hwm21's picture
Training in progress, step 10500, checkpoint
eab3ea0 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 53.84615384615385,
"eval_steps": 100,
"global_step": 10500,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.5128205128205128,
"grad_norm": 73.68424224853516,
"learning_rate": 9.9907e-06,
"loss": 3.1457,
"step": 100
},
{
"epoch": 0.5128205128205128,
"eval_loss": 2.2357213497161865,
"eval_runtime": 36.4689,
"eval_samples_per_second": 10.776,
"eval_steps_per_second": 1.371,
"step": 100
},
{
"epoch": 1.0256410256410255,
"grad_norm": 57.47202682495117,
"learning_rate": 9.980800000000001e-06,
"loss": 2.1614,
"step": 200
},
{
"epoch": 1.0256410256410255,
"eval_loss": 2.0913825035095215,
"eval_runtime": 36.349,
"eval_samples_per_second": 10.812,
"eval_steps_per_second": 1.376,
"step": 200
},
{
"epoch": 1.5384615384615383,
"grad_norm": 176.88357543945312,
"learning_rate": 9.970800000000001e-06,
"loss": 2.0388,
"step": 300
},
{
"epoch": 1.5384615384615383,
"eval_loss": 2.003911018371582,
"eval_runtime": 36.2551,
"eval_samples_per_second": 10.84,
"eval_steps_per_second": 1.379,
"step": 300
},
{
"epoch": 2.051282051282051,
"grad_norm": 55.31132507324219,
"learning_rate": 9.960800000000001e-06,
"loss": 1.9285,
"step": 400
},
{
"epoch": 2.051282051282051,
"eval_loss": 1.9796075820922852,
"eval_runtime": 36.4437,
"eval_samples_per_second": 10.784,
"eval_steps_per_second": 1.372,
"step": 400
},
{
"epoch": 2.564102564102564,
"grad_norm": 41.900753021240234,
"learning_rate": 9.9508e-06,
"loss": 1.9523,
"step": 500
},
{
"epoch": 2.564102564102564,
"eval_loss": 1.936122179031372,
"eval_runtime": 36.5845,
"eval_samples_per_second": 10.742,
"eval_steps_per_second": 1.367,
"step": 500
},
{
"epoch": 3.076923076923077,
"grad_norm": 50.21903991699219,
"learning_rate": 9.9408e-06,
"loss": 1.8452,
"step": 600
},
{
"epoch": 3.076923076923077,
"eval_loss": 1.8883634805679321,
"eval_runtime": 36.6015,
"eval_samples_per_second": 10.737,
"eval_steps_per_second": 1.366,
"step": 600
},
{
"epoch": 3.58974358974359,
"grad_norm": 45.193939208984375,
"learning_rate": 9.930900000000002e-06,
"loss": 1.8403,
"step": 700
},
{
"epoch": 3.58974358974359,
"eval_loss": 1.8506474494934082,
"eval_runtime": 36.454,
"eval_samples_per_second": 10.781,
"eval_steps_per_second": 1.372,
"step": 700
},
{
"epoch": 4.102564102564102,
"grad_norm": 27.302494049072266,
"learning_rate": 9.920900000000002e-06,
"loss": 1.7976,
"step": 800
},
{
"epoch": 4.102564102564102,
"eval_loss": 1.8370662927627563,
"eval_runtime": 36.834,
"eval_samples_per_second": 10.67,
"eval_steps_per_second": 1.357,
"step": 800
},
{
"epoch": 4.615384615384615,
"grad_norm": 52.6607666015625,
"learning_rate": 9.9109e-06,
"loss": 1.7508,
"step": 900
},
{
"epoch": 4.615384615384615,
"eval_loss": 1.8037244081497192,
"eval_runtime": 36.9939,
"eval_samples_per_second": 10.623,
"eval_steps_per_second": 1.352,
"step": 900
},
{
"epoch": 5.128205128205128,
"grad_norm": 59.508033752441406,
