ElenaHsieh's picture
End of training
eb67e68 verified
raw
history blame
35.3 kB
{
"best_metric": 0.5917085427135679,
"best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-eurosat/checkpoint-224",
"epoch": 50.0,
"eval_steps": 500,
"global_step": 1400,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.36,
"grad_norm": 2.619962215423584,
"learning_rate": 3.5714285714285714e-06,
"loss": 1.0415,
"step": 10
},
{
"epoch": 0.71,
"grad_norm": 2.402837038040161,
"learning_rate": 7.142857142857143e-06,
"loss": 1.0339,
"step": 20
},
{
"epoch": 1.0,
"eval_accuracy": 0.5640703517587939,
"eval_loss": 1.0541082620620728,
"eval_runtime": 5.8961,
"eval_samples_per_second": 270.009,
"eval_steps_per_second": 2.205,
"step": 28
},
{
"epoch": 1.07,
"grad_norm": 2.223407745361328,
"learning_rate": 1.0714285714285714e-05,
"loss": 1.025,
"step": 30
},
{
"epoch": 1.43,
"grad_norm": 2.5425541400909424,
"learning_rate": 1.4285714285714285e-05,
"loss": 1.0072,
"step": 40
},
{
"epoch": 1.79,
"grad_norm": 3.219869375228882,
"learning_rate": 1.785714285714286e-05,
"loss": 1.0193,
"step": 50
},
{
"epoch": 2.0,
"eval_accuracy": 0.5621859296482412,
"eval_loss": 1.0463842153549194,
"eval_runtime": 5.8033,
"eval_samples_per_second": 274.326,
"eval_steps_per_second": 2.24,
"step": 56
},
{
"epoch": 2.14,
"grad_norm": 3.941281318664551,
"learning_rate": 2.1428571428571428e-05,
"loss": 1.0133,
"step": 60
},
{
"epoch": 2.5,
"grad_norm": 4.169382095336914,
"learning_rate": 2.5e-05,
"loss": 1.0182,
"step": 70
},
{
"epoch": 2.86,
"grad_norm": 3.475947380065918,
"learning_rate": 2.857142857142857e-05,
"loss": 1.0348,
"step": 80
},
{
"epoch": 3.0,
"eval_accuracy": 0.5690954773869347,
"eval_loss": 1.0330955982208252,
"eval_runtime": 5.8594,
"eval_samples_per_second": 271.702,
"eval_steps_per_second": 2.219,
"step": 84
},
{
"epoch": 3.21,
"grad_norm": 3.1382639408111572,
"learning_rate": 3.2142857142857144e-05,
"loss": 1.0266,
"step": 90
},
{
"epoch": 3.57,
"grad_norm": 2.344207525253296,
"learning_rate": 3.571428571428572e-05,
"loss": 1.0075,
"step": 100
},
{
"epoch": 3.93,
"grad_norm": 2.2122068405151367,
"learning_rate": 3.928571428571429e-05,
"loss": 1.0072,
"step": 110
},
{
"epoch": 4.0,
"eval_accuracy": 0.5847989949748744,
"eval_loss": 1.0253527164459229,
"eval_runtime": 5.8726,
"eval_samples_per_second": 271.091,
"eval_steps_per_second": 2.214,
"step": 112
},
{
"epoch": 4.29,
"grad_norm": 2.9957728385925293,
"learning_rate": 4.2857142857142856e-05,
"loss": 1.0024,
"step": 120
},
{
"epoch": 4.64,
"grad_norm": 2.9330532550811768,
"learning_rate": 4.642857142857143e-05,
"loss": 0.991,
"step": 130
},
{
"epoch": 5.0,
"grad_norm": 2.355830430984497,
"learning_rate": 5e-05,
"loss": 0.9892,
"step": 140
},
{
"epoch": 5.0,
"eval_accuracy": 0.5753768844221105,
"eval_loss": 1.012104868888855,
"eval_runtime": 5.622,
"eval_samples_per_second": 283.175,
"eval_steps_per_second": 2.312,
"step": 140
},
{
"epoch": 5.36,
"grad_norm": 3.8320775032043457,
"learning_rate": 4.960317460317461e-05,
"loss": 0.9517,
"step": 150
},
{
"epoch": 5.71,
"grad_norm": 3.0432510375976562,
"learning_rate": 4.9206349206349204e-05,
"loss": 0.9379,
"step": 160
},
{
"epoch": 6.0,
"eval_accuracy": 0.5810301507537688,
"eval_loss": 1.017525553703308,
"eval_runtime": 6.0814,
"eval_samples_per_second": 261.781,
"eval_steps_per_second": 2.138,
"step": 168
},
{
"epoch": 6.07,
"grad_norm": 3.3289706707000732,
"learning_rate": 4.880952380952381e-05,
