,
"grad_norm": 0.14926906128927434,
"learning_rate": 1.5958249741012103e-05,
"loss": 0.9677,
"step": 3200
},
{
"epoch": 0.37,
"grad_norm": 0.1484202128310707,
"learning_rate": 1.5942153025472286e-05,
"loss": 1.0099,
"step": 3205
},
{
"epoch": 0.37,
"grad_norm": 0.15001769584155025,
"learning_rate": 1.592603247472093e-05,
"loss": 0.9921,
"step": 3210
},
{
"epoch": 0.37,
"grad_norm": 0.15083868442222803,
"learning_rate": 1.5909888153420915e-05,
"loss": 0.9695,
"step": 3215
},
{
"epoch": 0.37,
"grad_norm": 0.17293540142669137,
"learning_rate": 1.5893720126330476e-05,
"loss": 0.977,
"step": 3220
},
{
"epoch": 0.37,
"grad_norm": 0.16246334244113603,
"learning_rate": 1.5877528458302926e-05,
"loss": 0.9792,
"step": 3225
},
{
"epoch": 0.37,
"grad_norm": 0.1570674830092573,
"learning_rate": 1.586131321428642e-05,
"loss": 0.9938,
"step": 3230
},
{
"epoch": 0.37,
"grad_norm": 0.15735033714651314,
"learning_rate": 1.5845074459323667e-05,
"loss": 0.9956,
"step": 3235
},
{
"epoch": 0.37,
"grad_norm": 0.1571272611475555,
"learning_rate": 1.5828812258551694e-05,
"loss": 0.9813,
"step": 3240
},
{
"epoch": 0.37,
"grad_norm": 0.15918274692541856,
"learning_rate": 1.5812526677201567e-05,
"loss": 0.9945,
"step": 3245
},
{
"epoch": 0.37,
"grad_norm": 0.14436690745351366,
"learning_rate": 1.5796217780598144e-05,
"loss": 1.0062,
"step": 3250
},
{
"epoch": 0.37,
"grad_norm": 0.1485018074474233,
"learning_rate": 1.5779885634159802e-05,
"loss": 0.9714,
"step": 3255
},
{
"epoch": 0.37,
"grad_norm": 0.1407420449715402,
"learning_rate": 1.576353030339817e-05,
"loss": 0.9621,
"step": 3260
},
{
"epoch": 0.37,
"grad_norm": 0.16036869385142152,
"learning_rate": 1.5747151853917885e-05,
"loss": 1.0324,
"step": 3265
},
{
"epoch": 0.38,
"grad_norm": 0.14536659692515053,
"learning_rate": 1.5730750351416316e-05,
"loss": 0.9622,
"step": 3270
},
{
"epoch": 0.38,
"grad_norm": 0.15548878120595266,
"learning_rate": 1.5714325861683295e-05,
"loss": 0.9995,
"step": 3275
},
{
"epoch": 0.38,
"grad_norm": 0.13745464384568895,
"learning_rate": 1.569787845060087e-05,
"loss": 0.9584,
"step": 3280
},
{
"epoch": 0.38,
"grad_norm": 0.16080165926322829,
"learning_rate": 1.568140818414303e-05,
"loss": 1.0026,
"step": 3285
},
{
"epoch": 0.38,
"grad_norm": 0.15663911184681953,
"learning_rate": 1.566491512837543e-05,
"loss": 0.9881,
"step": 3290
},
{
"epoch": 0.38,
"grad_norm": 0.17512739668385027,
"learning_rate": 1.5648399349455157e-05,
"loss": 1.0009,
"step": 3295
},
{
"epoch": 0.38,
"grad_norm": 0.15071062817167954,
"learning_rate": 1.5631860913630424e-05,
"loss": 1.0061,
"step": 3300
},
{
"epoch": 0.38,
"grad_norm": 0.1555491965968726,
"learning_rate": 1.561529988724035e-05,
"loss": 1.0351,
"step": 3305
},
{
"epoch": 0.38,
"grad_norm": 0.15327441298717967,
"learning_rate": 1.5598716336714645e-05,
"loss": 1.0398,
"step": 3310
},
{
"epoch": 0.38,
"grad_norm": 0.14656905041649712,
"learning_rate": 1.5582110328573382e-05,
"loss": 0.9958,
"step": 3315
},
{
"epoch": 0.38,
"grad_norm": 0.14362922465804928,
"learning_rate": 1.5565481929426715e-05,
"loss": 0.9593,
"step": 3320
},
{
"epoch": 0.38,
"grad_norm": 0.1759534627433801,
"learning_rate": 1.554883120597461e-05,
"loss": 0.9784,
"step": 3325
},
{
"epoch": 0.38,
"grad_norm": 0.15786741271562332,
"learning_rate": 1.553215822500658e-05,
"loss": 1.0149,
"step": 3330
},
{
"epoch": 0.38,
"grad_norm": 0.15856257892106465,
"learning_rate": 1.5515463053401422e-05,
"loss": 1.0148,
"step": 3335
},
{
"epoch": 0.38,
"grad_norm": 0.17746198707508795,
"learning_rate": 1.5498745758126937e-05,
"loss": 0.9827,
"step": 3340
},
{
"epoch": 0.38,
"grad_norm": 0.15046797260703446,
"learning_rate": 1.5482006406239673e-05,
"loss": 0.9806,
"step": 3345
},
{
"epoch": 0.38,
"grad_norm": 0.16691641350237055,
"learning_rate": 1.546524506488466e-05,
"loss": 1.0116,
"step": 3350
},
{
"epoch": 0.38,
"grad_norm": 0.14286953078191655,
"learning_rate": 1.5448461801295116e-05,
"loss": 1.0016,
"step": 3355
},
{
"epoch": 0.39,
"grad_norm": 0.16776321662261176,
"learning_rate": 1.5431656682792205e-05,
"loss": 0.9713,
"step": 3360
},
{
"epoch": 0.39,
"grad_norm": 0.14946430730764748,
"learning_rate": 1.5414829776784747e-05,
"loss": 0.9377,
"step": 3365
},
{
"epoch": 0.39,
"grad_norm": 0.14479142510361898,
"learning_rate": 1.5397981150768966e-05,
"loss": 1.0123,
"step": 3370
},
{
"epoch": 0.39,
"grad_norm": 0.1464233036744324,
"learning_rate": 1.5381110872328205e-05,
"loss": 0.9964,
"step": 3375
},
{
"epoch": 0.39,
"grad_norm": 0.1447149314223894,
"learning_rate": 1.536421900913266e-05,
"loss": 0.9906,
"step": 3380
},
{
"epoch": 0.39,
"grad_norm": 0.13703951734401543,
"learning_rate": 1.5347305628939106e-05,
"loss": 0.9272,
"step": 3385
},
{
"epoch": 0.39,
"grad_norm": 0.1588255096699679,
"learning_rate": 1.533037079959062e-05,
"loss": 1.0233,
"step": 3390
},
{
"epoch": 0.39,
"grad_norm": 0.14464357013960558,
"learning_rate": 1.531341458901634e-05,
"loss": 0.9526,
"step": 3395
},
{
"epoch": 0.39,
"grad_norm": 0.15935442378104275,
"learning_rate": 1.5296437065231145e-05,
"loss": 0.9888,
"step": 3400
},
{
"epoch": 0.39,
"grad_norm": 0.16393215410729484,
"learning_rate": 1.527943829633541e-05,
"loss": 0.9793,
"step": 3405
},
{
"epoch": 0.39,
"grad_norm": 0.15326843295684908,
"learning_rate": 1.5262418350514736e-05,
"loss": 1.0055,
"step": 3410
},
{
"epoch": 0.39,
"grad_norm": 0.14249574273223028,
"learning_rate": 1.5245377296039665e-05,
"loss": 1.0038,
"step": 3415
},
{
"epoch": 0.39,
"grad_norm": 0.13917680151518602,
"learning_rate": 1.5228315201265412e-05,
"loss": 0.9589,
"step": 3420
},
{
"epoch": 0.39,
"grad_norm": 0.15223902584076057,
"learning_rate": 1.5211232134631586e-05,
"loss": 0.9908,
"step": 3425
},
{
"epoch": 0.39,
"grad_norm": 0.1525869530749806,
"learning_rate": 1.5194128164661922e-05,
"loss": 1.0664,
"step": 3430
},
{
"epoch": 0.39,
"grad_norm": 0.16860799332107007,
"learning_rate": 1.5177003359963994e-05,
"loss": 1.0206,
"step": 3435
},
{
"epoch": 0.39,
"grad_norm": 0.1530172217473774,
"learning_rate": 1.5159857789228958e-05,
"loss": 1.0424,
"step": 3440
},
{
"epoch": 0.4,
"grad_norm": 0.17541030700425464,
"learning_rate": 1.5142691521231267e-05,
"loss": 0.9846,
"step": 3445
},
{
"epoch": 0.4,
"grad_norm": 0.17359697438486013,
"learning_rate": 1.5125504624828392e-05,
"loss": 0.9393,
"step": 3450
},
{
"epoch": 0.4,
"grad_norm": 0.1449251427303983,
"learning_rate": 1.5108297168960544e-05,
"loss": 0.9865,
"step": 3455
},
{
"epoch": 0.4,
"grad_norm": 0.14941544195504816,
"learning_rate": 1.5091069222650413e-05,
"loss": 0.9567,
"step": 3460
},
{
"epoch": 0.4,
"grad_norm": 0.13848508013291608,
"learning_rate": 1.5073820855002868e-05,
"loss": 0.9805,
"step": 3465
},
{
"epoch": 0.4,
"grad_norm": 0.1463685230312249,
"learning_rate": 1.5056552135204705e-05,
"loss": 1.045,
"step": 3470
},
{
"epoch": 0.4,
"grad_norm": 0.1616079486890371,
"learning_rate": 1.5039263132524346e-05,
"loss": 1.0086,
"step": 3475
},
{
"epoch": 0.4,
"grad_norm": 0.17883261656286584,
"learning_rate": 1.5021953916311578e-05,
"loss": 0.9778,
"step": 3480
},
{
"epoch": 0.4,
"grad_norm": 0.17522782557586933,
"learning_rate": 1.5004624555997267e-05,
"loss": 0.968,
"step": 3485
},
{
"epoch": 0.4,
"grad_norm": 0.16765912145829395,
"learning_rate": 1.4987275121093078e-05,
"loss": 1.024,
"step": 3490
},
{
"epoch": 0.4,
"grad_norm": 0.14030176755001708,
"learning_rate": 1.4969905681191209e-05,
"loss": 1.0267,
"step": 3495
},
{
"epoch": 0.4,
"grad_norm": 0.16793857094726738,
"learning_rate": 1.4952516305964086e-05,
"loss": 0.9954,
"step": 3500
},
{
"epoch": 0.4,
"grad_norm": 0.1553990090424684,
"learning_rate": 1.4935107065164121e-05,
"loss": 0.9965,
"step": 3505
},
{
"epoch": 0.4,
"grad_norm": 0.13981625215916743,
"learning_rate": 1.4917678028623392e-05,
"loss": 0.9904,
"step": 3510
},
{
"epoch": 0.4,
"grad_norm": 0.1605072005530355,
"learning_rate": 1.490022926625339e-05,
"loss": 0.9491,
"step": 3515
},
{
"epoch": 0.4,
"grad_norm": 0.157695063769502,
"learning_rate": 1.4882760848044728e-05,
"loss": 0.9663,
"step": 3520
},
{
"epoch": 0.4,
"grad_norm": 0.17106319260303612,
"learning_rate": 1.486527284406687e-05,
"loss": 0.9726,
"step": 3525
},
{
"epoch": 0.41,
"grad_norm": 0.16184771934926728,
"learning_rate": 1.484776532446783e-05,
"loss": 1.0041,
"step": 3530
},
{
"epoch": 0.41,
"grad_norm": 0.15027907994594003,
"learning_rate": 1.4830238359473916e-05,
"loss": 0.9595,
"step": 3535
},
{
"epoch": 0.41,
"grad_norm": 0.14628119012701915,
"learning_rate": 1.4812692019389426e-05,
"loss": 0.9924,
"step": 3540
},
{
"epoch": 0.41,
"grad_norm": 0.14219474469521565,
"learning_rate": 1.4795126374596378e-05,
"loss": 0.9558,
"step": 3545
},
{
"epoch": 0.41,
"grad_norm": 0.1488424001685648,
"learning_rate": 1.4777541495554231e-05,
"loss": 0.9726,
"step": 3550
},
{
"epoch": 0.41,
"grad_norm": 0.16016469383625997,
"learning_rate": 1.475993745279959e-05,
"loss": 1.0547,
"step": 3555
},
{
"epoch": 0.41,
"grad_norm": 0.1627929095414222,
"learning_rate": 1.474231431694593e-05,
"loss": 0.9795,
"step": 3560
},
{
"epoch": 0.41,
"grad_norm": 0.14358770499484072,
"learning_rate": 1.4724672158683311e-05,
"loss": 0.9712,
"step": 3565
},
{
"epoch": 0.41,
"grad_norm": 0.15784116678071797,
"learning_rate": 1.4707011048778103e-05,
"loss": 1.0256,
"step": 3570
},
{
"epoch": 0.41,
"grad_norm": 0.15413701121111165,
"learning_rate": 1.4689331058072694e-05,
"loss": 0.9924,
"step": 3575
},
{
"epoch": 0.41,
"grad_norm": 0.15033940998910633,
"learning_rate": 1.4671632257485198e-05,
"loss": 0.986,
"step": 3580
},
{
"epoch": 0.41,
"grad_norm": 0.1693042686679888,
"learning_rate": 1.4653914718009186e-05,
"loss": 0.9982,
"step": 3585
},
{
"epoch": 0.41,
"grad_norm": 0.1579206489619635,
"learning_rate": 1.463617851071339e-05,
"loss": 0.9877,
"step": 3590
},
{
"epoch": 0.41,
"grad_norm": 0.1404325074073772,
"learning_rate": 1.4618423706741433e-05,
"loss": 0.9894,
"step": 3595
},
{
"epoch": 0.41,
"grad_norm": 0.15149708326140202,
"learning_rate": 1.4600650377311523e-05,
"loss": 0.9654,
"step": 3600
},
{
"epoch": 0.41,
"grad_norm": 0.1401347013029462,
"learning_rate": 1.458285859371618e-05,
"loss": 0.9909,
"step": 3605
},
{
"epoch": 0.41,
"grad_norm": 0.14022200107670948,
"learning_rate": 1.4565048427321949e-05,
"loss": 1.0068,
"step": 3610
},
{
"epoch": 0.41,
"grad_norm": 0.16196209679361157,
"learning_rate": 1.4547219949569108e-05,
"loss": 0.9752,
"step": 3615
},
{
"epoch": 0.42,
"grad_norm": 0.14601897828285645,
"learning_rate": 1.4529373231971394e-05,
"loss": 0.996,
"step": 3620
},
{
"epoch": 0.42,
"grad_norm": 0.14922547527545116,
"learning_rate": 1.45115083461157e-05,
"loss": 0.9888,
"step": 3625
},
{
"epoch": 0.42,
"grad_norm": 0.16080928014134555,
"learning_rate": 1.4493625363661798e-05,
"loss": 0.9888,
"step": 3630
},
{
"epoch": 0.42,
"grad_norm": 0.1543997211807014,
"learning_rate": 1.4475724356342053e-05,
"loss": 0.9545,
"step": 3635
},
{
"epoch": 0.42,
"grad_norm": 0.15571593296985692,
"learning_rate": 1.4457805395961124e-05,
"loss": 0.9995,
"step": 3640
},
{
"epoch": 0.42,
"grad_norm": 0.15849435745005291,
"learning_rate": 1.4439868554395688e-05,
"loss": 0.9754,
"step": 3645
},
{
"epoch": 0.42,
"grad_norm": 0.22636046115187175,
"learning_rate": 1.4421913903594149e-05,
"loss": 0.966,
"step": 3650
},
{
"epoch": 0.42,
"grad_norm": 0.15946790102223793,
"learning_rate": 1.4403941515576344e-05,
"loss": 0.9578,
"step": 3655
},
{
"epoch": 0.42,
"grad_norm": 0.14832051144269262,
"learning_rate": 1.438595146243326e-05,
"loss": 1.0057,
"step": 3660
},
{
"epoch": 0.42,
"grad_norm": 0.1569585270235798,
"learning_rate": 1.4367943816326742e-05,
"loss": 1.0089,
"step": 3665
},
{
"epoch": 0.42,
"grad_norm": 0.16007228766458445,
"learning_rate": 1.4349918649489198e-05,
"loss": 1.011,
"step": 3670
},
{
"epoch": 0.42,
"grad_norm": 0.14536059175561813,
"learning_rate": 1.4331876034223325e-05,
"loss": 1.0144,
"step": 3675
},
{
"epoch": 0.42,
"grad_norm": 0.1563610541229928,
"learning_rate": 1.4313816042901808e-05,
"loss": 0.9777,
"step": 3680
},
{
"epoch": 0.42,
"grad_norm": 0.16797283340763924,
"learning_rate": 1.4295738747967023e-05,
"loss": 0.9978,
"step": 3685
},
{
"epoch": 0.42,
"grad_norm": 0.14723591676593217,
"learning_rate": 1.4277644221930763e-05,
"loss": 0.9766,
"step": 3690
},
{
"epoch": 0.42,
"grad_norm": 0.16082902615877878,
"learning_rate": 1.425953253737393e-05,
"loss": 1.0053,
"step": 3695
},
{
"epoch": 0.42,
"grad_norm": 0.144584090639964,
"learning_rate": 1.4241403766946262e-05,
"loss": 1.0589,
"step": 3700
},
{
"epoch": 0.43,
"grad_norm": 0.15594026211568185,
"learning_rate": 1.4223257983366026e-05,
"loss": 0.9807,
"step": 3705
},
{
"epoch": 0.43,
"grad_norm": 0.17739984932663028,
"learning_rate": 1.4205095259419734e-05,
"loss": 1.0184,
"step": 3710
},
{
"epoch": 0.43,
"grad_norm": 0.1710157883931397,
"learning_rate": 1.418691566796185e-05,
"loss": 1.0358,
"step": 3715
},
{
"epoch": 0.43,
"grad_norm": 0.22505898733935215,
"learning_rate": 1.4168719281914492e-05,
"loss": 0.9603,
"step": 3720
},
{
"epoch": 0.43,
"grad_norm": 0.16633278023222484,
"learning_rate": 1.4150506174267148e-05,
"loss": 0.966,
"step": 3725
},
{
"epoch": 0.43,
"grad_norm": 0.1508418666390597,
"learning_rate": 1.4132276418076389e-05,
"loss": 0.9554,
"step": 3730
},
{
"epoch": 0.43,
"grad_norm": 0.1680071643045003,
"learning_rate": 1.4114030086465553e-05,
"loss": 0.992,
"step": 3735
},
{
"epoch": 0.43,
"grad_norm": 0.1502061066134189,
"learning_rate": 1.4095767252624466e-05,
"loss": 1.0212,
"step": 3740
},
{
"epoch": 0.43,
"grad_norm": 0.16070258944756927,
"learning_rate": 1.407748798980916e-05,
"loss": 1.021,
"step": 3745
},
{
"epoch": 0.43,
"grad_norm": 0.19286101676398434,
"learning_rate": 1.4059192371341555e-05,
"loss": 1.008,
"step": 3750
},
{
"epoch": 0.43,
"grad_norm": 0.14763598478200443,
"learning_rate": 1.4040880470609186e-05,
"loss": 0.9976,
"step": 3755
},
{
"epoch": 0.43,
"grad_norm": 0.17053359371582075,
"learning_rate": 1.402255236106489e-05,
"loss": 0.9916,
"step": 3760
},
{
"epoch": 0.43,
"grad_norm": 0.17198831999976935,
"learning_rate": 1.4004208116226528e-05,
"loss": 0.9348,
"step": 3765
},
{
"epoch": 0.43,
"grad_norm": 0.18553771594930055,
"learning_rate": 1.3985847809676681e-05,
"loss": 0.9865,
"step": 3770
},
{
"epoch": 0.43,
"grad_norm": 0.16449021125155008,
"learning_rate": 1.3967471515062353e-05,
"loss": 1.0062,
"step": 3775
},
{
"epoch": 0.43,
"grad_norm": 0.1636504146101553,
"learning_rate": 1.3949079306094692e-05,
"loss": 1.0615,
"step": 3780
},
{
"epoch": 0.43,
"grad_norm": 0.1434366481808659,
"learning_rate": 1.3930671256548668e-05,
"loss": 0.9473,
"step": 3785
},
{
"epoch": 0.43,
"grad_norm": 0.15752445867757173,
"learning_rate": 1.3912247440262797e-05,
"loss": 1.0083,
"step": 3790
},
{
"epoch": 0.44,
"grad_norm": 0.18654137912806318,
"learning_rate": 1.3893807931138836e-05,
"loss": 0.9697,
"step": 3795
},
{
"epoch": 0.44,
"grad_norm": 0.2062214930612226,
"learning_rate": 1.3875352803141489e-05,
"loss": 0.959,
"step": 3800
},
{
"epoch": 0.44,
"grad_norm": 0.14830224323318772,
"learning_rate": 1.3856882130298118e-05,
"loss": 0.9668,
"step": 3805
},
{
"epoch": 0.44,
"grad_norm": 0.14834169918624612,
"learning_rate": 1.3838395986698427e-05,
"loss": 0.9984,
"step": 3810
},
{
"epoch": 0.44,
"grad_norm": 0.1508110848962588,
"learning_rate": 1.3819894446494189e-05,
"loss": 1.0628,
"step": 3815
},
{
"epoch": 0.44,
"grad_norm": 0.13323443581294903,
"learning_rate": 1.3801377583898927e-05,
"loss": 1.0398,
"step": 3820
},
{
"epoch": 0.44,
"grad_norm": 0.14123180946144745,
"learning_rate": 1.3782845473187623e-05,
"loss": 0.9695,
"step": 3825
},
{
"epoch": 0.44,
"grad_norm": 0.15914721559172718,
"learning_rate": 1.376429818869643e-05,
"loss": 0.9806,
"step": 3830
},
{
"epoch": 0.44,
"grad_norm": 0.14481911026585031,
"learning_rate": 1.3745735804822367e-05,
"loss": 1.0132,
"step": 3835
},
{
"epoch": 0.44,
"grad_norm": 0.15344600368269243,
"learning_rate": 1.3727158396023012e-05,
"loss": 0.9909,
"step": 3840
},
{
"epoch": 0.44,
"grad_norm": 0.15227719890320587,
"learning_rate": 1.3708566036816217e-05,
"loss": 0.976,
"step": 3845
},
{
"epoch": 0.44,
"grad_norm": 0.13610577326036657,
"learning_rate": 1.36899588017798e-05,
"loss": 0.9572,
"step": 3850
},
{
"epoch": 0.44,
"grad_norm": 0.1504697412049455,
"learning_rate": 1.3671336765551248e-05,
"loss": 0.9685,
"step": 3855
},
{
"epoch": 0.44,
"grad_norm": 0.17643789341743507,
"learning_rate": 1.3652700002827425e-05,
"loss": 1.0092,
"step": 3860
},
{
"epoch": 0.44,
"grad_norm": 0.14946033992494584,
"learning_rate": 1.3634048588364264e-05,
"loss": 0.994,
"step": 3865
},
{
"epoch": 0.44,
"grad_norm": 0.15718910780920808,
"learning_rate": 1.3615382596976462e-05,
"loss": 1.0386,
"step": 3870
},
{
"epoch": 0.44,
"grad_norm": 0.19221776780769112,
"learning_rate": 1.3596702103537195e-05,
"loss": 0.996,
"step": 3875
},
{
"epoch": 0.45,
"grad_norm": 0.13738678316718522,
"learning_rate": 1.3578007182977804e-05,
"loss": 0.9846,
"step": 3880
},
{
"epoch": 0.45,
"grad_norm": 0.14603085622712667,
"learning_rate": 1.3559297910287509e-05,
"loss": 1.0316,
"step": 3885
},
{
"epoch": 0.45,
