cube_stack_allex_image8_cvla / trainer_state.json
arunos728's picture
Upload trainer_state.json
ec6b12f verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 76.92307692307692,
"eval_steps": 500,
"global_step": 20000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.038461538461538464,
"grad_norm": 0.8465695381164551,
"learning_rate": 2.7e-07,
"loss": 1.3641,
"step": 10
},
{
"epoch": 0.07692307692307693,
"grad_norm": 0.6664445400238037,
"learning_rate": 5.7e-07,
"loss": 1.3695,
"step": 20
},
{
"epoch": 0.11538461538461539,
"grad_norm": 0.7535883784294128,
"learning_rate": 8.7e-07,
"loss": 1.3618,
"step": 30
},
{
"epoch": 0.15384615384615385,
"grad_norm": 0.6306048631668091,
"learning_rate": 1.17e-06,
"loss": 1.3613,
"step": 40
},
{
"epoch": 0.19230769230769232,
"grad_norm": 0.778509795665741,
"learning_rate": 1.4700000000000001e-06,
"loss": 1.3601,
"step": 50
},
{
"epoch": 0.23076923076923078,
"grad_norm": 0.8616118431091309,
"learning_rate": 1.77e-06,
"loss": 1.3455,
"step": 60
},
{
"epoch": 0.2692307692307692,
"grad_norm": 0.9706582427024841,
"learning_rate": 2.07e-06,
"loss": 1.3393,
"step": 70
},
{
"epoch": 0.3076923076923077,
"grad_norm": 1.0790719985961914,
"learning_rate": 2.37e-06,
"loss": 1.338,
"step": 80
},
{
"epoch": 0.34615384615384615,
"grad_norm": 1.1439405679702759,
"learning_rate": 2.67e-06,
"loss": 1.3231,
"step": 90
},
{
"epoch": 0.38461538461538464,
"grad_norm": 1.5780725479125977,
"learning_rate": 2.9700000000000004e-06,
"loss": 1.3026,
"step": 100
},
{
"epoch": 0.4230769230769231,
"grad_norm": 1.6573809385299683,
"learning_rate": 3.27e-06,
"loss": 1.2959,
"step": 110
},
{
"epoch": 0.46153846153846156,
"grad_norm": 1.5089571475982666,
"learning_rate": 3.57e-06,
"loss": 1.2855,
"step": 120
},
{
"epoch": 0.5,
"grad_norm": 1.3819092512130737,
"learning_rate": 3.87e-06,
"loss": 1.2571,
"step": 130
},
{
"epoch": 0.5384615384615384,
"grad_norm": 1.3043365478515625,
"learning_rate": 4.170000000000001e-06,
"loss": 1.2411,
"step": 140
},
{
"epoch": 0.5769230769230769,
"grad_norm": 1.0912373065948486,
"learning_rate": 4.4699999999999996e-06,
"loss": 1.224,
"step": 150
},
{
"epoch": 0.6153846153846154,
"grad_norm": 1.026869535446167,
"learning_rate": 4.77e-06,
"loss": 1.1946,
"step": 160
},
{
"epoch": 0.6538461538461539,
"grad_norm": 0.9299987554550171,
"learning_rate": 5.070000000000001e-06,
"loss": 1.1772,
"step": 170
},
{
"epoch": 0.6923076923076923,
"grad_norm": 0.9173968434333801,
"learning_rate": 5.37e-06,
"loss": 1.1611,
"step": 180
},
{
"epoch": 0.7307692307692307,
"grad_norm": 0.9373317360877991,
"learning_rate": 5.67e-06,
"loss": 1.1386,
"step": 190
},
{
"epoch": 0.7692307692307693,
"grad_norm": 0.8152101635932922,
"learning_rate": 5.9700000000000004e-06,
"loss": 1.1378,
"step": 200
},
{
"epoch": 0.8076923076923077,
"grad_norm": 1.0374643802642822,
"learning_rate": 6.27e-06,
"loss": 1.1314,
"step": 210
},
{
"epoch": 0.8461538461538461,
"grad_norm": 0.9438158869743347,
"learning_rate": 6.57e-06,
"loss": 1.1076,
"step": 220
},
{
"epoch": 0.8846153846153846,
"grad_norm": 0.9103168845176697,
"learning_rate": 6.87e-06,
"loss": 1.0955,
"step": 230
},
{
"epoch": 0.9230769230769231,
"grad_norm": 0.743010401725769,
"learning_rate": 7.17e-06,
"loss": 1.093,
"step": 240
},
{
"epoch": 0.9615384615384616,
"grad_norm": 0.8534924983978271,
"learning_rate": 7.4700000000000005e-06,
"loss": 1.0831,
"step": 250
},
{
"epoch": 1.0,
"grad_norm": 1.0246775150299072,
"learning_rate": 7.77e-06,
"loss": 1.0725,
"step": 260
},
{
"epoch": 1.0384615384615385,
"grad_norm": 0.9404919743537903,
"learning_rate": 8.07e-06,
"loss": 1.0711,
"step": 270
},
{
"epoch": 1.0769230769230769,
"grad_norm": 1.093787431716919,
"learning_rate": 8.370000000000001e-06,
"loss": 1.0709,
"step": 280
},
{
"epoch": 1.1153846153846154,
"grad_norm": 0.870293378829956,
"learning_rate": 8.67e-06,
"loss": 1.0667,
"step": 290
},
{
"epoch": 1.1538461538461537,
"grad_norm": 0.9972339272499084,
"learning_rate": 8.97e-06,
"loss": 1.0592,
"step": 300
},
{
"epoch": 1.1923076923076923,
"grad_norm": 1.1312875747680664,
"learning_rate": 9.27e-06,
"loss": 1.0592,
"step": 310
},
{
"epoch": 1.2307692307692308,
"grad_norm": 1.0009738206863403,
"learning_rate": 9.57e-06,
"loss": 1.0524,
"step": 320
},
{
"epoch": 1.2692307692307692,
"grad_norm": 0.9962753057479858,
"learning_rate": 9.87e-06,
"loss": 1.0491,
"step": 330
},
{
"epoch": 1.3076923076923077,
"grad_norm": 1.0982589721679688,
"learning_rate": 1.0170000000000001e-05,
"loss": 1.051,
"step": 340
},
{
"epoch": 1.3461538461538463,
"grad_norm": 1.1114718914031982,
"learning_rate": 1.047e-05,
"loss": 1.0461,
"step": 350
},
{
"epoch": 1.3846153846153846,
"grad_norm": 1.2868746519088745,
"learning_rate": 1.077e-05,
"loss": 1.0319,
"step": 360
},
{
"epoch": 1.4230769230769231,
"grad_norm": 1.2511143684387207,
"learning_rate": 1.107e-05,
"loss": 1.0278,
"step": 370
},
{
"epoch": 1.4615384615384617,
"grad_norm": 1.0229848623275757,
"learning_rate": 1.137e-05,
"loss": 1.0187,
"step": 380
},
{
"epoch": 1.5,
"grad_norm": 1.2445800304412842,
"learning_rate": 1.167e-05,
"loss": 1.0057,
"step": 390
},
{
"epoch": 1.5384615384615383,
"grad_norm": 1.2295691967010498,
"learning_rate": 1.197e-05,
"loss": 0.9782,
"step": 400
},
{
"epoch": 1.5769230769230769,
"grad_norm": 1.7420762777328491,
"learning_rate": 1.227e-05,
"loss": 0.9389,
"step": 410
},
{
"epoch": 1.6153846153846154,
"grad_norm": 1.8028358221054077,
"learning_rate": 1.257e-05,
"loss": 0.8751,
"step": 420
},
{
"epoch": 1.6538461538461537,
"grad_norm": 2.2324907779693604,
"learning_rate": 1.287e-05,
"loss": 0.828,
"step": 430
},
{
"epoch": 1.6923076923076923,
"grad_norm": 2.3086061477661133,
"learning_rate": 1.3170000000000001e-05,
"loss": 0.7754,
"step": 440
},
{
"epoch": 1.7307692307692308,
"grad_norm": 2.051551342010498,
"learning_rate": 1.3470000000000001e-05,
"loss": 0.7396,
"step": 450
},
{
"epoch": 1.7692307692307692,
"grad_norm": 2.074948787689209,
"learning_rate": 1.377e-05,
"loss": 0.7048,
"step": 460
},
{
"epoch": 1.8076923076923077,
"grad_norm": 2.608893871307373,
"learning_rate": 1.4069999999999999e-05,
"loss": 0.684,
"step": 470
},
{
"epoch": 1.8461538461538463,
"grad_norm": 2.885021209716797,
"learning_rate": 1.437e-05,
"loss": 0.6542,
"step": 480
},
{
"epoch": 1.8846153846153846,
"grad_norm": 2.5614078044891357,
"learning_rate": 1.467e-05,
"loss": 0.6306,
"step": 490
},
{
"epoch": 1.9230769230769231,
"grad_norm": 2.7305147647857666,
"learning_rate": 1.497e-05,
"loss": 0.6043,
"step": 500
},
{
"epoch": 1.9615384615384617,
"grad_norm": 3.067171812057495,
"learning_rate": 1.527e-05,
"loss": 0.5919,
"step": 510
},
{
"epoch": 2.0,
"grad_norm": 3.1146395206451416,
"learning_rate": 1.5570000000000002e-05,
"loss": 0.5696,
"step": 520
},
{
"epoch": 2.0384615384615383,
"grad_norm": 3.35973858833313,
"learning_rate": 1.5870000000000002e-05,
"loss": 0.5662,
"step": 530
},
{
"epoch": 2.076923076923077,
"grad_norm": 3.7923669815063477,
"learning_rate": 1.6170000000000003e-05,
"loss": 0.5478,
"step": 540
},
{
"epoch": 2.1153846153846154,
"grad_norm": 3.8711159229278564,
"learning_rate": 1.6470000000000003e-05,
"loss": 0.5375,
"step": 550
},
{
"epoch": 2.1538461538461537,
"grad_norm": 3.2727854251861572,
"learning_rate": 1.677e-05,
"loss": 0.522,
"step": 560
},
{
"epoch": 2.1923076923076925,
"grad_norm": 3.385040521621704,
"learning_rate": 1.7069999999999998e-05,
"loss": 0.5089,
"step": 570
},
{
"epoch": 2.230769230769231,
"grad_norm": 3.5132815837860107,
"learning_rate": 1.7369999999999998e-05,
"loss": 0.5031,
"step": 580
},
{
"epoch": 2.269230769230769,
"grad_norm": 3.465324878692627,
"learning_rate": 1.767e-05,
"loss": 0.5019,
"step": 590
},
{
"epoch": 2.3076923076923075,
"grad_norm": 3.7787463665008545,
"learning_rate": 1.797e-05,
"loss": 0.502,
"step": 600
},
{
"epoch": 2.3461538461538463,
"grad_norm": 3.8140358924865723,
"learning_rate": 1.827e-05,
"loss": 0.4996,
"step": 610
},
{
"epoch": 2.3846153846153846,
"grad_norm": 3.4496121406555176,
"learning_rate": 1.857e-05,
"loss": 0.4875,
"step": 620
},
{
"epoch": 2.423076923076923,
"grad_norm": 3.9799156188964844,
"learning_rate": 1.887e-05,
"loss": 0.4834,
"step": 630
},
{
"epoch": 2.4615384615384617,
"grad_norm": 3.7426834106445312,
"learning_rate": 1.917e-05,
"loss": 0.4802,
"step": 640
},
{
"epoch": 2.5,
"grad_norm": 3.552870988845825,
"learning_rate": 1.947e-05,
"loss": 0.4839,
"step": 650
},
{
"epoch": 2.5384615384615383,
"grad_norm": 3.0882790088653564,
"learning_rate": 1.9770000000000002e-05,
"loss": 0.4777,
"step": 660
},
{
"epoch": 2.5769230769230766,
"grad_norm": 3.536569595336914,
"learning_rate": 2.0070000000000003e-05,
"loss": 0.4892,
"step": 670
},
{
"epoch": 2.6153846153846154,
"grad_norm": 3.4024746417999268,
"learning_rate": 2.0370000000000003e-05,
"loss": 0.4761,
"step": 680
},
{
"epoch": 2.6538461538461537,
"grad_norm": 3.4993839263916016,
"learning_rate": 2.067e-05,
"loss": 0.4699,
"step": 690
},
{
"epoch": 2.6923076923076925,
"grad_norm": 4.353096961975098,
"learning_rate": 2.097e-05,
"loss": 0.4741,
"step": 700
},
{
"epoch": 2.730769230769231,
"grad_norm": 4.202399730682373,
"learning_rate": 2.1269999999999998e-05,
"loss": 0.4794,
"step": 710
},
{
"epoch": 2.769230769230769,
"grad_norm": 4.3636980056762695,
"learning_rate": 2.157e-05,
"loss": 0.4712,
"step": 720
},
{
"epoch": 2.8076923076923075,
"grad_norm": 4.043900489807129,
"learning_rate": 2.187e-05,
"loss": 0.4676,
"step": 730
},
{
"epoch": 2.8461538461538463,
"grad_norm": 3.271538496017456,
"learning_rate": 2.217e-05,
"loss": 0.459,
"step": 740
},
{
"epoch": 2.8846153846153846,
"grad_norm": 3.215296506881714,
"learning_rate": 2.247e-05,
"loss": 0.4648,
"step": 750
},
{
"epoch": 2.9230769230769234,
"grad_norm": 3.489201784133911,
"learning_rate": 2.277e-05,
"loss": 0.4556,
"step": 760
},
{
"epoch": 2.9615384615384617,
"grad_norm": 3.2616641521453857,
"learning_rate": 2.307e-05,
"loss": 0.4613,
"step": 770
},
{
"epoch": 3.0,
"grad_norm": 3.8984408378601074,
"learning_rate": 2.337e-05,
"loss": 0.4573,
"step": 780
},
{
"epoch": 3.0384615384615383,
"grad_norm": 4.524011135101318,
"learning_rate": 2.3670000000000002e-05,
"loss": 0.4587,
"step": 790
},
{
"epoch": 3.076923076923077,
"grad_norm": 3.7047295570373535,
"learning_rate": 2.3970000000000003e-05,
"loss": 0.4582,
"step": 800
},
{
"epoch": 3.1153846153846154,
"grad_norm": 4.147561073303223,
"learning_rate": 2.4270000000000003e-05,
"loss": 0.4652,
"step": 810
},
{
"epoch": 3.1538461538461537,
"grad_norm": 3.836184024810791,
"learning_rate": 2.457e-05,
"loss": 0.4549,
"step": 820
},
{
"epoch": 3.1923076923076925,
"grad_norm": 4.417998313903809,
"learning_rate": 2.487e-05,
"loss": 0.4531,
"step": 830
},
{
"epoch": 3.230769230769231,
"grad_norm": 4.859844207763672,
"learning_rate": 2.517e-05,
"loss": 0.4527,
"step": 840
},
{
"epoch": 3.269230769230769,
"grad_norm": 3.956090211868286,
"learning_rate": 2.547e-05,
"loss": 0.4513,
"step": 850
},
{
"epoch": 3.3076923076923075,
"grad_norm": 3.7535667419433594,
"learning_rate": 2.577e-05,
"loss": 0.4543,
"step": 860
},
{
"epoch": 3.3461538461538463,
"grad_norm": 3.8291614055633545,
"learning_rate": 2.607e-05,
"loss": 0.448,
"step": 870
},
{
"epoch": 3.3846153846153846,
"grad_norm": 3.6051249504089355,
"learning_rate": 2.637e-05,
"loss": 0.4425,
"step": 880
},
{
"epoch": 3.423076923076923,
"grad_norm": 4.240597248077393,
"learning_rate": 2.667e-05,
"loss": 0.4432,
"step": 890
},
{
"epoch": 3.4615384615384617,
"grad_norm": 4.126594543457031,
"learning_rate": 2.697e-05,
"loss": 0.4481,
"step": 900
},
{
"epoch": 3.5,
"grad_norm": 3.919288396835327,
"learning_rate": 2.727e-05,
"loss": 0.4392,
"step": 910
},
{
"epoch": 3.5384615384615383,
"grad_norm": 3.442884683609009,
"learning_rate": 2.7570000000000002e-05,
"loss": 0.4453,
"step": 920
},
{
"epoch": 3.5769230769230766,
"grad_norm": 4.0917067527771,
"learning_rate": 2.7870000000000003e-05,
"loss": 0.4407,
"step": 930
},
{
"epoch": 3.6153846153846154,
"grad_norm": 4.487368106842041,
"learning_rate": 2.817e-05,
"loss": 0.4435,
"step": 940
},
{
"epoch": 3.6538461538461537,
"grad_norm": 3.6351025104522705,
"learning_rate": 2.847e-05,
"loss": 0.4509,
"step": 950
},
{
"epoch": 3.6923076923076925,
"grad_norm": 4.146681308746338,
"learning_rate": 2.877e-05,
"loss": 0.4346,
"step": 960
},
{
"epoch": 3.730769230769231,
"grad_norm": 4.30366849899292,
"learning_rate": 2.907e-05,
"loss": 0.4398,
"step": 970
},
{
"epoch": 3.769230769230769,
"grad_norm": 2.9849531650543213,
"learning_rate": 2.9370000000000002e-05,
"loss": 0.4414,
"step": 980
},
{
"epoch": 3.8076923076923075,
"grad_norm": 3.1285574436187744,
"learning_rate": 2.967e-05,
"loss": 0.4431,
"step": 990
},
{
"epoch": 3.8461538461538463,
"grad_norm": 3.2634713649749756,
"learning_rate": 2.997e-05,
"loss": 0.443,
"step": 1000
},
{
"epoch": 3.8846153846153846,
"grad_norm": 3.845669746398926,
"learning_rate": 2.9999983391181253e-05,
"loss": 0.4364,
"step": 1010
},
{
"epoch": 3.9230769230769234,
"grad_norm": 3.47135853767395,
"learning_rate": 2.9999925978027876e-05,
"loss": 0.4414,
"step": 1020
},
{
"epoch": 3.9615384615384617,
"grad_norm": 3.6643917560577393,
"learning_rate": 2.9999827555649637e-05,
"loss": 0.4331,
"step": 1030
},
{
"epoch": 4.0,
"grad_norm": 3.6928398609161377,
"learning_rate": 2.999968812431563e-05,
"loss": 0.4358,
"step": 1040
},
{
"epoch": 4.038461538461538,
"grad_norm": 3.31592059135437,
"learning_rate": 2.999950768440706e-05,
"loss": 0.4304,
"step": 1050
},
{
"epoch": 4.076923076923077,
"grad_norm": 3.2825427055358887,
"learning_rate": 2.999928623641723e-05,
"loss": 0.4372,
"step": 1060
},
{
"epoch": 4.115384615384615,
"grad_norm": 3.60921311378479,
"learning_rate": 2.9999023780951575e-05,
"loss": 0.4321,
"step": 1070
},
{
"epoch": 4.153846153846154,
"grad_norm": 3.924734354019165,
"learning_rate": 2.999872031872764e-05,
"loss": 0.4291,
"step": 1080
},
{
"epoch": 4.1923076923076925,
"grad_norm": 3.746192693710327,
"learning_rate": 2.999837585057508e-05,
"loss": 0.4212,
"step": 1090
},
{
"epoch": 4.230769230769231,
"grad_norm": 3.685960531234741,
"learning_rate": 2.999799037743565e-05,
"loss": 0.4186,
"step": 1100
},
{
"epoch": 4.269230769230769,
"grad_norm": 3.234938621520996,
"learning_rate": 2.999756390036323e-05,
"loss": 0.4166,
"step": 1110
},
{
"epoch": 4.3076923076923075,
"grad_norm": 4.075984001159668,
"learning_rate": 2.9997096420523788e-05,
"loss": 0.4191,
"step": 1120
},
{
"epoch": 4.346153846153846,
"grad_norm": 3.920379161834717,
"learning_rate": 2.9996587939195395e-05,
"loss": 0.4224,
"step": 1130
},
{
"epoch": 4.384615384615385,
"grad_norm": 3.652268171310425,
"learning_rate": 2.999603845776822e-05,
"loss": 0.4188,
"step": 1140
},
{
"epoch": 4.423076923076923,
"grad_norm": 4.860605716705322,
"learning_rate": 2.999544797774452e-05,
"loss": 0.4217,
"step": 1150
},
{
"epoch": 4.461538461538462,
"grad_norm": 4.3559088706970215,
"learning_rate": 2.9994816500738648e-05,
"loss": 0.4161,
"step": 1160
},
{
"epoch": 4.5,
"grad_norm": 3.8840413093566895,
"learning_rate": 2.999414402847704e-05,
"loss": 0.4172,
"step": 1170
},
{
"epoch": 4.538461538461538,
"grad_norm": 3.8383638858795166,
"learning_rate": 2.999343056279821e-05,
"loss": 0.4133,
"step": 1180
},
{
"epoch": 4.576923076923077,
"grad_norm": 3.845787286758423,
"learning_rate": 2.9992676105652746e-05,
"loss": 0.4143,
"step": 1190
},
{
"epoch": 4.615384615384615,
"grad_norm": 3.575833320617676,
"learning_rate": 2.9991880659103298e-05,
"loss": 0.4152,
"step": 1200
},
{
"epoch": 4.653846153846154,
"grad_norm": 3.3321430683135986,
"learning_rate": 2.9991044225324593e-05,
"loss": 0.4121,
"step": 1210
},
{
"epoch": 4.6923076923076925,
"grad_norm": 3.8986430168151855,
"learning_rate": 2.9990166806603407e-05,
"loss": 0.4084,
"step": 1220
},
{
"epoch": 4.730769230769231,
"grad_norm": 3.6424148082733154,
"learning_rate": 2.9989248405338573e-05,
"loss": 0.414,
"step": 1230
},
{
"epoch": 4.769230769230769,
"grad_norm": 4.0105671882629395,
"learning_rate": 2.9988289024040962e-05,
"loss": 0.4047,
"step": 1240
},
{
"epoch": 4.8076923076923075,
"grad_norm": 3.252894401550293,
"learning_rate": 2.998728866533348e-05,
"loss": 0.4123,
"step": 1250
},
{
"epoch": 4.846153846153846,
"grad_norm": 3.5156307220458984,
"learning_rate": 2.9986247331951083e-05,
"loss": 0.4068,
"step": 1260
},
{
"epoch": 4.884615384615385,
"grad_norm": 2.7843873500823975,
"learning_rate": 2.998516502674072e-05,
"loss": 0.4116,
"step": 1270
},
{
"epoch": 4.923076923076923,
"grad_norm": 3.0897672176361084,
"learning_rate": 2.9984041752661386e-05,
"loss": 0.4074,
"step": 1280
},
{
"epoch": 4.961538461538462,
"grad_norm": 3.780107021331787,
"learning_rate": 2.9982877512784067e-05,
"loss": 0.4063,
"step": 1290
},
{
"epoch": 5.0,
"grad_norm": 4.262648582458496,
"learning_rate": 2.998167231029174e-05,
"loss": 0.4051,
"step": 1300
},
{
"epoch": 5.038461538461538,
"grad_norm": 3.8252484798431396,
"learning_rate": 2.99804261484794e-05,
"loss": 0.4053,
"step": 1310
},
{
"epoch": 5.076923076923077,
"grad_norm": 3.634221076965332,
"learning_rate": 2.997913903075399e-05,
"loss": 0.4077,
"step": 1320
},
{
"epoch": 5.115384615384615,
"grad_norm": 2.8847663402557373,
"learning_rate": 2.997781096063445e-05,
"loss": 0.4022,
"step": 1330
},
{
"epoch": 5.153846153846154,
"grad_norm": 3.5783627033233643,
"learning_rate": 2.9976441941751663e-05,
"loss": 0.401,
"step": 1340
},
{
"epoch": 5.1923076923076925,
"grad_norm": 3.9379782676696777,
"learning_rate": 2.997503197784849e-05,
"loss": 0.3991,
"step": 1350
},
{
"epoch": 5.230769230769231,
"grad_norm": 3.22617506980896,
"learning_rate": 2.9973581072779702e-05,
"loss": 0.399,
"step": 1360
},
{
"epoch": 5.269230769230769,
"grad_norm": 3.1377646923065186,
"learning_rate": 2.9972089230512035e-05,
"loss": 0.3961,
"step": 1370
},
{
"epoch": 5.3076923076923075,
"grad_norm": 3.380993127822876,
"learning_rate": 2.997055645512411e-05,
"loss": 0.402,
"step": 1380
},
{
"epoch": 5.346153846153846,
"grad_norm": 3.271967649459839,
"learning_rate": 2.9968982750806492e-05,
"loss": 0.3929,
"step": 1390
},
{
"epoch": 5.384615384615385,
"grad_norm": 3.3165857791900635,
"learning_rate": 2.9967368121861623e-05,
"loss": 0.3996,
"step": 1400
},
{
"epoch": 5.423076923076923,
"grad_norm": 3.385725975036621,
"learning_rate": 2.9965712572703834e-05,
"loss": 0.3936,
"step": 1410
},
{
"epoch": 5.461538461538462,
"grad_norm": 3.0297584533691406,
"learning_rate": 2.996401610785934e-05,
"loss": 0.3991,
"step": 1420
},
{
"epoch": 5.5,
"grad_norm": 3.3578076362609863,
"learning_rate": 2.99622787319662e-05,
"loss": 0.3939,
"step": 1430
},
{
"epoch": 5.538461538461538,
"grad_norm": 3.9298784732818604,
"learning_rate": 2.9960500449774338e-05,
"loss": 0.3997,
"step": 1440
},
{
"epoch": 5.576923076923077,
"grad_norm": 3.5111310482025146,
"learning_rate": 2.9958681266145517e-05,
"loss": 0.3994,
"step": 1450
},
{
"epoch": 5.615384615384615,
"grad_norm": 3.912317991256714,
"learning_rate": 2.995682118605331e-05,
"loss": 0.392,
"step": 1460
},
{
"epoch": 5.653846153846154,
"grad_norm": 3.1432626247406006,
"learning_rate": 2.9954920214583107e-05,
"loss": 0.3951,
"step": 1470
},
{
"epoch": 5.6923076923076925,
"grad_norm": 3.084312915802002,
"learning_rate": 2.9952978356932084e-05,
"loss": 0.3941,
"step": 1480
},
{
"epoch": 5.730769230769231,
"grad_norm": 3.1436219215393066,
"learning_rate": 2.9950995618409215e-05,
"loss": 0.3924,
"step": 1490
},
{
"epoch": 5.769230769230769,
"grad_norm": 3.0541083812713623,
"learning_rate": 2.9948972004435228e-05,
"loss": 0.4004,
"step": 1500
},
{
"epoch": 5.8076923076923075,
"grad_norm": 3.0577759742736816,
"learning_rate": 2.9946907520542602e-05,
"loss": 0.3949,
"step": 1510
},
{
"epoch": 5.846153846153846,
"grad_norm": 3.1131491661071777,
"learning_rate": 2.9944802172375566e-05,
"loss": 0.3936,
"step": 1520
},
{
"epoch": 5.884615384615385,
"grad_norm": 2.9365406036376953,
"learning_rate": 2.9942655965690053e-05,
"loss": 0.3904,
"step": 1530
},
{
"epoch": 5.923076923076923,
"grad_norm": 3.253673791885376,
"learning_rate": 2.9940468906353712e-05,
"loss": 0.3916,
"step": 1540
},
{
"epoch": 5.961538461538462,
"grad_norm": 3.9712073802948,
"learning_rate": 2.9938241000345887e-05,
"loss": 0.3864,
"step": 1550
},
{
"epoch": 6.0,
"grad_norm": 3.6548357009887695,
"learning_rate": 2.993597225375758e-05,
"loss": 0.3842,
"step": 1560
},
{
"epoch": 6.038461538461538,
"grad_norm": 4.270135402679443,
"learning_rate": 2.993366267279146e-05,
"loss": 0.385,
"step": 1570
},
{
"epoch": 6.076923076923077,
"grad_norm": 4.103717803955078,
"learning_rate": 2.993131226376183e-05,
"loss": 0.382,
"step": 1580
},
{
"epoch": 6.115384615384615,
"grad_norm": 3.5214860439300537,
"learning_rate": 2.9928921033094626e-05,
"loss": 0.3818,
"step": 1590
},
{
"epoch": 6.153846153846154,
"grad_norm": 3.6836981773376465,
"learning_rate": 2.9926488987327376e-05,
"loss": 0.377,
"step": 1600
},
{
"epoch": 6.1923076923076925,
"grad_norm": 3.7300078868865967,
"learning_rate": 2.99240161331092e-05,
"loss": 0.3705,
"step": 1610
},
{
"epoch": 6.230769230769231,
"grad_norm": 4.4989118576049805,
"learning_rate": 2.992150247720079e-05,
"loss": 0.3683,
"step": 1620
},
{
"epoch": 6.269230769230769,
"grad_norm": 3.8246257305145264,
"learning_rate": 2.991894802647438e-05,
"loss": 0.3666,
"step": 1630
},
{
"epoch": 6.3076923076923075,
"grad_norm": 3.7448792457580566,
"learning_rate": 2.9916352787913746e-05,
"loss": 0.3657,
"step": 1640
},
{
"epoch": 6.346153846153846,
"grad_norm": 5.380867004394531,
"learning_rate": 2.991371676861417e-05,
"loss": 0.3704,
"step": 1650
},
{
"epoch": 6.384615384615385,
"grad_norm": 5.416983127593994,
"learning_rate": 2.991103997578243e-05,
"loss": 0.36,
"step": 1660
},
{
"epoch": 6.423076923076923,
"grad_norm": 4.012523174285889,
"learning_rate": 2.9908322416736767e-05,
"loss": 0.3645,
"step": 1670
},
{
"epoch": 6.461538461538462,
"grad_norm": 3.2570512294769287,
"learning_rate": 2.990556409890689e-05,
"loss": 0.3583,
"step": 1680
},
{
"epoch": 6.5,
"grad_norm": 3.389817714691162,
"learning_rate": 2.990276502983394e-05,
"loss": 0.3583,
"step": 1690
},
{
"epoch": 6.538461538461538,
"grad_norm": 4.151524066925049,
"learning_rate": 2.9899925217170455e-05,
"loss": 0.3525,
"step": 1700
},
{
"epoch": 6.576923076923077,
"grad_norm": 4.0402140617370605,
"learning_rate": 2.989704466868038e-05,
"loss": 0.3543,
"step": 1710
},
{
"epoch": 6.615384615384615,
"grad_norm": 3.9035301208496094,
"learning_rate": 2.9894123392239018e-05,
"loss": 0.3524,
"step": 1720
},
{
"epoch": 6.653846153846154,
"grad_norm": 4.771100044250488,
"learning_rate": 2.9891161395833037e-05,
"loss": 0.346,
"step": 1730
},
{
"epoch": 6.6923076923076925,
"grad_norm": 4.07216739654541,
"learning_rate": 2.988815868756042e-05,
"loss": 0.3412,
"step": 1740
},
{
"epoch": 6.730769230769231,
"grad_norm": 3.648852586746216,
"learning_rate": 2.9885115275630447e-05,
"loss": 0.3364,
"step": 1750
},
{
"epoch": 6.769230769230769,
"grad_norm": 4.100618362426758,
"learning_rate": 2.9882031168363703e-05,
"loss": 0.3314,
"step": 1760
},
{
"epoch": 6.8076923076923075,
"grad_norm": 4.023171901702881,
"learning_rate": 2.9878906374192013e-05,
"loss": 0.3298,
"step": 1770
},
{
"epoch": 6.846153846153846,
"grad_norm": 3.8320722579956055,
"learning_rate": 2.9875740901658446e-05,
"loss": 0.3208,
"step": 1780
},
{
"epoch": 6.884615384615385,
"grad_norm": 3.7699222564697266,
"learning_rate": 2.987253475941728e-05,
"loss": 0.3203,
"step": 1790
},
{
"epoch": 6.923076923076923,
"grad_norm": 4.349817752838135,
"learning_rate": 2.9869287956233986e-05,
"loss": 0.3092,
"step": 1800
},
{
"epoch": 6.961538461538462,
"grad_norm": 3.1222894191741943,
"learning_rate": 2.9866000500985207e-05,
"loss": 0.3076,
"step": 1810
},
{
"epoch": 7.0,
"grad_norm": 5.115797519683838,
"learning_rate": 2.9862672402658712e-05,
"loss": 0.3011,
"step": 1820
},
{
"epoch": 7.038461538461538,
"grad_norm": 4.312737941741943,
"learning_rate": 2.98593036703534e-05,
"loss": 0.3016,
"step": 1830
},
{
"epoch": 7.076923076923077,
"grad_norm": 5.156615734100342,
"learning_rate": 2.9855894313279256e-05,
"loss": 0.2959,
"step": 1840
},
{
"epoch": 7.115384615384615,
"grad_norm": 4.620377540588379,
"learning_rate": 2.9852444340757326e-05,
"loss": 0.2916,
"step": 1850
},
{
"epoch": 7.153846153846154,
"grad_norm": 4.231908321380615,
"learning_rate": 2.9848953762219707e-05,
"loss": 0.2947,
"step": 1860
},
{
"epoch": 7.1923076923076925,
"grad_norm": 4.510662078857422,
"learning_rate": 2.984542258720951e-05,
"loss": 0.2846,
"step": 1870
},
{
"epoch": 7.230769230769231,
"grad_norm": 5.093549728393555,
"learning_rate": 2.984185082538083e-05,
"loss": 0.2707,
"step": 1880
},
{
"epoch": 7.269230769230769,
"grad_norm": 5.482805252075195,
"learning_rate": 2.983823848649873e-05,
"loss": 0.2624,
"step": 1890
},
{
"epoch": 7.3076923076923075,
"grad_norm": 3.6456329822540283,
"learning_rate": 2.9834585580439203e-05,
"loss": 0.2769,
"step": 1900
},
{
"epoch": 7.346153846153846,
"grad_norm": 4.198169708251953,
"learning_rate": 2.9830892117189157e-05,
"loss": 0.2659,
"step": 1910
},
{
"epoch": 7.384615384615385,
"grad_norm": 4.8469929695129395,
"learning_rate": 2.982715810684638e-05,
"loss": 0.2612,
"step": 1920
},
{
"epoch": 7.423076923076923,
"grad_norm": 5.01643705368042,
"learning_rate": 2.982338355961951e-05,
"loss": 0.2562,
"step": 1930
},
{
"epoch": 7.461538461538462,
"grad_norm": 5.112571716308594,
"learning_rate": 2.981956848582802e-05,
"loss": 0.2523,
"step": 1940
},
{
"epoch": 7.5,
"grad_norm": 3.7991206645965576,
"learning_rate": 2.981571289590217e-05,
"loss": 0.2475,
"step": 1950
},
{
"epoch": 7.538461538461538,
"grad_norm": 4.921756267547607,
"learning_rate": 2.9811816800383003e-05,
"loss": 0.237,
"step": 1960
},
{
"epoch": 7.576923076923077,
"grad_norm": 4.657440662384033,
"learning_rate": 2.9807880209922288e-05,
"loss": 0.2464,
"step": 1970
},
{
"epoch": 7.615384615384615,
"grad_norm": 4.40252685546875,
"learning_rate": 2.9803903135282518e-05,
"loss": 0.245,
"step": 1980
},
{
"epoch": 7.653846153846154,
"grad_norm": 4.782217025756836,
"learning_rate": 2.9799885587336862e-05,
"loss": 0.2325,
"step": 1990
},
{
"epoch": 7.6923076923076925,
"grad_norm": 5.208028316497803,
"learning_rate": 2.9795827577069145e-05,
"loss": 0.2242,
"step": 2000
},
{
"epoch": 7.730769230769231,
"grad_norm": 4.525110244750977,
"learning_rate": 2.9791729115573808e-05,
"loss": 0.2211,
"step": 2010
},
{
"epoch": 7.769230769230769,
"grad_norm": 5.336142063140869,
"learning_rate": 2.9787590214055887e-05,
"loss": 0.2227,
"step": 2020
},
{
"epoch": 7.8076923076923075,
"grad_norm": 5.140171527862549,
"learning_rate": 2.9783410883830983e-05,
"loss": 0.2235,
"step": 2030
},
{
"epoch": 7.846153846153846,
"grad_norm": 5.8251142501831055,
"learning_rate": 2.9779191136325233e-05,
"loss": 0.2256,
"step": 2040
},
{
"epoch": 7.884615384615385,
"grad_norm": 4.06174373626709,
"learning_rate": 2.977493098307525e-05,
"loss": 0.2101,
"step": 2050
},
{
"epoch": 7.923076923076923,
"grad_norm": 5.9502739906311035,
"learning_rate": 2.9770630435728142e-05,
"loss": 0.1998,
"step": 2060
},
{
"epoch": 7.961538461538462,
"grad_norm": 5.872840881347656,
"learning_rate": 2.976628950604144e-05,
"loss": 0.2011,
"step": 2070
},
{
"epoch": 8.0,
"grad_norm": 5.389374256134033,
"learning_rate": 2.9761908205883073e-05,
"loss": 0.1917,
"step": 2080
},
{
"epoch": 8.038461538461538,
"grad_norm": 5.9729905128479,
"learning_rate": 2.9757486547231357e-05,
"loss": 0.1891,
"step": 2090
},
{
"epoch": 8.076923076923077,
"grad_norm": 4.516313076019287,
"learning_rate": 2.9753024542174934e-05,
"loss": 0.1779,
"step": 2100
},
{
"epoch": 8.115384615384615,
"grad_norm": 5.897557735443115,
"learning_rate": 2.9748522202912755e-05,
"loss": 0.1768,
"step": 2110
},
{
"epoch": 8.153846153846153,
"grad_norm": 5.554835796356201,
"learning_rate": 2.974397954175404e-05,
"loss": 0.1748,
"step": 2120
},
{
"epoch": 8.192307692307692,
"grad_norm": 4.796779632568359,
"learning_rate": 2.973939657111826e-05,
"loss": 0.1797,
"step": 2130
},
{
"epoch": 8.23076923076923,
"grad_norm": 4.482677936553955,
"learning_rate": 2.9734773303535078e-05,
"loss": 0.1756,
"step": 2140
},
{
"epoch": 8.26923076923077,
"grad_norm": 5.248786449432373,
"learning_rate": 2.9730109751644325e-05,
"loss": 0.1696,
"step": 2150
},
{
"epoch": 8.307692307692308,
"grad_norm": 6.799733638763428,
"learning_rate": 2.9725405928195985e-05,
"loss": 0.1645,
"step": 2160
},
{
"epoch": 8.346153846153847,
"grad_norm": 7.030920028686523,
"learning_rate": 2.9720661846050123e-05,
"loss": 0.1504,
"step": 2170
},
{
"epoch": 8.384615384615385,
"grad_norm": 4.81633996963501,
"learning_rate": 2.971587751817688e-05,
"loss": 0.1605,
"step": 2180
},
{
"epoch": 8.423076923076923,
"grad_norm": 4.238231658935547,
"learning_rate": 2.9711052957656425e-05,
"loss": 0.1455,
"step": 2190
},
{
"epoch": 8.461538461538462,
"grad_norm": 4.765488147735596,
"learning_rate": 2.9706188177678924e-05,
"loss": 0.141,
"step": 2200
},
{
"epoch": 8.5,
"grad_norm": 5.133832931518555,
"learning_rate": 2.97012831915445e-05,
"loss": 0.1367,
"step": 2210
},
{
"epoch": 8.538461538461538,
"grad_norm": 6.479177951812744,
"learning_rate": 2.96963380126632e-05,
"loss": 0.1356,
"step": 2220
},
{
"epoch": 8.576923076923077,
"grad_norm": 5.76138162612915,
"learning_rate": 2.9691352654554953e-05,
"loss": 0.1367,
"step": 2230
},
{
"epoch": 8.615384615384615,
"grad_norm": 4.94566011428833,
"learning_rate": 2.9686327130849536e-05,
"loss": 0.127,
"step": 2240
},
{
"epoch": 8.653846153846153,
"grad_norm": 4.86747407913208,
"learning_rate": 2.9681261455286538e-05,
"loss": 0.134,
"step": 2250
},
{
"epoch": 8.692307692307692,
"grad_norm": 4.12553071975708,
"learning_rate": 2.9676155641715318e-05,
"loss": 0.1293,
"step": 2260
},
{
"epoch": 8.73076923076923,
"grad_norm": 5.830256462097168,
"learning_rate": 2.9671009704094988e-05,
"loss": 0.1184,
"step": 2270
},
{
"epoch": 8.76923076923077,
"grad_norm": 4.664676666259766,
"learning_rate": 2.9665823656494335e-05,
"loss": 0.1248,
"step": 2280
},
{
"epoch": 8.807692307692308,
"grad_norm": 3.934089183807373,
"learning_rate": 2.9660597513091824e-05,
"loss": 0.1141,
"step": 2290
},
{
"epoch": 8.846153846153847,
"grad_norm": 3.1599950790405273,
"learning_rate": 2.965533128817552e-05,
"loss": 0.1055,
"step": 2300
},
{
"epoch": 8.884615384615385,
"grad_norm": 4.693429470062256,
"learning_rate": 2.9650024996143084e-05,
"loss": 0.1189,
"step": 2310
},
{
"epoch": 8.923076923076923,
"grad_norm": 4.345015048980713,
"learning_rate": 2.964467865150172e-05,
"loss": 0.1011,
"step": 2320
},
{
"epoch": 8.961538461538462,
"grad_norm": 4.986255645751953,
"learning_rate": 2.9639292268868133e-05,
"loss": 0.1146,
"step": 2330
},
{
"epoch": 9.0,
"grad_norm": 4.554662227630615,
"learning_rate": 2.9633865862968478e-05,
"loss": 0.1019,
"step": 2340
},
{
"epoch": 9.038461538461538,
"grad_norm": 4.482904434204102,
"learning_rate": 2.9628399448638352e-05,
"loss": 0.1091,
"step": 2350
},
{
"epoch": 9.076923076923077,
"grad_norm": 5.466559410095215,
"learning_rate": 2.9622893040822714e-05,
"loss": 0.0983,
"step": 2360
},
{
"epoch": 9.115384615384615,
"grad_norm": 5.750704765319824,
"learning_rate": 2.9617346654575875e-05,
"loss": 0.0986,
"step": 2370
},
{
"epoch": 9.153846153846153,
"grad_norm": 4.733994483947754,
"learning_rate": 2.9611760305061447e-05,
"loss": 0.0926,
"step": 2380
},
{
"epoch": 9.192307692307692,
"grad_norm": 4.952576637268066,
"learning_rate": 2.9606134007552292e-05,
"loss": 0.1036,
"step": 2390
},
{
"epoch": 9.23076923076923,
"grad_norm": 5.682345390319824,
"learning_rate": 2.9600467777430497e-05,
"loss": 0.0992,
"step": 2400
},
{
"epoch": 9.26923076923077,
"grad_norm": 4.997755527496338,
"learning_rate": 2.9594761630187312e-05,
"loss": 0.0886,
"step": 2410
},
{
"epoch": 9.307692307692308,
"grad_norm": 4.387738227844238,
"learning_rate": 2.9589015581423132e-05,
"loss": 0.0907,
"step": 2420
},
{
"epoch": 9.346153846153847,
"grad_norm": 4.436047077178955,
"learning_rate": 2.958322964684743e-05,
"loss": 0.0875,
"step": 2430
},
{
"epoch": 9.384615384615385,
"grad_norm": 4.745996952056885,
"learning_rate": 2.9577403842278735e-05,
"loss": 0.0949,
"step": 2440
},
{
"epoch": 9.423076923076923,
"grad_norm": 5.043275356292725,
"learning_rate": 2.957153818364457e-05,
"loss": 0.0904,
"step": 2450
},
{
"epoch": 9.461538461538462,
"grad_norm": 4.609716415405273,
"learning_rate": 2.9565632686981428e-05,
"loss": 0.0989,
"step": 2460
},
{
"epoch": 9.5,
"grad_norm": 4.066394329071045,
"learning_rate": 2.9559687368434702e-05,
"loss": 0.0872,
"step": 2470
},
{
"epoch": 9.538461538461538,
"grad_norm": 4.978208541870117,
"learning_rate": 2.9553702244258674e-05,
"loss": 0.0874,
"step": 2480
},
{
"epoch": 9.576923076923077,
"grad_norm": 4.116719722747803,
"learning_rate": 2.954767733081644e-05,
"loss": 0.0802,
"step": 2490
},
{
"epoch": 9.615384615384615,
"grad_norm": 4.07297420501709,
"learning_rate": 2.9541612644579887e-05,
"loss": 0.0899,
"step": 2500
},
{
"epoch": 9.653846153846153,
"grad_norm": 3.5808463096618652,
"learning_rate": 2.9535508202129634e-05,
"loss": 0.0858,
"step": 2510
},
{
"epoch": 9.692307692307692,
"grad_norm": 5.122105598449707,
"learning_rate": 2.9529364020154994e-05,
"loss": 0.08,
"step": 2520
},
{
"epoch": 9.73076923076923,
"grad_norm": 3.463164806365967,
"learning_rate": 2.9523180115453922e-05,
"loss": 0.0788,
"step": 2530
},
{
"epoch": 9.76923076923077,
"grad_norm": 4.045137405395508,
"learning_rate": 2.9516956504932984e-05,
"loss": 0.0883,
"step": 2540
},
{
"epoch": 9.807692307692308,
"grad_norm": 3.9384825229644775,
"learning_rate": 2.9510693205607286e-05,
"loss": 0.0952,
"step": 2550
},
{
"epoch": 9.846153846153847,
"grad_norm": 3.5259459018707275,
"learning_rate": 2.9504390234600456e-05,
"loss": 0.0862,
"step": 2560
},
{
"epoch": 9.884615384615385,
"grad_norm": 3.8916735649108887,
"learning_rate": 2.9498047609144577e-05,