"loss": 0.9475,
"step": 170
},
{
"epoch": 6.43,
"grad_norm": 3.335624933242798,
"learning_rate": 4.841269841269841e-05,
"loss": 0.9103,
"step": 180
},
{
"epoch": 6.79,
"grad_norm": 3.2209532260894775,
"learning_rate": 4.801587301587302e-05,
"loss": 0.9123,
"step": 190
},
{
"epoch": 7.0,
"eval_accuracy": 0.5866834170854272,
"eval_loss": 1.0119962692260742,
"eval_runtime": 5.8208,
"eval_samples_per_second": 273.502,
"eval_steps_per_second": 2.233,
"step": 196
},
{
"epoch": 7.14,
"grad_norm": 3.194115400314331,
"learning_rate": 4.761904761904762e-05,
"loss": 0.896,
"step": 200
},
{
"epoch": 7.5,
"grad_norm": 2.89758563041687,
"learning_rate": 4.722222222222222e-05,
"loss": 0.9013,
"step": 210
},
{
"epoch": 7.86,
"grad_norm": 3.0895237922668457,
"learning_rate": 4.682539682539683e-05,
"loss": 0.8865,
"step": 220
},
{
"epoch": 8.0,
"eval_accuracy": 0.5917085427135679,
"eval_loss": 1.0104140043258667,
"eval_runtime": 6.0762,
"eval_samples_per_second": 262.008,
"eval_steps_per_second": 2.14,
"step": 224
},
{
"epoch": 8.21,
"grad_norm": 2.715590715408325,
"learning_rate": 4.642857142857143e-05,
"loss": 0.8758,
"step": 230
},
{
"epoch": 8.57,
"grad_norm": 3.639813184738159,
"learning_rate": 4.603174603174603e-05,
"loss": 0.8516,
"step": 240
},
{
"epoch": 8.93,
"grad_norm": 3.1830241680145264,
"learning_rate": 4.563492063492064e-05,
"loss": 0.8668,
"step": 250
},
{
"epoch": 9.0,
"eval_accuracy": 0.5873115577889447,
"eval_loss": 1.0236343145370483,
"eval_runtime": 6.0393,
"eval_samples_per_second": 263.608,
"eval_steps_per_second": 2.153,
"step": 252
},
{
"epoch": 9.29,
"grad_norm": 3.1714346408843994,
"learning_rate": 4.523809523809524e-05,
"loss": 0.8352,
"step": 260
},
{
"epoch": 9.64,
"grad_norm": 3.9194183349609375,
"learning_rate": 4.4841269841269846e-05,
"loss": 0.8207,
"step": 270
},
{
"epoch": 10.0,
"grad_norm": 4.183494567871094,
"learning_rate": 4.4444444444444447e-05,
"loss": 0.8189,
"step": 280
},
{
"epoch": 10.0,
"eval_accuracy": 0.5829145728643216,
"eval_loss": 1.0359796285629272,
"eval_runtime": 6.1118,
"eval_samples_per_second": 260.481,
"eval_steps_per_second": 2.127,
"step": 280
},
{
"epoch": 10.36,
"grad_norm": 3.281873941421509,
"learning_rate": 4.404761904761905e-05,
"loss": 0.8051,
"step": 290
},
{
"epoch": 10.71,
"grad_norm": 2.9959723949432373,
"learning_rate": 4.3650793650793655e-05,
"loss": 0.7933,
"step": 300
},
{
"epoch": 11.0,
"eval_accuracy": 0.5835427135678392,
"eval_loss": 1.039525032043457,
"eval_runtime": 5.9823,
"eval_samples_per_second": 266.118,
"eval_steps_per_second": 2.173,
"step": 308
},
{
"epoch": 11.07,
"grad_norm": 3.470358371734619,
"learning_rate": 4.3253968253968256e-05,
"loss": 0.797,
"step": 310
},
{
"epoch": 11.43,
"grad_norm": 3.285288095474243,
"learning_rate": 4.2857142857142856e-05,
"loss": 0.7723,
"step": 320
},
{
"epoch": 11.79,
"grad_norm": 4.59849739074707,
"learning_rate": 4.2460317460317464e-05,
"loss": 0.7765,
"step": 330
},
{
"epoch": 12.0,
"eval_accuracy": 0.5728643216080402,
"eval_loss": 1.0594085454940796,
"eval_runtime": 5.9856,
"eval_samples_per_second": 265.972,
"eval_steps_per_second": 2.172,
"step": 336
},
{
"epoch": 12.14,
"grad_norm": 4.7830305099487305,
"learning_rate": 4.2063492063492065e-05,
"loss": 0.7567,
"step": 340
},
{
"epoch": 12.5,
"grad_norm": 3.7676315307617188,
"learning_rate": 4.166666666666667e-05,
"loss": 0.738,
"step": 350
},
{
"epoch": 12.86,
"grad_norm": 4.092833518981934,
