sheepy928's picture
Training in progress, step 5000
80d1116
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.8565368299267258,
"global_step": 5000,
"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
},
{
"ep_loss": 0.0,
"epoch": 0.78,
"learning_rate": 5.05e-05,
"loss": 1.5856,
"mlm_loss": 1.5856,
"step": 1010
},
{
"ep_loss": 0.0,
"epoch": 0.79,
"learning_rate": 5.1e-05,
"loss": 1.5641,
"mlm_loss": 1.5641,
"step": 1020
},
{
"ep_loss": 0.0,
"epoch": 0.79,
"learning_rate": 5.15e-05,
"loss": 1.5455,
"mlm_loss": 1.5455,
"step": 1030
},
{
"ep_loss": 0.0,
"epoch": 0.8,
"learning_rate": 5.2e-05,
"loss": 1.5587,
"mlm_loss": 1.5587,
"step": 1040
},
{
"ep_loss": 0.0,
"epoch": 0.81,
"learning_rate": 5.25e-05,
"loss": 1.5495,
"mlm_loss": 1.5495,
"step": 1050
},
{
"ep_loss": 0.0,
"epoch": 0.82,
"learning_rate": 5.3e-05,
"loss": 1.5388,
"mlm_loss": 1.5388,
"step": 1060
},
{
"ep_loss": 0.0,
"epoch": 0.83,
"learning_rate": 5.35e-05,
"loss": 1.5085,
"mlm_loss": 1.5085,
"step": 1070
},
{
"ep_loss": 0.0,
"epoch": 0.83,
"learning_rate": 5.4e-05,
"loss": 1.5258,
"mlm_loss": 1.5258,
"step": 1080
},
{
"ep_loss": 0.0,
"epoch": 0.84,
"learning_rate": 5.45e-05,
"loss": 1.5027,
"mlm_loss": 1.5027,
"step": 1090
},
{
"ep_loss": 0.0,
"epoch": 0.85,
"learning_rate": 5.5e-05,
"loss": 1.4991,
"mlm_loss": 1.4991,
"step": 1100
},
{
"epoch": 0.85,
"eval_ep_loss": -1.8915724754333496,
"eval_loss": 1.4428980350494385,
"eval_mlm_loss": 1.4428980350494385,
"eval_runtime": 60.2275,
"eval_samples_per_second": 1160.085,
"eval_steps_per_second": 0.581,
"step": 1100
},
{
"ep_loss": 0.0,
"epoch": 0.86,
"learning_rate": 5.55e-05,
"loss": 1.5068,
"mlm_loss": 1.5068,
"step": 1110
},
{
"ep_loss": 0.0,
"epoch": 0.86,
"learning_rate": 5.6e-05,
"loss": 1.4812,
"mlm_loss": 1.4812,
"step": 1120
},
{
"ep_loss": 0.0,
"epoch": 0.87,
"learning_rate": 5.6500000000000005e-05,
"loss": 1.4765,
"mlm_loss": 1.4765,
"step": 1130
},
{
"ep_loss": 0.0,
"epoch": 0.88,
"learning_rate": 5.7e-05,
"loss": 1.4679,
"mlm_loss": 1.4679,
"step": 1140
},
{
"ep_loss": 0.0,
"epoch": 0.89,
"learning_rate": 5.75e-05,
"loss": 1.4631,
"mlm_loss": 1.4631,
"step": 1150
},
{
"ep_loss": 0.0,
"epoch": 0.89,
"learning_rate": 5.800000000000001e-05,
"loss": 1.4562,
"mlm_loss": 1.4562,
"step": 1160
},
{
"ep_loss": 0.0,
"epoch": 0.9,
"learning_rate": 5.8500000000000006e-05,
"loss": 1.4542,
"mlm_loss": 1.4542,
"step": 1170
},
{
"ep_loss": 0.0,
"epoch": 0.91,
"learning_rate": 5.9e-05,
"loss": 1.4266,
"mlm_loss": 1.4266,
"step": 1180
},
{
"ep_loss": 0.0,
"epoch": 0.92,
"learning_rate": 5.9499999999999996e-05,
"loss": 1.4181,
"mlm_loss": 1.4181,
"step": 1190
},
{
"ep_loss": 0.0,
"epoch": 0.93,
"learning_rate": 6e-05,
"loss": 1.4133,
"mlm_loss": 1.4133,
"step": 1200
},
{
"epoch": 0.93,
"eval_ep_loss": -1.9453246593475342,
"eval_loss": 1.3670274019241333,
"eval_mlm_loss": 1.3670274019241333,
"eval_runtime": 61.6667,
"eval_samples_per_second": 1133.011,
"eval_steps_per_second": 0.568,
"step": 1200
},
{
"ep_loss": 0.0,
"epoch": 0.93,
"learning_rate": 6.05e-05,
"loss": 1.414,
"mlm_loss": 1.414,
"step": 1210
},
{
"ep_loss": 0.0,
"epoch": 0.94,
"learning_rate": 6.1e-05,
"loss": 1.4032,
"mlm_loss": 1.4032,
"step": 1220
},
{
"ep_loss": 0.0,
"epoch": 0.95,
"learning_rate": 6.15e-05,
"loss": 1.3958,
"mlm_loss": 1.3958,
"step": 1230
},
{
"ep_loss": 0.0,
"epoch": 0.96,
"learning_rate": 6.2e-05,
"loss": 1.3952,
"mlm_loss": 1.3952,
"step": 1240
},
{
"ep_loss": 0.0,
"epoch": 0.96,
"learning_rate": 6.25e-05,
"loss": 1.3829,
"mlm_loss": 1.3829,
"step": 1250
},
{
"ep_loss": 0.0,
"epoch": 0.97,
"learning_rate": 6.3e-05,
"loss": 1.3915,
"mlm_loss": 1.3915,
"step": 1260
},
{
"ep_loss": 0.0,
"epoch": 0.98,
"learning_rate": 6.35e-05,
"loss": 1.3881,
"mlm_loss": 1.3881,
"step": 1270
},
{
"ep_loss": 0.0,
"epoch": 0.99,
"learning_rate": 6.4e-05,
"loss": 1.3915,
"mlm_loss": 1.3915,
"step": 1280
},
{
"ep_loss": 0.0,
"epoch": 0.99,
"learning_rate": 6.450000000000001e-05,
"loss": 1.3728,
"mlm_loss": 1.3728,
"step": 1290
},
{
"ep_loss": 0.0,
"epoch": 1.0,
"learning_rate": 6.500000000000001e-05,
"loss": 1.3379,
"mlm_loss": 1.3379,
"step": 1300
},
{
"epoch": 1.0,
"eval_ep_loss": -2.041435480117798,
"eval_loss": 1.2956568002700806,
"eval_mlm_loss": 1.2956568002700806,
"eval_runtime": 58.5906,
"eval_samples_per_second": 1192.495,
"eval_steps_per_second": 0.597,
"step": 1300
},
{
"ep_loss": 0.0,
"epoch": 1.01,
"learning_rate": 6.55e-05,
"loss": 1.3488,
"mlm_loss": 1.3488,
"step": 1310
},
{
"ep_loss": 0.0,
"epoch": 1.02,
"learning_rate": 6.6e-05,
"loss": 1.3408,
"mlm_loss": 1.3408,
"step": 1320
},
{
"ep_loss": 0.0,
"epoch": 1.03,
"learning_rate": 6.65e-05,
"loss": 1.3318,
"mlm_loss": 1.3318,
"step": 1330
},
{
"ep_loss": 0.0,
"epoch": 1.03,
"learning_rate": 6.7e-05,
"loss": 1.3216,
"mlm_loss": 1.3216,
"step": 1340
},
{
"ep_loss": 0.0,
"epoch": 1.04,
"learning_rate": 6.75e-05,
"loss": 1.3142,
"mlm_loss": 1.3142,
"step": 1350
},
{
"ep_loss": 0.0,
"epoch": 1.05,
"learning_rate": 6.800000000000001e-05,
"loss": 1.3075,
"mlm_loss": 1.3075,
"step": 1360
},
{
"ep_loss": 0.0,
"epoch": 1.06,
"learning_rate": 6.850000000000001e-05,
"loss": 1.312,
"mlm_loss": 1.312,
"step": 1370
},
{
"ep_loss": 0.0,
"epoch": 1.06,
"learning_rate": 6.900000000000001e-05,
"loss": 1.2987,
"mlm_loss": 1.2987,
"step": 1380
},
{
"ep_loss": 0.0,
"epoch": 1.07,
"learning_rate": 6.950000000000001e-05,
"loss": 1.2822,
"mlm_loss": 1.2822,
"step": 1390
},
{
"ep_loss": 0.0,
"epoch": 1.08,
"learning_rate": 7.000000000000001e-05,
"loss": 1.2817,
"mlm_loss": 1.2817,
"step": 1400
},
{
"epoch": 1.08,
"eval_ep_loss": -2.044208288192749,
"eval_loss": 1.2174968719482422,
"eval_mlm_loss": 1.2174968719482422,
"eval_runtime": 62.7874,
"eval_samples_per_second": 1112.788,
"eval_steps_per_second": 0.557,
"step": 1400
},
{
"ep_loss": 0.0,
"epoch": 1.09,
"learning_rate": 7.049999999999999e-05,
"loss": 1.2673,
"mlm_loss": 1.2673,
"step": 1410
},
{
"ep_loss": 0.0,
"epoch": 1.1,
"learning_rate": 7.099999999999999e-05,
"loss": 1.28,
"mlm_loss": 1.28,
"step": 1420
},
{
"ep_loss": 0.0,
"epoch": 1.1,
"learning_rate": 7.149999999999999e-05,
"loss": 1.2606,
"mlm_loss": 1.2606,
"step": 1430
},
{
"ep_loss": 0.0,
"epoch": 1.11,
"learning_rate": 7.2e-05,
"loss": 1.2439,
"mlm_loss": 1.2439,
"step": 1440
},
{
"ep_loss": 0.0,
"epoch": 1.12,
"learning_rate": 7.25e-05,
"loss": 1.2428,
"mlm_loss": 1.2428,
"step": 1450
},
{
"ep_loss": 0.0,
"epoch": 1.13,
"learning_rate": 7.3e-05,
"loss": 1.2342,
"mlm_loss": 1.2342,
"step": 1460
},
{
"ep_loss": 0.0,
"epoch": 1.13,
"learning_rate": 7.35e-05,
"loss": 1.2391,
"mlm_loss": 1.2391,
"step": 1470
},
{
"ep_loss": 0.0,
"epoch": 1.14,
"learning_rate": 7.4e-05,
"loss": 1.2337,
"mlm_loss": 1.2337,
"step": 1480
},
{
"ep_loss": 0.0,
"epoch": 1.15,
"learning_rate": 7.45e-05,
"loss": 1.2161,
"mlm_loss": 1.2161,
"step": 1490
},
{
"ep_loss": 0.0,
"epoch": 1.16,
"learning_rate": 7.5e-05,
"loss": 1.2083,
"mlm_loss": 1.2083,
"step": 1500
},
{
"epoch": 1.16,
"eval_ep_loss": -2.247138738632202,
"eval_loss": 1.145897388458252,