"loss": 0.082,
"step": 2570
},
{
"epoch": 9.923076923076923,
"grad_norm": 4.234613418579102,
"learning_rate": 2.9491665346580134e-05,
"loss": 0.0844,
"step": 2580
},
{
"epoch": 9.961538461538462,
"grad_norm": 3.3096225261688232,
"learning_rate": 2.9485243464356e-05,
"loss": 0.0835,
"step": 2590
},
{
"epoch": 10.0,
"grad_norm": 3.675506353378296,
"learning_rate": 2.9478781980029352e-05,
"loss": 0.0798,
"step": 2600
},
{
"epoch": 10.038461538461538,
"grad_norm": 3.9946699142456055,
"learning_rate": 2.9472280911265642e-05,
"loss": 0.0764,
"step": 2610
},
{
"epoch": 10.076923076923077,
"grad_norm": 3.801903247833252,
"learning_rate": 2.9465740275838543e-05,
"loss": 0.0794,
"step": 2620
},
{
"epoch": 10.115384615384615,
"grad_norm": 4.382718086242676,
"learning_rate": 2.94591600916299e-05,
"loss": 0.0768,
"step": 2630
},
{
"epoch": 10.153846153846153,
"grad_norm": 4.071086406707764,
"learning_rate": 2.9452540376629692e-05,
"loss": 0.0722,
"step": 2640
},
{
"epoch": 10.192307692307692,
"grad_norm": 3.6516573429107666,
"learning_rate": 2.944588114893596e-05,
"loss": 0.0833,
"step": 2650
},
{
"epoch": 10.23076923076923,
"grad_norm": 3.51243257522583,
"learning_rate": 2.9439182426754784e-05,
"loss": 0.0891,
"step": 2660
},
{
"epoch": 10.26923076923077,
"grad_norm": 4.3780741691589355,
"learning_rate": 2.9432444228400208e-05,
"loss": 0.0781,
"step": 2670
},
{
"epoch": 10.307692307692308,
"grad_norm": 3.2519776821136475,
"learning_rate": 2.9425666572294218e-05,
"loss": 0.0808,
"step": 2680
},
{
"epoch": 10.346153846153847,
"grad_norm": 4.084770679473877,
"learning_rate": 2.941884947696666e-05,
"loss": 0.0759,
"step": 2690
},
{
"epoch": 10.384615384615385,
"grad_norm": 3.3953628540039062,
"learning_rate": 2.9411992961055214e-05,
"loss": 0.084,
"step": 2700
},
{
"epoch": 10.423076923076923,
"grad_norm": 3.6471056938171387,
"learning_rate": 2.9405097043305334e-05,
"loss": 0.0793,
"step": 2710
},
{
"epoch": 10.461538461538462,
"grad_norm": 3.899949789047241,
"learning_rate": 2.9398161742570196e-05,
"loss": 0.0774,
"step": 2720
},
{
"epoch": 10.5,
"grad_norm": 3.219599723815918,
"learning_rate": 2.9391187077810644e-05,
"loss": 0.081,
"step": 2730
},
{
"epoch": 10.538461538461538,
"grad_norm": 3.684591293334961,
"learning_rate": 2.9384173068095145e-05,
"loss": 0.0774,
"step": 2740
},
{
"epoch": 10.576923076923077,
"grad_norm": 3.664158582687378,
"learning_rate": 2.937711973259974e-05,
"loss": 0.0687,
"step": 2750
},
{
"epoch": 10.615384615384615,
"grad_norm": 3.140185832977295,
"learning_rate": 2.9370027090607974e-05,
"loss": 0.0741,
"step": 2760
},
{
"epoch": 10.653846153846153,
"grad_norm": 3.908612012863159,
"learning_rate": 2.936289516151086e-05,
"loss": 0.0716,
"step": 2770
},
{
"epoch": 10.692307692307692,
"grad_norm": 3.865504503250122,
"learning_rate": 2.935572396480682e-05,
"loss": 0.0745,
"step": 2780
},
{
"epoch": 10.73076923076923,
"grad_norm": 3.879666566848755,
"learning_rate": 2.9348513520101636e-05,
"loss": 0.0694,
"step": 2790
},
{
"epoch": 10.76923076923077,
"grad_norm": 3.2028911113739014,
"learning_rate": 2.9341263847108383e-05,
"loss": 0.0739,
"step": 2800
},
{
"epoch": 10.807692307692308,
"grad_norm": 3.3011929988861084,
"learning_rate": 2.933397496564739e-05,
"loss": 0.0743,
"step": 2810
},
{
"epoch": 10.846153846153847,
"grad_norm": 3.6709179878234863,
"learning_rate": 2.9326646895646178e-05,
"loss": 0.0675,
"step": 2820
},
{
"epoch": 10.884615384615385,
"grad_norm": 2.8137264251708984,
"learning_rate": 2.931927965713942e-05,
"loss": 0.0729,
"step": 2830
},
{
"epoch": 10.923076923076923,
"grad_norm": 4.2008819580078125,
"learning_rate": 2.931187327026886e-05,
"loss": 0.0702,
"step": 2840
},
{
"epoch": 10.961538461538462,
"grad_norm": 3.371169090270996,
"learning_rate": 2.9304427755283278e-05,
"loss": 0.0728,
"step": 2850
},
{
"epoch": 11.0,
"grad_norm": 3.6277854442596436,
"learning_rate": 2.9296943132538425e-05,
"loss": 0.0681,
"step": 2860
},
{
"epoch": 11.038461538461538,
"grad_norm": 3.383019208908081,
"learning_rate": 2.928941942249697e-05,
"loss": 0.0692,
"step": 2870
},
{
"epoch": 11.076923076923077,
"grad_norm": 4.068559169769287,
"learning_rate": 2.928185664572846e-05,
"loss": 0.07,
"step": 2880
},
{
"epoch": 11.115384615384615,
"grad_norm": 3.2356979846954346,
"learning_rate": 2.927425482290923e-05,
"loss": 0.072,
"step": 2890
},
{
"epoch": 11.153846153846153,
"grad_norm": 3.1939711570739746,
"learning_rate": 2.926661397482238e-05,
"loss": 0.0739,
"step": 2900
},
{
"epoch": 11.192307692307692,
"grad_norm": 3.55348539352417,
"learning_rate": 2.9258934122357685e-05,
"loss": 0.091,
"step": 2910
},
{
"epoch": 11.23076923076923,
"grad_norm": 3.496305227279663,
"learning_rate": 2.9251215286511574e-05,
"loss": 0.0699,
"step": 2920
},
{
"epoch": 11.26923076923077,
"grad_norm": 3.339102268218994,
"learning_rate": 2.924345748838706e-05,
"loss": 0.0688,
"step": 2930
},
{
"epoch": 11.307692307692308,
"grad_norm": 3.273207426071167,
"learning_rate": 2.923566074919365e-05,
"loss": 0.0687,
"step": 2940
},
{
"epoch": 11.346153846153847,
"grad_norm": 3.1789557933807373,
"learning_rate": 2.9227825090247346e-05,
"loss": 0.0659,
"step": 2950
},
{
"epoch": 11.384615384615385,
"grad_norm": 3.100209951400757,
"learning_rate": 2.9219950532970526e-05,
"loss": 0.0783,
"step": 2960
},
{
"epoch": 11.423076923076923,
"grad_norm": 3.196394920349121,
"learning_rate": 2.921203709889194e-05,
"loss": 0.0751,
"step": 2970
},
{
"epoch": 11.461538461538462,
"grad_norm": 3.14573073387146,
"learning_rate": 2.9204084809646607e-05,
"loss": 0.0663,
"step": 2980
},
{
"epoch": 11.5,
"grad_norm": 3.1672585010528564,
"learning_rate": 2.9196093686975793e-05,
"loss": 0.0644,
"step": 2990
},
{
"epoch": 11.538461538461538,
"grad_norm": 3.119452953338623,
"learning_rate": 2.918806375272691e-05,
"loss": 0.0752,
"step": 3000
},
{
"epoch": 11.576923076923077,
"grad_norm": 3.139831781387329,
"learning_rate": 2.9179995028853498e-05,
"loss": 0.0683,
"step": 3010
},
{
"epoch": 11.615384615384615,
"grad_norm": 3.1740639209747314,
"learning_rate": 2.917188753741514e-05,
"loss": 0.0772,
"step": 3020
},
{
"epoch": 11.653846153846153,
"grad_norm": 3.1087207794189453,
"learning_rate": 2.916374130057741e-05,
"loss": 0.0678,
"step": 3030
},
{
"epoch": 11.692307692307692,
"grad_norm": 3.149672269821167,
"learning_rate": 2.91555563406118e-05,
"loss": 0.0621,
"step": 3040
},
{
"epoch": 11.73076923076923,
"grad_norm": 3.219829559326172,
"learning_rate": 2.9147332679895683e-05,
"loss": 0.0717,
"step": 3050
},
{
"epoch": 11.76923076923077,
"grad_norm": 3.3537230491638184,
"learning_rate": 2.9139070340912236e-05,
"loss": 0.0705,
"step": 3060
},
{
"epoch": 11.807692307692308,
"grad_norm": 2.787032127380371,
"learning_rate": 2.9130769346250376e-05,
"loss": 0.0776,
"step": 3070
},
{
"epoch": 11.846153846153847,
"grad_norm": 3.3528003692626953,
"learning_rate": 2.9122429718604704e-05,
"loss": 0.07,
"step": 3080
},
{
"epoch": 11.884615384615385,
"grad_norm": 3.332606792449951,
"learning_rate": 2.911405148077545e-05,
"loss": 0.0687,
"step": 3090
},
{
"epoch": 11.923076923076923,
"grad_norm": 2.7732484340667725,
"learning_rate": 2.9105634655668385e-05,
"loss": 0.0714,
"step": 3100
},
{
"epoch": 11.961538461538462,
"grad_norm": 4.068169593811035,
"learning_rate": 2.9097179266294794e-05,
"loss": 0.0766,
"step": 3110
},
{
"epoch": 12.0,
"grad_norm": 3.828740358352661,
"learning_rate": 2.9088685335771396e-05,
"loss": 0.0625,
"step": 3120
},
{
"epoch": 12.038461538461538,
"grad_norm": 3.0695571899414062,
"learning_rate": 2.9080152887320255e-05,
"loss": 0.0716,
"step": 3130
},
{
"epoch": 12.076923076923077,
"grad_norm": 3.0535635948181152,
"learning_rate": 2.9071581944268778e-05,
"loss": 0.0696,
"step": 3140
},
{
"epoch": 12.115384615384615,
"grad_norm": 2.860771894454956,
"learning_rate": 2.906297253004958e-05,
"loss": 0.0691,
"step": 3150
},
{
"epoch": 12.153846153846153,
"grad_norm": 2.567913293838501,
"learning_rate": 2.9054324668200483e-05,
"loss": 0.0666,
"step": 3160
},
{
"epoch": 12.192307692307692,
"grad_norm": 3.2005672454833984,
"learning_rate": 2.9045638382364404e-05,
"loss": 0.0699,
"step": 3170
},
{
"epoch": 12.23076923076923,
"grad_norm": 3.3413379192352295,
"learning_rate": 2.9036913696289318e-05,
"loss": 0.0658,
"step": 3180
},
{
"epoch": 12.26923076923077,
"grad_norm": 3.1574761867523193,
"learning_rate": 2.9028150633828186e-05,
"loss": 0.0673,
"step": 3190
},
{
"epoch": 12.307692307692308,
"grad_norm": 3.2442660331726074,
"learning_rate": 2.9019349218938887e-05,
"loss": 0.0769,
"step": 3200
},
{
"epoch": 12.346153846153847,
"grad_norm": 2.781846046447754,
"learning_rate": 2.9010509475684146e-05,
"loss": 0.068,
"step": 3210
},
{
"epoch": 12.384615384615385,
"grad_norm": 3.350027084350586,
"learning_rate": 2.900163142823149e-05,
"loss": 0.0619,
"step": 3220
},
{
"epoch": 12.423076923076923,
"grad_norm": 3.0505483150482178,
"learning_rate": 2.8992715100853166e-05,
"loss": 0.0669,
"step": 3230
},
{
"epoch": 12.461538461538462,
"grad_norm": 3.5292305946350098,
"learning_rate": 2.898376051792606e-05,
"loss": 0.0591,
"step": 3240
},
{
"epoch": 12.5,
"grad_norm": 3.194944143295288,
"learning_rate": 2.897476770393167e-05,
"loss": 0.0574,
"step": 3250
},
{
"epoch": 12.538461538461538,
"grad_norm": 3.1942076683044434,
"learning_rate": 2.8965736683456e-05,
"loss": 0.0752,
"step": 3260
},
{
"epoch": 12.576923076923077,
"grad_norm": 2.857682943344116,
"learning_rate": 2.895666748118952e-05,
"loss": 0.0666,
"step": 3270
},
{
"epoch": 12.615384615384615,
"grad_norm": 3.969834566116333,
"learning_rate": 2.8947560121927077e-05,
"loss": 0.0607,
"step": 3280
},
{
"epoch": 12.653846153846153,
"grad_norm": 3.6644859313964844,
"learning_rate": 2.8938414630567852e-05,
"loss": 0.066,
"step": 3290
},
{
"epoch": 12.692307692307692,
"grad_norm": 2.7282423973083496,
"learning_rate": 2.892923103211526e-05,
"loss": 0.0596,
"step": 3300
},
{
"epoch": 12.73076923076923,
"grad_norm": 3.0733189582824707,
"learning_rate": 2.892000935167691e-05,
"loss": 0.0583,
"step": 3310
},
{
"epoch": 12.76923076923077,
"grad_norm": 3.726287364959717,
"learning_rate": 2.8910749614464536e-05,
"loss": 0.0594,
"step": 3320
},
{
"epoch": 12.807692307692308,
"grad_norm": 3.276707887649536,
"learning_rate": 2.890145184579389e-05,
"loss": 0.0665,
"step": 3330
},
{
"epoch": 12.846153846153847,
"grad_norm": 3.513742208480835,
"learning_rate": 2.8892116071084727e-05,
"loss": 0.0635,
"step": 3340
},
{
"epoch": 12.884615384615385,
"grad_norm": 3.981471300125122,
"learning_rate": 2.8882742315860692e-05,
"loss": 0.0683,
"step": 3350
},
{
"epoch": 12.923076923076923,
"grad_norm": 3.698300838470459,
"learning_rate": 2.8873330605749275e-05,
"loss": 0.0624,
"step": 3360
},
{
"epoch": 12.961538461538462,
"grad_norm": 2.87587571144104,
"learning_rate": 2.886388096648174e-05,
"loss": 0.0752,
"step": 3370
},
{
"epoch": 13.0,
"grad_norm": 3.0200350284576416,
"learning_rate": 2.8854393423893024e-05,
"loss": 0.0653,
"step": 3380
},
{
"epoch": 13.038461538461538,
"grad_norm": 2.7889177799224854,
"learning_rate": 2.8844868003921723e-05,
"loss": 0.0678,
"step": 3390
},
{
"epoch": 13.076923076923077,
"grad_norm": 3.0269007682800293,
"learning_rate": 2.8835304732609962e-05,
"loss": 0.0713,
"step": 3400
},
{
"epoch": 13.115384615384615,
"grad_norm": 2.714719295501709,
"learning_rate": 2.882570363610336e-05,
"loss": 0.0641,
"step": 3410
},
{
"epoch": 13.153846153846153,
"grad_norm": 3.500612497329712,
"learning_rate": 2.8816064740650954e-05,
"loss": 0.0676,
"step": 3420
},
{
"epoch": 13.192307692307692,
"grad_norm": 3.845914840698242,
"learning_rate": 2.880638807260511e-05,
"loss": 0.0609,
"step": 3430
},
{
"epoch": 13.23076923076923,
"grad_norm": 2.972393274307251,
"learning_rate": 2.8796673658421472e-05,
"loss": 0.0659,
"step": 3440
},
{
"epoch": 13.26923076923077,
"grad_norm": 2.625223159790039,
"learning_rate": 2.8786921524658877e-05,
"loss": 0.0573,
"step": 3450
},
{
"epoch": 13.307692307692308,
"grad_norm": 3.5500714778900146,
"learning_rate": 2.8777131697979283e-05,
"loss": 0.0694,
"step": 3460
},
{
"epoch": 13.346153846153847,
"grad_norm": 2.848158121109009,
"learning_rate": 2.876730420514771e-05,
"loss": 0.0574,
"step": 3470
},
{
"epoch": 13.384615384615385,
"grad_norm": 3.2372846603393555,
"learning_rate": 2.8757439073032136e-05,
"loss": 0.0638,
"step": 3480
},
{
"epoch": 13.423076923076923,
"grad_norm": 2.9966297149658203,
"learning_rate": 2.874753632860347e-05,
"loss": 0.0636,
"step": 3490
},
{
"epoch": 13.461538461538462,
"grad_norm": 2.996350049972534,
"learning_rate": 2.873759599893543e-05,
"loss": 0.0721,
"step": 3500
},
{
"epoch": 13.5,
"grad_norm": 2.873504400253296,
"learning_rate": 2.8727618111204494e-05,
"loss": 0.0656,
"step": 3510
},
{
"epoch": 13.538461538461538,
"grad_norm": 2.916673183441162,
"learning_rate": 2.871760269268983e-05,
"loss": 0.0594,
"step": 3520
},
{
"epoch": 13.576923076923077,
"grad_norm": 2.4159295558929443,
"learning_rate": 2.870754977077321e-05,
"loss": 0.061,
"step": 3530
},
{
"epoch": 13.615384615384615,
"grad_norm": 2.685166597366333,
"learning_rate": 2.869745937293894e-05,
"loss": 0.0671,
"step": 3540
},
{
"epoch": 13.653846153846153,
"grad_norm": 2.81882905960083,
"learning_rate": 2.8687331526773775e-05,
"loss": 0.0645,
"step": 3550
},
{
"epoch": 13.692307692307692,
"grad_norm": 2.4796597957611084,
"learning_rate": 2.867716625996687e-05,
"loss": 0.0647,
"step": 3560
},
{
"epoch": 13.73076923076923,
"grad_norm": 2.852337598800659,
"learning_rate": 2.8666963600309672e-05,
"loss": 0.0701,
"step": 3570
},
{
"epoch": 13.76923076923077,
"grad_norm": 2.938345432281494,
"learning_rate": 2.8656723575695862e-05,
"loss": 0.0616,
"step": 3580
},
{
"epoch": 13.807692307692308,
"grad_norm": 2.8633902072906494,
"learning_rate": 2.8646446214121276e-05,
"loss": 0.0586,
"step": 3590
},
{
"epoch": 13.846153846153847,
"grad_norm": 2.704404830932617,
"learning_rate": 2.8636131543683828e-05,
"loss": 0.0654,
"step": 3600
},
{
"epoch": 13.884615384615385,
"grad_norm": 2.430389642715454,
"learning_rate": 2.8625779592583436e-05,
"loss": 0.0563,
"step": 3610
},
{
"epoch": 13.923076923076923,
"grad_norm": 2.908712387084961,
"learning_rate": 2.861539038912193e-05,
"loss": 0.061,
"step": 3620
},
{
"epoch": 13.961538461538462,
"grad_norm": 2.6836538314819336,
"learning_rate": 2.860496396170301e-05,
"loss": 0.0608,
"step": 3630
},
{
"epoch": 14.0,
"grad_norm": 2.7360219955444336,
"learning_rate": 2.859450033883212e-05,
"loss": 0.0627,
"step": 3640
},
{
"epoch": 14.038461538461538,
"grad_norm": 3.1698520183563232,
"learning_rate": 2.8583999549116413e-05,
"loss": 0.0627,
"step": 3650
},
{
"epoch": 14.076923076923077,
"grad_norm": 2.8817367553710938,
"learning_rate": 2.857346162126464e-05,
"loss": 0.053,
"step": 3660
},
{
"epoch": 14.115384615384615,
"grad_norm": 3.117218017578125,
"learning_rate": 2.8562886584087092e-05,
"loss": 0.0634,
"step": 3670
},
{
"epoch": 14.153846153846153,
"grad_norm": 3.047295331954956,
"learning_rate": 2.8552274466495525e-05,
"loss": 0.0654,
"step": 3680
},
{
"epoch": 14.192307692307692,
"grad_norm": 2.8166685104370117,
"learning_rate": 2.8541625297503056e-05,
"loss": 0.0647,
"step": 3690
},
{
"epoch": 14.23076923076923,
"grad_norm": 2.7225589752197266,
"learning_rate": 2.8530939106224106e-05,
"loss": 0.0592,
"step": 3700
},
{
"epoch": 14.26923076923077,
"grad_norm": 3.0143346786499023,
"learning_rate": 2.8520215921874325e-05,
"loss": 0.0619,
"step": 3710
},
{
"epoch": 14.307692307692308,
"grad_norm": 3.057224988937378,
"learning_rate": 2.850945577377048e-05,
"loss": 0.0636,
"step": 3720
},
{
"epoch": 14.346153846153847,
"grad_norm": 2.7452213764190674,
"learning_rate": 2.8498658691330406e-05,
"loss": 0.0544,
"step": 3730
},
{
"epoch": 14.384615384615385,
"grad_norm": 2.056774377822876,
"learning_rate": 2.8487824704072913e-05,
"loss": 0.062,
"step": 3740
},
{
"epoch": 14.423076923076923,
"grad_norm": 3.2659449577331543,
"learning_rate": 2.8476953841617713e-05,
"loss": 0.0614,
"step": 3750
},
{
"epoch": 14.461538461538462,
"grad_norm": 2.6030311584472656,
"learning_rate": 2.846604613368532e-05,
"loss": 0.056,
"step": 3760
},
{
"epoch": 14.5,
"grad_norm": 2.8850746154785156,
"learning_rate": 2.8455101610097002e-05,
"loss": 0.0681,
"step": 3770
},
{
"epoch": 14.538461538461538,
"grad_norm": 3.431685209274292,
"learning_rate": 2.8444120300774666e-05,
"loss": 0.0614,
"step": 3780
},
{
"epoch": 14.576923076923077,
"grad_norm": 3.1018617153167725,
"learning_rate": 2.8433102235740788e-05,
"loss": 0.057,
"step": 3790
},
{
"epoch": 14.615384615384615,
"grad_norm": 3.772675037384033,
"learning_rate": 2.842204744511834e-05,
"loss": 0.0591,
"step": 3800
},
{
"epoch": 14.653846153846153,
"grad_norm": 2.6003100872039795,
"learning_rate": 2.8410955959130693e-05,
"loss": 0.0536,
"step": 3810
},
{
"epoch": 14.692307692307692,
"grad_norm": 2.9984867572784424,
"learning_rate": 2.8399827808101554e-05,
"loss": 0.0635,
"step": 3820
},
{
"epoch": 14.73076923076923,
"grad_norm": 2.4007630348205566,
"learning_rate": 2.8388663022454857e-05,
"loss": 0.0612,
"step": 3830
},
{
"epoch": 14.76923076923077,
"grad_norm": 2.7119133472442627,
"learning_rate": 2.83774616327147e-05,
"loss": 0.0575,
"step": 3840
},
{
"epoch": 14.807692307692308,
"grad_norm": 3.0233607292175293,
"learning_rate": 2.836622366950526e-05,
"loss": 0.0586,
"step": 3850
},
{
"epoch": 14.846153846153847,
"grad_norm": 2.880004405975342,
"learning_rate": 2.835494916355069e-05,
"loss": 0.0647,
"step": 3860
},
{
"epoch": 14.884615384615385,
"grad_norm": 3.2674670219421387,
"learning_rate": 2.8343638145675072e-05,
"loss": 0.0619,
"step": 3870
},
{
"epoch": 14.923076923076923,
"grad_norm": 2.571911573410034,
"learning_rate": 2.8332290646802282e-05,
"loss": 0.0588,
"step": 3880
},
{
"epoch": 14.961538461538462,
"grad_norm": 2.711045026779175,
"learning_rate": 2.8320906697955963e-05,
"loss": 0.0547,
"step": 3890
},
{
"epoch": 15.0,
"grad_norm": 2.6947715282440186,
"learning_rate": 2.8309486330259385e-05,
"loss": 0.0614,
"step": 3900
},
{
"epoch": 15.038461538461538,
"grad_norm": 3.1563644409179688,
"learning_rate": 2.82980295749354e-05,
"loss": 0.0619,
"step": 3910
},
{
"epoch": 15.076923076923077,
"grad_norm": 2.481595993041992,
"learning_rate": 2.828653646330634e-05,
"loss": 0.0603,
"step": 3920
},
{
"epoch": 15.115384615384615,
"grad_norm": 3.0050806999206543,
"learning_rate": 2.8275007026793938e-05,
"loss": 0.0594,
"step": 3930
},
{
"epoch": 15.153846153846153,
"grad_norm": 3.2291088104248047,
"learning_rate": 2.826344129691923e-05,
"loss": 0.0632,
"step": 3940
},
{
"epoch": 15.192307692307692,
"grad_norm": 3.0251619815826416,
"learning_rate": 2.8251839305302478e-05,
"loss": 0.063,
"step": 3950
},
{
"epoch": 15.23076923076923,
"grad_norm": 2.865499258041382,
"learning_rate": 2.8240201083663088e-05,
"loss": 0.0574,
"step": 3960
},
{
"epoch": 15.26923076923077,
"grad_norm": 2.642956018447876,
"learning_rate": 2.8228526663819504e-05,
"loss": 0.06,
"step": 3970
},
{
"epoch": 15.307692307692308,
"grad_norm": 2.802671194076538,
"learning_rate": 2.8216816077689158e-05,
"loss": 0.0568,
"step": 3980
},
{
"epoch": 15.346153846153847,
"grad_norm": 2.3889827728271484,
"learning_rate": 2.8205069357288337e-05,
"loss": 0.0553,
"step": 3990
},
{
"epoch": 15.384615384615385,
"grad_norm": 2.5534961223602295,
"learning_rate": 2.8193286534732128e-05,
"loss": 0.0609,
"step": 4000
},
{
"epoch": 15.423076923076923,
"grad_norm": 2.4432270526885986,
"learning_rate": 2.8181467642234317e-05,
"loss": 0.0571,
"step": 4010
},
{
"epoch": 15.461538461538462,
"grad_norm": 2.605041980743408,
"learning_rate": 2.8169612712107306e-05,
"loss": 0.0554,
"step": 4020
},
{
"epoch": 15.5,
"grad_norm": 2.6760783195495605,
"learning_rate": 2.8157721776762017e-05,
"loss": 0.0631,
"step": 4030
},
{
"epoch": 15.538461538461538,
"grad_norm": 3.092106819152832,
"learning_rate": 2.814579486870782e-05,
"loss": 0.0608,
"step": 4040
},
{
"epoch": 15.576923076923077,
"grad_norm": 2.732588052749634,
"learning_rate": 2.813383202055242e-05,
"loss": 0.0552,
"step": 4050
},
{
"epoch": 15.615384615384615,
"grad_norm": 2.6859591007232666,
"learning_rate": 2.8121833265001792e-05,
"loss": 0.0654,
"step": 4060
},
{
"epoch": 15.653846153846153,
"grad_norm": 1.9384580850601196,
"learning_rate": 2.8109798634860072e-05,
"loss": 0.059,
"step": 4070
},
{
"epoch": 15.692307692307692,
"grad_norm": 2.3426663875579834,
"learning_rate": 2.8097728163029482e-05,
"loss": 0.0527,
"step": 4080
},
{
"epoch": 15.73076923076923,
"grad_norm": 2.785004138946533,
"learning_rate": 2.8085621882510233e-05,
"loss": 0.0568,
"step": 4090
},
{
"epoch": 15.76923076923077,
"grad_norm": 3.2040162086486816,
"learning_rate": 2.8073479826400425e-05,
"loss": 0.0622,
"step": 4100
},
{
"epoch": 15.807692307692308,
"grad_norm": 2.6994717121124268,
"learning_rate": 2.806130202789598e-05,
"loss": 0.0596,
"step": 4110
},
{
"epoch": 15.846153846153847,
"grad_norm": 2.955667018890381,
"learning_rate": 2.804908852029054e-05,
"loss": 0.0576,
"step": 4120
},
{
"epoch": 15.884615384615385,
"grad_norm": 2.229564905166626,
"learning_rate": 2.8036839336975367e-05,
"loss": 0.0567,
"step": 4130
},
{
"epoch": 15.923076923076923,
"grad_norm": 2.8583757877349854,
"learning_rate": 2.8024554511439253e-05,
"loss": 0.0588,
"step": 4140
},
{
"epoch": 15.961538461538462,
"grad_norm": 3.233686685562134,
"learning_rate": 2.801223407726844e-05,
"loss": 0.055,
"step": 4150
},
{
"epoch": 16.0,
"grad_norm": 3.210627794265747,
"learning_rate": 2.7999878068146537e-05,
"loss": 0.0493,
"step": 4160
},
{
"epoch": 16.03846153846154,
"grad_norm": 2.979895830154419,
"learning_rate": 2.7987486517854396e-05,
"loss": 0.0604,
"step": 4170
},
{
"epoch": 16.076923076923077,
"grad_norm": 2.135582447052002,
"learning_rate": 2.7975059460270037e-05,
"loss": 0.0618,
"step": 4180
},
{
"epoch": 16.115384615384617,
"grad_norm": 2.363574743270874,
"learning_rate": 2.7962596929368566e-05,
"loss": 0.0581,
"step": 4190
},
{
"epoch": 16.153846153846153,
"grad_norm": 2.2562291622161865,
"learning_rate": 2.795009895922207e-05,
"loss": 0.0541,
"step": 4200
},
{
"epoch": 16.192307692307693,
"grad_norm": 2.4633421897888184,
"learning_rate": 2.7937565583999513e-05,
"loss": 0.0501,
"step": 4210
},
{
"epoch": 16.23076923076923,
"grad_norm": 2.5683891773223877,
"learning_rate": 2.792499683796667e-05,
"loss": 0.0537,
"step": 4220
},
{
"epoch": 16.26923076923077,
"grad_norm": 2.360839605331421,
"learning_rate": 2.791239275548601e-05,
"loss": 0.0478,
"step": 4230
},
{
"epoch": 16.307692307692307,
"grad_norm": 2.2146573066711426,
"learning_rate": 2.789975337101662e-05,
"loss": 0.0591,
"step": 4240
},
{
"epoch": 16.346153846153847,
"grad_norm": 2.4830048084259033,
"learning_rate": 2.788707871911409e-05,
"loss": 0.0597,
"step": 4250
},
{
"epoch": 16.384615384615383,
"grad_norm": 2.256363868713379,
"learning_rate": 2.7874368834430426e-05,
"loss": 0.0549,
"step": 4260
},
{
"epoch": 16.423076923076923,
"grad_norm": 2.8379452228546143,
"learning_rate": 2.7861623751713982e-05,
"loss": 0.0524,
"step": 4270
},
{
"epoch": 16.46153846153846,
"grad_norm": 2.0764126777648926,
"learning_rate": 2.7848843505809317e-05,
"loss": 0.0553,
"step": 4280
},
{
"epoch": 16.5,
"grad_norm": 2.321157217025757,
"learning_rate": 2.7836028131657142e-05,
"loss": 0.0512,
"step": 4290
},
{
"epoch": 16.53846153846154,
"grad_norm": 2.401200532913208,
"learning_rate": 2.7823177664294197e-05,
"loss": 0.0541,
"step": 4300
},
{
"epoch": 16.576923076923077,
"grad_norm": 2.6631157398223877,
"learning_rate": 2.7810292138853168e-05,
"loss": 0.0532,
"step": 4310
},
{
"epoch": 16.615384615384617,
"grad_norm": 2.2633821964263916,
"learning_rate": 2.779737159056259e-05,
"loss": 0.0536,
"step": 4320
},
{
"epoch": 16.653846153846153,
"grad_norm": 2.5215775966644287,
"learning_rate": 2.7784416054746753e-05,
"loss": 0.0526,
"step": 4330
},
{
"epoch": 16.692307692307693,
"grad_norm": 2.0003058910369873,
"learning_rate": 2.7771425566825593e-05,
"loss": 0.0531,
"step": 4340
},
{
"epoch": 16.73076923076923,
"grad_norm": 2.353093385696411,
"learning_rate": 2.7758400162314605e-05,
"loss": 0.0544,
"step": 4350
},
{
"epoch": 16.76923076923077,
"grad_norm": 2.478005886077881,
"learning_rate": 2.7745339876824756e-05,
"loss": 0.0503,
"step": 4360
},
{
"epoch": 16.807692307692307,
"grad_norm": 2.6122934818267822,
"learning_rate": 2.7732244746062363e-05,
"loss": 0.0543,
"step": 4370
},
{
"epoch": 16.846153846153847,
"grad_norm": 2.5369300842285156,
"learning_rate": 2.7719114805829015e-05,
"loss": 0.052,
"step": 4380
},
{
"epoch": 16.884615384615383,
"grad_norm": 2.575892925262451,
"learning_rate": 2.7705950092021465e-05,
"loss": 0.0577,
"step": 4390
},
{
"epoch": 16.923076923076923,
"grad_norm": 3.0408105850219727,
"learning_rate": 2.7692750640631533e-05,
"loss": 0.0659,
"step": 4400
},
{
"epoch": 16.96153846153846,
"grad_norm": 2.7433135509490967,
"learning_rate": 2.767951648774603e-05,
"loss": 0.0558,
"step": 4410
},
{
"epoch": 17.0,
"grad_norm": 2.37965989112854,
"learning_rate": 2.766624766954661e-05,
"loss": 0.0563,
"step": 4420
},
{
"epoch": 17.03846153846154,
"grad_norm": 2.1070268154144287,
"learning_rate": 2.7652944222309727e-05,
"loss": 0.0541,
"step": 4430
},
{
"epoch": 17.076923076923077,
"grad_norm": 2.4608118534088135,
"learning_rate": 2.7639606182406484e-05,
"loss": 0.0619,
"step": 4440
},
{
"epoch": 17.115384615384617,
"grad_norm": 2.633108615875244,
"learning_rate": 2.7626233586302583e-05,
"loss": 0.0496,
"step": 4450
},
{
"epoch": 17.153846153846153,
"grad_norm": 2.1537821292877197,
"learning_rate": 2.7612826470558192e-05,
"loss": 0.0578,
"step": 4460
},
{
"epoch": 17.192307692307693,
"grad_norm": 2.589418411254883,
"learning_rate": 2.7599384871827846e-05,
"loss": 0.0547,
"step": 4470
},
{
"epoch": 17.23076923076923,
"grad_norm": 3.257871389389038,
"learning_rate": 2.7585908826860368e-05,
"loss": 0.0537,
"step": 4480
},
{
"epoch": 17.26923076923077,
"grad_norm": 2.2862274646759033,
"learning_rate": 2.757239837249875e-05,
"loss": 0.0552,
"step": 4490
},
{
"epoch": 17.307692307692307,
"grad_norm": 2.6018054485321045,
"learning_rate": 2.7558853545680057e-05,
"loss": 0.0523,
"step": 4500
},
{
"epoch": 17.346153846153847,
"grad_norm": 2.4538543224334717,
"learning_rate": 2.754527438343533e-05,
"loss": 0.0621,
"step": 4510
},
{
"epoch": 17.384615384615383,
"grad_norm": 2.5169732570648193,
"learning_rate": 2.7531660922889477e-05,
"loss": 0.0532,
"step": 4520
},
{
"epoch": 17.423076923076923,
"grad_norm": 2.893049955368042,
"learning_rate": 2.751801320126118e-05,
"loss": 0.0559,
"step": 4530
},
{
"epoch": 17.46153846153846,
"grad_norm": 2.932993173599243,
"learning_rate": 2.750433125586279e-05,
"loss": 0.0512,
"step": 4540
},
{
"epoch": 17.5,
"grad_norm": 2.4409329891204834,
"learning_rate": 2.7490615124100225e-05,
"loss": 0.0478,
"step": 4550
},
{
"epoch": 17.53846153846154,
"grad_norm": 2.2984349727630615,
"learning_rate": 2.747686484347286e-05,
"loss": 0.0484,
"step": 4560
},
{
"epoch": 17.576923076923077,
"grad_norm": 2.0560920238494873,
"learning_rate": 2.7463080451573447e-05,
"loss": 0.0545,
"step": 4570
},
{
"epoch": 17.615384615384617,
"grad_norm": 2.1629302501678467,
"learning_rate": 2.744926198608798e-05,
"loss": 0.0522,
"step": 4580
},
{
"epoch": 17.653846153846153,
"grad_norm": 2.228933811187744,
"learning_rate": 2.743540948479561e-05,
"loss": 0.0534,
"step": 4590
},
{
"epoch": 17.692307692307693,
"grad_norm": 2.0013070106506348,
"learning_rate": 2.7421522985568562e-05,
"loss": 0.0538,
"step": 4600
},
{
"epoch": 17.73076923076923,
"grad_norm": 2.4562020301818848,
"learning_rate": 2.7407602526371983e-05,
"loss": 0.0549,
"step": 4610
},
{
"epoch": 17.76923076923077,
"grad_norm": 2.1591811180114746,
"learning_rate": 2.7393648145263873e-05,
"loss": 0.055,
"step": 4620
},
{
"epoch": 17.807692307692307,
"grad_norm": 2.4402129650115967,
"learning_rate": 2.7379659880394996e-05,
"loss": 0.0546,
"step": 4630
},
{
"epoch": 17.846153846153847,
"grad_norm": 2.294828176498413,
"learning_rate": 2.7365637770008717e-05,
"loss": 0.0538,
"step": 4640
},
{
"epoch": 17.884615384615383,
"grad_norm": 2.2330775260925293,
"learning_rate": 2.7351581852440953e-05,
"loss": 0.0527,
"step": 4650
},
{
"epoch": 17.923076923076923,
"grad_norm": 2.5576770305633545,
"learning_rate": 2.7337492166120053e-05,
"loss": 0.0633,
"step": 4660
},
{
"epoch": 17.96153846153846,
"grad_norm": 2.897631883621216,
"learning_rate": 2.732336874956667e-05,
"loss": 0.0551,
"step": 4670
},
{
"epoch": 18.0,
"grad_norm": 2.2748541831970215,
"learning_rate": 2.7309211641393696e-05,
"loss": 0.0444,
"step": 4680
},
{
"epoch": 18.03846153846154,
"grad_norm": 2.8125858306884766,
"learning_rate": 2.7295020880306123e-05,
"loss": 0.0575,
"step": 4690
},
{
"epoch": 18.076923076923077,
"grad_norm": 2.6854939460754395,
"learning_rate": 2.7280796505100946e-05,
"loss": 0.0586,
"step": 4700
},
{
"epoch": 18.115384615384617,
"grad_norm": 2.2231080532073975,
"learning_rate": 2.7266538554667065e-05,
"loss": 0.0524,
"step": 4710
},
{
"epoch": 18.153846153846153,
"grad_norm": 1.815577745437622,
"learning_rate": 2.725224706798517e-05,
"loss": 0.0568,
"step": 4720
},
{
"epoch": 18.192307692307693,
"grad_norm": 2.6000752449035645,
"learning_rate": 2.7237922084127643e-05,
"loss": 0.0534,
"step": 4730
},
{
"epoch": 18.23076923076923,
"grad_norm": 2.511378526687622,
"learning_rate": 2.7223563642258446e-05,
"loss": 0.0497,
"step": 4740
},
{
"epoch": 18.26923076923077,
"grad_norm": 2.3093998432159424,
"learning_rate": 2.7209171781633e-05,
"loss": 0.0528,
"step": 4750
},
{
"epoch": 18.307692307692307,
"grad_norm": 2.629744529724121,
"learning_rate": 2.7194746541598113e-05,
"loss": 0.0513,
"step": 4760
},
{
"epoch": 18.346153846153847,
"grad_norm": 2.0061843395233154,
"learning_rate": 2.7180287961591835e-05,
"loss": 0.0527,
"step": 4770
},
{
"epoch": 18.384615384615383,
"grad_norm": 2.1390366554260254,
"learning_rate": 2.7165796081143377e-05,
"loss": 0.0537,
"step": 4780
},
{
"epoch": 18.423076923076923,
"grad_norm": 1.9976611137390137,
"learning_rate": 2.715127093987298e-05,
"loss": 0.0498,
"step": 4790
},
{
"epoch": 18.46153846153846,
"grad_norm": 2.1415915489196777,
"learning_rate": 2.713671257749183e-05,
"loss": 0.0518,
"step": 4800
},
{
"epoch": 18.5,
"grad_norm": 2.861093044281006,
"learning_rate": 2.712212103380193e-05,
"loss": 0.0499,
"step": 4810
},
{
"epoch": 18.53846153846154,
"grad_norm": 2.4393386840820312,
"learning_rate": 2.7107496348696004e-05,
"loss": 0.0469,
"step": 4820
},
{
"epoch": 18.576923076923077,
"grad_norm": 2.497103691101074,
"learning_rate": 2.7092838562157386e-05,
"loss": 0.0608,
"step": 4830
},
{
"epoch": 18.615384615384617,
"grad_norm": 2.7781546115875244,
"learning_rate": 2.7078147714259905e-05,
"loss": 0.0613,
"step": 4840
},
{
"epoch": 18.653846153846153,
"grad_norm": 2.271496057510376,
"learning_rate": 2.7063423845167773e-05,
"loss": 0.0519,
"step": 4850
},
{
"epoch": 18.692307692307693,
"grad_norm": 2.31038761138916,
"learning_rate": 2.7048666995135494e-05,
"loss": 0.0521,
"step": 4860
},
{
"epoch": 18.73076923076923,
"grad_norm": 2.466082811355591,
"learning_rate": 2.7033877204507722e-05,
"loss": 0.0517,
"step": 4870
},
{
"epoch": 18.76923076923077,
"grad_norm": 2.2604198455810547,
"learning_rate": 2.701905451371919e-05,
"loss": 0.0554,
"step": 4880
},
{
"epoch": 18.807692307692307,
"grad_norm": 2.663639545440674,
"learning_rate": 2.7004198963294558e-05,
"loss": 0.0548,
"step": 4890
},
{
"epoch": 18.846153846153847,
"grad_norm": 2.3590447902679443,
"learning_rate": 2.6989310593848345e-05,
"loss": 0.0547,
"step": 4900
},
{
"epoch": 18.884615384615383,
"grad_norm": 2.3990252017974854,
"learning_rate": 2.6974389446084776e-05,
"loss": 0.0483,
"step": 4910
},
{
"epoch": 18.923076923076923,
"grad_norm": 2.5169942378997803,
"learning_rate": 2.6959435560797706e-05,
"loss": 0.0487,
"step": 4920
},
{
"epoch": 18.96153846153846,
"grad_norm": 2.5757737159729004,
"learning_rate": 2.6944448978870478e-05,
"loss": 0.0564,
"step": 4930
},
{
"epoch": 19.0,
"grad_norm": 2.7602148056030273,
"learning_rate": 2.6929429741275845e-05,
"loss": 0.0496,
"step": 4940
},
{
"epoch": 19.03846153846154,
"grad_norm": 2.638188123703003,
"learning_rate": 2.691437788907582e-05,
"loss": 0.0519,
"step": 4950
},
{
"epoch": 19.076923076923077,
"grad_norm": 2.3091073036193848,
"learning_rate": 2.689929346342159e-05,
"loss": 0.0457,
"step": 4960
},
{
"epoch": 19.115384615384617,
"grad_norm": 2.0856239795684814,
"learning_rate": 2.688417650555341e-05,
"loss": 0.0588,
"step": 4970
},
{
"epoch": 19.153846153846153,
"grad_norm": 1.9847303628921509,
"learning_rate": 2.686902705680046e-05,
"loss": 0.0571,
"step": 4980
},
{
"epoch": 19.192307692307693,
"grad_norm": 2.181013822555542,
"learning_rate": 2.6853845158580756e-05,
"loss": 0.0518,
"step": 4990
},
{
"epoch": 19.23076923076923,
"grad_norm": 2.1970889568328857,
"learning_rate": 2.6838630852401028e-05,
"loss": 0.0531,
"step": 5000
},
{
"epoch": 19.26923076923077,
"grad_norm": 2.3489420413970947,
"learning_rate": 2.6823384179856602e-05,
"loss": 0.052,
"step": 5010
},
{
"epoch": 19.307692307692307,
"grad_norm": 2.067884683609009,
"learning_rate": 2.6808105182631303e-05,
"loss": 0.0545,
"step": 5020
},
{
"epoch": 19.346153846153847,
"grad_norm": 2.4723005294799805,
"learning_rate": 2.6792793902497328e-05,
"loss": 0.0506,
"step": 5030
},
{
"epoch": 19.384615384615383,
"grad_norm": 2.328829050064087,
"learning_rate": 2.6777450381315133e-05,
"loss": 0.0499,
"step": 5040
},
{
"epoch": 19.423076923076923,
"grad_norm": 2.5263142585754395,
"learning_rate": 2.676207466103331e-05,
"loss": 0.052,
"step": 5050
},
{
"epoch": 19.46153846153846,
"grad_norm": 3.0785343647003174,
"learning_rate": 2.6746666783688503e-05,
"loss": 0.0507,
"step": 5060
},
{
"epoch": 19.5,
"grad_norm": 2.390286684036255,