"learning_rate": 4.126984126984127e-05,
"loss": 0.7538,
"step": 360
},
{
"epoch": 13.0,
"eval_accuracy": 0.5879396984924623,
"eval_loss": 1.0552036762237549,
"eval_runtime": 5.8071,
"eval_samples_per_second": 274.146,
"eval_steps_per_second": 2.239,
"step": 364
},
{
"epoch": 13.21,
"grad_norm": 4.582472801208496,
"learning_rate": 4.0873015873015874e-05,
"loss": 0.7175,
"step": 370
},
{
"epoch": 13.57,
"grad_norm": 4.479797840118408,
"learning_rate": 4.047619047619048e-05,
"loss": 0.7101,
"step": 380
},
{
"epoch": 13.93,
"grad_norm": 4.223387241363525,
"learning_rate": 4.007936507936508e-05,
"loss": 0.7146,
"step": 390
},
{
"epoch": 14.0,
"eval_accuracy": 0.5829145728643216,
"eval_loss": 1.0619994401931763,
"eval_runtime": 5.8703,
"eval_samples_per_second": 271.194,
"eval_steps_per_second": 2.215,
"step": 392
},
{
"epoch": 14.29,
"grad_norm": 4.616257190704346,
"learning_rate": 3.968253968253968e-05,
"loss": 0.7031,
"step": 400
},
{
"epoch": 14.64,
"grad_norm": 3.6395843029022217,
"learning_rate": 3.928571428571429e-05,
"loss": 0.6901,
"step": 410
},
{
"epoch": 15.0,
"grad_norm": 3.426710367202759,
"learning_rate": 3.888888888888889e-05,
"loss": 0.6885,
"step": 420
},
{
"epoch": 15.0,
"eval_accuracy": 0.5841708542713567,
"eval_loss": 1.0782768726348877,
"eval_runtime": 6.0171,
"eval_samples_per_second": 264.578,
"eval_steps_per_second": 2.16,
"step": 420
},
{
"epoch": 15.36,
"grad_norm": 3.5050458908081055,
"learning_rate": 3.84920634920635e-05,
"loss": 0.6697,
"step": 430
},
{
"epoch": 15.71,
"grad_norm": 4.314873218536377,
"learning_rate": 3.809523809523809e-05,
"loss": 0.6556,
"step": 440
},
{
"epoch": 16.0,
"eval_accuracy": 0.5816582914572864,
"eval_loss": 1.1010228395462036,
"eval_runtime": 6.002,
"eval_samples_per_second": 265.244,
"eval_steps_per_second": 2.166,
"step": 448
},
{
"epoch": 16.07,
"grad_norm": 3.6204850673675537,
"learning_rate": 3.76984126984127e-05,
"loss": 0.6621,
"step": 450
},
{
"epoch": 16.43,
"grad_norm": 5.87465238571167,
"learning_rate": 3.730158730158731e-05,
"loss": 0.6388,
"step": 460
},
{
"epoch": 16.79,
"grad_norm": 3.773968458175659,
"learning_rate": 3.690476190476191e-05,
"loss": 0.6453,
"step": 470
},
{
"epoch": 17.0,
"eval_accuracy": 0.5734924623115578,
"eval_loss": 1.1130640506744385,
"eval_runtime": 5.9355,
"eval_samples_per_second": 268.215,
"eval_steps_per_second": 2.19,
"step": 476
},
{
"epoch": 17.14,
"grad_norm": 3.7962214946746826,
"learning_rate": 3.650793650793651e-05,
"loss": 0.6504,
"step": 480
},
{
"epoch": 17.5,
"grad_norm": 4.2159013748168945,
"learning_rate": 3.611111111111111e-05,
"loss": 0.6058,
"step": 490
},
{
"epoch": 17.86,
"grad_norm": 4.258510112762451,
"learning_rate": 3.571428571428572e-05,
"loss": 0.6175,
"step": 500
},
{
"epoch": 18.0,
"eval_accuracy": 0.5891959798994975,
"eval_loss": 1.1074261665344238,
"eval_runtime": 5.4208,
"eval_samples_per_second": 293.686,
"eval_steps_per_second": 2.398,
"step": 504
},
{
"epoch": 18.21,
"grad_norm": 4.8816914558410645,
"learning_rate": 3.5317460317460324e-05,
"loss": 0.5942,
"step": 510
},
{
"epoch": 18.57,
"grad_norm": 4.609433650970459,
"learning_rate": 3.492063492063492e-05,
"loss": 0.5943,
"step": 520
},
{
"epoch": 18.93,
"grad_norm": 7.7561211585998535,
"learning_rate": 3.4523809523809526e-05,
"loss": 0.5993,
"step": 530
},
{
"epoch": 19.0,
"eval_accuracy": 0.5741206030150754,
"eval_loss": 1.1327540874481201,
"eval_runtime": 5.7963,