"grad_norm": 0.15028566899149973,
"learning_rate": 1.354057436051309e-05,
"loss": 0.9768,
"step": 3890
},
{
"epoch": 0.45,
"grad_norm": 0.146315428512196,
"learning_rate": 1.3521836608758596e-05,
"loss": 0.9897,
"step": 3895
},
{
"epoch": 0.45,
"grad_norm": 0.15739944777458292,
"learning_rate": 1.3503084730185054e-05,
"loss": 0.9842,
"step": 3900
},
{
"epoch": 0.45,
"grad_norm": 0.14891969207344075,
"learning_rate": 1.3484318800010138e-05,
"loss": 1.0198,
"step": 3905
},
{
"epoch": 0.45,
"grad_norm": 0.15292005024360214,
"learning_rate": 1.3465538893507908e-05,
"loss": 0.974,
"step": 3910
},
{
"epoch": 0.45,
"grad_norm": 0.15663606167393956,
"learning_rate": 1.3446745086008465e-05,
"loss": 0.951,
"step": 3915
},
{
"epoch": 0.45,
"grad_norm": 0.14685435704204747,
"learning_rate": 1.3427937452897689e-05,
"loss": 0.9386,
"step": 3920
},
{
"epoch": 0.45,
"grad_norm": 0.1570480689249208,
"learning_rate": 1.3409116069616898e-05,
"loss": 1.0526,
"step": 3925
},
{
"epoch": 0.45,
"grad_norm": 0.15052254358214684,
"learning_rate": 1.339028101166258e-05,
"loss": 0.9577,
"step": 3930
},
{
"epoch": 0.45,
"grad_norm": 0.2224310592911464,
"learning_rate": 1.3371432354586067e-05,
"loss": 0.9882,
"step": 3935
},
{
"epoch": 0.45,
"grad_norm": 0.14568639223236052,
"learning_rate": 1.3352570173993245e-05,
"loss": 0.9356,
"step": 3940
},
{
"epoch": 0.45,
"grad_norm": 0.13187133831027373,
"learning_rate": 1.333369454554424e-05,
"loss": 0.9584,
"step": 3945
},
{
"epoch": 0.45,
"grad_norm": 0.15503530001225607,
"learning_rate": 1.3314805544953126e-05,
"loss": 1.0144,
"step": 3950
},
{
"epoch": 0.45,
"grad_norm": 0.15982997478683678,
"learning_rate": 1.3295903247987609e-05,
"loss": 0.9754,
"step": 3955
},
{
"epoch": 0.45,
"grad_norm": 0.15423517237357112,
"learning_rate": 1.3276987730468738e-05,
"loss": 0.9794,
"step": 3960
},
{
"epoch": 0.45,
"grad_norm": 0.17474273547003924,
"learning_rate": 1.3258059068270583e-05,
"loss": 0.9848,
"step": 3965
},
{
"epoch": 0.46,
"grad_norm": 0.15365641708205371,
"learning_rate": 1.323911733731995e-05,
"loss": 1.0069,
"step": 3970
},
{
"epoch": 0.46,
"grad_norm": 0.17587475253924142,
"learning_rate": 1.3220162613596054e-05,
"loss": 1.0292,
"step": 3975
},
{
"epoch": 0.46,
"grad_norm": 0.14912819947658754,
"learning_rate": 1.3201194973130237e-05,
"loss": 1.0344,
"step": 3980
},
{
"epoch": 0.46,
"grad_norm": 0.16410892588519563,
"learning_rate": 1.318221449200565e-05,
"loss": 1.0469,
"step": 3985
},
{
"epoch": 0.46,
"grad_norm": 0.1358182010393516,
"learning_rate": 1.3163221246356951e-05,
"loss": 0.9806,
"step": 3990
},
{
"epoch": 0.46,
"grad_norm": 0.17736414108258527,
"learning_rate": 1.314421531237e-05,
"loss": 0.9211,
"step": 3995
},
{
"epoch": 0.46,
"grad_norm": 0.5296526344268576,
"learning_rate": 1.3125196766281546e-05,
"loss": 1.0051,
"step": 4000
},
{
"epoch": 0.46,
"grad_norm": 0.1485756493546117,
"learning_rate": 1.310616568437893e-05,
"loss": 0.9469,
"step": 4005
},
{
"epoch": 0.46,
"grad_norm": 0.14180550536948622,
"learning_rate": 1.3087122142999786e-05,
"loss": 0.9999,
"step": 4010
},
{
"epoch": 0.46,
"grad_norm": 0.17019971064025535,
"learning_rate": 1.3068066218531717e-05,
"loss": 1.0096,
"step": 4015
},
{
"epoch": 0.46,
"grad_norm": 0.15603608173848027,
"learning_rate": 1.3048997987412e-05,
"loss": 0.9932,
"step": 4020
},
{
"epoch": 0.46,
"grad_norm": 0.1441529491800985,
"learning_rate": 1.302991752612727e-05,
"loss": 0.9682,
"step": 4025
},
{
"epoch": 0.46,
"grad_norm": 0.1720466361142318,
"learning_rate": 1.3010824911213233e-05,
"loss": 0.9503,
"step": 4030
},
{
"epoch": 0.46,
"grad_norm": 0.1732950343280376,
"learning_rate": 1.2991720219254332e-05,
"loss": 0.9713,
"step": 4035
},
{
"epoch": 0.46,
"grad_norm": 0.1713358435364423,
"learning_rate": 1.2972603526883463e-05,
"loss": 0.9785,
"step": 4040
},
{
"epoch": 0.46,
"grad_norm": 0.16165510140919195,
"learning_rate": 1.2953474910781654e-05,
"loss": 0.9819,
"step": 4045
},
{
"epoch": 0.46,
"grad_norm": 0.1457172219259892,
"learning_rate": 1.2934334447677759e-05,
"loss": 1.0205,
"step": 4050
},
{
"epoch": 0.47,
"grad_norm": 0.14113796970427514,
"learning_rate": 1.2915182214348159e-05,
"loss": 0.9629,
"step": 4055
},
{
"epoch": 0.47,
"grad_norm": 0.13699132646468432,
"learning_rate": 1.2896018287616445e-05,
"loss": 1.0,
"step": 4060
},
{
"epoch": 0.47,
"grad_norm": 0.15694702528576743,
"learning_rate": 1.2876842744353113e-05,
"loss": 0.9611,
"step": 4065
},
{
"epoch": 0.47,
"grad_norm": 0.14964562757632074,
"learning_rate": 1.2857655661475253e-05,
"loss": 1.0261,
"step": 4070
},
{
"epoch": 0.47,
"grad_norm": 0.17828076225216555,
"learning_rate": 1.2838457115946247e-05,
"loss": 1.0127,
"step": 4075
},
{
"epoch": 0.47,
"grad_norm": 0.13826025034587244,
"learning_rate": 1.2819247184775451e-05,
"loss": 1.0079,
"step": 4080
},
{
"epoch": 0.47,
"grad_norm": 0.1709882240580372,
"learning_rate": 1.2800025945017896e-05,
"loss": 1.0168,
"step": 4085
},
{
"epoch": 0.47,
"grad_norm": 0.16657500070831102,
"learning_rate": 1.2780793473773971e-05,
"loss": 0.9967,
"step": 4090
},
{
"epoch": 0.47,
"grad_norm": 0.16655489053078543,
"learning_rate": 1.2761549848189119e-05,
"loss": 1.0223,
"step": 4095
},
{
"epoch": 0.47,
"grad_norm": 0.13298144908876694,
"learning_rate": 1.2742295145453522e-05,
"loss": 0.9972,
"step": 4100
},
{
"epoch": 0.47,
"grad_norm": 0.1516997102981016,
"learning_rate": 1.2723029442801797e-05,
"loss": 1.0165,
"step": 4105
},
{
"epoch": 0.47,
"grad_norm": 0.17333549965719228,
"learning_rate": 1.2703752817512688e-05,
"loss": 0.954,
"step": 4110
},
{
"epoch": 0.47,
"grad_norm": 0.15395508607083733,
"learning_rate": 1.2684465346908742e-05,
"loss": 0.9814,
"step": 4115
},
{
"epoch": 0.47,
"grad_norm": 0.1435650370376509,
"learning_rate": 1.2665167108356017e-05,
"loss": 0.9999,
"step": 4120
},
{
"epoch": 0.47,
"grad_norm": 0.16180239090999451,
"learning_rate": 1.2645858179263766e-05,
"loss": 0.993,
"step": 4125
},
{
"epoch": 0.47,
"grad_norm": 0.1420948785337966,
"learning_rate": 1.2626538637084114e-05,
"loss": 0.9787,
"step": 4130
},
{
"epoch": 0.47,
"grad_norm": 0.14750493093332478,
"learning_rate": 1.2607208559311761e-05,
"loss": 1.0035,
"step": 4135
},
{
"epoch": 0.48,
"grad_norm": 0.1467995264622431,
"learning_rate": 1.2587868023483671e-05,
"loss": 1.0263,
"step": 4140
},
{
"epoch": 0.48,
"grad_norm": 0.1417251408302051,
"learning_rate": 1.256851710717876e-05,
"loss": 1.0059,
"step": 4145
},
{
"epoch": 0.48,
"grad_norm": 0.20766022842530052,
"learning_rate": 1.254915588801757e-05,
"loss": 0.9553,
"step": 4150
},
{
"epoch": 0.48,
"grad_norm": 0.16963314573423616,
"learning_rate": 1.2529784443661981e-05,
"loss": 1.0005,
"step": 4155
},
{
"epoch": 0.48,
"grad_norm": 0.2091373157864917,
"learning_rate": 1.2510402851814884e-05,
"loss": 1.013,
"step": 4160
},
{
"epoch": 0.48,
"grad_norm": 0.15405021500970398,
"learning_rate": 1.2491011190219874e-05,
"loss": 0.9762,
"step": 4165
},
{
"epoch": 0.48,
"grad_norm": 0.14649023578721931,
"learning_rate": 1.2471609536660943e-05,
"loss": 1.0165,
"step": 4170
},
{
"epoch": 0.48,
"grad_norm": 0.14065669246800563,
"learning_rate": 1.2452197968962153e-05,
"loss": 0.9933,
"step": 4175
},
{
"epoch": 0.48,
"grad_norm": 0.14413563159918724,
"learning_rate": 1.243277656498734e-05,
"loss": 0.9443,
"step": 4180
},
{
"epoch": 0.48,
"grad_norm": 0.1654655225786306,
"learning_rate": 1.2413345402639798e-05,
"loss": 1.0036,
"step": 4185
},
{
"epoch": 0.48,
"grad_norm": 0.15748520117751613,
"learning_rate": 1.2393904559861954e-05,
"loss": 0.9755,
"step": 4190
},
{
"epoch": 0.48,
"grad_norm": 0.1473990244540755,
"learning_rate": 1.2374454114635077e-05,
"loss": 0.9745,
"step": 4195
},
{
"epoch": 0.48,
"grad_norm": 0.15212978129733198,
"learning_rate": 1.2354994144978947e-05,
"loss": 1.0139,
"step": 4200
},
{
"epoch": 0.48,
"grad_norm": 0.1396847513689492,
"learning_rate": 1.2335524728951543e-05,
"loss": 1.0162,
"step": 4205
},
{
"epoch": 0.48,
"grad_norm": 0.17299742875221621,
"learning_rate": 1.231604594464875e-05,
"loss": 0.9824,
"step": 4210
},
{
"epoch": 0.48,
"grad_norm": 0.1485481236268159,
"learning_rate": 1.2296557870204016e-05,
"loss": 0.9948,
"step": 4215
},
{
"epoch": 0.48,
"grad_norm": 0.14066749577420928,
"learning_rate": 1.2277060583788065e-05,
"loss": 1.0131,
"step": 4220
},
{
"epoch": 0.48,
"grad_norm": 0.15381025153069317,
"learning_rate": 1.2257554163608564e-05,
"loss": 1.0398,
"step": 4225
},
{
"epoch": 0.49,
"grad_norm": 0.14873800540056195,
"learning_rate": 1.2238038687909828e-05,
"loss": 1.0339,
"step": 4230
},
{
"epoch": 0.49,
"grad_norm": 0.1506024380670204,
"learning_rate": 1.221851423497248e-05,
"loss": 0.9867,
"step": 4235
},
{
"epoch": 0.49,
"grad_norm": 0.1410735112997371,
"learning_rate": 1.2198980883113164e-05,
"loss": 1.0048,
"step": 4240
},
{
"epoch": 0.49,
"grad_norm": 0.16228529238180794,
"learning_rate": 1.2179438710684217e-05,
"loss": 0.9644,
"step": 4245
},
{
"epoch": 0.49,
"grad_norm": 0.15257794679210096,
"learning_rate": 1.215988779607336e-05,
"loss": 0.9549,
"step": 4250
},
{
"epoch": 0.49,
"grad_norm": 0.1334233993137291,
"learning_rate": 1.2140328217703374e-05,
"loss": 1.0098,
"step": 4255
},
{
"epoch": 0.49,
"grad_norm": 0.1481883812729615,
"learning_rate": 1.2120760054031794e-05,
"loss": 0.9772,
"step": 4260
},
{
"epoch": 0.49,
"grad_norm": 0.15470254839106493,
"learning_rate": 1.2101183383550595e-05,
"loss": 0.9939,
"step": 4265
},
{
"epoch": 0.49,
"grad_norm": 0.16326943468961283,
"learning_rate": 1.2081598284785874e-05,
"loss": 1.0096,
"step": 4270
},
{
"epoch": 0.49,
"grad_norm": 0.1455309036154809,
"learning_rate": 1.2062004836297538e-05,
"loss": 0.9055,
"step": 4275
},
{
"epoch": 0.49,
"grad_norm": 0.16119677336912305,
"learning_rate": 1.204240311667898e-05,
"loss": 0.9675,
"step": 4280
},
{
"epoch": 0.49,
"grad_norm": 0.15748031735190754,
"learning_rate": 1.2022793204556775e-05,
"loss": 0.9521,
"step": 4285
},
{
"epoch": 0.49,
"grad_norm": 0.13681732561471252,
"learning_rate": 1.2003175178590358e-05,
"loss": 0.9379,
"step": 4290
},
{
"epoch": 0.49,
"grad_norm": 0.1638641503758061,
"learning_rate": 1.1983549117471708e-05,
"loss": 0.9636,
"step": 4295
},
{
"epoch": 0.49,
"grad_norm": 0.1491119508136759,
"learning_rate": 1.1963915099925048e-05,
"loss": 1.0032,
"step": 4300
},
{
"epoch": 0.49,
"grad_norm": 0.17168739691061927,
"learning_rate": 1.1944273204706496e-05,
"loss": 1.023,
"step": 4305
},
{
"epoch": 0.49,
"grad_norm": 0.1476945671591736,
"learning_rate": 1.192462351060378e-05,
"loss": 0.9962,
"step": 4310
},
{
"epoch": 0.5,
"grad_norm": 0.1496274609594804,
"learning_rate": 1.1904966096435913e-05,
"loss": 0.9698,
"step": 4315
},
{
"epoch": 0.5,
"grad_norm": 0.1601757963056008,
"learning_rate": 1.1885301041052864e-05,
"loss": 1.0003,
"step": 4320
},
{
"epoch": 0.5,
"grad_norm": 0.16422723586260526,
"learning_rate": 1.186562842333527e-05,
"loss": 0.9745,
"step": 4325
},
{
"epoch": 0.5,
"grad_norm": 0.15164475792325074,
"learning_rate": 1.1845948322194085e-05,
"loss": 1.0232,
"step": 4330
},
{
"epoch": 0.5,
"grad_norm": 0.14806086318972134,
"learning_rate": 1.1826260816570288e-05,
"loss": 0.9734,
"step": 4335
},
{
"epoch": 0.5,
"grad_norm": 0.15508915543958018,
"learning_rate": 1.1806565985434562e-05,
"loss": 1.0408,
"step": 4340
},
{
"epoch": 0.5,
"grad_norm": 0.1528517104160431,
"learning_rate": 1.1786863907786966e-05,
"loss": 0.9894,
"step": 4345
},
{
"epoch": 0.5,
"grad_norm": 0.1650502275344229,
"learning_rate": 1.1767154662656632e-05,
"loss": 0.9778,
"step": 4350
},
{
"epoch": 0.5,
"grad_norm": 0.1653287074238984,
"learning_rate": 1.1747438329101442e-05,
"loss": 1.0082,
"step": 4355
},
{
"epoch": 0.5,
"grad_norm": 0.13842183953926193,
"learning_rate": 1.172771498620771e-05,
"loss": 0.9272,
"step": 4360
},
{
"epoch": 0.5,
"grad_norm": 0.14367079342517913,
"learning_rate": 1.1707984713089869e-05,
"loss": 0.96,
"step": 4365
},
{
"epoch": 0.5,
"grad_norm": 0.15236301021955512,
"learning_rate": 1.168824758889014e-05,
"loss": 0.9935,
"step": 4370
},
{
"epoch": 0.5,
"grad_norm": 0.14942318434736018,
"learning_rate": 1.1668503692778239e-05,
"loss": 0.9731,
"step": 4375
},
{
"epoch": 0.5,
"grad_norm": 0.1503513728667334,
"learning_rate": 1.1648753103951038e-05,
"loss": 0.9741,
"step": 4380
},
{
"epoch": 0.5,
"grad_norm": 0.16642432652659186,
"learning_rate": 1.1628995901632257e-05,
"loss": 0.9393,
"step": 4385
},
{
"epoch": 0.5,
"grad_norm": 0.1485257389562868,
"learning_rate": 1.1609232165072142e-05,
"loss": 0.9577,
"step": 4390
},
{
"epoch": 0.5,
"grad_norm": 0.14498490691408294,
"learning_rate": 1.158946197354715e-05,
"loss": 0.9665,
"step": 4395
},
{
"epoch": 0.5,
"grad_norm": 0.1391572162333707,
"learning_rate": 1.1569685406359627e-05,
"loss": 0.9849,
"step": 4400
},
{
"epoch": 0.51,
"grad_norm": 0.1363357838739056,
"learning_rate": 1.1549902542837508e-05,
"loss": 0.9682,
"step": 4405
},
{
"epoch": 0.51,
"grad_norm": 0.14851175669073696,
"learning_rate": 1.1530113462333965e-05,
"loss": 0.9743,
"step": 4410
},
{
"epoch": 0.51,
"grad_norm": 0.15150315717787996,
"learning_rate": 1.1510318244227115e-05,
"loss": 1.0443,
"step": 4415
},
{
"epoch": 0.51,
"grad_norm": 0.14804690171666604,
"learning_rate": 1.1490516967919694e-05,
"loss": 0.9256,
"step": 4420
},
{
"epoch": 0.51,
"grad_norm": 0.16237757809000358,
"learning_rate": 1.1470709712838742e-05,
"loss": 1.0134,
"step": 4425
},
{
"epoch": 0.51,
"grad_norm": 0.17197755784321325,
"learning_rate": 1.1450896558435276e-05,
"loss": 0.9849,
"step": 4430
},
{
"epoch": 0.51,
"grad_norm": 0.15612773410802283,
"learning_rate": 1.1431077584183982e-05,
"loss": 0.9598,
"step": 4435
},
{
"epoch": 0.51,
"grad_norm": 0.13466732794295638,
"learning_rate": 1.1411252869582888e-05,
"loss": 0.9625,
"step": 4440
},
{
"epoch": 0.51,
"grad_norm": 0.13713530748097877,
"learning_rate": 1.1391422494153037e-05,
"loss": 0.9897,
"step": 4445
},
{
"epoch": 0.51,
"grad_norm": 0.16334826333471714,
"learning_rate": 1.13715865374382e-05,
"loss": 0.9714,
"step": 4450
},
{
"epoch": 0.51,
"grad_norm": 0.14902501867779766,
"learning_rate": 1.1351745079004523e-05,
"loss": 0.9769,
"step": 4455
},
{
"epoch": 0.51,
"grad_norm": 0.16804222405495303,
"learning_rate": 1.133189819844022e-05,
"loss": 1.0022,
"step": 4460
},
{
"epoch": 0.51,
"grad_norm": 0.1573471808966452,
"learning_rate": 1.1312045975355253e-05,
"loss": 0.9948,
"step": 4465
},
{
"epoch": 0.51,
"grad_norm": 0.1351325991746818,
"learning_rate": 1.1292188489381023e-05,
"loss": 0.9762,
"step": 4470
},
{
"epoch": 0.51,
"grad_norm": 0.14025166228074737,
"learning_rate": 1.1272325820170031e-05,
"loss": 0.9933,
"step": 4475
},
{
"epoch": 0.51,
"grad_norm": 0.15310624611604554,
"learning_rate": 1.1252458047395577e-05,
"loss": 0.9881,
"step": 4480
},
{
"epoch": 0.51,
"grad_norm": 0.15135553880037123,
"learning_rate": 1.1232585250751433e-05,
"loss": 0.9988,
"step": 4485
},
{
"epoch": 0.52,
"grad_norm": 0.12523474585811423,
"learning_rate": 1.121270750995151e-05,
"loss": 0.9292,
"step": 4490
},
{
"epoch": 0.52,
"grad_norm": 0.14936930962483547,
"learning_rate": 1.1192824904729568e-05,
"loss": 1.0397,
"step": 4495
},
{
"epoch": 0.52,
"grad_norm": 0.15744801232486724,
"learning_rate": 1.1172937514838866e-05,
"loss": 0.9702,
"step": 4500
},
{
"epoch": 0.52,
"grad_norm": 0.1735195003369352,
"learning_rate": 1.1153045420051866e-05,
"loss": 0.9946,
"step": 4505
},
{
"epoch": 0.52,
"grad_norm": 0.15084013341935223,
"learning_rate": 1.1133148700159892e-05,
"loss": 1.0113,
"step": 4510
},
{
"epoch": 0.52,
"grad_norm": 0.14328578360061417,
"learning_rate": 1.1113247434972832e-05,
"loss": 0.9604,
"step": 4515
},
{
"epoch": 0.52,
"grad_norm": 0.16104686414167652,
"learning_rate": 1.1093341704318792e-05,
"loss": 1.0124,
"step": 4520
},
{
"epoch": 0.52,
"grad_norm": 0.17211134627673846,
"learning_rate": 1.1073431588043799e-05,
"loss": 1.0079,
"step": 4525
},
{
"epoch": 0.52,
"grad_norm": 0.15016045102624198,
"learning_rate": 1.1053517166011471e-05,
"loss": 0.994,
"step": 4530
},
{
"epoch": 0.52,
"grad_norm": 0.14647371801557627,
"learning_rate": 1.10335985181027e-05,
"loss": 0.99,
"step": 4535
},
{
"epoch": 0.52,
"grad_norm": 0.15401688749011067,
"learning_rate": 1.101367572421532e-05,
"loss": 1.0032,
"step": 4540
},
{
"epoch": 0.52,
"grad_norm": 0.12425432198566704,
"learning_rate": 1.0993748864263802e-05,
"loss": 0.9464,
"step": 4545
},
{
"epoch": 0.52,
"grad_norm": 0.14348491126344948,
"learning_rate": 1.0973818018178925e-05,
"loss": 0.9962,
"step": 4550
},
{
"epoch": 0.52,
"grad_norm": 0.15787722063955972,
"learning_rate": 1.095388326590746e-05,
"loss": 0.9868,
"step": 4555
},
{
"epoch": 0.52,
"grad_norm": 0.14647008306231324,
"learning_rate": 1.0933944687411839e-05,
"loss": 0.986,
"step": 4560
},
{
"epoch": 0.52,
"grad_norm": 0.1414439674967068,
"learning_rate": 1.0914002362669853e-05,
"loss": 0.9808,
"step": 4565
},
{
"epoch": 0.52,
"grad_norm": 0.15491127061526816,
"learning_rate": 1.0894056371674311e-05,
"loss": 1.0293,
"step": 4570
},
{
"epoch": 0.52,
"grad_norm": 0.2095567923448607,
"learning_rate": 1.0874106794432728e-05,
"loss": 1.0149,
"step": 4575
},
{
"epoch": 0.53,
"grad_norm": 0.14373084798659963,