"learning_rate": 2.673122679140525e-05,
"loss": 0.0421,
"step": 5070
},
{
"epoch": 19.53846153846154,
"grad_norm": 2.432105302810669,
"learning_rate": 2.671575472639591e-05,
"loss": 0.0515,
"step": 5080
},
{
"epoch": 19.576923076923077,
"grad_norm": 1.704558253288269,
"learning_rate": 2.6700250630960506e-05,
"loss": 0.0527,
"step": 5090
},
{
"epoch": 19.615384615384617,
"grad_norm": 2.047685384750366,
"learning_rate": 2.6684714547486654e-05,
"loss": 0.0538,
"step": 5100
},
{
"epoch": 19.653846153846153,
"grad_norm": 2.3358631134033203,
"learning_rate": 2.6669146518449407e-05,
"loss": 0.0479,
"step": 5110
},
{
"epoch": 19.692307692307693,
"grad_norm": 1.9280929565429688,
"learning_rate": 2.665354658641117e-05,
"loss": 0.0599,
"step": 5120
},
{
"epoch": 19.73076923076923,
"grad_norm": 1.7903410196304321,
"learning_rate": 2.6637914794021552e-05,
"loss": 0.0457,
"step": 5130
},
{
"epoch": 19.76923076923077,
"grad_norm": 2.035609483718872,
"learning_rate": 2.6622251184017274e-05,
"loss": 0.0483,
"step": 5140
},
{
"epoch": 19.807692307692307,
"grad_norm": 2.2856998443603516,
"learning_rate": 2.660655579922206e-05,
"loss": 0.0488,
"step": 5150
},
{
"epoch": 19.846153846153847,
"grad_norm": 2.0502729415893555,
"learning_rate": 2.6590828682546487e-05,
"loss": 0.0545,
"step": 5160
},
{
"epoch": 19.884615384615383,
"grad_norm": 2.134788990020752,
"learning_rate": 2.657506987698789e-05,
"loss": 0.0409,
"step": 5170
},
{
"epoch": 19.923076923076923,
"grad_norm": 2.1987850666046143,
"learning_rate": 2.655927942563024e-05,
"loss": 0.0503,
"step": 5180
},
{
"epoch": 19.96153846153846,
"grad_norm": 2.169670581817627,
"learning_rate": 2.6543457371644027e-05,
"loss": 0.0515,
"step": 5190
},
{
"epoch": 20.0,
"grad_norm": 2.150855779647827,
"learning_rate": 2.652760375828615e-05,
"loss": 0.0554,
"step": 5200
},
{
"epoch": 20.03846153846154,
"grad_norm": 2.300340414047241,
"learning_rate": 2.651171862889978e-05,
"loss": 0.0507,
"step": 5210
},
{
"epoch": 20.076923076923077,
"grad_norm": 2.2516908645629883,
"learning_rate": 2.649580202691425e-05,
"loss": 0.0514,
"step": 5220
},
{
"epoch": 20.115384615384617,
"grad_norm": 2.0438835620880127,
"learning_rate": 2.6479853995844942e-05,
"loss": 0.052,
"step": 5230
},
{
"epoch": 20.153846153846153,
"grad_norm": 2.1609389781951904,
"learning_rate": 2.646387457929317e-05,
"loss": 0.0489,
"step": 5240
},
{
"epoch": 20.192307692307693,
"grad_norm": 1.979067087173462,
"learning_rate": 2.6447863820946047e-05,
"loss": 0.0455,
"step": 5250
},
{
"epoch": 20.23076923076923,
"grad_norm": 2.4219133853912354,
"learning_rate": 2.6431821764576367e-05,
"loss": 0.0505,
"step": 5260
},
{
"epoch": 20.26923076923077,
"grad_norm": 1.8512394428253174,
"learning_rate": 2.641574845404251e-05,
"loss": 0.0499,
"step": 5270
},
{
"epoch": 20.307692307692307,
"grad_norm": 1.9653434753417969,
"learning_rate": 2.639964393328829e-05,
"loss": 0.0533,
"step": 5280
},
{
"epoch": 20.346153846153847,
"grad_norm": 2.2867801189422607,
"learning_rate": 2.6383508246342844e-05,
"loss": 0.0503,
"step": 5290
},
{
"epoch": 20.384615384615383,
"grad_norm": 1.9846916198730469,
"learning_rate": 2.636734143732054e-05,
"loss": 0.0493,
"step": 5300
},
{
"epoch": 20.423076923076923,
"grad_norm": 2.2939069271087646,
"learning_rate": 2.63511435504208e-05,
"loss": 0.0499,
"step": 5310
},
{
"epoch": 20.46153846153846,
"grad_norm": 1.8951475620269775,
"learning_rate": 2.633491462992804e-05,
"loss": 0.0508,
"step": 5320
},
{
"epoch": 20.5,
"grad_norm": 2.4194393157958984,
"learning_rate": 2.63186547202115e-05,
"loss": 0.0461,
"step": 5330
},
{
"epoch": 20.53846153846154,
"grad_norm": 2.025209665298462,
"learning_rate": 2.6302363865725158e-05,
"loss": 0.0518,
"step": 5340
},
{
"epoch": 20.576923076923077,
"grad_norm": 3.0230369567871094,
"learning_rate": 2.628604211100759e-05,
"loss": 0.051,
"step": 5350
},
{
"epoch": 20.615384615384617,
"grad_norm": 2.486722707748413,
"learning_rate": 2.6269689500681846e-05,
"loss": 0.052,
"step": 5360
},
{
"epoch": 20.653846153846153,
"grad_norm": 2.9617080688476562,
"learning_rate": 2.6253306079455337e-05,
"loss": 0.0545,
"step": 5370
},
{
"epoch": 20.692307692307693,
"grad_norm": 2.7517824172973633,
"learning_rate": 2.6236891892119713e-05,
"loss": 0.0501,
"step": 5380
},
{
"epoch": 20.73076923076923,
"grad_norm": 2.343135118484497,
"learning_rate": 2.6220446983550738e-05,
"loss": 0.0486,
"step": 5390
},
{
"epoch": 20.76923076923077,
"grad_norm": 2.22013258934021,
"learning_rate": 2.6203971398708162e-05,
"loss": 0.0491,
"step": 5400
},
{
"epoch": 20.807692307692307,
"grad_norm": 2.3863282203674316,
"learning_rate": 2.6187465182635598e-05,
"loss": 0.052,
"step": 5410
},
{
"epoch": 20.846153846153847,
"grad_norm": 2.3278868198394775,
"learning_rate": 2.6170928380460424e-05,
"loss": 0.046,
"step": 5420
},
{
"epoch": 20.884615384615383,
"grad_norm": 2.228511095046997,
"learning_rate": 2.615436103739362e-05,
"loss": 0.0498,
"step": 5430
},
{
"epoch": 20.923076923076923,
"grad_norm": 2.5615031719207764,
"learning_rate": 2.6137763198729665e-05,
"loss": 0.0472,
"step": 5440
},
{
"epoch": 20.96153846153846,
"grad_norm": 2.5572619438171387,
"learning_rate": 2.6121134909846416e-05,
"loss": 0.0549,
"step": 5450
},
{
"epoch": 21.0,
"grad_norm": 2.7003588676452637,
"learning_rate": 2.6104476216204985e-05,
"loss": 0.0451,
"step": 5460
},
{
"epoch": 21.03846153846154,
"grad_norm": 2.0107014179229736,
"learning_rate": 2.6087787163349605e-05,
"loss": 0.0481,
"step": 5470
},
{
"epoch": 21.076923076923077,
"grad_norm": 2.2759485244750977,
"learning_rate": 2.60710677969075e-05,
"loss": 0.0473,
"step": 5480
},
{
"epoch": 21.115384615384617,
"grad_norm": 2.0311803817749023,
"learning_rate": 2.6054318162588792e-05,
"loss": 0.0437,
"step": 5490
},
{
"epoch": 21.153846153846153,
"grad_norm": 2.2505581378936768,
"learning_rate": 2.6037538306186337e-05,
"loss": 0.0445,
"step": 5500
},
{
"epoch": 21.192307692307693,
"grad_norm": 2.1031274795532227,
"learning_rate": 2.602072827357562e-05,
"loss": 0.0491,
"step": 5510
},
{
"epoch": 21.23076923076923,
"grad_norm": 2.2294819355010986,
"learning_rate": 2.6003888110714624e-05,
"loss": 0.0405,
"step": 5520
},
{
"epoch": 21.26923076923077,
"grad_norm": 2.206677198410034,
"learning_rate": 2.5987017863643714e-05,
"loss": 0.054,
"step": 5530
},
{
"epoch": 21.307692307692307,
"grad_norm": 1.9745968580245972,
"learning_rate": 2.5970117578485506e-05,
"loss": 0.0469,
"step": 5540
},
{
"epoch": 21.346153846153847,
"grad_norm": 1.8446788787841797,
"learning_rate": 2.5953187301444733e-05,
"loss": 0.0431,
"step": 5550
},
{
"epoch": 21.384615384615383,
"grad_norm": 2.375255584716797,
"learning_rate": 2.5936227078808123e-05,
"loss": 0.0489,
"step": 5560
},
{
"epoch": 21.423076923076923,
"grad_norm": 2.1657443046569824,
"learning_rate": 2.5919236956944277e-05,
"loss": 0.0442,
"step": 5570
},
{
"epoch": 21.46153846153846,
"grad_norm": 2.1673221588134766,
"learning_rate": 2.5902216982303544e-05,
"loss": 0.0489,
"step": 5580
},
{
"epoch": 21.5,
"grad_norm": 2.0828003883361816,
"learning_rate": 2.588516720141788e-05,
"loss": 0.0465,
"step": 5590
},
{
"epoch": 21.53846153846154,
"grad_norm": 2.3611319065093994,
"learning_rate": 2.5868087660900735e-05,
"loss": 0.0506,
"step": 5600
},
{
"epoch": 21.576923076923077,
"grad_norm": 2.302957773208618,
"learning_rate": 2.5850978407446924e-05,
"loss": 0.0542,
"step": 5610
},
{
"epoch": 21.615384615384617,
"grad_norm": 2.3799068927764893,
"learning_rate": 2.5833839487832488e-05,
"loss": 0.0513,
"step": 5620
},
{
"epoch": 21.653846153846153,
"grad_norm": 2.2304797172546387,
"learning_rate": 2.5816670948914583e-05,
"loss": 0.0439,
"step": 5630
},
{
"epoch": 21.692307692307693,
"grad_norm": 2.4063093662261963,
"learning_rate": 2.5799472837631338e-05,
"loss": 0.0521,
"step": 5640
},
{
"epoch": 21.73076923076923,
"grad_norm": 1.8401644229888916,
"learning_rate": 2.578224520100173e-05,
"loss": 0.0416,
"step": 5650
},
{
"epoch": 21.76923076923077,
"grad_norm": 2.4482624530792236,
"learning_rate": 2.576498808612546e-05,
"loss": 0.0477,
"step": 5660
},
{
"epoch": 21.807692307692307,
"grad_norm": 2.242112398147583,
"learning_rate": 2.5747701540182825e-05,
"loss": 0.051,
"step": 5670
},
{
"epoch": 21.846153846153847,
"grad_norm": 2.264697790145874,
"learning_rate": 2.573038561043458e-05,
"loss": 0.0463,
"step": 5680
},
{
"epoch": 21.884615384615383,
"grad_norm": 2.5730340480804443,
"learning_rate": 2.5713040344221815e-05,
"loss": 0.0539,
"step": 5690
},
{
"epoch": 21.923076923076923,
"grad_norm": 2.2590243816375732,
"learning_rate": 2.5695665788965823e-05,
"loss": 0.0423,
"step": 5700
},
{
"epoch": 21.96153846153846,
"grad_norm": 2.2034289836883545,
"learning_rate": 2.5678261992167978e-05,
"loss": 0.0539,
"step": 5710
},
{
"epoch": 22.0,
"grad_norm": 2.3551292419433594,
"learning_rate": 2.5660829001409594e-05,
"loss": 0.0555,
"step": 5720
},
{
"epoch": 22.03846153846154,
"grad_norm": 1.970501184463501,
"learning_rate": 2.5643366864351806e-05,
"loss": 0.0554,
"step": 5730
},
{
"epoch": 22.076923076923077,
"grad_norm": 2.2068798542022705,
"learning_rate": 2.5625875628735423e-05,
"loss": 0.0461,
"step": 5740
},
{
"epoch": 22.115384615384617,
"grad_norm": 2.198699712753296,
"learning_rate": 2.560835534238082e-05,
"loss": 0.0463,
"step": 5750
},
{
"epoch": 22.153846153846153,
"grad_norm": 1.864600658416748,
"learning_rate": 2.5590806053187793e-05,
"loss": 0.0474,
"step": 5760
},
{
"epoch": 22.192307692307693,
"grad_norm": 2.2763874530792236,
"learning_rate": 2.557322780913542e-05,
"loss": 0.0493,
"step": 5770
},
{
"epoch": 22.23076923076923,
"grad_norm": 2.336698293685913,
"learning_rate": 2.555562065828196e-05,
"loss": 0.0532,
"step": 5780
},
{
"epoch": 22.26923076923077,
"grad_norm": 2.111300230026245,
"learning_rate": 2.5537984648764684e-05,
"loss": 0.0461,
"step": 5790
},
{
"epoch": 22.307692307692307,
"grad_norm": 1.9061249494552612,
"learning_rate": 2.5520319828799766e-05,
"loss": 0.0478,
"step": 5800
},
{
"epoch": 22.346153846153847,
"grad_norm": 1.8777246475219727,
"learning_rate": 2.550262624668216e-05,
"loss": 0.0433,
"step": 5810
},
{
"epoch": 22.384615384615383,
"grad_norm": 1.811094880104065,
"learning_rate": 2.5484903950785432e-05,
"loss": 0.0443,
"step": 5820
},
{
"epoch": 22.423076923076923,
"grad_norm": 2.2233502864837646,
"learning_rate": 2.546715298956167e-05,
"loss": 0.0534,
"step": 5830
},
{
"epoch": 22.46153846153846,
"grad_norm": 1.968826174736023,
"learning_rate": 2.5449373411541322e-05,
"loss": 0.05,
"step": 5840
},
{
"epoch": 22.5,
"grad_norm": 2.2315022945404053,
"learning_rate": 2.5431565265333074e-05,
"loss": 0.0558,
"step": 5850
},
{
"epoch": 22.53846153846154,
"grad_norm": 1.852952480316162,
"learning_rate": 2.541372859962372e-05,
"loss": 0.0478,
"step": 5860
},
{
"epoch": 22.576923076923077,
"grad_norm": 2.2287561893463135,
"learning_rate": 2.5395863463178023e-05,
"loss": 0.0466,
"step": 5870
},
{
"epoch": 22.615384615384617,
"grad_norm": 2.396843671798706,
"learning_rate": 2.537796990483858e-05,
"loss": 0.0601,
"step": 5880
},
{
"epoch": 22.653846153846153,
"grad_norm": 2.18788480758667,
"learning_rate": 2.53600479735257e-05,
"loss": 0.0458,
"step": 5890
},
{
"epoch": 22.692307692307693,
"grad_norm": 1.890351414680481,
"learning_rate": 2.5342097718237262e-05,
"loss": 0.0492,
"step": 5900
},
{
"epoch": 22.73076923076923,
"grad_norm": 1.6278502941131592,
"learning_rate": 2.5324119188048567e-05,
"loss": 0.0476,
"step": 5910
},
{
"epoch": 22.76923076923077,
"grad_norm": 2.3576247692108154,
"learning_rate": 2.530611243211224e-05,
"loss": 0.0477,
"step": 5920
},
{
"epoch": 22.807692307692307,
"grad_norm": 1.9282854795455933,
"learning_rate": 2.5288077499658064e-05,
"loss": 0.0442,
"step": 5930
},
{
"epoch": 22.846153846153847,
"grad_norm": 1.774977684020996,
"learning_rate": 2.527001443999285e-05,
"loss": 0.0466,
"step": 5940
},
{
"epoch": 22.884615384615383,
"grad_norm": 1.9546281099319458,
"learning_rate": 2.5251923302500318e-05,
"loss": 0.0446,
"step": 5950
},
{
"epoch": 22.923076923076923,
"grad_norm": 2.2042019367218018,
"learning_rate": 2.523380413664095e-05,
"loss": 0.0459,
"step": 5960
},
{
"epoch": 22.96153846153846,
"grad_norm": 1.848617672920227,
"learning_rate": 2.5215656991951844e-05,
"loss": 0.0433,
"step": 5970
},
{
"epoch": 23.0,
"grad_norm": 2.180008888244629,
"learning_rate": 2.5197481918046606e-05,
"loss": 0.0476,
"step": 5980
},
{
"epoch": 23.03846153846154,
"grad_norm": 2.211040735244751,
"learning_rate": 2.5179278964615192e-05,
"loss": 0.0448,
"step": 5990
},
{
"epoch": 23.076923076923077,
"grad_norm": 2.2788994312286377,
"learning_rate": 2.516104818142379e-05,
"loss": 0.0494,
"step": 6000
},
{
"epoch": 23.115384615384617,
"grad_norm": 1.680962324142456,
"learning_rate": 2.5142789618314654e-05,
"loss": 0.0409,
"step": 6010
},
{
"epoch": 23.153846153846153,
"grad_norm": 1.8511399030685425,
"learning_rate": 2.5124503325206006e-05,
"loss": 0.0469,
"step": 6020
},
{
"epoch": 23.192307692307693,
"grad_norm": 2.0306243896484375,
"learning_rate": 2.5106189352091867e-05,
"loss": 0.0427,
"step": 6030
},
{
"epoch": 23.23076923076923,
"grad_norm": 1.6079951524734497,
"learning_rate": 2.5087847749041944e-05,
"loss": 0.0501,
"step": 6040
},
{
"epoch": 23.26923076923077,
"grad_norm": 1.870809555053711,
"learning_rate": 2.506947856620148e-05,
"loss": 0.04,
"step": 6050
},
{
"epoch": 23.307692307692307,
"grad_norm": 1.878944993019104,
"learning_rate": 2.505108185379111e-05,
"loss": 0.0463,
"step": 6060
},
{
"epoch": 23.346153846153847,
"grad_norm": 2.2728562355041504,
"learning_rate": 2.503265766210676e-05,
"loss": 0.0436,
"step": 6070
},
{
"epoch": 23.384615384615383,
"grad_norm": 2.2612860202789307,
"learning_rate": 2.5014206041519456e-05,
"loss": 0.0517,
"step": 6080
},
{
"epoch": 23.423076923076923,
"grad_norm": 1.6607364416122437,
"learning_rate": 2.499572704247523e-05,
"loss": 0.0456,
"step": 6090
},
{
"epoch": 23.46153846153846,
"grad_norm": 2.1083016395568848,
"learning_rate": 2.497722071549495e-05,
"loss": 0.0414,
"step": 6100
},
{
"epoch": 23.5,
"grad_norm": 1.410343050956726,
"learning_rate": 2.4958687111174216e-05,
"loss": 0.0413,
"step": 6110
},
{
"epoch": 23.53846153846154,
"grad_norm": 1.9180458784103394,
"learning_rate": 2.494012628018319e-05,
"loss": 0.0479,
"step": 6120
},
{
"epoch": 23.576923076923077,
"grad_norm": 1.7604724168777466,
"learning_rate": 2.4921538273266475e-05,
"loss": 0.0486,
"step": 6130
},
{
"epoch": 23.615384615384617,
"grad_norm": 2.1101067066192627,
"learning_rate": 2.490292314124298e-05,
"loss": 0.0413,
"step": 6140
},
{
"epoch": 23.653846153846153,
"grad_norm": 1.9359272718429565,
"learning_rate": 2.4884280935005755e-05,
"loss": 0.0396,
"step": 6150
},
{
"epoch": 23.692307692307693,
"grad_norm": 1.9525443315505981,
"learning_rate": 2.486561170552188e-05,
"loss": 0.0503,
"step": 6160
},
{
"epoch": 23.73076923076923,
"grad_norm": 2.2428786754608154,
"learning_rate": 2.4846915503832326e-05,
"loss": 0.05,
"step": 6170
},
{
"epoch": 23.76923076923077,
"grad_norm": 2.0034830570220947,
"learning_rate": 2.4828192381051787e-05,
"loss": 0.0492,
"step": 6180
},
{
"epoch": 23.807692307692307,
"grad_norm": 2.311720132827759,
"learning_rate": 2.480944238836857e-05,
"loss": 0.0531,
"step": 6190
},
{
"epoch": 23.846153846153847,
"grad_norm": 1.5703109502792358,
"learning_rate": 2.4790665577044428e-05,
"loss": 0.0453,
"step": 6200
},
{
"epoch": 23.884615384615383,
"grad_norm": 1.7515316009521484,
"learning_rate": 2.4771861998414458e-05,
"loss": 0.046,
"step": 6210
},
{
"epoch": 23.923076923076923,
"grad_norm": 2.0526859760284424,
"learning_rate": 2.475303170388692e-05,
"loss": 0.0445,
"step": 6220
},
{
"epoch": 23.96153846153846,
"grad_norm": 1.8120594024658203,
"learning_rate": 2.4734174744943122e-05,
"loss": 0.0451,
"step": 6230
},
{
"epoch": 24.0,
"grad_norm": 2.8347058296203613,
"learning_rate": 2.471529117313727e-05,
"loss": 0.0456,
"step": 6240
},
{
"epoch": 24.03846153846154,
"grad_norm": 2.466557502746582,
"learning_rate": 2.4696381040096335e-05,
"loss": 0.0481,
"step": 6250
},
{
"epoch": 24.076923076923077,
"grad_norm": 2.347093105316162,
"learning_rate": 2.4677444397519883e-05,
"loss": 0.0475,
"step": 6260
},
{
"epoch": 24.115384615384617,
"grad_norm": 1.7908564805984497,
"learning_rate": 2.4658481297179987e-05,
"loss": 0.0468,
"step": 6270
},
{
"epoch": 24.153846153846153,
"grad_norm": 2.0475430488586426,
"learning_rate": 2.4639491790921028e-05,
"loss": 0.0474,
"step": 6280
},
{
"epoch": 24.192307692307693,
"grad_norm": 2.225431442260742,
"learning_rate": 2.4620475930659596e-05,
"loss": 0.0488,
"step": 6290
},
{
"epoch": 24.23076923076923,
"grad_norm": 1.9803553819656372,
"learning_rate": 2.4601433768384327e-05,
"loss": 0.0421,
"step": 6300
},
{
"epoch": 24.26923076923077,
"grad_norm": 2.0825839042663574,
"learning_rate": 2.4582365356155766e-05,
"loss": 0.0452,
"step": 6310
},
{
"epoch": 24.307692307692307,
"grad_norm": 1.9435968399047852,
"learning_rate": 2.4563270746106224e-05,
"loss": 0.0423,
"step": 6320
},
{
"epoch": 24.346153846153847,
"grad_norm": 2.0983896255493164,
"learning_rate": 2.4544149990439632e-05,
"loss": 0.0431,
"step": 6330
},
{
"epoch": 24.384615384615383,
"grad_norm": 1.7998305559158325,
"learning_rate": 2.4525003141431413e-05,
"loss": 0.0434,
"step": 6340
},
{
"epoch": 24.423076923076923,
"grad_norm": 2.194087028503418,
"learning_rate": 2.450583025142831e-05,
"loss": 0.0406,
"step": 6350
},
{
"epoch": 24.46153846153846,
"grad_norm": 2.305424690246582,
"learning_rate": 2.4486631372848286e-05,
"loss": 0.0461,
"step": 6360
},
{
"epoch": 24.5,
"grad_norm": 2.211747646331787,
"learning_rate": 2.4467406558180328e-05,
"loss": 0.0454,
"step": 6370
},
{
"epoch": 24.53846153846154,
"grad_norm": 2.2912757396698,
"learning_rate": 2.4448155859984357e-05,
"loss": 0.0425,
"step": 6380
},
{
"epoch": 24.576923076923077,
"grad_norm": 1.930274248123169,
"learning_rate": 2.442887933089104e-05,
"loss": 0.0495,
"step": 6390
},
{
"epoch": 24.615384615384617,
"grad_norm": 2.166144609451294,
"learning_rate": 2.440957702360167e-05,
"loss": 0.0453,
"step": 6400
},
{
"epoch": 24.653846153846153,
"grad_norm": 1.9380121231079102,
"learning_rate": 2.4390248990888026e-05,
"loss": 0.0467,
"step": 6410
},
{
"epoch": 24.692307692307693,
"grad_norm": 1.9483615159988403,
"learning_rate": 2.4370895285592202e-05,
"loss": 0.0461,
"step": 6420
},
{
"epoch": 24.73076923076923,
"grad_norm": 1.7530697584152222,
"learning_rate": 2.43515159606265e-05,
"loss": 0.0404,
"step": 6430
},
{
"epoch": 24.76923076923077,
"grad_norm": 2.1312994956970215,
"learning_rate": 2.4332111068973243e-05,
"loss": 0.0447,
"step": 6440
},
{
"epoch": 24.807692307692307,
"grad_norm": 1.7659395933151245,
"learning_rate": 2.4312680663684674e-05,
"loss": 0.0409,
"step": 6450
},
{
"epoch": 24.846153846153847,
"grad_norm": 1.9349921941757202,
"learning_rate": 2.429322479788277e-05,
"loss": 0.042,
"step": 6460
},
{
"epoch": 24.884615384615383,
"grad_norm": 1.856867790222168,
"learning_rate": 2.4273743524759132e-05,
"loss": 0.0494,
"step": 6470
},
{
"epoch": 24.923076923076923,
"grad_norm": 2.1651554107666016,
"learning_rate": 2.4254236897574818e-05,
"loss": 0.0423,
"step": 6480
},
{
"epoch": 24.96153846153846,
"grad_norm": 1.993500828742981,
"learning_rate": 2.4234704969660192e-05,
"loss": 0.0415,
"step": 6490
},
{
"epoch": 25.0,
"grad_norm": 1.9535496234893799,
"learning_rate": 2.4215147794414806e-05,
"loss": 0.0425,
"step": 6500
},
{
"epoch": 25.03846153846154,
"grad_norm": 1.6581015586853027,
"learning_rate": 2.419556542530723e-05,
"loss": 0.0461,
"step": 6510
},
{
"epoch": 25.076923076923077,
"grad_norm": 1.8992817401885986,
"learning_rate": 2.4175957915874916e-05,
"loss": 0.0488,
"step": 6520
},
{
"epoch": 25.115384615384617,
"grad_norm": 1.6976550817489624,
"learning_rate": 2.4156325319724037e-05,
"loss": 0.0471,
"step": 6530
},
{
"epoch": 25.153846153846153,
"grad_norm": 1.9944517612457275,
"learning_rate": 2.4136667690529372e-05,
"loss": 0.0429,
"step": 6540
},
{
"epoch": 25.192307692307693,
"grad_norm": 1.7885385751724243,
"learning_rate": 2.4116985082034126e-05,
"loss": 0.0444,
"step": 6550
},
{
"epoch": 25.23076923076923,
"grad_norm": 1.9043548107147217,
"learning_rate": 2.409727754804979e-05,
"loss": 0.0483,
"step": 6560
},
{
"epoch": 25.26923076923077,
"grad_norm": 1.5960673093795776,
"learning_rate": 2.4077545142456025e-05,
"loss": 0.0481,
"step": 6570
},
{
"epoch": 25.307692307692307,
"grad_norm": 1.7031124830245972,
"learning_rate": 2.405778791920046e-05,
"loss": 0.0569,
"step": 6580
},
{
"epoch": 25.346153846153847,
"grad_norm": 2.0520358085632324,
"learning_rate": 2.4038005932298594e-05,
"loss": 0.0444,
"step": 6590
},
{
"epoch": 25.384615384615383,
"grad_norm": 2.1107587814331055,
"learning_rate": 2.4018199235833624e-05,
"loss": 0.0429,
"step": 6600
},
{
"epoch": 25.423076923076923,
"grad_norm": 1.6484862565994263,
"learning_rate": 2.3998367883956306e-05,
"loss": 0.0463,
"step": 6610
},
{
"epoch": 25.46153846153846,
"grad_norm": 1.5690255165100098,
"learning_rate": 2.3978511930884795e-05,
"loss": 0.0423,
"step": 6620
},
{
"epoch": 25.5,
"grad_norm": 1.9984314441680908,
"learning_rate": 2.3958631430904504e-05,
"loss": 0.0494,
"step": 6630
},
{
"epoch": 25.53846153846154,
"grad_norm": 1.9124306440353394,
"learning_rate": 2.393872643836797e-05,
"loss": 0.0469,
"step": 6640
},
{
"epoch": 25.576923076923077,
"grad_norm": 1.6842821836471558,
"learning_rate": 2.3918797007694675e-05,
"loss": 0.041,
"step": 6650
},
{
"epoch": 25.615384615384617,
"grad_norm": 2.1914467811584473,
"learning_rate": 2.3898843193370923e-05,
"loss": 0.0452,
"step": 6660
},
{
"epoch": 25.653846153846153,
"grad_norm": 1.7262556552886963,
"learning_rate": 2.387886504994969e-05,
"loss": 0.0436,
"step": 6670
},
{
"epoch": 25.692307692307693,
"grad_norm": 2.0586323738098145,
"learning_rate": 2.385886263205044e-05,
"loss": 0.0465,
"step": 6680
},
{
"epoch": 25.73076923076923,
"grad_norm": 1.9069286584854126,
"learning_rate": 2.3838835994359036e-05,
"loss": 0.0443,
"step": 6690
},
{
"epoch": 25.76923076923077,
"grad_norm": 2.2853593826293945,
"learning_rate": 2.3818785191627525e-05,
"loss": 0.0445,
"step": 6700
},
{
"epoch": 25.807692307692307,
"grad_norm": 1.947430968284607,
"learning_rate": 2.379871027867405e-05,
"loss": 0.0449,
"step": 6710
},
{
"epoch": 25.846153846153847,
"grad_norm": 1.848595142364502,
"learning_rate": 2.3778611310382653e-05,
"loss": 0.0433,
"step": 6720
},
{
"epoch": 25.884615384615383,
"grad_norm": 2.047389507293701,
"learning_rate": 2.3758488341703137e-05,
"loss": 0.0451,
"step": 6730
},
{
"epoch": 25.923076923076923,
"grad_norm": 1.6822788715362549,
"learning_rate": 2.3738341427650945e-05,
"loss": 0.0504,
"step": 6740
},
{
"epoch": 25.96153846153846,
"grad_norm": 1.673630714416504,
"learning_rate": 2.3718170623306955e-05,
"loss": 0.0422,
"step": 6750
},
{
"epoch": 26.0,
"grad_norm": 1.6114610433578491,
"learning_rate": 2.369797598381739e-05,
"loss": 0.041,
"step": 6760
},
{
"epoch": 26.03846153846154,
"grad_norm": 1.9758553504943848,
"learning_rate": 2.3677757564393612e-05,
"loss": 0.0398,
"step": 6770
},
{
"epoch": 26.076923076923077,
"grad_norm": 1.816084623336792,
"learning_rate": 2.3657515420312015e-05,
"loss": 0.0408,
"step": 6780
},
{
"epoch": 26.115384615384617,
"grad_norm": 1.7615054845809937,
"learning_rate": 2.3637249606913847e-05,
"loss": 0.0445,
"step": 6790
},
{
"epoch": 26.153846153846153,
"grad_norm": 2.0626277923583984,
"learning_rate": 2.3616960179605064e-05,
"loss": 0.0527,
"step": 6800
},
{
"epoch": 26.192307692307693,
"grad_norm": 1.707031011581421,
"learning_rate": 2.3596647193856188e-05,
"loss": 0.04,
"step": 6810
},
{
"epoch": 26.23076923076923,
"grad_norm": 1.8059414625167847,
"learning_rate": 2.3576310705202143e-05,
"loss": 0.0394,
"step": 6820
},
{
"epoch": 26.26923076923077,
"grad_norm": 2.0285634994506836,
"learning_rate": 2.3555950769242122e-05,
"loss": 0.037,
"step": 6830
},
{
"epoch": 26.307692307692307,
"grad_norm": 1.6764180660247803,
"learning_rate": 2.3535567441639396e-05,
"loss": 0.041,
"step": 6840
},
{
"epoch": 26.346153846153847,
"grad_norm": 1.839801549911499,
"learning_rate": 2.351516077812122e-05,
"loss": 0.0385,
"step": 6850
},
{
"epoch": 26.384615384615383,
"grad_norm": 2.2130534648895264,
"learning_rate": 2.349473083447863e-05,
"loss": 0.0481,
"step": 6860
},
{
"epoch": 26.423076923076923,
"grad_norm": 1.7864869832992554,
"learning_rate": 2.3474277666566307e-05,
"loss": 0.0495,
"step": 6870
},
{
"epoch": 26.46153846153846,
"grad_norm": 1.2948951721191406,
"learning_rate": 2.345380133030243e-05,
"loss": 0.0503,
"step": 6880
},
{
"epoch": 26.5,
"grad_norm": 1.299521803855896,
"learning_rate": 2.343330188166853e-05,
"loss": 0.0441,
"step": 6890
},
{
"epoch": 26.53846153846154,
"grad_norm": 2.135817289352417,
"learning_rate": 2.3412779376709304e-05,
"loss": 0.0435,
"step": 6900
},
{
"epoch": 26.576923076923077,
"grad_norm": 1.7558770179748535,
"learning_rate": 2.3392233871532504e-05,
"loss": 0.0471,
"step": 6910
},
{
"epoch": 26.615384615384617,
"grad_norm": 2.0203568935394287,
"learning_rate": 2.337166542230876e-05,
"loss": 0.0517,
"step": 6920
},
{
"epoch": 26.653846153846153,
"grad_norm": 2.188866138458252,
"learning_rate": 2.335107408527142e-05,
"loss": 0.0398,
"step": 6930
},
{
"epoch": 26.692307692307693,
"grad_norm": 1.6940990686416626,
"learning_rate": 2.3330459916716417e-05,
"loss": 0.0392,
"step": 6940
},
{
"epoch": 26.73076923076923,
"grad_norm": 1.8685543537139893,
"learning_rate": 2.3309822973002097e-05,
"loss": 0.0482,
"step": 6950
},
{
"epoch": 26.76923076923077,
"grad_norm": 1.6639153957366943,
"learning_rate": 2.328916331054908e-05,
"loss": 0.044,
"step": 6960
},
{
"epoch": 26.807692307692307,
"grad_norm": 1.8613269329071045,
"learning_rate": 2.3268480985840093e-05,
"loss": 0.0431,
"step": 6970
},
{
"epoch": 26.846153846153847,
"grad_norm": 1.5119969844818115,
"learning_rate": 2.3247776055419826e-05,
"loss": 0.0422,
"step": 6980
},
{
"epoch": 26.884615384615383,
"grad_norm": 1.6979011297225952,
"learning_rate": 2.3227048575894758e-05,
"loss": 0.0446,
"step": 6990
},
{
"epoch": 26.923076923076923,
"grad_norm": 2.1201868057250977,
"learning_rate": 2.3206298603933037e-05,
"loss": 0.0394,
"step": 7000
},
{
"epoch": 26.96153846153846,
"grad_norm": 1.884023666381836,
"learning_rate": 2.3185526196264288e-05,
"loss": 0.0433,
"step": 7010
},
{
"epoch": 27.0,
"grad_norm": 1.637145757675171,
"learning_rate": 2.3164731409679476e-05,
"loss": 0.0359,
"step": 7020
},
{
"epoch": 27.03846153846154,
"grad_norm": 1.3160382509231567,
"learning_rate": 2.3143914301030765e-05,
"loss": 0.0439,
"step": 7030
},
{
"epoch": 27.076923076923077,
"grad_norm": 1.8100475072860718,
"learning_rate": 2.3123074927231332e-05,
"loss": 0.0452,
"step": 7040
},
{
"epoch": 27.115384615384617,
"grad_norm": 2.004098653793335,
"learning_rate": 2.310221334525522e-05,
"loss": 0.0513,
"step": 7050
},
{
"epoch": 27.153846153846153,
"grad_norm": 2.4186182022094727,
"learning_rate": 2.3081329612137207e-05,
"loss": 0.0411,
"step": 7060
},
{
"epoch": 27.192307692307693,
"grad_norm": 1.9002735614776611,
"learning_rate": 2.3060423784972625e-05,
"loss": 0.0456,
"step": 7070
},
{
"epoch": 27.23076923076923,
"grad_norm": 2.0039029121398926,
"learning_rate": 2.3039495920917193e-05,
"loss": 0.0395,
"step": 7080
},
{
"epoch": 27.26923076923077,
"grad_norm": 1.979544758796692,
"learning_rate": 2.301854607718691e-05,
"loss": 0.039,
"step": 7090
},
{
"epoch": 27.307692307692307,
"grad_norm": 1.5195621252059937,
"learning_rate": 2.299757431105783e-05,
"loss": 0.0426,
"step": 7100
},
{
"epoch": 27.346153846153847,
"grad_norm": 1.6798497438430786,
"learning_rate": 2.2976580679865972e-05,
"loss": 0.0433,
"step": 7110
},
{
"epoch": 27.384615384615383,
"grad_norm": 1.4978078603744507,
"learning_rate": 2.2955565241007123e-05,
"loss": 0.045,
"step": 7120
},
{
"epoch": 27.423076923076923,
"grad_norm": 1.5086406469345093,
"learning_rate": 2.293452805193669e-05,
"loss": 0.0403,
"step": 7130
},
{
"epoch": 27.46153846153846,
"grad_norm": 1.614099144935608,
"learning_rate": 2.291346917016954e-05,
"loss": 0.0455,
"step": 7140
},
{
"epoch": 27.5,
"grad_norm": 1.391572117805481,
"learning_rate": 2.289238865327985e-05,
"loss": 0.0446,
"step": 7150
},
{
"epoch": 27.53846153846154,
"grad_norm": 1.4410821199417114,
"learning_rate": 2.2871286558900956e-05,
"loss": 0.0456,
"step": 7160
},
{
"epoch": 27.576923076923077,
"grad_norm": 1.736544132232666,
"learning_rate": 2.285016294472517e-05,
"loss": 0.0448,
"step": 7170
},
{
"epoch": 27.615384615384617,
"grad_norm": 1.6949403285980225,
"learning_rate": 2.2829017868503658e-05,
"loss": 0.0417,
"step": 7180
},
{
"epoch": 27.653846153846153,
"grad_norm": 1.5383449792861938,
"learning_rate": 2.280785138804624e-05,
"loss": 0.0511,
"step": 7190
},
{
"epoch": 27.692307692307693,
"grad_norm": 1.8487229347229004,
"learning_rate": 2.2786663561221265e-05,
"loss": 0.0414,
"step": 7200
},
{
"epoch": 27.73076923076923,
"grad_norm": 1.7891006469726562,
"learning_rate": 2.2765454445955452e-05,
"loss": 0.0377,
"step": 7210
},
{
"epoch": 27.76923076923077,
"grad_norm": 1.737947940826416,
"learning_rate": 2.2744224100233705e-05,
"loss": 0.0492,
"step": 7220
},
{
"epoch": 27.807692307692307,
"grad_norm": 1.625235915184021,
"learning_rate": 2.2722972582098984e-05,
"loss": 0.0412,
"step": 7230
},
{
"epoch": 27.846153846153847,
"grad_norm": 1.8091816902160645,
"learning_rate": 2.2701699949652118e-05,
"loss": 0.0417,
"step": 7240
},
{
"epoch": 27.884615384615383,
"grad_norm": 2.394031047821045,
"learning_rate": 2.2680406261051685e-05,
"loss": 0.0388,
"step": 7250
},
{
"epoch": 27.923076923076923,
"grad_norm": 1.3932080268859863,
"learning_rate": 2.2659091574513805e-05,
"loss": 0.0438,
"step": 7260
},
{
"epoch": 27.96153846153846,
"grad_norm": 1.760535717010498,
"learning_rate": 2.263775594831202e-05,
"loss": 0.0432,
"step": 7270
},
{
"epoch": 28.0,
"grad_norm": 1.6721904277801514,
"learning_rate": 2.2616399440777128e-05,
"loss": 0.0384,
"step": 7280
},
{
"epoch": 28.03846153846154,
"grad_norm": 1.8300305604934692,
"learning_rate": 2.2595022110296988e-05,
"loss": 0.0415,
"step": 7290
},
{
"epoch": 28.076923076923077,
"grad_norm": 1.8963747024536133,
"learning_rate": 2.2573624015316418e-05,
"loss": 0.0457,
"step": 7300
},
{
"epoch": 28.115384615384617,
"grad_norm": 1.676701545715332,
"learning_rate": 2.2552205214336986e-05,
"loss": 0.0448,
"step": 7310
},
{
"epoch": 28.153846153846153,
"grad_norm": 1.6721583604812622,
"learning_rate": 2.253076576591688e-05,
"loss": 0.0415,
"step": 7320
},
{
"epoch": 28.192307692307693,
"grad_norm": 2.002814531326294,
"learning_rate": 2.2509305728670733e-05,
"loss": 0.0462,
"step": 7330
},
{
"epoch": 28.23076923076923,
"grad_norm": 1.5034431219100952,
"learning_rate": 2.2487825161269463e-05,
"loss": 0.0416,
"step": 7340
},
{
"epoch": 28.26923076923077,
"grad_norm": 2.0287587642669678,
"learning_rate": 2.2466324122440125e-05,
"loss": 0.0401,
"step": 7350
},
{
"epoch": 28.307692307692307,
"grad_norm": 1.6312346458435059,
"learning_rate": 2.2444802670965732e-05,
"loss": 0.0403,
"step": 7360
},
{
"epoch": 28.346153846153847,
"grad_norm": 1.8766647577285767,
"learning_rate": 2.2423260865685124e-05,
"loss": 0.0404,
"step": 7370
},
{
"epoch": 28.384615384615383,
"grad_norm": 1.6648188829421997,
"learning_rate": 2.2401698765492762e-05,
"loss": 0.0437,
"step": 7380
},
{
"epoch": 28.423076923076923,
"grad_norm": 1.7376813888549805,
"learning_rate": 2.2380116429338612e-05,
"loss": 0.0443,
"step": 7390
},
{
"epoch": 28.46153846153846,
"grad_norm": 1.6222645044326782,
"learning_rate": 2.2358513916227945e-05,
"loss": 0.0414,
"step": 7400
},
{
"epoch": 28.5,
"grad_norm": 1.5098398923873901,
"learning_rate": 2.233689128522122e-05,
"loss": 0.0369,
"step": 7410
},
{
"epoch": 28.53846153846154,
"grad_norm": 1.603192925453186,
"learning_rate": 2.2315248595433883e-05,
"loss": 0.0457,
"step": 7420
},
{
"epoch": 28.576923076923077,
"grad_norm": 1.890964388847351,
"learning_rate": 2.2293585906036214e-05,
"loss": 0.0427,
"step": 7430
},
{
"epoch": 28.615384615384617,
"grad_norm": 1.8507992029190063,
"learning_rate": 2.2271903276253183e-05,
"loss": 0.0426,
"step": 7440
},
{
"epoch": 28.653846153846153,
"grad_norm": 1.867964744567871,
"learning_rate": 2.2250200765364273e-05,
"loss": 0.0391,
"step": 7450
},
{
"epoch": 28.692307692307693,
"grad_norm": 1.5608159303665161,
"learning_rate": 2.2228478432703317e-05,
"loss": 0.043,
"step": 7460
},
{
"epoch": 28.73076923076923,
"grad_norm": 1.7745044231414795,
"learning_rate": 2.2206736337658348e-05,
"loss": 0.0421,
"step": 7470
},
{
"epoch": 28.76923076923077,
"grad_norm": 1.6739052534103394,
"learning_rate": 2.2184974539671417e-05,
"loss": 0.0419,
"step": 7480
},
{
"epoch": 28.807692307692307,
"grad_norm": 2.010806083679199,
"learning_rate": 2.2163193098238453e-05,
"loss": 0.0465,
"step": 7490
},
{
"epoch": 28.846153846153847,
"grad_norm": 1.8989545106887817,
"learning_rate": 2.2141392072909082e-05,
"loss": 0.0423,
"step": 7500
},
{
"epoch": 28.884615384615383,
"grad_norm": 1.8697062730789185,
"learning_rate": 2.2119571523286484e-05,
"loss": 0.0371,
"step": 7510
},
{
"epoch": 28.923076923076923,
"grad_norm": 1.5613651275634766,
"learning_rate": 2.2097731509027196e-05,
"loss": 0.0444,
"step": 7520
},
{
"epoch": 28.96153846153846,
"grad_norm": 1.36122465133667,
"learning_rate": 2.207587208984099e-05,
"loss": 0.0442,
"step": 7530
},
{
"epoch": 29.0,
"grad_norm": 1.5111128091812134,
"learning_rate": 2.205399332549068e-05,
"loss": 0.0383,
"step": 7540
},
{
"epoch": 29.03846153846154,
"grad_norm": 1.6100459098815918,
"learning_rate": 2.2032095275791974e-05,
"loss": 0.0403,
"step": 7550
},
{
"epoch": 29.076923076923077,
"grad_norm": 1.7022507190704346,
"learning_rate": 2.2010178000613307e-05,
"loss": 0.0408,
"step": 7560
},
{
"epoch": 29.115384615384617,