"eval_samples_per_second": 274.658,
"eval_steps_per_second": 2.243,
"step": 532
},
{
"epoch": 19.29,
"grad_norm": 4.787009239196777,
"learning_rate": 3.412698412698413e-05,
"loss": 0.5741,
"step": 540
},
{
"epoch": 19.64,
"grad_norm": 7.622641086578369,
"learning_rate": 3.3730158730158734e-05,
"loss": 0.5803,
"step": 550
},
{
"epoch": 20.0,
"grad_norm": 6.530401706695557,
"learning_rate": 3.3333333333333335e-05,
"loss": 0.5683,
"step": 560
},
{
"epoch": 20.0,
"eval_accuracy": 0.5791457286432161,
"eval_loss": 1.1423227787017822,
"eval_runtime": 5.7458,
"eval_samples_per_second": 277.073,
"eval_steps_per_second": 2.263,
"step": 560
},
{
"epoch": 20.36,
"grad_norm": 4.021312236785889,
"learning_rate": 3.2936507936507936e-05,
"loss": 0.5457,
"step": 570
},
{
"epoch": 20.71,
"grad_norm": 5.022521018981934,
"learning_rate": 3.253968253968254e-05,
"loss": 0.5524,
"step": 580
},
{
"epoch": 21.0,
"eval_accuracy": 0.5873115577889447,
"eval_loss": 1.1516565084457397,
"eval_runtime": 6.4553,
"eval_samples_per_second": 246.619,
"eval_steps_per_second": 2.014,
"step": 588
},
{
"epoch": 21.07,
"grad_norm": 4.585721969604492,
"learning_rate": 3.2142857142857144e-05,
"loss": 0.5493,
"step": 590
},
{
"epoch": 21.43,
"grad_norm": 3.6502797603607178,
"learning_rate": 3.1746031746031745e-05,
"loss": 0.5393,
"step": 600
},
{
"epoch": 21.79,
"grad_norm": 4.494349002838135,
"learning_rate": 3.134920634920635e-05,
"loss": 0.5151,
"step": 610
},
{
"epoch": 22.0,
"eval_accuracy": 0.5766331658291457,
"eval_loss": 1.1673014163970947,
"eval_runtime": 6.0498,
"eval_samples_per_second": 263.15,
"eval_steps_per_second": 2.149,
"step": 616
},
{
"epoch": 22.14,
"grad_norm": 4.292243957519531,
"learning_rate": 3.095238095238095e-05,
"loss": 0.5241,
"step": 620
},
{
"epoch": 22.5,
"grad_norm": 4.669267177581787,
"learning_rate": 3.055555555555556e-05,
"loss": 0.5095,
"step": 630
},
{
"epoch": 22.86,
"grad_norm": 4.2560882568359375,
"learning_rate": 3.0158730158730158e-05,
"loss": 0.5096,
"step": 640
},
{
"epoch": 23.0,
"eval_accuracy": 0.5797738693467337,
"eval_loss": 1.17599618434906,
"eval_runtime": 5.8279,
"eval_samples_per_second": 273.169,
"eval_steps_per_second": 2.231,
"step": 644
},
{
"epoch": 23.21,
"grad_norm": 4.331119537353516,
"learning_rate": 2.9761904761904762e-05,
"loss": 0.5092,
"step": 650
},
{
"epoch": 23.57,
"grad_norm": 5.246380805969238,
"learning_rate": 2.9365079365079366e-05,
"loss": 0.5016,
"step": 660
},
{
"epoch": 23.93,
"grad_norm": 4.598259449005127,
"learning_rate": 2.8968253968253974e-05,
"loss": 0.4937,
"step": 670
},
{
"epoch": 24.0,
"eval_accuracy": 0.5816582914572864,
"eval_loss": 1.193081259727478,
"eval_runtime": 5.7093,
"eval_samples_per_second": 278.844,
"eval_steps_per_second": 2.277,
"step": 672
},
{
"epoch": 24.29,
"grad_norm": 4.1697893142700195,
"learning_rate": 2.857142857142857e-05,
"loss": 0.4713,
"step": 680
},
{
"epoch": 24.64,
"grad_norm": 3.761561870574951,
"learning_rate": 2.8174603174603175e-05,
"loss": 0.469,
"step": 690
},
{
"epoch": 25.0,
"grad_norm": 3.7730281352996826,
"learning_rate": 2.777777777777778e-05,
"loss": 0.487,
"step": 700
},
{
"epoch": 25.0,
"eval_accuracy": 0.5734924623115578,
"eval_loss": 1.2083638906478882,
"eval_runtime": 6.0189,
"eval_samples_per_second": 264.498,
"eval_steps_per_second": 2.16,
"step": 700
},
{
"epoch": 25.36,
"grad_norm": 4.0658674240112305,
"learning_rate": 2.7380952380952383e-05,
"loss": 0.4587,
"step": 710