"eval_mlm_loss": 1.145897388458252,
"eval_runtime": 60.0275,
"eval_samples_per_second": 1163.951,
"eval_steps_per_second": 0.583,
"step": 1500
},
{
"ep_loss": 0.0,
"epoch": 1.16,
"learning_rate": 7.55e-05,
"loss": 1.2048,
"mlm_loss": 1.2048,
"step": 1510
},
{
"ep_loss": 0.0,
"epoch": 1.17,
"learning_rate": 7.6e-05,
"loss": 1.1934,
"mlm_loss": 1.1934,
"step": 1520
},
{
"ep_loss": 0.0,
"epoch": 1.18,
"learning_rate": 7.65e-05,
"loss": 1.1741,
"mlm_loss": 1.1741,
"step": 1530
},
{
"ep_loss": 0.0,
"epoch": 1.19,
"learning_rate": 7.7e-05,
"loss": 1.1832,
"mlm_loss": 1.1832,
"step": 1540
},
{
"ep_loss": 0.0,
"epoch": 1.2,
"learning_rate": 7.75e-05,
"loss": 1.1759,
"mlm_loss": 1.1759,
"step": 1550
},
{
"ep_loss": 0.0,
"epoch": 1.2,
"learning_rate": 7.8e-05,
"loss": 1.1776,
"mlm_loss": 1.1776,
"step": 1560
},
{
"ep_loss": 0.0,
"epoch": 1.21,
"learning_rate": 7.85e-05,
"loss": 1.164,
"mlm_loss": 1.164,
"step": 1570
},
{
"ep_loss": 0.0,
"epoch": 1.22,
"learning_rate": 7.9e-05,
"loss": 1.1447,
"mlm_loss": 1.1447,
"step": 1580
},
{
"ep_loss": 0.0,
"epoch": 1.23,
"learning_rate": 7.950000000000001e-05,
"loss": 1.1669,
"mlm_loss": 1.1669,
"step": 1590
},
{
"ep_loss": 0.0,
"epoch": 1.23,
"learning_rate": 8e-05,
"loss": 1.1331,
"mlm_loss": 1.1331,
"step": 1600
},
{
"epoch": 1.23,
"eval_ep_loss": -2.236558198928833,
"eval_loss": 1.0778173208236694,
"eval_mlm_loss": 1.0778173208236694,
"eval_runtime": 61.8774,
"eval_samples_per_second": 1129.152,
"eval_steps_per_second": 0.566,
"step": 1600
},
{
"ep_loss": 0.0,
"epoch": 1.24,
"learning_rate": 8.05e-05,
"loss": 1.1332,
"mlm_loss": 1.1332,
"step": 1610
},
{
"ep_loss": 0.0,
"epoch": 1.25,
"learning_rate": 8.1e-05,
"loss": 1.1288,
"mlm_loss": 1.1288,
"step": 1620
},
{
"ep_loss": 0.0,
"epoch": 1.26,
"learning_rate": 8.15e-05,
"loss": 1.1321,
"mlm_loss": 1.1321,
"step": 1630
},
{
"ep_loss": 0.0,
"epoch": 1.26,
"learning_rate": 8.2e-05,
"loss": 1.11,
"mlm_loss": 1.11,
"step": 1640
},
{
"ep_loss": 0.0,
"epoch": 1.27,
"learning_rate": 8.25e-05,
"loss": 1.1083,
"mlm_loss": 1.1083,
"step": 1650
},
{
"ep_loss": 0.0,
"epoch": 1.28,
"learning_rate": 8.300000000000001e-05,
"loss": 1.0999,
"mlm_loss": 1.0999,
"step": 1660
},
{
"ep_loss": 0.0,
"epoch": 1.29,
"learning_rate": 8.350000000000001e-05,
"loss": 1.1041,
"mlm_loss": 1.1041,
"step": 1670
},
{
"ep_loss": 0.0,
"epoch": 1.3,
"learning_rate": 8.400000000000001e-05,
"loss": 1.0823,
"mlm_loss": 1.0823,
"step": 1680
},
{
"ep_loss": 0.0,
"epoch": 1.3,
"learning_rate": 8.450000000000001e-05,
"loss": 1.0809,
"mlm_loss": 1.0809,
"step": 1690
},
{
"ep_loss": 0.0,
"epoch": 1.31,
"learning_rate": 8.5e-05,
"loss": 1.0859,
"mlm_loss": 1.0859,
"step": 1700
},
{
"epoch": 1.31,
"eval_ep_loss": -2.2500903606414795,
"eval_loss": 1.0151546001434326,
"eval_mlm_loss": 1.0151546001434326,
"eval_runtime": 60.0928,
"eval_samples_per_second": 1162.685,
"eval_steps_per_second": 0.582,
"step": 1700
},
{
"ep_loss": 0.0,
"epoch": 1.32,
"learning_rate": 8.55e-05,
"loss": 1.0757,
"mlm_loss": 1.0757,
"step": 1710
},
{
"ep_loss": 0.0,
"epoch": 1.33,
"learning_rate": 8.599999999999999e-05,
"loss": 1.0638,
"mlm_loss": 1.0638,
"step": 1720
},
{
"ep_loss": 0.0,
"epoch": 1.33,
"learning_rate": 8.65e-05,
"loss": 1.0554,
"mlm_loss": 1.0554,
"step": 1730
},
{
"ep_loss": 0.0,
"epoch": 1.34,
"learning_rate": 8.7e-05,
"loss": 1.0624,
"mlm_loss": 1.0624,
"step": 1740
},
{
"ep_loss": 0.0,
"epoch": 1.35,
"learning_rate": 8.75e-05,
"loss": 1.0483,
"mlm_loss": 1.0483,
"step": 1750
},
{
"ep_loss": 0.0,
"epoch": 1.36,
"learning_rate": 8.8e-05,
"loss": 1.0432,
"mlm_loss": 1.0432,
"step": 1760
},
{
"ep_loss": 0.0,
"epoch": 1.37,
"learning_rate": 8.85e-05,
"loss": 1.0464,
"mlm_loss": 1.0464,
"step": 1770
},
{
"ep_loss": 0.0,
"epoch": 1.37,
"learning_rate": 8.9e-05,
"loss": 1.0379,
"mlm_loss": 1.0379,
"step": 1780
},
{
"ep_loss": 0.0,
"epoch": 1.38,
"learning_rate": 8.95e-05,
"loss": 1.0416,
"mlm_loss": 1.0416,
"step": 1790
},
{
"ep_loss": 0.0,
"epoch": 1.39,
"learning_rate": 8.999999999999999e-05,
"loss": 1.0178,
"mlm_loss": 1.0178,
"step": 1800
},
{
"epoch": 1.39,
"eval_ep_loss": -2.3243699073791504,
"eval_loss": 0.9662984013557434,
"eval_mlm_loss": 0.9662984013557434,
"eval_runtime": 61.639,
"eval_samples_per_second": 1133.519,
"eval_steps_per_second": 0.568,
"step": 1800
},
{
"ep_loss": 0.0,
"epoch": 1.4,
"learning_rate": 9.05e-05,
"loss": 1.0189,
"mlm_loss": 1.0189,
"step": 1810
},
{
"ep_loss": 0.0,
"epoch": 1.4,
"learning_rate": 9.1e-05,
"loss": 1.0181,
"mlm_loss": 1.0181,
"step": 1820
},
{
"ep_loss": 0.0,
"epoch": 1.41,
"learning_rate": 9.15e-05,
"loss": 1.006,
"mlm_loss": 1.006,
"step": 1830
},
{
"ep_loss": 0.0,
"epoch": 1.42,
"learning_rate": 9.2e-05,
"loss": 1.0088,
"mlm_loss": 1.0088,
"step": 1840
},
{
"ep_loss": 0.0,
"epoch": 1.43,
"learning_rate": 9.25e-05,
"loss": 1.008,
"mlm_loss": 1.008,
"step": 1850
},
{
"ep_loss": 0.0,
"epoch": 1.43,
"learning_rate": 9.3e-05,
"loss": 0.9891,
"mlm_loss": 0.9891,
"step": 1860
},
{
"ep_loss": 0.0,
"epoch": 1.44,
"learning_rate": 9.35e-05,
"loss": 0.9998,
"mlm_loss": 0.9998,
"step": 1870
},
{
"ep_loss": 0.0,
"epoch": 1.45,
"learning_rate": 9.400000000000001e-05,
"loss": 0.9927,
"mlm_loss": 0.9927,
"step": 1880
},
{
"ep_loss": 0.0,
"epoch": 1.46,
"learning_rate": 9.45e-05,
"loss": 0.9825,
"mlm_loss": 0.9825,
"step": 1890
},
{
"ep_loss": 0.0,
"epoch": 1.47,
"learning_rate": 9.5e-05,
"loss": 0.979,
"mlm_loss": 0.979,
"step": 1900
},
{
"epoch": 1.47,
"eval_ep_loss": -2.297435760498047,
"eval_loss": 0.9245826005935669,
"eval_mlm_loss": 0.9245826005935669,
"eval_runtime": 60.9274,
"eval_samples_per_second": 1146.759,
"eval_steps_per_second": 0.574,
"step": 1900
},
{
"ep_loss": 0.0,
"epoch": 1.47,
"learning_rate": 9.55e-05,
"loss": 0.9782,
"mlm_loss": 0.9782,
"step": 1910
},
{
"ep_loss": 0.0,
"epoch": 1.48,
"learning_rate": 9.6e-05,
"loss": 0.9802,
"mlm_loss": 0.9802,
"step": 1920
},
{
"ep_loss": 0.0,
"epoch": 1.49,
"learning_rate": 9.65e-05,
"loss": 0.9648,
"mlm_loss": 0.9648,
"step": 1930
},
{
"ep_loss": 0.0,
"epoch": 1.5,
"learning_rate": 9.7e-05,
"loss": 0.9576,
"mlm_loss": 0.9576,
"step": 1940
},
{
"ep_loss": 0.0,
"epoch": 1.5,
"learning_rate": 9.750000000000001e-05,
"loss": 0.9629,
"mlm_loss": 0.9629,
"step": 1950
},
{
"ep_loss": 0.0,
"epoch": 1.51,
"learning_rate": 9.800000000000001e-05,
"loss": 0.9585,
"mlm_loss": 0.9585,
"step": 1960
},
{
"ep_loss": 0.0,
"epoch": 1.52,
"learning_rate": 9.850000000000001e-05,
"loss": 0.9404,
"mlm_loss": 0.9404,
"step": 1970
},
{
"ep_loss": 0.0,
"epoch": 1.53,
"learning_rate": 9.900000000000001e-05,
"loss": 0.9538,
"mlm_loss": 0.9538,
"step": 1980
},
{
"ep_loss": 0.0,
"epoch": 1.53,
"learning_rate": 9.95e-05,
"loss": 0.9591,
"mlm_loss": 0.9591,
"step": 1990
},
{
"ep_loss": 0.0,
"epoch": 1.54,
"learning_rate": 0.0001,
"loss": 0.9467,
"mlm_loss": 0.9467,
"step": 2000
},
{
"epoch": 1.54,
"eval_ep_loss": -2.388266086578369,
"eval_loss": 0.8887887597084045,
"eval_mlm_loss": 0.8887887597084045,
"eval_runtime": 59.4206,
"eval_samples_per_second": 1175.838,