"grad_norm": 1.8633601665496826,
"learning_rate": 2.1988241559875666e-05,
"loss": 0.0422,
"step": 7570
},
{
"epoch": 29.153846153846153,
"grad_norm": 1.7628028392791748,
"learning_rate": 2.1966286013552448e-05,
"loss": 0.0476,
"step": 7580
},
{
"epoch": 29.192307692307693,
"grad_norm": 1.9117072820663452,
"learning_rate": 2.1944311421669274e-05,
"loss": 0.0436,
"step": 7590
},
{
"epoch": 29.23076923076923,
"grad_norm": 1.4103853702545166,
"learning_rate": 2.1922317844303846e-05,
"loss": 0.0456,
"step": 7600
},
{
"epoch": 29.26923076923077,
"grad_norm": 1.4408249855041504,
"learning_rate": 2.1900305341585756e-05,
"loss": 0.037,
"step": 7610
},
{
"epoch": 29.307692307692307,
"grad_norm": 1.3809062242507935,
"learning_rate": 2.187827397369635e-05,
"loss": 0.0385,
"step": 7620
},
{
"epoch": 29.346153846153847,
"grad_norm": 1.9578503370285034,
"learning_rate": 2.1856223800868542e-05,
"loss": 0.041,
"step": 7630
},
{
"epoch": 29.384615384615383,
"grad_norm": 1.955604910850525,
"learning_rate": 2.183415488338667e-05,
"loss": 0.0424,
"step": 7640
},
{
"epoch": 29.423076923076923,
"grad_norm": 1.5779365301132202,
"learning_rate": 2.1812067281586312e-05,
"loss": 0.0444,
"step": 7650
},
{
"epoch": 29.46153846153846,
"grad_norm": 1.7006980180740356,
"learning_rate": 2.178996105585412e-05,
"loss": 0.0413,
"step": 7660
},
{
"epoch": 29.5,
"grad_norm": 1.7158124446868896,
"learning_rate": 2.1767836266627676e-05,
"loss": 0.0418,
"step": 7670
},
{
"epoch": 29.53846153846154,
"grad_norm": 1.5377243757247925,
"learning_rate": 2.174569297439531e-05,
"loss": 0.0379,
"step": 7680
},
{
"epoch": 29.576923076923077,
"grad_norm": 1.7341413497924805,
"learning_rate": 2.1723531239695932e-05,
"loss": 0.0411,
"step": 7690
},
{
"epoch": 29.615384615384617,
"grad_norm": 1.6391249895095825,
"learning_rate": 2.1701351123118886e-05,
"loss": 0.0403,
"step": 7700
},
{
"epoch": 29.653846153846153,
"grad_norm": 1.8113683462142944,
"learning_rate": 2.167915268530376e-05,
"loss": 0.0433,
"step": 7710
},
{
"epoch": 29.692307692307693,
"grad_norm": 1.8385941982269287,
"learning_rate": 2.165693598694023e-05,
"loss": 0.0396,
"step": 7720
},
{
"epoch": 29.73076923076923,
"grad_norm": 1.5339360237121582,
"learning_rate": 2.163470108876791e-05,
"loss": 0.0473,
"step": 7730
},
{
"epoch": 29.76923076923077,
"grad_norm": 1.451357126235962,
"learning_rate": 2.161244805157616e-05,
"loss": 0.0363,
"step": 7740
},
{
"epoch": 29.807692307692307,
"grad_norm": 1.3908246755599976,
"learning_rate": 2.159017693620393e-05,
"loss": 0.0415,
"step": 7750
},
{
"epoch": 29.846153846153847,
"grad_norm": 1.707660436630249,
"learning_rate": 2.15678878035396e-05,
"loss": 0.0398,
"step": 7760
},
{
"epoch": 29.884615384615383,
"grad_norm": 1.7854512929916382,
"learning_rate": 2.1545580714520817e-05,
"loss": 0.0355,
"step": 7770
},
{
"epoch": 29.923076923076923,
"grad_norm": 1.6504265069961548,
"learning_rate": 2.1523255730134294e-05,
"loss": 0.0428,
"step": 7780
},
{
"epoch": 29.96153846153846,
"grad_norm": 1.5774441957473755,
"learning_rate": 2.15009129114157e-05,
"loss": 0.0371,
"step": 7790
},
{
"epoch": 30.0,
"grad_norm": 1.6256803274154663,
"learning_rate": 2.1478552319449443e-05,
"loss": 0.0405,
"step": 7800
},
{
"epoch": 30.03846153846154,
"grad_norm": 1.8955668210983276,
"learning_rate": 2.1456174015368527e-05,
"loss": 0.0426,
"step": 7810
},
{
"epoch": 30.076923076923077,
"grad_norm": 1.6124283075332642,
"learning_rate": 2.1433778060354375e-05,
"loss": 0.0466,
"step": 7820
},
{
"epoch": 30.115384615384617,
"grad_norm": 1.5108370780944824,
"learning_rate": 2.1411364515636685e-05,
"loss": 0.0513,
"step": 7830
},
{
"epoch": 30.153846153846153,
"grad_norm": 1.566898226737976,
"learning_rate": 2.1388933442493232e-05,
"loss": 0.0437,
"step": 7840
},
{
"epoch": 30.192307692307693,
"grad_norm": 1.4218852519989014,
"learning_rate": 2.13664849022497e-05,
"loss": 0.0394,
"step": 7850
},
{
"epoch": 30.23076923076923,
"grad_norm": 1.6596649885177612,
"learning_rate": 2.1344018956279547e-05,
"loss": 0.0365,
"step": 7860
},
{
"epoch": 30.26923076923077,
"grad_norm": 1.4985148906707764,
"learning_rate": 2.1321535666003817e-05,
"loss": 0.0389,
"step": 7870
},
{
"epoch": 30.307692307692307,
"grad_norm": 1.470139980316162,
"learning_rate": 2.1299035092890966e-05,
"loss": 0.0446,
"step": 7880
},
{
"epoch": 30.346153846153847,
"grad_norm": 1.7060062885284424,
"learning_rate": 2.12765172984567e-05,
"loss": 0.0415,
"step": 7890
},
{
"epoch": 30.384615384615383,
"grad_norm": 1.4624581336975098,
"learning_rate": 2.1253982344263803e-05,
"loss": 0.0375,
"step": 7900
},
{
"epoch": 30.423076923076923,
"grad_norm": 1.9595462083816528,
"learning_rate": 2.1231430291921987e-05,
"loss": 0.0461,
"step": 7910
},
{
"epoch": 30.46153846153846,
"grad_norm": 1.7338889837265015,
"learning_rate": 2.1208861203087695e-05,
"loss": 0.0406,
"step": 7920
},
{
"epoch": 30.5,
"grad_norm": 1.601374864578247,
"learning_rate": 2.1186275139463967e-05,
"loss": 0.0418,
"step": 7930
},
{
"epoch": 30.53846153846154,
"grad_norm": 1.823254108428955,
"learning_rate": 2.1163672162800222e-05,
"loss": 0.0403,
"step": 7940
},
{
"epoch": 30.576923076923077,
"grad_norm": 1.9945504665374756,
"learning_rate": 2.114105233489215e-05,
"loss": 0.0508,
"step": 7950
},
{
"epoch": 30.615384615384617,
"grad_norm": 1.810983657836914,
"learning_rate": 2.1118415717581487e-05,
"loss": 0.0509,
"step": 7960
},
{
"epoch": 30.653846153846153,
"grad_norm": 1.639147400856018,
"learning_rate": 2.1095762372755885e-05,
"loss": 0.043,
"step": 7970
},
{
"epoch": 30.692307692307693,
"grad_norm": 1.810479998588562,
"learning_rate": 2.1073092362348716e-05,
"loss": 0.0496,
"step": 7980
},
{
"epoch": 30.73076923076923,
"grad_norm": 1.596810221672058,
"learning_rate": 2.1050405748338933e-05,
"loss": 0.0423,
"step": 7990
},
{
"epoch": 30.76923076923077,
"grad_norm": 1.752320408821106,
"learning_rate": 2.102770259275087e-05,
"loss": 0.0391,
"step": 8000
},
{
"epoch": 30.807692307692307,
"grad_norm": 1.5737982988357544,
"learning_rate": 2.100498295765408e-05,
"loss": 0.0447,
"step": 8010
},
{
"epoch": 30.846153846153847,
"grad_norm": 1.6139024496078491,
"learning_rate": 2.098224690516319e-05,
"loss": 0.0404,
"step": 8020
},
{
"epoch": 30.884615384615383,
"grad_norm": 1.4000976085662842,
"learning_rate": 2.0959494497437688e-05,
"loss": 0.0424,
"step": 8030
},
{
"epoch": 30.923076923076923,
"grad_norm": 1.5702110528945923,
"learning_rate": 2.0936725796681796e-05,
"loss": 0.0379,
"step": 8040
},
{
"epoch": 30.96153846153846,
"grad_norm": 1.8499011993408203,
"learning_rate": 2.0913940865144266e-05,
"loss": 0.0373,
"step": 8050
},
{
"epoch": 31.0,
"grad_norm": 1.890944242477417,
"learning_rate": 2.0891139765118235e-05,
"loss": 0.0416,
"step": 8060
},
{
"epoch": 31.03846153846154,
"grad_norm": 1.509231448173523,
"learning_rate": 2.086832255894104e-05,
"loss": 0.0383,
"step": 8070
},
{
"epoch": 31.076923076923077,
"grad_norm": 1.7421714067459106,
"learning_rate": 2.084548930899405e-05,
"loss": 0.0386,
"step": 8080
},
{
"epoch": 31.115384615384617,
"grad_norm": 1.293750524520874,
"learning_rate": 2.08226400777025e-05,
"loss": 0.0365,
"step": 8090
},
{
"epoch": 31.153846153846153,
"grad_norm": 1.332698106765747,
"learning_rate": 2.0799774927535313e-05,
"loss": 0.0353,
"step": 8100
},
{
"epoch": 31.192307692307693,
"grad_norm": 1.566386342048645,
"learning_rate": 2.0776893921004936e-05,
"loss": 0.0386,
"step": 8110
},
{
"epoch": 31.23076923076923,
"grad_norm": 1.5467346906661987,
"learning_rate": 2.0753997120667172e-05,
"loss": 0.0423,
"step": 8120
},
{
"epoch": 31.26923076923077,
"grad_norm": 1.2984790802001953,
"learning_rate": 2.0731084589120995e-05,
"loss": 0.0351,
"step": 8130
},
{
"epoch": 31.307692307692307,
"grad_norm": 1.6691569089889526,
"learning_rate": 2.070815638900839e-05,
"loss": 0.0441,
"step": 8140
},
{
"epoch": 31.346153846153847,
"grad_norm": 1.67073392868042,
"learning_rate": 2.0685212583014186e-05,
"loss": 0.0408,
"step": 8150
},
{
"epoch": 31.384615384615383,
"grad_norm": 1.7874646186828613,
"learning_rate": 2.0662253233865866e-05,
"loss": 0.0401,
"step": 8160
},
{
"epoch": 31.423076923076923,
"grad_norm": 1.6154216527938843,
"learning_rate": 2.063927840433342e-05,
"loss": 0.0395,
"step": 8170
},
{
"epoch": 31.46153846153846,
"grad_norm": 1.4242808818817139,
"learning_rate": 2.0616288157229154e-05,
"loss": 0.0424,
"step": 8180
},
{
"epoch": 31.5,
"grad_norm": 1.4706525802612305,
"learning_rate": 2.0593282555407522e-05,
"loss": 0.0405,
"step": 8190
},
{
"epoch": 31.53846153846154,
"grad_norm": 1.7018382549285889,
"learning_rate": 2.057026166176496e-05,
"loss": 0.0397,
"step": 8200
},
{
"epoch": 31.576923076923077,
"grad_norm": 1.3753076791763306,
"learning_rate": 2.0547225539239715e-05,
"loss": 0.0398,
"step": 8210
},
{
"epoch": 31.615384615384617,
"grad_norm": 1.6274869441986084,
"learning_rate": 2.0524174250811665e-05,
"loss": 0.0442,
"step": 8220
},
{
"epoch": 31.653846153846153,
"grad_norm": 1.7797141075134277,
"learning_rate": 2.050110785950216e-05,
"loss": 0.041,
"step": 8230
},
{
"epoch": 31.692307692307693,
"grad_norm": 1.6109033823013306,
"learning_rate": 2.047802642837382e-05,
"loss": 0.0377,
"step": 8240
},
{
"epoch": 31.73076923076923,
"grad_norm": 1.7530934810638428,
"learning_rate": 2.0454930020530403e-05,
"loss": 0.038,
"step": 8250
},
{
"epoch": 31.76923076923077,
"grad_norm": 1.6273490190505981,
"learning_rate": 2.0431818699116606e-05,
"loss": 0.0372,
"step": 8260
},
{
"epoch": 31.807692307692307,
"grad_norm": 1.5764533281326294,
"learning_rate": 2.04086925273179e-05,
"loss": 0.0515,
"step": 8270
},
{
"epoch": 31.846153846153847,
"grad_norm": 1.7474285364151,
"learning_rate": 2.0385551568360357e-05,
"loss": 0.0389,
"step": 8280
},
{
"epoch": 31.884615384615383,
"grad_norm": 1.3446240425109863,
"learning_rate": 2.036239588551047e-05,
"loss": 0.0404,
"step": 8290
},
{
"epoch": 31.923076923076923,
"grad_norm": 1.7598605155944824,
"learning_rate": 2.0339225542074996e-05,
"loss": 0.0351,
"step": 8300
},
{
"epoch": 31.96153846153846,
"grad_norm": 1.6149331331253052,
"learning_rate": 2.0316040601400765e-05,
"loss": 0.0431,
"step": 8310
},
{
"epoch": 32.0,
"grad_norm": 2.087171792984009,
"learning_rate": 2.029284112687453e-05,
"loss": 0.0421,
"step": 8320
},
{
"epoch": 32.03846153846154,
"grad_norm": 1.3845360279083252,
"learning_rate": 2.0269627181922752e-05,
"loss": 0.0416,
"step": 8330
},
{
"epoch": 32.07692307692308,
"grad_norm": 1.316144347190857,
"learning_rate": 2.0246398830011482e-05,
"loss": 0.038,
"step": 8340
},
{
"epoch": 32.11538461538461,
"grad_norm": 1.228372573852539,
"learning_rate": 2.0223156134646142e-05,
"loss": 0.0416,
"step": 8350
},
{
"epoch": 32.15384615384615,
"grad_norm": 1.584571361541748,
"learning_rate": 2.019989915937138e-05,
"loss": 0.0395,
"step": 8360
},
{
"epoch": 32.19230769230769,
"grad_norm": 1.7417460680007935,
"learning_rate": 2.0176627967770873e-05,
"loss": 0.0343,
"step": 8370
},
{
"epoch": 32.23076923076923,
"grad_norm": 1.6731925010681152,
"learning_rate": 2.015334262346717e-05,
"loss": 0.0407,
"step": 8380
},
{
"epoch": 32.26923076923077,
"grad_norm": 1.829028844833374,
"learning_rate": 2.0130043190121515e-05,
"loss": 0.0402,
"step": 8390
},
{
"epoch": 32.30769230769231,
"grad_norm": 1.5785481929779053,
"learning_rate": 2.0106729731433663e-05,
"loss": 0.0342,
"step": 8400
},
{
"epoch": 32.34615384615385,
"grad_norm": 1.5879954099655151,
"learning_rate": 2.008340231114173e-05,
"loss": 0.0388,
"step": 8410
},
{
"epoch": 32.38461538461539,
"grad_norm": 1.334164023399353,
"learning_rate": 2.006006099302199e-05,
"loss": 0.0377,
"step": 8420
},
{
"epoch": 32.42307692307692,
"grad_norm": 1.397067904472351,
"learning_rate": 2.003670584088871e-05,
"loss": 0.0381,
"step": 8430
},
{
"epoch": 32.46153846153846,
"grad_norm": 1.5849800109863281,
"learning_rate": 2.001333691859399e-05,
"loss": 0.0352,
"step": 8440
},
{
"epoch": 32.5,
"grad_norm": 1.4086744785308838,
"learning_rate": 1.9989954290027565e-05,
"loss": 0.0381,
"step": 8450
},
{
"epoch": 32.53846153846154,
"grad_norm": 1.6298279762268066,
"learning_rate": 1.9966558019116654e-05,
"loss": 0.0345,
"step": 8460
},
{
"epoch": 32.57692307692308,
"grad_norm": 1.541572093963623,
"learning_rate": 1.9943148169825766e-05,
"loss": 0.0393,
"step": 8470
},
{
"epoch": 32.61538461538461,
"grad_norm": 1.4838348627090454,
"learning_rate": 1.991972480615653e-05,
"loss": 0.0461,
"step": 8480
},
{
"epoch": 32.65384615384615,
"grad_norm": 1.341556429862976,
"learning_rate": 1.989628799214754e-05,
"loss": 0.0411,
"step": 8490
},
{
"epoch": 32.69230769230769,
"grad_norm": 1.6740994453430176,
"learning_rate": 1.987283779187414e-05,
"loss": 0.0354,
"step": 8500
},
{
"epoch": 32.73076923076923,
"grad_norm": 1.841543436050415,
"learning_rate": 1.9849374269448288e-05,
"loss": 0.0407,
"step": 8510
},
{
"epoch": 32.76923076923077,
"grad_norm": 1.9429466724395752,
"learning_rate": 1.982589748901836e-05,
"loss": 0.0389,
"step": 8520
},
{
"epoch": 32.80769230769231,
"grad_norm": 1.6010141372680664,
"learning_rate": 1.9802407514768964e-05,
"loss": 0.0367,
"step": 8530
},
{
"epoch": 32.84615384615385,
"grad_norm": 1.6420562267303467,
"learning_rate": 1.9778904410920808e-05,
"loss": 0.0353,
"step": 8540
},
{
"epoch": 32.88461538461539,
"grad_norm": 1.6107007265090942,
"learning_rate": 1.9755388241730475e-05,
"loss": 0.0411,
"step": 8550
},
{
"epoch": 32.92307692307692,
"grad_norm": 1.331921935081482,
"learning_rate": 1.973185907149027e-05,
"loss": 0.0356,
"step": 8560
},
{
"epoch": 32.96153846153846,
"grad_norm": 2.029869556427002,
"learning_rate": 1.970831696452805e-05,
"loss": 0.0426,
"step": 8570
},
{
"epoch": 33.0,
"grad_norm": 1.8805733919143677,
"learning_rate": 1.9684761985207038e-05,
"loss": 0.0369,
"step": 8580
},
{
"epoch": 33.03846153846154,
"grad_norm": 1.8220821619033813,
"learning_rate": 1.9661194197925644e-05,
"loss": 0.0357,
"step": 8590
},
{
"epoch": 33.07692307692308,
"grad_norm": 1.6471298933029175,
"learning_rate": 1.9637613667117303e-05,
"loss": 0.0432,
"step": 8600
},
{
"epoch": 33.11538461538461,
"grad_norm": 1.5916423797607422,
"learning_rate": 1.961402045725028e-05,
"loss": 0.0405,
"step": 8610
},
{
"epoch": 33.15384615384615,
"grad_norm": 1.7723677158355713,
"learning_rate": 1.9590414632827513e-05,
"loss": 0.041,
"step": 8620
},
{
"epoch": 33.19230769230769,
"grad_norm": 1.4833166599273682,
"learning_rate": 1.9566796258386424e-05,
"loss": 0.0368,
"step": 8630
},
{
"epoch": 33.23076923076923,
"grad_norm": 1.5924839973449707,
"learning_rate": 1.9543165398498743e-05,
"loss": 0.0378,
"step": 8640
},
{
"epoch": 33.26923076923077,
"grad_norm": 1.5793626308441162,
"learning_rate": 1.9519522117770355e-05,
"loss": 0.041,
"step": 8650
},
{
"epoch": 33.30769230769231,
"grad_norm": 1.8710663318634033,
"learning_rate": 1.9495866480841063e-05,
"loss": 0.0384,
"step": 8660
},
{
"epoch": 33.34615384615385,
"grad_norm": 1.8821028470993042,
"learning_rate": 1.9472198552384494e-05,
"loss": 0.0377,
"step": 8670
},
{
"epoch": 33.38461538461539,
"grad_norm": 1.6655566692352295,
"learning_rate": 1.9448518397107848e-05,
"loss": 0.0443,
"step": 8680
},
{
"epoch": 33.42307692307692,
"grad_norm": 1.593344807624817,
"learning_rate": 1.942482607975177e-05,
"loss": 0.0384,
"step": 8690
},
{
"epoch": 33.46153846153846,
"grad_norm": 1.651685118675232,
"learning_rate": 1.940112166509016e-05,
"loss": 0.0383,
"step": 8700
},
{
"epoch": 33.5,
"grad_norm": 1.4256179332733154,
"learning_rate": 1.937740521792996e-05,
"loss": 0.0343,
"step": 8710
},
{
"epoch": 33.53846153846154,
"grad_norm": 1.3333659172058105,
"learning_rate": 1.935367680311106e-05,
"loss": 0.0383,
"step": 8720
},
{
"epoch": 33.57692307692308,
"grad_norm": 1.2656036615371704,
"learning_rate": 1.9329936485506012e-05,
"loss": 0.0402,
"step": 8730
},
{
"epoch": 33.61538461538461,
"grad_norm": 1.6089822053909302,
"learning_rate": 1.930618433001996e-05,
"loss": 0.0392,
"step": 8740
},
{
"epoch": 33.65384615384615,
"grad_norm": 1.4962538480758667,
"learning_rate": 1.9282420401590377e-05,
"loss": 0.0408,
"step": 8750
},
{
"epoch": 33.69230769230769,
"grad_norm": 2.0058789253234863,
"learning_rate": 1.925864476518694e-05,
"loss": 0.043,
"step": 8760
},
{
"epoch": 33.73076923076923,
"grad_norm": 1.5615150928497314,
"learning_rate": 1.9234857485811336e-05,
"loss": 0.0369,
"step": 8770
},
{
"epoch": 33.76923076923077,
"grad_norm": 1.595367193222046,
"learning_rate": 1.9211058628497066e-05,
"loss": 0.041,
"step": 8780
},
{
"epoch": 33.80769230769231,
"grad_norm": 1.1622068881988525,
"learning_rate": 1.918724825830931e-05,
"loss": 0.0438,
"step": 8790
},
{
"epoch": 33.84615384615385,
"grad_norm": 1.4720163345336914,
"learning_rate": 1.9163426440344702e-05,
"loss": 0.0397,
"step": 8800
},
{
"epoch": 33.88461538461539,
"grad_norm": 1.658189058303833,
"learning_rate": 1.913959323973119e-05,
"loss": 0.0411,
"step": 8810
},
{
"epoch": 33.92307692307692,
"grad_norm": 1.8597990274429321,
"learning_rate": 1.9115748721627827e-05,
"loss": 0.0419,
"step": 8820
},
{
"epoch": 33.96153846153846,
"grad_norm": 1.6139363050460815,
"learning_rate": 1.9091892951224614e-05,
"loss": 0.0394,
"step": 8830
},
{
"epoch": 34.0,
"grad_norm": 1.4286227226257324,
"learning_rate": 1.906802599374233e-05,
"loss": 0.0352,
"step": 8840
},
{
"epoch": 34.03846153846154,
"grad_norm": 1.7807533740997314,
"learning_rate": 1.904414791443231e-05,
"loss": 0.0425,
"step": 8850
},
{
"epoch": 34.07692307692308,
"grad_norm": 1.8693132400512695,
"learning_rate": 1.9020258778576324e-05,
"loss": 0.0416,
"step": 8860
},
{
"epoch": 34.11538461538461,
"grad_norm": 1.5255342721939087,
"learning_rate": 1.8996358651486347e-05,
"loss": 0.0364,
"step": 8870
},
{
"epoch": 34.15384615384615,
"grad_norm": 1.4123107194900513,
"learning_rate": 1.8972447598504417e-05,
"loss": 0.0409,
"step": 8880
},
{
"epoch": 34.19230769230769,
"grad_norm": 1.831511378288269,
"learning_rate": 1.8948525685002438e-05,
"loss": 0.0353,
"step": 8890
},
{
"epoch": 34.23076923076923,
"grad_norm": 1.5092872381210327,
"learning_rate": 1.892459297638201e-05,
"loss": 0.043,
"step": 8900
},
{
"epoch": 34.26923076923077,
"grad_norm": 1.5909144878387451,
"learning_rate": 1.890064953807425e-05,
"loss": 0.034,
"step": 8910
},
{
"epoch": 34.30769230769231,
"grad_norm": 1.4315061569213867,
"learning_rate": 1.8876695435539596e-05,
"loss": 0.0371,
"step": 8920
},
{
"epoch": 34.34615384615385,
"grad_norm": 1.372524380683899,
"learning_rate": 1.8852730734267653e-05,
"loss": 0.0348,
"step": 8930
},
{
"epoch": 34.38461538461539,
"grad_norm": 1.4289276599884033,
"learning_rate": 1.8828755499776997e-05,
"loss": 0.0376,
"step": 8940
},
{
"epoch": 34.42307692307692,
"grad_norm": 1.375580072402954,
"learning_rate": 1.8804769797615007e-05,
"loss": 0.0386,
"step": 8950
},
{
"epoch": 34.46153846153846,
"grad_norm": 1.6561743021011353,
"learning_rate": 1.8780773693357675e-05,
"loss": 0.0415,
"step": 8960
},
{
"epoch": 34.5,
"grad_norm": 1.3034884929656982,
"learning_rate": 1.8756767252609433e-05,
"loss": 0.0337,
"step": 8970
},
{
"epoch": 34.53846153846154,
"grad_norm": 1.5481013059616089,
"learning_rate": 1.8732750541002974e-05,
"loss": 0.0389,
"step": 8980
},
{
"epoch": 34.57692307692308,
"grad_norm": 1.4725347757339478,
"learning_rate": 1.870872362419907e-05,
"loss": 0.0392,
"step": 8990
},
{
"epoch": 34.61538461538461,
"grad_norm": 1.3678796291351318,
"learning_rate": 1.8684686567886398e-05,
"loss": 0.0432,
"step": 9000
},
{
"epoch": 34.65384615384615,
"grad_norm": 1.5329874753952026,
"learning_rate": 1.8660639437781344e-05,
"loss": 0.0354,
"step": 9010
},
{
"epoch": 34.69230769230769,
"grad_norm": 1.523851990699768,
"learning_rate": 1.8636582299627854e-05,
"loss": 0.0356,
"step": 9020
},
{
"epoch": 34.73076923076923,
"grad_norm": 1.2790080308914185,
"learning_rate": 1.8612515219197215e-05,
"loss": 0.0354,
"step": 9030
},
{
"epoch": 34.76923076923077,
"grad_norm": 1.50131356716156,
"learning_rate": 1.858843826228791e-05,
"loss": 0.035,
"step": 9040
},
{
"epoch": 34.80769230769231,
"grad_norm": 1.5479196310043335,
"learning_rate": 1.8564351494725423e-05,
"loss": 0.0364,
"step": 9050
},
{
"epoch": 34.84615384615385,
"grad_norm": 1.5020997524261475,
"learning_rate": 1.8540254982362053e-05,
"loss": 0.0375,
"step": 9060
},
{
"epoch": 34.88461538461539,
"grad_norm": 1.6902251243591309,
"learning_rate": 1.8516148791076743e-05,
"loss": 0.0386,
"step": 9070
},
{
"epoch": 34.92307692307692,
"grad_norm": 1.5879284143447876,
"learning_rate": 1.8492032986774904e-05,
"loss": 0.0357,
"step": 9080
},
{
"epoch": 34.96153846153846,
"grad_norm": 1.6842821836471558,
"learning_rate": 1.8467907635388225e-05,
"loss": 0.037,
"step": 9090
},
{
"epoch": 35.0,
"grad_norm": 1.4614428281784058,
"learning_rate": 1.844377280287449e-05,
"loss": 0.038,
"step": 9100
},
{
"epoch": 35.03846153846154,
"grad_norm": 1.5030885934829712,
"learning_rate": 1.8419628555217407e-05,
"loss": 0.0402,
"step": 9110
},
{
"epoch": 35.07692307692308,
"grad_norm": 1.419340968132019,
"learning_rate": 1.839547495842644e-05,
"loss": 0.0385,
"step": 9120
},
{
"epoch": 35.11538461538461,
"grad_norm": 1.6496049165725708,
"learning_rate": 1.8371312078536587e-05,
"loss": 0.0315,
"step": 9130
},
{
"epoch": 35.15384615384615,
"grad_norm": 1.2874219417572021,
"learning_rate": 1.834713998160825e-05,
"loss": 0.0388,
"step": 9140
},
{
"epoch": 35.19230769230769,
"grad_norm": 1.6375945806503296,
"learning_rate": 1.832295873372701e-05,
"loss": 0.0419,
"step": 9150
},
{
"epoch": 35.23076923076923,
"grad_norm": 1.6813386678695679,
"learning_rate": 1.8298768401003477e-05,
"loss": 0.0379,
"step": 9160
},
{
"epoch": 35.26923076923077,
"grad_norm": 1.5185751914978027,
"learning_rate": 1.8274569049573103e-05,
"loss": 0.0415,
"step": 9170
},
{
"epoch": 35.30769230769231,
"grad_norm": 1.4227298498153687,
"learning_rate": 1.8250360745595983e-05,
"loss": 0.0394,
"step": 9180
},
{
"epoch": 35.34615384615385,
"grad_norm": 1.1683084964752197,
"learning_rate": 1.8226143555256703e-05,
"loss": 0.0341,
"step": 9190
},
{
"epoch": 35.38461538461539,
"grad_norm": 1.2086158990859985,
"learning_rate": 1.820191754476413e-05,
"loss": 0.0395,
"step": 9200
},
{
"epoch": 35.42307692307692,
"grad_norm": 1.3567473888397217,
"learning_rate": 1.8177682780351256e-05,
"loss": 0.0374,
"step": 9210
},
{
"epoch": 35.46153846153846,
"grad_norm": 1.413490653038025,
"learning_rate": 1.8153439328275e-05,
"loss": 0.0435,
"step": 9220
},
{
"epoch": 35.5,
"grad_norm": 1.8671621084213257,
"learning_rate": 1.8129187254816035e-05,
"loss": 0.0396,
"step": 9230
},
{
"epoch": 35.53846153846154,
"grad_norm": 1.7603775262832642,
"learning_rate": 1.81049266262786e-05,
"loss": 0.0367,
"step": 9240
},
{
"epoch": 35.57692307692308,
"grad_norm": 1.467278003692627,
"learning_rate": 1.808065750899033e-05,
"loss": 0.041,
"step": 9250
},
{
"epoch": 35.61538461538461,
"grad_norm": 1.8476868867874146,
"learning_rate": 1.8056379969302066e-05,
"loss": 0.0373,
"step": 9260
},
{
"epoch": 35.65384615384615,
"grad_norm": 1.9753694534301758,
"learning_rate": 1.8032094073587675e-05,
"loss": 0.0374,
"step": 9270
},
{
"epoch": 35.69230769230769,
"grad_norm": 1.9146374464035034,
"learning_rate": 1.800779988824387e-05,
"loss": 0.0344,
"step": 9280
},
{
"epoch": 35.73076923076923,
"grad_norm": 1.2956942319869995,
"learning_rate": 1.7983497479690018e-05,
"loss": 0.0348,
"step": 9290
},
{
"epoch": 35.76923076923077,
"grad_norm": 1.381451964378357,
"learning_rate": 1.795918691436798e-05,
"loss": 0.0394,
"step": 9300
},
{
"epoch": 35.80769230769231,
"grad_norm": 1.3659331798553467,
"learning_rate": 1.7934868258741917e-05,
"loss": 0.0366,
"step": 9310
},
{
"epoch": 35.84615384615385,
"grad_norm": 1.5138591527938843,
"learning_rate": 1.79105415792981e-05,
"loss": 0.037,
"step": 9320
},
{
"epoch": 35.88461538461539,
"grad_norm": 1.5586953163146973,
"learning_rate": 1.788620694254475e-05,
"loss": 0.0409,
"step": 9330
},
{
"epoch": 35.92307692307692,
"grad_norm": 1.2028257846832275,
"learning_rate": 1.7861864415011827e-05,
"loss": 0.0409,
"step": 9340
},
{
"epoch": 35.96153846153846,
"grad_norm": 1.3556921482086182,
"learning_rate": 1.783751406325087e-05,
"loss": 0.0334,
"step": 9350
},
{
"epoch": 36.0,
"grad_norm": 1.3308838605880737,
"learning_rate": 1.7813155953834814e-05,
"loss": 0.0397,
"step": 9360
},
{
"epoch": 36.03846153846154,
"grad_norm": 1.681963324546814,
"learning_rate": 1.7788790153357803e-05,
"loss": 0.0436,
"step": 9370
},
{
"epoch": 36.07692307692308,
"grad_norm": 1.1227318048477173,
"learning_rate": 1.7764416728435e-05,
"loss": 0.0422,
"step": 9380
},
{
"epoch": 36.11538461538461,
"grad_norm": 1.438764214515686,
"learning_rate": 1.774003574570242e-05,
"loss": 0.0415,
"step": 9390
},
{
"epoch": 36.15384615384615,
"grad_norm": 1.7580561637878418,
"learning_rate": 1.7715647271816744e-05,
"loss": 0.0349,
"step": 9400
},
{
"epoch": 36.19230769230769,
"grad_norm": 1.7516335248947144,
"learning_rate": 1.769125137345512e-05,
"loss": 0.0359,
"step": 9410
},
{
"epoch": 36.23076923076923,
"grad_norm": 1.7185901403427124,
"learning_rate": 1.7666848117315008e-05,
"loss": 0.0396,
"step": 9420
},
{
"epoch": 36.26923076923077,
"grad_norm": 1.6472246646881104,
"learning_rate": 1.7642437570113974e-05,
"loss": 0.0354,
"step": 9430
},
{
"epoch": 36.30769230769231,
"grad_norm": 1.1611000299453735,
"learning_rate": 1.7618019798589525e-05,
"loss": 0.0426,
"step": 9440
},
{
"epoch": 36.34615384615385,
"grad_norm": 1.5789954662322998,
"learning_rate": 1.7593594869498915e-05,
"loss": 0.0346,
"step": 9450
},
{
"epoch": 36.38461538461539,
"grad_norm": 1.232185959815979,
"learning_rate": 1.7569162849618966e-05,
"loss": 0.0392,
"step": 9460
},
{
"epoch": 36.42307692307692,
"grad_norm": 1.2570834159851074,
"learning_rate": 1.75447238057459e-05,
"loss": 0.0379,
"step": 9470
},
{
"epoch": 36.46153846153846,
"grad_norm": 1.4947396516799927,
"learning_rate": 1.752027780469511e-05,
"loss": 0.0367,
"step": 9480
},
{
"epoch": 36.5,
"grad_norm": 1.4545466899871826,
"learning_rate": 1.7495824913301043e-05,
"loss": 0.0419,
"step": 9490
},
{
"epoch": 36.53846153846154,
"grad_norm": 1.4352824687957764,
"learning_rate": 1.7471365198416957e-05,
"loss": 0.0381,
"step": 9500
},
{
"epoch": 36.57692307692308,
"grad_norm": 1.5588959455490112,
"learning_rate": 1.7446898726914797e-05,
"loss": 0.0366,
"step": 9510
},
{
"epoch": 36.61538461538461,
"grad_norm": 1.6907953023910522,
"learning_rate": 1.742242556568495e-05,
"loss": 0.0425,
"step": 9520
},
{
"epoch": 36.65384615384615,
"grad_norm": 1.4294540882110596,
"learning_rate": 1.73979457816361e-05,
"loss": 0.0387,
"step": 9530
},
{
"epoch": 36.69230769230769,
"grad_norm": 1.4833825826644897,
"learning_rate": 1.7373459441695058e-05,
"loss": 0.0427,
"step": 9540
},
{
"epoch": 36.73076923076923,
"grad_norm": 1.466599702835083,
"learning_rate": 1.7348966612806524e-05,
"loss": 0.0419,
"step": 9550
},
{
"epoch": 36.76923076923077,
"grad_norm": 1.2378714084625244,
"learning_rate": 1.7324467361932973e-05,
"loss": 0.0392,
"step": 9560
},
{
"epoch": 36.80769230769231,
"grad_norm": 1.413780689239502,
"learning_rate": 1.729996175605441e-05,
"loss": 0.0402,
"step": 9570
},
{
"epoch": 36.84615384615385,
"grad_norm": 1.374563217163086,
"learning_rate": 1.7275449862168235e-05,
"loss": 0.0395,
"step": 9580
},
{
"epoch": 36.88461538461539,
"grad_norm": 1.3563233613967896,
"learning_rate": 1.725093174728902e-05,
"loss": 0.0323,
"step": 9590
},
{
"epoch": 36.92307692307692,
"grad_norm": 1.4940588474273682,
"learning_rate": 1.7226407478448357e-05,
"loss": 0.0384,
"step": 9600
},
{
"epoch": 36.96153846153846,
"grad_norm": 1.593131422996521,
"learning_rate": 1.7201877122694666e-05,
"loss": 0.0421,
"step": 9610
},
{
"epoch": 37.0,
"grad_norm": 1.7007324695587158,
"learning_rate": 1.7177340747093e-05,
"loss": 0.0394,
"step": 9620
},
{
"epoch": 37.03846153846154,
"grad_norm": 1.4172247648239136,
"learning_rate": 1.7152798418724873e-05,
"loss": 0.0414,
"step": 9630
},
{
"epoch": 37.07692307692308,
"grad_norm": 1.3566648960113525,
"learning_rate": 1.712825020468807e-05,
"loss": 0.0356,
"step": 9640
},
{
"epoch": 37.11538461538461,
"grad_norm": 1.1719372272491455,
"learning_rate": 1.710369617209648e-05,
"loss": 0.0407,
"step": 9650
},
{
"epoch": 37.15384615384615,
"grad_norm": 1.3853756189346313,
"learning_rate": 1.7079136388079884e-05,
"loss": 0.0379,
"step": 9660
},
{
"epoch": 37.19230769230769,
"grad_norm": 1.4619250297546387,
"learning_rate": 1.7054570919783796e-05,
"loss": 0.0378,
"step": 9670
},
{
"epoch": 37.23076923076923,
"grad_norm": 1.4194920063018799,
"learning_rate": 1.7029999834369264e-05,
"loss": 0.0381,
"step": 9680
},
{
"epoch": 37.26923076923077,
"grad_norm": 1.2735564708709717,
"learning_rate": 1.7005423199012696e-05,
"loss": 0.0339,
"step": 9690
},
{
"epoch": 37.30769230769231,
"grad_norm": 1.3071863651275635,
"learning_rate": 1.6980841080905687e-05,
"loss": 0.0344,
"step": 9700
},
{
"epoch": 37.34615384615385,
"grad_norm": 1.2849153280258179,
"learning_rate": 1.6956253547254798e-05,
"loss": 0.0406,
"step": 9710
},
{
"epoch": 37.38461538461539,
"grad_norm": 1.2601699829101562,
"learning_rate": 1.693166066528141e-05,
"loss": 0.0373,
"step": 9720
},
{
"epoch": 37.42307692307692,
"grad_norm": 1.2954739332199097,
"learning_rate": 1.690706250222152e-05,
"loss": 0.0404,
"step": 9730
},
{
"epoch": 37.46153846153846,
"grad_norm": 1.5108919143676758,
"learning_rate": 1.6882459125325573e-05,
"loss": 0.0372,
"step": 9740
},
{
"epoch": 37.5,
"grad_norm": 1.3737967014312744,
"learning_rate": 1.685785060185826e-05,
"loss": 0.036,
"step": 9750
},
{
"epoch": 37.53846153846154,
"grad_norm": 1.3580888509750366,
"learning_rate": 1.683323699909834e-05,
"loss": 0.0366,
"step": 9760
},
{
"epoch": 37.57692307692308,
"grad_norm": 1.485553503036499,
"learning_rate": 1.6808618384338472e-05,
"loss": 0.0369,
"step": 9770
},
{
"epoch": 37.61538461538461,
"grad_norm": 1.2370328903198242,
"learning_rate": 1.6783994824885e-05,
"loss": 0.0355,
"step": 9780
},
{
"epoch": 37.65384615384615,
"grad_norm": 1.3363810777664185,
"learning_rate": 1.6759366388057795e-05,
"loss": 0.0399,
"step": 9790
},
{
"epoch": 37.69230769230769,
"grad_norm": 1.4240365028381348,
"learning_rate": 1.6734733141190073e-05,
"loss": 0.0322,
"step": 9800
},
{
"epoch": 37.73076923076923,
"grad_norm": 1.1581053733825684,
"learning_rate": 1.6710095151628182e-05,
"loss": 0.0398,
"step": 9810
},
{
"epoch": 37.76923076923077,
"grad_norm": 1.2987163066864014,
"learning_rate": 1.668545248673144e-05,
"loss": 0.0326,
"step": 9820
},
{
"epoch": 37.80769230769231,
"grad_norm": 1.334222674369812,
"learning_rate": 1.6660805213871962e-05,
"loss": 0.0434,
"step": 9830
},
{
"epoch": 37.84615384615385,
"grad_norm": 1.4126960039138794,
"learning_rate": 1.663615340043445e-05,
"loss": 0.0402,
"step": 9840
},
{
"epoch": 37.88461538461539,
"grad_norm": 1.3437000513076782,
"learning_rate": 1.6611497113816014e-05,
"loss": 0.0424,
"step": 9850
},
{
"epoch": 37.92307692307692,
"grad_norm": 1.3839226961135864,
"learning_rate": 1.6586836421426007e-05,
"loss": 0.038,
"step": 9860
},
{
"epoch": 37.96153846153846,
"grad_norm": 1.3781486749649048,
"learning_rate": 1.6562171390685815e-05,
"loss": 0.0328,
"step": 9870
},
{
"epoch": 38.0,
"grad_norm": 1.212611436843872,
"learning_rate": 1.653750208902869e-05,
"loss": 0.0372,
"step": 9880
},
{
"epoch": 38.03846153846154,
"grad_norm": 1.7444937229156494,
"learning_rate": 1.6512828583899562e-05,
"loss": 0.042,
"step": 9890
},
{
"epoch": 38.07692307692308,
"grad_norm": 1.4450607299804688,
"learning_rate": 1.648815094275486e-05,
"loss": 0.0368,
"step": 9900
},
{
"epoch": 38.11538461538461,
"grad_norm": 1.3216570615768433,
"learning_rate": 1.6463469233062302e-05,
"loss": 0.0377,
"step": 9910
},
{
"epoch": 38.15384615384615,
"grad_norm": 1.340665578842163,
"learning_rate": 1.6438783522300742e-05,
"loss": 0.0398,
"step": 9920
},
{
"epoch": 38.19230769230769,
"grad_norm": 1.386986255645752,
"learning_rate": 1.641409387795997e-05,
"loss": 0.0314,
"step": 9930
},
{
"epoch": 38.23076923076923,
"grad_norm": 1.4249019622802734,
"learning_rate": 1.6389400367540534e-05,
"loss": 0.0438,
"step": 9940
},
{
"epoch": 38.26923076923077,
"grad_norm": 1.537424087524414,
"learning_rate": 1.6364703058553552e-05,
"loss": 0.0404,
"step": 9950
},
{
"epoch": 38.30769230769231,
"grad_norm": 1.465016484260559,
"learning_rate": 1.6340002018520512e-05,
"loss": 0.0371,
"step": 9960
},
{
"epoch": 38.34615384615385,
"grad_norm": 1.5465450286865234,
"learning_rate": 1.6315297314973126e-05,
"loss": 0.0351,
"step": 9970
},
{
"epoch": 38.38461538461539,
"grad_norm": 1.1357992887496948,
"learning_rate": 1.6290589015453102e-05,
"loss": 0.0358,
"step": 9980
},
{
"epoch": 38.42307692307692,
"grad_norm": 1.0869678258895874,
"learning_rate": 1.6265877187511993e-05,
"loss": 0.0376,
"step": 9990
},
{
"epoch": 38.46153846153846,
"grad_norm": 1.5823304653167725,
"learning_rate": 1.6241161898710993e-05,
"loss": 0.0408,
"step": 10000
},
{
"epoch": 38.5,
"grad_norm": 1.1973952054977417,
"learning_rate": 1.6216443216620752e-05,
"loss": 0.0364,
"step": 10010
},
{
"epoch": 38.53846153846154,
"grad_norm": 1.229051947593689,
"learning_rate": 1.6191721208821208e-05,
"loss": 0.0372,
"step": 10020
},
{
"epoch": 38.57692307692308,
"grad_norm": 1.330804467201233,
"learning_rate": 1.6166995942901382e-05,
"loss": 0.0371,
"step": 10030
},
{
"epoch": 38.61538461538461,
"grad_norm": 1.4124090671539307,
"learning_rate": 1.614226748645921e-05,
"loss": 0.0433,
"step": 10040
},
{
"epoch": 38.65384615384615,
"grad_norm": 1.0600825548171997,
"learning_rate": 1.6117535907101354e-05,
"loss": 0.0326,
"step": 10050
},
{
"epoch": 38.69230769230769,
"grad_norm": 1.338943600654602,
"learning_rate": 1.6092801272442996e-05,
"loss": 0.0392,
"step": 10060
},
{
"epoch": 38.73076923076923,
"grad_norm": 1.3412127494812012,
"learning_rate": 1.606806365010769e-05,