},
{
"epoch": 25.71,
"grad_norm": 3.6383256912231445,
"learning_rate": 2.6984126984126984e-05,
"loss": 0.4597,
"step": 720
},
{
"epoch": 26.0,
"eval_accuracy": 0.571608040201005,
"eval_loss": 1.2270249128341675,
"eval_runtime": 6.2134,
"eval_samples_per_second": 256.222,
"eval_steps_per_second": 2.092,
"step": 728
},
{
"epoch": 26.07,
"grad_norm": 4.403080940246582,
"learning_rate": 2.6587301587301588e-05,
"loss": 0.4461,
"step": 730
},
{
"epoch": 26.43,
"grad_norm": 4.59806489944458,
"learning_rate": 2.6190476190476192e-05,
"loss": 0.4487,
"step": 740
},
{
"epoch": 26.79,
"grad_norm": 3.9561307430267334,
"learning_rate": 2.5793650793650796e-05,
"loss": 0.4482,
"step": 750
},
{
"epoch": 27.0,
"eval_accuracy": 0.5829145728643216,
"eval_loss": 1.2389401197433472,
"eval_runtime": 5.923,
"eval_samples_per_second": 268.783,
"eval_steps_per_second": 2.195,
"step": 756
},
{
"epoch": 27.14,
"grad_norm": 4.864688396453857,
"learning_rate": 2.5396825396825397e-05,
"loss": 0.4678,
"step": 760
},
{
"epoch": 27.5,
"grad_norm": 11.65251636505127,
"learning_rate": 2.5e-05,
"loss": 0.4347,
"step": 770
},
{
"epoch": 27.86,
"grad_norm": 4.065871715545654,
"learning_rate": 2.4603174603174602e-05,
"loss": 0.4183,
"step": 780
},
{
"epoch": 28.0,
"eval_accuracy": 0.5772613065326633,
"eval_loss": 1.2430446147918701,
"eval_runtime": 5.8698,
"eval_samples_per_second": 271.218,
"eval_steps_per_second": 2.215,
"step": 784
},
{
"epoch": 28.21,
"grad_norm": 3.8666417598724365,
"learning_rate": 2.4206349206349206e-05,
"loss": 0.42,
"step": 790
},
{
"epoch": 28.57,
"grad_norm": 4.105574131011963,
"learning_rate": 2.380952380952381e-05,
"loss": 0.4359,
"step": 800
},
{
"epoch": 28.93,
"grad_norm": 4.677915573120117,
"learning_rate": 2.3412698412698414e-05,
"loss": 0.4228,
"step": 810
},
{
"epoch": 29.0,
"eval_accuracy": 0.5741206030150754,
"eval_loss": 1.2637208700180054,
"eval_runtime": 5.5801,
"eval_samples_per_second": 285.302,
"eval_steps_per_second": 2.33,
"step": 812
},
{
"epoch": 29.29,
"grad_norm": 4.680498123168945,
"learning_rate": 2.3015873015873015e-05,
"loss": 0.4268,
"step": 820
},
{
"epoch": 29.64,
"grad_norm": 3.9155895709991455,
"learning_rate": 2.261904761904762e-05,
"loss": 0.3908,
"step": 830
},
{
"epoch": 30.0,
"grad_norm": 4.745033264160156,
"learning_rate": 2.2222222222222223e-05,
"loss": 0.4116,
"step": 840
},
{
"epoch": 30.0,
"eval_accuracy": 0.5778894472361809,
"eval_loss": 1.2687839269638062,
"eval_runtime": 5.7938,
"eval_samples_per_second": 274.776,
"eval_steps_per_second": 2.244,
"step": 840
},
{
"epoch": 30.36,
"grad_norm": 4.2474236488342285,
"learning_rate": 2.1825396825396827e-05,
"loss": 0.3889,
"step": 850
},
{
"epoch": 30.71,
"grad_norm": 4.243285655975342,
"learning_rate": 2.1428571428571428e-05,
"loss": 0.3942,
"step": 860
},
{
"epoch": 31.0,
"eval_accuracy": 0.5879396984924623,
"eval_loss": 1.2986161708831787,
"eval_runtime": 5.7249,
"eval_samples_per_second": 278.081,
"eval_steps_per_second": 2.271,
"step": 868
},
{
"epoch": 31.07,
"grad_norm": 4.418118000030518,
"learning_rate": 2.1031746031746032e-05,
"loss": 0.3868,
"step": 870
},
{
"epoch": 31.43,
"grad_norm": 4.63828182220459,
"learning_rate": 2.0634920634920636e-05,
"loss": 0.3935,
"step": 880
},
{
"epoch": 31.79,
"grad_norm": 4.932127952575684,
"learning_rate": 2.023809523809524e-05,
"loss": 0.3815,
"step": 890
},
{
"epoch": 32.0,
"eval_accuracy": 0.5766331658291457,
"eval_loss": 1.2911109924316406,
