sheepy928's picture
Training in progress, step 1000
5a0f26e
raw
history blame
18.9 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.7713073659853451,
"global_step": 1000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"ep_loss": 0.0,
"epoch": 0.01,
"learning_rate": 5e-07,
"loss": 10.4763,
"mlm_loss": 10.4763,
"step": 10
},
{
"ep_loss": 0.0,
"epoch": 0.02,
"learning_rate": 1e-06,
"loss": 10.3052,
"mlm_loss": 10.3052,
"step": 20
},
{
"ep_loss": 0.0,
"epoch": 0.02,
"learning_rate": 1.5e-06,
"loss": 9.9641,
"mlm_loss": 9.9641,
"step": 30
},
{
"ep_loss": 0.0,
"epoch": 0.03,
"learning_rate": 2e-06,
"loss": 9.5354,
"mlm_loss": 9.5354,
"step": 40
},
{
"ep_loss": 0.0,
"epoch": 0.04,
"learning_rate": 2.5e-06,
"loss": 9.1346,
"mlm_loss": 9.1346,
"step": 50
},
{
"ep_loss": 0.0,
"epoch": 0.05,
"learning_rate": 3e-06,
"loss": 8.8162,
"mlm_loss": 8.8162,
"step": 60
},
{
"ep_loss": 0.0,
"epoch": 0.05,
"learning_rate": 3.5e-06,
"loss": 8.558,
"mlm_loss": 8.558,
"step": 70
},
{
"ep_loss": 0.0,
"epoch": 0.06,
"learning_rate": 4e-06,
"loss": 8.3459,
"mlm_loss": 8.3459,
"step": 80
},
{
"ep_loss": 0.0,
"epoch": 0.07,
"learning_rate": 4.5e-06,
"loss": 8.175,
"mlm_loss": 8.175,
"step": 90
},
{
"ep_loss": 0.0,
"epoch": 0.08,
"learning_rate": 5e-06,
"loss": 8.0255,
"mlm_loss": 8.0255,
"step": 100
},
{
"epoch": 0.08,
"eval_ep_loss": -0.14438027143478394,
"eval_loss": 7.826408386230469,
"eval_mlm_loss": 7.826408386230469,
"eval_runtime": 63.5165,
"eval_samples_per_second": 1100.014,
"eval_steps_per_second": 0.551,
"step": 100
},
{
"ep_loss": 0.0,
"epoch": 0.08,
"learning_rate": 5.5e-06,
"loss": 7.8979,
"mlm_loss": 7.8979,
"step": 110
},
{
"ep_loss": 0.0,
"epoch": 0.09,
"learning_rate": 6e-06,
"loss": 7.7543,
"mlm_loss": 7.7543,
"step": 120
},
{
"ep_loss": 0.0,
"epoch": 0.1,
"learning_rate": 6.5e-06,
"loss": 7.5942,
"mlm_loss": 7.5942,
"step": 130
},
{
"ep_loss": 0.0,
"epoch": 0.11,
"learning_rate": 7e-06,
"loss": 7.4304,
"mlm_loss": 7.4304,
"step": 140
},
{
"ep_loss": 0.0,
"epoch": 0.12,
"learning_rate": 7.5e-06,
"loss": 7.1862,
"mlm_loss": 7.1862,
"step": 150
},
{
"ep_loss": 0.0,
"epoch": 0.12,
"learning_rate": 8e-06,
"loss": 6.924,
"mlm_loss": 6.924,
"step": 160
},
{
"ep_loss": 0.0,
"epoch": 0.13,
"learning_rate": 8.500000000000002e-06,
"loss": 6.6521,
"mlm_loss": 6.6521,
"step": 170
},
{
"ep_loss": 0.0,
"epoch": 0.14,
"learning_rate": 9e-06,
"loss": 6.3639,
"mlm_loss": 6.3639,
"step": 180
},
{
"ep_loss": 0.0,
"epoch": 0.15,
"learning_rate": 9.5e-06,
"loss": 6.0915,
"mlm_loss": 6.0915,
"step": 190
},
{
"ep_loss": 0.0,
"epoch": 0.15,
"learning_rate": 1e-05,
"loss": 5.8272,
"mlm_loss": 5.8272,
"step": 200
},
{
"epoch": 0.15,
"eval_ep_loss": -0.4913683831691742,
"eval_loss": 5.531888008117676,
"eval_mlm_loss": 5.531888008117676,
"eval_runtime": 60.9938,