"learning_rate": 1.0854153710967011e-05,
"loss": 0.9837,
"step": 4580
},
{
"epoch": 0.53,
"grad_norm": 0.1495493474755723,
"learning_rate": 1.0834197201313128e-05,
"loss": 0.9613,
"step": 4585
},
{
"epoch": 0.53,
"grad_norm": 0.14050436108360081,
"learning_rate": 1.0814237345520785e-05,
"loss": 0.9763,
"step": 4590
},
{
"epoch": 0.53,
"grad_norm": 0.2203173180254517,
"learning_rate": 1.079427422365312e-05,
"loss": 0.9877,
"step": 4595
},
{
"epoch": 0.53,
"grad_norm": 0.14846974819518338,
"learning_rate": 1.0774307915786361e-05,
"loss": 1.0278,
"step": 4600
},
{
"epoch": 0.53,
"grad_norm": 0.15972382233862675,
"learning_rate": 1.0754338502009525e-05,
"loss": 1.0394,
"step": 4605
},
{
"epoch": 0.53,
"grad_norm": 0.14205543455092337,
"learning_rate": 1.0734366062424082e-05,
"loss": 0.9532,
"step": 4610
},
{
"epoch": 0.53,
"grad_norm": 0.16257627416094947,
"learning_rate": 1.0714390677143642e-05,
"loss": 0.9888,
"step": 4615
},
{
"epoch": 0.53,
"grad_norm": 0.15463838056129522,
"learning_rate": 1.0694412426293625e-05,
"loss": 0.9901,
"step": 4620
},
{
"epoch": 0.53,
"grad_norm": 0.157473743731946,
"learning_rate": 1.0674431390010958e-05,
"loss": 0.9738,
"step": 4625
},
{
"epoch": 0.53,
"grad_norm": 0.17185346360347878,
"learning_rate": 1.0654447648443724e-05,
"loss": 0.977,
"step": 4630
},
{
"epoch": 0.53,
"grad_norm": 0.1669057831363489,
"learning_rate": 1.0634461281750874e-05,
"loss": 0.9659,
"step": 4635
},
{
"epoch": 0.53,
"grad_norm": 0.15756537418235952,
"learning_rate": 1.0614472370101874e-05,
"loss": 0.9527,
"step": 4640
},
{
"epoch": 0.53,
"grad_norm": 0.15710413690275918,
"learning_rate": 1.0594480993676412e-05,
"loss": 0.9246,
"step": 4645
},
{
"epoch": 0.53,
"grad_norm": 0.15049007080290014,
"learning_rate": 1.0574487232664055e-05,
"loss": 0.967,
"step": 4650
},
{
"epoch": 0.53,
"grad_norm": 0.16424813052911172,
"learning_rate": 1.0554491167263934e-05,
"loss": 0.9872,
"step": 4655
},
{
"epoch": 0.53,
"grad_norm": 0.13533485006418303,
"learning_rate": 1.0534492877684425e-05,
"loss": 0.988,
"step": 4660
},
{
"epoch": 0.54,
"grad_norm": 0.14540963578112445,
"learning_rate": 1.051449244414283e-05,
"loss": 0.9769,
"step": 4665
},
{
"epoch": 0.54,
"grad_norm": 0.16544658930091816,
"learning_rate": 1.0494489946865047e-05,
"loss": 1.0061,
"step": 4670
},
{
"epoch": 0.54,
"grad_norm": 0.15419620797124292,
"learning_rate": 1.0474485466085251e-05,
"loss": 0.9781,
"step": 4675
},
{
"epoch": 0.54,
"grad_norm": 0.1480980258764122,
"learning_rate": 1.0454479082045572e-05,
"loss": 1.0045,
"step": 4680
},
{
"epoch": 0.54,
"grad_norm": 0.1319121901929664,
"learning_rate": 1.0434470874995781e-05,
"loss": 1.0011,
"step": 4685
},
{
"epoch": 0.54,
"grad_norm": 0.16294310053411337,
"learning_rate": 1.0414460925192957e-05,
"loss": 1.0355,
"step": 4690
},
{
"epoch": 0.54,
"grad_norm": 0.14002433497408856,
"learning_rate": 1.039444931290117e-05,
"loss": 0.9583,
"step": 4695
},
{
"epoch": 0.54,
"grad_norm": 0.14598774446970095,
"learning_rate": 1.0374436118391156e-05,
"loss": 1.02,
"step": 4700
},
{
"epoch": 0.54,
"grad_norm": 0.17054803538911295,
"learning_rate": 1.0354421421940002e-05,
"loss": 0.9715,
"step": 4705
},
{
"epoch": 0.54,
"grad_norm": 0.15798536121334328,
"learning_rate": 1.0334405303830817e-05,
"loss": 0.988,
"step": 4710
},
{
"epoch": 0.54,
"grad_norm": 0.15812352855721043,
"learning_rate": 1.0314387844352413e-05,
"loss": 0.9882,
"step": 4715
},
{
"epoch": 0.54,
"grad_norm": 0.1760777524893954,
"learning_rate": 1.029436912379899e-05,
"loss": 1.0047,
"step": 4720
},
{
"epoch": 0.54,
"grad_norm": 0.15269266048039007,
"learning_rate": 1.0274349222469788e-05,
"loss": 1.0045,
"step": 4725
},
{
"epoch": 0.54,
"grad_norm": 0.14681913980842487,
"learning_rate": 1.02543282206688e-05,
"loss": 1.0152,
"step": 4730
},
{
"epoch": 0.54,
"grad_norm": 0.17079280497983096,
"learning_rate": 1.0234306198704428e-05,
"loss": 1.0272,
"step": 4735
},
{
"epoch": 0.54,
"grad_norm": 0.22616437447155613,
"learning_rate": 1.0214283236889169e-05,
"loss": 1.008,
"step": 4740
},
{
"epoch": 0.54,
"grad_norm": 0.1699898357871866,
"learning_rate": 1.0194259415539281e-05,
"loss": 0.9582,
"step": 4745
},
{
"epoch": 0.55,
"grad_norm": 0.16501284429199523,
"learning_rate": 1.0174234814974478e-05,
"loss": 0.9327,
"step": 4750
},
{
"epoch": 0.55,
"grad_norm": 0.14390406797469468,
"learning_rate": 1.0154209515517599e-05,
"loss": 0.9547,
"step": 4755
},
{
"epoch": 0.55,
"grad_norm": 0.13923695783946935,
"learning_rate": 1.013418359749428e-05,
"loss": 0.9366,
"step": 4760
},
{
"epoch": 0.55,
"grad_norm": 0.1665568281121691,
"learning_rate": 1.011415714123265e-05,
"loss": 1.0001,
"step": 4765
},
{
"epoch": 0.55,
"grad_norm": 0.14948678997856832,
"learning_rate": 1.0094130227062986e-05,
"loss": 0.9805,
"step": 4770
},
{
"epoch": 0.55,
"grad_norm": 0.155662172571138,
"learning_rate": 1.0074102935317402e-05,
"loss": 0.9379,
"step": 4775
},
{
"epoch": 0.55,
"grad_norm": 0.1536058019155457,
"learning_rate": 1.005407534632954e-05,
"loss": 0.992,
"step": 4780
},
{
"epoch": 0.55,
"grad_norm": 0.1529728135576244,
"learning_rate": 1.0034047540434211e-05,
"loss": 0.9668,
"step": 4785
},
{
"epoch": 0.55,
"grad_norm": 0.13414403702329025,
"learning_rate": 1.001401959796712e-05,
"loss": 0.987,
"step": 4790
},
{
"epoch": 0.55,
"grad_norm": 0.13452626961263964,
"learning_rate": 9.993991599264506e-06,
"loss": 0.9444,
"step": 4795
},
{
"epoch": 0.55,
"grad_norm": 0.16505817547042026,
"learning_rate": 9.97396362466284e-06,
"loss": 0.9685,
"step": 4800
},
{
"epoch": 0.55,
"grad_norm": 0.16308471898366522,
"learning_rate": 9.953935754498484e-06,
"loss": 1.0164,
"step": 4805
},
{
"epoch": 0.55,
"grad_norm": 0.1567129781661088,
"learning_rate": 9.933908069107397e-06,
"loss": 0.9935,
"step": 4810
},
{
"epoch": 0.55,
"grad_norm": 0.16325836607492852,
"learning_rate": 9.913880648824793e-06,
"loss": 0.9558,
"step": 4815
},
{
"epoch": 0.55,
"grad_norm": 0.14862892665763258,
"learning_rate": 9.893853573984814e-06,
"loss": 0.971,
"step": 4820
},
{
"epoch": 0.55,
"grad_norm": 0.1397478333828406,
"learning_rate": 9.873826924920229e-06,
"loss": 1.0398,
"step": 4825
},
{
"epoch": 0.55,
"grad_norm": 0.17375307566696258,
"learning_rate": 9.853800781962081e-06,
"loss": 1.0116,
"step": 4830
},
{
"epoch": 0.55,
"grad_norm": 0.13961951652778562,
"learning_rate": 9.833775225439403e-06,
"loss": 0.988,
"step": 4835
},
{
"epoch": 0.56,
"grad_norm": 0.1735715103872482,
"learning_rate": 9.813750335678866e-06,
"loss": 0.9929,
"step": 4840
},
{
"epoch": 0.56,
"grad_norm": 0.161333261837155,
"learning_rate": 9.793726193004463e-06,
"loss": 0.9744,
"step": 4845
},
{
"epoch": 0.56,
"grad_norm": 0.151411122286772,
"learning_rate": 9.773702877737198e-06,
"loss": 0.9793,
"step": 4850
},
{
"epoch": 0.56,
"grad_norm": 0.12961908613659245,
"learning_rate": 9.753680470194757e-06,
"loss": 0.9251,
"step": 4855
},
{
"epoch": 0.56,
"grad_norm": 0.1396913275936924,
"learning_rate": 9.73365905069117e-06,
"loss": 0.9509,
"step": 4860
},
{
"epoch": 0.56,
"grad_norm": 0.16315659722720055,
"learning_rate": 9.713638699536524e-06,
"loss": 1.0024,
"step": 4865
},
{
"epoch": 0.56,
"grad_norm": 0.14949442678959507,
"learning_rate": 9.693619497036604e-06,
"loss": 0.9525,
"step": 4870
},
{
"epoch": 0.56,
"grad_norm": 0.1509161402687968,
"learning_rate": 9.673601523492597e-06,
"loss": 0.969,
"step": 4875
},
{
"epoch": 0.56,
"grad_norm": 0.18884499178448425,
"learning_rate": 9.653584859200762e-06,
"loss": 0.9821,
"step": 4880
},
{
"epoch": 0.56,
"grad_norm": 0.14943384616494287,
"learning_rate": 9.633569584452096e-06,
"loss": 0.9981,
"step": 4885
},
{
"epoch": 0.56,
"grad_norm": 0.14718426517579003,
"learning_rate": 9.613555779532032e-06,
"loss": 1.0126,
"step": 4890
},
{
"epoch": 0.56,
"grad_norm": 0.14346964739274728,
"learning_rate": 9.5935435247201e-06,
"loss": 0.9873,
"step": 4895
},
{
"epoch": 0.56,
"grad_norm": 0.14269118083814264,
"learning_rate": 9.57353290028962e-06,
"loss": 0.9689,
"step": 4900
},
{
"epoch": 0.56,
"grad_norm": 0.1452735891214661,
"learning_rate": 9.55352398650737e-06,
"loss": 0.9855,
"step": 4905
},
{
"epoch": 0.56,
"grad_norm": 0.1512469411100847,
"learning_rate": 9.533516863633258e-06,
"loss": 1.0124,
"step": 4910
},
{
"epoch": 0.56,
"grad_norm": 0.1588613180012733,
"learning_rate": 9.513511611920016e-06,
"loss": 0.9554,
"step": 4915
},
{
"epoch": 0.56,
"grad_norm": 0.13919230680904357,
"learning_rate": 9.493508311612874e-06,
"loss": 0.9619,
"step": 4920
},
{
"epoch": 0.57,
"grad_norm": 0.25231214027906207,
"learning_rate": 9.473507042949223e-06,
"loss": 0.94,
"step": 4925
},
{
"epoch": 0.57,
"grad_norm": 0.17506179266894803,
"learning_rate": 9.45350788615832e-06,
"loss": 0.995,
"step": 4930
},
{
"epoch": 0.57,
"grad_norm": 0.14659385985042395,
"learning_rate": 9.43351092146093e-06,
"loss": 1.0203,
"step": 4935
},
{
"epoch": 0.57,
"grad_norm": 0.16325214031455623,
"learning_rate": 9.413516229069045e-06,
"loss": 1.0016,
"step": 4940
},
{
"epoch": 0.57,
"grad_norm": 0.14967891818616197,
"learning_rate": 9.393523889185535e-06,
"loss": 0.9453,
"step": 4945
},
{
"epoch": 0.57,
"grad_norm": 0.1632689714187689,
"learning_rate": 9.373533982003827e-06,
"loss": 0.9679,
"step": 4950
},
{
"epoch": 0.57,
"grad_norm": 0.15005729699906664,
"learning_rate": 9.353546587707605e-06,
"loss": 0.9841,
"step": 4955
},
{
"epoch": 0.57,
"grad_norm": 0.1538189394310828,
"learning_rate": 9.333561786470454e-06,
"loss": 0.9865,
"step": 4960
},
{
"epoch": 0.57,
"grad_norm": 0.15335965458089096,
"learning_rate": 9.313579658455572e-06,
"loss": 0.9598,
"step": 4965
},
{
"epoch": 0.57,
"grad_norm": 0.14878012061335097,
"learning_rate": 9.293600283815433e-06,
"loss": 0.99,
"step": 4970
},
{
"epoch": 0.57,
"grad_norm": 0.1366362316176452,
"learning_rate": 9.27362374269146e-06,
"loss": 0.9912,
"step": 4975
},
{
"epoch": 0.57,
"grad_norm": 0.15412663482980257,
"learning_rate": 9.253650115213719e-06,
"loss": 1.0154,
"step": 4980
},
{
"epoch": 0.57,
"grad_norm": 0.15719848318968258,
"learning_rate": 9.233679481500575e-06,
"loss": 0.9475,
"step": 4985
},
{
"epoch": 0.57,
"grad_norm": 0.16570261853854998,
"learning_rate": 9.213711921658398e-06,
"loss": 0.9678,
"step": 4990
},
{
"epoch": 0.57,
"grad_norm": 0.14303649866416773,
"learning_rate": 9.193747515781225e-06,
"loss": 1.0032,
"step": 4995
},
{
"epoch": 0.57,
"grad_norm": 0.14908836332689762,
"learning_rate": 9.173786343950436e-06,
"loss": 0.9861,
"step": 5000
},
{
"epoch": 0.57,
"grad_norm": 0.14677058351546596,
"learning_rate": 9.153828486234442e-06,
"loss": 0.9657,
"step": 5005
},
{
"epoch": 0.57,
"grad_norm": 0.16735853358172254,
"learning_rate": 9.133874022688368e-06,
"loss": 0.9787,
"step": 5010
},
{
"epoch": 0.58,
"grad_norm": 0.14392109266937986,
"learning_rate": 9.113923033353708e-06,
"loss": 0.9855,
"step": 5015
},
{
"epoch": 0.58,
"grad_norm": 0.13749767713833144,
"learning_rate": 9.093975598258035e-06,
"loss": 0.9408,
"step": 5020
},
{
"epoch": 0.58,
"grad_norm": 0.13844256507999125,
"learning_rate": 9.074031797414658e-06,
"loss": 0.9672,
"step": 5025
},
{
"epoch": 0.58,
"grad_norm": 0.1505903550368753,
"learning_rate": 9.054091710822306e-06,
"loss": 0.9503,
"step": 5030
},
{
"epoch": 0.58,
"grad_norm": 0.1432964826403832,
"learning_rate": 9.034155418464823e-06,
"loss": 0.9544,
"step": 5035
},
{
"epoch": 0.58,
"grad_norm": 0.14913143397603465,
"learning_rate": 9.014223000310814e-06,
"loss": 0.9893,
"step": 5040
},
{
"epoch": 0.58,
"grad_norm": 0.14355161784836412,
"learning_rate": 8.994294536313356e-06,
"loss": 0.9685,
"step": 5045
},
{
"epoch": 0.58,
"grad_norm": 0.15130767601658926,
"learning_rate": 8.974370106409664e-06,
"loss": 0.9822,
"step": 5050
},
{
"epoch": 0.58,
"grad_norm": 0.1659015928013058,
"learning_rate": 8.954449790520767e-06,
"loss": 0.9902,
"step": 5055
},
{
"epoch": 0.58,
"grad_norm": 0.279115701281848,
"learning_rate": 8.934533668551202e-06,
"loss": 0.9904,
"step": 5060
},
{
"epoch": 0.58,
"grad_norm": 0.13907419466412566,
"learning_rate": 8.914621820388667e-06,
"loss": 0.9937,
"step": 5065
},
{
"epoch": 0.58,
"grad_norm": 0.15118661193712304,
"learning_rate": 8.89471432590373e-06,
"loss": 0.991,
"step": 5070
},
{
"epoch": 0.58,
"grad_norm": 0.16086086936369975,
"learning_rate": 8.874811264949494e-06,
"loss": 0.9687,
"step": 5075
},
{
"epoch": 0.58,
"grad_norm": 0.15362628456938304,
"learning_rate": 8.85491271736127e-06,
"loss": 0.9793,
"step": 5080
},
{
"epoch": 0.58,
"grad_norm": 0.132616393183217,
"learning_rate": 8.835018762956278e-06,
"loss": 0.9254,
"step": 5085
},
{
"epoch": 0.58,
"grad_norm": 0.13187561052179048,
"learning_rate": 8.815129481533296e-06,
"loss": 0.9828,
"step": 5090
},
{
"epoch": 0.58,
"grad_norm": 0.15137852472294444,
"learning_rate": 8.795244952872376e-06,
"loss": 0.9827,
"step": 5095
},
{
"epoch": 0.59,
"grad_norm": 0.1477255563681887,
"learning_rate": 8.775365256734497e-06,
"loss": 1.0086,
"step": 5100
},
{
"epoch": 0.59,
"grad_norm": 0.14825159316658662,
"learning_rate": 8.755490472861249e-06,
"loss": 0.9451,
"step": 5105
},
{
"epoch": 0.59,
"grad_norm": 0.15240480505538903,
"learning_rate": 8.735620680974535e-06,
"loss": 1.0283,
"step": 5110
},
{
"epoch": 0.59,
"grad_norm": 0.16020471219051496,
"learning_rate": 8.715755960776208e-06,
"loss": 0.9489,
"step": 5115
},
{
"epoch": 0.59,
"grad_norm": 0.14075989592719426,
"learning_rate": 8.695896391947801e-06,
"loss": 1.0015,
"step": 5120
},
{
"epoch": 0.59,
"grad_norm": 0.16294134847877653,
"learning_rate": 8.676042054150174e-06,
"loss": 0.9506,
"step": 5125
},
{
"epoch": 0.59,
"grad_norm": 0.14782698521008092,
"learning_rate": 8.656193027023202e-06,
"loss": 0.9854,
"step": 5130
},
{
"epoch": 0.59,
"grad_norm": 0.17230717196439183,
"learning_rate": 8.63634939018546e-06,
"loss": 0.9588,
"step": 5135
},
{
"epoch": 0.59,
"grad_norm": 0.15433067984559318,
"learning_rate": 8.616511223233908e-06,
"loss": 1.0119,
"step": 5140
},
{
"epoch": 0.59,
"grad_norm": 0.17938360493297748,
"learning_rate": 8.596678605743552e-06,
"loss": 0.9882,
"step": 5145
},
{
"epoch": 0.59,
"grad_norm": 0.15626840006847958,
"learning_rate": 8.576851617267151e-06,
"loss": 1.0036,
"step": 5150
},
{
"epoch": 0.59,
"grad_norm": 0.1609284379141364,
"learning_rate": 8.557030337334874e-06,
"loss": 0.9905,
"step": 5155
},
{
"epoch": 0.59,
"grad_norm": 0.1792796743233978,
"learning_rate": 8.537214845454e-06,
"loss": 0.9869,
"step": 5160
},
{
"epoch": 0.59,
"grad_norm": 0.1478004851716894,
"learning_rate": 8.51740522110859e-06,
"loss": 0.9407,
"step": 5165
},
{
"epoch": 0.59,
"grad_norm": 0.1711436302530027,
"learning_rate": 8.497601543759163e-06,
"loss": 1.0088,
"step": 5170
},
{
"epoch": 0.59,
"grad_norm": 0.15473145302958452,
"learning_rate": 8.477803892842388e-06,
"loss": 1.0731,
"step": 5175
},
{
"epoch": 0.59,
"grad_norm": 0.144296243379969,
"learning_rate": 8.458012347770761e-06,
"loss": 0.9975,
"step": 5180
},
{
"epoch": 0.59,
"grad_norm": 0.17501178995030744,
"learning_rate": 8.438226987932283e-06,
"loss": 1.0172,
"step": 5185
},
{
"epoch": 0.6,
"grad_norm": 0.14730519462646782,
"learning_rate": 8.418447892690155e-06,
"loss": 0.9702,
"step": 5190
},
{
"epoch": 0.6,
"grad_norm": 0.16210116921169443,
"learning_rate": 8.398675141382428e-06,
"loss": 0.9561,
"step": 5195
},
{
"epoch": 0.6,
"grad_norm": 0.15484285425579644,
"learning_rate": 8.37890881332173e-06,
"loss": 0.9841,
"step": 5200
},
{
"epoch": 0.6,
"grad_norm": 0.15242888841867855,
"learning_rate": 8.359148987794911e-06,
"loss": 1.0019,
"step": 5205
},
{
"epoch": 0.6,
"grad_norm": 0.14800116150906875,
"learning_rate": 8.339395744062736e-06,
"loss": 0.9709,
"step": 5210
},
{
"epoch": 0.6,
"grad_norm": 0.14330824227437114,
"learning_rate": 8.319649161359585e-06,
"loss": 0.9626,
"step": 5215
},
{
"epoch": 0.6,
"grad_norm": 0.17989850320118725,
"learning_rate": 8.299909318893093e-06,
"loss": 1.0095,
"step": 5220
},
{
"epoch": 0.6,
"grad_norm": 0.16440923586227832,
"learning_rate": 8.28017629584388e-06,
"loss": 0.9952,
"step": 5225
},
{
"epoch": 0.6,
"grad_norm": 0.14348214952317598,
"learning_rate": 8.260450171365209e-06,
"loss": 0.9775,
"step": 5230
},
{
"epoch": 0.6,
"grad_norm": 0.17269857804858033,
"learning_rate": 8.240731024582663e-06,
"loss": 0.9974,
"step": 5235
},
{
"epoch": 0.6,
"grad_norm": 0.1614298283931525,
"learning_rate": 8.221018934593845e-06,
"loss": 0.9814,
"step": 5240
},
{
"epoch": 0.6,
"grad_norm": 0.14285379790301184,
"learning_rate": 8.20131398046804e-06,
"loss": 1.0027,
"step": 5245
},
{
"epoch": 0.6,
"grad_norm": 0.1514218295811995,
"learning_rate": 8.181616241245922e-06,
"loss": 1.0202,
"step": 5250
},
{
"epoch": 0.6,
"grad_norm": 0.161603998329929,
"learning_rate": 8.16192579593922e-06,
"loss": 1.0082,
"step": 5255
},
{
"epoch": 0.6,
"grad_norm": 0.16842415660081853,
"learning_rate": 8.142242723530402e-06,
"loss": 0.9834,
"step": 5260
},
{
"epoch": 0.6,
"grad_norm": 0.1457561912310776,
"learning_rate": 8.12256710297237e-06,
"loss": 0.9782,
"step": 5265
},
{
"epoch": 0.6,
"grad_norm": 0.1641518398035216,
"learning_rate": 8.10289901318812e-06,
"loss": 1.0271,
"step": 5270
},
{
"epoch": 0.61,
"grad_norm": 0.1427288449992825,