"eval_steps_per_second": 0.589,
"step": 2000
},
{
"ep_loss": 0.0,
"epoch": 1.55,
"learning_rate": 0.0001005,
"loss": 0.947,
"mlm_loss": 0.947,
"step": 2010
},
{
"ep_loss": 0.0,
"epoch": 1.56,
"learning_rate": 0.000101,
"loss": 0.9407,
"mlm_loss": 0.9407,
"step": 2020
},
{
"ep_loss": 0.0,
"epoch": 1.57,
"learning_rate": 0.00010150000000000001,
"loss": 0.9352,
"mlm_loss": 0.9352,
"step": 2030
},
{
"ep_loss": 0.0,
"epoch": 1.57,
"learning_rate": 0.000102,
"loss": 0.9234,
"mlm_loss": 0.9234,
"step": 2040
},
{
"ep_loss": 0.0,
"epoch": 1.58,
"learning_rate": 0.0001025,
"loss": 0.9252,
"mlm_loss": 0.9252,
"step": 2050
},
{
"ep_loss": 0.0,
"epoch": 1.59,
"learning_rate": 0.000103,
"loss": 0.9271,
"mlm_loss": 0.9271,
"step": 2060
},
{
"ep_loss": 0.0,
"epoch": 1.6,
"learning_rate": 0.0001035,
"loss": 0.9254,
"mlm_loss": 0.9254,
"step": 2070
},
{
"ep_loss": 0.0,
"epoch": 1.6,
"learning_rate": 0.000104,
"loss": 0.9232,
"mlm_loss": 0.9232,
"step": 2080
},
{
"ep_loss": 0.0,
"epoch": 1.61,
"learning_rate": 0.00010449999999999999,
"loss": 0.919,
"mlm_loss": 0.919,
"step": 2090
},
{
"ep_loss": 0.0,
"epoch": 1.62,
"learning_rate": 0.000105,
"loss": 0.9127,
"mlm_loss": 0.9127,
"step": 2100
},
{
"epoch": 1.62,
"eval_ep_loss": -2.262110948562622,
"eval_loss": 0.8549456000328064,
"eval_mlm_loss": 0.8549456000328064,
"eval_runtime": 59.0907,
"eval_samples_per_second": 1182.403,
"eval_steps_per_second": 0.592,
"step": 2100
},
{
"ep_loss": 0.0,
"epoch": 1.63,
"learning_rate": 0.0001055,
"loss": 0.9023,
"mlm_loss": 0.9023,
"step": 2110
},
{
"ep_loss": 0.0,
"epoch": 1.64,
"learning_rate": 0.000106,
"loss": 0.9031,
"mlm_loss": 0.9031,
"step": 2120
},
{
"ep_loss": 0.0,
"epoch": 1.64,
"learning_rate": 0.0001065,
"loss": 0.9039,
"mlm_loss": 0.9039,
"step": 2130
},
{
"ep_loss": 0.0,
"epoch": 1.65,
"learning_rate": 0.000107,
"loss": 0.9022,
"mlm_loss": 0.9022,
"step": 2140
},
{
"ep_loss": 0.0,
"epoch": 1.66,
"learning_rate": 0.0001075,
"loss": 0.8978,
"mlm_loss": 0.8978,
"step": 2150
},
{
"ep_loss": 0.0,
"epoch": 1.67,
"learning_rate": 0.000108,
"loss": 0.8893,
"mlm_loss": 0.8893,
"step": 2160
},
{
"ep_loss": 0.0,
"epoch": 1.67,
"learning_rate": 0.00010850000000000001,
"loss": 0.8796,
"mlm_loss": 0.8796,
"step": 2170
},
{
"ep_loss": 0.0,
"epoch": 1.68,
"learning_rate": 0.000109,
"loss": 0.8925,
"mlm_loss": 0.8925,
"step": 2180
},
{
"ep_loss": 0.0,
"epoch": 1.69,
"learning_rate": 0.0001095,
"loss": 0.8873,
"mlm_loss": 0.8873,
"step": 2190
},
{
"ep_loss": 0.0,
"epoch": 1.7,
"learning_rate": 0.00011,
"loss": 0.8858,
"mlm_loss": 0.8858,
"step": 2200
},
{
"epoch": 1.7,
"eval_ep_loss": -2.3254947662353516,
"eval_loss": 0.8307343125343323,
"eval_mlm_loss": 0.8307343125343323,
"eval_runtime": 62.4113,
"eval_samples_per_second": 1119.493,
"eval_steps_per_second": 0.561,
"step": 2200
},
{
"ep_loss": 0.0,
"epoch": 1.7,
"learning_rate": 0.0001105,
"loss": 0.888,
"mlm_loss": 0.888,
"step": 2210
},
{
"ep_loss": 0.0,
"epoch": 1.71,
"learning_rate": 0.000111,
"loss": 0.8706,
"mlm_loss": 0.8706,
"step": 2220
},
{
"ep_loss": 0.0,
"epoch": 1.72,
"learning_rate": 0.0001115,
"loss": 0.8854,
"mlm_loss": 0.8854,
"step": 2230
},
{
"ep_loss": 0.0,
"epoch": 1.73,
"learning_rate": 0.000112,
"loss": 0.864,
"mlm_loss": 0.864,
"step": 2240
},
{
"ep_loss": 0.0,
"epoch": 1.74,
"learning_rate": 0.00011250000000000001,
"loss": 0.8675,
"mlm_loss": 0.8675,
"step": 2250
},
{
"ep_loss": 0.0,
"epoch": 1.74,
"learning_rate": 0.00011300000000000001,
"loss": 0.8654,
"mlm_loss": 0.8654,
"step": 2260
},
{
"ep_loss": 0.0,
"epoch": 1.75,
"learning_rate": 0.00011350000000000001,
"loss": 0.8593,
"mlm_loss": 0.8593,
"step": 2270
},
{
"ep_loss": 0.0,
"epoch": 1.76,
"learning_rate": 0.000114,
"loss": 0.8586,
"mlm_loss": 0.8586,
"step": 2280
},
{
"ep_loss": 0.0,
"epoch": 1.77,
"learning_rate": 0.0001145,
"loss": 0.8494,
"mlm_loss": 0.8494,
"step": 2290
},
{
"ep_loss": 0.0,
"epoch": 1.77,
"learning_rate": 0.000115,
"loss": 0.8517,
"mlm_loss": 0.8517,
"step": 2300
},
{
"epoch": 1.77,
"eval_ep_loss": -2.2645294666290283,
"eval_loss": 0.803054928779602,
"eval_mlm_loss": 0.803054928779602,
"eval_runtime": 61.7973,
"eval_samples_per_second": 1130.616,
"eval_steps_per_second": 0.566,
"step": 2300
},
{
"ep_loss": 0.0,
"epoch": 1.78,
"learning_rate": 0.0001155,
"loss": 0.8396,
"mlm_loss": 0.8396,
"step": 2310
},
{
"ep_loss": 0.0,
"epoch": 1.79,
"learning_rate": 0.00011600000000000001,
"loss": 0.8478,
"mlm_loss": 0.8478,
"step": 2320
},
{
"ep_loss": 0.0,
"epoch": 1.8,
"learning_rate": 0.00011650000000000001,
"loss": 0.8495,
"mlm_loss": 0.8495,
"step": 2330
},
{
"ep_loss": 0.0,
"epoch": 1.8,
"learning_rate": 0.00011700000000000001,
"loss": 0.8441,
"mlm_loss": 0.8441,
"step": 2340
},
{
"ep_loss": 0.0,
"epoch": 1.81,
"learning_rate": 0.0001175,
"loss": 0.8441,
"mlm_loss": 0.8441,
"step": 2350
},
{
"ep_loss": 0.0,
"epoch": 1.82,
"learning_rate": 0.000118,
"loss": 0.8406,
"mlm_loss": 0.8406,
"step": 2360
},
{
"ep_loss": 0.0,
"epoch": 1.83,
"learning_rate": 0.0001185,
"loss": 0.8322,
"mlm_loss": 0.8322,
"step": 2370
},
{
"ep_loss": 0.0,
"epoch": 1.84,
"learning_rate": 0.00011899999999999999,
"loss": 0.8282,
"mlm_loss": 0.8282,
"step": 2380
},
{
"ep_loss": 0.0,
"epoch": 1.84,
"learning_rate": 0.00011949999999999999,
"loss": 0.8437,
"mlm_loss": 0.8437,
"step": 2390
},
{
"ep_loss": 0.0,
"epoch": 1.85,
"learning_rate": 0.00012,
"loss": 0.8415,
"mlm_loss": 0.8415,
"step": 2400
},
{
"epoch": 1.85,
"eval_ep_loss": -2.278646945953369,
"eval_loss": 0.7815272212028503,
"eval_mlm_loss": 0.7815272212028503,
"eval_runtime": 60.3734,
"eval_samples_per_second": 1157.281,
"eval_steps_per_second": 0.58,
"step": 2400
},
{
"ep_loss": 0.0,
"epoch": 1.86,
"learning_rate": 0.0001205,
"loss": 0.8232,
"mlm_loss": 0.8232,
"step": 2410
},
{
"ep_loss": 0.0,
"epoch": 1.87,
"learning_rate": 0.000121,
"loss": 0.8204,
"mlm_loss": 0.8204,
"step": 2420
},
{
"ep_loss": 0.0,
"epoch": 1.87,
"learning_rate": 0.0001215,
"loss": 0.8246,
"mlm_loss": 0.8246,
"step": 2430
},
{
"ep_loss": 0.0,
"epoch": 1.88,
"learning_rate": 0.000122,
"loss": 0.8092,
"mlm_loss": 0.8092,
"step": 2440
},
{
"ep_loss": 0.0,
"epoch": 1.89,
"learning_rate": 0.0001225,
"loss": 0.812,
"mlm_loss": 0.812,
"step": 2450
},
{
"ep_loss": 0.0,
"epoch": 1.9,
"learning_rate": 0.000123,
"loss": 0.8148,
"mlm_loss": 0.8148,
"step": 2460
},
{
"ep_loss": 0.0,
"epoch": 1.91,
"learning_rate": 0.0001235,
"loss": 0.8065,
"mlm_loss": 0.8065,
"step": 2470
},
{
"ep_loss": 0.0,
"epoch": 1.91,
"learning_rate": 0.000124,
"loss": 0.8029,
"mlm_loss": 0.8029,
"step": 2480
},
{
"ep_loss": 0.0,
"epoch": 1.92,
"learning_rate": 0.0001245,
"loss": 0.8148,
"mlm_loss": 0.8148,
"step": 2490
},
{
"ep_loss": 0.0,
"epoch": 1.93,
"learning_rate": 0.000125,
"loss": 0.806,
"mlm_loss": 0.806,
"step": 2500
},
{
"epoch": 1.93,
"eval_ep_loss": -2.1858584880828857,
"eval_loss": 0.7560575604438782,
"eval_mlm_loss": 0.7560575604438782,
"eval_runtime": 60.8551,
"eval_samples_per_second": 1148.121,
"eval_steps_per_second": 0.575,
"step": 2500
},
{
"ep_loss": 0.0,
