Beybars's picture
End of training
c996d27
{
"best_metric": 0.9857142857142858,
"best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-525",
"epoch": 7.125,
"eval_steps": 500,
"global_step": 600,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.02,
"learning_rate": 8.333333333333334e-06,
"loss": 2.4415,
"step": 10
},
{
"epoch": 0.03,
"learning_rate": 1.6666666666666667e-05,
"loss": 2.2673,
"step": 20
},
{
"epoch": 0.05,
"learning_rate": 2.5e-05,
"loss": 2.2863,
"step": 30
},
{
"epoch": 0.07,
"learning_rate": 3.3333333333333335e-05,
"loss": 2.1765,
"step": 40
},
{
"epoch": 0.08,
"learning_rate": 4.166666666666667e-05,
"loss": 2.2121,
"step": 50
},
{
"epoch": 0.1,
"learning_rate": 5e-05,
"loss": 2.0465,
"step": 60
},
{
"epoch": 0.12,
"learning_rate": 4.9074074074074075e-05,
"loss": 1.9543,
"step": 70
},
{
"epoch": 0.12,
"eval_accuracy": 0.6,
"eval_loss": 1.5145318508148193,
"eval_runtime": 7.6774,
"eval_samples_per_second": 9.118,
"eval_steps_per_second": 2.345,
"step": 75
},
{
"epoch": 1.01,
"learning_rate": 4.814814814814815e-05,
"loss": 1.4717,
"step": 80
},
{
"epoch": 1.02,
"learning_rate": 4.722222222222222e-05,
"loss": 1.2937,
"step": 90
},
{
"epoch": 1.04,
"learning_rate": 4.62962962962963e-05,
"loss": 1.0696,
"step": 100
},
{
"epoch": 1.06,
"learning_rate": 4.5370370370370374e-05,
"loss": 0.8328,
"step": 110
},
{
"epoch": 1.07,
"learning_rate": 4.4444444444444447e-05,
"loss": 0.7714,
"step": 120
},
{
"epoch": 1.09,
"learning_rate": 4.351851851851852e-05,
"loss": 0.6604,
"step": 130
},
{
"epoch": 1.11,
"learning_rate": 4.259259259259259e-05,
"loss": 0.4873,
"step": 140
},
{
"epoch": 1.12,
"learning_rate": 4.166666666666667e-05,
"loss": 0.7175,
"step": 150
},
{
"epoch": 1.12,
"eval_accuracy": 0.6571428571428571,
"eval_loss": 0.9734774827957153,
"eval_runtime": 7.7823,
"eval_samples_per_second": 8.995,
"eval_steps_per_second": 2.313,
"step": 150
},
{
"epoch": 2.02,
"learning_rate": 4.074074074074074e-05,
"loss": 0.5271,
"step": 160
},
{
"epoch": 2.03,
"learning_rate": 3.981481481481482e-05,
"loss": 0.5743,
"step": 170
},
{
"epoch": 2.05,
"learning_rate": 3.888888888888889e-05,
"loss": 0.3496,
"step": 180
},
{
"epoch": 2.07,
"learning_rate": 3.7962962962962964e-05,
"loss": 0.379,
"step": 190
},
{
"epoch": 2.08,
"learning_rate": 3.7037037037037037e-05,
"loss": 0.4612,
"step": 200
},
{
"epoch": 2.1,
"learning_rate": 3.611111111111111e-05,
"loss": 0.4579,
"step": 210
},
{
"epoch": 2.12,
"learning_rate": 3.518518518518519e-05,
"loss": 0.2914,
"step": 220
},
{
"epoch": 2.12,
"eval_accuracy": 0.8,
"eval_loss": 0.47639983892440796,
"eval_runtime": 8.028,
"eval_samples_per_second": 8.719,
"eval_steps_per_second": 2.242,
"step": 225
},
{
"epoch": 3.01,
"learning_rate": 3.425925925925926e-05,
"loss": 0.2305,
"step": 230
},
{
"epoch": 3.02,
"learning_rate": 3.3333333333333335e-05,
"loss": 0.1102,
"step": 240
},
{
"epoch": 3.04,
"learning_rate": 3.240740740740741e-05,
"loss": 0.0558,
"step": 250
},
{
"epoch": 3.06,
"learning_rate": 3.148148148148148e-05,
"loss": 0.4813,
"step": 260
},
{
"epoch": 3.08,
"learning_rate": 3.055555555555556e-05,
"loss": 0.4584,
"step": 270
},
{
"epoch": 3.09,
"learning_rate": 2.962962962962963e-05,
"loss": 0.288,
"step": 280
},
{
"epoch": 3.11,
"learning_rate": 2.8703703703703706e-05,
"loss": 0.4266,
"step": 290
},
{
"epoch": 3.12,
"learning_rate": 2.777777777777778e-05,
"loss": 0.54,
"step": 300
},
{
"epoch": 3.12,
"eval_accuracy": 0.8,
"eval_loss": 0.6035991907119751,
"eval_runtime": 7.9026,
"eval_samples_per_second": 8.858,
"eval_steps_per_second": 2.278,
"step": 300
},
{
"epoch": 4.02,
"learning_rate": 2.6851851851851855e-05,
"loss": 0.3826,
"step": 310
},
{
"epoch": 4.03,
"learning_rate": 2.5925925925925925e-05,
"loss": 0.4745,
"step": 320
},
{
"epoch": 4.05,
"learning_rate": 2.5e-05,
"loss": 0.2059,
"step": 330
},
{
"epoch": 4.07,
"learning_rate": 2.4074074074074074e-05,