"learning_rate": 8.083238533070461e-06,
"loss": 0.9834,
"step": 5275
},
{
"epoch": 0.61,
"grad_norm": 0.15917425490481338,
"learning_rate": 8.063585741481662e-06,
"loss": 1.0123,
"step": 5280
},
{
"epoch": 0.61,
"grad_norm": 0.18132122105623633,
"learning_rate": 8.043940717253157e-06,
"loss": 0.936,
"step": 5285
},
{
"epoch": 0.61,
"grad_norm": 0.14085230120806444,
"learning_rate": 8.024303539185225e-06,
"loss": 0.9887,
"step": 5290
},
{
"epoch": 0.61,
"grad_norm": 0.1588469763534457,
"learning_rate": 8.004674286046676e-06,
"loss": 1.0133,
"step": 5295
},
{
"epoch": 0.61,
"grad_norm": 0.1777549186427469,
"learning_rate": 7.985053036574516e-06,
"loss": 0.9889,
"step": 5300
},
{
"epoch": 0.61,
"grad_norm": 0.14784080375679456,
"learning_rate": 7.965439869473664e-06,
"loss": 0.9854,
"step": 5305
},
{
"epoch": 0.61,
"grad_norm": 0.15511836834003834,
"learning_rate": 7.94583486341661e-06,
"loss": 1.0,
"step": 5310
},
{
"epoch": 0.61,
"grad_norm": 0.13808570872287415,
"learning_rate": 7.926238097043109e-06,
"loss": 0.9784,
"step": 5315
},
{
"epoch": 0.61,
"grad_norm": 0.13828695649291495,
"learning_rate": 7.906649648959875e-06,
"loss": 0.9549,
"step": 5320
},
{
"epoch": 0.61,
"grad_norm": 0.1518719654753116,
"learning_rate": 7.887069597740236e-06,
"loss": 0.9991,
"step": 5325
},
{
"epoch": 0.61,
"grad_norm": 0.20469308889897195,
"learning_rate": 7.867498021923854e-06,
"loss": 0.9483,
"step": 5330
},
{
"epoch": 0.61,
"grad_norm": 0.14557444227074315,
"learning_rate": 7.847935000016389e-06,
"loss": 0.9652,
"step": 5335
},
{
"epoch": 0.61,
"grad_norm": 0.14514159511658356,
"learning_rate": 7.82838061048919e-06,
"loss": 1.0003,
"step": 5340
},
{
"epoch": 0.61,
"grad_norm": 0.133079137886686,
"learning_rate": 7.808834931778985e-06,
"loss": 1.0082,
"step": 5345
},
{
"epoch": 0.61,
"grad_norm": 0.14492771934825913,
"learning_rate": 7.789298042287547e-06,
"loss": 0.991,
"step": 5350
},
{
"epoch": 0.61,
"grad_norm": 0.1544443241966409,
"learning_rate": 7.76977002038141e-06,
"loss": 0.9894,
"step": 5355
},
{
"epoch": 0.62,
"grad_norm": 0.22987284477129463,
"learning_rate": 7.750250944391528e-06,
"loss": 0.9803,
"step": 5360
},
{
"epoch": 0.62,
"grad_norm": 0.17557961742866732,
"learning_rate": 7.730740892612974e-06,
"loss": 0.9753,
"step": 5365
},
{
"epoch": 0.62,
"grad_norm": 0.14803910215555913,
"learning_rate": 7.711239943304628e-06,
"loss": 0.9966,
"step": 5370
},
{
"epoch": 0.62,
"grad_norm": 0.16258995491992698,
"learning_rate": 7.691748174688845e-06,
"loss": 0.9911,
"step": 5375
},
{
"epoch": 0.62,
"grad_norm": 0.14383742236889785,
"learning_rate": 7.672265664951166e-06,
"loss": 0.9792,
"step": 5380
},
{
"epoch": 0.62,
"grad_norm": 0.1590583110030591,
"learning_rate": 7.65279249223999e-06,
"loss": 0.9972,
"step": 5385
},
{
"epoch": 0.62,
"grad_norm": 0.14709674725626556,
"learning_rate": 7.633328734666261e-06,
"loss": 0.9672,
"step": 5390
},
{
"epoch": 0.62,
"grad_norm": 0.14038131387269753,
"learning_rate": 7.613874470303165e-06,
"loss": 1.0277,
"step": 5395
},
{
"epoch": 0.62,
"grad_norm": 0.14702776262265455,
"learning_rate": 7.594429777185792e-06,
"loss": 0.9804,
"step": 5400
},
{
"epoch": 0.62,
"grad_norm": 0.15497999781089547,
"learning_rate": 7.574994733310854e-06,
"loss": 0.9701,
"step": 5405
},
{
"epoch": 0.62,
"grad_norm": 0.1464797447039427,
"learning_rate": 7.555569416636356e-06,
"loss": 1.0321,
"step": 5410
},
{
"epoch": 0.62,
"grad_norm": 0.1448780726732443,
"learning_rate": 7.5361539050812795e-06,
"loss": 0.9717,
"step": 5415
},
{
"epoch": 0.62,
"grad_norm": 0.144477997201547,
"learning_rate": 7.5167482765252805e-06,
"loss": 0.9504,
"step": 5420
},
{
"epoch": 0.62,
"grad_norm": 0.14679907622299043,
"learning_rate": 7.497352608808376e-06,
"loss": 0.9448,
"step": 5425
},
{
"epoch": 0.62,
"grad_norm": 0.14245243667635163,
"learning_rate": 7.477966979730613e-06,
"loss": 0.9517,
"step": 5430
},
{
"epoch": 0.62,
"grad_norm": 0.14440247801852485,
"learning_rate": 7.458591467051789e-06,
"loss": 1.0102,
"step": 5435
},
{
"epoch": 0.62,
"grad_norm": 0.1551242104275944,
"learning_rate": 7.43922614849111e-06,
"loss": 0.9726,
"step": 5440
},
{
"epoch": 0.62,
"grad_norm": 0.1587874985801824,
"learning_rate": 7.4198711017268966e-06,
"loss": 0.9954,
"step": 5445
},
{
"epoch": 0.63,
"grad_norm": 0.14607029810196093,
"learning_rate": 7.4005264043962735e-06,
"loss": 0.9619,
"step": 5450
},
{
"epoch": 0.63,
"grad_norm": 0.14304579760186378,
"learning_rate": 7.381192134094833e-06,
"loss": 0.9739,
"step": 5455
},
{
"epoch": 0.63,
"grad_norm": 0.16088690012358306,
"learning_rate": 7.361868368376364e-06,
"loss": 1.0368,
"step": 5460
},
{
"epoch": 0.63,
"grad_norm": 0.15739831273345975,
"learning_rate": 7.342555184752502e-06,
"loss": 0.9773,
"step": 5465
},
{
"epoch": 0.63,
"grad_norm": 0.17052648974139123,
"learning_rate": 7.323252660692448e-06,
"loss": 0.9699,
"step": 5470
},
{
"epoch": 0.63,
"grad_norm": 0.15882652484770424,
"learning_rate": 7.3039608736226435e-06,
"loss": 0.9406,
"step": 5475
},
{
"epoch": 0.63,
"grad_norm": 0.14284068634662567,
"learning_rate": 7.284679900926448e-06,
"loss": 0.9916,
"step": 5480
},
{
"epoch": 0.63,
"grad_norm": 0.14216871015452523,
"learning_rate": 7.265409819943861e-06,
"loss": 1.0457,
"step": 5485
},
{
"epoch": 0.63,
"grad_norm": 0.15667413291002677,
"learning_rate": 7.24615070797118e-06,
"loss": 1.0025,
"step": 5490
},
{
"epoch": 0.63,
"grad_norm": 0.15141607719736466,
"learning_rate": 7.226902642260711e-06,
"loss": 0.9367,
"step": 5495
},
{
"epoch": 0.63,
"grad_norm": 0.1459328673698501,
"learning_rate": 7.207665700020452e-06,
"loss": 0.9545,
"step": 5500
},
{
"epoch": 0.63,
"grad_norm": 0.14855103153215518,
"learning_rate": 7.188439958413771e-06,
"loss": 0.995,
"step": 5505
},
{
"epoch": 0.63,
"grad_norm": 0.14583298532549416,
"learning_rate": 7.16922549455912e-06,
"loss": 1.0213,
"step": 5510
},
{
"epoch": 0.63,
"grad_norm": 0.1660327762628555,
"learning_rate": 7.150022385529711e-06,
"loss": 1.0075,
"step": 5515
},
{
"epoch": 0.63,
"grad_norm": 0.17128062649209386,
"learning_rate": 7.130830708353205e-06,
"loss": 0.9976,
"step": 5520
},
{
"epoch": 0.63,
"grad_norm": 0.14875911294165256,
"learning_rate": 7.111650540011414e-06,
"loss": 0.991,
"step": 5525
},
{
"epoch": 0.63,
"grad_norm": 0.15890984772837433,
"learning_rate": 7.092481957439976e-06,
"loss": 0.9788,
"step": 5530
},
{
"epoch": 0.64,
"grad_norm": 0.1608222722834038,
"learning_rate": 7.073325037528066e-06,
"loss": 0.9541,
"step": 5535
},
{
"epoch": 0.64,
"grad_norm": 0.16406131873366578,
"learning_rate": 7.054179857118075e-06,
"loss": 0.9781,
"step": 5540
},
{
"epoch": 0.64,
"grad_norm": 0.14968895280856412,
"learning_rate": 7.0350464930052994e-06,
"loss": 0.9717,
"step": 5545
},
{
"epoch": 0.64,
"grad_norm": 0.14952310355850104,
"learning_rate": 7.015925021937649e-06,
"loss": 0.9419,
"step": 5550
},
{
"epoch": 0.64,
"grad_norm": 0.15974595952405887,
"learning_rate": 6.9968155206153085e-06,
"loss": 0.957,
"step": 5555
},
{
"epoch": 0.64,
"grad_norm": 0.1814693520174185,
"learning_rate": 6.977718065690471e-06,
"loss": 0.9883,
"step": 5560
},
{
"epoch": 0.64,
"grad_norm": 0.1472629551922816,
"learning_rate": 6.958632733766999e-06,
"loss": 0.9625,
"step": 5565
},
{
"epoch": 0.64,
"grad_norm": 0.14950957898121478,
"learning_rate": 6.939559601400125e-06,
"loss": 0.9745,
"step": 5570
},
{
"epoch": 0.64,
"grad_norm": 0.16584859974742167,
"learning_rate": 6.920498745096149e-06,
"loss": 1.0041,
"step": 5575
},
{
"epoch": 0.64,
"grad_norm": 0.16418338414514283,
"learning_rate": 6.901450241312138e-06,
"loss": 0.9864,
"step": 5580
},
{
"epoch": 0.64,
"grad_norm": 0.1438306767559837,
"learning_rate": 6.88241416645559e-06,
"loss": 1.0129,
"step": 5585
},
{
"epoch": 0.64,
"grad_norm": 0.18471446407680997,
"learning_rate": 6.863390596884168e-06,
"loss": 0.9925,
"step": 5590
},
{
"epoch": 0.64,
"grad_norm": 0.15508912959026236,
"learning_rate": 6.844379608905363e-06,
"loss": 1.0001,
"step": 5595
},
{
"epoch": 0.64,
"grad_norm": 0.15915885491400988,
"learning_rate": 6.8253812787762e-06,
"loss": 0.9938,
"step": 5600
},
{
"epoch": 0.64,
"grad_norm": 0.15499851201185355,
"learning_rate": 6.806395682702937e-06,
"loss": 1.0216,
"step": 5605
},
{
"epoch": 0.64,
"grad_norm": 0.16736463018683265,
"learning_rate": 6.787422896840744e-06,
"loss": 1.0287,
"step": 5610
},
{
"epoch": 0.64,
"grad_norm": 0.18987709656170776,
"learning_rate": 6.768462997293412e-06,
"loss": 1.0181,
"step": 5615
},
{
"epoch": 0.64,
"grad_norm": 0.14668306358716413,
"learning_rate": 6.74951606011304e-06,
"loss": 0.9406,
"step": 5620
},
{
"epoch": 0.65,
"grad_norm": 0.15412889647580086,
"learning_rate": 6.730582161299733e-06,
"loss": 0.9792,
"step": 5625
},
{
"epoch": 0.65,
"grad_norm": 0.15006800647203158,
"learning_rate": 6.711661376801304e-06,
"loss": 1.0017,
"step": 5630
},
{
"epoch": 0.65,
"grad_norm": 0.14390708339212177,
"learning_rate": 6.692753782512942e-06,
"loss": 0.9931,
"step": 5635
},
{
"epoch": 0.65,
"grad_norm": 0.13660082876660443,
"learning_rate": 6.673859454276948e-06,
"loss": 0.9742,
"step": 5640
},
{
"epoch": 0.65,
"grad_norm": 0.1565118192165576,
"learning_rate": 6.6549784678824005e-06,
"loss": 0.9714,
"step": 5645
},
{
"epoch": 0.65,
"grad_norm": 0.158586835149041,
"learning_rate": 6.63611089906486e-06,
"loss": 0.9726,
"step": 5650
},
{
"epoch": 0.65,
"grad_norm": 0.15455288911592918,
"learning_rate": 6.6172568235060775e-06,
"loss": 1.0701,
"step": 5655
},
{
"epoch": 0.65,
"grad_norm": 0.16484229606426973,
"learning_rate": 6.598416316833658e-06,
"loss": 0.9269,
"step": 5660
},
{
"epoch": 0.65,
"grad_norm": 0.1828304365280764,
"learning_rate": 6.5795894546207995e-06,
"loss": 0.9558,
"step": 5665
},
{
"epoch": 0.65,
"grad_norm": 0.15126042217111035,
"learning_rate": 6.560776312385961e-06,
"loss": 1.0145,
"step": 5670
},
{
"epoch": 0.65,
"grad_norm": 0.22973461813902313,
"learning_rate": 6.541976965592567e-06,
"loss": 0.9849,
"step": 5675
},
{
"epoch": 0.65,
"grad_norm": 0.15628403688664821,
"learning_rate": 6.5231914896487125e-06,
"loss": 0.9818,
"step": 5680
},
{
"epoch": 0.65,
"grad_norm": 0.14462988458312,
"learning_rate": 6.504419959906839e-06,
"loss": 0.9852,
"step": 5685
},
{
"epoch": 0.65,
"grad_norm": 0.16460609889411057,
"learning_rate": 6.485662451663461e-06,
"loss": 0.9872,
"step": 5690
},
{
"epoch": 0.65,
"grad_norm": 0.15424485905857424,
"learning_rate": 6.466919040158848e-06,
"loss": 1.0329,
"step": 5695
},
{
"epoch": 0.65,
"grad_norm": 0.14963598517774093,
"learning_rate": 6.448189800576714e-06,
"loss": 0.9731,
"step": 5700
},
{
"epoch": 0.65,
"grad_norm": 0.1617225497146229,
"learning_rate": 6.4294748080439425e-06,
"loss": 1.0108,
"step": 5705
},
{
"epoch": 0.66,
"grad_norm": 0.1452728421827715,
"learning_rate": 6.410774137630254e-06,
"loss": 0.9906,
"step": 5710
},
{
"epoch": 0.66,
"grad_norm": 0.15225967259470957,
"learning_rate": 6.3920878643479275e-06,
"loss": 0.9916,
"step": 5715
},
{
"epoch": 0.66,
"grad_norm": 0.16032130201920372,
"learning_rate": 6.373416063151495e-06,
"loss": 0.9772,
"step": 5720
},
{
"epoch": 0.66,
"grad_norm": 0.1647727278503259,
"learning_rate": 6.35475880893743e-06,
"loss": 0.9645,
"step": 5725
},
{
"epoch": 0.66,
"grad_norm": 0.15661046319964683,
"learning_rate": 6.336116176543861e-06,
"loss": 0.962,
"step": 5730
},
{
"epoch": 0.66,
"grad_norm": 0.1548890947323481,
"learning_rate": 6.317488240750267e-06,
"loss": 1.0052,
"step": 5735
},
{
"epoch": 0.66,
"grad_norm": 0.14728109217858765,
"learning_rate": 6.298875076277166e-06,
"loss": 1.0019,
"step": 5740
},
{
"epoch": 0.66,
"grad_norm": 0.2137079790693736,
"learning_rate": 6.280276757785835e-06,
"loss": 1.0361,
"step": 5745
},
{
"epoch": 0.66,
"grad_norm": 0.32683091920833074,
"learning_rate": 6.261693359877996e-06,
"loss": 1.0015,
"step": 5750
},
{
"epoch": 0.66,
"grad_norm": 0.1440649858622469,
"learning_rate": 6.243124957095522e-06,
"loss": 1.02,
"step": 5755
},
{
"epoch": 0.66,
"grad_norm": 0.162049198082731,
"learning_rate": 6.224571623920145e-06,
"loss": 0.9722,
"step": 5760
},
{
"epoch": 0.66,
"grad_norm": 0.15399257569496416,
"learning_rate": 6.206033434773132e-06,
"loss": 0.9187,
"step": 5765
},
{
"epoch": 0.66,
"grad_norm": 0.16736776286614488,
"learning_rate": 6.187510464015023e-06,
"loss": 1.0096,
"step": 5770
},
{
"epoch": 0.66,
"grad_norm": 0.14161836623472104,
"learning_rate": 6.1690027859453015e-06,
"loss": 0.9951,
"step": 5775
},
{
"epoch": 0.66,
"grad_norm": 0.14528448430125562,
"learning_rate": 6.150510474802113e-06,
"loss": 0.9731,
"step": 5780
},
{
"epoch": 0.66,
"grad_norm": 0.14720970071154277,
"learning_rate": 6.132033604761971e-06,
"loss": 1.0082,
"step": 5785
},
{
"epoch": 0.66,
"grad_norm": 0.1496431204587347,
"learning_rate": 6.113572249939432e-06,
"loss": 0.9922,
"step": 5790
},
{
"epoch": 0.66,
"grad_norm": 0.16230033964233959,
"learning_rate": 6.095126484386838e-06,
"loss": 0.9431,
"step": 5795
},
{
"epoch": 0.67,
"grad_norm": 0.17072846352162332,
"learning_rate": 6.076696382093987e-06,
"loss": 0.9791,
"step": 5800
},
{
"epoch": 0.67,
"grad_norm": 0.16766036079039287,
"learning_rate": 6.0582820169878506e-06,
"loss": 0.9799,
"step": 5805
},
{
"epoch": 0.67,
"grad_norm": 0.14612565817096235,
"learning_rate": 6.039883462932286e-06,
"loss": 0.9934,
"step": 5810
},
{
"epoch": 0.67,
"grad_norm": 0.15230722191961354,
"learning_rate": 6.021500793727707e-06,
"loss": 1.0002,
"step": 5815
},
{
"epoch": 0.67,
"grad_norm": 0.1579299467211445,
"learning_rate": 6.003134083110829e-06,
"loss": 1.0295,
"step": 5820
},
{
"epoch": 0.67,
"grad_norm": 0.14157421547913424,
"learning_rate": 5.98478340475435e-06,
"loss": 0.9841,
"step": 5825
},
{
"epoch": 0.67,
"grad_norm": 0.1676476042504541,
"learning_rate": 5.966448832266651e-06,
"loss": 0.9399,
"step": 5830
},
{
"epoch": 0.67,
"grad_norm": 0.1627557866586007,
"learning_rate": 5.948130439191524e-06,
"loss": 0.978,
"step": 5835
},
{
"epoch": 0.67,
"grad_norm": 0.13590132012830658,
"learning_rate": 5.929828299007845e-06,
"loss": 1.0321,
"step": 5840
},
{
"epoch": 0.67,
"grad_norm": 0.1604664945790097,
"learning_rate": 5.911542485129307e-06,
"loss": 0.9374,
"step": 5845
},
{
"epoch": 0.67,
"grad_norm": 0.15426067571236451,
"learning_rate": 5.893273070904113e-06,
"loss": 0.9517,
"step": 5850
},
{
"epoch": 0.67,
"grad_norm": 0.16742392780919216,
"learning_rate": 5.875020129614683e-06,
"loss": 1.0109,
"step": 5855
},
{
"epoch": 0.67,
"grad_norm": 0.15422157327740102,
"learning_rate": 5.856783734477358e-06,
"loss": 0.9931,
"step": 5860
},
{
"epoch": 0.67,
"grad_norm": 0.15888282220841574,
"learning_rate": 5.838563958642118e-06,
"loss": 0.9704,
"step": 5865
},
{
"epoch": 0.67,
"grad_norm": 0.1537655018620826,
"learning_rate": 5.820360875192264e-06,
"loss": 0.9831,
"step": 5870
},
{
"epoch": 0.67,
"grad_norm": 0.1843725344614515,
"learning_rate": 5.80217455714416e-06,
"loss": 1.0212,
"step": 5875
},
{
"epoch": 0.67,
"grad_norm": 0.15929927386592213,
"learning_rate": 5.784005077446897e-06,
"loss": 1.0352,
"step": 5880
},
{
"epoch": 0.68,
"grad_norm": 0.1419758411864639,
"learning_rate": 5.765852508982046e-06,
"loss": 0.9213,
"step": 5885
},
{
"epoch": 0.68,
"grad_norm": 0.1557576281287212,
"learning_rate": 5.747716924563331e-06,
"loss": 0.9852,
"step": 5890
},
{
"epoch": 0.68,
"grad_norm": 0.16547186333918965,
"learning_rate": 5.72959839693635e-06,
"loss": 1.0518,
"step": 5895
},
{
"epoch": 0.68,
"grad_norm": 0.1558393156118298,
"learning_rate": 5.711496998778297e-06,
"loss": 0.9799,
"step": 5900
},
{
"epoch": 0.68,
"grad_norm": 0.1587417446308069,
"learning_rate": 5.693412802697626e-06,
"loss": 0.9779,
"step": 5905
},
{
"epoch": 0.68,
"grad_norm": 0.15439478354192476,
"learning_rate": 5.675345881233822e-06,
"loss": 0.9747,
"step": 5910
},
{
"epoch": 0.68,
"grad_norm": 0.15648461139755826,
"learning_rate": 5.657296306857061e-06,
"loss": 1.0085,
"step": 5915
},
{
"epoch": 0.68,
"grad_norm": 0.15208915622541327,
"learning_rate": 5.6392641519679355e-06,
"loss": 0.9679,
"step": 5920
},
{
"epoch": 0.68,
"grad_norm": 0.16576456459062394,
"learning_rate": 5.621249488897176e-06,
"loss": 0.9439,
"step": 5925
},
{
"epoch": 0.68,
"grad_norm": 0.1547373111349878,
"learning_rate": 5.603252389905338e-06,
"loss": 0.9327,
"step": 5930
},
{
"epoch": 0.68,
"grad_norm": 0.14787955509305042,
"learning_rate": 5.585272927182528e-06,
"loss": 0.9857,
"step": 5935
},
{
"epoch": 0.68,
"grad_norm": 0.15555983681358038,
"learning_rate": 5.567311172848113e-06,
"loss": 1.0101,
"step": 5940
},
{
"epoch": 0.68,
"grad_norm": 0.1466253526964909,
"learning_rate": 5.549367198950417e-06,
"loss": 0.9655,
"step": 5945
},
{
"epoch": 0.68,
"grad_norm": 0.14829416877447038,
"learning_rate": 5.531441077466463e-06,
"loss": 0.9749,
"step": 5950
},
{
"epoch": 0.68,
"grad_norm": 0.15441162548907375,
"learning_rate": 5.513532880301645e-06,
"loss": 1.0185,
"step": 5955
},
{
"epoch": 0.68,
"grad_norm": 0.13726023899056045,
"learning_rate": 5.495642679289468e-06,
"loss": 0.994,
"step": 5960
},
{
"epoch": 0.68,
"grad_norm": 0.15613086631486045,
"learning_rate": 5.477770546191248e-06,
"loss": 1.0494,
"step": 5965
},
{
"epoch": 0.69,