"epoch": 1.94,
"learning_rate": 0.00012550000000000001,
"loss": 0.8048,
"mlm_loss": 0.8048,
"step": 2510
},
{
"ep_loss": 0.0,
"epoch": 1.94,
"learning_rate": 0.000126,
"loss": 0.8027,
"mlm_loss": 0.8027,
"step": 2520
},
{
"ep_loss": 0.0,
"epoch": 1.95,
"learning_rate": 0.0001265,
"loss": 0.797,
"mlm_loss": 0.797,
"step": 2530
},
{
"ep_loss": 0.0,
"epoch": 1.96,
"learning_rate": 0.000127,
"loss": 0.8041,
"mlm_loss": 0.8041,
"step": 2540
},
{
"ep_loss": 0.0,
"epoch": 1.97,
"learning_rate": 0.0001275,
"loss": 0.8,
"mlm_loss": 0.8,
"step": 2550
},
{
"ep_loss": 0.0,
"epoch": 1.97,
"learning_rate": 0.000128,
"loss": 0.7955,
"mlm_loss": 0.7955,
"step": 2560
},
{
"ep_loss": 0.0,
"epoch": 1.98,
"learning_rate": 0.0001285,
"loss": 0.7943,
"mlm_loss": 0.7943,
"step": 2570
},
{
"ep_loss": 0.0,
"epoch": 1.99,
"learning_rate": 0.00012900000000000002,
"loss": 0.7934,
"mlm_loss": 0.7934,
"step": 2580
},
{
"ep_loss": 0.0,
"epoch": 2.0,
"learning_rate": 0.0001295,
"loss": 0.7884,
"mlm_loss": 0.7884,
"step": 2590
},
{
"ep_loss": 0.0,
"epoch": 2.01,
"learning_rate": 0.00013000000000000002,
"loss": 0.7892,
"mlm_loss": 0.7892,
"step": 2600
},
{
"epoch": 2.01,
"eval_ep_loss": -2.1713080406188965,
"eval_loss": 0.7403181195259094,
"eval_mlm_loss": 0.7403181195259094,
"eval_runtime": 64.561,
"eval_samples_per_second": 1082.217,
"eval_steps_per_second": 0.542,
"step": 2600
},
{
"ep_loss": 0.0,
"epoch": 2.01,
"learning_rate": 0.0001305,
"loss": 0.7818,
"mlm_loss": 0.7818,
"step": 2610
},
{
"ep_loss": 0.0,
"epoch": 2.02,
"learning_rate": 0.000131,
"loss": 0.7771,
"mlm_loss": 0.7771,
"step": 2620
},
{
"ep_loss": 0.0,
"epoch": 2.03,
"learning_rate": 0.0001315,
"loss": 0.767,
"mlm_loss": 0.767,
"step": 2630
},
{
"ep_loss": 0.0,
"epoch": 2.04,
"learning_rate": 0.000132,
"loss": 0.7826,
"mlm_loss": 0.7826,
"step": 2640
},
{
"ep_loss": 0.0,
"epoch": 2.04,
"learning_rate": 0.00013250000000000002,
"loss": 0.771,
"mlm_loss": 0.771,
"step": 2650
},
{
"ep_loss": 0.0,
"epoch": 2.05,
"learning_rate": 0.000133,
"loss": 0.7774,
"mlm_loss": 0.7774,
"step": 2660
},
{
"ep_loss": 0.0,
"epoch": 2.06,
"learning_rate": 0.00013350000000000002,
"loss": 0.7674,
"mlm_loss": 0.7674,
"step": 2670
},
{
"ep_loss": 0.0,
"epoch": 2.07,
"learning_rate": 0.000134,
"loss": 0.7737,
"mlm_loss": 0.7737,
"step": 2680
},
{
"ep_loss": 0.0,
"epoch": 2.07,
"learning_rate": 0.00013450000000000002,
"loss": 0.7639,
"mlm_loss": 0.7639,
"step": 2690
},
{
"ep_loss": 0.0,
"epoch": 2.08,
"learning_rate": 0.000135,
"loss": 0.7655,
"mlm_loss": 0.7655,
"step": 2700
},
{
"epoch": 2.08,
"eval_ep_loss": -2.082960605621338,
"eval_loss": 0.7183709144592285,
"eval_mlm_loss": 0.7183709144592285,
"eval_runtime": 61.749,
"eval_samples_per_second": 1131.5,
"eval_steps_per_second": 0.567,
"step": 2700
},
{
"ep_loss": 0.0,
"epoch": 2.09,
"learning_rate": 0.00013550000000000001,
"loss": 0.7611,
"mlm_loss": 0.7611,
"step": 2710
},
{
"ep_loss": 0.0,
"epoch": 2.1,
"learning_rate": 0.00013600000000000003,
"loss": 0.7586,
"mlm_loss": 0.7586,
"step": 2720
},
{
"ep_loss": 0.0,
"epoch": 2.11,
"learning_rate": 0.0001365,
"loss": 0.762,
"mlm_loss": 0.762,
"step": 2730
},
{
"ep_loss": 0.0,
"epoch": 2.11,
"learning_rate": 0.00013700000000000002,
"loss": 0.7479,
"mlm_loss": 0.7479,
"step": 2740
},
{
"ep_loss": 0.0,
"epoch": 2.12,
"learning_rate": 0.0001375,
"loss": 0.7514,
"mlm_loss": 0.7514,
"step": 2750
},
{
"ep_loss": 0.0,
"epoch": 2.13,
"learning_rate": 0.00013800000000000002,
"loss": 0.7476,
"mlm_loss": 0.7476,
"step": 2760
},
{
"ep_loss": 0.0,
"epoch": 2.14,
"learning_rate": 0.0001385,
"loss": 0.7629,
"mlm_loss": 0.7629,
"step": 2770
},
{
"ep_loss": 0.0,
"epoch": 2.14,
"learning_rate": 0.00013900000000000002,
"loss": 0.7432,
"mlm_loss": 0.7432,
"step": 2780
},
{
"ep_loss": 0.0,
"epoch": 2.15,
"learning_rate": 0.0001395,
"loss": 0.7449,
"mlm_loss": 0.7449,
"step": 2790
},
{
"ep_loss": 0.0,
"epoch": 2.16,
"learning_rate": 0.00014000000000000001,
"loss": 0.7379,
"mlm_loss": 0.7379,
"step": 2800
},
{
"epoch": 2.16,
"eval_ep_loss": -2.261542558670044,
"eval_loss": 0.7018134593963623,
"eval_mlm_loss": 0.7018134593963623,
"eval_runtime": 60.9379,
"eval_samples_per_second": 1146.561,
"eval_steps_per_second": 0.574,
"step": 2800
},
{
"ep_loss": 0.0,
"epoch": 2.17,
"learning_rate": 0.00014050000000000003,
"loss": 0.7465,
"mlm_loss": 0.7465,
"step": 2810
},
{
"ep_loss": 0.0,
"epoch": 2.18,
"learning_rate": 0.00014099999999999998,
"loss": 0.737,
"mlm_loss": 0.737,
"step": 2820
},
{
"ep_loss": 0.0,
"epoch": 2.18,
"learning_rate": 0.0001415,
"loss": 0.7377,
"mlm_loss": 0.7377,
"step": 2830
},
{
"ep_loss": 0.0,
"epoch": 2.19,
"learning_rate": 0.00014199999999999998,
"loss": 0.7401,
"mlm_loss": 0.7401,
"step": 2840
},
{
"ep_loss": 0.0,
"epoch": 2.2,
"learning_rate": 0.0001425,
"loss": 0.7266,
"mlm_loss": 0.7266,
"step": 2850
},
{
"ep_loss": 0.0,
"epoch": 2.21,
"learning_rate": 0.00014299999999999998,
"loss": 0.7443,
"mlm_loss": 0.7443,
"step": 2860
},
{
"ep_loss": 0.0,
"epoch": 2.21,
"learning_rate": 0.0001435,
"loss": 0.7301,
"mlm_loss": 0.7301,
"step": 2870
},
{
"ep_loss": 0.0,
"epoch": 2.22,
"learning_rate": 0.000144,
"loss": 0.7427,
"mlm_loss": 0.7427,
"step": 2880
},
{
"ep_loss": 0.0,
"epoch": 2.23,
"learning_rate": 0.0001445,
"loss": 0.7387,
"mlm_loss": 0.7387,
"step": 2890
},
{
"ep_loss": 0.0,
"epoch": 2.24,
"learning_rate": 0.000145,
"loss": 0.721,
"mlm_loss": 0.721,
"step": 2900
},
{
"epoch": 2.24,
"eval_ep_loss": -2.377054214477539,
"eval_loss": 0.6826241612434387,
"eval_mlm_loss": 0.6826241612434387,
"eval_runtime": 60.9503,
"eval_samples_per_second": 1146.327,
"eval_steps_per_second": 0.574,
"step": 2900
},
{
"ep_loss": 0.0,
"epoch": 2.24,
"learning_rate": 0.00014549999999999999,
"loss": 0.7352,
"mlm_loss": 0.7352,
"step": 2910
},
{
"ep_loss": 0.0,
"epoch": 2.25,
"learning_rate": 0.000146,
"loss": 0.7219,
"mlm_loss": 0.7219,
"step": 2920
},
{
"ep_loss": 0.0,
"epoch": 2.26,
"learning_rate": 0.00014649999999999998,
"loss": 0.729,
"mlm_loss": 0.729,
"step": 2930
},
{
"ep_loss": 0.0,
"epoch": 2.27,
"learning_rate": 0.000147,
"loss": 0.7148,
"mlm_loss": 0.7148,
"step": 2940
},
{
"ep_loss": 0.0,
"epoch": 2.28,
"learning_rate": 0.0001475,
"loss": 0.721,
"mlm_loss": 0.721,
"step": 2950
},
{
"ep_loss": 0.0,
"epoch": 2.28,
"learning_rate": 0.000148,
"loss": 0.7113,
"mlm_loss": 0.7113,
"step": 2960
},
{
"ep_loss": 0.0,
"epoch": 2.29,
"learning_rate": 0.0001485,
"loss": 0.711,
"mlm_loss": 0.711,
"step": 2970
},
{
"ep_loss": 0.0,
"epoch": 2.3,
"learning_rate": 0.000149,
"loss": 0.7109,
"mlm_loss": 0.7109,
"step": 2980
},
{
"ep_loss": 0.0,
"epoch": 2.31,
"learning_rate": 0.0001495,
"loss": 0.7121,
"mlm_loss": 0.7121,
"step": 2990
},
{
"ep_loss": 0.0,
"epoch": 2.31,
"learning_rate": 0.00015,
"loss": 0.7106,
"mlm_loss": 0.7106,
"step": 3000
},
{
"epoch": 2.31,
"eval_ep_loss": -2.0994629859924316,
"eval_loss": 0.6672462821006775,
"eval_mlm_loss": 0.6672462821006775,
"eval_runtime": 63.0197,
"eval_samples_per_second": 1108.686,
"eval_steps_per_second": 0.555,
"step": 3000
},