"eval_runtime": 5.5086,
"eval_samples_per_second": 289.004,
"eval_steps_per_second": 2.36,
"step": 896
},
{
"epoch": 32.14,
"grad_norm": 5.120588779449463,
"learning_rate": 1.984126984126984e-05,
"loss": 0.3814,
"step": 900
},
{
"epoch": 32.5,
"grad_norm": 4.348486423492432,
"learning_rate": 1.9444444444444445e-05,
"loss": 0.3891,
"step": 910
},
{
"epoch": 32.86,
"grad_norm": 5.042387008666992,
"learning_rate": 1.9047619047619046e-05,
"loss": 0.3828,
"step": 920
},
{
"epoch": 33.0,
"eval_accuracy": 0.5772613065326633,
"eval_loss": 1.311328411102295,
"eval_runtime": 5.8619,
"eval_samples_per_second": 271.584,
"eval_steps_per_second": 2.218,
"step": 924
},
{
"epoch": 33.21,
"grad_norm": 4.366072177886963,
"learning_rate": 1.8650793650793654e-05,
"loss": 0.3692,
"step": 930
},
{
"epoch": 33.57,
"grad_norm": 3.8073909282684326,
"learning_rate": 1.8253968253968254e-05,
"loss": 0.3617,
"step": 940
},
{
"epoch": 33.93,
"grad_norm": 4.39243221282959,
"learning_rate": 1.785714285714286e-05,
"loss": 0.3791,
"step": 950
},
{
"epoch": 34.0,
"eval_accuracy": 0.5766331658291457,
"eval_loss": 1.3316831588745117,
"eval_runtime": 5.7518,
"eval_samples_per_second": 276.782,
"eval_steps_per_second": 2.26,
"step": 952
},
{
"epoch": 34.29,
"grad_norm": 4.339027404785156,
"learning_rate": 1.746031746031746e-05,
"loss": 0.3716,
"step": 960
},
{
"epoch": 34.64,
"grad_norm": 4.211394786834717,
"learning_rate": 1.7063492063492063e-05,
"loss": 0.3644,
"step": 970
},
{
"epoch": 35.0,
"grad_norm": 4.398725509643555,
"learning_rate": 1.6666666666666667e-05,
"loss": 0.3701,
"step": 980
},
{
"epoch": 35.0,
"eval_accuracy": 0.5772613065326633,
"eval_loss": 1.3383643627166748,
"eval_runtime": 5.7692,
"eval_samples_per_second": 275.948,
"eval_steps_per_second": 2.253,
"step": 980
},
{
"epoch": 35.36,
"grad_norm": 5.39536714553833,
"learning_rate": 1.626984126984127e-05,
"loss": 0.3625,
"step": 990
},
{
"epoch": 35.71,
"grad_norm": 4.096094131469727,
"learning_rate": 1.5873015873015872e-05,
"loss": 0.3566,
"step": 1000
},
{
"epoch": 36.0,
"eval_accuracy": 0.5753768844221105,
"eval_loss": 1.3406319618225098,
"eval_runtime": 5.7402,
"eval_samples_per_second": 277.342,
"eval_steps_per_second": 2.265,
"step": 1008
},
{
"epoch": 36.07,
"grad_norm": 4.44684362411499,
"learning_rate": 1.5476190476190476e-05,
"loss": 0.3621,
"step": 1010
},
{
"epoch": 36.43,
"grad_norm": 4.413670063018799,
"learning_rate": 1.5079365079365079e-05,
"loss": 0.3435,
"step": 1020
},
{
"epoch": 36.79,
"grad_norm": 4.409304618835449,
"learning_rate": 1.4682539682539683e-05,
"loss": 0.3551,
"step": 1030
},
{
"epoch": 37.0,
"eval_accuracy": 0.5766331658291457,
"eval_loss": 1.34104323387146,
"eval_runtime": 5.5746,
"eval_samples_per_second": 285.58,
"eval_steps_per_second": 2.332,
"step": 1036
},
{
"epoch": 37.14,
"grad_norm": 4.187081813812256,
"learning_rate": 1.4285714285714285e-05,
"loss": 0.3558,
"step": 1040
},
{
"epoch": 37.5,
"grad_norm": 4.235673427581787,
"learning_rate": 1.388888888888889e-05,
"loss": 0.3392,
"step": 1050
},
{
"epoch": 37.86,
"grad_norm": 4.1030497550964355,
"learning_rate": 1.3492063492063492e-05,
"loss": 0.3487,
"step": 1060
},
{
"epoch": 38.0,
"eval_accuracy": 0.5866834170854272,
"eval_loss": 1.3364226818084717,
"eval_runtime": 5.6497,
"eval_samples_per_second": 281.783,
"eval_steps_per_second": 2.301,
"step": 1064
},
{
"epoch": 38.21,
"grad_norm": 5.5322394371032715,