"learning_rate": 9.9009e-06,
"loss": 1.7383,
"step": 1000
},
{
"epoch": 5.128205128205128,
"eval_loss": 1.7986633777618408,
"eval_runtime": 36.8356,
"eval_samples_per_second": 10.669,
"eval_steps_per_second": 1.357,
"step": 1000
},
{
"epoch": 5.641025641025641,
"grad_norm": 71.58872985839844,
"learning_rate": 9.8909e-06,
"loss": 1.7361,
"step": 1100
},
{
"epoch": 5.641025641025641,
"eval_loss": 1.7810852527618408,
"eval_runtime": 37.0957,
"eval_samples_per_second": 10.594,
"eval_steps_per_second": 1.348,
"step": 1100
},
{
"epoch": 6.153846153846154,
"grad_norm": 41.782066345214844,
"learning_rate": 9.8809e-06,
"loss": 1.682,
"step": 1200
},
{
"epoch": 6.153846153846154,
"eval_loss": 1.777554988861084,
"eval_runtime": 36.9173,
"eval_samples_per_second": 10.645,
"eval_steps_per_second": 1.354,
"step": 1200
},
{
"epoch": 6.666666666666667,
"grad_norm": 40.28728485107422,
"learning_rate": 9.8709e-06,
"loss": 1.7216,
"step": 1300
},
{
"epoch": 6.666666666666667,
"eval_loss": 1.7382572889328003,
"eval_runtime": 36.8918,
"eval_samples_per_second": 10.653,
"eval_steps_per_second": 1.355,
"step": 1300
},
{
"epoch": 7.17948717948718,
"grad_norm": 104.99211883544922,
"learning_rate": 9.8609e-06,
"loss": 1.6534,
"step": 1400
},
{
"epoch": 7.17948717948718,
"eval_loss": 1.76559579372406,
"eval_runtime": 36.3398,
"eval_samples_per_second": 10.815,
"eval_steps_per_second": 1.376,
"step": 1400
},
{
"epoch": 7.6923076923076925,
"grad_norm": 29.326631546020508,
"learning_rate": 9.8509e-06,
"loss": 1.707,
"step": 1500
},
{
"epoch": 7.6923076923076925,
"eval_loss": 1.750089406967163,
"eval_runtime": 36.4098,
"eval_samples_per_second": 10.794,
"eval_steps_per_second": 1.373,
"step": 1500
},
{
"epoch": 8.205128205128204,
"grad_norm": 37.941261291503906,
"learning_rate": 9.840900000000001e-06,
"loss": 1.6554,
"step": 1600
},
{
"epoch": 8.205128205128204,
"eval_loss": 1.6900651454925537,
"eval_runtime": 36.3226,
"eval_samples_per_second": 10.82,
"eval_steps_per_second": 1.377,
"step": 1600
},
{
"epoch": 8.717948717948717,
"grad_norm": 44.60703659057617,
"learning_rate": 9.830900000000001e-06,
"loss": 1.6334,
"step": 1700
},
{
"epoch": 8.717948717948717,
"eval_loss": 1.7162973880767822,
"eval_runtime": 36.2672,
"eval_samples_per_second": 10.836,
"eval_steps_per_second": 1.379,
"step": 1700
},
{
"epoch": 9.23076923076923,
"grad_norm": 34.127254486083984,
"learning_rate": 9.820900000000001e-06,
"loss": 1.6345,
"step": 1800
},
{
"epoch": 9.23076923076923,
"eval_loss": 1.6906001567840576,
"eval_runtime": 36.2264,
"eval_samples_per_second": 10.848,
"eval_steps_per_second": 1.38,
"step": 1800
},
{
"epoch": 9.743589743589745,
"grad_norm": 60.377540588378906,
"learning_rate": 9.810900000000001e-06,
"loss": 1.598,
"step": 1900
},
{
"epoch": 9.743589743589745,
"eval_loss": 1.6555503606796265,
"eval_runtime": 36.3896,
"eval_samples_per_second": 10.8,
"eval_steps_per_second": 1.374,
"step": 1900
},
{
"epoch": 10.256410256410255,
"grad_norm": 20.264404296875,