"loss": 0.0368,
"step": 10070
},
{
"epoch": 38.76923076923077,
"grad_norm": 1.6434335708618164,
"learning_rate": 1.6043323107727143e-05,
"loss": 0.0362,
"step": 10080
},
{
"epoch": 38.80769230769231,
"grad_norm": 1.3812055587768555,
"learning_rate": 1.6018579712941064e-05,
"loss": 0.0343,
"step": 10090
},
{
"epoch": 38.84615384615385,
"grad_norm": 1.2723709344863892,
"learning_rate": 1.599383353339694e-05,
"loss": 0.0361,
"step": 10100
},
{
"epoch": 38.88461538461539,
"grad_norm": 1.6107410192489624,
"learning_rate": 1.596908463674989e-05,
"loss": 0.0385,
"step": 10110
},
{
"epoch": 38.92307692307692,
"grad_norm": 1.566305160522461,
"learning_rate": 1.5944333090662442e-05,
"loss": 0.0364,
"step": 10120
},
{
"epoch": 38.96153846153846,
"grad_norm": 1.3718688488006592,
"learning_rate": 1.5919578962804386e-05,
"loss": 0.0349,
"step": 10130
},
{
"epoch": 39.0,
"grad_norm": 1.5631097555160522,
"learning_rate": 1.5894822320852563e-05,
"loss": 0.0347,
"step": 10140
},
{
"epoch": 39.03846153846154,
"grad_norm": 1.1739816665649414,
"learning_rate": 1.5870063232490677e-05,
"loss": 0.0339,
"step": 10150
},
{
"epoch": 39.07692307692308,
"grad_norm": 1.4080265760421753,
"learning_rate": 1.5845301765409144e-05,
"loss": 0.0341,
"step": 10160
},
{
"epoch": 39.11538461538461,
"grad_norm": 1.1734020709991455,
"learning_rate": 1.5820537987304856e-05,
"loss": 0.0375,
"step": 10170
},
{
"epoch": 39.15384615384615,
"grad_norm": 1.4352694749832153,
"learning_rate": 1.5795771965881044e-05,
"loss": 0.0369,
"step": 10180
},
{
"epoch": 39.19230769230769,
"grad_norm": 1.4638075828552246,
"learning_rate": 1.577100376884707e-05,
"loss": 0.0376,
"step": 10190
},
{
"epoch": 39.23076923076923,
"grad_norm": 1.5176154375076294,
"learning_rate": 1.5746233463918226e-05,
"loss": 0.0351,
"step": 10200
},
{
"epoch": 39.26923076923077,
"grad_norm": 1.6033369302749634,
"learning_rate": 1.572146111881559e-05,
"loss": 0.0384,
"step": 10210
},
{
"epoch": 39.30769230769231,
"grad_norm": 1.4504886865615845,
"learning_rate": 1.56966868012658e-05,
"loss": 0.0355,
"step": 10220
},
{
"epoch": 39.34615384615385,
"grad_norm": 1.1136956214904785,
"learning_rate": 1.56719105790009e-05,
"loss": 0.0376,
"step": 10230
},
{
"epoch": 39.38461538461539,
"grad_norm": 1.2752516269683838,
"learning_rate": 1.5647132519758135e-05,
"loss": 0.034,
"step": 10240
},
{
"epoch": 39.42307692307692,
"grad_norm": 1.298332929611206,
"learning_rate": 1.562235269127977e-05,
"loss": 0.0348,
"step": 10250
},
{
"epoch": 39.46153846153846,
"grad_norm": 1.5165218114852905,
"learning_rate": 1.5597571161312914e-05,
"loss": 0.0365,
"step": 10260
},
{
"epoch": 39.5,
"grad_norm": 1.3248893022537231,
"learning_rate": 1.557278799760932e-05,
"loss": 0.0345,
"step": 10270
},
{
"epoch": 39.53846153846154,
"grad_norm": 1.1644551753997803,
"learning_rate": 1.5548003267925214e-05,
"loss": 0.045,
"step": 10280
},
{
"epoch": 39.57692307692308,
"grad_norm": 1.4049543142318726,
"learning_rate": 1.5523217040021094e-05,
"loss": 0.0405,
"step": 10290
},
{
"epoch": 39.61538461538461,
"grad_norm": 1.3734569549560547,
"learning_rate": 1.549842938166157e-05,
"loss": 0.0342,
"step": 10300
},
{
"epoch": 39.65384615384615,
"grad_norm": 1.476423978805542,
"learning_rate": 1.5473640360615146e-05,
"loss": 0.0358,
"step": 10310
},
{
"epoch": 39.69230769230769,
"grad_norm": 1.3189231157302856,
"learning_rate": 1.5448850044654063e-05,
"loss": 0.0391,
"step": 10320
},
{
"epoch": 39.73076923076923,
"grad_norm": 1.5631731748580933,
"learning_rate": 1.5424058501554102e-05,
"loss": 0.0347,
"step": 10330
},
{
"epoch": 39.76923076923077,
"grad_norm": 1.3164517879486084,
"learning_rate": 1.5399265799094383e-05,
"loss": 0.0333,
"step": 10340
},
{
"epoch": 39.80769230769231,
"grad_norm": 1.293961524963379,
"learning_rate": 1.537447200505722e-05,
"loss": 0.0358,
"step": 10350
},
{
"epoch": 39.84615384615385,
"grad_norm": 1.3249964714050293,
"learning_rate": 1.5349677187227892e-05,
"loss": 0.0297,
"step": 10360
},
{
"epoch": 39.88461538461539,
"grad_norm": 1.552517056465149,
"learning_rate": 1.532488141339449e-05,
"loss": 0.0396,
"step": 10370
},
{
"epoch": 39.92307692307692,
"grad_norm": 1.540193796157837,
"learning_rate": 1.5300084751347703e-05,
"loss": 0.0368,
"step": 10380
},
{
"epoch": 39.96153846153846,
"grad_norm": 1.5178617238998413,
"learning_rate": 1.527528726888067e-05,
"loss": 0.032,
"step": 10390
},
{
"epoch": 40.0,
"grad_norm": 1.5166221857070923,
"learning_rate": 1.5250489033788757e-05,
"loss": 0.0371,
"step": 10400
},
{
"epoch": 40.03846153846154,
"grad_norm": 1.6149933338165283,
"learning_rate": 1.5225690113869383e-05,
"loss": 0.0355,
"step": 10410
},
{
"epoch": 40.07692307692308,
"grad_norm": 1.4362809658050537,
"learning_rate": 1.5200890576921863e-05,
"loss": 0.0354,
"step": 10420
},
{
"epoch": 40.11538461538461,
"grad_norm": 1.6698564291000366,
"learning_rate": 1.5176090490747174e-05,
"loss": 0.0343,
"step": 10430
},
{
"epoch": 40.15384615384615,
"grad_norm": 1.6468229293823242,
"learning_rate": 1.5151289923147806e-05,
"loss": 0.0376,
"step": 10440
},
{
"epoch": 40.19230769230769,
"grad_norm": 1.571349859237671,
"learning_rate": 1.5126488941927568e-05,
"loss": 0.0365,
"step": 10450
},
{
"epoch": 40.23076923076923,
"grad_norm": 1.5099705457687378,
"learning_rate": 1.5101687614891385e-05,
"loss": 0.0349,
"step": 10460
},
{
"epoch": 40.26923076923077,
"grad_norm": 1.4673430919647217,
"learning_rate": 1.5076886009845157e-05,
"loss": 0.0357,
"step": 10470
},
{
"epoch": 40.30769230769231,
"grad_norm": 1.2761073112487793,
"learning_rate": 1.5052084194595507e-05,
"loss": 0.0353,
"step": 10480
},
{
"epoch": 40.34615384615385,
"grad_norm": 1.3340052366256714,
"learning_rate": 1.5027282236949662e-05,
"loss": 0.0337,
"step": 10490
},
{
"epoch": 40.38461538461539,
"grad_norm": 1.1699473857879639,
"learning_rate": 1.5002480204715218e-05,
"loss": 0.0321,
"step": 10500
},
{
"epoch": 40.42307692307692,
"grad_norm": 1.432531714439392,
"learning_rate": 1.4977678165699992e-05,
"loss": 0.038,
"step": 10510
},
{
"epoch": 40.46153846153846,
"grad_norm": 1.7699759006500244,
"learning_rate": 1.4952876187711806e-05,
"loss": 0.0413,
"step": 10520
},
{
"epoch": 40.5,
"grad_norm": 1.4600393772125244,
"learning_rate": 1.4928074338558326e-05,
"loss": 0.039,
"step": 10530
},
{
"epoch": 40.53846153846154,
"grad_norm": 1.0337119102478027,
"learning_rate": 1.4903272686046857e-05,
"loss": 0.0335,
"step": 10540
},
{
"epoch": 40.57692307692308,
"grad_norm": 1.0465848445892334,
"learning_rate": 1.4878471297984174e-05,
"loss": 0.0327,
"step": 10550
},
{
"epoch": 40.61538461538461,
"grad_norm": 1.2741106748580933,
"learning_rate": 1.4853670242176318e-05,
"loss": 0.0359,
"step": 10560
},
{
"epoch": 40.65384615384615,
"grad_norm": 1.2669447660446167,
"learning_rate": 1.4828869586428433e-05,
"loss": 0.0355,
"step": 10570
},
{
"epoch": 40.69230769230769,
"grad_norm": 1.4156800508499146,
"learning_rate": 1.4804069398544563e-05,
"loss": 0.0332,
"step": 10580
},
{
"epoch": 40.73076923076923,
"grad_norm": 1.3298932313919067,
"learning_rate": 1.477926974632748e-05,
"loss": 0.0373,
"step": 10590
},
{
"epoch": 40.76923076923077,
"grad_norm": 1.3701391220092773,
"learning_rate": 1.4754470697578478e-05,
"loss": 0.035,
"step": 10600
},
{
"epoch": 40.80769230769231,
"grad_norm": 1.5692551136016846,
"learning_rate": 1.4729672320097214e-05,
"loss": 0.0356,
"step": 10610
},
{
"epoch": 40.84615384615385,
"grad_norm": 1.1283255815505981,
"learning_rate": 1.4704874681681504e-05,
"loss": 0.0339,
"step": 10620
},
{
"epoch": 40.88461538461539,
"grad_norm": 1.3027880191802979,
"learning_rate": 1.4680077850127146e-05,
"loss": 0.0373,
"step": 10630
},
{
"epoch": 40.92307692307692,
"grad_norm": 1.37151300907135,
"learning_rate": 1.465528189322773e-05,
"loss": 0.0337,
"step": 10640
},
{
"epoch": 40.96153846153846,
"grad_norm": 1.3166391849517822,
"learning_rate": 1.4630486878774455e-05,
"loss": 0.0318,
"step": 10650
},
{
"epoch": 41.0,
"grad_norm": 1.0613034963607788,
"learning_rate": 1.4605692874555942e-05,
"loss": 0.0355,
"step": 10660
},
{
"epoch": 41.03846153846154,
"grad_norm": 1.1301933526992798,
"learning_rate": 1.4580899948358054e-05,
"loss": 0.0335,
"step": 10670
},
{
"epoch": 41.07692307692308,
"grad_norm": 1.2418525218963623,
"learning_rate": 1.4556108167963702e-05,
"loss": 0.0368,
"step": 10680
},
{
"epoch": 41.11538461538461,
"grad_norm": 1.166088342666626,
"learning_rate": 1.4531317601152675e-05,
"loss": 0.0327,
"step": 10690
},
{
"epoch": 41.15384615384615,
"grad_norm": 1.1724998950958252,
"learning_rate": 1.4506528315701425e-05,
"loss": 0.0347,
"step": 10700
},
{
"epoch": 41.19230769230769,
"grad_norm": 1.4540597200393677,
"learning_rate": 1.4481740379382916e-05,
"loss": 0.0395,
"step": 10710
},
{
"epoch": 41.23076923076923,
"grad_norm": 1.3581476211547852,
"learning_rate": 1.445695385996642e-05,
"loss": 0.0305,
"step": 10720
},
{
"epoch": 41.26923076923077,
"grad_norm": 0.9946508407592773,
"learning_rate": 1.4432168825217344e-05,
"loss": 0.0321,
"step": 10730
},
{
"epoch": 41.30769230769231,
"grad_norm": 1.1900794506072998,
"learning_rate": 1.4407385342897005e-05,
"loss": 0.038,
"step": 10740
},
{
"epoch": 41.34615384615385,
"grad_norm": 1.0426830053329468,
"learning_rate": 1.4382603480762514e-05,
"loss": 0.0325,
"step": 10750
},
{
"epoch": 41.38461538461539,
"grad_norm": 1.3630422353744507,
"learning_rate": 1.4357823306566529e-05,
"loss": 0.0376,
"step": 10760
},
{
"epoch": 41.42307692307692,
"grad_norm": 1.1893590688705444,
"learning_rate": 1.4333044888057104e-05,
"loss": 0.036,
"step": 10770
},
{
"epoch": 41.46153846153846,
"grad_norm": 1.2472355365753174,
"learning_rate": 1.4308268292977496e-05,
"loss": 0.0351,
"step": 10780
},
{
"epoch": 41.5,
"grad_norm": 1.2175953388214111,
"learning_rate": 1.4283493589065948e-05,
"loss": 0.0363,
"step": 10790
},
{
"epoch": 41.53846153846154,
"grad_norm": 1.1068720817565918,
"learning_rate": 1.4258720844055573e-05,
"loss": 0.0353,
"step": 10800
},
{
"epoch": 41.57692307692308,
"grad_norm": 1.0755929946899414,
"learning_rate": 1.4233950125674105e-05,
"loss": 0.0351,
"step": 10810
},
{
"epoch": 41.61538461538461,
"grad_norm": 1.193617820739746,
"learning_rate": 1.420918150164374e-05,
"loss": 0.0384,
"step": 10820
},
{
"epoch": 41.65384615384615,
"grad_norm": 1.4359036684036255,
"learning_rate": 1.4184415039680958e-05,
"loss": 0.0387,
"step": 10830
},
{
"epoch": 41.69230769230769,
"grad_norm": 1.1195001602172852,
"learning_rate": 1.415965080749631e-05,
"loss": 0.0368,
"step": 10840
},
{
"epoch": 41.73076923076923,
"grad_norm": 1.4046603441238403,
"learning_rate": 1.4134888872794265e-05,
"loss": 0.0339,
"step": 10850
},
{
"epoch": 41.76923076923077,
"grad_norm": 0.9813646078109741,
"learning_rate": 1.411012930327301e-05,
"loss": 0.0338,
"step": 10860
},
{
"epoch": 41.80769230769231,
"grad_norm": 1.283523678779602,
"learning_rate": 1.4085372166624263e-05,
"loss": 0.0331,
"step": 10870
},
{
"epoch": 41.84615384615385,
"grad_norm": 1.3945362567901611,
"learning_rate": 1.406061753053308e-05,
"loss": 0.0342,
"step": 10880
},
{
"epoch": 41.88461538461539,
"grad_norm": 1.5126872062683105,
"learning_rate": 1.40358654626777e-05,
"loss": 0.0336,
"step": 10890
},
{
"epoch": 41.92307692307692,
"grad_norm": 1.2869038581848145,
"learning_rate": 1.4011116030729333e-05,
"loss": 0.0349,
"step": 10900
},
{
"epoch": 41.96153846153846,
"grad_norm": 1.316064715385437,
"learning_rate": 1.3986369302351974e-05,
"loss": 0.0329,
"step": 10910
},
{
"epoch": 42.0,
"grad_norm": 1.2845642566680908,
"learning_rate": 1.3961625345202245e-05,
"loss": 0.0303,
"step": 10920
},
{
"epoch": 42.03846153846154,
"grad_norm": 1.0855242013931274,
"learning_rate": 1.3936884226929163e-05,
"loss": 0.038,
"step": 10930
},
{
"epoch": 42.07692307692308,
"grad_norm": 1.4729291200637817,
"learning_rate": 1.391214601517401e-05,
"loss": 0.0336,
"step": 10940
},
{
"epoch": 42.11538461538461,
"grad_norm": 1.260464072227478,
"learning_rate": 1.3887410777570116e-05,
"loss": 0.0355,
"step": 10950
},
{
"epoch": 42.15384615384615,
"grad_norm": 1.0688891410827637,
"learning_rate": 1.3862678581742667e-05,
"loss": 0.0342,
"step": 10960
},
{
"epoch": 42.19230769230769,
"grad_norm": 1.4798998832702637,
"learning_rate": 1.3837949495308558e-05,
"loss": 0.033,
"step": 10970
},
{
"epoch": 42.23076923076923,
"grad_norm": 1.2672547101974487,
"learning_rate": 1.3813223585876145e-05,
"loss": 0.0362,
"step": 10980
},
{
"epoch": 42.26923076923077,
"grad_norm": 1.104485034942627,
"learning_rate": 1.3788500921045135e-05,
"loss": 0.0355,
"step": 10990
},
{
"epoch": 42.30769230769231,
"grad_norm": 1.1904023885726929,
"learning_rate": 1.3763781568406343e-05,
"loss": 0.0423,
"step": 11000
},
{
"epoch": 42.34615384615385,
"grad_norm": 1.222254991531372,
"learning_rate": 1.3739065595541548e-05,
"loss": 0.0317,
"step": 11010
},
{
"epoch": 42.38461538461539,
"grad_norm": 1.4406765699386597,
"learning_rate": 1.3714353070023257e-05,
"loss": 0.0347,
"step": 11020
},
{
"epoch": 42.42307692307692,
"grad_norm": 1.3593086004257202,
"learning_rate": 1.368964405941458e-05,
"loss": 0.0402,
"step": 11030
},
{
"epoch": 42.46153846153846,
"grad_norm": 1.4239052534103394,
"learning_rate": 1.366493863126901e-05,
"loss": 0.0313,
"step": 11040
},
{
"epoch": 42.5,
"grad_norm": 1.1864207983016968,
"learning_rate": 1.3640236853130243e-05,
"loss": 0.0363,
"step": 11050
},
{
"epoch": 42.53846153846154,
"grad_norm": 1.0722951889038086,
"learning_rate": 1.3615538792532002e-05,
"loss": 0.0297,
"step": 11060
},
{
"epoch": 42.57692307692308,
"grad_norm": 1.3497544527053833,
"learning_rate": 1.3590844516997832e-05,
"loss": 0.0397,
"step": 11070
},
{
"epoch": 42.61538461538461,
"grad_norm": 1.445272445678711,
"learning_rate": 1.356615409404094e-05,
"loss": 0.036,
"step": 11080
},
{
"epoch": 42.65384615384615,
"grad_norm": 1.157558798789978,
"learning_rate": 1.354146759116401e-05,
"loss": 0.0345,
"step": 11090
},
{
"epoch": 42.69230769230769,
"grad_norm": 1.1287713050842285,
"learning_rate": 1.3516785075858988e-05,
"loss": 0.0357,
"step": 11100
},
{
"epoch": 42.73076923076923,
"grad_norm": 1.0962657928466797,
"learning_rate": 1.3492106615606941e-05,
"loss": 0.0415,
"step": 11110
},
{
"epoch": 42.76923076923077,
"grad_norm": 1.3343199491500854,
"learning_rate": 1.346743227787782e-05,
"loss": 0.0351,
"step": 11120
},
{
"epoch": 42.80769230769231,
"grad_norm": 1.0660537481307983,
"learning_rate": 1.344276213013033e-05,
"loss": 0.033,
"step": 11130
},
{
"epoch": 42.84615384615385,
"grad_norm": 1.3008800745010376,
"learning_rate": 1.3418096239811712e-05,
"loss": 0.0387,
"step": 11140
},
{
"epoch": 42.88461538461539,
"grad_norm": 1.3870251178741455,
"learning_rate": 1.3393434674357579e-05,
"loss": 0.0424,
"step": 11150
},
{
"epoch": 42.92307692307692,
"grad_norm": 1.0750961303710938,
"learning_rate": 1.3368777501191692e-05,
"loss": 0.0383,
"step": 11160
},
{
"epoch": 42.96153846153846,
"grad_norm": 1.3516203165054321,
"learning_rate": 1.334412478772583e-05,
"loss": 0.0366,
"step": 11170
},
{
"epoch": 43.0,
"grad_norm": 1.1265571117401123,
"learning_rate": 1.3319476601359565e-05,
"loss": 0.0341,
"step": 11180
},
{
"epoch": 43.03846153846154,
"grad_norm": 1.2286325693130493,
"learning_rate": 1.3294833009480105e-05,
"loss": 0.0407,
"step": 11190
},
{
"epoch": 43.07692307692308,
"grad_norm": 1.0944629907608032,
"learning_rate": 1.3270194079462091e-05,
"loss": 0.0393,
"step": 11200
},
{
"epoch": 43.11538461538461,
"grad_norm": 1.2205157279968262,
"learning_rate": 1.3245559878667405e-05,
"loss": 0.0428,
"step": 11210
},
{
"epoch": 43.15384615384615,
"grad_norm": 1.2515414953231812,
"learning_rate": 1.3220930474445019e-05,
"loss": 0.0343,
"step": 11220
},
{
"epoch": 43.19230769230769,
"grad_norm": 1.3750489950180054,
"learning_rate": 1.3196305934130778e-05,
"loss": 0.0401,
"step": 11230
},
{
"epoch": 43.23076923076923,
"grad_norm": 1.156083583831787,
"learning_rate": 1.3171686325047241e-05,
"loss": 0.0327,
"step": 11240
},
{
"epoch": 43.26923076923077,
"grad_norm": 1.4203742742538452,
"learning_rate": 1.3147071714503484e-05,
"loss": 0.0399,
"step": 11250
},
{
"epoch": 43.30769230769231,
"grad_norm": 1.0962517261505127,
"learning_rate": 1.3122462169794903e-05,
"loss": 0.0337,
"step": 11260
},
{
"epoch": 43.34615384615385,
"grad_norm": 1.2661428451538086,
"learning_rate": 1.3097857758203053e-05,
"loss": 0.0329,
"step": 11270
},
{
"epoch": 43.38461538461539,
"grad_norm": 1.1555863618850708,
"learning_rate": 1.3073258546995455e-05,
"loss": 0.0344,
"step": 11280
},
{
"epoch": 43.42307692307692,
"grad_norm": 1.220564842224121,
"learning_rate": 1.3048664603425429e-05,
"loss": 0.0366,
"step": 11290
},
{
"epoch": 43.46153846153846,
"grad_norm": 1.1790637969970703,
"learning_rate": 1.3024075994731859e-05,
"loss": 0.0363,
"step": 11300
},
{
"epoch": 43.5,
"grad_norm": 1.1303735971450806,
"learning_rate": 1.2999492788139068e-05,
"loss": 0.0337,
"step": 11310
},
{
"epoch": 43.53846153846154,
"grad_norm": 0.850884199142456,
"learning_rate": 1.2974915050856605e-05,
"loss": 0.0315,
"step": 11320
},
{
"epoch": 43.57692307692308,
"grad_norm": 1.1740418672561646,
"learning_rate": 1.2950342850079061e-05,
"loss": 0.0371,
"step": 11330
},
{
"epoch": 43.61538461538461,
"grad_norm": 1.2208044528961182,
"learning_rate": 1.292577625298591e-05,
"loss": 0.0381,
"step": 11340
},
{
"epoch": 43.65384615384615,
"grad_norm": 1.4097691774368286,
"learning_rate": 1.2901215326741273e-05,
"loss": 0.0321,
"step": 11350
},
{
"epoch": 43.69230769230769,
"grad_norm": 1.2222874164581299,
"learning_rate": 1.287666013849379e-05,
"loss": 0.0324,
"step": 11360
},
{
"epoch": 43.73076923076923,
"grad_norm": 1.394254207611084,
"learning_rate": 1.285211075537641e-05,
"loss": 0.0391,
"step": 11370
},
{
"epoch": 43.76923076923077,
"grad_norm": 1.0281349420547485,
"learning_rate": 1.2827567244506203e-05,
"loss": 0.0314,
"step": 11380
},
{
"epoch": 43.80769230769231,
"grad_norm": 1.1789246797561646,
"learning_rate": 1.2803029672984208e-05,
"loss": 0.0315,
"step": 11390
},
{
"epoch": 43.84615384615385,
"grad_norm": 1.0867520570755005,
"learning_rate": 1.2778498107895186e-05,
"loss": 0.0295,
"step": 11400
},
{
"epoch": 43.88461538461539,
"grad_norm": 1.2576417922973633,
"learning_rate": 1.275397261630751e-05,
"loss": 0.0364,
"step": 11410
},
{
"epoch": 43.92307692307692,
"grad_norm": 1.128233551979065,
"learning_rate": 1.2729453265272935e-05,
"loss": 0.0372,
"step": 11420
},
{
"epoch": 43.96153846153846,
"grad_norm": 1.1675398349761963,
"learning_rate": 1.270494012182644e-05,
"loss": 0.0333,
"step": 11430
},
{
"epoch": 44.0,
"grad_norm": 1.2023512125015259,
"learning_rate": 1.268043325298601e-05,
"loss": 0.0329,
"step": 11440
},
{
"epoch": 44.03846153846154,
"grad_norm": 1.3402878046035767,
"learning_rate": 1.2655932725752494e-05,
"loss": 0.0367,
"step": 11450
},
{
"epoch": 44.07692307692308,
"grad_norm": 1.0831363201141357,
"learning_rate": 1.26314386071094e-05,
"loss": 0.043,
"step": 11460
},
{
"epoch": 44.11538461538461,
"grad_norm": 1.265649676322937,
"learning_rate": 1.2606950964022701e-05,
"loss": 0.0376,
"step": 11470
},
{
"epoch": 44.15384615384615,
"grad_norm": 1.0680443048477173,
"learning_rate": 1.2582469863440704e-05,
"loss": 0.0344,
"step": 11480
},
{
"epoch": 44.19230769230769,
"grad_norm": 1.0312443971633911,
"learning_rate": 1.2557995372293778e-05,
"loss": 0.0308,
"step": 11490
},
{
"epoch": 44.23076923076923,
"grad_norm": 1.2013682126998901,
"learning_rate": 1.2533527557494257e-05,
"loss": 0.0361,
"step": 11500
},
{
"epoch": 44.26923076923077,
"grad_norm": 1.2492719888687134,
"learning_rate": 1.250906648593621e-05,
"loss": 0.0326,
"step": 11510
},
{
"epoch": 44.30769230769231,
"grad_norm": 0.8632230162620544,
"learning_rate": 1.2484612224495275e-05,
"loss": 0.0342,
"step": 11520
},
{
"epoch": 44.34615384615385,
"grad_norm": 1.1540902853012085,
"learning_rate": 1.2460164840028477e-05,
"loss": 0.0365,
"step": 11530
},
{
"epoch": 44.38461538461539,
"grad_norm": 1.1001176834106445,
"learning_rate": 1.2435724399374016e-05,
"loss": 0.0311,
"step": 11540
},
{
"epoch": 44.42307692307692,
"grad_norm": 1.1563735008239746,
"learning_rate": 1.2411290969351129e-05,
"loss": 0.0363,
"step": 11550
},
{
"epoch": 44.46153846153846,
"grad_norm": 1.2087491750717163,
"learning_rate": 1.2386864616759883e-05,
"loss": 0.0348,
"step": 11560
},
{
"epoch": 44.5,
"grad_norm": 1.2683261632919312,
"learning_rate": 1.2362445408380996e-05,
"loss": 0.0329,
"step": 11570
},
{
"epoch": 44.53846153846154,
"grad_norm": 1.2558414936065674,
"learning_rate": 1.2338033410975644e-05,
"loss": 0.0303,
"step": 11580
},
{
"epoch": 44.57692307692308,
"grad_norm": 1.2375128269195557,
"learning_rate": 1.2313628691285301e-05,
"loss": 0.0362,
"step": 11590
},
{
"epoch": 44.61538461538461,
"grad_norm": 1.0311369895935059,
"learning_rate": 1.2289231316031536e-05,
"loss": 0.033,
"step": 11600
},
{
"epoch": 44.65384615384615,
"grad_norm": 1.0355188846588135,
"learning_rate": 1.2264841351915842e-05,
"loss": 0.0335,
"step": 11610
},
{
"epoch": 44.69230769230769,
"grad_norm": 1.3842549324035645,
"learning_rate": 1.2240458865619455e-05,
"loss": 0.0371,
"step": 11620
},
{
"epoch": 44.73076923076923,
"grad_norm": 1.252090573310852,
"learning_rate": 1.2216083923803152e-05,
"loss": 0.0316,
"step": 11630
},
{
"epoch": 44.76923076923077,
"grad_norm": 1.099759817123413,
"learning_rate": 1.2191716593107097e-05,
"loss": 0.036,
"step": 11640
},
{
"epoch": 44.80769230769231,
"grad_norm": 1.2019200325012207,
"learning_rate": 1.2167356940150645e-05,
"loss": 0.0329,
"step": 11650
},
{
"epoch": 44.84615384615385,
"grad_norm": 1.1980955600738525,
"learning_rate": 1.2143005031532152e-05,
"loss": 0.0317,
"step": 11660
},
{
"epoch": 44.88461538461539,
"grad_norm": 1.3130910396575928,
"learning_rate": 1.2118660933828813e-05,
"loss": 0.0357,
"step": 11670
},
{
"epoch": 44.92307692307692,
"grad_norm": 1.326754093170166,
"learning_rate": 1.2094324713596453e-05,
"loss": 0.0317,
"step": 11680
},
{
"epoch": 44.96153846153846,
"grad_norm": 1.204189419746399,
"learning_rate": 1.2069996437369374e-05,
"loss": 0.0355,
"step": 11690
},
{
"epoch": 45.0,
"grad_norm": 0.9756190180778503,
"learning_rate": 1.2045676171660154e-05,
"loss": 0.035,
"step": 11700
},
{
"epoch": 45.03846153846154,
"grad_norm": 1.243010401725769,
"learning_rate": 1.2021363982959472e-05,
"loss": 0.0399,
"step": 11710
},
{
"epoch": 45.07692307692308,
"grad_norm": 1.2061409950256348,
"learning_rate": 1.1997059937735919e-05,
"loss": 0.0308,
"step": 11720
},
{
"epoch": 45.11538461538461,
"grad_norm": 1.214267611503601,
"learning_rate": 1.197276410243583e-05,
"loss": 0.0321,
"step": 11730
},
{
"epoch": 45.15384615384615,
"grad_norm": 1.2562459707260132,
"learning_rate": 1.194847654348309e-05,
"loss": 0.0322,
"step": 11740
},
{
"epoch": 45.19230769230769,
"grad_norm": 1.1353240013122559,
"learning_rate": 1.1924197327278957e-05,
"loss": 0.0334,
"step": 11750
},
{
"epoch": 45.23076923076923,
"grad_norm": 1.0406856536865234,
"learning_rate": 1.1899926520201885e-05,
"loss": 0.0348,
"step": 11760
},
{
"epoch": 45.26923076923077,
"grad_norm": 1.4781416654586792,
"learning_rate": 1.1875664188607327e-05,
"loss": 0.0287,
"step": 11770
},
{
"epoch": 45.30769230769231,
"grad_norm": 1.0640053749084473,
"learning_rate": 1.1851410398827578e-05,
"loss": 0.0311,
"step": 11780
},
{
"epoch": 45.34615384615385,
"grad_norm": 1.1978460550308228,
"learning_rate": 1.1827165217171567e-05,
"loss": 0.0345,
"step": 11790
},
{
"epoch": 45.38461538461539,
"grad_norm": 1.0306118726730347,
"learning_rate": 1.18029287099247e-05,
"loss": 0.0328,
"step": 11800
},
{
"epoch": 45.42307692307692,
"grad_norm": 1.2475486993789673,
"learning_rate": 1.1778700943348662e-05,
"loss": 0.0321,
"step": 11810
},
{
"epoch": 45.46153846153846,
"grad_norm": 1.11435067653656,
"learning_rate": 1.1754481983681238e-05,
"loss": 0.0371,
"step": 11820
},
{
"epoch": 45.5,
"grad_norm": 1.4166765213012695,
"learning_rate": 1.173027189713614e-05,
"loss": 0.0348,
"step": 11830
},
{
"epoch": 45.53846153846154,
"grad_norm": 1.6289052963256836,
"learning_rate": 1.170607074990282e-05,
"loss": 0.0305,
"step": 11840
},
{
"epoch": 45.57692307692308,
"grad_norm": 1.254708170890808,
"learning_rate": 1.1681878608146297e-05,
"loss": 0.0345,
"step": 11850
},
{
"epoch": 45.61538461538461,
"grad_norm": 1.2652217149734497,
"learning_rate": 1.1657695538006952e-05,
"loss": 0.0389,
"step": 11860
},
{
"epoch": 45.65384615384615,
"grad_norm": 1.1574680805206299,
"learning_rate": 1.163352160560038e-05,
"loss": 0.0382,
"step": 11870
},
{
"epoch": 45.69230769230769,
"grad_norm": 1.2947102785110474,
"learning_rate": 1.1609356877017191e-05,
"loss": 0.0323,
"step": 11880
},
{
"epoch": 45.73076923076923,
"grad_norm": 1.23800790309906,
"learning_rate": 1.1585201418322828e-05,
"loss": 0.0364,
"step": 11890
},
{
"epoch": 45.76923076923077,
"grad_norm": 1.157383918762207,
"learning_rate": 1.1561055295557397e-05,
"loss": 0.0365,
"step": 11900
},
{
"epoch": 45.80769230769231,
"grad_norm": 1.0947489738464355,
"learning_rate": 1.1536918574735469e-05,
"loss": 0.0347,
"step": 11910
},
{
"epoch": 45.84615384615385,
"grad_norm": 1.083691120147705,
"learning_rate": 1.1512791321845921e-05,
"loss": 0.033,
"step": 11920
},
{
"epoch": 45.88461538461539,
"grad_norm": 0.9972357749938965,
"learning_rate": 1.148867360285174e-05,
"loss": 0.0296,
"step": 11930
},
{
"epoch": 45.92307692307692,
"grad_norm": 1.1867698431015015,
"learning_rate": 1.1464565483689853e-05,
"loss": 0.0357,
"step": 11940
},
{
"epoch": 45.96153846153846,
"grad_norm": 1.1101211309432983,
"learning_rate": 1.144046703027093e-05,
"loss": 0.0318,
"step": 11950
},
{
"epoch": 46.0,
"grad_norm": 0.9498093128204346,
"learning_rate": 1.1416378308479223e-05,
"loss": 0.0326,
"step": 11960
},
{
"epoch": 46.03846153846154,
"grad_norm": 0.9711620807647705,
"learning_rate": 1.1392299384172383e-05,
"loss": 0.0354,
"step": 11970
},
{
"epoch": 46.07692307692308,
"grad_norm": 1.2946242094039917,
"learning_rate": 1.1368230323181267e-05,
"loss": 0.0331,
"step": 11980
},
{
"epoch": 46.11538461538461,
"grad_norm": 1.1568892002105713,
"learning_rate": 1.1344171191309772e-05,
"loss": 0.032,
"step": 11990
},
{
"epoch": 46.15384615384615,
"grad_norm": 1.16463303565979,
"learning_rate": 1.1320122054334636e-05,
"loss": 0.0314,
"step": 12000
},
{
"epoch": 46.19230769230769,
"grad_norm": 1.1152621507644653,
"learning_rate": 1.1296082978005292e-05,
"loss": 0.0357,
"step": 12010
},
{
"epoch": 46.23076923076923,
"grad_norm": 1.2109414339065552,
"learning_rate": 1.127205402804365e-05,
"loss": 0.0385,
"step": 12020
},
{
"epoch": 46.26923076923077,
"grad_norm": 1.1345036029815674,
"learning_rate": 1.1248035270143946e-05,
"loss": 0.0335,
"step": 12030
},
{
"epoch": 46.30769230769231,
"grad_norm": 0.9837266802787781,
"learning_rate": 1.1224026769972545e-05,
"loss": 0.0303,
"step": 12040
},
{
"epoch": 46.34615384615385,
"grad_norm": 1.2618961334228516,
"learning_rate": 1.1200028593167769e-05,
"loss": 0.0343,
"step": 12050
},
{
"epoch": 46.38461538461539,
"grad_norm": 1.2574706077575684,
"learning_rate": 1.1176040805339718e-05,
"loss": 0.0371,
"step": 12060
},
{
"epoch": 46.42307692307692,
"grad_norm": 1.127816081047058,
"learning_rate": 1.1152063472070086e-05,
"loss": 0.0335,
"step": 12070
},
{
"epoch": 46.46153846153846,
"grad_norm": 1.2891298532485962,
"learning_rate": 1.1128096658911992e-05,
"loss": 0.0313,
"step": 12080
},
{
"epoch": 46.5,
"grad_norm": 1.1167298555374146,
"learning_rate": 1.1104140431389782e-05,
"loss": 0.0291,
"step": 12090
},
{
"epoch": 46.53846153846154,
"grad_norm": 1.112906575202942,
"learning_rate": 1.1080194854998868e-05,
"loss": 0.0357,
"step": 12100
},
{
"epoch": 46.57692307692308,
"grad_norm": 1.082861304283142,
"learning_rate": 1.1056259995205545e-05,
"loss": 0.0323,
"step": 12110
},
{
"epoch": 46.61538461538461,
"grad_norm": 1.2894811630249023,
"learning_rate": 1.1032335917446803e-05,
"loss": 0.0348,
"step": 12120
},
{
"epoch": 46.65384615384615,
"grad_norm": 1.0083926916122437,
"learning_rate": 1.100842268713016e-05,
"loss": 0.0316,
"step": 12130
},
{
"epoch": 46.69230769230769,
"grad_norm": 1.1639894247055054,
"learning_rate": 1.098452036963347e-05,
"loss": 0.0332,
"step": 12140
},
{
"epoch": 46.73076923076923,
"grad_norm": 1.2733254432678223,
"learning_rate": 1.096062903030476e-05,
"loss": 0.0353,
"step": 12150
},
{
"epoch": 46.76923076923077,
"grad_norm": 1.142136812210083,
"learning_rate": 1.0936748734462036e-05,
"loss": 0.0402,
"step": 12160
},
{
"epoch": 46.80769230769231,
"grad_norm": 1.0598315000534058,
"learning_rate": 1.0912879547393119e-05,
"loss": 0.033,
"step": 12170
},
{
"epoch": 46.84615384615385,
"grad_norm": 0.8986493945121765,
"learning_rate": 1.0889021534355456e-05,
"loss": 0.0361,
"step": 12180
},
{
"epoch": 46.88461538461539,
"grad_norm": 1.1551433801651,
"learning_rate": 1.0865174760575936e-05,
"loss": 0.0367,
"step": 12190
},
{
"epoch": 46.92307692307692,
"grad_norm": 1.0737844705581665,
"learning_rate": 1.0841339291250733e-05,
"loss": 0.0295,
"step": 12200
},
{
"epoch": 46.96153846153846,
"grad_norm": 1.2422358989715576,
"learning_rate": 1.081751519154511e-05,
"loss": 0.0379,
"step": 12210
},
{
"epoch": 47.0,
"grad_norm": 1.1002851724624634,
"learning_rate": 1.079370252659325e-05,
"loss": 0.0335,
"step": 12220
},
{
"epoch": 47.03846153846154,
"grad_norm": 1.1535612344741821,
"learning_rate": 1.076990136149806e-05,
"loss": 0.0389,
"step": 12230
},
{
"epoch": 47.07692307692308,
"grad_norm": 1.1302235126495361,
"learning_rate": 1.0746111761331021e-05,
"loss": 0.0317,
"step": 12240
},
{
"epoch": 47.11538461538461,
"grad_norm": 1.2010834217071533,
"learning_rate": 1.0722333791131996e-05,
"loss": 0.0329,
"step": 12250
},
{
"epoch": 47.15384615384615,
"grad_norm": 1.2279101610183716,
"learning_rate": 1.0698567515909041e-05,
"loss": 0.0352,
"step": 12260
},
{
"epoch": 47.19230769230769,
"grad_norm": 1.3264673948287964,
"learning_rate": 1.0674813000638252e-05,
"loss": 0.0338,
"step": 12270
},
{
"epoch": 47.23076923076923,
"grad_norm": 1.1362801790237427,
"learning_rate": 1.0651070310263559e-05,
"loss": 0.0333,
"step": 12280
},
{
"epoch": 47.26923076923077,
"grad_norm": 1.2410087585449219,
"learning_rate": 1.0627339509696574e-05,
"loss": 0.0302,
"step": 12290
},
{
"epoch": 47.30769230769231,
"grad_norm": 0.9610795378684998,
"learning_rate": 1.06036206638164e-05,
"loss": 0.0318,
"step": 12300
},
{
"epoch": 47.34615384615385,
"grad_norm": 0.9377633929252625,
"learning_rate": 1.0579913837469455e-05,
"loss": 0.0313,
"step": 12310
},
{
"epoch": 47.38461538461539,
"grad_norm": 1.0323609113693237,
"learning_rate": 1.0556219095469303e-05,
"loss": 0.0322,
"step": 12320
},
{
"epoch": 47.42307692307692,
"grad_norm": 0.9905584454536438,
"learning_rate": 1.0532536502596455e-05,
"loss": 0.0304,
"step": 12330
},
{
"epoch": 47.46153846153846,
"grad_norm": 1.1498249769210815,
"learning_rate": 1.0508866123598218e-05,
"loss": 0.0315,
"step": 12340
},
{
"epoch": 47.5,
"grad_norm": 0.8419849872589111,
"learning_rate": 1.0485208023188505e-05,
"loss": 0.0331,
"step": 12350
},
{
"epoch": 47.53846153846154,
"grad_norm": 1.1979166269302368,
"learning_rate": 1.0461562266047668e-05,
"loss": 0.0295,
"step": 12360
},
{
"epoch": 47.57692307692308,
"grad_norm": 1.025926947593689,
"learning_rate": 1.0437928916822286e-05,
"loss": 0.0285,
"step": 12370
},
{
"epoch": 47.61538461538461,
"grad_norm": 1.1680309772491455,
"learning_rate": 1.0414308040125043e-05,
"loss": 0.0391,
"step": 12380
},
{
"epoch": 47.65384615384615,
"grad_norm": 1.0929526090621948,
"learning_rate": 1.0390699700534517e-05,
"loss": 0.0306,
"step": 12390
},
{
"epoch": 47.69230769230769,
"grad_norm": 1.0418400764465332,
"learning_rate": 1.0367103962595003e-05,
"loss": 0.0375,
"step": 12400
},
{
"epoch": 47.73076923076923,
"grad_norm": 1.1047964096069336,
"learning_rate": 1.0343520890816356e-05,
"loss": 0.0352,
"step": 12410
},
{
"epoch": 47.76923076923077,
"grad_norm": 1.0839290618896484,
"learning_rate": 1.0319950549673779e-05,
"loss": 0.0346,
"step": 12420
},
{
"epoch": 47.80769230769231,
"grad_norm": 1.1870118379592896,
"learning_rate": 1.0296393003607692e-05,
"loss": 0.0339,
"step": 12430
},
{
"epoch": 47.84615384615385,
"grad_norm": 1.1544498205184937,
"learning_rate": 1.0272848317023526e-05,
"loss": 0.0316,
"step": 12440
},
{
"epoch": 47.88461538461539,
"grad_norm": 1.0045503377914429,
"learning_rate": 1.0249316554291556e-05,
"loss": 0.0295,
"step": 12450
},
{
"epoch": 47.92307692307692,
"grad_norm": 1.0486705303192139,
"learning_rate": 1.022579777974673e-05,
"loss": 0.0348,
"step": 12460
},
{
"epoch": 47.96153846153846,
"grad_norm": 1.0861692428588867,
"learning_rate": 1.0202292057688462e-05,
"loss": 0.031,
"step": 12470
},
{
"epoch": 48.0,
"grad_norm": 0.9276844263076782,
"learning_rate": 1.0178799452380511e-05,
"loss": 0.0356,
"step": 12480
},
{
"epoch": 48.03846153846154,
"grad_norm": 1.1706113815307617,
"learning_rate": 1.0155320028050757e-05,
"loss": 0.0316,
"step": 12490
},
{
"epoch": 48.07692307692308,
"grad_norm": 0.9763121008872986,
"learning_rate": 1.0131853848891063e-05,
"loss": 0.0318,
"step": 12500
},
{
"epoch": 48.11538461538461,
"grad_norm": 0.9173984527587891,
"learning_rate": 1.0108400979057048e-05,
"loss": 0.03,
"step": 12510
},
{
"epoch": 48.15384615384615,
"grad_norm": 0.9412733316421509,
"learning_rate": 1.008496148266797e-05,
"loss": 0.0302,
"step": 12520
},
{
"epoch": 48.19230769230769,
"grad_norm": 1.0302488803863525,
"learning_rate": 1.0061535423806519e-05,
"loss": 0.0299,
"step": 12530
},
{
"epoch": 48.23076923076923,
"grad_norm": 1.2443172931671143,
"learning_rate": 1.0038122866518647e-05,
"loss": 0.0306,
"step": 12540
},
{
"epoch": 48.26923076923077,
"grad_norm": 1.1415942907333374,
"learning_rate": 1.0014723874813394e-05,
"loss": 0.0327,
"step": 12550
},
{
"epoch": 48.30769230769231,