"learning_rate": 1.3095238095238096e-05,
"loss": 0.3359,
"step": 1070
},
{
"epoch": 38.57,
"grad_norm": 4.478665828704834,
"learning_rate": 1.2698412698412699e-05,
"loss": 0.3445,
"step": 1080
},
{
"epoch": 38.93,
"grad_norm": 4.051063060760498,
"learning_rate": 1.2301587301587301e-05,
"loss": 0.3463,
"step": 1090
},
{
"epoch": 39.0,
"eval_accuracy": 0.5810301507537688,
"eval_loss": 1.3495649099349976,
"eval_runtime": 5.6668,
"eval_samples_per_second": 280.934,
"eval_steps_per_second": 2.294,
"step": 1092
},
{
"epoch": 39.29,
"grad_norm": 3.2469401359558105,
"learning_rate": 1.1904761904761905e-05,
"loss": 0.3348,
"step": 1100
},
{
"epoch": 39.64,
"grad_norm": 3.4843826293945312,
"learning_rate": 1.1507936507936508e-05,
"loss": 0.3162,
"step": 1110
},
{
"epoch": 40.0,
"grad_norm": 4.144495010375977,
"learning_rate": 1.1111111111111112e-05,
"loss": 0.3242,
"step": 1120
},
{
"epoch": 40.0,
"eval_accuracy": 0.574748743718593,
"eval_loss": 1.3639838695526123,
"eval_runtime": 5.8819,
"eval_samples_per_second": 270.661,
"eval_steps_per_second": 2.21,
"step": 1120
},
{
"epoch": 40.36,
"grad_norm": 4.064784526824951,
"learning_rate": 1.0714285714285714e-05,
"loss": 0.3376,
"step": 1130
},
{
"epoch": 40.71,
"grad_norm": 3.94600248336792,
"learning_rate": 1.0317460317460318e-05,
"loss": 0.3308,
"step": 1140
},
{
"epoch": 41.0,
"eval_accuracy": 0.571608040201005,
"eval_loss": 1.3626537322998047,
"eval_runtime": 5.5788,
"eval_samples_per_second": 285.366,
"eval_steps_per_second": 2.33,
"step": 1148
},
{
"epoch": 41.07,
"grad_norm": 3.8364109992980957,
"learning_rate": 9.92063492063492e-06,
"loss": 0.3196,
"step": 1150
},
{
"epoch": 41.43,
"grad_norm": 3.929502010345459,
"learning_rate": 9.523809523809523e-06,
"loss": 0.3224,
"step": 1160
},
{
"epoch": 41.79,
"grad_norm": 4.359261989593506,
"learning_rate": 9.126984126984127e-06,
"loss": 0.3255,
"step": 1170
},
{
"epoch": 42.0,
"eval_accuracy": 0.5804020100502513,
"eval_loss": 1.379508376121521,
"eval_runtime": 5.7679,
"eval_samples_per_second": 276.011,
"eval_steps_per_second": 2.254,
"step": 1176
},
{
"epoch": 42.14,
"grad_norm": 4.006887912750244,
"learning_rate": 8.73015873015873e-06,
"loss": 0.3268,
"step": 1180
},
{
"epoch": 42.5,
"grad_norm": 3.5152969360351562,
"learning_rate": 8.333333333333334e-06,
"loss": 0.3166,
"step": 1190
},
{
"epoch": 42.86,
"grad_norm": 3.868173122406006,
"learning_rate": 7.936507936507936e-06,
"loss": 0.3295,
"step": 1200
},
{
"epoch": 43.0,
"eval_accuracy": 0.5797738693467337,
"eval_loss": 1.374683141708374,
"eval_runtime": 5.5853,
"eval_samples_per_second": 285.032,
"eval_steps_per_second": 2.328,
"step": 1204
},
{
"epoch": 43.21,
"grad_norm": 3.3808319568634033,
"learning_rate": 7.5396825396825394e-06,
"loss": 0.3201,
"step": 1210
},
{
"epoch": 43.57,
"grad_norm": 4.094415187835693,
"learning_rate": 7.142857142857143e-06,
"loss": 0.3223,
"step": 1220
},
{
"epoch": 43.93,
"grad_norm": 3.389286518096924,
"learning_rate": 6.746031746031746e-06,
"loss": 0.3147,
"step": 1230
},
{
"epoch": 44.0,
"eval_accuracy": 0.5860552763819096,
"eval_loss": 1.3746650218963623,
"eval_runtime": 6.0707,
"eval_samples_per_second": 262.244,
"eval_steps_per_second": 2.141,
"step": 1232
},
{
"epoch": 44.29,
"grad_norm": 3.764704942703247,
"learning_rate": 6.349206349206349e-06,
"loss": 0.3107,
"step": 1240
},
{
"epoch": 44.64,
"grad_norm": 4.65610933303833,
"learning_rate": 5.9523809523809525e-06,
"loss": 0.3058,