{
"ep_loss": 0.0,
"epoch": 2.32,
"learning_rate": 0.0001505,
"loss": 0.7067,
"mlm_loss": 0.7067,
"step": 3010
},
{
"ep_loss": 0.0,
"epoch": 2.33,
"learning_rate": 0.000151,
"loss": 0.7082,
"mlm_loss": 0.7082,
"step": 3020
},
{
"ep_loss": 0.0,
"epoch": 2.34,
"learning_rate": 0.0001515,
"loss": 0.7097,
"mlm_loss": 0.7097,
"step": 3030
},
{
"ep_loss": 0.0,
"epoch": 2.34,
"learning_rate": 0.000152,
"loss": 0.6972,
"mlm_loss": 0.6972,
"step": 3040
},
{
"ep_loss": 0.0,
"epoch": 2.35,
"learning_rate": 0.0001525,
"loss": 0.7051,
"mlm_loss": 0.7051,
"step": 3050
},
{
"ep_loss": 0.0,
"epoch": 2.36,
"learning_rate": 0.000153,
"loss": 0.7029,
"mlm_loss": 0.7029,
"step": 3060
},
{
"ep_loss": 0.0,
"epoch": 2.37,
"learning_rate": 0.0001535,
"loss": 0.7037,
"mlm_loss": 0.7037,
"step": 3070
},
{
"ep_loss": 0.0,
"epoch": 2.38,
"learning_rate": 0.000154,
"loss": 0.6966,
"mlm_loss": 0.6966,
"step": 3080
},
{
"ep_loss": 0.0,
"epoch": 2.38,
"learning_rate": 0.00015450000000000001,
"loss": 0.6948,
"mlm_loss": 0.6948,
"step": 3090
},
{
"ep_loss": 0.0,
"epoch": 2.39,
"learning_rate": 0.000155,
"loss": 0.6974,
"mlm_loss": 0.6974,
"step": 3100
},
{
"epoch": 2.39,
"eval_ep_loss": -2.283140182495117,
"eval_loss": 0.6495695114135742,
"eval_mlm_loss": 0.6495695114135742,
"eval_runtime": 59.8554,
"eval_samples_per_second": 1167.296,
"eval_steps_per_second": 0.585,
"step": 3100
},
{
"ep_loss": 0.0,
"epoch": 2.4,
"learning_rate": 0.0001555,
"loss": 0.6877,
"mlm_loss": 0.6877,
"step": 3110
},
{
"ep_loss": 0.0,
"epoch": 2.41,
"learning_rate": 0.000156,
"loss": 0.6862,
"mlm_loss": 0.6862,
"step": 3120
},
{
"ep_loss": 0.0,
"epoch": 2.41,
"learning_rate": 0.0001565,
"loss": 0.6893,
"mlm_loss": 0.6893,
"step": 3130
},
{
"ep_loss": 0.0,
"epoch": 2.42,
"learning_rate": 0.000157,
"loss": 0.6943,
"mlm_loss": 0.6943,
"step": 3140
},
{
"ep_loss": 0.0,
"epoch": 2.43,
"learning_rate": 0.0001575,
"loss": 0.6804,
"mlm_loss": 0.6804,
"step": 3150
},
{
"ep_loss": 0.0,
"epoch": 2.44,
"learning_rate": 0.000158,
"loss": 0.6813,
"mlm_loss": 0.6813,
"step": 3160
},
{
"ep_loss": 0.0,
"epoch": 2.45,
"learning_rate": 0.0001585,
"loss": 0.6828,
"mlm_loss": 0.6828,
"step": 3170
},
{
"ep_loss": 0.0,
"epoch": 2.45,
"learning_rate": 0.00015900000000000002,
"loss": 0.6736,
"mlm_loss": 0.6736,
"step": 3180
},
{
"ep_loss": 0.0,
"epoch": 2.46,
"learning_rate": 0.0001595,
"loss": 0.6763,
"mlm_loss": 0.6763,
"step": 3190
},
{
"ep_loss": 0.0,
"epoch": 2.47,
"learning_rate": 0.00016,
"loss": 0.6796,
"mlm_loss": 0.6796,
"step": 3200
},
{
"epoch": 2.47,
"eval_ep_loss": -2.57122540473938,
"eval_loss": 0.6341940760612488,
"eval_mlm_loss": 0.6341940760612488,
"eval_runtime": 63.1531,
"eval_samples_per_second": 1106.343,
"eval_steps_per_second": 0.554,
"step": 3200
},
{
"ep_loss": 0.0,
"epoch": 2.48,
"learning_rate": 0.0001605,
"loss": 0.6688,
"mlm_loss": 0.6688,
"step": 3210
},
{
"ep_loss": 0.0,
"epoch": 2.48,
"learning_rate": 0.000161,
"loss": 0.6725,
"mlm_loss": 0.6725,
"step": 3220
},
{
"ep_loss": 0.0,
"epoch": 2.49,
"learning_rate": 0.0001615,
"loss": 0.6718,
"mlm_loss": 0.6718,
"step": 3230
},
{
"ep_loss": 0.0,
"epoch": 2.5,
"learning_rate": 0.000162,
"loss": 0.6675,
"mlm_loss": 0.6675,
"step": 3240
},
{
"ep_loss": 0.0,
"epoch": 2.51,
"learning_rate": 0.00016250000000000002,
"loss": 0.6665,
"mlm_loss": 0.6665,
"step": 3250
},
{
"ep_loss": 0.0,
"epoch": 2.51,
"learning_rate": 0.000163,
"loss": 0.6615,
"mlm_loss": 0.6615,
"step": 3260
},
{
"ep_loss": 0.0,
"epoch": 2.52,
"learning_rate": 0.00016350000000000002,
"loss": 0.6784,
"mlm_loss": 0.6784,
"step": 3270
},
{
"ep_loss": 0.0,
"epoch": 2.53,
"learning_rate": 0.000164,
"loss": 0.6686,
"mlm_loss": 0.6686,
"step": 3280
},
{
"ep_loss": 0.0,
"epoch": 2.54,
"learning_rate": 0.00016450000000000001,
"loss": 0.6697,
"mlm_loss": 0.6697,
"step": 3290
},
{
"ep_loss": 0.0,
"epoch": 2.55,
"learning_rate": 0.000165,
"loss": 0.6621,
"mlm_loss": 0.6621,
"step": 3300
},
{
"epoch": 2.55,
"eval_ep_loss": -2.320197820663452,
"eval_loss": 0.6192212104797363,
"eval_mlm_loss": 0.6192212104797363,
"eval_runtime": 60.3086,
"eval_samples_per_second": 1158.524,
"eval_steps_per_second": 0.58,
"step": 3300
},
{
"ep_loss": 0.0,
"epoch": 2.55,
"learning_rate": 0.0001655,
"loss": 0.6645,
"mlm_loss": 0.6645,
"step": 3310
},
{
"ep_loss": 0.0,
"epoch": 2.56,
"learning_rate": 0.00016600000000000002,
"loss": 0.6594,
"mlm_loss": 0.6594,
"step": 3320
},
{
"ep_loss": 0.0,
"epoch": 2.57,
"learning_rate": 0.0001665,
"loss": 0.6585,
"mlm_loss": 0.6585,
"step": 3330
},
{
"ep_loss": 0.0,
"epoch": 2.58,
"learning_rate": 0.00016700000000000002,
"loss": 0.6499,
"mlm_loss": 0.6499,
"step": 3340
},
{
"ep_loss": 0.0,
"epoch": 2.58,
"learning_rate": 0.0001675,
"loss": 0.641,
"mlm_loss": 0.641,
"step": 3350
},
{
"ep_loss": 0.0,
"epoch": 2.59,
"learning_rate": 0.00016800000000000002,
"loss": 0.653,
"mlm_loss": 0.653,
"step": 3360
},
{
"ep_loss": 0.0,
"epoch": 2.6,
"learning_rate": 0.0001685,
"loss": 0.6526,
"mlm_loss": 0.6526,
"step": 3370
},
{
"ep_loss": 0.0,
"epoch": 2.61,
"learning_rate": 0.00016900000000000002,
"loss": 0.6588,
"mlm_loss": 0.6588,
"step": 3380
},
{
"ep_loss": 0.0,
"epoch": 2.61,
"learning_rate": 0.00016950000000000003,
"loss": 0.6457,
"mlm_loss": 0.6457,
"step": 3390
},
{
"ep_loss": 0.0,
"epoch": 2.62,
"learning_rate": 0.00017,
"loss": 0.6464,
"mlm_loss": 0.6464,
"step": 3400
},
{
"epoch": 2.62,
"eval_ep_loss": -2.441110610961914,
"eval_loss": 0.6078351736068726,
"eval_mlm_loss": 0.6078351736068726,
"eval_runtime": 60.4806,
"eval_samples_per_second": 1155.23,
"eval_steps_per_second": 0.579,
"step": 3400
},
{
"ep_loss": 0.0,
"epoch": 2.63,
"learning_rate": 0.00017050000000000002,
"loss": 0.6517,
"mlm_loss": 0.6517,
"step": 3410
},
{
"ep_loss": 0.0,
"epoch": 2.64,
"learning_rate": 0.000171,
"loss": 0.6473,
"mlm_loss": 0.6473,
"step": 3420
},
{
"ep_loss": 0.0,
"epoch": 2.65,
"learning_rate": 0.00017150000000000002,
"loss": 0.6475,
"mlm_loss": 0.6475,
"step": 3430
},
{
"ep_loss": 0.0,
"epoch": 2.65,
"learning_rate": 0.00017199999999999998,
"loss": 0.649,
"mlm_loss": 0.649,
"step": 3440
},
{
"ep_loss": 0.0,
"epoch": 2.66,
"learning_rate": 0.0001725,
"loss": 0.652,
"mlm_loss": 0.652,
"step": 3450
},
{
"ep_loss": 0.0,
"epoch": 2.67,
"learning_rate": 0.000173,
"loss": 0.6484,
"mlm_loss": 0.6484,
"step": 3460
},
{
"ep_loss": 0.0,
"epoch": 2.68,
"learning_rate": 0.0001735,
"loss": 0.6405,
"mlm_loss": 0.6405,
"step": 3470
},
{
"ep_loss": 0.0,
"epoch": 2.68,
"learning_rate": 0.000174,
"loss": 0.6421,
"mlm_loss": 0.6421,
"step": 3480
},
{
"ep_loss": 0.0,
"epoch": 2.69,
"learning_rate": 0.00017449999999999999,
"loss": 0.6313,
"mlm_loss": 0.6313,
"step": 3490
},
{
"ep_loss": 0.0,
"epoch": 2.7,
"learning_rate": 0.000175,
"loss": 0.6228,
"mlm_loss": 0.6228,
"step": 3500
},
{
"epoch": 2.7,
"eval_ep_loss": -2.3252084255218506,
"eval_loss": 0.5944607257843018,
"eval_mlm_loss": 0.5944607257843018,
"eval_runtime": 59.8998,
"eval_samples_per_second": 1166.431,
"eval_steps_per_second": 0.584,
"step": 3500
},
{
"ep_loss": 0.0,