"grad_norm": 1.1814067363739014,
"learning_rate": 9.991338512662696e-06,
"loss": 0.0283,
"step": 12560
},
{
"epoch": 48.34615384615385,
"grad_norm": 1.1458048820495605,
"learning_rate": 9.96796684400125e-06,
"loss": 0.0351,
"step": 12570
},
{
"epoch": 48.38461538461539,
"grad_norm": 1.0160274505615234,
"learning_rate": 9.944608932726306e-06,
"loss": 0.0329,
"step": 12580
},
{
"epoch": 48.42307692307692,
"grad_norm": 1.0080182552337646,
"learning_rate": 9.921264842697501e-06,
"loss": 0.0303,
"step": 12590
},
{
"epoch": 48.46153846153846,
"grad_norm": 1.039967656135559,
"learning_rate": 9.897934637736692e-06,
"loss": 0.034,
"step": 12600
},
{
"epoch": 48.5,
"grad_norm": 1.1883785724639893,
"learning_rate": 9.874618381627751e-06,
"loss": 0.0298,
"step": 12610
},
{
"epoch": 48.53846153846154,
"grad_norm": 0.9066627621650696,
"learning_rate": 9.851316138116446e-06,
"loss": 0.0317,
"step": 12620
},
{
"epoch": 48.57692307692308,
"grad_norm": 1.3488620519638062,
"learning_rate": 9.828027970910217e-06,
"loss": 0.034,
"step": 12630
},
{
"epoch": 48.61538461538461,
"grad_norm": 1.255429744720459,
"learning_rate": 9.80475394367803e-06,
"loss": 0.031,
"step": 12640
},
{
"epoch": 48.65384615384615,
"grad_norm": 1.0005918741226196,
"learning_rate": 9.781494120050176e-06,
"loss": 0.0319,
"step": 12650
},
{
"epoch": 48.69230769230769,
"grad_norm": 1.221487045288086,
"learning_rate": 9.758248563618126e-06,
"loss": 0.0339,
"step": 12660
},
{
"epoch": 48.73076923076923,
"grad_norm": 1.1094986200332642,
"learning_rate": 9.735017337934349e-06,
"loss": 0.0351,
"step": 12670
},
{
"epoch": 48.76923076923077,
"grad_norm": 0.8227080702781677,
"learning_rate": 9.71180050651213e-06,
"loss": 0.0284,
"step": 12680
},
{
"epoch": 48.80769230769231,
"grad_norm": 0.9973031282424927,
"learning_rate": 9.688598132825402e-06,
"loss": 0.039,
"step": 12690
},
{
"epoch": 48.84615384615385,
"grad_norm": 1.0556597709655762,
"learning_rate": 9.665410280308555e-06,
"loss": 0.037,
"step": 12700
},
{
"epoch": 48.88461538461539,
"grad_norm": 0.8174852132797241,
"learning_rate": 9.642237012356302e-06,
"loss": 0.0278,
"step": 12710
},
{
"epoch": 48.92307692307692,
"grad_norm": 1.2452213764190674,
"learning_rate": 9.619078392323471e-06,
"loss": 0.0337,
"step": 12720
},
{
"epoch": 48.96153846153846,
"grad_norm": 1.3187814950942993,
"learning_rate": 9.595934483524847e-06,
"loss": 0.0317,
"step": 12730
},
{
"epoch": 49.0,
"grad_norm": 1.147231936454773,
"learning_rate": 9.572805349234997e-06,
"loss": 0.0321,
"step": 12740
},
{
"epoch": 49.03846153846154,
"grad_norm": 0.9327066540718079,
"learning_rate": 9.54969105268808e-06,
"loss": 0.0309,
"step": 12750
},
{
"epoch": 49.07692307692308,
"grad_norm": 1.0206278562545776,
"learning_rate": 9.526591657077701e-06,
"loss": 0.0334,
"step": 12760
},
{
"epoch": 49.11538461538461,
"grad_norm": 0.7847154140472412,
"learning_rate": 9.503507225556734e-06,
"loss": 0.0301,
"step": 12770
},
{
"epoch": 49.15384615384615,
"grad_norm": 0.9998318552970886,
"learning_rate": 9.480437821237134e-06,
"loss": 0.0379,
"step": 12780
},
{
"epoch": 49.19230769230769,
"grad_norm": 1.1885015964508057,
"learning_rate": 9.457383507189763e-06,
"loss": 0.0367,
"step": 12790
},
{
"epoch": 49.23076923076923,
"grad_norm": 0.9944354295730591,
"learning_rate": 9.434344346444237e-06,
"loss": 0.0355,
"step": 12800
},
{
"epoch": 49.26923076923077,
"grad_norm": 1.2915024757385254,
"learning_rate": 9.411320401988744e-06,
"loss": 0.0369,
"step": 12810
},
{
"epoch": 49.30769230769231,
"grad_norm": 1.0678672790527344,
"learning_rate": 9.388311736769867e-06,
"loss": 0.0354,
"step": 12820
},
{
"epoch": 49.34615384615385,
"grad_norm": 1.4048540592193604,
"learning_rate": 9.365318413692429e-06,
"loss": 0.0328,
"step": 12830
},
{
"epoch": 49.38461538461539,
"grad_norm": 1.215981364250183,
"learning_rate": 9.34234049561928e-06,
"loss": 0.0337,
"step": 12840
},
{
"epoch": 49.42307692307692,
"grad_norm": 1.0116572380065918,
"learning_rate": 9.31937804537118e-06,
"loss": 0.0298,
"step": 12850
},
{
"epoch": 49.46153846153846,
"grad_norm": 1.2934813499450684,
"learning_rate": 9.296431125726587e-06,
"loss": 0.0328,
"step": 12860
},
{
"epoch": 49.5,
"grad_norm": 1.2118585109710693,
"learning_rate": 9.27349979942151e-06,
"loss": 0.0332,
"step": 12870
},
{
"epoch": 49.53846153846154,
"grad_norm": 1.0535308122634888,
"learning_rate": 9.250584129149321e-06,
"loss": 0.0326,
"step": 12880
},
{
"epoch": 49.57692307692308,
"grad_norm": 1.0874793529510498,
"learning_rate": 9.227684177560575e-06,
"loss": 0.0328,
"step": 12890
},
{
"epoch": 49.61538461538461,
"grad_norm": 0.9859468936920166,
"learning_rate": 9.204800007262874e-06,
"loss": 0.0298,
"step": 12900
},
{
"epoch": 49.65384615384615,
"grad_norm": 0.9996634721755981,
"learning_rate": 9.18193168082066e-06,
"loss": 0.0307,
"step": 12910
},
{
"epoch": 49.69230769230769,
"grad_norm": 1.0833901166915894,
"learning_rate": 9.159079260755079e-06,
"loss": 0.0357,
"step": 12920
},
{
"epoch": 49.73076923076923,
"grad_norm": 1.2677006721496582,
"learning_rate": 9.136242809543754e-06,
"loss": 0.0325,
"step": 12930
},
{
"epoch": 49.76923076923077,
"grad_norm": 1.1370261907577515,
"learning_rate": 9.113422389620685e-06,
"loss": 0.0269,
"step": 12940
},
{
"epoch": 49.80769230769231,
"grad_norm": 0.9824889898300171,
"learning_rate": 9.090618063376021e-06,
"loss": 0.033,
"step": 12950
},
{
"epoch": 49.84615384615385,
"grad_norm": 1.1071754693984985,
"learning_rate": 9.067829893155922e-06,
"loss": 0.0324,
"step": 12960
},
{
"epoch": 49.88461538461539,
"grad_norm": 1.0568147897720337,
"learning_rate": 9.045057941262384e-06,
"loss": 0.0341,
"step": 12970
},
{
"epoch": 49.92307692307692,
"grad_norm": 1.1383357048034668,
"learning_rate": 9.02230226995304e-06,
"loss": 0.034,
"step": 12980
},
{
"epoch": 49.96153846153846,
"grad_norm": 1.0124088525772095,
"learning_rate": 8.999562941441031e-06,
"loss": 0.0329,
"step": 12990
},
{
"epoch": 50.0,
"grad_norm": 1.2345850467681885,
"learning_rate": 8.976840017894814e-06,
"loss": 0.0346,
"step": 13000
},
{
"epoch": 50.03846153846154,
"grad_norm": 1.0567175149917603,
"learning_rate": 8.954133561437993e-06,
"loss": 0.0362,
"step": 13010
},
{
"epoch": 50.07692307692308,
"grad_norm": 1.0732157230377197,
"learning_rate": 8.931443634149163e-06,
"loss": 0.0321,
"step": 13020
},
{
"epoch": 50.11538461538461,
"grad_norm": 1.1036709547042847,
"learning_rate": 8.908770298061702e-06,
"loss": 0.0335,
"step": 13030
},
{
"epoch": 50.15384615384615,
"grad_norm": 0.9429640173912048,
"learning_rate": 8.886113615163655e-06,
"loss": 0.0304,
"step": 13040
},
{
"epoch": 50.19230769230769,
"grad_norm": 0.9867258667945862,
"learning_rate": 8.863473647397522e-06,
"loss": 0.0319,
"step": 13050
},
{
"epoch": 50.23076923076923,
"grad_norm": 0.8746216297149658,
"learning_rate": 8.840850456660121e-06,
"loss": 0.0342,
"step": 13060
},
{
"epoch": 50.26923076923077,
"grad_norm": 0.8380419611930847,
"learning_rate": 8.818244104802384e-06,
"loss": 0.0246,
"step": 13070
},
{
"epoch": 50.30769230769231,
"grad_norm": 1.0145636796951294,
"learning_rate": 8.79565465362921e-06,
"loss": 0.0308,
"step": 13080
},
{
"epoch": 50.34615384615385,
"grad_norm": 1.0411081314086914,
"learning_rate": 8.773082164899305e-06,
"loss": 0.0334,
"step": 13090
},
{
"epoch": 50.38461538461539,
"grad_norm": 1.0380659103393555,
"learning_rate": 8.750526700324984e-06,
"loss": 0.0332,
"step": 13100
},
{
"epoch": 50.42307692307692,
"grad_norm": 1.0183615684509277,
"learning_rate": 8.72798832157203e-06,
"loss": 0.0321,
"step": 13110
},
{
"epoch": 50.46153846153846,
"grad_norm": 1.1746301651000977,
"learning_rate": 8.705467090259507e-06,
"loss": 0.0348,
"step": 13120
},
{
"epoch": 50.5,
"grad_norm": 1.037621259689331,
"learning_rate": 8.682963067959607e-06,
"loss": 0.0342,
"step": 13130
},
{
"epoch": 50.53846153846154,
"grad_norm": 1.1670657396316528,
"learning_rate": 8.660476316197457e-06,
"loss": 0.0306,
"step": 13140
},
{
"epoch": 50.57692307692308,
"grad_norm": 0.877946674823761,
"learning_rate": 8.638006896450991e-06,
"loss": 0.0302,
"step": 13150
},
{
"epoch": 50.61538461538461,
"grad_norm": 0.8860836625099182,
"learning_rate": 8.61555487015074e-06,
"loss": 0.0261,
"step": 13160
},
{
"epoch": 50.65384615384615,
"grad_norm": 0.7965879440307617,
"learning_rate": 8.593120298679676e-06,
"loss": 0.0337,
"step": 13170
},
{
"epoch": 50.69230769230769,
"grad_norm": 0.8773680329322815,
"learning_rate": 8.570703243373076e-06,
"loss": 0.0311,
"step": 13180
},
{
"epoch": 50.73076923076923,
"grad_norm": 0.9037876725196838,
"learning_rate": 8.5483037655183e-06,
"loss": 0.0278,
"step": 13190
},
{
"epoch": 50.76923076923077,
"grad_norm": 1.0279417037963867,
"learning_rate": 8.525921926354686e-06,
"loss": 0.0396,
"step": 13200
},
{
"epoch": 50.80769230769231,
"grad_norm": 1.109102487564087,
"learning_rate": 8.5035577870733e-06,
"loss": 0.0347,
"step": 13210
},
{
"epoch": 50.84615384615385,
"grad_norm": 1.1594183444976807,
"learning_rate": 8.481211408816858e-06,
"loss": 0.0339,
"step": 13220
},
{
"epoch": 50.88461538461539,
"grad_norm": 1.18452787399292,
"learning_rate": 8.45888285267951e-06,
"loss": 0.0344,
"step": 13230
},
{
"epoch": 50.92307692307692,
"grad_norm": 0.88536137342453,
"learning_rate": 8.436572179706666e-06,
"loss": 0.0358,
"step": 13240
},
{
"epoch": 50.96153846153846,
"grad_norm": 1.2674336433410645,
"learning_rate": 8.41427945089487e-06,
"loss": 0.0337,
"step": 13250
},
{
"epoch": 51.0,
"grad_norm": 0.9828641414642334,
"learning_rate": 8.39200472719157e-06,
"loss": 0.0371,
"step": 13260
},
{
"epoch": 51.03846153846154,
"grad_norm": 0.9601910710334778,
"learning_rate": 8.369748069495017e-06,
"loss": 0.0387,
"step": 13270
},
{
"epoch": 51.07692307692308,
"grad_norm": 1.1602641344070435,
"learning_rate": 8.347509538654074e-06,
"loss": 0.0327,
"step": 13280
},
{
"epoch": 51.11538461538461,
"grad_norm": 0.9831449389457703,
"learning_rate": 8.325289195468023e-06,
"loss": 0.0359,
"step": 13290
},
{
"epoch": 51.15384615384615,
"grad_norm": 1.0050350427627563,
"learning_rate": 8.303087100686449e-06,
"loss": 0.0377,
"step": 13300
},
{
"epoch": 51.19230769230769,
"grad_norm": 0.9317255616188049,
"learning_rate": 8.280903315009009e-06,
"loss": 0.0281,
"step": 13310
},
{
"epoch": 51.23076923076923,
"grad_norm": 1.0531281232833862,
"learning_rate": 8.258737899085338e-06,
"loss": 0.0324,
"step": 13320
},
{
"epoch": 51.26923076923077,
"grad_norm": 0.9911993741989136,
"learning_rate": 8.236590913514841e-06,
"loss": 0.0346,
"step": 13330
},
{
"epoch": 51.30769230769231,
"grad_norm": 0.8987839221954346,
"learning_rate": 8.214462418846529e-06,
"loss": 0.0314,
"step": 13340
},
{
"epoch": 51.34615384615385,
"grad_norm": 0.8279987573623657,
"learning_rate": 8.192352475578857e-06,
"loss": 0.0273,
"step": 13350
},
{
"epoch": 51.38461538461539,
"grad_norm": 0.9038864970207214,
"learning_rate": 8.170261144159563e-06,
"loss": 0.0316,
"step": 13360
},
{
"epoch": 51.42307692307692,
"grad_norm": 0.9167174100875854,
"learning_rate": 8.148188484985505e-06,
"loss": 0.0298,
"step": 13370
},
{
"epoch": 51.46153846153846,
"grad_norm": 1.0757681131362915,
"learning_rate": 8.126134558402501e-06,
"loss": 0.0327,
"step": 13380
},
{
"epoch": 51.5,
"grad_norm": 1.0499067306518555,
"learning_rate": 8.104099424705135e-06,
"loss": 0.0324,
"step": 13390
},
{
"epoch": 51.53846153846154,
"grad_norm": 0.9457489252090454,
"learning_rate": 8.082083144136625e-06,
"loss": 0.0282,
"step": 13400
},
{
"epoch": 51.57692307692308,
"grad_norm": 1.1252720355987549,
"learning_rate": 8.060085776888634e-06,
"loss": 0.0321,
"step": 13410
},
{
"epoch": 51.61538461538461,
"grad_norm": 1.111789584159851,
"learning_rate": 8.038107383101126e-06,
"loss": 0.0339,
"step": 13420
},
{
"epoch": 51.65384615384615,
"grad_norm": 0.7417734861373901,
"learning_rate": 8.016148022862201e-06,
"loss": 0.0351,
"step": 13430
},
{
"epoch": 51.69230769230769,
"grad_norm": 1.0458786487579346,
"learning_rate": 7.994207756207903e-06,
"loss": 0.0363,
"step": 13440
},
{
"epoch": 51.73076923076923,
"grad_norm": 1.1386157274246216,
"learning_rate": 7.972286643122083e-06,
"loss": 0.0339,
"step": 13450
},
{
"epoch": 51.76923076923077,
"grad_norm": 1.0273871421813965,
"learning_rate": 7.950384743536225e-06,
"loss": 0.034,
"step": 13460
},
{
"epoch": 51.80769230769231,
"grad_norm": 1.0202971696853638,
"learning_rate": 7.928502117329285e-06,
"loss": 0.0341,
"step": 13470
},
{
"epoch": 51.84615384615385,
"grad_norm": 0.9581331014633179,
"learning_rate": 7.906638824327545e-06,
"loss": 0.0284,
"step": 13480
},
{
"epoch": 51.88461538461539,
"grad_norm": 1.155298113822937,
"learning_rate": 7.884794924304392e-06,
"loss": 0.0417,
"step": 13490
},
{
"epoch": 51.92307692307692,
"grad_norm": 0.8334839940071106,
"learning_rate": 7.862970476980218e-06,
"loss": 0.0356,
"step": 13500
},
{
"epoch": 51.96153846153846,
"grad_norm": 1.1004962921142578,
"learning_rate": 7.841165542022242e-06,
"loss": 0.0381,
"step": 13510
},
{
"epoch": 52.0,
"grad_norm": 1.0550732612609863,
"learning_rate": 7.819380179044307e-06,
"loss": 0.0277,
"step": 13520
},
{
"epoch": 52.03846153846154,
"grad_norm": 0.8858346939086914,
"learning_rate": 7.797614447606783e-06,
"loss": 0.0303,
"step": 13530
},
{
"epoch": 52.07692307692308,
"grad_norm": 1.1065115928649902,
"learning_rate": 7.775868407216326e-06,
"loss": 0.0379,
"step": 13540
},
{
"epoch": 52.11538461538461,
"grad_norm": 0.9579010009765625,
"learning_rate": 7.754142117325792e-06,
"loss": 0.0326,
"step": 13550
},
{
"epoch": 52.15384615384615,
"grad_norm": 1.0247814655303955,
"learning_rate": 7.73243563733403e-06,
"loss": 0.0356,
"step": 13560
},
{
"epoch": 52.19230769230769,
"grad_norm": 0.9102254509925842,
"learning_rate": 7.710749026585726e-06,
"loss": 0.0275,
"step": 13570
},
{
"epoch": 52.23076923076923,
"grad_norm": 0.701599657535553,
"learning_rate": 7.689082344371244e-06,
"loss": 0.0353,
"step": 13580
},
{
"epoch": 52.26923076923077,
"grad_norm": 0.9362711906433105,
"learning_rate": 7.667435649926459e-06,
"loss": 0.0314,
"step": 13590
},
{
"epoch": 52.30769230769231,
"grad_norm": 0.8040527701377869,
"learning_rate": 7.645809002432616e-06,
"loss": 0.0331,
"step": 13600
},
{
"epoch": 52.34615384615385,
"grad_norm": 0.7203378081321716,
"learning_rate": 7.6242024610161454e-06,
"loss": 0.0318,
"step": 13610
},
{
"epoch": 52.38461538461539,
"grad_norm": 0.9988213777542114,
"learning_rate": 7.602616084748501e-06,
"loss": 0.0323,
"step": 13620
},
{
"epoch": 52.42307692307692,
"grad_norm": 0.9332322478294373,
"learning_rate": 7.5810499326460145e-06,
"loss": 0.0356,
"step": 13630
},
{
"epoch": 52.46153846153846,
"grad_norm": 1.0301990509033203,
"learning_rate": 7.5595040636697145e-06,
"loss": 0.034,
"step": 13640
},
{
"epoch": 52.5,
"grad_norm": 0.9222707152366638,
"learning_rate": 7.537978536725189e-06,
"loss": 0.029,
"step": 13650
},
{
"epoch": 52.53846153846154,
"grad_norm": 1.0901727676391602,
"learning_rate": 7.5164734106624135e-06,
"loss": 0.0343,
"step": 13660
},
{
"epoch": 52.57692307692308,
"grad_norm": 0.9981234073638916,
"learning_rate": 7.494988744275575e-06,
"loss": 0.0283,
"step": 13670
},
{
"epoch": 52.61538461538461,
"grad_norm": 0.922626256942749,
"learning_rate": 7.473524596302931e-06,
"loss": 0.037,
"step": 13680
},
{
"epoch": 52.65384615384615,
"grad_norm": 0.9884187579154968,
"learning_rate": 7.452081025426639e-06,
"loss": 0.031,
"step": 13690
},
{
"epoch": 52.69230769230769,
"grad_norm": 1.013758659362793,
"learning_rate": 7.430658090272606e-06,
"loss": 0.0333,
"step": 13700
},
{
"epoch": 52.73076923076923,
"grad_norm": 1.1022049188613892,
"learning_rate": 7.40925584941033e-06,
"loss": 0.0358,
"step": 13710
},
{
"epoch": 52.76923076923077,
"grad_norm": 0.9264612197875977,
"learning_rate": 7.3878743613527e-06,
"loss": 0.0284,
"step": 13720
},
{
"epoch": 52.80769230769231,
"grad_norm": 1.111238718032837,
"learning_rate": 7.366513684555903e-06,
"loss": 0.0342,
"step": 13730
},
{
"epoch": 52.84615384615385,
"grad_norm": 1.2126115560531616,
"learning_rate": 7.345173877419204e-06,
"loss": 0.031,
"step": 13740
},
{
"epoch": 52.88461538461539,
"grad_norm": 1.1086037158966064,
"learning_rate": 7.323854998284823e-06,
"loss": 0.0348,
"step": 13750
},
{
"epoch": 52.92307692307692,
"grad_norm": 1.1465508937835693,
"learning_rate": 7.302557105437775e-06,
"loss": 0.0304,
"step": 13760
},
{
"epoch": 52.96153846153846,
"grad_norm": 0.9518792033195496,
"learning_rate": 7.281280257105666e-06,
"loss": 0.0306,
"step": 13770
},
{
"epoch": 53.0,
"grad_norm": 1.0208402872085571,
"learning_rate": 7.260024511458599e-06,
"loss": 0.0296,
"step": 13780
},
{
"epoch": 53.03846153846154,
"grad_norm": 0.9359050393104553,
"learning_rate": 7.238789926608963e-06,
"loss": 0.0331,
"step": 13790
},
{
"epoch": 53.07692307692308,
"grad_norm": 1.0515170097351074,
"learning_rate": 7.217576560611305e-06,
"loss": 0.0345,
"step": 13800
},
{
"epoch": 53.11538461538461,
"grad_norm": 0.9996568560600281,
"learning_rate": 7.196384471462172e-06,
"loss": 0.0364,
"step": 13810
},
{
"epoch": 53.15384615384615,
"grad_norm": 1.300593614578247,
"learning_rate": 7.175213717099903e-06,
"loss": 0.0321,
"step": 13820
},
{
"epoch": 53.19230769230769,
"grad_norm": 1.1695557832717896,
"learning_rate": 7.154064355404547e-06,
"loss": 0.0327,
"step": 13830
},
{
"epoch": 53.23076923076923,
"grad_norm": 0.9456730484962463,
"learning_rate": 7.132936444197641e-06,
"loss": 0.0319,
"step": 13840
},
{
"epoch": 53.26923076923077,
"grad_norm": 0.7543469071388245,
"learning_rate": 7.111830041242101e-06,
"loss": 0.029,
"step": 13850
},
{
"epoch": 53.30769230769231,
"grad_norm": 0.9508689045906067,
"learning_rate": 7.090745204242018e-06,
"loss": 0.0353,
"step": 13860
},
{
"epoch": 53.34615384615385,
"grad_norm": 0.8398585319519043,
"learning_rate": 7.069681990842527e-06,
"loss": 0.0276,
"step": 13870
},
{
"epoch": 53.38461538461539,
"grad_norm": 0.7608618140220642,
"learning_rate": 7.04864045862966e-06,
"loss": 0.0313,
"step": 13880
},
{
"epoch": 53.42307692307692,
"grad_norm": 1.0014206171035767,
"learning_rate": 7.027620665130152e-06,
"loss": 0.0331,
"step": 13890
},
{
"epoch": 53.46153846153846,
"grad_norm": 0.7588220834732056,
"learning_rate": 7.006622667811326e-06,
"loss": 0.0351,
"step": 13900
},
{
"epoch": 53.5,
"grad_norm": 0.9668156504631042,
"learning_rate": 6.9856465240809e-06,
"loss": 0.0292,
"step": 13910
},
{
"epoch": 53.53846153846154,
"grad_norm": 1.0986829996109009,
"learning_rate": 6.964692291286844e-06,
"loss": 0.0347,
"step": 13920
},
{
"epoch": 53.57692307692308,
"grad_norm": 0.9580195546150208,
"learning_rate": 6.943760026717243e-06,
"loss": 0.0311,
"step": 13930
},
{
"epoch": 53.61538461538461,
"grad_norm": 1.2510932683944702,
"learning_rate": 6.922849787600097e-06,
"loss": 0.0319,
"step": 13940
},
{
"epoch": 53.65384615384615,
"grad_norm": 1.139769434928894,
"learning_rate": 6.90196163110321e-06,
"loss": 0.0357,
"step": 13950
},
{
"epoch": 53.69230769230769,
"grad_norm": 0.9529709815979004,
"learning_rate": 6.881095614334002e-06,
"loss": 0.0312,
"step": 13960
},
{
"epoch": 53.73076923076923,
"grad_norm": 0.8552069067955017,
"learning_rate": 6.860251794339359e-06,
"loss": 0.0295,
"step": 13970
},
{
"epoch": 53.76923076923077,
"grad_norm": 0.783135175704956,
"learning_rate": 6.839430228105501e-06,
"loss": 0.0294,
"step": 13980
},
{
"epoch": 53.80769230769231,
"grad_norm": 0.7776913046836853,
"learning_rate": 6.818630972557788e-06,
"loss": 0.0309,
"step": 13990
},
{
"epoch": 53.84615384615385,
"grad_norm": 0.7431850433349609,
"learning_rate": 6.797854084560585e-06,
"loss": 0.0286,
"step": 14000
},
{
"epoch": 53.88461538461539,
"grad_norm": 0.7169124484062195,
"learning_rate": 6.777099620917124e-06,
"loss": 0.0315,
"step": 14010
},
{
"epoch": 53.92307692307692,
"grad_norm": 1.0331408977508545,
"learning_rate": 6.756367638369301e-06,
"loss": 0.0337,
"step": 14020
},
{
"epoch": 53.96153846153846,
"grad_norm": 0.8683583736419678,
"learning_rate": 6.735658193597579e-06,
"loss": 0.0308,
"step": 14030
},
{
"epoch": 54.0,
"grad_norm": 0.9476208090782166,
"learning_rate": 6.7149713432207825e-06,
"loss": 0.0294,
"step": 14040
},
{
"epoch": 54.03846153846154,
"grad_norm": 0.8476566076278687,
"learning_rate": 6.694307143795966e-06,
"loss": 0.0268,
"step": 14050
},
{
"epoch": 54.07692307692308,
"grad_norm": 0.9866935014724731,
"learning_rate": 6.6736656518182704e-06,
"loss": 0.0315,
"step": 14060
},
{
"epoch": 54.11538461538461,
"grad_norm": 0.9184465408325195,
"learning_rate": 6.6530469237207375e-06,
"loss": 0.0328,
"step": 14070
},
{
"epoch": 54.15384615384615,
"grad_norm": 0.872621476650238,
"learning_rate": 6.632451015874193e-06,
"loss": 0.0286,
"step": 14080
},
{
"epoch": 54.19230769230769,
"grad_norm": 0.8452703356742859,
"learning_rate": 6.611877984587058e-06,
"loss": 0.029,
"step": 14090
},
{
"epoch": 54.23076923076923,
"grad_norm": 0.9648960828781128,
"learning_rate": 6.591327886105207e-06,
"loss": 0.0314,
"step": 14100
},
{
"epoch": 54.26923076923077,
"grad_norm": 1.0145609378814697,
"learning_rate": 6.570800776611836e-06,
"loss": 0.0264,
"step": 14110
},
{
"epoch": 54.30769230769231,
"grad_norm": 0.7880923748016357,
"learning_rate": 6.55029671222727e-06,
"loss": 0.0291,
"step": 14120
},
{
"epoch": 54.34615384615385,
"grad_norm": 0.9834394454956055,
"learning_rate": 6.529815749008846e-06,
"loss": 0.0321,
"step": 14130
},
{
"epoch": 54.38461538461539,
"grad_norm": 0.9993003606796265,
"learning_rate": 6.50935794295073e-06,
"loss": 0.0341,
"step": 14140
},
{
"epoch": 54.42307692307692,
"grad_norm": 0.8572825789451599,
"learning_rate": 6.488923349983779e-06,
"loss": 0.0292,
"step": 14150
},
{
"epoch": 54.46153846153846,
"grad_norm": 0.9555164575576782,
"learning_rate": 6.468512025975401e-06,
"loss": 0.0298,
"step": 14160
},
{
"epoch": 54.5,
"grad_norm": 0.8961513638496399,
"learning_rate": 6.448124026729363e-06,
"loss": 0.032,
"step": 14170
},
{
"epoch": 54.53846153846154,
"grad_norm": 0.9303173422813416,
"learning_rate": 6.427759407985691e-06,
"loss": 0.0308,
"step": 14180
},
{
"epoch": 54.57692307692308,
"grad_norm": 0.9030703902244568,
"learning_rate": 6.407418225420465e-06,
"loss": 0.0306,
"step": 14190
},
{
"epoch": 54.61538461538461,
"grad_norm": 1.0203205347061157,
"learning_rate": 6.387100534645698e-06,
"loss": 0.0312,
"step": 14200
},
{
"epoch": 54.65384615384615,
"grad_norm": 0.9905704259872437,
"learning_rate": 6.366806391209194e-06,
"loss": 0.0317,
"step": 14210
},
{
"epoch": 54.69230769230769,
"grad_norm": 0.8993654251098633,
"learning_rate": 6.346535850594352e-06,
"loss": 0.033,
"step": 14220
},
{
"epoch": 54.73076923076923,
"grad_norm": 0.888511061668396,
"learning_rate": 6.326288968220069e-06,
"loss": 0.0337,
"step": 14230
},
{
"epoch": 54.76923076923077,
"grad_norm": 0.8489452600479126,
"learning_rate": 6.306065799440542e-06,
"loss": 0.0283,
"step": 14240
},
{
"epoch": 54.80769230769231,
"grad_norm": 1.16481351852417,
"learning_rate": 6.285866399545137e-06,
"loss": 0.0328,
"step": 14250
},
{
"epoch": 54.84615384615385,
"grad_norm": 0.9366856813430786,
"learning_rate": 6.2656908237582515e-06,
"loss": 0.0341,
"step": 14260
},
{
"epoch": 54.88461538461539,
"grad_norm": 0.7751666307449341,
"learning_rate": 6.245539127239135e-06,
"loss": 0.0284,
"step": 14270
},
{
"epoch": 54.92307692307692,
"grad_norm": 1.0874589681625366,
"learning_rate": 6.225411365081752e-06,
"loss": 0.0303,
"step": 14280
},
{
"epoch": 54.96153846153846,
"grad_norm": 0.7190085053443909,
"learning_rate": 6.205307592314645e-06,
"loss": 0.0254,
"step": 14290
},
{
"epoch": 55.0,
"grad_norm": 0.9163426756858826,
"learning_rate": 6.185227863900751e-06,
"loss": 0.0299,
"step": 14300
},
{
"epoch": 55.03846153846154,
"grad_norm": 0.8473014831542969,
"learning_rate": 6.165172234737291e-06,
"loss": 0.0277,
"step": 14310
},
{
"epoch": 55.07692307692308,
"grad_norm": 0.6970618367195129,
"learning_rate": 6.145140759655586e-06,
"loss": 0.0369,
"step": 14320
},
{
"epoch": 55.11538461538461,
"grad_norm": 0.8919505476951599,
"learning_rate": 6.125133493420914e-06,
"loss": 0.0342,
"step": 14330
},
{
"epoch": 55.15384615384615,
"grad_norm": 0.9447165727615356,
"learning_rate": 6.1051504907323915e-06,
"loss": 0.0335,
"step": 14340
},
{
"epoch": 55.19230769230769,
"grad_norm": 0.7969279289245605,
"learning_rate": 6.085191806222774e-06,
"loss": 0.0297,
"step": 14350
},
{
"epoch": 55.23076923076923,
"grad_norm": 0.8206071853637695,
"learning_rate": 6.065257494458352e-06,
"loss": 0.0347,
"step": 14360
},
{
"epoch": 55.26923076923077,
"grad_norm": 0.7703580856323242,
"learning_rate": 6.045347609938767e-06,
"loss": 0.0317,
"step": 14370
},
{
"epoch": 55.30769230769231,
"grad_norm": 0.6906372904777527,
"learning_rate": 6.025462207096879e-06,
"loss": 0.0326,
"step": 14380
},
{
"epoch": 55.34615384615385,
"grad_norm": 0.7366732954978943,
"learning_rate": 6.005601340298631e-06,
"loss": 0.0297,
"step": 14390
},
{
"epoch": 55.38461538461539,
"grad_norm": 0.9685719609260559,
"learning_rate": 5.985765063842862e-06,
"loss": 0.031,
"step": 14400
},
{
"epoch": 55.42307692307692,
"grad_norm": 0.7533148527145386,
"learning_rate": 5.965953431961206e-06,
"loss": 0.0322,
"step": 14410
},
{
"epoch": 55.46153846153846,
"grad_norm": 0.7429641485214233,
"learning_rate": 5.946166498817903e-06,
"loss": 0.0303,
"step": 14420
},
{
"epoch": 55.5,
"grad_norm": 0.6613432765007019,
"learning_rate": 5.926404318509668e-06,
"loss": 0.0263,
"step": 14430
},
{
"epoch": 55.53846153846154,
"grad_norm": 0.8684971332550049,
"learning_rate": 5.906666945065556e-06,
"loss": 0.0278,
"step": 14440
},
{
"epoch": 55.57692307692308,
"grad_norm": 0.8969733119010925,
"learning_rate": 5.886954432446784e-06,
"loss": 0.0381,
"step": 14450
},
{
"epoch": 55.61538461538461,
"grad_norm": 0.8112402558326721,
"learning_rate": 5.867266834546617e-06,
"loss": 0.0286,
"step": 14460
},
{
"epoch": 55.65384615384615,
"grad_norm": 1.0435789823532104,
"learning_rate": 5.847604205190192e-06,
"loss": 0.0328,
"step": 14470
},
{
"epoch": 55.69230769230769,
"grad_norm": 0.8987309336662292,
"learning_rate": 5.827966598134383e-06,
"loss": 0.0339,
"step": 14480
},
{
"epoch": 55.73076923076923,
"grad_norm": 0.8202237486839294,
"learning_rate": 5.808354067067665e-06,
"loss": 0.0319,
"step": 14490
},
{
"epoch": 55.76923076923077,
"grad_norm": 0.8230668306350708,
"learning_rate": 5.788766665609941e-06,
"loss": 0.0255,
"step": 14500
},
{
"epoch": 55.80769230769231,
"grad_norm": 0.9060791730880737,
"learning_rate": 5.7692044473124276e-06,
"loss": 0.0308,
"step": 14510
},
{
"epoch": 55.84615384615385,
"grad_norm": 1.0444339513778687,
"learning_rate": 5.749667465657479e-06,
"loss": 0.027,
"step": 14520
},
{
"epoch": 55.88461538461539,
"grad_norm": 0.7329652309417725,
"learning_rate": 5.730155774058451e-06,
"loss": 0.0313,
"step": 14530
},
{
"epoch": 55.92307692307692,
"grad_norm": 1.0695486068725586,
"learning_rate": 5.710669425859575e-06,
"loss": 0.0357,
"step": 14540
},
{
"epoch": 55.96153846153846,
"grad_norm": 0.780979335308075,
"learning_rate": 5.691208474335774e-06,
"loss": 0.0326,
"step": 14550
},
{
"epoch": 56.0,
"grad_norm": 0.8867959380149841,
"learning_rate": 5.6717729726925446e-06,
"loss": 0.0308,
"step": 14560
},
{
"epoch": 56.03846153846154,
"grad_norm": 0.8821061849594116,
"learning_rate": 5.652362974065816e-06,
"loss": 0.0321,
"step": 14570
},
{
"epoch": 56.07692307692308,
"grad_norm": 0.7196897268295288,
"learning_rate": 5.6329785315217726e-06,
"loss": 0.0316,
"step": 14580
},
{
"epoch": 56.11538461538461,
"grad_norm": 0.7857591509819031,
"learning_rate": 5.6136196980567495e-06,
"loss": 0.0286,
"step": 14590
},
{
"epoch": 56.15384615384615,
"grad_norm": 0.7501246929168701,
"learning_rate": 5.594286526597054e-06,
"loss": 0.0301,
"step": 14600
},
{
"epoch": 56.19230769230769,
"grad_norm": 0.9493558406829834,
"learning_rate": 5.574979069998833e-06,
"loss": 0.0313,
"step": 14610
},
{
"epoch": 56.23076923076923,
"grad_norm": 0.8972458839416504,
"learning_rate": 5.5556973810479486e-06,
"loss": 0.0293,
"step": 14620
},
{
"epoch": 56.26923076923077,
"grad_norm": 0.7602220773696899,
"learning_rate": 5.536441512459787e-06,
"loss": 0.036,
"step": 14630
},
{
"epoch": 56.30769230769231,
"grad_norm": 1.1218823194503784,
"learning_rate": 5.517211516879172e-06,
"loss": 0.0319,
"step": 14640
},
{
"epoch": 56.34615384615385,
"grad_norm": 0.908366858959198,
"learning_rate": 5.49800744688017e-06,
"loss": 0.0302,
"step": 14650
},
{
"epoch": 56.38461538461539,
"grad_norm": 0.8256409764289856,
"learning_rate": 5.4788293549659694e-06,
"loss": 0.0289,
"step": 14660
},
{
"epoch": 56.42307692307692,
"grad_norm": 0.7066758275032043,
"learning_rate": 5.459677293568753e-06,
"loss": 0.0279,
"step": 14670
},
{
"epoch": 56.46153846153846,
"grad_norm": 0.6973010301589966,
"learning_rate": 5.440551315049515e-06,
"loss": 0.0307,
"step": 14680
},
{
"epoch": 56.5,
"grad_norm": 0.7377819418907166,
"learning_rate": 5.421451471697966e-06,
"loss": 0.0281,
"step": 14690
},
{
"epoch": 56.53846153846154,
"grad_norm": 0.8551780581474304,
"learning_rate": 5.402377815732326e-06,
"loss": 0.0308,
"step": 14700
},
{
"epoch": 56.57692307692308,
"grad_norm": 0.8823787569999695,
"learning_rate": 5.383330399299253e-06,
"loss": 0.0412,
"step": 14710
},
{
"epoch": 56.61538461538461,
"grad_norm": 0.9362679719924927,
"learning_rate": 5.364309274473663e-06,
"loss": 0.0339,
"step": 14720
},
{
"epoch": 56.65384615384615,
"grad_norm": 1.0090643167495728,
"learning_rate": 5.345314493258573e-06,
"loss": 0.031,
"step": 14730
},
{
"epoch": 56.69230769230769,
"grad_norm": 1.0990750789642334,
"learning_rate": 5.326346107585e-06,
"loss": 0.0381,
"step": 14740
},
{
"epoch": 56.73076923076923,
"grad_norm": 1.0385936498641968,
"learning_rate": 5.307404169311782e-06,
"loss": 0.0328,
"step": 14750
},
{
"epoch": 56.76923076923077,
"grad_norm": 1.0119373798370361,
"learning_rate": 5.288488730225449e-06,
"loss": 0.03,
"step": 14760
},
{
"epoch": 56.80769230769231,
"grad_norm": 0.8520657420158386,
"learning_rate": 5.2695998420401e-06,
"loss": 0.0367,
"step": 14770
},
{
"epoch": 56.84615384615385,
"grad_norm": 0.8646618127822876,
"learning_rate": 5.2507375563972236e-06,
"loss": 0.0306,
"step": 14780
},
{
"epoch": 56.88461538461539,
"grad_norm": 0.890970766544342,
"learning_rate": 5.231901924865596e-06,
"loss": 0.0324,
"step": 14790
},
{
"epoch": 56.92307692307692,
"grad_norm": 0.7873624563217163,
"learning_rate": 5.213092998941113e-06,
"loss": 0.0246,
"step": 14800
},
{
"epoch": 56.96153846153846,
"grad_norm": 0.8681284189224243,
"learning_rate": 5.1943108300466555e-06,
"loss": 0.0284,
"step": 14810
},
{
"epoch": 57.0,
"grad_norm": 0.9095093607902527,
"learning_rate": 5.175555469531964e-06,
"loss": 0.0304,
"step": 14820
},
{
"epoch": 57.03846153846154,
"grad_norm": 0.9900658130645752,
"learning_rate": 5.1568269686734716e-06,
"loss": 0.0312,
"step": 14830
},
{
"epoch": 57.07692307692308,
"grad_norm": 1.1000691652297974,
"learning_rate": 5.138125378674182e-06,
"loss": 0.0297,
"step": 14840
},
{
"epoch": 57.11538461538461,
"grad_norm": 0.9001245498657227,
"learning_rate": 5.119450750663539e-06,
"loss": 0.0286,
"step": 14850
},
{
"epoch": 57.15384615384615,
"grad_norm": 0.6249764561653137,
"learning_rate": 5.100803135697248e-06,
"loss": 0.0303,
"step": 14860
},
{
"epoch": 57.19230769230769,
"grad_norm": 0.7594456076622009,
"learning_rate": 5.0821825847571904e-06,
"loss": 0.0309,
"step": 14870
},
{
"epoch": 57.23076923076923,
"grad_norm": 0.7925312519073486,
"learning_rate": 5.063589148751236e-06,
"loss": 0.0309,
"step": 14880
},
{
"epoch": 57.26923076923077,
"grad_norm": 0.7072622776031494,
"learning_rate": 5.045022878513122e-06,
"loss": 0.0283,
"step": 14890
},
{
"epoch": 57.30769230769231,
"grad_norm": 0.8310312628746033,
"learning_rate": 5.026483824802333e-06,
"loss": 0.0277,
"step": 14900
},
{
"epoch": 57.34615384615385,
"grad_norm": 0.8580349683761597,
"learning_rate": 5.0079720383039245e-06,
"loss": 0.028,
"step": 14910
},
{
"epoch": 57.38461538461539,
"grad_norm": 0.9539895057678223,
"learning_rate": 4.989487569628425e-06,
"loss": 0.0378,
"step": 14920
},
{
"epoch": 57.42307692307692,
"grad_norm": 0.6495718359947205,
"learning_rate": 4.971030469311658e-06,
"loss": 0.0283,
"step": 14930
},
{
"epoch": 57.46153846153846,
"grad_norm": 0.9080126285552979,
"learning_rate": 4.952600787814628e-06,
"loss": 0.0321,
"step": 14940
},
{
"epoch": 57.5,
"grad_norm": 1.1204878091812134,
"learning_rate": 4.934198575523391e-06,
"loss": 0.0302,
"step": 14950
},
{
"epoch": 57.53846153846154,
"grad_norm": 0.8599961996078491,
"learning_rate": 4.915823882748882e-06,
"loss": 0.0263,
"step": 14960
},
{
"epoch": 57.57692307692308,
"grad_norm": 0.7675483226776123,
"learning_rate": 4.897476759726823e-06,
"loss": 0.0269,
"step": 14970
},
{
"epoch": 57.61538461538461,
"grad_norm": 0.7085859179496765,
"learning_rate": 4.87915725661753e-06,
"loss": 0.028,
"step": 14980
},
{
"epoch": 57.65384615384615,
"grad_norm": 0.9856685400009155,
"learning_rate": 4.860865423505833e-06,
"loss": 0.0348,
"step": 14990
},
{
"epoch": 57.69230769230769,
"grad_norm": 0.9436843991279602,
"learning_rate": 4.842601310400912e-06,
"loss": 0.0287,
"step": 15000
},
{
"epoch": 57.73076923076923,
"grad_norm": 0.8575741648674011,
"learning_rate": 4.824364967236145e-06,
"loss": 0.0299,
"step": 15010
},
{
"epoch": 57.76923076923077,
"grad_norm": 0.7252999544143677,
"learning_rate": 4.8061564438690095e-06,
"loss": 0.0268,
"step": 15020
},
{
"epoch": 57.80769230769231,
"grad_norm": 0.9593229293823242,
"learning_rate": 4.787975790080896e-06,
"loss": 0.033,
"step": 15030
},
{
"epoch": 57.84615384615385,
"grad_norm": 0.7520903944969177,
"learning_rate": 4.769823055577029e-06,
"loss": 0.0319,
"step": 15040
},
{
"epoch": 57.88461538461539,
"grad_norm": 0.6410161256790161,
"learning_rate": 4.7516982899862934e-06,
"loss": 0.0307,
"step": 15050
},
{