"step": 1250
},
{
"epoch": 45.0,
"grad_norm": 4.179033279418945,
"learning_rate": 5.555555555555556e-06,
"loss": 0.3125,
"step": 1260
},
{
"epoch": 45.0,
"eval_accuracy": 0.5816582914572864,
"eval_loss": 1.3839383125305176,
"eval_runtime": 5.8047,
"eval_samples_per_second": 274.262,
"eval_steps_per_second": 2.24,
"step": 1260
},
{
"epoch": 45.36,
"grad_norm": 7.927102088928223,
"learning_rate": 5.158730158730159e-06,
"loss": 0.3077,
"step": 1270
},
{
"epoch": 45.71,
"grad_norm": 3.8612356185913086,
"learning_rate": 4.7619047619047615e-06,
"loss": 0.3276,
"step": 1280
},
{
"epoch": 46.0,
"eval_accuracy": 0.5841708542713567,
"eval_loss": 1.3805787563323975,
"eval_runtime": 5.5264,
"eval_samples_per_second": 288.072,
"eval_steps_per_second": 2.352,
"step": 1288
},
{
"epoch": 46.07,
"grad_norm": 4.099473476409912,
"learning_rate": 4.365079365079365e-06,
"loss": 0.3076,
"step": 1290
},
{
"epoch": 46.43,
"grad_norm": 3.8270063400268555,
"learning_rate": 3.968253968253968e-06,
"loss": 0.3097,
"step": 1300
},
{
"epoch": 46.79,
"grad_norm": 3.907658576965332,
"learning_rate": 3.5714285714285714e-06,
"loss": 0.2989,
"step": 1310
},
{
"epoch": 47.0,
"eval_accuracy": 0.5885678391959799,
"eval_loss": 1.3905527591705322,
"eval_runtime": 5.6646,
"eval_samples_per_second": 281.046,
"eval_steps_per_second": 2.295,
"step": 1316
},
{
"epoch": 47.14,
"grad_norm": 4.716944694519043,
"learning_rate": 3.1746031746031746e-06,
"loss": 0.3173,
"step": 1320
},
{
"epoch": 47.5,
"grad_norm": 3.6047604084014893,
"learning_rate": 2.777777777777778e-06,
"loss": 0.3172,
"step": 1330
},
{
"epoch": 47.86,
"grad_norm": 4.362003803253174,
"learning_rate": 2.3809523809523808e-06,
"loss": 0.2941,
"step": 1340
},
{
"epoch": 48.0,
"eval_accuracy": 0.5866834170854272,
"eval_loss": 1.3876359462738037,
"eval_runtime": 5.5384,
"eval_samples_per_second": 287.447,
"eval_steps_per_second": 2.347,
"step": 1344
},
{
"epoch": 48.21,
"grad_norm": 4.251241207122803,
"learning_rate": 1.984126984126984e-06,
"loss": 0.3138,
"step": 1350
},
{
"epoch": 48.57,
"grad_norm": 3.9117441177368164,
"learning_rate": 1.5873015873015873e-06,
"loss": 0.3042,
"step": 1360
},
{
"epoch": 48.93,
"grad_norm": 3.703327178955078,
"learning_rate": 1.1904761904761904e-06,
"loss": 0.3131,
"step": 1370
},
{
"epoch": 49.0,
"eval_accuracy": 0.582286432160804,
"eval_loss": 1.3895643949508667,
"eval_runtime": 5.5003,
"eval_samples_per_second": 289.44,
"eval_steps_per_second": 2.364,
"step": 1372
},
{
"epoch": 49.29,
"grad_norm": 6.085123538970947,
"learning_rate": 7.936507936507937e-07,
"loss": 0.2984,
"step": 1380
},
{
"epoch": 49.64,
"grad_norm": 3.707207679748535,
"learning_rate": 3.9682539682539683e-07,
"loss": 0.3092,
"step": 1390
},
{
"epoch": 50.0,
"grad_norm": 4.588694095611572,
"learning_rate": 0.0,
"loss": 0.2975,
"step": 1400
},
{
"epoch": 50.0,
"eval_accuracy": 0.5835427135678392,
"eval_loss": 1.3905625343322754,
"eval_runtime": 5.7976,
"eval_samples_per_second": 274.595,
"eval_steps_per_second": 2.242,
"step": 1400
},
{
"epoch": 50.0,
"step": 1400,
"total_flos": 1.7807825640923136e+19,
"train_loss": 0.5565686808313642,
"train_runtime": 3248.0342,
"train_samples_per_second": 220.564,
"train_steps_per_second": 0.431
}
],
"logging_steps": 10,
"max_steps": 1400,
"num_input_tokens_seen": 0,
"num_train_epochs": 50,
"save_steps": 500,
"total_flos": 1.7807825640923136e+19,
"train_batch_size": 128,
"trial_name": null,
"trial_params": null
}