"learning_rate": 9.800900000000001e-06,
"loss": 1.5466,
"step": 2000
},
{
"epoch": 10.256410256410255,
"eval_loss": 1.648037075996399,
"eval_runtime": 36.4136,
"eval_samples_per_second": 10.793,
"eval_steps_per_second": 1.373,
"step": 2000
},
{
"epoch": 10.76923076923077,
"grad_norm": 27.18608856201172,
"learning_rate": 9.790900000000001e-06,
"loss": 1.5865,
"step": 2100
},
{
"epoch": 10.76923076923077,
"eval_loss": 1.6171936988830566,
"eval_runtime": 36.1051,
"eval_samples_per_second": 10.885,
"eval_steps_per_second": 1.385,
"step": 2100
},
{
"epoch": 11.282051282051283,
"grad_norm": 32.486331939697266,
"learning_rate": 9.780900000000002e-06,
"loss": 1.5284,
"step": 2200
},
{
"epoch": 11.282051282051283,
"eval_loss": 1.5915095806121826,
"eval_runtime": 36.1781,
"eval_samples_per_second": 10.863,
"eval_steps_per_second": 1.382,
"step": 2200
},
{
"epoch": 11.794871794871796,
"grad_norm": 65.88719940185547,
"learning_rate": 9.770900000000002e-06,
"loss": 1.5514,
"step": 2300
},
{
"epoch": 11.794871794871796,
"eval_loss": 1.5879931449890137,
"eval_runtime": 36.3934,
"eval_samples_per_second": 10.799,
"eval_steps_per_second": 1.374,
"step": 2300
},
{
"epoch": 12.307692307692308,
"grad_norm": 31.737024307250977,
"learning_rate": 9.760900000000002e-06,
"loss": 1.4941,
"step": 2400
},
{
"epoch": 12.307692307692308,
"eval_loss": 1.583853006362915,
"eval_runtime": 36.3797,
"eval_samples_per_second": 10.803,
"eval_steps_per_second": 1.374,
"step": 2400
},
{
"epoch": 12.820512820512821,
"grad_norm": 45.48268508911133,
"learning_rate": 9.7509e-06,
"loss": 1.5097,
"step": 2500
},
{
"epoch": 12.820512820512821,
"eval_loss": 1.5559026002883911,
"eval_runtime": 36.1605,
"eval_samples_per_second": 10.868,
"eval_steps_per_second": 1.383,
"step": 2500
},
{
"epoch": 13.333333333333334,
"grad_norm": 27.500398635864258,
"learning_rate": 9.7409e-06,
"loss": 1.5018,
"step": 2600
},
{
"epoch": 13.333333333333334,
"eval_loss": 1.5453521013259888,
"eval_runtime": 36.2887,
"eval_samples_per_second": 10.83,
"eval_steps_per_second": 1.378,
"step": 2600
},
{
"epoch": 13.846153846153847,
"grad_norm": 32.49728775024414,
"learning_rate": 9.7309e-06,
"loss": 1.4804,
"step": 2700
},
{
"epoch": 13.846153846153847,
"eval_loss": 1.5424816608428955,
"eval_runtime": 36.359,
"eval_samples_per_second": 10.809,
"eval_steps_per_second": 1.375,
"step": 2700
},
{
"epoch": 14.35897435897436,
"grad_norm": 38.46280288696289,
"learning_rate": 9.7209e-06,
"loss": 1.4826,
"step": 2800
},
{
"epoch": 14.35897435897436,
"eval_loss": 1.5317177772521973,
"eval_runtime": 36.3362,
"eval_samples_per_second": 10.816,
"eval_steps_per_second": 1.376,
"step": 2800
},
{
"epoch": 14.871794871794872,
"grad_norm": 16.075960159301758,
"learning_rate": 9.7109e-06,
"loss": 1.4568,
"step": 2900
},
{
"epoch": 14.871794871794872,
"eval_loss": 1.5241832733154297,
"eval_runtime": 36.2025,
"eval_samples_per_second": 10.856,
"eval_steps_per_second": 1.381,
"step": 2900
},
{
"epoch": 15.384615384615385,