"eval_samples_per_second": 1145.51,
"eval_steps_per_second": 0.574,
"step": 200
},
{
"ep_loss": 0.0,
"epoch": 0.16,
"learning_rate": 1.0500000000000001e-05,
"loss": 5.5749,
"mlm_loss": 5.5749,
"step": 210
},
{
"ep_loss": 0.0,
"epoch": 0.17,
"learning_rate": 1.1e-05,
"loss": 5.3413,
"mlm_loss": 5.3413,
"step": 220
},
{
"ep_loss": 0.0,
"epoch": 0.18,
"learning_rate": 1.15e-05,
"loss": 5.142,
"mlm_loss": 5.142,
"step": 230
},
{
"ep_loss": 0.0,
"epoch": 0.19,
"learning_rate": 1.2e-05,
"loss": 4.925,
"mlm_loss": 4.925,
"step": 240
},
{
"ep_loss": 0.0,
"epoch": 0.19,
"learning_rate": 1.25e-05,
"loss": 4.7334,
"mlm_loss": 4.7334,
"step": 250
},
{
"ep_loss": 0.0,
"epoch": 0.2,
"learning_rate": 1.3e-05,
"loss": 4.5702,
"mlm_loss": 4.5702,
"step": 260
},
{
"ep_loss": 0.0,
"epoch": 0.21,
"learning_rate": 1.35e-05,
"loss": 4.4388,
"mlm_loss": 4.4388,
"step": 270
},
{
"ep_loss": 0.0,
"epoch": 0.22,
"learning_rate": 1.4e-05,
"loss": 4.288,
"mlm_loss": 4.288,
"step": 280
},
{
"ep_loss": 0.0,
"epoch": 0.22,
"learning_rate": 1.4500000000000002e-05,
"loss": 4.1403,
"mlm_loss": 4.1403,
"step": 290
},
{
"ep_loss": 0.0,
"epoch": 0.23,
"learning_rate": 1.5e-05,
"loss": 4.0232,
"mlm_loss": 4.0232,
"step": 300
},
{
"epoch": 0.23,
"eval_ep_loss": -1.1010401248931885,
"eval_loss": 3.8623688220977783,
"eval_mlm_loss": 3.8623688220977783,
"eval_runtime": 61.8594,
"eval_samples_per_second": 1129.481,
"eval_steps_per_second": 0.566,
"step": 300
},
{
"ep_loss": 0.0,
"epoch": 0.24,
"learning_rate": 1.55e-05,
"loss": 3.9135,
"mlm_loss": 3.9135,
"step": 310
},
{
"ep_loss": 0.0,
"epoch": 0.25,
"learning_rate": 1.6e-05,
"loss": 3.8252,
"mlm_loss": 3.8252,
"step": 320
},
{
"ep_loss": 0.0,
"epoch": 0.25,
"learning_rate": 1.65e-05,
"loss": 3.7462,
"mlm_loss": 3.7462,
"step": 330
},
{
"ep_loss": 0.0,
"epoch": 0.26,
"learning_rate": 1.7000000000000003e-05,
"loss": 3.6491,
"mlm_loss": 3.6491,
"step": 340
},
{
"ep_loss": 0.0,
"epoch": 0.27,
"learning_rate": 1.7500000000000002e-05,
"loss": 3.5815,
"mlm_loss": 3.5815,
"step": 350
},
{
"ep_loss": 0.0,
"epoch": 0.28,
"learning_rate": 1.8e-05,
"loss": 3.5211,
"mlm_loss": 3.5211,
"step": 360
},
{
"ep_loss": 0.0,
"epoch": 0.29,
"learning_rate": 1.85e-05,
"loss": 3.452,
"mlm_loss": 3.452,
"step": 370
},
{
"ep_loss": 0.0,
"epoch": 0.29,
"learning_rate": 1.9e-05,
"loss": 3.3747,
"mlm_loss": 3.3747,
"step": 380
},
{
"ep_loss": 0.0,
"epoch": 0.3,
"learning_rate": 1.95e-05,
"loss": 3.3163,
"mlm_loss": 3.3163,
"step": 390
},
{
"ep_loss": 0.0,
"epoch": 0.31,
"learning_rate": 2e-05,
"loss": 3.2679,
"mlm_loss": 3.2679,
"step": 400
},
{
"epoch": 0.31,
"eval_ep_loss": -1.4546582698822021,
"eval_loss": 3.1360955238342285,
"eval_mlm_loss": 3.1360955238342285,
"eval_runtime": 59.2888,
"eval_samples_per_second": 1178.451,
"eval_steps_per_second": 0.59,
"step": 400
},