"epoch": 57.92307692307692,
"grad_norm": 0.7836447358131409,
"learning_rate": 4.733601542861098e-06,
"loss": 0.0288,
"step": 15060
},
{
"epoch": 57.96153846153846,
"grad_norm": 0.7842332124710083,
"learning_rate": 4.7155328636772735e-06,
"loss": 0.0311,
"step": 15070
},
{
"epoch": 58.0,
"grad_norm": 0.7095579504966736,
"learning_rate": 4.697492301833878e-06,
"loss": 0.027,
"step": 15080
},
{
"epoch": 58.03846153846154,
"grad_norm": 0.6566604375839233,
"learning_rate": 4.679479906653128e-06,
"loss": 0.0308,
"step": 15090
},
{
"epoch": 58.07692307692308,
"grad_norm": 0.7291385531425476,
"learning_rate": 4.661495727380232e-06,
"loss": 0.0282,
"step": 15100
},
{
"epoch": 58.11538461538461,
"grad_norm": 0.7904289960861206,
"learning_rate": 4.64353981318324e-06,
"loss": 0.0363,
"step": 15110
},
{
"epoch": 58.15384615384615,
"grad_norm": 0.7773877382278442,
"learning_rate": 4.62561221315294e-06,
"loss": 0.0357,
"step": 15120
},
{
"epoch": 58.19230769230769,
"grad_norm": 0.7698177695274353,
"learning_rate": 4.6077129763026995e-06,
"loss": 0.0314,
"step": 15130
},
{
"epoch": 58.23076923076923,
"grad_norm": 0.8109748959541321,
"learning_rate": 4.589842151568354e-06,
"loss": 0.0267,
"step": 15140
},
{
"epoch": 58.26923076923077,
"grad_norm": 0.606159508228302,
"learning_rate": 4.571999787808057e-06,
"loss": 0.0282,
"step": 15150
},
{
"epoch": 58.30769230769231,
"grad_norm": 0.7719907760620117,
"learning_rate": 4.554185933802151e-06,
"loss": 0.028,
"step": 15160
},
{
"epoch": 58.34615384615385,
"grad_norm": 0.8883272409439087,
"learning_rate": 4.5364006382530285e-06,
"loss": 0.0287,
"step": 15170
},
{
"epoch": 58.38461538461539,
"grad_norm": 0.782050609588623,
"learning_rate": 4.518643949785004e-06,
"loss": 0.0277,
"step": 15180
},
{
"epoch": 58.42307692307692,
"grad_norm": 0.7449761033058167,
"learning_rate": 4.500915916944193e-06,
"loss": 0.03,
"step": 15190
},
{
"epoch": 58.46153846153846,
"grad_norm": 0.7998998761177063,
"learning_rate": 4.483216588198366e-06,
"loss": 0.0325,
"step": 15200
},
{
"epoch": 58.5,
"grad_norm": 0.6946242451667786,
"learning_rate": 4.465546011936797e-06,
"loss": 0.0274,
"step": 15210
},
{
"epoch": 58.53846153846154,
"grad_norm": 0.6466119885444641,
"learning_rate": 4.447904236470177e-06,
"loss": 0.0284,
"step": 15220
},
{
"epoch": 58.57692307692308,
"grad_norm": 0.7280353903770447,
"learning_rate": 4.43029131003044e-06,
"loss": 0.0293,
"step": 15230
},
{
"epoch": 58.61538461538461,
"grad_norm": 0.8519057631492615,
"learning_rate": 4.412707280770658e-06,
"loss": 0.0313,
"step": 15240
},
{
"epoch": 58.65384615384615,
"grad_norm": 0.9241837859153748,
"learning_rate": 4.395152196764905e-06,
"loss": 0.0338,
"step": 15250
},
{
"epoch": 58.69230769230769,
"grad_norm": 0.8291633129119873,
"learning_rate": 4.3776261060080916e-06,
"loss": 0.0295,
"step": 15260
},
{
"epoch": 58.73076923076923,
"grad_norm": 0.8491876721382141,
"learning_rate": 4.360129056415895e-06,
"loss": 0.0324,
"step": 15270
},
{
"epoch": 58.76923076923077,
"grad_norm": 0.8303586840629578,
"learning_rate": 4.34266109582457e-06,
"loss": 0.0285,
"step": 15280
},
{
"epoch": 58.80769230769231,
"grad_norm": 1.0164424180984497,
"learning_rate": 4.325222271990861e-06,
"loss": 0.0326,
"step": 15290
},
{
"epoch": 58.84615384615385,
"grad_norm": 0.967418372631073,
"learning_rate": 4.307812632591853e-06,
"loss": 0.0295,
"step": 15300
},
{
"epoch": 58.88461538461539,
"grad_norm": 0.6895191073417664,
"learning_rate": 4.2904322252248186e-06,
"loss": 0.0329,
"step": 15310
},
{
"epoch": 58.92307692307692,
"grad_norm": 0.9052116870880127,
"learning_rate": 4.273081097407142e-06,
"loss": 0.037,
"step": 15320
},
{
"epoch": 58.96153846153846,
"grad_norm": 1.0200068950653076,
"learning_rate": 4.255759296576133e-06,
"loss": 0.0335,
"step": 15330
},
{
"epoch": 59.0,
"grad_norm": 0.899724006652832,
"learning_rate": 4.238466870088945e-06,
"loss": 0.0305,
"step": 15340
},
{
"epoch": 59.03846153846154,
"grad_norm": 0.9742430448532104,
"learning_rate": 4.221203865222405e-06,
"loss": 0.0289,
"step": 15350
},
{
"epoch": 59.07692307692308,
"grad_norm": 0.8280800580978394,
"learning_rate": 4.203970329172907e-06,
"loss": 0.0286,
"step": 15360
},
{
"epoch": 59.11538461538461,
"grad_norm": 0.7113635540008545,
"learning_rate": 4.186766309056286e-06,
"loss": 0.0274,
"step": 15370
},
{
"epoch": 59.15384615384615,
"grad_norm": 0.7944599390029907,
"learning_rate": 4.16959185190767e-06,
"loss": 0.0285,
"step": 15380
},
{
"epoch": 59.19230769230769,
"grad_norm": 0.8309881091117859,
"learning_rate": 4.152447004681379e-06,
"loss": 0.0363,
"step": 15390
},
{
"epoch": 59.23076923076923,
"grad_norm": 0.8225540518760681,
"learning_rate": 4.135331814250764e-06,
"loss": 0.0345,
"step": 15400
},
{
"epoch": 59.26923076923077,
"grad_norm": 0.603495180606842,
"learning_rate": 4.118246327408095e-06,
"loss": 0.0277,
"step": 15410
},
{
"epoch": 59.30769230769231,
"grad_norm": 0.8480184078216553,
"learning_rate": 4.101190590864457e-06,
"loss": 0.0369,
"step": 15420
},
{
"epoch": 59.34615384615385,
"grad_norm": 0.7788599729537964,
"learning_rate": 4.084164651249566e-06,
"loss": 0.0275,
"step": 15430
},
{
"epoch": 59.38461538461539,
"grad_norm": 0.7458527088165283,
"learning_rate": 4.0671685551117035e-06,
"loss": 0.0274,
"step": 15440
},
{
"epoch": 59.42307692307692,
"grad_norm": 0.7309711575508118,
"learning_rate": 4.050202348917544e-06,
"loss": 0.0303,
"step": 15450
},
{
"epoch": 59.46153846153846,
"grad_norm": 0.7994899749755859,
"learning_rate": 4.033266079052039e-06,
"loss": 0.0291,
"step": 15460
},
{
"epoch": 59.5,
"grad_norm": 0.891418993473053,
"learning_rate": 4.016359791818314e-06,
"loss": 0.0302,
"step": 15470
},
{
"epoch": 59.53846153846154,
"grad_norm": 1.124557375907898,
"learning_rate": 3.999483533437511e-06,
"loss": 0.0272,
"step": 15480
},
{
"epoch": 59.57692307692308,
"grad_norm": 0.9168891310691833,
"learning_rate": 3.982637350048669e-06,
"loss": 0.0337,
"step": 15490
},
{
"epoch": 59.61538461538461,
"grad_norm": 0.8245673775672913,
"learning_rate": 3.965821287708619e-06,
"loss": 0.0289,
"step": 15500
},
{
"epoch": 59.65384615384615,
"grad_norm": 0.7510520815849304,
"learning_rate": 3.949035392391825e-06,
"loss": 0.0278,
"step": 15510
},
{
"epoch": 59.69230769230769,
"grad_norm": 0.7584001421928406,
"learning_rate": 3.932279709990293e-06,
"loss": 0.0306,
"step": 15520
},
{
"epoch": 59.73076923076923,
"grad_norm": 0.7586838006973267,
"learning_rate": 3.915554286313413e-06,
"loss": 0.0369,
"step": 15530
},
{
"epoch": 59.76923076923077,
"grad_norm": 0.9130982160568237,
"learning_rate": 3.898859167087853e-06,
"loss": 0.0343,
"step": 15540
},
{
"epoch": 59.80769230769231,
"grad_norm": 0.8298722505569458,
"learning_rate": 3.882194397957437e-06,
"loss": 0.0307,
"step": 15550
},
{
"epoch": 59.84615384615385,
"grad_norm": 0.7778029441833496,
"learning_rate": 3.865560024483002e-06,
"loss": 0.0285,
"step": 15560
},
{
"epoch": 59.88461538461539,
"grad_norm": 0.6167328953742981,
"learning_rate": 3.848956092142294e-06,
"loss": 0.0271,
"step": 15570
},
{
"epoch": 59.92307692307692,
"grad_norm": 0.7541554570198059,
"learning_rate": 3.832382646329831e-06,
"loss": 0.0297,
"step": 15580
},
{
"epoch": 59.96153846153846,
"grad_norm": 0.7008756399154663,
"learning_rate": 3.8158397323567725e-06,
"loss": 0.0311,
"step": 15590
},
{
"epoch": 60.0,
"grad_norm": 0.6774673461914062,
"learning_rate": 3.7993273954508262e-06,
"loss": 0.0281,
"step": 15600
},
{
"epoch": 60.03846153846154,
"grad_norm": 0.7706513404846191,
"learning_rate": 3.782845680756078e-06,
"loss": 0.03,
"step": 15610
},
{
"epoch": 60.07692307692308,
"grad_norm": 0.6021534204483032,
"learning_rate": 3.7663946333329186e-06,
"loss": 0.0287,
"step": 15620
},
{
"epoch": 60.11538461538461,
"grad_norm": 0.8721636533737183,
"learning_rate": 3.7499742981578753e-06,
"loss": 0.0247,
"step": 15630
},
{
"epoch": 60.15384615384615,
"grad_norm": 0.6210073828697205,
"learning_rate": 3.7335847201235166e-06,
"loss": 0.0288,
"step": 15640
},
{
"epoch": 60.19230769230769,
"grad_norm": 0.8508103489875793,
"learning_rate": 3.717225944038331e-06,
"loss": 0.0323,
"step": 15650
},
{
"epoch": 60.23076923076923,
"grad_norm": 0.6650373339653015,
"learning_rate": 3.7008980146265776e-06,
"loss": 0.0291,
"step": 15660
},
{
"epoch": 60.26923076923077,
"grad_norm": 0.5672306418418884,
"learning_rate": 3.6846009765282013e-06,
"loss": 0.0293,
"step": 15670
},
{
"epoch": 60.30769230769231,
"grad_norm": 0.6800737977027893,
"learning_rate": 3.6683348742986784e-06,
"loss": 0.0285,
"step": 15680
},
{
"epoch": 60.34615384615385,
"grad_norm": 0.8498525619506836,
"learning_rate": 3.6520997524089057e-06,
"loss": 0.0303,
"step": 15690
},
{
"epoch": 60.38461538461539,
"grad_norm": 0.6448276042938232,
"learning_rate": 3.635895655245096e-06,
"loss": 0.0276,
"step": 15700
},
{
"epoch": 60.42307692307692,
"grad_norm": 0.7374983429908752,
"learning_rate": 3.619722627108624e-06,
"loss": 0.0305,
"step": 15710
},
{
"epoch": 60.46153846153846,
"grad_norm": 0.9034025073051453,
"learning_rate": 3.603580712215937e-06,
"loss": 0.0301,
"step": 15720
},
{
"epoch": 60.5,
"grad_norm": 0.6876755356788635,
"learning_rate": 3.587469954698413e-06,
"loss": 0.0272,
"step": 15730
},
{
"epoch": 60.53846153846154,
"grad_norm": 0.628259539604187,
"learning_rate": 3.5713903986022425e-06,
"loss": 0.0309,
"step": 15740
},
{
"epoch": 60.57692307692308,
"grad_norm": 0.8934166431427002,
"learning_rate": 3.555342087888326e-06,
"loss": 0.0286,
"step": 15750
},
{
"epoch": 60.61538461538461,
"grad_norm": 0.7909426689147949,
"learning_rate": 3.539325066432127e-06,
"loss": 0.0308,
"step": 15760
},
{
"epoch": 60.65384615384615,
"grad_norm": 0.8825272917747498,
"learning_rate": 3.523339378023569e-06,
"loss": 0.0322,
"step": 15770
},
{
"epoch": 60.69230769230769,
"grad_norm": 0.7353061437606812,
"learning_rate": 3.5073850663669193e-06,
"loss": 0.0325,
"step": 15780
},
{
"epoch": 60.73076923076923,
"grad_norm": 0.5934805274009705,
"learning_rate": 3.4914621750806503e-06,
"loss": 0.0306,
"step": 15790
},
{
"epoch": 60.76923076923077,
"grad_norm": 0.7742717266082764,
"learning_rate": 3.475570747697346e-06,
"loss": 0.0305,
"step": 15800
},
{
"epoch": 60.80769230769231,
"grad_norm": 0.8100197911262512,
"learning_rate": 3.4597108276635577e-06,
"loss": 0.0281,
"step": 15810
},
{
"epoch": 60.84615384615385,
"grad_norm": 0.5236333012580872,
"learning_rate": 3.443882458339699e-06,
"loss": 0.0336,
"step": 15820
},
{
"epoch": 60.88461538461539,
"grad_norm": 0.8024904131889343,
"learning_rate": 3.4280856829999323e-06,
"loss": 0.0299,
"step": 15830
},
{
"epoch": 60.92307692307692,
"grad_norm": 0.654270350933075,
"learning_rate": 3.412320544832033e-06,
"loss": 0.027,
"step": 15840
},
{
"epoch": 60.96153846153846,
"grad_norm": 0.6151342988014221,
"learning_rate": 3.396587086937294e-06,
"loss": 0.0311,
"step": 15850
},
{
"epoch": 61.0,
"grad_norm": 0.7870935201644897,
"learning_rate": 3.380885352330383e-06,
"loss": 0.0271,
"step": 15860
},
{
"epoch": 61.03846153846154,
"grad_norm": 0.5324817299842834,
"learning_rate": 3.3652153839392414e-06,
"loss": 0.0276,
"step": 15870
},
{
"epoch": 61.07692307692308,
"grad_norm": 0.6367908716201782,
"learning_rate": 3.34957722460497e-06,
"loss": 0.0352,
"step": 15880
},
{
"epoch": 61.11538461538461,
"grad_norm": 0.8792974948883057,
"learning_rate": 3.333970917081691e-06,
"loss": 0.0275,
"step": 15890
},
{
"epoch": 61.15384615384615,
"grad_norm": 0.7430537939071655,
"learning_rate": 3.318396504036465e-06,
"loss": 0.0325,
"step": 15900
},
{
"epoch": 61.19230769230769,
"grad_norm": 0.5728083848953247,
"learning_rate": 3.302854028049134e-06,
"loss": 0.034,
"step": 15910
},
{
"epoch": 61.23076923076923,
"grad_norm": 0.6408311724662781,
"learning_rate": 3.287343531612233e-06,
"loss": 0.0301,
"step": 15920
},
{
"epoch": 61.26923076923077,
"grad_norm": 0.7218260169029236,
"learning_rate": 3.271865057130874e-06,
"loss": 0.0386,
"step": 15930
},
{
"epoch": 61.30769230769231,
"grad_norm": 0.7336766719818115,
"learning_rate": 3.2564186469226064e-06,
"loss": 0.0279,
"step": 15940
},
{
"epoch": 61.34615384615385,
"grad_norm": 0.7276951670646667,
"learning_rate": 3.2410043432173353e-06,
"loss": 0.0299,
"step": 15950
},
{
"epoch": 61.38461538461539,
"grad_norm": 0.8573492765426636,
"learning_rate": 3.2256221881571734e-06,
"loss": 0.0343,
"step": 15960
},
{
"epoch": 61.42307692307692,
"grad_norm": 0.7509027719497681,
"learning_rate": 3.210272223796341e-06,
"loss": 0.0303,
"step": 15970
},
{
"epoch": 61.46153846153846,
"grad_norm": 0.741800844669342,
"learning_rate": 3.1949544921010637e-06,
"loss": 0.0299,
"step": 15980
},
{
"epoch": 61.5,
"grad_norm": 0.580657958984375,
"learning_rate": 3.1796690349494273e-06,
"loss": 0.028,
"step": 15990
},
{
"epoch": 61.53846153846154,
"grad_norm": 0.7470554709434509,
"learning_rate": 3.1644158941312935e-06,
"loss": 0.0278,
"step": 16000
},
{
"epoch": 61.57692307692308,
"grad_norm": 0.6147487163543701,
"learning_rate": 3.149195111348166e-06,
"loss": 0.0278,
"step": 16010
},
{
"epoch": 61.61538461538461,
"grad_norm": 0.9255980849266052,
"learning_rate": 3.1340067282130765e-06,
"loss": 0.0248,
"step": 16020
},
{
"epoch": 61.65384615384615,
"grad_norm": 0.6000899076461792,
"learning_rate": 3.118850786250495e-06,
"loss": 0.027,
"step": 16030
},
{
"epoch": 61.69230769230769,
"grad_norm": 0.5992773175239563,
"learning_rate": 3.1037273268961836e-06,
"loss": 0.0319,
"step": 16040
},
{
"epoch": 61.73076923076923,
"grad_norm": 0.5813210606575012,
"learning_rate": 3.0886363914970994e-06,
"loss": 0.0312,
"step": 16050
},
{
"epoch": 61.76923076923077,
"grad_norm": 0.6334101557731628,
"learning_rate": 3.0735780213112896e-06,
"loss": 0.0266,
"step": 16060
},
{
"epoch": 61.80769230769231,
"grad_norm": 0.6605973243713379,
"learning_rate": 3.0585522575077558e-06,
"loss": 0.0302,
"step": 16070
},
{
"epoch": 61.84615384615385,
"grad_norm": 0.716547429561615,
"learning_rate": 3.043559141166372e-06,
"loss": 0.0328,
"step": 16080
},
{
"epoch": 61.88461538461539,
"grad_norm": 0.7207222580909729,
"learning_rate": 3.02859871327774e-06,
"loss": 0.0299,
"step": 16090
},
{
"epoch": 61.92307692307692,
"grad_norm": 0.5931048393249512,
"learning_rate": 3.0136710147430945e-06,
"loss": 0.0287,
"step": 16100
},
{
"epoch": 61.96153846153846,
"grad_norm": 0.6544404625892639,
"learning_rate": 2.998776086374202e-06,
"loss": 0.0286,
"step": 16110
},
{
"epoch": 62.0,
"grad_norm": 0.7510085701942444,
"learning_rate": 2.983913968893221e-06,
"loss": 0.0299,
"step": 16120
},
{
"epoch": 62.03846153846154,
"grad_norm": 0.6854420900344849,
"learning_rate": 2.969084702932619e-06,
"loss": 0.0283,
"step": 16130
},
{
"epoch": 62.07692307692308,
"grad_norm": 0.7260907292366028,
"learning_rate": 2.95428832903504e-06,
"loss": 0.0311,
"step": 16140
},
{
"epoch": 62.11538461538461,
"grad_norm": 0.6026385426521301,
"learning_rate": 2.939524887653201e-06,
"loss": 0.0288,
"step": 16150
},
{
"epoch": 62.15384615384615,
"grad_norm": 0.5275495052337646,
"learning_rate": 2.924794419149796e-06,
"loss": 0.0326,
"step": 16160
},
{
"epoch": 62.19230769230769,
"grad_norm": 0.597240149974823,
"learning_rate": 2.910096963797354e-06,
"loss": 0.0272,
"step": 16170
},
{
"epoch": 62.23076923076923,
"grad_norm": 0.6819244623184204,
"learning_rate": 2.895432561778164e-06,
"loss": 0.0344,
"step": 16180
},
{
"epoch": 62.26923076923077,
"grad_norm": 0.624423623085022,
"learning_rate": 2.8808012531841393e-06,
"loss": 0.0297,
"step": 16190
},
{
"epoch": 62.30769230769231,
"grad_norm": 0.5835046768188477,
"learning_rate": 2.8662030780167138e-06,
"loss": 0.033,
"step": 16200
},
{
"epoch": 62.34615384615385,
"grad_norm": 0.6391960382461548,
"learning_rate": 2.85163807618675e-06,
"loss": 0.0254,
"step": 16210
},
{
"epoch": 62.38461538461539,
"grad_norm": 0.6837576031684875,
"learning_rate": 2.837106287514397e-06,
"loss": 0.029,
"step": 16220
},
{
"epoch": 62.42307692307692,
"grad_norm": 0.8031293153762817,
"learning_rate": 2.822607751729018e-06,
"loss": 0.0304,
"step": 16230
},
{
"epoch": 62.46153846153846,
"grad_norm": 0.7235280871391296,
"learning_rate": 2.808142508469054e-06,
"loss": 0.0302,
"step": 16240
},
{
"epoch": 62.5,
"grad_norm": 0.7115002870559692,
"learning_rate": 2.7937105972819237e-06,
"loss": 0.0287,
"step": 16250
},
{
"epoch": 62.53846153846154,
"grad_norm": 0.7001069784164429,
"learning_rate": 2.7793120576239285e-06,
"loss": 0.0312,
"step": 16260
},
{
"epoch": 62.57692307692308,
"grad_norm": 0.6503643989562988,
"learning_rate": 2.7649469288601175e-06,
"loss": 0.0292,
"step": 16270
},
{
"epoch": 62.61538461538461,
"grad_norm": 0.5469167232513428,
"learning_rate": 2.7506152502642125e-06,
"loss": 0.0271,
"step": 16280
},
{
"epoch": 62.65384615384615,
"grad_norm": 0.6128395199775696,
"learning_rate": 2.7363170610184716e-06,
"loss": 0.0321,
"step": 16290
},
{
"epoch": 62.69230769230769,
"grad_norm": 0.726383626461029,
"learning_rate": 2.722052400213595e-06,
"loss": 0.0298,
"step": 16300
},
{
"epoch": 62.73076923076923,
"grad_norm": 0.7536424398422241,
"learning_rate": 2.707821306848627e-06,
"loss": 0.0309,
"step": 16310
},
{
"epoch": 62.76923076923077,
"grad_norm": 0.7245988249778748,
"learning_rate": 2.6936238198308318e-06,
"loss": 0.0283,
"step": 16320
},
{
"epoch": 62.80769230769231,
"grad_norm": 0.6714974641799927,
"learning_rate": 2.67945997797559e-06,
"loss": 0.0323,
"step": 16330
},
{
"epoch": 62.84615384615385,
"grad_norm": 0.5724020600318909,
"learning_rate": 2.665329820006314e-06,
"loss": 0.0282,
"step": 16340
},
{
"epoch": 62.88461538461539,
"grad_norm": 0.7046947479248047,
"learning_rate": 2.6512333845543086e-06,
"loss": 0.0274,
"step": 16350
},
{
"epoch": 62.92307692307692,
"grad_norm": 0.6365247368812561,
"learning_rate": 2.637170710158697e-06,
"loss": 0.0315,
"step": 16360
},
{
"epoch": 62.96153846153846,
"grad_norm": 0.7064822316169739,
"learning_rate": 2.6231418352662895e-06,
"loss": 0.0296,
"step": 16370
},
{
"epoch": 63.0,
"grad_norm": 0.6772658228874207,
"learning_rate": 2.609146798231493e-06,
"loss": 0.033,
"step": 16380
},
{
"epoch": 63.03846153846154,
"grad_norm": 0.48029395937919617,
"learning_rate": 2.5951856373162097e-06,
"loss": 0.0295,
"step": 16390
},
{
"epoch": 63.07692307692308,
"grad_norm": 0.6232547163963318,
"learning_rate": 2.581258390689712e-06,
"loss": 0.0288,
"step": 16400
},
{
"epoch": 63.11538461538461,
"grad_norm": 0.8133201599121094,
"learning_rate": 2.5673650964285718e-06,
"loss": 0.0326,
"step": 16410
},
{
"epoch": 63.15384615384615,
"grad_norm": 0.8444731831550598,
"learning_rate": 2.553505792516518e-06,
"loss": 0.0265,
"step": 16420
},
{
"epoch": 63.19230769230769,
"grad_norm": 0.4815030097961426,
"learning_rate": 2.539680516844356e-06,
"loss": 0.0249,
"step": 16430
},
{
"epoch": 63.23076923076923,
"grad_norm": 0.6386057138442993,
"learning_rate": 2.5258893072098678e-06,
"loss": 0.0282,
"step": 16440
},
{
"epoch": 63.26923076923077,
"grad_norm": 0.701069176197052,
"learning_rate": 2.512132201317688e-06,
"loss": 0.0269,
"step": 16450
},
{
"epoch": 63.30769230769231,
"grad_norm": 1.0269007682800293,
"learning_rate": 2.4984092367792272e-06,
"loss": 0.0301,
"step": 16460
},
{
"epoch": 63.34615384615385,
"grad_norm": 0.6145013570785522,
"learning_rate": 2.484720451112536e-06,
"loss": 0.0282,
"step": 16470
},
{
"epoch": 63.38461538461539,
"grad_norm": 0.6777820587158203,
"learning_rate": 2.471065881742236e-06,
"loss": 0.0306,
"step": 16480
},
{
"epoch": 63.42307692307692,
"grad_norm": 0.7301444411277771,
"learning_rate": 2.4574455659994023e-06,
"loss": 0.0326,
"step": 16490
},
{
"epoch": 63.46153846153846,
"grad_norm": 0.6660694479942322,
"learning_rate": 2.4438595411214528e-06,
"loss": 0.029,
"step": 16500
},
{
"epoch": 63.5,
"grad_norm": 0.6402868032455444,
"learning_rate": 2.430307844252069e-06,
"loss": 0.0262,
"step": 16510
},
{
"epoch": 63.53846153846154,
"grad_norm": 0.7359407544136047,
"learning_rate": 2.4167905124410587e-06,
"loss": 0.0276,
"step": 16520
},
{
"epoch": 63.57692307692308,
"grad_norm": 0.5874005556106567,
"learning_rate": 2.4033075826442995e-06,
"loss": 0.0279,
"step": 16530
},
{
"epoch": 63.61538461538461,
"grad_norm": 0.7026081085205078,
"learning_rate": 2.389859091723608e-06,
"loss": 0.0284,
"step": 16540
},
{
"epoch": 63.65384615384615,
"grad_norm": 0.7727426886558533,
"learning_rate": 2.376445076446641e-06,
"loss": 0.0307,
"step": 16550
},
{
"epoch": 63.69230769230769,
"grad_norm": 0.7311943769454956,
"learning_rate": 2.3630655734868117e-06,
"loss": 0.0275,
"step": 16560
},
{
"epoch": 63.73076923076923,
"grad_norm": 0.7108755111694336,
"learning_rate": 2.349720619423158e-06,
"loss": 0.0293,
"step": 16570
},
{
"epoch": 63.76923076923077,
"grad_norm": 0.981634795665741,
"learning_rate": 2.3364102507402817e-06,
"loss": 0.0329,
"step": 16580
},
{
"epoch": 63.80769230769231,
"grad_norm": 0.7809610366821289,
"learning_rate": 2.3231345038282243e-06,
"loss": 0.0287,
"step": 16590
},
{
"epoch": 63.84615384615385,
"grad_norm": 0.9093500971794128,
"learning_rate": 2.3098934149823686e-06,
"loss": 0.0291,
"step": 16600
},
{
"epoch": 63.88461538461539,
"grad_norm": 0.6164126992225647,
"learning_rate": 2.296687020403346e-06,
"loss": 0.0306,
"step": 16610
},
{
"epoch": 63.92307692307692,
"grad_norm": 0.7200257778167725,
"learning_rate": 2.2835153561969322e-06,
"loss": 0.0309,
"step": 16620
},
{
"epoch": 63.96153846153846,
"grad_norm": 0.5784043669700623,
"learning_rate": 2.270378458373956e-06,
"loss": 0.027,
"step": 16630
},
{
"epoch": 64.0,
"grad_norm": 0.5341262221336365,
"learning_rate": 2.257276362850199e-06,
"loss": 0.0283,
"step": 16640
},
{
"epoch": 64.03846153846153,
"grad_norm": 0.47131168842315674,
"learning_rate": 2.244209105446286e-06,
"loss": 0.0304,
"step": 16650
},
{
"epoch": 64.07692307692308,
"grad_norm": 0.5508049726486206,
"learning_rate": 2.2311767218875995e-06,
"loss": 0.0285,
"step": 16660
},
{
"epoch": 64.11538461538461,
"grad_norm": 0.6628520488739014,
"learning_rate": 2.218179247804177e-06,
"loss": 0.0306,
"step": 16670
},
{
"epoch": 64.15384615384616,
"grad_norm": 0.6644569635391235,
"learning_rate": 2.2052167187306167e-06,
"loss": 0.0269,
"step": 16680
},
{
"epoch": 64.1923076923077,
"grad_norm": 0.8234758973121643,
"learning_rate": 2.192289170105989e-06,
"loss": 0.0286,
"step": 16690
},
{
"epoch": 64.23076923076923,
"grad_norm": 0.7737919688224792,
"learning_rate": 2.1793966372737003e-06,
"loss": 0.0339,
"step": 16700
},
{
"epoch": 64.26923076923077,
"grad_norm": 0.5969371795654297,
"learning_rate": 2.166539155481455e-06,
"loss": 0.025,
"step": 16710
},
{
"epoch": 64.3076923076923,
"grad_norm": 0.5700958371162415,
"learning_rate": 2.1537167598811118e-06,
"loss": 0.0321,
"step": 16720
},
{
"epoch": 64.34615384615384,
"grad_norm": 0.625689685344696,
"learning_rate": 2.140929485528612e-06,
"loss": 0.0267,
"step": 16730
},
{
"epoch": 64.38461538461539,
"grad_norm": 0.504895031452179,
"learning_rate": 2.1281773673838838e-06,
"loss": 0.0244,
"step": 16740
},
{
"epoch": 64.42307692307692,
"grad_norm": 0.5708885788917542,
"learning_rate": 2.1154604403107175e-06,
"loss": 0.0261,
"step": 16750
},
{
"epoch": 64.46153846153847,
"grad_norm": 0.7020063400268555,
"learning_rate": 2.102778739076715e-06,
"loss": 0.0299,
"step": 16760
},
{
"epoch": 64.5,
"grad_norm": 0.5950106978416443,
"learning_rate": 2.0901322983531574e-06,
"loss": 0.0323,
"step": 16770
},
{
"epoch": 64.53846153846153,
"grad_norm": 0.6007116436958313,
"learning_rate": 2.0775211527149357e-06,
"loss": 0.0272,
"step": 16780
},
{
"epoch": 64.57692307692308,
"grad_norm": 0.7846446633338928,
"learning_rate": 2.0649453366404438e-06,
"loss": 0.03,
"step": 16790
},
{
"epoch": 64.61538461538461,
"grad_norm": 0.6095095276832581,
"learning_rate": 2.052404884511472e-06,
"loss": 0.0284,
"step": 16800
},
{
"epoch": 64.65384615384616,
"grad_norm": 0.6284276247024536,
"learning_rate": 2.039899830613145e-06,
"loss": 0.0331,
"step": 16810
},
{
"epoch": 64.6923076923077,
"grad_norm": 0.6151442527770996,
"learning_rate": 2.0274302091337987e-06,
"loss": 0.0297,
"step": 16820
},
{
"epoch": 64.73076923076923,
"grad_norm": 0.6171903610229492,
"learning_rate": 2.0149960541649076e-06,
"loss": 0.0279,
"step": 16830
},
{
"epoch": 64.76923076923077,
"grad_norm": 0.6822643876075745,
"learning_rate": 2.002597399700974e-06,
"loss": 0.0295,
"step": 16840
},
{
"epoch": 64.8076923076923,
"grad_norm": 0.6247174143791199,
"learning_rate": 1.990234279639441e-06,
"loss": 0.029,
"step": 16850
},
{
"epoch": 64.84615384615384,
"grad_norm": 0.6442417502403259,
"learning_rate": 1.977906727780614e-06,
"loss": 0.0281,
"step": 16860
},
{
"epoch": 64.88461538461539,
"grad_norm": 0.45084184408187866,
"learning_rate": 1.9656147778275423e-06,
"loss": 0.0276,
"step": 16870
},
{
"epoch": 64.92307692307692,
"grad_norm": 0.5375664830207825,
"learning_rate": 1.953358463385954e-06,
"loss": 0.0329,
"step": 16880
},
{
"epoch": 64.96153846153847,
"grad_norm": 0.6008625626564026,
"learning_rate": 1.9411378179641435e-06,
"loss": 0.0349,
"step": 16890
},
{
"epoch": 65.0,
"grad_norm": 0.7332527041435242,
"learning_rate": 1.9289528749728834e-06,
"loss": 0.0374,
"step": 16900
},
{
"epoch": 65.03846153846153,
"grad_norm": 0.6559941172599792,
"learning_rate": 1.916803667725351e-06,
"loss": 0.0289,
"step": 16910
},
{
"epoch": 65.07692307692308,
"grad_norm": 0.6436246633529663,
"learning_rate": 1.9046902294370044e-06,
"loss": 0.0286,
"step": 16920
},
{
"epoch": 65.11538461538461,
"grad_norm": 0.7137200832366943,
"learning_rate": 1.8926125932255328e-06,
"loss": 0.0278,
"step": 16930
},
{
"epoch": 65.15384615384616,
"grad_norm": 0.7713292837142944,
"learning_rate": 1.8805707921107262e-06,
"loss": 0.0296,
"step": 16940
},
{
"epoch": 65.1923076923077,
"grad_norm": 0.4665347635746002,
"learning_rate": 1.8685648590144066e-06,
"loss": 0.0272,
"step": 16950
},
{
"epoch": 65.23076923076923,
"grad_norm": 0.5581397414207458,
"learning_rate": 1.8565948267603444e-06,
"loss": 0.0327,
"step": 16960
},
{
"epoch": 65.26923076923077,
"grad_norm": 0.4912269711494446,
"learning_rate": 1.8446607280741435e-06,
"loss": 0.0339,
"step": 16970
},
{
"epoch": 65.3076923076923,
"grad_norm": 0.6458134651184082,
"learning_rate": 1.8327625955831763e-06,
"loss": 0.0272,
"step": 16980
},
{
"epoch": 65.34615384615384,
"grad_norm": 0.6138810515403748,
"learning_rate": 1.8209004618164837e-06,
"loss": 0.0306,
"step": 16990
},
{
"epoch": 65.38461538461539,
"grad_norm": 0.6538389921188354,
"learning_rate": 1.8090743592046843e-06,
"loss": 0.0342,
"step": 17000
},
{
"epoch": 65.42307692307692,
"grad_norm": 0.5684748888015747,
"learning_rate": 1.7972843200798932e-06,
"loss": 0.0281,
"step": 17010
},
{
"epoch": 65.46153846153847,
"grad_norm": 0.6487337350845337,
"learning_rate": 1.7855303766756316e-06,
"loss": 0.0303,
"step": 17020
},
{
"epoch": 65.5,
"grad_norm": 0.6523866057395935,
"learning_rate": 1.7738125611267204e-06,
"loss": 0.029,
"step": 17030
},
{
"epoch": 65.53846153846153,
"grad_norm": 0.624439001083374,
"learning_rate": 1.7621309054692302e-06,
"loss": 0.027,
"step": 17040
},
{
"epoch": 65.57692307692308,
"grad_norm": 0.45990708470344543,
"learning_rate": 1.7504854416403542e-06,
"loss": 0.0288,
"step": 17050
},
{
"epoch": 65.61538461538461,
"grad_norm": 0.7313023209571838,
"learning_rate": 1.7388762014783493e-06,
"loss": 0.0287,
"step": 17060
},
{
"epoch": 65.65384615384616,
"grad_norm": 0.7057033777236938,
"learning_rate": 1.7273032167224418e-06,
"loss": 0.0322,
"step": 17070
},
{
"epoch": 65.6923076923077,
"grad_norm": 0.6121751070022583,
"learning_rate": 1.7157665190127154e-06,
"loss": 0.0325,
"step": 17080
},
{
"epoch": 65.73076923076923,
"grad_norm": 0.4792765974998474,
"learning_rate": 1.7042661398900733e-06,
"loss": 0.0253,
"step": 17090
},
{
"epoch": 65.76923076923077,
"grad_norm": 0.6455287933349609,
"learning_rate": 1.692802110796105e-06,
"loss": 0.0292,
"step": 17100
},
{
"epoch": 65.8076923076923,
"grad_norm": 0.8495745658874512,
"learning_rate": 1.6813744630730343e-06,
"loss": 0.0326,
"step": 17110
},
{
"epoch": 65.84615384615384,
"grad_norm": 0.7053533792495728,
"learning_rate": 1.6699832279636113e-06,
"loss": 0.0312,
"step": 17120
},
{
"epoch": 65.88461538461539,
"grad_norm": 0.6184305548667908,
"learning_rate": 1.6586284366110355e-06,
"loss": 0.0362,
"step": 17130
},
{
"epoch": 65.92307692307692,
"grad_norm": 0.4716685712337494,
"learning_rate": 1.647310120058878e-06,
"loss": 0.0339,
"step": 17140
},
{
"epoch": 65.96153846153847,
"grad_norm": 0.7297793030738831,
"learning_rate": 1.6360283092509765e-06,
"loss": 0.0292,
"step": 17150
},
{
"epoch": 66.0,
"grad_norm": 0.6194086074829102,
"learning_rate": 1.6247830350313797e-06,
"loss": 0.0309,
"step": 17160
},
{
"epoch": 66.03846153846153,
"grad_norm": 0.5371058583259583,
"learning_rate": 1.6135743281442333e-06,
"loss": 0.028,
"step": 17170
},
{
"epoch": 66.07692307692308,
"grad_norm": 0.6116629838943481,
"learning_rate": 1.6024022192337112e-06,
"loss": 0.0293,
"step": 17180
},
{
"epoch": 66.11538461538461,
"grad_norm": 0.5863935947418213,
"learning_rate": 1.591266738843939e-06,
"loss": 0.0276,
"step": 17190
},
{
"epoch": 66.15384615384616,
"grad_norm": 0.7205275893211365,
"learning_rate": 1.5801679174188888e-06,
"loss": 0.0295,
"step": 17200
},
{
"epoch": 66.1923076923077,
"grad_norm": 0.5224143266677856,
"learning_rate": 1.5691057853023199e-06,
"loss": 0.0313,
"step": 17210
},
{
"epoch": 66.23076923076923,
"grad_norm": 0.49154120683670044,
"learning_rate": 1.5580803727376786e-06,
"loss": 0.0278,
"step": 17220
},
{
"epoch": 66.26923076923077,
"grad_norm": 0.43938741087913513,
"learning_rate": 1.5470917098680142e-06,
"loss": 0.0275,
"step": 17230
},
{
"epoch": 66.3076923076923,
"grad_norm": 0.5548313856124878,
"learning_rate": 1.5361398267359205e-06,
"loss": 0.0291,
"step": 17240
},
{
"epoch": 66.34615384615384,
"grad_norm": 0.5422559976577759,
"learning_rate": 1.5252247532834246e-06,
"loss": 0.0318,
"step": 17250
},
{
"epoch": 66.38461538461539,
"grad_norm": 0.6111128330230713,
"learning_rate": 1.5143465193519173e-06,
"loss": 0.0294,
"step": 17260
},
{
"epoch": 66.42307692307692,
"grad_norm": 0.5277599692344666,
"learning_rate": 1.5035051546820821e-06,
"loss": 0.0297,
"step": 17270
},
{
"epoch": 66.46153846153847,
"grad_norm": 0.7245263457298279,
"learning_rate": 1.4927006889137862e-06,
"loss": 0.0317,
"step": 17280
},
{
"epoch": 66.5,
"grad_norm": 0.5456480383872986,
"learning_rate": 1.4819331515860357e-06,
"loss": 0.0279,
"step": 17290
},
{
"epoch": 66.53846153846153,
"grad_norm": 0.5236904621124268,
"learning_rate": 1.4712025721368644e-06,
"loss": 0.0324,
"step": 17300
},
{
"epoch": 66.57692307692308,
"grad_norm": 0.45201101899147034,
"learning_rate": 1.46050897990326e-06,
"loss": 0.0316,
"step": 17310
},
{
"epoch": 66.61538461538461,
"grad_norm": 0.41527774930000305,
"learning_rate": 1.449852404121103e-06,
"loss": 0.026,
"step": 17320
},
{
"epoch": 66.65384615384616,
"grad_norm": 0.5794553756713867,
"learning_rate": 1.4392328739250615e-06,
"loss": 0.0311,
"step": 17330
},
{
"epoch": 66.6923076923077,
"grad_norm": 0.4518781006336212,
"learning_rate": 1.4286504183485277e-06,
"loss": 0.0306,
"step": 17340
},
{
"epoch": 66.73076923076923,
"grad_norm": 0.7854360342025757,
"learning_rate": 1.4181050663235284e-06,
"loss": 0.0315,
"step": 17350
},
{
"epoch": 66.76923076923077,
"grad_norm": 0.6389064788818359,
"learning_rate": 1.4075968466806533e-06,
"loss": 0.0302,
"step": 17360
},
{
"epoch": 66.8076923076923,
"grad_norm": 0.4904123544692993,
"learning_rate": 1.3971257881489762e-06,
"loss": 0.0279,
"step": 17370
},
{
"epoch": 66.84615384615384,
"grad_norm": 0.5139650702476501,
"learning_rate": 1.386691919355968e-06,
"loss": 0.0258,
"step": 17380
},
{
"epoch": 66.88461538461539,
"grad_norm": 0.5151085257530212,
"learning_rate": 1.3762952688274316e-06,
"loss": 0.0256,
"step": 17390
},
{
"epoch": 66.92307692307692,
"grad_norm": 0.5301920175552368,
"learning_rate": 1.3659358649874104e-06,
"loss": 0.026,
"step": 17400
},
{
"epoch": 66.96153846153847,
"grad_norm": 0.7160604000091553,
"learning_rate": 1.3556137361581155e-06,
"loss": 0.0295,
"step": 17410
},
{
"epoch": 67.0,
"grad_norm": 0.7122892141342163,
"learning_rate": 1.3453289105598616e-06,
"loss": 0.0316,
"step": 17420
},
{
"epoch": 67.03846153846153,
"grad_norm": 0.6626392602920532,
"learning_rate": 1.3350814163109592e-06,
"loss": 0.0333,
"step": 17430
},
{
"epoch": 67.07692307692308,
"grad_norm": 0.5471773147583008,
"learning_rate": 1.3248712814276732e-06,
"loss": 0.026,
"step": 17440
},
{
"epoch": 67.11538461538461,
"grad_norm": 0.4416269361972809,
"learning_rate": 1.3146985338241207e-06,
"loss": 0.0281,
"step": 17450
},
{
"epoch": 67.15384615384616,
"grad_norm": 0.5858498811721802,
"learning_rate": 1.3045632013122032e-06,
"loss": 0.0295,
"step": 17460
},
{
"epoch": 67.1923076923077,
"grad_norm": 0.5736910104751587,
"learning_rate": 1.294465311601537e-06,
"loss": 0.0285,
"step": 17470
},
{
"epoch": 67.23076923076923,
"grad_norm": 0.547957718372345,
"learning_rate": 1.2844048922993602e-06,
"loss": 0.0311,
"step": 17480
},
{
"epoch": 67.26923076923077,
"grad_norm": 0.5907455682754517,
"learning_rate": 1.2743819709104826e-06,
"loss": 0.0282,
"step": 17490
},
{
"epoch": 67.3076923076923,
"grad_norm": 0.5498229265213013,
"learning_rate": 1.264396574837185e-06,
"loss": 0.0239,
"step": 17500
},
{
"epoch": 67.34615384615384,
"grad_norm": 0.3337724804878235,
"learning_rate": 1.2544487313791564e-06,
"loss": 0.0257,
"step": 17510
},
{
"epoch": 67.38461538461539,
"grad_norm": 0.5026791095733643,
"learning_rate": 1.2445384677334282e-06,
"loss": 0.0293,
"step": 17520
},
{
"epoch": 67.42307692307692,
"grad_norm": 0.5449739098548889,
"learning_rate": 1.2346658109942755e-06,
"loss": 0.0312,
"step": 17530
},
{
"epoch": 67.46153846153847,
"grad_norm": 0.5108739733695984,
"learning_rate": 1.2248307881531656e-06,
"loss": 0.0338,
"step": 17540
},
{