"epoch": 2.71,
"learning_rate": 0.00017549999999999998,
"loss": 0.6276,
"mlm_loss": 0.6276,
"step": 3510
},
{
"ep_loss": 0.0,
"epoch": 2.72,
"learning_rate": 0.000176,
"loss": 0.632,
"mlm_loss": 0.632,
"step": 3520
},
{
"ep_loss": 0.0,
"epoch": 2.72,
"learning_rate": 0.00017649999999999998,
"loss": 0.629,
"mlm_loss": 0.629,
"step": 3530
},
{
"ep_loss": 0.0,
"epoch": 2.73,
"learning_rate": 0.000177,
"loss": 0.6327,
"mlm_loss": 0.6327,
"step": 3540
},
{
"ep_loss": 0.0,
"epoch": 2.74,
"learning_rate": 0.0001775,
"loss": 0.6307,
"mlm_loss": 0.6307,
"step": 3550
},
{
"ep_loss": 0.0,
"epoch": 2.75,
"learning_rate": 0.000178,
"loss": 0.6156,
"mlm_loss": 0.6156,
"step": 3560
},
{
"ep_loss": 0.0,
"epoch": 2.75,
"learning_rate": 0.0001785,
"loss": 0.6246,
"mlm_loss": 0.6246,
"step": 3570
},
{
"ep_loss": 0.0,
"epoch": 2.76,
"learning_rate": 0.000179,
"loss": 0.6223,
"mlm_loss": 0.6223,
"step": 3580
},
{
"ep_loss": 0.0,
"epoch": 2.77,
"learning_rate": 0.0001795,
"loss": 0.6359,
"mlm_loss": 0.6359,
"step": 3590
},
{
"ep_loss": 0.0,
"epoch": 2.78,
"learning_rate": 0.00017999999999999998,
"loss": 0.6192,
"mlm_loss": 0.6192,
"step": 3600
},
{
"epoch": 2.78,
"eval_ep_loss": -2.543795585632324,
"eval_loss": 0.5830370783805847,
"eval_mlm_loss": 0.5830370783805847,
"eval_runtime": 62.3115,
"eval_samples_per_second": 1121.285,
"eval_steps_per_second": 0.562,
"step": 3600
},
{
"ep_loss": 0.0,
"epoch": 2.78,
"learning_rate": 0.0001805,
"loss": 0.619,
"mlm_loss": 0.619,
"step": 3610
},
{
"ep_loss": 0.0,
"epoch": 2.79,
"learning_rate": 0.000181,
"loss": 0.6107,
"mlm_loss": 0.6107,
"step": 3620
},
{
"ep_loss": 0.0,
"epoch": 2.8,
"learning_rate": 0.0001815,
"loss": 0.6287,
"mlm_loss": 0.6287,
"step": 3630
},
{
"ep_loss": 0.0,
"epoch": 2.81,
"learning_rate": 0.000182,
"loss": 0.616,
"mlm_loss": 0.616,
"step": 3640
},
{
"ep_loss": 0.0,
"epoch": 2.82,
"learning_rate": 0.0001825,
"loss": 0.6199,
"mlm_loss": 0.6199,
"step": 3650
},
{
"ep_loss": 0.0,
"epoch": 2.82,
"learning_rate": 0.000183,
"loss": 0.6179,
"mlm_loss": 0.6179,
"step": 3660
},
{
"ep_loss": 0.0,
"epoch": 2.83,
"learning_rate": 0.0001835,
"loss": 0.6159,
"mlm_loss": 0.6159,
"step": 3670
},
{
"ep_loss": 0.0,
"epoch": 2.84,
"learning_rate": 0.000184,
"loss": 0.6067,
"mlm_loss": 0.6067,
"step": 3680
},
{
"ep_loss": 0.0,
"epoch": 2.85,
"learning_rate": 0.0001845,
"loss": 0.6079,
"mlm_loss": 0.6079,
"step": 3690
},
{
"ep_loss": 0.0,
"epoch": 2.85,
"learning_rate": 0.000185,
"loss": 0.6063,
"mlm_loss": 0.6063,
"step": 3700
},
{
"epoch": 2.85,
"eval_ep_loss": -2.2521820068359375,
"eval_loss": 0.5688546299934387,
"eval_mlm_loss": 0.5688546299934387,
"eval_runtime": 60.0349,
"eval_samples_per_second": 1163.807,
"eval_steps_per_second": 0.583,
"step": 3700
},
{
"ep_loss": 0.0,
"epoch": 2.86,
"learning_rate": 0.0001855,
"loss": 0.6156,
"mlm_loss": 0.6156,
"step": 3710
},
{
"ep_loss": 0.0,
"epoch": 2.87,
"learning_rate": 0.000186,
"loss": 0.6093,
"mlm_loss": 0.6093,
"step": 3720
},
{
"ep_loss": 0.0,
"epoch": 2.88,
"learning_rate": 0.0001865,
"loss": 0.6073,
"mlm_loss": 0.6073,
"step": 3730
},
{
"ep_loss": 0.0,
"epoch": 2.88,
"learning_rate": 0.000187,
"loss": 0.6039,
"mlm_loss": 0.6039,
"step": 3740
},
{
"ep_loss": 0.0,
"epoch": 2.89,
"learning_rate": 0.0001875,
"loss": 0.5977,
"mlm_loss": 0.5977,
"step": 3750
},
{
"ep_loss": 0.0,
"epoch": 2.9,
"learning_rate": 0.00018800000000000002,
"loss": 0.6,
"mlm_loss": 0.6,
"step": 3760
},
{
"ep_loss": 0.0,
"epoch": 2.91,
"learning_rate": 0.0001885,
"loss": 0.6022,
"mlm_loss": 0.6022,
"step": 3770
},
{
"ep_loss": 0.0,
"epoch": 2.92,
"learning_rate": 0.000189,
"loss": 0.6015,
"mlm_loss": 0.6015,
"step": 3780
},
{
"ep_loss": 0.0,
"epoch": 2.92,
"learning_rate": 0.0001895,
"loss": 0.6016,
"mlm_loss": 0.6016,
"step": 3790
},
{
"ep_loss": 0.0,
"epoch": 2.93,
"learning_rate": 0.00019,
"loss": 0.5959,
"mlm_loss": 0.5959,
"step": 3800
},
{
"epoch": 2.93,
"eval_ep_loss": -2.366133213043213,
"eval_loss": 0.5573724508285522,
"eval_mlm_loss": 0.5573724508285522,
"eval_runtime": 61.8753,
"eval_samples_per_second": 1129.191,
"eval_steps_per_second": 0.566,
"step": 3800
},
{
"ep_loss": 0.0,
"epoch": 2.94,
"learning_rate": 0.0001905,
"loss": 0.6003,
"mlm_loss": 0.6003,
"step": 3810
},
{
"ep_loss": 0.0,
"epoch": 2.95,
"learning_rate": 0.000191,
"loss": 0.5984,
"mlm_loss": 0.5984,
"step": 3820
},
{
"ep_loss": 0.0,
"epoch": 2.95,
"learning_rate": 0.00019150000000000002,
"loss": 0.6005,
"mlm_loss": 0.6005,
"step": 3830
},
{
"ep_loss": 0.0,
"epoch": 2.96,
"learning_rate": 0.000192,
"loss": 0.5895,
"mlm_loss": 0.5895,
"step": 3840
},
{
"ep_loss": 0.0,
"epoch": 2.97,
"learning_rate": 0.00019250000000000002,
"loss": 0.5912,
"mlm_loss": 0.5912,
"step": 3850
},
{
"ep_loss": 0.0,
"epoch": 2.98,
"learning_rate": 0.000193,
"loss": 0.5828,
"mlm_loss": 0.5828,
"step": 3860
},
{
"ep_loss": 0.0,
"epoch": 2.98,
"learning_rate": 0.00019350000000000001,
"loss": 0.5865,
"mlm_loss": 0.5865,
"step": 3870
},
{
"ep_loss": 0.0,
"epoch": 2.99,
"learning_rate": 0.000194,
"loss": 0.5854,
"mlm_loss": 0.5854,
"step": 3880
},
{
"ep_loss": 0.0,
"epoch": 3.0,
"learning_rate": 0.0001945,
"loss": 0.5883,
"mlm_loss": 0.5883,
"step": 3890
},
{
"ep_loss": 0.0,
"epoch": 3.01,
"learning_rate": 0.00019500000000000002,
"loss": 0.5743,
"mlm_loss": 0.5743,
"step": 3900
},
{
"epoch": 3.01,
"eval_ep_loss": -2.332629680633545,
"eval_loss": 0.551331102848053,
"eval_mlm_loss": 0.551331102848053,
"eval_runtime": 61.0219,
"eval_samples_per_second": 1144.983,
"eval_steps_per_second": 0.574,
"step": 3900
},
{
"ep_loss": 0.0,
"epoch": 3.02,
"learning_rate": 0.0001955,
"loss": 0.5813,
"mlm_loss": 0.5813,
"step": 3910
},
{
"ep_loss": 0.0,
"epoch": 3.02,
"learning_rate": 0.00019600000000000002,
"loss": 0.5813,
"mlm_loss": 0.5813,
"step": 3920
},
{
"ep_loss": 0.0,
"epoch": 3.03,
"learning_rate": 0.0001965,
"loss": 0.5899,
"mlm_loss": 0.5899,
"step": 3930
},
{
"ep_loss": 0.0,
"epoch": 3.04,
"learning_rate": 0.00019700000000000002,
"loss": 0.5849,
"mlm_loss": 0.5849,
"step": 3940
},
{
"ep_loss": 0.0,
"epoch": 3.05,
"learning_rate": 0.0001975,
"loss": 0.5823,
"mlm_loss": 0.5823,
"step": 3950
},
{
"ep_loss": 0.0,
"epoch": 3.05,
"learning_rate": 0.00019800000000000002,
"loss": 0.5794,
"mlm_loss": 0.5794,
"step": 3960
},
{
"ep_loss": 0.0,
"epoch": 3.06,
"learning_rate": 0.00019850000000000003,
"loss": 0.5837,
"mlm_loss": 0.5837,
"step": 3970
},
{
"ep_loss": 0.0,
"epoch": 3.07,
"learning_rate": 0.000199,
"loss": 0.5783,
"mlm_loss": 0.5783,
"step": 3980
},
{
"ep_loss": 0.0,
"epoch": 3.08,
"learning_rate": 0.00019950000000000002,
"loss": 0.5733,
"mlm_loss": 0.5733,
"step": 3990
},
{
"ep_loss": 0.0,
"epoch": 3.09,
"learning_rate": 0.0002,
"loss": 0.5701,
"mlm_loss": 0.5701,
"step": 4000
},
{
"epoch": 3.09,
"eval_ep_loss": -2.3982014656066895,
"eval_loss": 0.5421211123466492,
"eval_mlm_loss": 0.5421211123466492,
"eval_runtime": 62.0302,
"eval_samples_per_second": 1126.37,
"eval_steps_per_second": 0.564,
"step": 4000
},
{
"ep_loss": 0.0,
"epoch": 3.09,
"learning_rate": 0.00020050000000000002,