{
"ep_loss": 0.0,
"epoch": 0.32,
"learning_rate": 2.05e-05,
"loss": 3.1831,
"mlm_loss": 3.1831,
"step": 410
},
{
"ep_loss": 0.0,
"epoch": 0.32,
"learning_rate": 2.1000000000000002e-05,
"loss": 3.1246,
"mlm_loss": 3.1246,
"step": 420
},
{
"ep_loss": 0.0,
"epoch": 0.33,
"learning_rate": 2.1499999999999997e-05,
"loss": 3.0658,
"mlm_loss": 3.0658,
"step": 430
},
{
"ep_loss": 0.0,
"epoch": 0.34,
"learning_rate": 2.2e-05,
"loss": 3.0032,
"mlm_loss": 3.0032,
"step": 440
},
{
"ep_loss": 0.0,
"epoch": 0.35,
"learning_rate": 2.2499999999999998e-05,
"loss": 2.931,
"mlm_loss": 2.931,
"step": 450
},
{
"ep_loss": 0.0,
"epoch": 0.35,
"learning_rate": 2.3e-05,
"loss": 2.904,
"mlm_loss": 2.904,
"step": 460
},
{
"ep_loss": 0.0,
"epoch": 0.36,
"learning_rate": 2.3500000000000002e-05,
"loss": 2.8526,
"mlm_loss": 2.8526,
"step": 470
},
{
"ep_loss": 0.0,
"epoch": 0.37,
"learning_rate": 2.4e-05,
"loss": 2.8336,
"mlm_loss": 2.8336,
"step": 480
},
{
"ep_loss": 0.0,
"epoch": 0.38,
"learning_rate": 2.4500000000000003e-05,
"loss": 2.7953,
"mlm_loss": 2.7953,
"step": 490
},
{
"ep_loss": 0.0,
"epoch": 0.39,
"learning_rate": 2.5e-05,
"loss": 2.7569,
"mlm_loss": 2.7569,
"step": 500
},
{
"epoch": 0.39,
"eval_ep_loss": -1.5984127521514893,
"eval_loss": 2.6636624336242676,
"eval_mlm_loss": 2.6636624336242676,
"eval_runtime": 64.3177,
"eval_samples_per_second": 1086.31,
"eval_steps_per_second": 0.544,
"step": 500
},
{
"ep_loss": 0.0,
"epoch": 0.39,
"learning_rate": 2.55e-05,
"loss": 2.7157,
"mlm_loss": 2.7157,
"step": 510
},
{
"ep_loss": 0.0,
"epoch": 0.4,
"learning_rate": 2.6e-05,
"loss": 2.6857,
"mlm_loss": 2.6857,
"step": 520
},
{
"ep_loss": 0.0,
"epoch": 0.41,
"learning_rate": 2.65e-05,
"loss": 2.6796,
"mlm_loss": 2.6796,
"step": 530
},
{
"ep_loss": 0.0,
"epoch": 0.42,
"learning_rate": 2.7e-05,
"loss": 2.6449,
"mlm_loss": 2.6449,
"step": 540
},
{
"ep_loss": 0.0,
"epoch": 0.42,
"learning_rate": 2.75e-05,
"loss": 2.5932,
"mlm_loss": 2.5932,
"step": 550
},
{
"ep_loss": 0.0,
"epoch": 0.43,
"learning_rate": 2.8e-05,
"loss": 2.5536,
"mlm_loss": 2.5536,
"step": 560
},
{
"ep_loss": 0.0,
"epoch": 0.44,
"learning_rate": 2.85e-05,
"loss": 2.5271,
"mlm_loss": 2.5271,
"step": 570
},
{
"ep_loss": 0.0,
"epoch": 0.45,
"learning_rate": 2.9000000000000004e-05,
"loss": 2.5057,
"mlm_loss": 2.5057,
"step": 580
},
{
"ep_loss": 0.0,
"epoch": 0.46,
"learning_rate": 2.95e-05,
"loss": 2.476,
"mlm_loss": 2.476,
"step": 590
},
{
"ep_loss": 0.0,
"epoch": 0.46,
"learning_rate": 3e-05,
"loss": 2.4265,
"mlm_loss": 2.4265,
"step": 600
},
{
"epoch": 0.46,
"eval_ep_loss": -1.7573626041412354,
"eval_loss": 2.359508752822876,
"eval_mlm_loss": 2.359508752822876,
"eval_runtime": 59.8647,
"eval_samples_per_second": 1167.114,
"eval_steps_per_second": 0.585,
"step": 600
},
{
"ep_loss": 0.0,
"epoch": 0.47,