"grad_norm": 0.15251277270363645,
"learning_rate": 5.459916552695823e-06,
"loss": 0.9713,
"step": 5970
},
{
"epoch": 0.69,
"grad_norm": 0.1548610693245031,
"learning_rate": 5.442080770419283e-06,
"loss": 1.0362,
"step": 5975
},
{
"epoch": 0.69,
"grad_norm": 0.16717602054661884,
"learning_rate": 5.424263270904657e-06,
"loss": 0.9902,
"step": 5980
},
{
"epoch": 0.69,
"grad_norm": 0.1454603349002882,
"learning_rate": 5.4064641256216375e-06,
"loss": 1.032,
"step": 5985
},
{
"epoch": 0.69,
"grad_norm": 0.15778631648027955,
"learning_rate": 5.388683405966301e-06,
"loss": 1.0214,
"step": 5990
},
{
"epoch": 0.69,
"grad_norm": 0.14539036367636313,
"learning_rate": 5.3709211832608065e-06,
"loss": 1.034,
"step": 5995
},
{
"epoch": 0.69,
"grad_norm": 0.1507584442194237,
"learning_rate": 5.3531775287531305e-06,
"loss": 0.9652,
"step": 6000
},
{
"epoch": 0.69,
"grad_norm": 0.14383203795478258,
"learning_rate": 5.3354525136167586e-06,
"loss": 1.0047,
"step": 6005
},
{
"epoch": 0.69,
"grad_norm": 0.15043649859226577,
"learning_rate": 5.3177462089504115e-06,
"loss": 0.994,
"step": 6010
},
{
"epoch": 0.69,
"grad_norm": 0.15775442511793325,
"learning_rate": 5.300058685777758e-06,
"loss": 0.9919,
"step": 6015
},
{
"epoch": 0.69,
"grad_norm": 0.16838150546819958,
"learning_rate": 5.282390015047136e-06,
"loss": 1.0443,
"step": 6020
},
{
"epoch": 0.69,
"grad_norm": 0.15056794621281444,
"learning_rate": 5.264740267631257e-06,
"loss": 0.9517,
"step": 6025
},
{
"epoch": 0.69,
"grad_norm": 0.15525335577158048,
"learning_rate": 5.247109514326928e-06,
"loss": 1.0269,
"step": 6030
},
{
"epoch": 0.69,
"grad_norm": 0.14568849635771233,
"learning_rate": 5.229497825854766e-06,
"loss": 0.9804,
"step": 6035
},
{
"epoch": 0.69,
"grad_norm": 0.1511784864972891,
"learning_rate": 5.211905272858912e-06,
"loss": 0.9605,
"step": 6040
},
{
"epoch": 0.69,
"grad_norm": 0.15002269425759746,
"learning_rate": 5.194331925906766e-06,
"loss": 0.9645,
"step": 6045
},
{
"epoch": 0.69,
"grad_norm": 0.15762931702469565,
"learning_rate": 5.176777855488669e-06,
"loss": 0.9324,
"step": 6050
},
{
"epoch": 0.69,
"grad_norm": 0.14301981638842692,
"learning_rate": 5.15924313201765e-06,
"loss": 0.9402,
"step": 6055
},
{
"epoch": 0.7,
"grad_norm": 0.17275421135150412,
"learning_rate": 5.141727825829132e-06,
"loss": 0.9285,
"step": 6060
},
{
"epoch": 0.7,
"grad_norm": 0.15712919854999133,
"learning_rate": 5.124232007180645e-06,
"loss": 0.9827,
"step": 6065
},
{
"epoch": 0.7,
"grad_norm": 0.1590005618816746,
"learning_rate": 5.106755746251565e-06,
"loss": 0.9868,
"step": 6070
},
{
"epoch": 0.7,
"grad_norm": 0.1484437292456365,
"learning_rate": 5.089299113142805e-06,
"loss": 1.0019,
"step": 6075
},
{
"epoch": 0.7,
"grad_norm": 0.14277927460817288,
"learning_rate": 5.071862177876548e-06,
"loss": 0.962,
"step": 6080
},
{
"epoch": 0.7,
"grad_norm": 0.16565419227901507,
"learning_rate": 5.054445010395977e-06,
"loss": 0.9831,
"step": 6085
},
{
"epoch": 0.7,
"grad_norm": 0.15469481765452062,
"learning_rate": 5.037047680564958e-06,
"loss": 0.9889,
"step": 6090
},
{
"epoch": 0.7,
"grad_norm": 0.15482578520553406,
"learning_rate": 5.019670258167812e-06,
"loss": 0.9866,
"step": 6095
},
{
"epoch": 0.7,
"grad_norm": 0.15687465535566547,
"learning_rate": 5.002312812908989e-06,
"loss": 0.9999,
"step": 6100
},
{
"epoch": 0.7,
"grad_norm": 0.12860070371219362,
"learning_rate": 4.984975414412808e-06,
"loss": 0.9355,
"step": 6105
},
{
"epoch": 0.7,
"grad_norm": 0.1537988817813586,
"learning_rate": 4.967658132223192e-06,
"loss": 0.9612,
"step": 6110
},
{
"epoch": 0.7,
"grad_norm": 0.15561905582680047,
"learning_rate": 4.950361035803345e-06,
"loss": 1.0241,
"step": 6115
},
{
"epoch": 0.7,
"grad_norm": 0.15111909695185752,
"learning_rate": 4.93308419453553e-06,
"loss": 0.953,
"step": 6120
},
{
"epoch": 0.7,
"grad_norm": 0.1436421717216217,
"learning_rate": 4.915827677720747e-06,
"loss": 0.988,
"step": 6125
},
{
"epoch": 0.7,
"grad_norm": 0.1407583666595833,
"learning_rate": 4.898591554578469e-06,
"loss": 0.9845,
"step": 6130
},
{
"epoch": 0.7,
"grad_norm": 0.14041182390738136,
"learning_rate": 4.881375894246382e-06,
"loss": 1.0037,
"step": 6135
},
{
"epoch": 0.7,
"grad_norm": 0.15419402288162185,
"learning_rate": 4.864180765780064e-06,
"loss": 1.0362,
"step": 6140
},
{
"epoch": 0.71,
"grad_norm": 0.1528728965214611,
"learning_rate": 4.8470062381527636e-06,
"loss": 1.0425,
"step": 6145
},
{
"epoch": 0.71,
"grad_norm": 0.1620174423478394,
"learning_rate": 4.829852380255079e-06,
"loss": 1.0024,
"step": 6150
},
{
"epoch": 0.71,
"grad_norm": 0.14226449339732855,
"learning_rate": 4.812719260894697e-06,
"loss": 1.0461,
"step": 6155
},
{
"epoch": 0.71,
"grad_norm": 0.15159678247385458,
"learning_rate": 4.795606948796135e-06,
"loss": 0.9955,
"step": 6160
},
{
"epoch": 0.71,
"grad_norm": 0.17176020672870604,
"learning_rate": 4.77851551260042e-06,
"loss": 1.0087,
"step": 6165
},
{
"epoch": 0.71,
"grad_norm": 0.149628134992534,
"learning_rate": 4.761445020864867e-06,
"loss": 0.976,
"step": 6170
},
{
"epoch": 0.71,
"grad_norm": 0.15329101050801877,
"learning_rate": 4.744395542062766e-06,
"loss": 0.9809,
"step": 6175
},
{
"epoch": 0.71,
"grad_norm": 0.15075943547247797,
"learning_rate": 4.72736714458312e-06,
"loss": 0.9604,
"step": 6180
},
{
"epoch": 0.71,
"grad_norm": 0.14214587479825333,
"learning_rate": 4.710359896730379e-06,
"loss": 0.9303,
"step": 6185
},
{
"epoch": 0.71,
"grad_norm": 0.17367794782492055,
"learning_rate": 4.69337386672414e-06,
"loss": 1.0248,
"step": 6190
},
{
"epoch": 0.71,
"grad_norm": 0.16464022686101376,
"learning_rate": 4.676409122698913e-06,
"loss": 0.9562,
"step": 6195
},
{
"epoch": 0.71,
"grad_norm": 0.15085503549503335,
"learning_rate": 4.659465732703809e-06,
"loss": 0.9533,
"step": 6200
},
{
"epoch": 0.71,
"grad_norm": 0.20687792821646106,
"learning_rate": 4.642543764702285e-06,
"loss": 1.01,
"step": 6205
},
{
"epoch": 0.71,
"grad_norm": 0.15709771072126685,
"learning_rate": 4.625643286571887e-06,
"loss": 0.9879,
"step": 6210
},
{
"epoch": 0.71,
"grad_norm": 0.1475435517920879,
"learning_rate": 4.6087643661039305e-06,
"loss": 0.9761,
"step": 6215
},
{
"epoch": 0.71,
"grad_norm": 0.16027006120678375,
"learning_rate": 4.591907071003287e-06,
"loss": 1.0196,
"step": 6220
},
{
"epoch": 0.71,
"grad_norm": 0.15981481611205986,
"learning_rate": 4.5750714688880716e-06,
"loss": 0.9956,
"step": 6225
},
{
"epoch": 0.71,
"grad_norm": 0.16292566435498934,
"learning_rate": 4.55825762728938e-06,
"loss": 0.9671,
"step": 6230
},
{
"epoch": 0.72,
"grad_norm": 0.15366310349082962,
"learning_rate": 4.541465613651034e-06,
"loss": 0.9912,
"step": 6235
},
{
"epoch": 0.72,
"grad_norm": 0.13385400945297377,
"learning_rate": 4.5246954953292895e-06,
"loss": 0.9582,
"step": 6240
},
{
"epoch": 0.72,
"grad_norm": 0.15233161252549499,
"learning_rate": 4.507947339592581e-06,
"loss": 0.9733,
"step": 6245
},
{
"epoch": 0.72,
"grad_norm": 0.13360150013621647,
"learning_rate": 4.491221213621241e-06,
"loss": 0.9875,
"step": 6250
},
{
"epoch": 0.72,
"grad_norm": 0.15882871426575904,
"learning_rate": 4.474517184507237e-06,
"loss": 0.9831,
"step": 6255
},
{
"epoch": 0.72,
"grad_norm": 0.15118357643757155,
"learning_rate": 4.457835319253909e-06,
"loss": 0.9606,
"step": 6260
},
{
"epoch": 0.72,
"grad_norm": 0.13571464800157987,
"learning_rate": 4.441175684775687e-06,
"loss": 0.962,
"step": 6265
},
{
"epoch": 0.72,
"grad_norm": 0.16564062308636374,
"learning_rate": 4.424538347897829e-06,
"loss": 0.9873,
"step": 6270
},
{
"epoch": 0.72,
"grad_norm": 0.15519345192382866,
"learning_rate": 4.4079233753561515e-06,
"loss": 0.993,
"step": 6275
},
{
"epoch": 0.72,
"grad_norm": 0.1523263745344713,
"learning_rate": 4.391330833796762e-06,
"loss": 1.0243,
"step": 6280
},
{
"epoch": 0.72,
"grad_norm": 0.15069166306638793,
"learning_rate": 4.374760789775802e-06,
"loss": 1.0076,
"step": 6285
},
{
"epoch": 0.72,
"grad_norm": 0.16394837192448947,
"learning_rate": 4.358213309759162e-06,
"loss": 1.003,
"step": 6290
},
{
"epoch": 0.72,
"grad_norm": 0.17611348702383625,
"learning_rate": 4.341688460122223e-06,
"loss": 0.9707,
"step": 6295
},
{
"epoch": 0.72,
"grad_norm": 0.1572408366216261,
"learning_rate": 4.325186307149593e-06,
"loss": 0.9865,
"step": 6300
},
{
"epoch": 0.72,
"grad_norm": 0.1831412882448159,
"learning_rate": 4.308706917034845e-06,
"loss": 0.9622,
"step": 6305
},
{
"epoch": 0.72,
"grad_norm": 0.14885628515319285,
"learning_rate": 4.292250355880235e-06,
"loss": 0.9667,
"step": 6310
},
{
"epoch": 0.72,
"grad_norm": 0.14978852193094377,
"learning_rate": 4.275816689696455e-06,
"loss": 1.0035,
"step": 6315
},
{
"epoch": 0.73,
"grad_norm": 0.14551972518060166,
"learning_rate": 4.259405984402355e-06,
"loss": 1.013,
"step": 6320
},
{
"epoch": 0.73,
"grad_norm": 0.17573222581592202,
"learning_rate": 4.243018305824685e-06,
"loss": 0.9556,
"step": 6325
},
{
"epoch": 0.73,
"grad_norm": 0.14465921778032356,
"learning_rate": 4.226653719697838e-06,
"loss": 0.9877,
"step": 6330
},
{
"epoch": 0.73,
"grad_norm": 0.15090546354584142,
"learning_rate": 4.21031229166357e-06,
"loss": 1.0186,
"step": 6335
},
{
"epoch": 0.73,
"grad_norm": 0.15038037644798075,
"learning_rate": 4.193994087270745e-06,
"loss": 0.9895,
"step": 6340
},
{
"epoch": 0.73,
"grad_norm": 0.14723555727309964,
"learning_rate": 4.177699171975077e-06,
"loss": 0.9893,
"step": 6345
},
{
"epoch": 0.73,
"grad_norm": 0.14288190587726013,
"learning_rate": 4.161427611138854e-06,
"loss": 0.9431,
"step": 6350
},
{
"epoch": 0.73,
"grad_norm": 0.16976097381378574,
"learning_rate": 4.145179470030699e-06,
"loss": 0.9866,
"step": 6355
},
{
"epoch": 0.73,
"grad_norm": 0.1527916367632068,
"learning_rate": 4.128954813825281e-06,
"loss": 0.9485,
"step": 6360
},
{
"epoch": 0.73,
"grad_norm": 0.1488773289483128,
"learning_rate": 4.112753707603067e-06,
"loss": 0.9558,
"step": 6365
},
{
"epoch": 0.73,
"grad_norm": 0.14539581127082574,
"learning_rate": 4.096576216350072e-06,
"loss": 0.9847,
"step": 6370
},
{
"epoch": 0.73,
"grad_norm": 0.1496826729339569,
"learning_rate": 4.080422404957567e-06,
"loss": 0.9802,
"step": 6375
},
{
"epoch": 0.73,
"grad_norm": 0.15493888823813523,
"learning_rate": 4.0642923382218576e-06,
"loss": 0.9745,
"step": 6380
},
{
"epoch": 0.73,
"grad_norm": 0.14555635892358196,
"learning_rate": 4.048186080843992e-06,
"loss": 0.9407,
"step": 6385
},
{
"epoch": 0.73,
"grad_norm": 0.1540087701423811,
"learning_rate": 4.032103697429516e-06,
"loss": 0.9682,
"step": 6390
},
{
"epoch": 0.73,
"grad_norm": 0.14892691191651208,
"learning_rate": 4.0160452524882234e-06,
"loss": 0.9676,
"step": 6395
},
{
"epoch": 0.73,
"grad_norm": 0.17081738531831414,
"learning_rate": 4.000010810433862e-06,
"loss": 0.9392,
"step": 6400
},
{
"epoch": 0.73,
"grad_norm": 0.1474766749470663,
"learning_rate": 3.9840004355839245e-06,
"loss": 0.9755,
"step": 6405
},
{
"epoch": 0.74,
"grad_norm": 0.175400126403777,
"learning_rate": 3.968014192159352e-06,
"loss": 0.9593,
"step": 6410
},
{
"epoch": 0.74,
"grad_norm": 0.16205893591599618,
"learning_rate": 3.952052144284285e-06,
"loss": 1.0342,
"step": 6415
},
{
"epoch": 0.74,
"grad_norm": 0.15344356507366966,
"learning_rate": 3.93611435598583e-06,
"loss": 1.016,
"step": 6420
},
{
"epoch": 0.74,
"grad_norm": 0.1886184259617471,
"learning_rate": 3.920200891193756e-06,
"loss": 1.0324,
"step": 6425
},
{
"epoch": 0.74,
"grad_norm": 0.14740903761058527,
"learning_rate": 3.9043118137402894e-06,
"loss": 1.0043,
"step": 6430
},
{
"epoch": 0.74,
"grad_norm": 0.16493725640688142,
"learning_rate": 3.8884471873598205e-06,
"loss": 1.0069,
"step": 6435
},
{
"epoch": 0.74,
"grad_norm": 0.14932940025133742,
"learning_rate": 3.872607075688663e-06,
"loss": 0.9617,
"step": 6440
},
{
"epoch": 0.74,
"grad_norm": 0.14165548616407353,
"learning_rate": 3.856791542264807e-06,
"loss": 0.9438,
"step": 6445
},
{
"epoch": 0.74,
"grad_norm": 0.15223795528082146,
"learning_rate": 3.841000650527635e-06,
"loss": 0.9781,
"step": 6450
},
{
"epoch": 0.74,
"grad_norm": 0.17401523643530528,
"learning_rate": 3.8252344638177065e-06,
"loss": 1.0118,
"step": 6455
},
{
"epoch": 0.74,
"grad_norm": 0.1470655452205851,
"learning_rate": 3.809493045376472e-06,
"loss": 0.9656,
"step": 6460
},
{
"epoch": 0.74,
"grad_norm": 0.15531973093039966,
"learning_rate": 3.7937764583460303e-06,
"loss": 0.9416,
"step": 6465
},
{
"epoch": 0.74,
"grad_norm": 0.1446467305648632,
"learning_rate": 3.7780847657688925e-06,
"loss": 0.9724,
"step": 6470
},
{
"epoch": 0.74,
"grad_norm": 0.16729870376414768,
"learning_rate": 3.7624180305876856e-06,
"loss": 0.9356,
"step": 6475
},
{
"epoch": 0.74,
"grad_norm": 0.1500686318930753,
"learning_rate": 3.746776315644954e-06,
"loss": 0.9826,
"step": 6480
},
{
"epoch": 0.74,
"grad_norm": 0.16643532315320334,
"learning_rate": 3.7311596836828643e-06,
"loss": 1.0124,
"step": 6485
},
{
"epoch": 0.74,
"grad_norm": 0.17644384676787686,
"learning_rate": 3.7155681973429736e-06,
"loss": 0.9826,
"step": 6490
},
{
"epoch": 0.75,
"grad_norm": 0.15054337514922506,
"learning_rate": 3.7000019191659854e-06,
"loss": 0.9828,
"step": 6495
},
{
"epoch": 0.75,
"grad_norm": 0.14132209272244406,
"learning_rate": 3.684460911591463e-06,
"loss": 0.9645,
"step": 6500
},
{
"epoch": 0.75,
"grad_norm": 0.14186347797933052,
"learning_rate": 3.6689452369576338e-06,
"loss": 0.9561,
"step": 6505
},
{
"epoch": 0.75,
"grad_norm": 0.15417940212294304,
"learning_rate": 3.6534549575010893e-06,
"loss": 0.9476,
"step": 6510
},
{
"epoch": 0.75,
"grad_norm": 0.14617575219617132,
"learning_rate": 3.637990135356558e-06,
"loss": 1.0118,
"step": 6515
},
{
"epoch": 0.75,
"grad_norm": 0.13545242335084928,
"learning_rate": 3.6225508325566673e-06,
"loss": 0.95,
"step": 6520
},
{
"epoch": 0.75,
"grad_norm": 0.15509606873611675,
"learning_rate": 3.6071371110316657e-06,
"loss": 1.0034,
"step": 6525
},
{
"epoch": 0.75,
"grad_norm": 0.15127616734361077,
"learning_rate": 3.5917490326091974e-06,
"loss": 1.0211,
"step": 6530
},
{
"epoch": 0.75,
"grad_norm": 0.15539116346374643,
"learning_rate": 3.5763866590140427e-06,
"loss": 0.9888,
"step": 6535
},
{
"epoch": 0.75,
"grad_norm": 0.15235763009045522,
"learning_rate": 3.5610500518678757e-06,
"loss": 0.9826,
"step": 6540
},
{
"epoch": 0.75,
"grad_norm": 0.14593435900689278,
"learning_rate": 3.545739272689024e-06,
"loss": 0.9556,
"step": 6545
},
{
"epoch": 0.75,
"grad_norm": 0.1643768760615994,
"learning_rate": 3.530454382892202e-06,
"loss": 0.9276,
"step": 6550
},
{
"epoch": 0.75,
"grad_norm": 0.15958231280802787,
"learning_rate": 3.515195443788283e-06,
"loss": 0.9905,
"step": 6555
},
{
"epoch": 0.75,
"grad_norm": 0.16957025851393506,
"learning_rate": 3.4999625165840444e-06,
"loss": 0.9386,
"step": 6560
},
{
"epoch": 0.75,
"grad_norm": 0.14851191249106294,
"learning_rate": 3.48475566238192e-06,
"loss": 0.9625,
"step": 6565
},
{
"epoch": 0.75,
"grad_norm": 0.13975248759432002,
"learning_rate": 3.469574942179774e-06,
"loss": 0.9776,
"step": 6570
},
{
"epoch": 0.75,
"grad_norm": 0.15033652798399189,
"learning_rate": 3.4544204168706286e-06,
"loss": 0.987,
"step": 6575
},
{
"epoch": 0.76,
"grad_norm": 0.156817146283973,
"learning_rate": 3.4392921472424334e-06,
"loss": 0.9878,
"step": 6580
},
{
"epoch": 0.76,
"grad_norm": 0.15269793740648327,
"learning_rate": 3.4241901939778242e-06,
"loss": 0.9683,
"step": 6585
},
{
"epoch": 0.76,
"grad_norm": 0.14570486327154092,
"learning_rate": 3.4091146176538816e-06,
"loss": 0.9795,
"step": 6590
},
{
"epoch": 0.76,
"grad_norm": 0.15420243853390458,
"learning_rate": 3.3940654787418727e-06,
"loss": 0.9352,
"step": 6595
},
{
"epoch": 0.76,
"grad_norm": 0.15743994076106987,
"learning_rate": 3.379042837607025e-06,
"loss": 0.9951,
"step": 6600
},
{
"epoch": 0.76,
"grad_norm": 0.14173494089969463,
"learning_rate": 3.364046754508277e-06,
"loss": 0.9606,
"step": 6605
},
{
"epoch": 0.76,
"grad_norm": 0.16455341288029054,
"learning_rate": 3.3490772895980306e-06,
"loss": 0.9865,
"step": 6610
},
{
"epoch": 0.76,
"grad_norm": 0.14920247429409314,
"learning_rate": 3.3341345029219305e-06,
"loss": 0.9809,
"step": 6615
},
{
"epoch": 0.76,
"grad_norm": 0.15573378905783714,
"learning_rate": 3.3192184544186e-06,
"loss": 0.9848,
"step": 6620
},
{
"epoch": 0.76,
"grad_norm": 0.16324932110188517,
"learning_rate": 3.30432920391941e-06,
"loss": 0.9623,
"step": 6625
},
{
"epoch": 0.76,
"grad_norm": 0.157862940194976,
"learning_rate": 3.2894668111482397e-06,
"loss": 0.9782,
"step": 6630
},
{
"epoch": 0.76,
"grad_norm": 0.15383251841048987,
"learning_rate": 3.2746313357212365e-06,
"loss": 0.945,
"step": 6635
},
{
"epoch": 0.76,
"grad_norm": 0.1474779368950937,
"learning_rate": 3.2598228371465835e-06,
"loss": 0.9387,
"step": 6640
},
{
"epoch": 0.76,
"grad_norm": 0.16309071606735917,
"learning_rate": 3.2450413748242437e-06,
"loss": 0.9591,
"step": 6645
},
{
"epoch": 0.76,
"grad_norm": 0.15016109224524307,
"learning_rate": 3.2302870080457327e-06,
"loss": 0.9363,
"step": 6650
},
{
"epoch": 0.76,
"grad_norm": 0.1486781853762874,
"learning_rate": 3.2155597959938967e-06,
"loss": 0.9915,
"step": 6655
},
{
"epoch": 0.76,
"grad_norm": 0.16894394038538,
"learning_rate": 3.2008597977426305e-06,
"loss": 1.0173,
"step": 6660