"loss": 0.1123,
"step": 340
},
{
"epoch": 4.08,
"learning_rate": 2.314814814814815e-05,
"loss": 0.0443,
"step": 350
},
{
"epoch": 4.1,
"learning_rate": 2.2222222222222223e-05,
"loss": 0.1167,
"step": 360
},
{
"epoch": 4.12,
"learning_rate": 2.1296296296296296e-05,
"loss": 0.1074,
"step": 370
},
{
"epoch": 4.12,
"eval_accuracy": 0.9285714285714286,
"eval_loss": 0.19949281215667725,
"eval_runtime": 7.9455,
"eval_samples_per_second": 8.81,
"eval_steps_per_second": 2.265,
"step": 375
},
{
"epoch": 5.01,
"learning_rate": 2.037037037037037e-05,
"loss": 0.1591,
"step": 380
},
{
"epoch": 5.03,
"learning_rate": 1.9444444444444445e-05,
"loss": 0.1139,
"step": 390
},
{
"epoch": 5.04,
"learning_rate": 1.8518518518518518e-05,
"loss": 0.028,
"step": 400
},
{
"epoch": 5.06,
"learning_rate": 1.7592592592592595e-05,
"loss": 0.0256,
"step": 410
},
{
"epoch": 5.08,
"learning_rate": 1.6666666666666667e-05,
"loss": 0.0579,
"step": 420
},
{
"epoch": 5.09,
"learning_rate": 1.574074074074074e-05,
"loss": 0.0093,
"step": 430
},
{
"epoch": 5.11,
"learning_rate": 1.4814814814814815e-05,
"loss": 0.0308,
"step": 440
},
{
"epoch": 5.12,
"learning_rate": 1.388888888888889e-05,
"loss": 0.0087,
"step": 450
},
{
"epoch": 5.12,
"eval_accuracy": 0.9714285714285714,
"eval_loss": 0.12557251751422882,
"eval_runtime": 7.9052,
"eval_samples_per_second": 8.855,
"eval_steps_per_second": 2.277,
"step": 450
},
{
"epoch": 6.02,
"learning_rate": 1.2962962962962962e-05,
"loss": 0.1549,
"step": 460
},
{
"epoch": 6.03,
"learning_rate": 1.2037037037037037e-05,
"loss": 0.0059,
"step": 470
},
{
"epoch": 6.05,
"learning_rate": 1.1111111111111112e-05,
"loss": 0.0332,
"step": 480
},
{
"epoch": 6.07,
"learning_rate": 1.0185185185185185e-05,
"loss": 0.0143,
"step": 490
},
{
"epoch": 6.08,
"learning_rate": 9.259259259259259e-06,
"loss": 0.0056,
"step": 500
},
{
"epoch": 6.1,
"learning_rate": 8.333333333333334e-06,
"loss": 0.0063,
"step": 510
},
{
"epoch": 6.12,
"learning_rate": 7.4074074074074075e-06,
"loss": 0.0061,
"step": 520
},
{
"epoch": 6.12,
"eval_accuracy": 0.9857142857142858,
"eval_loss": 0.04664895310997963,
"eval_runtime": 8.079,
"eval_samples_per_second": 8.664,
"eval_steps_per_second": 2.228,
"step": 525
},
{
"epoch": 7.01,
"learning_rate": 6.481481481481481e-06,
"loss": 0.005,
"step": 530
},
{
"epoch": 7.03,
"learning_rate": 5.555555555555556e-06,
"loss": 0.0056,
"step": 540
},
{
"epoch": 7.04,
"learning_rate": 4.6296296296296296e-06,
"loss": 0.0048,
"step": 550
},
{
"epoch": 7.06,
"learning_rate": 3.7037037037037037e-06,
"loss": 0.0101,
"step": 560
},
{
"epoch": 7.08,
"learning_rate": 2.777777777777778e-06,
"loss": 0.0046,
"step": 570
},
{
"epoch": 7.09,
"learning_rate": 1.8518518518518519e-06,
"loss": 0.0098,
"step": 580
},
{
"epoch": 7.11,
"learning_rate": 9.259259259259259e-07,
"loss": 0.011,
"step": 590
},
{
"epoch": 7.12,
"learning_rate": 0.0,
"loss": 0.0082,
"step": 600
},
{
"epoch": 7.12,
"eval_accuracy": 0.9857142857142858,
"eval_loss": 0.03631773963570595,
"eval_runtime": 7.8675,
"eval_samples_per_second": 8.897,
"eval_steps_per_second": 2.288,
"step": 600
},
{
"epoch": 7.12,
"step": 600,
"total_flos": 2.990768376250368e+18,
"train_loss": 0.5080375221868356,
"train_runtime": 676.6462,
"train_samples_per_second": 3.547,
"train_steps_per_second": 0.887
},
{
"epoch": 7.12,
"eval_accuracy": 0.9483870967741935,
"eval_loss": 0.2060869336128235,
"eval_runtime": 16.6655,
"eval_samples_per_second": 9.301,
"eval_steps_per_second": 2.34,
"step": 600
},
{
"epoch": 7.12,
"eval_accuracy": 0.9483870967741935,
"eval_loss": 0.2060869336128235,
"eval_runtime": 16.8116,
"eval_samples_per_second": 9.22,
"eval_steps_per_second": 2.32,
"step": 600
}
],
"logging_steps": 10,
"max_steps": 600,
"num_input_tokens_seen": 0,
"num_train_epochs": 9223372036854775807,
"save_steps": 500,
"total_flos": 2.990768376250368e+18,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}