"learning_rate": 3.05e-05,
"loss": 2.4061,
"mlm_loss": 2.4061,
"step": 610
},
{
"ep_loss": 0.0,
"epoch": 0.48,
"learning_rate": 3.1e-05,
"loss": 2.4071,
"mlm_loss": 2.4071,
"step": 620
},
{
"ep_loss": 0.0,
"epoch": 0.49,
"learning_rate": 3.15e-05,
"loss": 2.3728,
"mlm_loss": 2.3728,
"step": 630
},
{
"ep_loss": 0.0,
"epoch": 0.49,
"learning_rate": 3.2e-05,
"loss": 2.3465,
"mlm_loss": 2.3465,
"step": 640
},
{
"ep_loss": 0.0,
"epoch": 0.5,
"learning_rate": 3.2500000000000004e-05,
"loss": 2.3065,
"mlm_loss": 2.3065,
"step": 650
},
{
"ep_loss": 0.0,
"epoch": 0.51,
"learning_rate": 3.3e-05,
"loss": 2.312,
"mlm_loss": 2.312,
"step": 660
},
{
"ep_loss": 0.0,
"epoch": 0.52,
"learning_rate": 3.35e-05,
"loss": 2.29,
"mlm_loss": 2.29,
"step": 670
},
{
"ep_loss": 0.0,
"epoch": 0.52,
"learning_rate": 3.4000000000000007e-05,
"loss": 2.2939,
"mlm_loss": 2.2939,
"step": 680
},
{
"ep_loss": 0.0,
"epoch": 0.53,
"learning_rate": 3.4500000000000005e-05,
"loss": 2.2496,
"mlm_loss": 2.2496,
"step": 690
},
{
"ep_loss": 0.0,
"epoch": 0.54,
"learning_rate": 3.5000000000000004e-05,
"loss": 2.2482,
"mlm_loss": 2.2482,
"step": 700
},
{
"epoch": 0.54,
"eval_ep_loss": -1.7850463390350342,
"eval_loss": 2.1652746200561523,
"eval_mlm_loss": 2.1652746200561523,
"eval_runtime": 63.4238,
"eval_samples_per_second": 1101.621,
"eval_steps_per_second": 0.552,
"step": 700
},
{
"ep_loss": 0.0,
"epoch": 0.55,
"learning_rate": 3.5499999999999996e-05,
"loss": 2.2379,
"mlm_loss": 2.2379,
"step": 710
},
{
"ep_loss": 0.0,
"epoch": 0.56,
"learning_rate": 3.6e-05,
"loss": 2.2088,
"mlm_loss": 2.2088,
"step": 720
},
{
"ep_loss": 0.0,
"epoch": 0.56,
"learning_rate": 3.65e-05,
"loss": 2.1767,
"mlm_loss": 2.1767,
"step": 730
},
{
"ep_loss": 0.0,
"epoch": 0.57,
"learning_rate": 3.7e-05,
"loss": 2.1589,
"mlm_loss": 2.1589,
"step": 740
},
{
"ep_loss": 0.0,
"epoch": 0.58,
"learning_rate": 3.75e-05,
"loss": 2.1297,
"mlm_loss": 2.1297,
"step": 750
},
{
"ep_loss": 0.0,
"epoch": 0.59,
"learning_rate": 3.8e-05,
"loss": 2.1128,
"mlm_loss": 2.1128,
"step": 760
},
{
"ep_loss": 0.0,
"epoch": 0.59,
"learning_rate": 3.85e-05,
"loss": 2.0861,
"mlm_loss": 2.0861,
"step": 770
},
{
"ep_loss": 0.0,
"epoch": 0.6,
"learning_rate": 3.9e-05,
"loss": 2.0557,
"mlm_loss": 2.0557,
"step": 780
},
{
"ep_loss": 0.0,
"epoch": 0.61,
"learning_rate": 3.95e-05,
"loss": 2.0596,
"mlm_loss": 2.0596,
"step": 790
},
{
"ep_loss": 0.0,
"epoch": 0.62,
"learning_rate": 4e-05,
"loss": 2.0349,
"mlm_loss": 2.0349,
"step": 800
},
{
"epoch": 0.62,
"eval_ep_loss": -1.7692768573760986,
"eval_loss": 1.916082501411438,
"eval_mlm_loss": 1.916082501411438,
"eval_runtime": 62.7772,
"eval_samples_per_second": 1112.968,
"eval_steps_per_second": 0.558,
"step": 800
},
{
"ep_loss": 0.0,
"epoch": 0.62,
"learning_rate": 4.05e-05,
"loss": 1.9934,
"mlm_loss": 1.9934,
"step": 810