"epoch": 67.5,
"grad_norm": 0.6045510768890381,
"learning_rate": 1.2150334260986818e-06,
"loss": 0.0329,
"step": 17550
},
{
"epoch": 67.53846153846153,
"grad_norm": 0.6545273661613464,
"learning_rate": 1.2052737516164292e-06,
"loss": 0.0303,
"step": 17560
},
{
"epoch": 67.57692307692308,
"grad_norm": 0.558460533618927,
"learning_rate": 1.1955517913889924e-06,
"loss": 0.0286,
"step": 17570
},
{
"epoch": 67.61538461538461,
"grad_norm": 0.704518735408783,
"learning_rate": 1.185867571995835e-06,
"loss": 0.0304,
"step": 17580
},
{
"epoch": 67.65384615384616,
"grad_norm": 0.7233688831329346,
"learning_rate": 1.1762211199132433e-06,
"loss": 0.0262,
"step": 17590
},
{
"epoch": 67.6923076923077,
"grad_norm": 0.45494726300239563,
"learning_rate": 1.1666124615142525e-06,
"loss": 0.0253,
"step": 17600
},
{
"epoch": 67.73076923076923,
"grad_norm": 0.4010249972343445,
"learning_rate": 1.1570416230685627e-06,
"loss": 0.0254,
"step": 17610
},
{
"epoch": 67.76923076923077,
"grad_norm": 0.4817158877849579,
"learning_rate": 1.147508630742486e-06,
"loss": 0.0273,
"step": 17620
},
{
"epoch": 67.8076923076923,
"grad_norm": 0.5177744030952454,
"learning_rate": 1.1380135105988576e-06,
"loss": 0.0287,
"step": 17630
},
{
"epoch": 67.84615384615384,
"grad_norm": 0.5829024314880371,
"learning_rate": 1.128556288596969e-06,
"loss": 0.0316,
"step": 17640
},
{
"epoch": 67.88461538461539,
"grad_norm": 0.5057424902915955,
"learning_rate": 1.1191369905925096e-06,
"loss": 0.0276,
"step": 17650
},
{
"epoch": 67.92307692307692,
"grad_norm": 0.6341408491134644,
"learning_rate": 1.1097556423374765e-06,
"loss": 0.0304,
"step": 17660
},
{
"epoch": 67.96153846153847,
"grad_norm": 0.6067168116569519,
"learning_rate": 1.1004122694801233e-06,
"loss": 0.0364,
"step": 17670
},
{
"epoch": 68.0,
"grad_norm": 0.4588167071342468,
"learning_rate": 1.0911068975648697e-06,
"loss": 0.026,
"step": 17680
},
{
"epoch": 68.03846153846153,
"grad_norm": 0.5819829106330872,
"learning_rate": 1.0818395520322456e-06,
"loss": 0.0262,
"step": 17690
},
{
"epoch": 68.07692307692308,
"grad_norm": 0.5707022547721863,
"learning_rate": 1.072610258218825e-06,
"loss": 0.0361,
"step": 17700
},
{
"epoch": 68.11538461538461,
"grad_norm": 0.5363244414329529,
"learning_rate": 1.0634190413571415e-06,
"loss": 0.0281,
"step": 17710
},
{
"epoch": 68.15384615384616,
"grad_norm": 0.6520892381668091,
"learning_rate": 1.0542659265756337e-06,
"loss": 0.0303,
"step": 17720
},
{
"epoch": 68.1923076923077,
"grad_norm": 0.6424570083618164,
"learning_rate": 1.0451509388985663e-06,
"loss": 0.0327,
"step": 17730
},
{
"epoch": 68.23076923076923,
"grad_norm": 0.5023139715194702,
"learning_rate": 1.0360741032459636e-06,
"loss": 0.0309,
"step": 17740
},
{
"epoch": 68.26923076923077,
"grad_norm": 0.4547223448753357,
"learning_rate": 1.027035444433555e-06,
"loss": 0.0261,
"step": 17750
},
{
"epoch": 68.3076923076923,
"grad_norm": 0.440268337726593,
"learning_rate": 1.0180349871726819e-06,
"loss": 0.0304,
"step": 17760
},
{
"epoch": 68.34615384615384,
"grad_norm": 0.4488914906978607,
"learning_rate": 1.0090727560702572e-06,
"loss": 0.0321,
"step": 17770
},
{
"epoch": 68.38461538461539,
"grad_norm": 0.47670313715934753,
"learning_rate": 1.0001487756286748e-06,
"loss": 0.0292,
"step": 17780
},
{
"epoch": 68.42307692307692,
"grad_norm": 0.6052711009979248,
"learning_rate": 9.912630702457548e-07,
"loss": 0.0279,
"step": 17790
},
{
"epoch": 68.46153846153847,
"grad_norm": 0.38048213720321655,
"learning_rate": 9.824156642146798e-07,
"loss": 0.032,
"step": 17800
},
{
"epoch": 68.5,
"grad_norm": 0.5576040744781494,
"learning_rate": 9.736065817239192e-07,
"loss": 0.0292,
"step": 17810
},
{
"epoch": 68.53846153846153,
"grad_norm": 0.8681022524833679,
"learning_rate": 9.648358468571667e-07,
"loss": 0.0283,
"step": 17820
},
{
"epoch": 68.57692307692308,
"grad_norm": 0.6247050166130066,
"learning_rate": 9.561034835932774e-07,
"loss": 0.0285,
"step": 17830
},
{
"epoch": 68.61538461538461,
"grad_norm": 0.5081879496574402,
"learning_rate": 9.474095158061996e-07,
"loss": 0.032,
"step": 17840
},
{
"epoch": 68.65384615384616,
"grad_norm": 0.631948709487915,
"learning_rate": 9.387539672649082e-07,
"loss": 0.0398,
"step": 17850
},
{
"epoch": 68.6923076923077,
"grad_norm": 0.5832291841506958,
"learning_rate": 9.301368616333456e-07,
"loss": 0.0268,
"step": 17860
},
{
"epoch": 68.73076923076923,
"grad_norm": 0.5857760310173035,
"learning_rate": 9.215582224703417e-07,
"loss": 0.0268,
"step": 17870
},
{
"epoch": 68.76923076923077,
"grad_norm": 0.63872891664505,
"learning_rate": 9.13018073229579e-07,
"loss": 0.0329,
"step": 17880
},
{
"epoch": 68.8076923076923,
"grad_norm": 0.34974420070648193,
"learning_rate": 9.045164372594889e-07,
"loss": 0.0268,
"step": 17890
},
{
"epoch": 68.84615384615384,
"grad_norm": 0.4673135578632355,
"learning_rate": 8.960533378032288e-07,
"loss": 0.0308,
"step": 17900
},
{
"epoch": 68.88461538461539,
"grad_norm": 0.496360719203949,
"learning_rate": 8.876287979985853e-07,
"loss": 0.0268,
"step": 17910
},
{
"epoch": 68.92307692307692,
"grad_norm": 0.4563571810722351,
"learning_rate": 8.792428408779246e-07,
"loss": 0.033,
"step": 17920
},
{
"epoch": 68.96153846153847,
"grad_norm": 0.5708230137825012,
"learning_rate": 8.708954893681421e-07,
"loss": 0.0303,
"step": 17930
},
{
"epoch": 69.0,
"grad_norm": 0.624557614326477,
"learning_rate": 8.62586766290569e-07,
"loss": 0.0317,
"step": 17940
},
{
"epoch": 69.03846153846153,
"grad_norm": 0.3131464123725891,
"learning_rate": 8.543166943609448e-07,
"loss": 0.027,
"step": 17950
},
{
"epoch": 69.07692307692308,
"grad_norm": 0.8714306950569153,
"learning_rate": 8.460852961893234e-07,
"loss": 0.0309,
"step": 17960
},
{
"epoch": 69.11538461538461,
"grad_norm": 0.6325943470001221,
"learning_rate": 8.378925942800364e-07,
"loss": 0.0349,
"step": 17970
},
{
"epoch": 69.15384615384616,
"grad_norm": 0.44092488288879395,
"learning_rate": 8.297386110316202e-07,
"loss": 0.0275,
"step": 17980
},
{
"epoch": 69.1923076923077,
"grad_norm": 0.5496897101402283,
"learning_rate": 8.216233687367491e-07,
"loss": 0.0317,
"step": 17990
},
{
"epoch": 69.23076923076923,
"grad_norm": 0.615792453289032,
"learning_rate": 8.135468895821924e-07,
"loss": 0.0319,
"step": 18000
},
{
"epoch": 69.26923076923077,
"grad_norm": 0.6132825016975403,
"learning_rate": 8.05509195648727e-07,
"loss": 0.0296,
"step": 18010
},
{
"epoch": 69.3076923076923,
"grad_norm": 0.4234136641025543,
"learning_rate": 7.975103089111052e-07,
"loss": 0.0329,
"step": 18020
},
{
"epoch": 69.34615384615384,
"grad_norm": 0.5253662467002869,
"learning_rate": 7.895502512379805e-07,
"loss": 0.0327,
"step": 18030
},
{
"epoch": 69.38461538461539,
"grad_norm": 0.40497887134552,
"learning_rate": 7.816290443918411e-07,
"loss": 0.0297,
"step": 18040
},
{
"epoch": 69.42307692307692,
"grad_norm": 0.5554369688034058,
"learning_rate": 7.737467100289725e-07,
"loss": 0.0329,
"step": 18050
},
{
"epoch": 69.46153846153847,
"grad_norm": 0.5565347671508789,
"learning_rate": 7.659032696993661e-07,
"loss": 0.0307,
"step": 18060
},
{
"epoch": 69.5,
"grad_norm": 0.3294142484664917,
"learning_rate": 7.580987448466925e-07,
"loss": 0.0279,
"step": 18070
},
{
"epoch": 69.53846153846153,
"grad_norm": 0.5592007040977478,
"learning_rate": 7.503331568082267e-07,
"loss": 0.029,
"step": 18080
},
{
"epoch": 69.57692307692308,
"grad_norm": 0.6052271127700806,
"learning_rate": 7.426065268147875e-07,
"loss": 0.027,
"step": 18090
},
{
"epoch": 69.61538461538461,
"grad_norm": 0.3888969421386719,
"learning_rate": 7.349188759906889e-07,
"loss": 0.0294,
"step": 18100
},
{
"epoch": 69.65384615384616,
"grad_norm": 0.5859863758087158,
"learning_rate": 7.272702253536683e-07,
"loss": 0.0289,
"step": 18110
},
{
"epoch": 69.6923076923077,
"grad_norm": 0.5358383655548096,
"learning_rate": 7.196605958148505e-07,
"loss": 0.0288,
"step": 18120
},
{
"epoch": 69.73076923076923,
"grad_norm": 0.39235493540763855,
"learning_rate": 7.120900081786719e-07,
"loss": 0.0242,
"step": 18130
},
{
"epoch": 69.76923076923077,
"grad_norm": 0.564791202545166,
"learning_rate": 7.045584831428276e-07,
"loss": 0.0338,
"step": 18140
},
{
"epoch": 69.8076923076923,
"grad_norm": 0.6125440001487732,
"learning_rate": 6.970660412982199e-07,
"loss": 0.0254,
"step": 18150
},
{
"epoch": 69.84615384615384,
"grad_norm": 0.6010186076164246,
"learning_rate": 6.896127031288985e-07,
"loss": 0.0328,
"step": 18160
},
{
"epoch": 69.88461538461539,
"grad_norm": 0.45994505286216736,
"learning_rate": 6.821984890120064e-07,
"loss": 0.0257,
"step": 18170
},
{
"epoch": 69.92307692307692,
"grad_norm": 0.5627428889274597,
"learning_rate": 6.748234192177227e-07,
"loss": 0.0323,
"step": 18180
},
{
"epoch": 69.96153846153847,
"grad_norm": 0.6051352024078369,
"learning_rate": 6.674875139092051e-07,
"loss": 0.0307,
"step": 18190
},
{
"epoch": 70.0,
"grad_norm": 0.3997330963611603,
"learning_rate": 6.601907931425388e-07,
"loss": 0.0247,
"step": 18200
},
{
"epoch": 70.03846153846153,
"grad_norm": 0.4721720516681671,
"learning_rate": 6.529332768666779e-07,
"loss": 0.026,
"step": 18210
},
{
"epoch": 70.07692307692308,
"grad_norm": 0.536614716053009,
"learning_rate": 6.457149849233973e-07,
"loss": 0.0282,
"step": 18220
},
{
"epoch": 70.11538461538461,
"grad_norm": 0.4795149564743042,
"learning_rate": 6.385359370472343e-07,
"loss": 0.0309,
"step": 18230
},
{
"epoch": 70.15384615384616,
"grad_norm": 0.6099283695220947,
"learning_rate": 6.313961528654239e-07,
"loss": 0.0295,
"step": 18240
},
{
"epoch": 70.1923076923077,
"grad_norm": 0.6247525215148926,
"learning_rate": 6.242956518978682e-07,
"loss": 0.0325,
"step": 18250
},
{
"epoch": 70.23076923076923,
"grad_norm": 0.3841114640235901,
"learning_rate": 6.172344535570673e-07,
"loss": 0.0253,
"step": 18260
},
{
"epoch": 70.26923076923077,
"grad_norm": 0.4295583963394165,
"learning_rate": 6.102125771480655e-07,
"loss": 0.0283,
"step": 18270
},
{
"epoch": 70.3076923076923,
"grad_norm": 0.41051971912384033,
"learning_rate": 6.032300418684062e-07,
"loss": 0.0255,
"step": 18280
},
{
"epoch": 70.34615384615384,
"grad_norm": 0.5418064594268799,
"learning_rate": 5.962868668080706e-07,
"loss": 0.0311,
"step": 18290
},
{
"epoch": 70.38461538461539,
"grad_norm": 0.6504908800125122,
"learning_rate": 5.89383070949438e-07,
"loss": 0.0319,
"step": 18300
},
{
"epoch": 70.42307692307692,
"grad_norm": 0.4007869064807892,
"learning_rate": 5.825186731672217e-07,
"loss": 0.0287,
"step": 18310
},
{
"epoch": 70.46153846153847,
"grad_norm": 0.5333831310272217,
"learning_rate": 5.756936922284228e-07,
"loss": 0.0239,
"step": 18320
},
{
"epoch": 70.5,
"grad_norm": 0.5186364650726318,
"learning_rate": 5.689081467922791e-07,
"loss": 0.03,
"step": 18330
},
{
"epoch": 70.53846153846153,
"grad_norm": 0.7441152334213257,
"learning_rate": 5.621620554102108e-07,
"loss": 0.0267,
"step": 18340
},
{
"epoch": 70.57692307692308,
"grad_norm": 0.4066428244113922,
"learning_rate": 5.554554365257747e-07,
"loss": 0.0256,
"step": 18350
},
{
"epoch": 70.61538461538461,
"grad_norm": 0.3941934108734131,
"learning_rate": 5.487883084746137e-07,
"loss": 0.031,
"step": 18360
},
{
"epoch": 70.65384615384616,
"grad_norm": 0.5388003587722778,
"learning_rate": 5.421606894843989e-07,
"loss": 0.0335,
"step": 18370
},
{
"epoch": 70.6923076923077,
"grad_norm": 0.8006777167320251,
"learning_rate": 5.355725976747878e-07,
"loss": 0.0293,
"step": 18380
},
{
"epoch": 70.73076923076923,
"grad_norm": 0.48611146211624146,
"learning_rate": 5.290240510573707e-07,
"loss": 0.0271,
"step": 18390
},
{
"epoch": 70.76923076923077,
"grad_norm": 0.40488916635513306,
"learning_rate": 5.22515067535625e-07,
"loss": 0.0328,
"step": 18400
},
{
"epoch": 70.8076923076923,
"grad_norm": 0.5412066578865051,
"learning_rate": 5.160456649048656e-07,
"loss": 0.0338,
"step": 18410
},
{
"epoch": 70.84615384615384,
"grad_norm": 0.33006104826927185,
"learning_rate": 5.096158608521878e-07,
"loss": 0.0286,
"step": 18420
},
{
"epoch": 70.88461538461539,
"grad_norm": 0.5399073362350464,
"learning_rate": 5.032256729564349e-07,
"loss": 0.0305,
"step": 18430
},
{
"epoch": 70.92307692307692,
"grad_norm": 0.5007637739181519,
"learning_rate": 4.968751186881321e-07,
"loss": 0.0296,
"step": 18440
},
{
"epoch": 70.96153846153847,
"grad_norm": 0.4356684982776642,
"learning_rate": 4.905642154094526e-07,
"loss": 0.0281,
"step": 18450
},
{
"epoch": 71.0,
"grad_norm": 0.5559306740760803,
"learning_rate": 4.842929803741713e-07,
"loss": 0.0332,
"step": 18460
},
{
"epoch": 71.03846153846153,
"grad_norm": 0.5105197429656982,
"learning_rate": 4.780614307275987e-07,
"loss": 0.0281,
"step": 18470
},
{
"epoch": 71.07692307692308,
"grad_norm": 0.6882253885269165,
"learning_rate": 4.7186958350655506e-07,
"loss": 0.0257,
"step": 18480
},
{
"epoch": 71.11538461538461,
"grad_norm": 0.483347088098526,
"learning_rate": 4.65717455639314e-07,
"loss": 0.0297,
"step": 18490
},
{
"epoch": 71.15384615384616,
"grad_norm": 0.4287182092666626,
"learning_rate": 4.5960506394555956e-07,
"loss": 0.0301,
"step": 18500
},
{
"epoch": 71.1923076923077,
"grad_norm": 0.6860320568084717,
"learning_rate": 4.53532425136341e-07,
"loss": 0.0326,
"step": 18510
},
{
"epoch": 71.23076923076923,
"grad_norm": 0.379873126745224,
"learning_rate": 4.4749955581401103e-07,
"loss": 0.0276,
"step": 18520
},
{
"epoch": 71.26923076923077,
"grad_norm": 0.47224199771881104,
"learning_rate": 4.415064724722129e-07,
"loss": 0.0298,
"step": 18530
},
{
"epoch": 71.3076923076923,
"grad_norm": 0.5363563299179077,
"learning_rate": 4.355531914958016e-07,
"loss": 0.037,
"step": 18540
},
{
"epoch": 71.34615384615384,
"grad_norm": 0.4464838206768036,
"learning_rate": 4.2963972916082286e-07,
"loss": 0.0271,
"step": 18550
},
{
"epoch": 71.38461538461539,
"grad_norm": 0.39452752470970154,
"learning_rate": 4.2376610163446074e-07,
"loss": 0.0323,
"step": 18560
},
{
"epoch": 71.42307692307692,
"grad_norm": 0.46284347772598267,
"learning_rate": 4.1793232497498167e-07,
"loss": 0.03,
"step": 18570
},
{
"epoch": 71.46153846153847,
"grad_norm": 0.501216471195221,
"learning_rate": 4.1213841513171257e-07,
"loss": 0.0277,
"step": 18580
},
{
"epoch": 71.5,
"grad_norm": 0.6250320672988892,
"learning_rate": 4.0638438794497743e-07,
"loss": 0.0314,
"step": 18590
},
{
"epoch": 71.53846153846153,
"grad_norm": 0.5748885273933411,
"learning_rate": 4.0067025914607257e-07,
"loss": 0.0306,
"step": 18600
},
{
"epoch": 71.57692307692308,
"grad_norm": 0.7396159172058105,
"learning_rate": 3.9499604435720483e-07,
"loss": 0.0305,
"step": 18610
},
{
"epoch": 71.61538461538461,
"grad_norm": 0.47810420393943787,
"learning_rate": 3.8936175909146e-07,
"loss": 0.0255,
"step": 18620
},
{
"epoch": 71.65384615384616,
"grad_norm": 0.41282424330711365,
"learning_rate": 3.837674187527629e-07,
"loss": 0.0291,
"step": 18630
},
{
"epoch": 71.6923076923077,
"grad_norm": 0.3942478895187378,
"learning_rate": 3.7821303863581904e-07,
"loss": 0.0253,
"step": 18640
},
{
"epoch": 71.73076923076923,
"grad_norm": 0.5066636800765991,
"learning_rate": 3.726986339260996e-07,
"loss": 0.028,
"step": 18650
},
{
"epoch": 71.76923076923077,
"grad_norm": 0.6203503012657166,
"learning_rate": 3.672242196997733e-07,
"loss": 0.0283,
"step": 18660
},
{
"epoch": 71.8076923076923,
"grad_norm": 0.5737688541412354,
"learning_rate": 3.6178981092367615e-07,
"loss": 0.0384,
"step": 18670
},
{
"epoch": 71.84615384615384,
"grad_norm": 0.43425294756889343,
"learning_rate": 3.5639542245527847e-07,
"loss": 0.0291,
"step": 18680
},
{
"epoch": 71.88461538461539,
"grad_norm": 0.5544036030769348,
"learning_rate": 3.5104106904263134e-07,
"loss": 0.0303,
"step": 18690
},
{
"epoch": 71.92307692307692,
"grad_norm": 0.3179159164428711,
"learning_rate": 3.4572676532433345e-07,
"loss": 0.0311,
"step": 18700
},
{
"epoch": 71.96153846153847,
"grad_norm": 0.36036983132362366,
"learning_rate": 3.4045252582948603e-07,
"loss": 0.0306,
"step": 18710
},
{
"epoch": 72.0,
"grad_norm": 0.4262266159057617,
"learning_rate": 3.3521836497765803e-07,
"loss": 0.027,
"step": 18720
},
{
"epoch": 72.03846153846153,
"grad_norm": 0.4187345802783966,
"learning_rate": 3.300242970788492e-07,
"loss": 0.0269,
"step": 18730
},
{
"epoch": 72.07692307692308,
"grad_norm": 0.5532457232475281,
"learning_rate": 3.248703363334404e-07,
"loss": 0.0282,
"step": 18740
},
{
"epoch": 72.11538461538461,
"grad_norm": 0.4422209560871124,
"learning_rate": 3.197564968321637e-07,
"loss": 0.0257,
"step": 18750
},
{
"epoch": 72.15384615384616,
"grad_norm": 0.38217607140541077,
"learning_rate": 3.1468279255606027e-07,
"loss": 0.0267,
"step": 18760
},
{
"epoch": 72.1923076923077,
"grad_norm": 0.48791781067848206,
"learning_rate": 3.096492373764442e-07,
"loss": 0.0284,
"step": 18770
},
{
"epoch": 72.23076923076923,
"grad_norm": 0.48523175716400146,
"learning_rate": 3.046558450548642e-07,
"loss": 0.0287,
"step": 18780
},
{
"epoch": 72.26923076923077,
"grad_norm": 0.5189900994300842,
"learning_rate": 2.997026292430632e-07,
"loss": 0.0294,
"step": 18790
},
{
"epoch": 72.3076923076923,
"grad_norm": 0.5620434880256653,
"learning_rate": 2.9478960348294393e-07,
"loss": 0.0297,
"step": 18800
},
{
"epoch": 72.34615384615384,
"grad_norm": 0.3768644332885742,
"learning_rate": 2.8991678120653343e-07,
"loss": 0.0305,
"step": 18810
},
{
"epoch": 72.38461538461539,
"grad_norm": 0.4504743218421936,
"learning_rate": 2.850841757359385e-07,
"loss": 0.0287,
"step": 18820
},
{
"epoch": 72.42307692307692,
"grad_norm": 0.5062809586524963,
"learning_rate": 2.802918002833188e-07,
"loss": 0.0273,
"step": 18830
},
{
"epoch": 72.46153846153847,
"grad_norm": 0.4553619623184204,
"learning_rate": 2.7553966795084875e-07,
"loss": 0.0263,
"step": 18840
},
{
"epoch": 72.5,
"grad_norm": 0.5206001996994019,
"learning_rate": 2.708277917306723e-07,
"loss": 0.0289,
"step": 18850
},
{
"epoch": 72.53846153846153,
"grad_norm": 0.4859105050563812,
"learning_rate": 2.661561845048832e-07,
"loss": 0.0255,
"step": 18860
},
{
"epoch": 72.57692307692308,
"grad_norm": 0.36909204721450806,
"learning_rate": 2.615248590454733e-07,
"loss": 0.0282,
"step": 18870
},
{
"epoch": 72.61538461538461,
"grad_norm": 0.44843748211860657,
"learning_rate": 2.569338280143124e-07,
"loss": 0.0298,
"step": 18880
},
{
"epoch": 72.65384615384616,
"grad_norm": 0.4249100089073181,
"learning_rate": 2.523831039631036e-07,
"loss": 0.0258,
"step": 18890
},
{
"epoch": 72.6923076923077,
"grad_norm": 0.38302505016326904,
"learning_rate": 2.478726993333513e-07,
"loss": 0.0297,
"step": 18900
},
{
"epoch": 72.73076923076923,
"grad_norm": 0.4525497555732727,
"learning_rate": 2.434026264563299e-07,
"loss": 0.026,
"step": 18910
},
{
"epoch": 72.76923076923077,
"grad_norm": 0.47983717918395996,
"learning_rate": 2.389728975530486e-07,
"loss": 0.0334,
"step": 18920
},
{
"epoch": 72.8076923076923,
"grad_norm": 0.35710352659225464,
"learning_rate": 2.3458352473421986e-07,
"loss": 0.0292,
"step": 18930
},
{
"epoch": 72.84615384615384,
"grad_norm": 0.3811783790588379,
"learning_rate": 2.3023452000021594e-07,
"loss": 0.0289,
"step": 18940
},
{
"epoch": 72.88461538461539,
"grad_norm": 0.38314539194107056,
"learning_rate": 2.2592589524105255e-07,
"loss": 0.0258,
"step": 18950
},
{
"epoch": 72.92307692307692,
"grad_norm": 0.5492963194847107,
"learning_rate": 2.216576622363453e-07,
"loss": 0.0312,
"step": 18960
},
{
"epoch": 72.96153846153847,
"grad_norm": 0.4219627380371094,
"learning_rate": 2.1742983265527984e-07,
"loss": 0.0271,
"step": 18970
},
{
"epoch": 73.0,
"grad_norm": 0.2676997780799866,
"learning_rate": 2.1324241805658006e-07,
"loss": 0.0273,
"step": 18980
},
{
"epoch": 73.03846153846153,
"grad_norm": 0.5187320113182068,
"learning_rate": 2.0909542988848007e-07,
"loss": 0.0235,
"step": 18990
},
{
"epoch": 73.07692307692308,
"grad_norm": 0.39366570115089417,
"learning_rate": 2.0498887948868395e-07,
"loss": 0.0256,
"step": 19000
},
{
"epoch": 73.11538461538461,
"grad_norm": 0.5478989481925964,
"learning_rate": 2.009227780843459e-07,
"loss": 0.0271,
"step": 19010
},
{
"epoch": 73.15384615384616,
"grad_norm": 0.4115489423274994,
"learning_rate": 1.968971367920319e-07,
"loss": 0.0314,
"step": 19020
},
{
"epoch": 73.1923076923077,
"grad_norm": 0.4727810025215149,
"learning_rate": 1.9291196661768984e-07,
"loss": 0.0332,
"step": 19030
},
{
"epoch": 73.23076923076923,
"grad_norm": 0.5472187995910645,
"learning_rate": 1.8896727845662432e-07,
"loss": 0.0272,
"step": 19040
},
{
"epoch": 73.26923076923077,
"grad_norm": 0.2943132817745209,
"learning_rate": 1.8506308309346022e-07,
"loss": 0.0247,
"step": 19050
},
{
"epoch": 73.3076923076923,
"grad_norm": 0.47855880856513977,
"learning_rate": 1.811993912021226e-07,
"loss": 0.0258,
"step": 19060
},
{
"epoch": 73.34615384615384,
"grad_norm": 0.44629132747650146,
"learning_rate": 1.7737621334579346e-07,
"loss": 0.0279,
"step": 19070
},
{
"epoch": 73.38461538461539,
"grad_norm": 0.4611893892288208,
"learning_rate": 1.735935599768951e-07,
"loss": 0.0281,
"step": 19080
},
{
"epoch": 73.42307692307692,
"grad_norm": 0.4612509310245514,
"learning_rate": 1.6985144143706166e-07,
"loss": 0.0273,
"step": 19090
},
{
"epoch": 73.46153846153847,
"grad_norm": 0.651528537273407,
"learning_rate": 1.6614986795709774e-07,
"loss": 0.0288,
"step": 19100
},
{
"epoch": 73.5,
"grad_norm": 0.3900488317012787,
"learning_rate": 1.6248884965696654e-07,
"loss": 0.0268,
"step": 19110
},
{
"epoch": 73.53846153846153,
"grad_norm": 0.393657386302948,
"learning_rate": 1.5886839654575158e-07,
"loss": 0.0266,
"step": 19120
},
{
"epoch": 73.57692307692308,
"grad_norm": 0.5446521639823914,
"learning_rate": 1.5528851852163183e-07,
"loss": 0.0312,
"step": 19130
},
{
"epoch": 73.61538461538461,
"grad_norm": 0.37836071848869324,
"learning_rate": 1.5174922537185997e-07,
"loss": 0.0278,
"step": 19140
},
{
"epoch": 73.65384615384616,
"grad_norm": 0.40086400508880615,
"learning_rate": 1.4825052677272576e-07,
"loss": 0.0302,
"step": 19150
},
{
"epoch": 73.6923076923077,
"grad_norm": 0.3947674334049225,
"learning_rate": 1.4479243228953942e-07,
"loss": 0.0308,
"step": 19160
},
{
"epoch": 73.73076923076923,
"grad_norm": 0.5151625275611877,
"learning_rate": 1.4137495137659827e-07,
"loss": 0.0316,
"step": 19170
},
{
"epoch": 73.76923076923077,
"grad_norm": 0.539281964302063,
"learning_rate": 1.3799809337716517e-07,
"loss": 0.0268,
"step": 19180
},
{
"epoch": 73.8076923076923,
"grad_norm": 0.48684024810791016,
"learning_rate": 1.3466186752344178e-07,
"loss": 0.0298,
"step": 19190
},
{
"epoch": 73.84615384615384,
"grad_norm": 0.4839021861553192,
"learning_rate": 1.3136628293653863e-07,
"loss": 0.0277,
"step": 19200
},
{
"epoch": 73.88461538461539,
"grad_norm": 0.39492732286453247,
"learning_rate": 1.2811134862646178e-07,
"loss": 0.0257,
"step": 19210
},
{
"epoch": 73.92307692307692,
"grad_norm": 0.41039136052131653,
"learning_rate": 1.2489707349207623e-07,
"loss": 0.0277,
"step": 19220
},
{
"epoch": 73.96153846153847,
"grad_norm": 0.6431099772453308,
"learning_rate": 1.2172346632108754e-07,
"loss": 0.0309,
"step": 19230
},
{
"epoch": 74.0,
"grad_norm": 0.41975292563438416,
"learning_rate": 1.1859053579001688e-07,
"loss": 0.0274,
"step": 19240
},
{
"epoch": 74.03846153846153,
"grad_norm": 0.5424035787582397,
"learning_rate": 1.1549829046417437e-07,
"loss": 0.0304,
"step": 19250
},
{
"epoch": 74.07692307692308,
"grad_norm": 0.25909972190856934,
"learning_rate": 1.1244673879764411e-07,
"loss": 0.0275,
"step": 19260
},
{
"epoch": 74.11538461538461,
"grad_norm": 0.3769975006580353,
"learning_rate": 1.094358891332492e-07,
"loss": 0.0274,
"step": 19270
},
{
"epoch": 74.15384615384616,
"grad_norm": 0.46477359533309937,
"learning_rate": 1.0646574970253842e-07,
"loss": 0.0329,
"step": 19280
},
{
"epoch": 74.1923076923077,
"grad_norm": 0.5041369199752808,
"learning_rate": 1.0353632862576124e-07,
"loss": 0.031,
"step": 19290
},
{
"epoch": 74.23076923076923,
"grad_norm": 0.34948527812957764,
"learning_rate": 1.0064763391184118e-07,
"loss": 0.0266,
"step": 19300
},
{
"epoch": 74.26923076923077,
"grad_norm": 0.49228963255882263,
"learning_rate": 9.779967345835917e-08,
"loss": 0.0269,
"step": 19310
},
{
"epoch": 74.3076923076923,
"grad_norm": 0.30661070346832275,
"learning_rate": 9.49924550515302e-08,
"loss": 0.0266,
"step": 19320
},
{
"epoch": 74.34615384615384,
"grad_norm": 0.546813428401947,
"learning_rate": 9.22259863661834e-08,
"loss": 0.0282,
"step": 19330
},
{
"epoch": 74.38461538461539,
"grad_norm": 0.34433984756469727,
"learning_rate": 8.950027496573865e-08,
"loss": 0.0285,
"step": 19340
},
{
"epoch": 74.42307692307692,
"grad_norm": 0.6200993061065674,
"learning_rate": 8.681532830218497e-08,
"loss": 0.0302,
"step": 19350
},
{
"epoch": 74.46153846153847,
"grad_norm": 0.283530592918396,
"learning_rate": 8.417115371606554e-08,
"loss": 0.0258,
"step": 19360
},
{
"epoch": 74.5,
"grad_norm": 0.4857964515686035,
"learning_rate": 8.15677584364527e-08,
"loss": 0.0275,
"step": 19370
},
{
"epoch": 74.53846153846153,
"grad_norm": 0.4263962209224701,
"learning_rate": 7.900514958092964e-08,
"loss": 0.0295,
"step": 19380
},
{
"epoch": 74.57692307692308,
"grad_norm": 0.4442691206932068,
"learning_rate": 7.64833341555704e-08,
"loss": 0.0314,
"step": 19390
},
{
"epoch": 74.61538461538461,
"grad_norm": 0.4575265944004059,
"learning_rate": 7.400231905492328e-08,
"loss": 0.0275,
"step": 19400
},
{
"epoch": 74.65384615384616,
"grad_norm": 0.5982808470726013,
"learning_rate": 7.15621110619874e-08,
"loss": 0.0318,
"step": 19410
},
{
"epoch": 74.6923076923077,
"grad_norm": 0.47400107979774475,
"learning_rate": 6.916271684819787e-08,
"loss": 0.0268,
"step": 19420
},
{
"epoch": 74.73076923076923,
"grad_norm": 0.3319554030895233,
"learning_rate": 6.680414297340897e-08,
"loss": 0.0267,
"step": 19430
},
{
"epoch": 74.76923076923077,
"grad_norm": 0.4506201148033142,
"learning_rate": 6.448639588587103e-08,
"loss": 0.0296,
"step": 19440
},
{
"epoch": 74.8076923076923,
"grad_norm": 0.4350897967815399,
"learning_rate": 6.22094819222152e-08,
"loss": 0.0279,
"step": 19450
},
{
"epoch": 74.84615384615384,
"grad_norm": 0.46058401465415955,
"learning_rate": 5.997340730743705e-08,
"loss": 0.0247,
"step": 19460
},
{
"epoch": 74.88461538461539,
"grad_norm": 0.29316025972366333,
"learning_rate": 5.7778178154879736e-08,
"loss": 0.0309,
"step": 19470
},
{
"epoch": 74.92307692307692,
"grad_norm": 0.4575360119342804,
"learning_rate": 5.56238004662174e-08,
"loss": 0.0346,
"step": 19480
},
{
"epoch": 74.96153846153847,
"grad_norm": 0.4061368703842163,
"learning_rate": 5.351028013143355e-08,
"loss": 0.0345,
"step": 19490
},
{
"epoch": 75.0,
"grad_norm": 0.38843148946762085,
"learning_rate": 5.1437622928814374e-08,
"loss": 0.027,
"step": 19500
},
{
"epoch": 75.03846153846153,
"grad_norm": 0.46850109100341797,
"learning_rate": 4.940583452492542e-08,
"loss": 0.0289,
"step": 19510
},
{
"epoch": 75.07692307692308,
"grad_norm": 0.2569507658481598,
"learning_rate": 4.741492047459661e-08,
"loss": 0.0273,
"step": 19520
},
{
"epoch": 75.11538461538461,
"grad_norm": 0.546170175075531,
"learning_rate": 4.5464886220912275e-08,
"loss": 0.0295,
"step": 19530
},
{
"epoch": 75.15384615384616,
"grad_norm": 0.35281962156295776,
"learning_rate": 4.355573709519112e-08,
"loss": 0.0277,
"step": 19540
},
{
"epoch": 75.1923076923077,
"grad_norm": 0.5647571086883545,
"learning_rate": 4.168747831697628e-08,
"loss": 0.0327,
"step": 19550
},
{
"epoch": 75.23076923076923,
"grad_norm": 0.48984843492507935,
"learning_rate": 3.986011499401199e-08,
"loss": 0.0296,
"step": 19560
},
{
"epoch": 75.26923076923077,
"grad_norm": 0.4074559807777405,
"learning_rate": 3.80736521222419e-08,
"loss": 0.0323,
"step": 19570
},
{
"epoch": 75.3076923076923,
"grad_norm": 0.35387930274009705,
"learning_rate": 3.6328094585789116e-08,
"loss": 0.0303,
"step": 19580
},
{
"epoch": 75.34615384615384,
"grad_norm": 0.37408000230789185,
"learning_rate": 3.462344715693788e-08,
"loss": 0.0254,
"step": 19590
},
{
"epoch": 75.38461538461539,
"grad_norm": 0.3567971885204315,
"learning_rate": 3.295971449613022e-08,
"loss": 0.0259,
"step": 19600
},
{
"epoch": 75.42307692307692,
"grad_norm": 0.2978547215461731,
"learning_rate": 3.1336901151949316e-08,
"loss": 0.0242,
"step": 19610
},
{
"epoch": 75.46153846153847,
"grad_norm": 0.4735611379146576,
"learning_rate": 2.975501156110283e-08,
"loss": 0.0284,
"step": 19620
},
{
"epoch": 75.5,
"grad_norm": 0.3551243245601654,
"learning_rate": 2.821405004841793e-08,
"loss": 0.0271,
"step": 19630
},
{
"epoch": 75.53846153846153,
"grad_norm": 0.478957861661911,
"learning_rate": 2.671402082682295e-08,
"loss": 0.0263,
"step": 19640
},
{
"epoch": 75.57692307692308,
"grad_norm": 0.4333643913269043,
"learning_rate": 2.5254927997342415e-08,
"loss": 0.0255,
"step": 19650
},
{
"epoch": 75.61538461538461,
"grad_norm": 0.456312358379364,
"learning_rate": 2.3836775549078703e-08,
"loss": 0.0317,
"step": 19660
},
{
"epoch": 75.65384615384616,
"grad_norm": 0.5187574625015259,
"learning_rate": 2.245956735920873e-08,
"loss": 0.0313,
"step": 19670
},
{
"epoch": 75.6923076923077,
"grad_norm": 0.3721175491809845,
"learning_rate": 2.1123307192965625e-08,
"loss": 0.0258,
"step": 19680
},
{
"epoch": 75.73076923076923,
"grad_norm": 0.42539578676223755,
"learning_rate": 1.9827998703632056e-08,
"loss": 0.0263,
"step": 19690
},
{
"epoch": 75.76923076923077,
"grad_norm": 0.594091534614563,
"learning_rate": 1.8573645432535258e-08,
"loss": 0.0347,
"step": 19700
},
{
"epoch": 75.8076923076923,
"grad_norm": 0.30675655603408813,
"learning_rate": 1.7360250809027034e-08,
"loss": 0.0275,
"step": 19710
},
{
"epoch": 75.84615384615384,
"grad_norm": 0.37368330359458923,
"learning_rate": 1.618781815048209e-08,
"loss": 0.0287,
"step": 19720
},
{
"epoch": 75.88461538461539,
"grad_norm": 0.3517482280731201,
"learning_rate": 1.5056350662286388e-08,
"loss": 0.03,
"step": 19730
},
{
"epoch": 75.92307692307692,
"grad_norm": 0.5828191041946411,
"learning_rate": 1.3965851437830468e-08,
"loss": 0.0295,
"step": 19740
},
{
"epoch": 75.96153846153847,
"grad_norm": 0.359904408454895,
"learning_rate": 1.2916323458494473e-08,
"loss": 0.0305,
"step": 19750
},
{
"epoch": 76.0,
"grad_norm": 0.5236489772796631,
"learning_rate": 1.1907769593651474e-08,
"loss": 0.0294,
"step": 19760
},
{
"epoch": 76.03846153846153,
"grad_norm": 0.4129619300365448,
"learning_rate": 1.0940192600647491e-08,
"loss": 0.03,
"step": 19770
},
{
"epoch": 76.07692307692308,
"grad_norm": 0.4157169461250305,
"learning_rate": 1.0013595124801488e-08,
"loss": 0.0294,
"step": 19780
},
{
"epoch": 76.11538461538461,
"grad_norm": 0.5382236242294312,
"learning_rate": 9.127979699393719e-09,
"loss": 0.0307,
"step": 19790
},
{
"epoch": 76.15384615384616,
"grad_norm": 0.36726507544517517,
"learning_rate": 8.283348745665719e-09,
"loss": 0.0273,
"step": 19800
},
{
"epoch": 76.1923076923077,
"grad_norm": 0.35563984513282776,
"learning_rate": 7.479704572805336e-09,
"loss": 0.0299,
"step": 19810
},
{
"epoch": 76.23076923076923,
"grad_norm": 0.5569225549697876,
"learning_rate": 6.717049377943374e-09,
"loss": 0.0284,
"step": 19820
},
{
"epoch": 76.26923076923077,
"grad_norm": 0.37364256381988525,
"learning_rate": 5.995385246151952e-09,
"loss": 0.0308,
"step": 19830
},
{
"epoch": 76.3076923076923,
"grad_norm": 0.3298107385635376,
"learning_rate": 5.314714150432831e-09,
"loss": 0.0343,
"step": 19840
},
{
"epoch": 76.34615384615384,
"grad_norm": 0.6490257978439331,
"learning_rate": 4.67503795171409e-09,
"loss": 0.0333,
"step": 19850
},
{
"epoch": 76.38461538461539,
"grad_norm": 0.6045204401016235,
"learning_rate": 4.076358398846791e-09,
"loss": 0.0331,
"step": 19860
},
{
"epoch": 76.42307692307692,
"grad_norm": 0.40658727288246155,
"learning_rate": 3.518677128598324e-09,
"loss": 0.0314,
"step": 19870
},
{
"epoch": 76.46153846153847,
"grad_norm": 0.40208566188812256,
"learning_rate": 3.0019956656457404e-09,
"loss": 0.027,
"step": 19880
},
{
"epoch": 76.5,
"grad_norm": 0.3167416453361511,
"learning_rate": 2.526315422579084e-09,
"loss": 0.0299,
"step": 19890
},
{
"epoch": 76.53846153846153,
"grad_norm": 0.5065469145774841,
"learning_rate": 2.091637699889737e-09,
"loss": 0.0288,
"step": 19900
},
{
"epoch": 76.57692307692308,
"grad_norm": 0.3822614550590515,
"learning_rate": 1.6979636859687509e-09,
"loss": 0.027,
"step": 19910
},
{
"epoch": 76.61538461538461,
"grad_norm": 0.48988813161849976,
"learning_rate": 1.3452944571051839e-09,
"loss": 0.0262,
"step": 19920
},
{
"epoch": 76.65384615384616,
"grad_norm": 0.34604522585868835,
"learning_rate": 1.0336309774860998e-09,
"loss": 0.0247,
"step": 19930
},
{
"epoch": 76.6923076923077,
"grad_norm": 0.5055863261222839,
"learning_rate": 7.629740991849099e-10,
"loss": 0.0274,
"step": 19940
},
{
"epoch": 76.73076923076923,
"grad_norm": 0.45108962059020996,
"learning_rate": 5.333245621680361e-10,
"loss": 0.0313,
"step": 19950
},
{
"epoch": 76.76923076923077,
"grad_norm": 0.4111419916152954,
"learning_rate": 3.446829942882479e-10,
"loss": 0.0278,
"step": 19960
},
{
"epoch": 76.8076923076923,
"grad_norm": 0.5759734511375427,
"learning_rate": 1.9704991128632887e-10,
"loss": 0.0312,
"step": 19970
},
{
"epoch": 76.84615384615384,
"grad_norm": 0.4901687204837799,
"learning_rate": 9.042571678274936e-11,
"loss": 0.026,
"step": 19980
},
{
"epoch": 76.88461538461539,
"grad_norm": 0.47623711824417114,
"learning_rate": 2.4810702284328024e-11,
"loss": 0.0309,
"step": 19990
},
{
"epoch": 76.92307692307692,
"grad_norm": 0.5158125162124634,
"learning_rate": 2.050471825665312e-13,
"loss": 0.0254,
"step": 20000
},
{
"epoch": 76.92307692307692,
"step": 20000,
"total_flos": 0.0,
"train_loss": 0.09504265115857125,
"train_runtime": 20232.6485,
"train_samples_per_second": 63.264,
"train_steps_per_second": 0.989
}
],
"logging_steps": 10,
"max_steps": 20000,
"num_input_tokens_seen": 0,
"num_train_epochs": 77,
"save_steps": 10000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 32,
"trial_name": null,
"trial_params": null
}