"loss": 0.5737,
"mlm_loss": 0.5737,
"step": 4010
},
{
"ep_loss": 0.0,
"epoch": 3.1,
"learning_rate": 0.000201,
"loss": 0.5645,
"mlm_loss": 0.5645,
"step": 4020
},
{
"ep_loss": 0.0,
"epoch": 3.11,
"learning_rate": 0.00020150000000000002,
"loss": 0.5758,
"mlm_loss": 0.5758,
"step": 4030
},
{
"ep_loss": 0.0,
"epoch": 3.12,
"learning_rate": 0.000202,
"loss": 0.5849,
"mlm_loss": 0.5849,
"step": 4040
},
{
"ep_loss": 0.0,
"epoch": 3.12,
"learning_rate": 0.00020250000000000002,
"loss": 0.5667,
"mlm_loss": 0.5667,
"step": 4050
},
{
"ep_loss": 0.0,
"epoch": 3.13,
"learning_rate": 0.00020300000000000003,
"loss": 0.5726,
"mlm_loss": 0.5726,
"step": 4060
},
{
"ep_loss": 0.0,
"epoch": 3.14,
"learning_rate": 0.00020349999999999999,
"loss": 0.5645,
"mlm_loss": 0.5645,
"step": 4070
},
{
"ep_loss": 0.0,
"epoch": 3.15,
"learning_rate": 0.000204,
"loss": 0.5681,
"mlm_loss": 0.5681,
"step": 4080
},
{
"ep_loss": 0.0,
"epoch": 3.15,
"learning_rate": 0.00020449999999999998,
"loss": 0.564,
"mlm_loss": 0.564,
"step": 4090
},
{
"ep_loss": 0.0,
"epoch": 3.16,
"learning_rate": 0.000205,
"loss": 0.5655,
"mlm_loss": 0.5655,
"step": 4100
},
{
"epoch": 3.16,
"eval_ep_loss": -2.218219041824341,
"eval_loss": 0.5314173698425293,
"eval_mlm_loss": 0.5314173698425293,
"eval_runtime": 60.8804,
"eval_samples_per_second": 1147.644,
"eval_steps_per_second": 0.575,
"step": 4100
},
{
"ep_loss": 0.0,
"epoch": 3.17,
"learning_rate": 0.00020549999999999998,
"loss": 0.5654,
"mlm_loss": 0.5654,
"step": 4110
},
{
"ep_loss": 0.0,
"epoch": 3.18,
"learning_rate": 0.000206,
"loss": 0.5647,
"mlm_loss": 0.5647,
"step": 4120
},
{
"ep_loss": 0.0,
"epoch": 3.19,
"learning_rate": 0.0002065,
"loss": 0.568,
"mlm_loss": 0.568,
"step": 4130
},
{
"ep_loss": 0.0,
"epoch": 3.19,
"learning_rate": 0.000207,
"loss": 0.5538,
"mlm_loss": 0.5538,
"step": 4140
},
{
"ep_loss": 0.0,
"epoch": 3.2,
"learning_rate": 0.0002075,
"loss": 0.5585,
"mlm_loss": 0.5585,
"step": 4150
},
{
"ep_loss": 0.0,
"epoch": 3.21,
"learning_rate": 0.000208,
"loss": 0.5578,
"mlm_loss": 0.5578,
"step": 4160
},
{
"ep_loss": 0.0,
"epoch": 3.22,
"learning_rate": 0.0002085,
"loss": 0.5508,
"mlm_loss": 0.5508,
"step": 4170
},
{
"ep_loss": 0.0,
"epoch": 3.22,
"learning_rate": 0.00020899999999999998,
"loss": 0.5644,
"mlm_loss": 0.5644,
"step": 4180
},
{
"ep_loss": 0.0,
"epoch": 3.23,
"learning_rate": 0.0002095,
"loss": 0.5642,
"mlm_loss": 0.5642,
"step": 4190
},
{
"ep_loss": 0.0,
"epoch": 3.24,
"learning_rate": 0.00021,
"loss": 0.561,
"mlm_loss": 0.561,
"step": 4200
},
{
"epoch": 3.24,
"eval_ep_loss": -2.287240743637085,
"eval_loss": 0.5224212408065796,
"eval_mlm_loss": 0.5224212408065796,
"eval_runtime": 61.3268,
"eval_samples_per_second": 1139.291,
"eval_steps_per_second": 0.571,
"step": 4200
},
{
"ep_loss": 0.0,
"epoch": 3.25,
"learning_rate": 0.0002105,
"loss": 0.5564,
"mlm_loss": 0.5564,
"step": 4210
},
{
"ep_loss": 0.0,
"epoch": 3.25,
"learning_rate": 0.000211,
"loss": 0.5511,
"mlm_loss": 0.5511,
"step": 4220
},
{
"ep_loss": 0.0,
"epoch": 3.26,
"learning_rate": 0.0002115,
"loss": 0.5578,
"mlm_loss": 0.5578,
"step": 4230
},
{
"ep_loss": 0.0,
"epoch": 3.27,
"learning_rate": 0.000212,
"loss": 0.5531,
"mlm_loss": 0.5531,
"step": 4240
},
{
"ep_loss": 0.0,
"epoch": 3.28,
"learning_rate": 0.0002125,
"loss": 0.5569,
"mlm_loss": 0.5569,
"step": 4250
},
{
"ep_loss": 0.0,
"epoch": 3.29,
"learning_rate": 0.000213,
"loss": 0.5499,
"mlm_loss": 0.5499,
"step": 4260
},
{
"ep_loss": 0.0,
"epoch": 3.29,
"learning_rate": 0.0002135,
"loss": 0.5514,
"mlm_loss": 0.5514,
"step": 4270
},
{
"ep_loss": 0.0,
"epoch": 3.3,
"learning_rate": 0.000214,
"loss": 0.554,
"mlm_loss": 0.554,
"step": 4280
},
{
"ep_loss": 0.0,
"epoch": 3.31,
"learning_rate": 0.0002145,
"loss": 0.5444,
"mlm_loss": 0.5444,
"step": 4290
},
{
"ep_loss": 0.0,
"epoch": 3.32,
"learning_rate": 0.000215,
"loss": 0.5555,
"mlm_loss": 0.5555,
"step": 4300
},
{
"epoch": 3.32,
"eval_ep_loss": -2.407238245010376,
"eval_loss": 0.5153276920318604,
"eval_mlm_loss": 0.5153276920318604,
"eval_runtime": 62.01,
"eval_samples_per_second": 1126.738,
"eval_steps_per_second": 0.564,
"step": 4300
},
{
"ep_loss": 0.0,
"epoch": 3.32,
"learning_rate": 0.0002155,
"loss": 0.5374,
"mlm_loss": 0.5374,
"step": 4310
},
{
"ep_loss": 0.0,
"epoch": 3.33,
"learning_rate": 0.000216,
"loss": 0.5526,
"mlm_loss": 0.5526,
"step": 4320
},
{
"ep_loss": 0.0,
"epoch": 3.34,
"learning_rate": 0.0002165,
"loss": 0.5451,
"mlm_loss": 0.5451,
"step": 4330
},
{
"ep_loss": 0.0,
"epoch": 3.35,
"learning_rate": 0.00021700000000000002,
"loss": 0.5406,
"mlm_loss": 0.5406,
"step": 4340
},
{
"ep_loss": 0.0,
"epoch": 3.36,
"learning_rate": 0.0002175,
"loss": 0.5339,
"mlm_loss": 0.5339,
"step": 4350
},
{
"ep_loss": 0.0,
"epoch": 3.36,
"learning_rate": 0.000218,
"loss": 0.5434,
"mlm_loss": 0.5434,
"step": 4360
},
{
"ep_loss": 0.0,
"epoch": 3.37,
"learning_rate": 0.0002185,
"loss": 0.546,
"mlm_loss": 0.546,
"step": 4370
},
{
"ep_loss": 0.0,
"epoch": 3.38,
"learning_rate": 0.000219,
"loss": 0.5359,
"mlm_loss": 0.5359,
"step": 4380
},
{
"ep_loss": 0.0,
"epoch": 3.39,
"learning_rate": 0.0002195,
"loss": 0.5405,
"mlm_loss": 0.5405,
"step": 4390
},
{
"ep_loss": 0.0,
"epoch": 3.39,
"learning_rate": 0.00022,
"loss": 0.5506,
"mlm_loss": 0.5506,
"step": 4400
},
{
"epoch": 3.39,
"eval_ep_loss": -2.1430230140686035,
"eval_loss": 0.5089389681816101,
"eval_mlm_loss": 0.5089389681816101,
"eval_runtime": 60.2829,
"eval_samples_per_second": 1159.018,
"eval_steps_per_second": 0.581,
"step": 4400
},
{
"ep_loss": 0.0,
"epoch": 3.4,
"learning_rate": 0.0002205,
"loss": 0.5425,
"mlm_loss": 0.5425,
"step": 4410
},
{
"ep_loss": 0.0,
"epoch": 3.41,
"learning_rate": 0.000221,
"loss": 0.5443,
"mlm_loss": 0.5443,
"step": 4420
},
{
"ep_loss": 0.0,
"epoch": 3.42,
"learning_rate": 0.00022150000000000002,
"loss": 0.5433,
"mlm_loss": 0.5433,
"step": 4430
},
{
"ep_loss": 0.0,
"epoch": 3.42,
"learning_rate": 0.000222,
"loss": 0.5379,
"mlm_loss": 0.5379,
"step": 4440
},
{
"ep_loss": 0.0,
"epoch": 3.43,
"learning_rate": 0.00022250000000000001,
"loss": 0.5349,
"mlm_loss": 0.5349,
"step": 4450
},
{
"ep_loss": 0.0,
"epoch": 3.44,
"learning_rate": 0.000223,
"loss": 0.5385,
"mlm_loss": 0.5385,
"step": 4460
},
{
"ep_loss": 0.0,
"epoch": 3.45,
"learning_rate": 0.0002235,
"loss": 0.5337,
"mlm_loss": 0.5337,
"step": 4470
},
{
"ep_loss": 0.0,
"epoch": 3.46,
"learning_rate": 0.000224,
"loss": 0.5372,
"mlm_loss": 0.5372,
"step": 4480
},
{
"ep_loss": 0.0,
"epoch": 3.46,
"learning_rate": 0.0002245,
"loss": 0.5342,
"mlm_loss": 0.5342,
"step": 4490
},
{
"ep_loss": 0.0,
"epoch": 3.47,
"learning_rate": 0.00022500000000000002,
"loss": 0.5329,
"mlm_loss": 0.5329,
"step": 4500
},
{
"epoch": 3.47,
"eval_ep_loss": -2.252265691757202,
"eval_loss": 0.501177966594696,
"eval_mlm_loss": 0.501177966594696,
"eval_runtime": 61.1269,
"eval_samples_per_second": 1143.016,
"eval_steps_per_second": 0.573,
"step": 4500
},
{
"ep_loss": 0.0,
"epoch": 3.48,
"learning_rate": 0.0002255,
"loss": 0.5321,
"mlm_loss": 0.5321,
"step": 4510