},
{
"ep_loss": 0.0,
"epoch": 0.63,
"learning_rate": 4.1e-05,
"loss": 1.9554,
"mlm_loss": 1.9554,
"step": 820
},
{
"ep_loss": 0.0,
"epoch": 0.64,
"learning_rate": 4.1500000000000006e-05,
"loss": 1.9348,
"mlm_loss": 1.9348,
"step": 830
},
{
"ep_loss": 0.0,
"epoch": 0.65,
"learning_rate": 4.2000000000000004e-05,
"loss": 1.8978,
"mlm_loss": 1.8978,
"step": 840
},
{
"ep_loss": 0.0,
"epoch": 0.66,
"learning_rate": 4.25e-05,
"loss": 1.8752,
"mlm_loss": 1.8752,
"step": 850
},
{
"ep_loss": 0.0,
"epoch": 0.66,
"learning_rate": 4.2999999999999995e-05,
"loss": 1.8466,
"mlm_loss": 1.8466,
"step": 860
},
{
"ep_loss": 0.0,
"epoch": 0.67,
"learning_rate": 4.35e-05,
"loss": 1.8168,
"mlm_loss": 1.8168,
"step": 870
},
{
"ep_loss": 0.0,
"epoch": 0.68,
"learning_rate": 4.4e-05,
"loss": 1.8161,
"mlm_loss": 1.8161,
"step": 880
},
{
"ep_loss": 0.0,
"epoch": 0.69,
"learning_rate": 4.45e-05,
"loss": 1.7637,
"mlm_loss": 1.7637,
"step": 890
},
{
"ep_loss": 0.0,
"epoch": 0.69,
"learning_rate": 4.4999999999999996e-05,
"loss": 1.7292,
"mlm_loss": 1.7292,
"step": 900
},
{
"epoch": 0.69,
"eval_ep_loss": -1.673869252204895,
"eval_loss": 1.6480634212493896,
"eval_mlm_loss": 1.6480634212493896,
"eval_runtime": 61.0031,
"eval_samples_per_second": 1145.335,
"eval_steps_per_second": 0.574,
"step": 900
},
{
"ep_loss": 0.0,
"epoch": 0.7,
"learning_rate": 4.55e-05,
"loss": 1.7031,
"mlm_loss": 1.7031,
"step": 910
},
{
"ep_loss": 0.0,
"epoch": 0.71,
"learning_rate": 4.6e-05,
"loss": 1.6951,
"mlm_loss": 1.6951,
"step": 920
},
{
"ep_loss": 0.0,
"epoch": 0.72,
"learning_rate": 4.65e-05,
"loss": 1.6706,
"mlm_loss": 1.6706,
"step": 930
},
{
"ep_loss": 0.0,
"epoch": 0.73,
"learning_rate": 4.7000000000000004e-05,
"loss": 1.6674,
"mlm_loss": 1.6674,
"step": 940
},
{
"ep_loss": 0.0,
"epoch": 0.73,
"learning_rate": 4.75e-05,
"loss": 1.6446,
"mlm_loss": 1.6446,
"step": 950
},
{
"ep_loss": 0.0,
"epoch": 0.74,
"learning_rate": 4.8e-05,
"loss": 1.6344,
"mlm_loss": 1.6344,
"step": 960
},
{
"ep_loss": 0.0,
"epoch": 0.75,
"learning_rate": 4.85e-05,
"loss": 1.6182,
"mlm_loss": 1.6182,
"step": 970
},
{
"ep_loss": 0.0,
"epoch": 0.76,
"learning_rate": 4.9000000000000005e-05,
"loss": 1.602,
"mlm_loss": 1.602,
"step": 980
},
{
"ep_loss": 0.0,
"epoch": 0.76,
"learning_rate": 4.9500000000000004e-05,
"loss": 1.5965,
"mlm_loss": 1.5965,
"step": 990
},
{
"ep_loss": 0.0,
"epoch": 0.77,
"learning_rate": 5e-05,
"loss": 1.6022,
"mlm_loss": 1.6022,
"step": 1000
},
{
"epoch": 0.77,
"eval_ep_loss": -1.7339978218078613,
"eval_loss": 1.521234393119812,
"eval_mlm_loss": 1.521234393119812,
"eval_runtime": 61.977,
"eval_samples_per_second": 1127.338,
"eval_steps_per_second": 0.565,
"step": 1000
}
],
"max_steps": 25920,
"num_train_epochs": 20,
"total_flos": 2.713795755507712e+17,
"trial_name": null,
"trial_params": null
}