},
{
"epoch": 0.76,
"grad_norm": 0.1495730982568493,
"learning_rate": 3.1861870722566944e-06,
"loss": 0.9749,
"step": 6665
},
{
"epoch": 0.77,
"grad_norm": 0.16602455425138257,
"learning_rate": 3.1715416783914387e-06,
"loss": 1.0063,
"step": 6670
},
{
"epoch": 0.77,
"grad_norm": 0.1470880510010936,
"learning_rate": 3.1569236748925804e-06,
"loss": 0.9467,
"step": 6675
},
{
"epoch": 0.77,
"grad_norm": 0.15574127699378634,
"learning_rate": 3.142333120395982e-06,
"loss": 0.9892,
"step": 6680
},
{
"epoch": 0.77,
"grad_norm": 0.16119497545570802,
"learning_rate": 3.1277700734273786e-06,
"loss": 0.9862,
"step": 6685
},
{
"epoch": 0.77,
"grad_norm": 0.15950903067751668,
"learning_rate": 3.113234592402189e-06,
"loss": 0.9928,
"step": 6690
},
{
"epoch": 0.77,
"grad_norm": 0.14925696451539675,
"learning_rate": 3.098726735625248e-06,
"loss": 0.9414,
"step": 6695
},
{
"epoch": 0.77,
"grad_norm": 0.1454036404081469,
"learning_rate": 3.084246561290584e-06,
"loss": 0.9578,
"step": 6700
},
{
"epoch": 0.77,
"grad_norm": 0.1541625455587587,
"learning_rate": 3.069794127481195e-06,
"loss": 0.9612,
"step": 6705
},
{
"epoch": 0.77,
"grad_norm": 0.13709653928451979,
"learning_rate": 3.055369492168788e-06,
"loss": 0.9776,
"step": 6710
},
{
"epoch": 0.77,
"grad_norm": 0.1543662522819474,
"learning_rate": 3.0409727132135825e-06,
"loss": 0.9927,
"step": 6715
},
{
"epoch": 0.77,
"grad_norm": 0.1564536021341216,
"learning_rate": 3.0266038483640513e-06,
"loss": 1.0023,
"step": 6720
},
{
"epoch": 0.77,
"grad_norm": 0.15358460677991279,
"learning_rate": 3.0122629552566962e-06,
"loss": 0.9958,
"step": 6725
},
{
"epoch": 0.77,
"grad_norm": 0.14583463520698572,
"learning_rate": 2.997950091415831e-06,
"loss": 1.0267,
"step": 6730
},
{
"epoch": 0.77,
"grad_norm": 0.17352607189247246,
"learning_rate": 2.9836653142533167e-06,
"loss": 1.0002,
"step": 6735
},
{
"epoch": 0.77,
"grad_norm": 0.16248501190619724,
"learning_rate": 2.969408681068374e-06,
"loss": 1.0108,
"step": 6740
},
{
"epoch": 0.77,
"grad_norm": 0.15909294729771317,
"learning_rate": 2.9551802490473215e-06,
"loss": 0.9921,
"step": 6745
},
{
"epoch": 0.77,
"grad_norm": 0.16029664550130318,
"learning_rate": 2.9409800752633545e-06,
"loss": 0.9845,
"step": 6750
},
{
"epoch": 0.78,
"grad_norm": 0.14254871496435523,
"learning_rate": 2.9268082166763336e-06,
"loss": 0.967,
"step": 6755
},
{
"epoch": 0.78,
"grad_norm": 0.15460265219205102,
"learning_rate": 2.9126647301325173e-06,
"loss": 0.9382,
"step": 6760
},
{
"epoch": 0.78,
"grad_norm": 0.1466358602039194,
"learning_rate": 2.8985496723643826e-06,
"loss": 0.9796,
"step": 6765
},
{
"epoch": 0.78,
"grad_norm": 0.15192828478168516,
"learning_rate": 2.8844630999903566e-06,
"loss": 0.9738,
"step": 6770
},
{
"epoch": 0.78,
"grad_norm": 0.1729198618225192,
"learning_rate": 2.870405069514608e-06,
"loss": 0.9682,
"step": 6775
},
{
"epoch": 0.78,
"grad_norm": 0.14535310250530914,
"learning_rate": 2.8563756373268314e-06,
"loss": 0.9951,
"step": 6780
},
{
"epoch": 0.78,
"grad_norm": 0.140057021577601,
"learning_rate": 2.842374859701982e-06,
"loss": 0.9771,
"step": 6785
},
{
"epoch": 0.78,
"grad_norm": 0.15314909977794092,
"learning_rate": 2.8284027928001025e-06,
"loss": 1.0689,
"step": 6790
},
{
"epoch": 0.78,
"grad_norm": 0.1461597837563946,
"learning_rate": 2.814459492666054e-06,
"loss": 0.9355,
"step": 6795
},
{
"epoch": 0.78,
"grad_norm": 0.15281484904034245,
"learning_rate": 2.80054501522931e-06,
"loss": 0.9342,
"step": 6800
},
{
"epoch": 0.78,
"grad_norm": 0.14712836705133592,
"learning_rate": 2.7866594163037398e-06,
"loss": 0.965,
"step": 6805
},
{
"epoch": 0.78,
"grad_norm": 0.16922276145173143,
"learning_rate": 2.7728027515873655e-06,
"loss": 0.9958,
"step": 6810
},
{
"epoch": 0.78,
"grad_norm": 0.16075177834455678,
"learning_rate": 2.7589750766621515e-06,
"loss": 1.0072,
"step": 6815
},
{
"epoch": 0.78,
"grad_norm": 0.1703453588403547,
"learning_rate": 2.7451764469937758e-06,
"loss": 0.978,
"step": 6820
},
{
"epoch": 0.78,
"grad_norm": 0.15388589543860512,
"learning_rate": 2.7314069179314108e-06,
"loss": 0.954,
"step": 6825
},
{
"epoch": 0.78,
"grad_norm": 0.1419355513919545,
"learning_rate": 2.7176665447075046e-06,
"loss": 0.9517,
"step": 6830
},
{
"epoch": 0.78,
"grad_norm": 0.16982167970589784,
"learning_rate": 2.703955382437551e-06,
"loss": 1.0503,
"step": 6835
},
{
"epoch": 0.78,
"grad_norm": 0.15724949450493386,
"learning_rate": 2.690273486119872e-06,
"loss": 0.9638,
"step": 6840
},
{
"epoch": 0.79,
"grad_norm": 0.156381108142604,
"learning_rate": 2.6766209106354002e-06,
"loss": 0.9955,
"step": 6845
},
{
"epoch": 0.79,
"grad_norm": 0.1906102548083453,
"learning_rate": 2.6629977107474512e-06,
"loss": 1.0156,
"step": 6850
},
{
"epoch": 0.79,
"grad_norm": 0.15576212674634393,
"learning_rate": 2.6494039411015192e-06,
"loss": 0.9722,
"step": 6855
},
{
"epoch": 0.79,
"grad_norm": 0.1504241476466268,
"learning_rate": 2.6358396562250377e-06,
"loss": 1.0075,
"step": 6860
},
{
"epoch": 0.79,
"grad_norm": 0.16754272047048907,
"learning_rate": 2.6223049105271748e-06,
"loss": 0.9813,
"step": 6865
},
{
"epoch": 0.79,
"grad_norm": 0.15009514492743162,
"learning_rate": 2.6087997582986067e-06,
"loss": 1.0038,
"step": 6870
},
{
"epoch": 0.79,
"grad_norm": 0.15403797952592127,
"learning_rate": 2.5953242537113143e-06,
"loss": 0.9778,
"step": 6875
},
{
"epoch": 0.79,
"grad_norm": 0.1571636732793357,
"learning_rate": 2.581878450818346e-06,
"loss": 0.9862,
"step": 6880
},
{
"epoch": 0.79,
"grad_norm": 0.15110304417659273,
"learning_rate": 2.5684624035536133e-06,
"loss": 0.955,
"step": 6885
},
{
"epoch": 0.79,
"grad_norm": 0.14495142205754752,
"learning_rate": 2.55507616573167e-06,
"loss": 1.0287,
"step": 6890
},
{
"epoch": 0.79,
"grad_norm": 0.15646034757520777,
"learning_rate": 2.5417197910475e-06,
"loss": 0.9589,
"step": 6895
},
{
"epoch": 0.79,
"grad_norm": 0.14270660369725968,
"learning_rate": 2.5283933330763033e-06,
"loss": 1.0479,
"step": 6900
},
{
"epoch": 0.79,
"grad_norm": 0.14596925751765133,
"learning_rate": 2.515096845273273e-06,
"loss": 1.026,
"step": 6905
},
{
"epoch": 0.79,
"grad_norm": 0.1644497351186129,
"learning_rate": 2.5018303809733844e-06,
"loss": 0.995,
"step": 6910
},
{
"epoch": 0.79,
"grad_norm": 0.16927366551390255,
"learning_rate": 2.4885939933911885e-06,
"loss": 0.9772,
"step": 6915
},
{
"epoch": 0.79,
"grad_norm": 0.14614467442515344,
"learning_rate": 2.4753877356205826e-06,
"loss": 0.943,
"step": 6920
},
{
"epoch": 0.79,
"grad_norm": 0.17779479922456906,
"learning_rate": 2.462211660634619e-06,
"loss": 1.0236,
"step": 6925
},
{
"epoch": 0.8,
"grad_norm": 0.1633635698612729,
"learning_rate": 2.4490658212852735e-06,
"loss": 0.9733,
"step": 6930
},
{
"epoch": 0.8,
"grad_norm": 0.16812464842220606,
"learning_rate": 2.435950270303241e-06,
"loss": 0.949,
"step": 6935
},
{
"epoch": 0.8,
"grad_norm": 0.14368425575669092,
"learning_rate": 2.422865060297723e-06,
"loss": 0.9856,
"step": 6940
},
{
"epoch": 0.8,
"grad_norm": 0.16393164478523037,
"learning_rate": 2.4098102437562164e-06,
"loss": 0.9772,
"step": 6945
},
{
"epoch": 0.8,
"grad_norm": 0.1508090015912488,
"learning_rate": 2.3967858730443095e-06,
"loss": 1.0287,
"step": 6950
},
{
"epoch": 0.8,
"grad_norm": 0.16245985110432845,
"learning_rate": 2.3837920004054617e-06,
"loss": 0.9866,
"step": 6955
},
{
"epoch": 0.8,
"grad_norm": 0.1447019480924854,
"learning_rate": 2.3708286779607924e-06,
"loss": 1.0045,
"step": 6960
},
{
"epoch": 0.8,
"grad_norm": 0.15480051743174753,
"learning_rate": 2.3578959577088933e-06,
"loss": 1.0078,
"step": 6965
},
{
"epoch": 0.8,
"grad_norm": 0.1462126270393594,
"learning_rate": 2.3449938915255856e-06,
"loss": 1.0195,
"step": 6970
},
{
"epoch": 0.8,
"grad_norm": 0.15568840577623588,
"learning_rate": 2.3321225311637465e-06,
"loss": 0.9989,
"step": 6975
},
{
"epoch": 0.8,
"grad_norm": 0.15120792436415278,
"learning_rate": 2.3192819282530753e-06,
"loss": 0.9853,
"step": 6980
},
{
"epoch": 0.8,
"grad_norm": 0.18590505405328653,
"learning_rate": 2.3064721342999007e-06,
"loss": 0.956,
"step": 6985
},
{
"epoch": 0.8,
"grad_norm": 0.15964502006936362,
"learning_rate": 2.293693200686976e-06,
"loss": 0.9426,
"step": 6990
},
{
"epoch": 0.8,
"grad_norm": 0.14982190356205696,
"learning_rate": 2.280945178673253e-06,
"loss": 1.0457,
"step": 6995
},
{
"epoch": 0.8,
"grad_norm": 0.16716330142656413,
"learning_rate": 2.268228119393705e-06,
"loss": 0.9912,
"step": 7000
},
{
"epoch": 0.8,
"grad_norm": 0.15873990637567342,
"learning_rate": 2.2555420738590992e-06,
"loss": 0.9556,
"step": 7005
},
{
"epoch": 0.8,
"grad_norm": 0.14942706666413866,
"learning_rate": 2.2428870929558012e-06,
"loss": 1.0056,
"step": 7010
},
{
"epoch": 0.8,
"grad_norm": 0.14528494282216067,
"learning_rate": 2.230263227445577e-06,
"loss": 1.0092,
"step": 7015
},
{
"epoch": 0.81,
"grad_norm": 0.15436812813602724,
"learning_rate": 2.217670527965369e-06,
"loss": 0.9574,
"step": 7020
},
{
"epoch": 0.81,
"grad_norm": 0.17547054372938378,
"learning_rate": 2.2051090450271173e-06,
"loss": 1.0165,
"step": 7025
},
{
"epoch": 0.81,
"grad_norm": 0.17023213362503592,
"learning_rate": 2.192578829017543e-06,
"loss": 1.004,
"step": 7030
},
{
"epoch": 0.81,
"grad_norm": 0.17268947032692641,
"learning_rate": 2.180079930197946e-06,
"loss": 0.9588,
"step": 7035
},
{
"epoch": 0.81,
"grad_norm": 0.14257011553269627,
"learning_rate": 2.1676123987040145e-06,
"loss": 0.9633,
"step": 7040
},
{
"epoch": 0.81,
"grad_norm": 0.20780066411054987,
"learning_rate": 2.1551762845455994e-06,
"loss": 0.9967,
"step": 7045
},
{
"epoch": 0.81,
"grad_norm": 0.15343761293964775,
"learning_rate": 2.142771637606551e-06,
"loss": 0.9164,
"step": 7050
},
{
"epoch": 0.81,
"grad_norm": 0.19043035802719563,
"learning_rate": 2.1303985076444834e-06,
"loss": 1.059,
"step": 7055
},
{
"epoch": 0.81,
"grad_norm": 0.15393658038837044,
"learning_rate": 2.118056944290593e-06,
"loss": 0.9436,
"step": 7060
},
{
"epoch": 0.81,
"grad_norm": 0.1556496480954147,
"learning_rate": 2.105746997049464e-06,
"loss": 1.0069,
"step": 7065
},
{
"epoch": 0.81,
"grad_norm": 0.1604380349154524,
"learning_rate": 2.093468715298843e-06,
"loss": 0.9941,
"step": 7070
},
{
"epoch": 0.81,
"grad_norm": 0.14349663226609585,
"learning_rate": 2.0812221482894824e-06,
"loss": 0.9793,
"step": 7075
},
{
"epoch": 0.81,
"grad_norm": 0.15897846814042985,
"learning_rate": 2.0690073451449035e-06,
"loss": 0.9958,
"step": 7080
},
{
"epoch": 0.81,
"grad_norm": 0.15622240119338981,
"learning_rate": 2.056824354861222e-06,
"loss": 1.0169,
"step": 7085
},
{
"epoch": 0.81,
"grad_norm": 0.14766992410187638,
"learning_rate": 2.0446732263069492e-06,
"loss": 1.0254,
"step": 7090
},
{
"epoch": 0.81,
"grad_norm": 0.17468522904776687,
"learning_rate": 2.0325540082227856e-06,
"loss": 1.0161,
"step": 7095
},
{
"epoch": 0.81,
"grad_norm": 0.18963796049069814,
"learning_rate": 2.020466749221438e-06,
"loss": 1.0343,
"step": 7100
},
{
"epoch": 0.82,
"grad_norm": 0.14059050693197184,
"learning_rate": 2.0084114977874135e-06,
"loss": 0.9367,
"step": 7105
},
{
"epoch": 0.82,
"grad_norm": 0.15952116676700964,
"learning_rate": 1.9963883022768305e-06,
"loss": 0.9381,
"step": 7110
},
{
"epoch": 0.82,
"grad_norm": 0.1459010910630693,
"learning_rate": 1.9843972109172315e-06,
"loss": 0.985,
"step": 7115
},
{
"epoch": 0.82,
"grad_norm": 0.14524953090759543,
"learning_rate": 1.9724382718073765e-06,
"loss": 0.9902,
"step": 7120
},
{
"epoch": 0.82,
"grad_norm": 0.1525728530451523,
"learning_rate": 1.9605115329170563e-06,
"loss": 0.988,
"step": 7125
},
{
"epoch": 0.82,
"grad_norm": 0.14974802897038045,
"learning_rate": 1.948617042086901e-06,
"loss": 0.9789,
"step": 7130
},
{
"epoch": 0.82,
"grad_norm": 0.16760910271310145,
"learning_rate": 1.936754847028185e-06,
"loss": 1.0427,
"step": 7135
},
{
"epoch": 0.82,
"grad_norm": 0.1514444936843788,
"learning_rate": 1.9249249953226434e-06,
"loss": 0.9411,
"step": 7140
},
{
"epoch": 0.82,
"grad_norm": 0.18819708925940495,
"learning_rate": 1.91312753442227e-06,
"loss": 1.0294,
"step": 7145
},
{
"epoch": 0.82,
"grad_norm": 0.15580563311393808,
"learning_rate": 1.901362511649133e-06,
"loss": 1.0069,
"step": 7150
},
{
"epoch": 0.82,
"grad_norm": 0.1674336734465148,
"learning_rate": 1.8896299741951873e-06,
"loss": 1.0046,
"step": 7155
},
{
"epoch": 0.82,
"grad_norm": 0.14556105337367084,
"learning_rate": 1.8779299691220753e-06,
"loss": 0.962,
"step": 7160
},
{
"epoch": 0.82,
"grad_norm": 0.14510602260541935,
"learning_rate": 1.8662625433609582e-06,
"loss": 0.9263,
"step": 7165
},
{
"epoch": 0.82,
"grad_norm": 0.15137109736849935,
"learning_rate": 1.8546277437123038e-06,
"loss": 0.9561,
"step": 7170
},
{
"epoch": 0.82,
"grad_norm": 0.1590179499444243,
"learning_rate": 1.8430256168457128e-06,
"loss": 1.0201,
"step": 7175
},
{
"epoch": 0.82,
"grad_norm": 0.14290336347820384,
"learning_rate": 1.8314562092997256e-06,
"loss": 0.9676,
"step": 7180
},
{
"epoch": 0.82,
"grad_norm": 0.150654138841851,
"learning_rate": 1.8199195674816484e-06,
"loss": 0.9735,
"step": 7185
},
{
"epoch": 0.83,
"grad_norm": 0.14649512552230376,
"learning_rate": 1.8084157376673473e-06,
"loss": 0.9153,
"step": 7190
},
{
"epoch": 0.83,
"grad_norm": 0.15184167762859505,
"learning_rate": 1.7969447660010752e-06,
"loss": 0.9255,
"step": 7195
},
{
"epoch": 0.83,
"grad_norm": 0.14411291108406019,
"learning_rate": 1.7855066984952873e-06,
"loss": 1.0162,
"step": 7200
},
{
"epoch": 0.83,
"grad_norm": 0.17412245799452614,
"learning_rate": 1.7741015810304474e-06,
"loss": 0.9995,
"step": 7205
},
{
"epoch": 0.83,
"grad_norm": 0.18223983842210378,
"learning_rate": 1.7627294593548582e-06,
"loss": 0.9915,
"step": 7210
},
{
"epoch": 0.83,
"grad_norm": 0.16226661934581404,
"learning_rate": 1.7513903790844634e-06,
"loss": 0.9676,
"step": 7215
},
{
"epoch": 0.83,
"grad_norm": 0.14794433641860003,
"learning_rate": 1.7400843857026707e-06,
"loss": 1.0161,
"step": 7220
},
{
"epoch": 0.83,
"grad_norm": 0.16498158730198073,
"learning_rate": 1.728811524560171e-06,
"loss": 0.97,
"step": 7225
},
{
"epoch": 0.83,
"grad_norm": 0.16191458658719166,
"learning_rate": 1.717571840874752e-06,
"loss": 0.9321,
"step": 7230
},
{
"epoch": 0.83,
"grad_norm": 0.15337354178612872,
"learning_rate": 1.7063653797311285e-06,
"loss": 0.9319,
"step": 7235
},
{
"epoch": 0.83,
"grad_norm": 0.15565300165878881,
"learning_rate": 1.6951921860807442e-06,
"loss": 0.9765,
"step": 7240
},
{
"epoch": 0.83,
"grad_norm": 0.15814394649292565,
"learning_rate": 1.684052304741598e-06,
"loss": 0.9976,
"step": 7245
},
{
"epoch": 0.83,
"grad_norm": 0.15943089539188177,
"learning_rate": 1.6729457803980809e-06,
"loss": 0.9337,
"step": 7250
},
{
"epoch": 0.83,
"grad_norm": 0.15015342210719318,
"learning_rate": 1.6618726576007593e-06,
"loss": 0.931,
"step": 7255
},
{
"epoch": 0.83,
"grad_norm": 0.13787072868532266,
"learning_rate": 1.6508329807662405e-06,
"loss": 0.9752,
"step": 7260
},
{
"epoch": 0.83,
"grad_norm": 0.14311904794393374,
"learning_rate": 1.6398267941769619e-06,
"loss": 0.9557,
"step": 7265
},
{
"epoch": 0.83,
"grad_norm": 0.16231784698756183,
"learning_rate": 1.6288541419810234e-06,
"loss": 1.0467,
"step": 7270
},
{
"epoch": 0.83,
"grad_norm": 0.16377631157347217,
"learning_rate": 1.617915068192022e-06,
"loss": 1.0489,
"step": 7275
},
{
"epoch": 0.84,
"grad_norm": 0.1407448509942888,
"learning_rate": 1.6070096166888482e-06,
"loss": 0.9888,
"step": 7280
},
{
"epoch": 0.84,
"grad_norm": 0.15178686813455647,
"learning_rate": 1.596137831215544e-06,
"loss": 0.9955,
"step": 7285
},
{
"epoch": 0.84,
"grad_norm": 0.15466318451299674,
"learning_rate": 1.585299755381099e-06,
"loss": 1.0199,
"step": 7290
},
{
"epoch": 0.84,
"grad_norm": 0.16466165859041162,
"learning_rate": 1.5744954326592865e-06,
"loss": 0.9767,
"step": 7295
},
{
"epoch": 0.84,
"grad_norm": 0.16818996765281718,
"learning_rate": 1.5637249063884985e-06,
"loss": 1.0057,
"step": 7300
},
{
"epoch": 0.84,
"grad_norm": 0.14757049504837866,
"learning_rate": 1.5529882197715474e-06,
"loss": 0.9796,
"step": 7305
},
{
"epoch": 0.84,
"grad_norm": 0.1487102377323519,
"learning_rate": 1.5422854158755229e-06,
"loss": 0.9552,
"step": 7310
},
{
"epoch": 0.84,
"grad_norm": 0.14015777449395478,
"learning_rate": 1.531616537631595e-06,
"loss": 0.9977,
"step": 7315
},
{
"epoch": 0.84,
"grad_norm": 0.15256800056558062,
"learning_rate": 1.520981627834851e-06,
"loss": 1.0223,
"step": 7320
},
{
"epoch": 0.84,
"grad_norm": 0.1448323784612338,
"learning_rate": 1.5103807291441308e-06,
"loss": 0.9488,
"step": 7325
},
{
"epoch": 0.84,
"grad_norm": 0.16090208682168544,
"learning_rate": 1.4998138840818365e-06,
"loss": 0.9798,
"step": 7330
},
{
"epoch": 0.84,
"grad_norm": 0.14197854422944003,
"learning_rate": 1.4892811350337877e-06,
"loss": 0.9152,
"step": 7335
},
{
"epoch": 0.84,
"grad_norm": 0.1504938124145082,
"learning_rate": 1.4787825242490284e-06,
"loss": 1.0177,
"step": 7340
},
{
"epoch": 0.84,
"grad_norm": 0.14362190426990482,
"learning_rate": 1.4683180938396679e-06,
"loss": 0.9935,
"step": 7345
},
{
"epoch": 0.84,
"grad_norm": 0.14565229962907184,
"learning_rate": 1.4578878857807198e-06,
"loss": 0.9205,
"step": 7350
},
{
"epoch": 0.84,