},
{
"ep_loss": 0.0,
"epoch": 3.49,
"learning_rate": 0.00022600000000000002,
"loss": 0.5413,
"mlm_loss": 0.5413,
"step": 4520
},
{
"ep_loss": 0.0,
"epoch": 3.49,
"learning_rate": 0.0002265,
"loss": 0.5364,
"mlm_loss": 0.5364,
"step": 4530
},
{
"ep_loss": 0.0,
"epoch": 3.5,
"learning_rate": 0.00022700000000000002,
"loss": 0.5326,
"mlm_loss": 0.5326,
"step": 4540
},
{
"ep_loss": 0.0,
"epoch": 3.51,
"learning_rate": 0.0002275,
"loss": 0.5327,
"mlm_loss": 0.5327,
"step": 4550
},
{
"ep_loss": 0.0,
"epoch": 3.52,
"learning_rate": 0.000228,
"loss": 0.5334,
"mlm_loss": 0.5334,
"step": 4560
},
{
"ep_loss": 0.0,
"epoch": 3.52,
"learning_rate": 0.00022850000000000002,
"loss": 0.5344,
"mlm_loss": 0.5344,
"step": 4570
},
{
"ep_loss": 0.0,
"epoch": 3.53,
"learning_rate": 0.000229,
"loss": 0.5259,
"mlm_loss": 0.5259,
"step": 4580
},
{
"ep_loss": 0.0,
"epoch": 3.54,
"learning_rate": 0.00022950000000000002,
"loss": 0.519,
"mlm_loss": 0.519,
"step": 4590
},
{
"ep_loss": 0.0,
"epoch": 3.55,
"learning_rate": 0.00023,
"loss": 0.5243,
"mlm_loss": 0.5243,
"step": 4600
},
{
"epoch": 3.55,
"eval_ep_loss": -2.4623284339904785,
"eval_loss": 0.49688172340393066,
"eval_mlm_loss": 0.49688172340393066,
"eval_runtime": 60.8838,
"eval_samples_per_second": 1147.579,
"eval_steps_per_second": 0.575,
"step": 4600
},
{
"ep_loss": 0.0,
"epoch": 3.56,
"learning_rate": 0.00023050000000000002,
"loss": 0.5237,
"mlm_loss": 0.5237,
"step": 4610
},
{
"ep_loss": 0.0,
"epoch": 3.56,
"learning_rate": 0.000231,
"loss": 0.5249,
"mlm_loss": 0.5249,
"step": 4620
},
{
"ep_loss": 0.0,
"epoch": 3.57,
"learning_rate": 0.00023150000000000002,
"loss": 0.5208,
"mlm_loss": 0.5208,
"step": 4630
},
{
"ep_loss": 0.0,
"epoch": 3.58,
"learning_rate": 0.00023200000000000003,
"loss": 0.5213,
"mlm_loss": 0.5213,
"step": 4640
},
{
"ep_loss": 0.0,
"epoch": 3.59,
"learning_rate": 0.0002325,
"loss": 0.5316,
"mlm_loss": 0.5316,
"step": 4650
},
{
"ep_loss": 0.0,
"epoch": 3.59,
"learning_rate": 0.00023300000000000003,
"loss": 0.5231,
"mlm_loss": 0.5231,
"step": 4660
},
{
"ep_loss": 0.0,
"epoch": 3.6,
"learning_rate": 0.0002335,
"loss": 0.524,
"mlm_loss": 0.524,
"step": 4670
},
{
"ep_loss": 0.0,
"epoch": 3.61,
"learning_rate": 0.00023400000000000002,
"loss": 0.5274,
"mlm_loss": 0.5274,
"step": 4680
},
{
"ep_loss": 0.0,
"epoch": 3.62,
"learning_rate": 0.00023449999999999998,
"loss": 0.5186,
"mlm_loss": 0.5186,
"step": 4690
},
{
"ep_loss": 0.0,
"epoch": 3.63,
"learning_rate": 0.000235,
"loss": 0.5201,
"mlm_loss": 0.5201,
"step": 4700
},
{
"epoch": 3.63,
"eval_ep_loss": -2.275129556655884,
"eval_loss": 0.48909062147140503,
"eval_mlm_loss": 0.48909062147140503,
"eval_runtime": 63.7661,
"eval_samples_per_second": 1095.708,
"eval_steps_per_second": 0.549,
"step": 4700
},
{
"ep_loss": 0.0,
"epoch": 3.63,
"learning_rate": 0.0002355,
"loss": 0.5167,
"mlm_loss": 0.5167,
"step": 4710
},
{
"ep_loss": 0.0,
"epoch": 3.64,
"learning_rate": 0.000236,
"loss": 0.5152,
"mlm_loss": 0.5152,
"step": 4720
},
{
"ep_loss": 0.0,
"epoch": 3.65,
"learning_rate": 0.0002365,
"loss": 0.5201,
"mlm_loss": 0.5201,
"step": 4730
},
{
"ep_loss": 0.0,
"epoch": 3.66,
"learning_rate": 0.000237,
"loss": 0.5234,
"mlm_loss": 0.5234,
"step": 4740
},
{
"ep_loss": 0.0,
"epoch": 3.66,
"learning_rate": 0.0002375,
"loss": 0.5156,
"mlm_loss": 0.5156,
"step": 4750
},
{
"ep_loss": 0.0,
"epoch": 3.67,
"learning_rate": 0.00023799999999999998,
"loss": 0.5189,
"mlm_loss": 0.5189,
"step": 4760
},
{
"ep_loss": 0.0,
"epoch": 3.68,
"learning_rate": 0.0002385,
"loss": 0.5189,
"mlm_loss": 0.5189,
"step": 4770
},
{
"ep_loss": 0.0,
"epoch": 3.69,
"learning_rate": 0.00023899999999999998,
"loss": 0.51,
"mlm_loss": 0.51,
"step": 4780
},
{
"ep_loss": 0.0,
"epoch": 3.69,
"learning_rate": 0.0002395,
"loss": 0.5158,
"mlm_loss": 0.5158,
"step": 4790
},
{
"ep_loss": 0.0,
"epoch": 3.7,
"learning_rate": 0.00024,
"loss": 0.5121,
"mlm_loss": 0.5121,
"step": 4800
},
{
"epoch": 3.7,
"eval_ep_loss": -2.0928449630737305,
"eval_loss": 0.48291251063346863,
"eval_mlm_loss": 0.48291251063346863,
"eval_runtime": 60.133,
"eval_samples_per_second": 1161.908,
"eval_steps_per_second": 0.582,
"step": 4800
},
{
"ep_loss": 0.0,
"epoch": 3.71,
"learning_rate": 0.0002405,
"loss": 0.5108,
"mlm_loss": 0.5108,
"step": 4810
},
{
"ep_loss": 0.0,
"epoch": 3.72,
"learning_rate": 0.000241,
"loss": 0.5176,
"mlm_loss": 0.5176,
"step": 4820
},
{
"ep_loss": 0.0,
"epoch": 3.73,
"learning_rate": 0.0002415,
"loss": 0.5102,
"mlm_loss": 0.5102,
"step": 4830
},
{
"ep_loss": 0.0,
"epoch": 3.73,
"learning_rate": 0.000242,
"loss": 0.514,
"mlm_loss": 0.514,
"step": 4840
},
{
"ep_loss": 0.0,
"epoch": 3.74,
"learning_rate": 0.00024249999999999999,
"loss": 0.5086,
"mlm_loss": 0.5086,
"step": 4850
},
{
"ep_loss": 0.0,
"epoch": 3.75,
"learning_rate": 0.000243,
"loss": 0.515,
"mlm_loss": 0.515,
"step": 4860
},
{
"ep_loss": 0.0,
"epoch": 3.76,
"learning_rate": 0.0002435,
"loss": 0.5091,
"mlm_loss": 0.5091,
"step": 4870
},
{
"ep_loss": 0.0,
"epoch": 3.76,
"learning_rate": 0.000244,
"loss": 0.514,
"mlm_loss": 0.514,
"step": 4880
},
{
"ep_loss": 0.0,
"epoch": 3.77,
"learning_rate": 0.0002445,
"loss": 0.5051,
"mlm_loss": 0.5051,
"step": 4890
},
{
"ep_loss": 0.0,
"epoch": 3.78,
"learning_rate": 0.000245,
"loss": 0.526,
"mlm_loss": 0.526,
"step": 4900
},
{
"epoch": 3.78,
"eval_ep_loss": -2.452838182449341,
"eval_loss": 0.4754766523838043,
"eval_mlm_loss": 0.4754766523838043,
"eval_runtime": 60.9895,
"eval_samples_per_second": 1145.591,
"eval_steps_per_second": 0.574,
"step": 4900
},
{
"ep_loss": 0.0,
"epoch": 3.79,
"learning_rate": 0.0002455,
"loss": 0.5104,
"mlm_loss": 0.5104,
"step": 4910
},
{
"ep_loss": 0.0,
"epoch": 3.79,
"learning_rate": 0.000246,
"loss": 0.4975,
"mlm_loss": 0.4975,
"step": 4920
},
{
"ep_loss": 0.0,
"epoch": 3.8,
"learning_rate": 0.00024650000000000003,
"loss": 0.511,
"mlm_loss": 0.511,
"step": 4930
},
{
"ep_loss": 0.0,
"epoch": 3.81,
"learning_rate": 0.000247,
"loss": 0.5081,
"mlm_loss": 0.5081,
"step": 4940
},
{
"ep_loss": 0.0,
"epoch": 3.82,
"learning_rate": 0.0002475,
"loss": 0.5067,
"mlm_loss": 0.5067,
"step": 4950
},
{
"ep_loss": 0.0,
"epoch": 3.83,
"learning_rate": 0.000248,
"loss": 0.5062,
"mlm_loss": 0.5062,
"step": 4960
},
{
"ep_loss": 0.0,
"epoch": 3.83,
"learning_rate": 0.0002485,
"loss": 0.5054,
"mlm_loss": 0.5054,
"step": 4970
},
{
"ep_loss": 0.0,
"epoch": 3.84,
"learning_rate": 0.000249,
"loss": 0.5032,
"mlm_loss": 0.5032,
"step": 4980
},
{
"ep_loss": 0.0,
"epoch": 3.85,
"learning_rate": 0.0002495,
"loss": 0.5022,
"mlm_loss": 0.5022,
"step": 4990
},
{
"ep_loss": 0.0,
"epoch": 3.86,
"learning_rate": 0.00025,
"loss": 0.5077,
"mlm_loss": 0.5077,
"step": 5000
},
{
"epoch": 3.86,
"eval_ep_loss": -2.5484063625335693,
"eval_loss": 0.47381114959716797,
"eval_mlm_loss": 0.47381114959716797,
"eval_runtime": 59.7879,
"eval_samples_per_second": 1168.615,
"eval_steps_per_second": 0.585,
"step": 5000
}
],
"max_steps": 25920,
"num_train_epochs": 20,
"total_flos": 1.356897877753856e+18,
"trial_name": null,
"trial_params": null
}