"grad_norm": 0.15596655757836664,
"learning_rate": 1.4474919419099076e-06,
"loss": 0.9908,
"step": 7355
},
{
"epoch": 0.84,
"grad_norm": 0.17804469349135607,
"learning_rate": 1.4371303039275319e-06,
"loss": 0.9753,
"step": 7360
},
{
"epoch": 0.85,
"grad_norm": 0.16370943290596937,
"learning_rate": 1.4268030133962751e-06,
"loss": 0.9747,
"step": 7365
},
{
"epoch": 0.85,
"grad_norm": 0.15145253437921677,
"learning_rate": 1.416510111741045e-06,
"loss": 0.9893,
"step": 7370
},
{
"epoch": 0.85,
"grad_norm": 0.1378249529700924,
"learning_rate": 1.4062516402488148e-06,
"loss": 0.9783,
"step": 7375
},
{
"epoch": 0.85,
"grad_norm": 0.1989269003045399,
"learning_rate": 1.3960276400684457e-06,
"loss": 1.0219,
"step": 7380
},
{
"epoch": 0.85,
"grad_norm": 0.13984535642614174,
"learning_rate": 1.385838152210527e-06,
"loss": 0.9446,
"step": 7385
},
{
"epoch": 0.85,
"grad_norm": 0.14752367548406115,
"learning_rate": 1.3756832175472157e-06,
"loss": 0.9548,
"step": 7390
},
{
"epoch": 0.85,
"grad_norm": 0.14176389354415508,
"learning_rate": 1.365562876812062e-06,
"loss": 0.9893,
"step": 7395
},
{
"epoch": 0.85,
"grad_norm": 0.1623284756198545,
"learning_rate": 1.3554771705998638e-06,
"loss": 0.9634,
"step": 7400
},
{
"epoch": 0.85,
"grad_norm": 0.1889904494595951,
"learning_rate": 1.3454261393664835e-06,
"loss": 1.0281,
"step": 7405
},
{
"epoch": 0.85,
"grad_norm": 0.14869474727032403,
"learning_rate": 1.335409823428696e-06,
"loss": 0.9477,
"step": 7410
},
{
"epoch": 0.85,
"grad_norm": 0.14975687931999646,
"learning_rate": 1.3254282629640303e-06,
"loss": 0.9671,
"step": 7415
},
{
"epoch": 0.85,
"grad_norm": 0.15690215972286173,
"learning_rate": 1.3154814980105957e-06,
"loss": 0.9598,
"step": 7420
},
{
"epoch": 0.85,
"grad_norm": 0.16315319897197336,
"learning_rate": 1.3055695684669423e-06,
"loss": 0.9984,
"step": 7425
},
{
"epoch": 0.85,
"grad_norm": 0.16676238506436128,
"learning_rate": 1.2956925140918775e-06,
"loss": 0.9149,
"step": 7430
},
{
"epoch": 0.85,
"grad_norm": 0.16148925200161718,
"learning_rate": 1.2858503745043205e-06,
"loss": 0.9913,
"step": 7435
},
{
"epoch": 0.85,
"grad_norm": 0.16083560082902632,
"learning_rate": 1.2760431891831404e-06,
"loss": 0.9565,
"step": 7440
},
{
"epoch": 0.85,
"grad_norm": 0.168202802828684,
"learning_rate": 1.266270997466994e-06,
"loss": 0.9505,
"step": 7445
},
{
"epoch": 0.85,
"grad_norm": 0.14329700185882965,
"learning_rate": 1.2565338385541792e-06,
"loss": 0.9796,
"step": 7450
},
{
"epoch": 0.86,
"grad_norm": 0.1707666646526025,
"learning_rate": 1.246831751502463e-06,
"loss": 1.0213,
"step": 7455
},
{
"epoch": 0.86,
"grad_norm": 0.14780012517015967,
"learning_rate": 1.237164775228935e-06,
"loss": 1.0108,
"step": 7460
},
{
"epoch": 0.86,
"grad_norm": 0.15159526837014772,
"learning_rate": 1.2275329485098452e-06,
"loss": 0.9818,
"step": 7465
},
{
"epoch": 0.86,
"grad_norm": 0.16164236856639663,
"learning_rate": 1.2179363099804575e-06,
"loss": 0.9661,
"step": 7470
},
{
"epoch": 0.86,
"grad_norm": 0.17373172901153616,
"learning_rate": 1.2083748981348832e-06,
"loss": 1.0211,
"step": 7475
},
{
"epoch": 0.86,
"grad_norm": 0.14723017017798046,
"learning_rate": 1.1988487513259328e-06,
"loss": 0.943,
"step": 7480
},
{
"epoch": 0.86,
"grad_norm": 0.14395658907263886,
"learning_rate": 1.1893579077649632e-06,
"loss": 0.9684,
"step": 7485
},
{
"epoch": 0.86,
"grad_norm": 0.16164130250738235,
"learning_rate": 1.179902405521718e-06,
"loss": 1.0073,
"step": 7490
},
{
"epoch": 0.86,
"grad_norm": 0.1449208079997686,
"learning_rate": 1.170482282524189e-06,
"loss": 0.9624,
"step": 7495
},
{
"epoch": 0.86,
"grad_norm": 0.14790749091039576,
"learning_rate": 1.1610975765584453e-06,
"loss": 0.9782,
"step": 7500
},
{
"epoch": 0.86,
"grad_norm": 0.16978776159053743,
"learning_rate": 1.1517483252684959e-06,
"loss": 0.9853,
"step": 7505
},
{
"epoch": 0.86,
"grad_norm": 0.14209048639467994,
"learning_rate": 1.1424345661561299e-06,
"loss": 0.9516,
"step": 7510
},
{
"epoch": 0.86,
"grad_norm": 0.17600710969928976,
"learning_rate": 1.1331563365807707e-06,
"loss": 1.0266,
"step": 7515
},
{
"epoch": 0.86,
"grad_norm": 0.15479222777832077,
"learning_rate": 1.1239136737593315e-06,
"loss": 1.0006,
"step": 7520
},
{
"epoch": 0.86,
"grad_norm": 0.25086186520233095,
"learning_rate": 1.1147066147660523e-06,
"loss": 0.9885,
"step": 7525
},
{
"epoch": 0.86,
"grad_norm": 0.15489146367341117,
"learning_rate": 1.1055351965323602e-06,
"loss": 1.0268,
"step": 7530
},
{
"epoch": 0.86,
"grad_norm": 0.15395935335630304,
"learning_rate": 1.0963994558467261e-06,
"loss": 0.9293,
"step": 7535
},
{
"epoch": 0.87,
"grad_norm": 0.17502482594663862,
"learning_rate": 1.087299429354498e-06,
"loss": 0.9799,
"step": 7540
},
{
"epoch": 0.87,
"grad_norm": 0.18707303736129866,
"learning_rate": 1.0782351535577806e-06,
"loss": 1.0076,
"step": 7545
},
{
"epoch": 0.87,
"grad_norm": 0.14531050939641618,
"learning_rate": 1.0692066648152676e-06,
"loss": 1.0355,
"step": 7550
},
{
"epoch": 0.87,
"grad_norm": 0.19747607037912387,
"learning_rate": 1.0602139993421034e-06,
"loss": 1.0087,
"step": 7555
},
{
"epoch": 0.87,
"grad_norm": 0.17290280985657078,
"learning_rate": 1.0512571932097438e-06,
"loss": 0.9697,
"step": 7560
},
{
"epoch": 0.87,
"grad_norm": 0.14344820861788155,
"learning_rate": 1.0423362823457939e-06,
"loss": 0.9484,
"step": 7565
},
{
"epoch": 0.87,
"grad_norm": 0.1490420489573449,
"learning_rate": 1.0334513025338888e-06,
"loss": 0.9507,
"step": 7570
},
{
"epoch": 0.87,
"grad_norm": 0.1375800845603591,
"learning_rate": 1.0246022894135277e-06,
"loss": 0.9675,
"step": 7575
},
{
"epoch": 0.87,
"grad_norm": 0.16472492766321487,
"learning_rate": 1.0157892784799407e-06,
"loss": 0.9904,
"step": 7580
},
{
"epoch": 0.87,
"grad_norm": 0.19482805209016107,
"learning_rate": 1.0070123050839519e-06,
"loss": 1.0614,
"step": 7585
},
{
"epoch": 0.87,
"grad_norm": 0.15549945160391881,
"learning_rate": 9.982714044318221e-07,
"loss": 0.9753,
"step": 7590
},
{
"epoch": 0.87,
"grad_norm": 0.1559125430765744,
"learning_rate": 9.895666115851254e-07,
"loss": 0.9827,
"step": 7595
},
{
"epoch": 0.87,
"grad_norm": 0.15225908648441513,
"learning_rate": 9.80897961460594e-07,
"loss": 0.988,
"step": 7600
},
{
"epoch": 0.87,
"grad_norm": 0.148809482402579,
"learning_rate": 9.722654888299853e-07,
"loss": 0.9857,
"step": 7605
},
{
"epoch": 0.87,
"grad_norm": 0.15134732145261556,
"learning_rate": 9.636692283199478e-07,
"loss": 1.0061,
"step": 7610
},
{
"epoch": 0.87,
"grad_norm": 0.1564809030230025,
"learning_rate": 9.551092144118623e-07,
"loss": 1.0184,
"step": 7615
},
{
"epoch": 0.87,
"grad_norm": 0.14754787002713818,
"learning_rate": 9.465854814417308e-07,
"loss": 0.9882,
"step": 7620
},
{
"epoch": 0.87,
"grad_norm": 0.14815991771979375,
"learning_rate": 9.380980636000158e-07,
"loss": 0.9546,
"step": 7625
},
{
"epoch": 0.88,
"grad_norm": 0.14399710850714126,
"learning_rate": 9.296469949315157e-07,
"loss": 1.0044,
"step": 7630
},
{
"epoch": 0.88,
"grad_norm": 0.15894659228151953,
"learning_rate": 9.212323093352282e-07,
"loss": 1.0145,
"step": 7635
},
{
"epoch": 0.88,
"grad_norm": 0.1606975794161914,
"learning_rate": 9.128540405641995e-07,
"loss": 0.9863,
"step": 7640
},
{
"epoch": 0.88,
"grad_norm": 0.14445650778792482,
"learning_rate": 9.045122222254154e-07,
"loss": 1.0003,
"step": 7645
},
{
"epoch": 0.88,
"grad_norm": 0.15407341650960119,
"learning_rate": 8.962068877796403e-07,
"loss": 0.9303,
"step": 7650
},
{
"epoch": 0.88,
"grad_norm": 0.156856844454904,
"learning_rate": 8.879380705412977e-07,
"loss": 0.9971,
"step": 7655
},
{
"epoch": 0.88,
"grad_norm": 0.15899673406353104,
"learning_rate": 8.797058036783379e-07,
"loss": 0.9567,
"step": 7660
},
{
"epoch": 0.88,
"grad_norm": 0.164975186018842,
"learning_rate": 8.715101202120868e-07,
"loss": 0.9947,
"step": 7665
},
{
"epoch": 0.88,
"grad_norm": 0.1680181700534043,
"learning_rate": 8.63351053017143e-07,
"loss": 0.9874,
"step": 7670
},
{
"epoch": 0.88,
"grad_norm": 0.16173732160821808,
"learning_rate": 8.55228634821218e-07,
"loss": 0.9896,
"step": 7675
},
{
"epoch": 0.88,
"grad_norm": 0.1514586397500581,
"learning_rate": 8.471428982050201e-07,
"loss": 0.9447,
"step": 7680
},
{
"epoch": 0.88,
"grad_norm": 0.1484391589569159,
"learning_rate": 8.390938756021227e-07,
"loss": 0.9954,
"step": 7685
},
{
"epoch": 0.88,
"grad_norm": 0.15643703876438503,
"learning_rate": 8.310815992988286e-07,
"loss": 1.0181,
"step": 7690
},
{
"epoch": 0.88,
"grad_norm": 0.15570909005909317,
"learning_rate": 8.231061014340436e-07,
"loss": 0.9415,
"step": 7695
},
{
"epoch": 0.88,
"grad_norm": 0.16886783625004367,
"learning_rate": 8.151674139991483e-07,
"loss": 0.9953,
"step": 7700
},
{
"epoch": 0.88,
"grad_norm": 0.15278959440201248,
"learning_rate": 8.072655688378683e-07,
"loss": 0.9736,
"step": 7705
},
{
"epoch": 0.88,
"grad_norm": 0.1577880997639766,
"learning_rate": 7.994005976461494e-07,
"loss": 1.0047,
"step": 7710
},
{
"epoch": 0.89,
"grad_norm": 0.15213121645538813,
"learning_rate": 7.915725319720257e-07,
"loss": 0.9574,
"step": 7715
},
{
"epoch": 0.89,
"grad_norm": 0.20507394873600635,
"learning_rate": 7.837814032154978e-07,
"loss": 0.9303,
"step": 7720
},
{
"epoch": 0.89,
"grad_norm": 0.1595768820575974,
"learning_rate": 7.760272426284021e-07,
"loss": 0.9818,
"step": 7725
},
{
"epoch": 0.89,
"grad_norm": 0.142774194172215,
"learning_rate": 7.683100813142886e-07,
"loss": 1.0231,
"step": 7730
},
{
"epoch": 0.89,
"grad_norm": 0.15624847694172178,
"learning_rate": 7.606299502282977e-07,
"loss": 0.9878,
"step": 7735
},
{
"epoch": 0.89,
"grad_norm": 0.1584452180446109,
"learning_rate": 7.529868801770324e-07,
"loss": 0.9799,
"step": 7740
},
{
"epoch": 0.89,
"grad_norm": 0.1576982253551169,
"learning_rate": 7.453809018184365e-07,
"loss": 1.009,
"step": 7745
},
{
"epoch": 0.89,
"grad_norm": 0.13601273591742913,
"learning_rate": 7.378120456616677e-07,
"loss": 1.032,
"step": 7750
},
{
"epoch": 0.89,
"grad_norm": 0.1350632831555549,
"learning_rate": 7.302803420669868e-07,
"loss": 0.9484,
"step": 7755
},
{
"epoch": 0.89,
"grad_norm": 0.15438762375533158,
"learning_rate": 7.227858212456207e-07,
"loss": 0.9747,
"step": 7760
},
{
"epoch": 0.89,
"grad_norm": 0.15607621061238972,
"learning_rate": 7.15328513259651e-07,
"loss": 1.0129,
"step": 7765
},
{
"epoch": 0.89,
"grad_norm": 0.15667049838586844,
"learning_rate": 7.079084480218912e-07,
"loss": 0.979,
"step": 7770
},
{
"epoch": 0.89,
"grad_norm": 0.1528497547857842,
"learning_rate": 7.005256552957651e-07,
"loss": 0.9477,
"step": 7775
},
{
"epoch": 0.89,
"grad_norm": 0.15172926224108654,
"learning_rate": 6.931801646951919e-07,
"loss": 0.9948,
"step": 7780
},
{
"epoch": 0.89,
"grad_norm": 0.14987536970341647,
"learning_rate": 6.858720056844614e-07,
"loss": 0.9541,
"step": 7785
},
{
"epoch": 0.89,
"grad_norm": 0.1466647661726728,
"learning_rate": 6.786012075781179e-07,
"loss": 1.0185,
"step": 7790
},
{
"epoch": 0.89,
"grad_norm": 0.15635878981478488,
"learning_rate": 6.713677995408452e-07,
"loss": 1.0102,
"step": 7795
},
{
"epoch": 0.9,
"grad_norm": 0.15509405699868806,
"learning_rate": 6.641718105873462e-07,
"loss": 0.9772,
"step": 7800
},
{
"epoch": 0.9,
"grad_norm": 0.1436230896716419,
"learning_rate": 6.570132695822318e-07,
"loss": 0.9565,
"step": 7805
},
{
"epoch": 0.9,
"grad_norm": 0.15271872082790874,
"learning_rate": 6.49892205239897e-07,
"loss": 0.9601,
"step": 7810
},
{
"epoch": 0.9,
"grad_norm": 0.15375246981059396,
"learning_rate": 6.42808646124411e-07,
"loss": 0.9743,
"step": 7815
},
{
"epoch": 0.9,
"grad_norm": 0.14509077800106734,
"learning_rate": 6.357626206494061e-07,
"loss": 0.9555,
"step": 7820
},
{
"epoch": 0.9,
"grad_norm": 0.1458168477588119,
"learning_rate": 6.287541570779499e-07,
"loss": 0.9957,
"step": 7825
},
{
"epoch": 0.9,
"grad_norm": 0.18070625349676658,
"learning_rate": 6.217832835224524e-07,
"loss": 0.9449,
"step": 7830
},
{
"epoch": 0.9,
"grad_norm": 0.18047403698472697,
"learning_rate": 6.148500279445358e-07,
"loss": 0.9488,
"step": 7835
},
{
"epoch": 0.9,
"grad_norm": 0.17691205808463678,
"learning_rate": 6.079544181549269e-07,
"loss": 1.0276,
"step": 7840
},
{
"epoch": 0.9,
"grad_norm": 0.15116229852284016,
"learning_rate": 6.010964818133569e-07,
"loss": 0.9453,
"step": 7845
},
{
"epoch": 0.9,
"grad_norm": 0.15590129725507204,
"learning_rate": 5.94276246428428e-07,
"loss": 0.9769,
"step": 7850
},
{
"epoch": 0.9,
"grad_norm": 0.1437746502074229,
"learning_rate": 5.874937393575276e-07,
"loss": 0.9693,
"step": 7855
},
{
"epoch": 0.9,
"grad_norm": 0.15831764192442618,
"learning_rate": 5.807489878067008e-07,
"loss": 1.0076,
"step": 7860
},
{
"epoch": 0.9,
"grad_norm": 0.1529276718971069,
"learning_rate": 5.740420188305451e-07,
"loss": 0.9514,
"step": 7865
},
{
"epoch": 0.9,
"grad_norm": 0.15016913525105133,
"learning_rate": 5.673728593321148e-07,
"loss": 0.9567,
"step": 7870
},
{
"epoch": 0.9,
"grad_norm": 0.15948800098996255,
"learning_rate": 5.607415360627899e-07,
"loss": 1.0123,
"step": 7875
},
{
"epoch": 0.9,
"grad_norm": 0.1450871521957898,
"learning_rate": 5.541480756221884e-07,
"loss": 1.0018,
"step": 7880
},
{
"epoch": 0.9,
"grad_norm": 0.16456194248403688,
"learning_rate": 5.475925044580533e-07,
"loss": 0.9339,
"step": 7885
},
{
"epoch": 0.91,
"grad_norm": 0.16132528012657435,
"learning_rate": 5.410748488661411e-07,
"loss": 0.9691,
"step": 7890
},
{
"epoch": 0.91,
"grad_norm": 0.15874755825037057,
"learning_rate": 5.345951349901268e-07,
"loss": 1.0051,
"step": 7895
},
{
"epoch": 0.91,
"grad_norm": 0.16750449188671507,
"learning_rate": 5.281533888214862e-07,
"loss": 1.0105,
"step": 7900
},
{
"epoch": 0.91,
"grad_norm": 0.15332731716866801,
"learning_rate": 5.217496361994078e-07,
"loss": 0.9977,
"step": 7905
},
{
"epoch": 0.91,
"grad_norm": 0.1659473369453083,
"learning_rate": 5.153839028106711e-07,
"loss": 1.0256,
"step": 7910
},
{
"epoch": 0.91,
"grad_norm": 0.16570079039096247,
"learning_rate": 5.090562141895583e-07,
"loss": 0.9517,
"step": 7915
},
{
"epoch": 0.91,
"grad_norm": 0.16577327224089683,
"learning_rate": 5.027665957177453e-07,
"loss": 1.0242,
"step": 7920
},
{
"epoch": 0.91,
"grad_norm": 0.14845384532507355,
"learning_rate": 4.965150726241974e-07,
"loss": 1.0188,
"step": 7925
},
{
"epoch": 0.91,
"grad_norm": 0.13792632882246528,
"learning_rate": 4.903016699850749e-07,
"loss": 0.959,
"step": 7930
},
{
"epoch": 0.91,
"grad_norm": 0.14497758515385686,
"learning_rate": 4.84126412723629e-07,
"loss": 1.015,
"step": 7935
},
{
"epoch": 0.91,
"grad_norm": 0.13825721094885277,
"learning_rate": 4.779893256100987e-07,
"loss": 0.9924,
"step": 7940
},
{
"epoch": 0.91,
"grad_norm": 0.14540764250457264,
"learning_rate": 4.7189043326162143e-07,
"loss": 0.9464,
"step": 7945
},
{
"epoch": 0.91,
"grad_norm": 0.17298644277821076,
"learning_rate": 4.6582976014211467e-07,
"loss": 0.9356,
"step": 7950
},
{
"epoch": 0.91,
"grad_norm": 0.1466172919586259,
"learning_rate": 4.598073305622064e-07,
"loss": 0.9585,
"step": 7955
},
{
"epoch": 0.91,
"grad_norm": 0.1526386541405698,
"learning_rate": 4.538231686791106e-07,
"loss": 0.9481,
"step": 7960
},
{
"epoch": 0.91,
"grad_norm": 0.16457238893585135,
"learning_rate": 4.4787729849654295e-07,
"loss": 0.9603,
"step": 7965
},
{
"epoch": 0.91,
"grad_norm": 0.1886575559060366,
"learning_rate": 4.4196974386463085e-07,
"loss": 1.0029,
"step": 7970
},
{
"epoch": 0.92,
"grad_norm": 0.1469274330265399,
"learning_rate": 4.3610052847980033e-07,
"loss": 0.8772,
"step": 7975
},
{
"epoch": 0.92,
"grad_norm": 0.16318962371851223,
"learning_rate": 4.3026967588469694e-07,
"loss": 0.9909,
"step": 7980
},
{
"epoch": 0.92,
"grad_norm": 0.14232725457186637,
"learning_rate": 4.2447720946808156e-07,
"loss": 0.9741,
"step": 7985
},
{
"epoch": 0.92,
"grad_norm": 0.15374261055956523,
"learning_rate": 4.187231524647406e-07,
"loss": 0.9778,
"step": 7990
},
{
"epoch": 0.92,
"grad_norm": 0.17738824379627186,
"learning_rate": 4.130075279553969e-07,
"loss": 0.9695,
"step": 7995
},
{
"epoch": 0.92,
"grad_norm": 0.1484765025474173,
"learning_rate": 4.0733035886660556e-07,
"loss": 1.0142,
"step": 8000
},
{
"epoch": 0.92,
"grad_norm": 0.1729080857316514,
"learning_rate": 4.0169166797067504e-07,
"loss": 1.0291,
"step": 8005
},
{
"epoch": 0.92,
"grad_norm": 0.14473446399217213,
"learning_rate": 3.9609147788556623e-07,
"loss": 0.9685,
"step": 8010
},
{
"epoch": 0.92,
"grad_norm": 0.16326484843700462,
"learning_rate": 3.905298110748046e-07,
"loss": 0.9775,
"step": 8015
},
{
"epoch": 0.92,
"grad_norm": 0.1649722355177796,
"learning_rate": 3.850066898473948e-07,
"loss": 1.0144,
"step": 8020
},
{
"epoch": 0.92,
"grad_norm": 0.14740811569768472,
"learning_rate": 3.7952213635772395e-07,
"loss": 0.9647,
"step": 8025
},
{
"epoch": 0.92,
"grad_norm": 0.14455541089990448,
"learning_rate": 3.7407617260547536e-07,
"loss": 0.9721,
"step": 8030
},
{
"epoch": 0.92,
"grad_norm": 0.1583879507287872,
"learning_rate": 3.6866882043554375e-07,
"loss": 0.9908,
"step": 8035
},
{
"epoch": 0.92,
"grad_norm": 0.15999382464192402,
"learning_rate": 3.633001015379423e-07,
"loss": 0.9784,
"step": 8040
},
{
"epoch": 0.92,
"grad_norm": 0.15988392540313173,
"learning_rate": 3.5797003744772017e-07,
"loss": 0.9874,
"step": 8045
},
{
"epoch": 0.92,
"grad_norm": 0.1539978466978224,
"learning_rate": 3.5267864954487197e-07,
"loss": 0.9793,
"step": 8050
},
{
"epoch": 0.92,
"grad_norm": 0.1446777000895735,
"learning_rate": 3.4742595905425504e-07,
"loss": 0.9901,
"step": 8055
},
{
"epoch": 0.92,
"grad_norm": 0.16048050796535274,
"learning_rate": 3.422119870455021e-07,
"loss": 0.9807,
"step": 8060
},
{
"epoch": 0.93,
"grad_norm": 0.1543701246761387,
"learning_rate": 3.3703675443294113e-07,
"loss": 0.9766,
"step": 8065
},
{
"epoch": 0.93,
"grad_norm": 0.18166977897055536,
"learning_rate": 3.3190028197550684e-07,
"loss": 0.997,
"step": 8070
},
{
"epoch": 0.93,
"grad_norm": 0.15231836615615033,
"learning_rate": 3.2680259027665696e-07,
"loss": 0.9774,
"step": 8075
},
{
"epoch": 0.93,
"grad_norm": 0.18309368883486862,
"learning_rate": 3.2174369978429374e-07,
"loss": 0.998,
"step": 8080
},
{
"epoch": 0.93,
"grad_norm": 0.15688811679884562,
"learning_rate": 3.1672363079067734e-07,
"loss": 0.9586,
"step": 8085
},
{
"epoch": 0.93,
"grad_norm": 0.15255732035618555,
"learning_rate": 3.1174240343235017e-07,
"loss": 0.9481,
"step": 8090
},
{
"epoch": 0.93,
"grad_norm": 0.1644021992742689,
"learning_rate": 3.0680003769005153e-07,
"loss": 0.9435,
"step": 8095
},
{
"epoch": 0.93,
"grad_norm": 0.21534856866624286,
"learning_rate": 3.018965533886364e-07,
"loss": 0.9757,
"step": 8100
},
{
"epoch": 0.93,
"grad_norm": 0.15848360232339143,
"learning_rate": 2.9703197019700367e-07,
"loss": 0.9442,
"step": 8105
},
{
"epoch": 0.93,
"grad_norm": 0.15207816917734795,
"learning_rate": 2.922063076280035e-07,
"loss": 0.9376,
"step": 8110
},
{
"epoch": 0.93,
"grad_norm": 0.15776984810390274,
"learning_rate": 2.874195850383743e-07,
"loss": 0.9792,
"step": 8115
},
{
"epoch": 0.93,
"grad_norm": 0.1518008313490981,
"learning_rate": 2.826718216286561e-07,
"loss": 0.9481,
"step": 8120
},
{
"epoch": 0.93,
"grad_norm": 0.13844474891090167,
"learning_rate": 2.7796303644311405e-07,
"loss": 0.9858,
"step": 8125
},
{
"epoch": 0.93,
"grad_norm": 0.15301269010882665,
"learning_rate": 2.732932483696682e-07,
"loss": 0.9552,
"step": 8130
},
{
"epoch": 0.93,
"grad_norm": 0.15430343626041787,
"learning_rate": 2.686624761398049e-07,
"loss": 1.0399,
"step": 8135
},
{
"epoch": 0.93,
"grad_norm": 0.174745739212343,
"learning_rate": 2.6407073832851684e-07,
"loss": 0.9964,
"step": 8140
},
{
"epoch": 0.93,
"grad_norm": 0.15885723640260319,
"learning_rate": 2.5951805335421966e-07,
"loss": 1.0068,
"step": 8145
},
{
"epoch": 0.94,
"grad_norm": 0.17928986207788275,
"learning_rate": 2.550044394786788e-07,
"loss": 1.001,
"step": 8150
},
{
"epoch": 0.94,
"grad_norm": 0.1708911027395856,
"learning_rate": 2.5052991480694067e-07,
"loss": 0.9682,
"step": 8155
},
{
"epoch": 0.94,
"grad_norm": 0.165778187909741,
"learning_rate": 2.460944972872492e-07,
"loss": 1.0144,
"step": 8160
},
{
"epoch": 0.94,
"grad_norm": 0.14324536645636454,
"learning_rate": 2.416982047109917e-07,
"loss": 1.003,
"step": 8165
},
{
"epoch": 0.94,
"grad_norm": 0.18739075539822905,
"learning_rate": 2.373410547126087e-07,
"loss": 0.9763,
"step": 8170
},
{
"epoch": 0.94,
"grad_norm": 0.15908576729401405,
"learning_rate": 2.3302306476953528e-07,
"loss": 0.9676,
"step": 8175
},
{
"epoch": 0.94,
"grad_norm": 0.14300810923211452,
"learning_rate": 2.2874425220212992e-07,
"loss": 1.0044,
"step": 8180
},
{
"epoch": 0.94,
"grad_norm": 0.14097048421641786,
"learning_rate": 2.245046341735957e-07,
"loss": 0.9442,
"step": 8185
},
{
"epoch": 0.94,
"grad_norm": 0.1424806875998385,
"learning_rate": 2.2030422768992476e-07,
"loss": 0.9518,
"step": 8190
},
{
"epoch": 0.94,
"grad_norm": 0.15779607172909144,
"learning_rate": 2.161430495998207e-07,
"loss": 0.9465,
"step": 8195
},
{
"epoch": 0.94,
"grad_norm": 0.14655229106958392,
"learning_rate": 2.1202111659463175e-07,
"loss": 0.9238,
"step": 8200
},
{
"epoch": 0.94,
"grad_norm": 0.15659674434447207,
"learning_rate": 2.079384452082933e-07,
"loss": 0.9874,
"step": 8205
},
{
"epoch": 0.94,
"grad_norm": 0.15362428128750674,
"learning_rate": 2.0389505181724445e-07,
"loss": 1.0195,
"step": 8210
},
{
"epoch": 0.94,
"grad_norm": 0.14402140289254575,
"learning_rate": 1.9989095264038027e-07,
"loss": 0.9765,
"step": 8215
},
{
"epoch": 0.94,
"grad_norm": 0.14684868831803946,
"learning_rate": 1.9592616373897421e-07,
"loss": 0.9651,
"step": 8220
},
{
"epoch": 0.94,
"grad_norm": 0.16473086700167075,
"learning_rate": 1.9200070101661805e-07,
"loss": 0.9977,
"step": 8225
},
{
"epoch": 0.94,
"grad_norm": 0.1537244430943279,
"learning_rate": 1.8811458021916085e-07,
"loss": 0.9998,
"step": 8230
},
{
"epoch": 0.94,
"grad_norm": 0.15760055795852243,
"learning_rate": 1.8426781693464012e-07,
"loss": 0.97,
"step": 8235
},
{
"epoch": 0.95,
"grad_norm": 0.14813580581300442,
"learning_rate": 1.8046042659322304e-07,
"loss": 0.9592,
"step": 8240
},
{
"epoch": 0.95,
"grad_norm": 0.15896124229085856,
"learning_rate": 1.766924244671453e-07,
"loss": 0.9529,
"step": 8245
},
{
"epoch": 0.95,
"grad_norm": 0.1821836692462499,
"learning_rate": 1.7296382567064674e-07,
"loss": 1.0003,
"step": 8250
},
{
"epoch": 0.95,
"grad_norm": 0.15539371297632606,
"learning_rate": 1.6927464515991142e-07,
"loss": 0.9786,
"step": 8255
},
{
"epoch": 0.95,
"grad_norm": 0.15219630124346206,
"learning_rate": 1.6562489773301215e-07,
"loss": 0.9514,
"step": 8260
},
{
"epoch": 0.95,
"grad_norm": 0.16066988731571322,
"learning_rate": 1.6201459802984154e-07,
"loss": 1.0214,
"step": 8265
},
{
"epoch": 0.95,
"grad_norm": 0.16329280667065701,
"learning_rate": 1.5844376053206546e-07,
"loss": 0.9755,
"step": 8270
},
{
"epoch": 0.95,
"grad_norm": 0.1666492068813255,
"learning_rate": 1.5491239956305305e-07,
"loss": 0.971,
"step": 8275
},
{
"epoch": 0.95,
"grad_norm": 0.15599542702758698,
"learning_rate": 1.5142052928782902e-07,
"loss": 0.9548,
"step": 8280
},
{
"epoch": 0.95,
"grad_norm": 0.18065911339777607,
"learning_rate": 1.479681637130126e-07,
"loss": 0.9761,
"step": 8285
},
{
"epoch": 0.95,
"grad_norm": 0.14331275918483952,
"learning_rate": 1.4455531668675747e-07,
"loss": 0.9745,
"step": 8290
},
{
"epoch": 0.95,
"grad_norm": 0.16303294628487286,
"learning_rate": 1.4118200189870424e-07,
"loss": 0.9874,
"step": 8295
},
{
"epoch": 0.95,
"grad_norm": 0.1458185163827089,
"learning_rate": 1.378482328799191e-07,
"loss": 0.997,
"step": 8300
},
{
"epoch": 0.95,
"grad_norm": 0.1742311973838547,
"learning_rate": 1.3455402300284416e-07,
"loss": 1.0286,
"step": 8305
},
{
"epoch": 0.95,
"grad_norm": 0.16384279681165212,
"learning_rate": 1.3129938548123945e-07,
"loss": 1.0069,
"step": 8310
},
{
"epoch": 0.95,
"grad_norm": 0.14936147036875858,
"learning_rate": 1.2808433337013316e-07,
"loss": 1.0013,
"step": 8315
},
{
"epoch": 0.95,
"grad_norm": 0.16632188124151814,
"learning_rate": 1.2490887956576825e-07,
"loss": 0.949,
"step": 8320
},
{
"epoch": 0.96,
"grad_norm": 0.1591284349016137,
"learning_rate": 1.217730368055492e-07,
"loss": 1.0261,
"step": 8325
},
{
"epoch": 0.96,
"grad_norm": 0.15135545171795708,
"learning_rate": 1.1867681766799532e-07,
"loss": 0.976,
"step": 8330
},
{
"epoch": 0.96,
"grad_norm": 0.16182276286127703,
"learning_rate": 1.1562023457268423e-07,
"loss": 0.9972,
"step": 8335
},
{
"epoch": 0.96,
"grad_norm": 0.17547902107932337,
"learning_rate": 1.1260329978020624e-07,
"loss": 1.0,
"step": 8340
},
{
"epoch": 0.96,
"grad_norm": 0.17142814065291342,
"learning_rate": 1.0962602539211331e-07,
"loss": 0.9771,
"step": 8345
},
{
"epoch": 0.96,
"grad_norm": 0.16777274726965624,
"learning_rate": 1.066884233508747e-07,
"loss": 0.9648,
"step": 8350
},
{
"epoch": 0.96,
"grad_norm": 0.17071206777952214,
"learning_rate": 1.0379050543982139e-07,
"loss": 1.0062,
"step": 8355
},
{
"epoch": 0.96,
"grad_norm": 0.1681542128710368,
"learning_rate": 1.0093228328310611e-07,
"loss": 0.9626,
"step": 8360
},
{
"epoch": 0.96,
"grad_norm": 0.148473295631225,
"learning_rate": 9.811376834565123e-08,
"loss": 0.9629,
"step": 8365
},
{
"epoch": 0.96,
"grad_norm": 0.1546463807015304,
"learning_rate": 9.533497193310537e-08,
"loss": 1.028,
"step": 8370
},
{
"epoch": 0.96,
"grad_norm": 0.14826358292564817,
"learning_rate": 9.259590519180017e-08,
"loss": 1.013,
"step": 8375
},
{
"epoch": 0.96,
"grad_norm": 0.1358573900291209,
"learning_rate": 8.989657910870253e-08,
"loss": 1.008,
"step": 8380
},
{
"epoch": 0.96,
"grad_norm": 0.16295931046497253,
"learning_rate": 8.723700451136907e-08,
"loss": 0.9659,
"step": 8385
},
{
"epoch": 0.96,
"grad_norm": 0.13999582980217742,
"learning_rate": 8.461719206790842e-08,
"loss": 1.0064,
"step": 8390
},
{
"epoch": 0.96,
"grad_norm": 0.15511954406988068,
"learning_rate": 8.203715228693121e-08,
"loss": 0.9447,
"step": 8395
},
{
"epoch": 0.96,
"grad_norm": 0.1515514246412033,
"learning_rate": 7.949689551751572e-08,
"loss": 1.0056,
"step": 8400
},
{
"epoch": 0.96,
"grad_norm": 0.16221882454456843,
"learning_rate": 7.699643194915784e-08,
"loss": 0.9462,
"step": 8405
},
{
"epoch": 0.97,
"grad_norm": 0.18551463233245163,
"learning_rate": 7.453577161173786e-08,
"loss": 1.0052,
"step": 8410
},
{
"epoch": 0.97,
"grad_norm": 0.15317068737827436,
"learning_rate": 7.211492437547817e-08,
"loss": 0.9868,
"step": 8415
},
{
"epoch": 0.97,
"grad_norm": 0.1529156593411024,
"learning_rate": 6.973389995089675e-08,
"loss": 1.0024,
"step": 8420
},
{
"epoch": 0.97,
"grad_norm": 0.16578368013819816,
"learning_rate": 6.739270788878038e-08,
"loss": 0.9799,
"step": 8425
},
{
"epoch": 0.97,
"grad_norm": 0.15052136559331336,
"learning_rate": 6.509135758013596e-08,
"loss": 1.0144,
"step": 8430
},
{
"epoch": 0.97,
"grad_norm": 0.14049753240709992,
"learning_rate": 6.282985825615928e-08,
"loss": 1.03,
"step": 8435
},
{
"epoch": 0.97,
"grad_norm": 0.1617981840823852,
"learning_rate": 6.060821898819402e-08,
"loss": 0.9503,
"step": 8440
},
{
"epoch": 0.97,
"grad_norm": 0.16362025767916583,
"learning_rate": 5.8426448687696204e-08,
"loss": 0.9716,
"step": 8445
},
{
"epoch": 0.97,
"grad_norm": 0.17277131093229767,
"learning_rate": 5.628455610620309e-08,
"loss": 1.0116,
"step": 8450
},
{
"epoch": 0.97,
"grad_norm": 0.16354457993420854,
"learning_rate": 5.41825498352877e-08,
"loss": 0.9366,
"step": 8455
},
{
"epoch": 0.97,
"grad_norm": 0.14809370501603067,
"learning_rate": 5.212043830653657e-08,
"loss": 0.9779,
"step": 8460
},
{
"epoch": 0.97,
"grad_norm": 0.14662321872127645,
"learning_rate": 5.009822979150758e-08,
"loss": 1.0012,
"step": 8465
},
{
"epoch": 0.97,
"grad_norm": 0.16227017046094616,
"learning_rate": 4.8115932401698874e-08,
"loss": 1.0095,
"step": 8470
},
{
"epoch": 0.97,
"grad_norm": 0.1698386872512351,
"learning_rate": 4.617355408851776e-08,
"loss": 0.9684,
"step": 8475
},
{
"epoch": 0.97,
"grad_norm": 0.15450705328620457,
"learning_rate": 4.427110264324852e-08,
"loss": 0.9034,
"step": 8480
},
{
"epoch": 0.97,
"grad_norm": 0.15035138838459883,
"learning_rate": 4.2408585697019114e-08,
"loss": 0.9238,
"step": 8485
},
{
"epoch": 0.97,
"grad_norm": 0.1494510194203988,
"learning_rate": 4.0586010720772284e-08,
"loss": 1.0039,
"step": 8490
},
{
"epoch": 0.97,
"grad_norm": 0.15651267651716128,
"learning_rate": 3.880338502523451e-08,
"loss": 0.9559,
"step": 8495
},
{
"epoch": 0.98,
"grad_norm": 0.14935437132807455,
"learning_rate": 3.706071576088821e-08,
"loss": 0.9814,
"step": 8500
},
{
"epoch": 0.98,
"grad_norm": 0.17038859828693223,
"learning_rate": 3.535800991794402e-08,
"loss": 0.984,
"step": 8505
},
{
"epoch": 0.98,
"grad_norm": 0.14244079221296307,
"learning_rate": 3.369527432630748e-08,
"loss": 0.9747,
"step": 8510
},
{
"epoch": 0.98,
"grad_norm": 0.15469240243761584,
"learning_rate": 3.207251565555791e-08,
"loss": 0.9576,
"step": 8515
},
{
"epoch": 0.98,
"grad_norm": 0.14843445877357878,
"learning_rate": 3.048974041491626e-08,
"loss": 0.9818,
"step": 8520
},
{
"epoch": 0.98,
"grad_norm": 0.1540221873981517,
"learning_rate": 2.8946954953223972e-08,
"loss": 0.9663,
"step": 8525
},
{
"epoch": 0.98,
"grad_norm": 0.15940209006836195,
"learning_rate": 2.744416545891526e-08,
"loss": 0.9718,
"step": 8530
},
{
"epoch": 0.98,
"grad_norm": 0.16464259110836912,
"learning_rate": 2.598137795999045e-08,
"loss": 0.9651,
"step": 8535
},
{
"epoch": 0.98,
"grad_norm": 0.15982550191164843,
"learning_rate": 2.4558598323994876e-08,
"loss": 0.9849,
"step": 8540
},
{
"epoch": 0.98,
"grad_norm": 0.1532348406465396,
"learning_rate": 2.3175832257992248e-08,
"loss": 0.9618,
"step": 8545
},
{
"epoch": 0.98,
"grad_norm": 0.17064142600350107,
"learning_rate": 2.1833085308546887e-08,
"loss": 1.0002,
"step": 8550
},
{
"epoch": 0.98,
"grad_norm": 0.15842395161909745,
"learning_rate": 2.053036286169485e-08,
"loss": 0.9544,
"step": 8555
},
{
"epoch": 0.98,
"grad_norm": 0.14139424539155257,
"learning_rate": 1.9267670142926187e-08,
"loss": 0.9914,
"step": 8560
},
{
"epoch": 0.98,
"grad_norm": 0.15847300804017894,
"learning_rate": 1.8045012217164925e-08,
"loss": 0.9383,
"step": 8565
},
{
"epoch": 0.98,
"grad_norm": 0.16899308169946436,
"learning_rate": 1.6862393988746895e-08,
"loss": 0.9939,
"step": 8570
},
{
"epoch": 0.98,
"grad_norm": 0.14940061690292197,
"learning_rate": 1.5719820201398617e-08,
"loss": 0.9417,
"step": 8575
},
{
"epoch": 0.98,
"grad_norm": 0.15944810502081894,
"learning_rate": 1.4617295438220657e-08,
"loss": 1.002,
"step": 8580
},
{
"epoch": 0.99,
"grad_norm": 0.15241774916033593,
"learning_rate": 1.3554824121669862e-08,
"loss": 0.9246,
"step": 8585
},
{
"epoch": 0.99,
"grad_norm": 0.16641284050730326,
"learning_rate": 1.2532410513540483e-08,
"loss": 0.9255,
"step": 8590
},
{
"epoch": 0.99,
"grad_norm": 0.14601375039200526,
"learning_rate": 1.15500587149453e-08,
"loss": 0.937,
"step": 8595
},
{
"epoch": 0.99,
"grad_norm": 0.1621965223626885,
"learning_rate": 1.0607772666302308e-08,
"loss": 1.0129,
"step": 8600
},
{
"epoch": 0.99,
"grad_norm": 0.1435022924006593,
"learning_rate": 9.705556147314721e-09,
"loss": 0.9793,
"step": 8605
},
{
"epoch": 0.99,
"grad_norm": 0.14798192771688623,
"learning_rate": 8.843412776963211e-09,
"loss": 0.9579,
"step": 8610
},
{
"epoch": 0.99,
"grad_norm": 0.15862211866390744,
"learning_rate": 8.02134601348259e-09,
"loss": 1.0096,
"step": 8615
},
{
"epoch": 0.99,
"grad_norm": 0.15145671497853636,
"learning_rate": 7.239359154354031e-09,
"loss": 0.9725,
"step": 8620
},
{
"epoch": 0.99,
"grad_norm": 0.14104006822291443,
"learning_rate": 6.497455336289537e-09,
"loss": 0.9971,
"step": 8625
},
{
"epoch": 0.99,
"grad_norm": 0.16535917440584985,
"learning_rate": 5.795637535219722e-09,
"loss": 0.967,
"step": 8630
},
{
"epoch": 0.99,
"grad_norm": 0.15152431418624868,
"learning_rate": 5.133908566281598e-09,
"loss": 0.9911,
"step": 8635
},
{
"epoch": 0.99,
"grad_norm": 0.15273678452685552,
"learning_rate": 4.5122710838074735e-09,
"loss": 1.0159,
"step": 8640
},
{
"epoch": 0.99,
"grad_norm": 0.15736855705370478,
"learning_rate": 3.930727581314963e-09,
"loss": 0.994,
"step": 8645
},
{
"epoch": 0.99,
"grad_norm": 0.1427230482941389,
"learning_rate": 3.3892803914958862e-09,
"loss": 0.994,
"step": 8650
},
{
"epoch": 0.99,
"grad_norm": 0.1523616525750465,
"learning_rate": 2.8879316862073792e-09,
"loss": 0.9313,
"step": 8655
},
{
"epoch": 0.99,
"grad_norm": 0.13415504410501358,
"learning_rate": 2.426683476464131e-09,
"loss": 1.0082,
"step": 8660
},
{
"epoch": 0.99,
"grad_norm": 0.14466296831569572,
"learning_rate": 2.005537612426167e-09,
"loss": 0.9347,
"step": 8665
},
{
"epoch": 0.99,
"grad_norm": 0.14336944049823355,
"learning_rate": 1.6244957833999598e-09,
"loss": 1.0241,
"step": 8670
},
{
"epoch": 1.0,
"grad_norm": 0.18071345955146673,
"learning_rate": 1.2835595178217752e-09,
"loss": 0.9723,
"step": 8675
},
{
"epoch": 1.0,
"grad_norm": 0.17655800078764253,
"learning_rate": 9.827301832598945e-10,
"loss": 0.9892,
"step": 8680
},
{
"epoch": 1.0,
"grad_norm": 0.1561528705190895,
"learning_rate": 7.220089864012903e-10,
"loss": 0.9619,
"step": 8685
},
{
"epoch": 1.0,
"grad_norm": 0.16720009270662858,
"learning_rate": 5.013969730527368e-10,
"loss": 1.024,
"step": 8690
},
{
"epoch": 1.0,
"grad_norm": 0.23675211325845877,
"learning_rate": 3.2089502813636987e-10,
"loss": 0.9946,
"step": 8695
},
{
"epoch": 1.0,
"grad_norm": 0.1664760617901683,
"learning_rate": 1.8050387568302463e-10,
"loss": 0.9711,
"step": 8700
},
{
"epoch": 1.0,
"grad_norm": 0.15503154684167228,
"learning_rate": 8.022407882890549e-11,
"loss": 0.9637,
"step": 8705
},
{
"epoch": 1.0,
"grad_norm": 0.14590541249765324,
"learning_rate": 2.0056039820026684e-11,
"loss": 0.9911,
"step": 8710
},
{
"epoch": 1.0,
"grad_norm": 0.15090805904067295,
"learning_rate": 0.0,
"loss": 0.9598,
"step": 8715
},
{
"epoch": 1.0,
"eval_loss": 0.9879439473152161,
"eval_runtime": 1766.0344,
"eval_samples_per_second": 8.738,
"eval_steps_per_second": 1.092,
"step": 8715
},
{
"epoch": 1.0,
"step": 8715,
"total_flos": 456185924812800.0,
"train_loss": 0.9989171136335196,
"train_runtime": 47811.9362,
"train_samples_per_second": 2.916,
"train_steps_per_second": 0.182
}
],
"logging_steps": 5,
"max_steps": 8715,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"total_flos": 456185924812800.0,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}