"grad_norm": 27.334318161010742,
"learning_rate": 9.7009e-06,
"loss": 1.4176,
"step": 3000
},
{
"epoch": 15.384615384615385,
"eval_loss": 1.520580768585205,
"eval_runtime": 36.398,
"eval_samples_per_second": 10.797,
"eval_steps_per_second": 1.374,
"step": 3000
},
{
"epoch": 15.897435897435898,
"grad_norm": 94.90784454345703,
"learning_rate": 9.6909e-06,
"loss": 1.4681,
"step": 3100
},
{
"epoch": 15.897435897435898,
"eval_loss": 1.5268648862838745,
"eval_runtime": 36.0541,
"eval_samples_per_second": 10.9,
"eval_steps_per_second": 1.387,
"step": 3100
},
{
"epoch": 16.41025641025641,
"grad_norm": 16.697856903076172,
"learning_rate": 9.6809e-06,
"loss": 1.454,
"step": 3200
},
{
"epoch": 16.41025641025641,
"eval_loss": 1.5157753229141235,
"eval_runtime": 36.3172,
"eval_samples_per_second": 10.821,
"eval_steps_per_second": 1.377,
"step": 3200
},
{
"epoch": 16.923076923076923,
"grad_norm": 54.05553436279297,
"learning_rate": 9.670900000000001e-06,
"loss": 1.4309,
"step": 3300
},
{
"epoch": 16.923076923076923,
"eval_loss": 1.516249179840088,
"eval_runtime": 36.2632,
"eval_samples_per_second": 10.837,
"eval_steps_per_second": 1.379,
"step": 3300
},
{
"epoch": 17.435897435897434,
"grad_norm": 47.010475158691406,
"learning_rate": 9.660900000000001e-06,
"loss": 1.4571,
"step": 3400
},
{
"epoch": 17.435897435897434,
"eval_loss": 1.5018247365951538,
"eval_runtime": 36.2118,
"eval_samples_per_second": 10.853,
"eval_steps_per_second": 1.381,
"step": 3400
},
{
"epoch": 17.94871794871795,
"grad_norm": 52.865718841552734,
"learning_rate": 9.650900000000001e-06,
"loss": 1.4168,
"step": 3500
},
{
"epoch": 17.94871794871795,
"eval_loss": 1.4993616342544556,
"eval_runtime": 36.2572,
"eval_samples_per_second": 10.839,
"eval_steps_per_second": 1.379,
"step": 3500
},
{
"epoch": 18.46153846153846,
"grad_norm": 30.117380142211914,
"learning_rate": 9.640900000000001e-06,
"loss": 1.4275,
"step": 3600
},
{
"epoch": 18.46153846153846,
"eval_loss": 1.4899998903274536,
"eval_runtime": 36.4696,
"eval_samples_per_second": 10.776,
"eval_steps_per_second": 1.371,
"step": 3600
},
{
"epoch": 18.974358974358974,
"grad_norm": 31.10028076171875,
"learning_rate": 9.630900000000001e-06,
"loss": 1.4148,
"step": 3700
},
{
"epoch": 18.974358974358974,
"eval_loss": 1.5231629610061646,
"eval_runtime": 36.4604,
"eval_samples_per_second": 10.779,
"eval_steps_per_second": 1.371,
"step": 3700
},
{
"epoch": 19.487179487179485,
"grad_norm": 44.06697082519531,
"learning_rate": 9.620900000000001e-06,
"loss": 1.4057,
"step": 3800
},
{
"epoch": 19.487179487179485,
"eval_loss": 1.4841217994689941,
"eval_runtime": 36.3382,
"eval_samples_per_second": 10.815,
"eval_steps_per_second": 1.376,
"step": 3800
},
{
"epoch": 20.0,
"grad_norm": 53.86429214477539,
"learning_rate": 9.610900000000001e-06,
"loss": 1.4302,
"step": 3900
},
{
"epoch": 20.0,
"eval_loss": 1.477772831916809,
"eval_runtime": 36.1794,
"eval_samples_per_second": 10.863,
"eval_steps_per_second": 1.382,
"step": 3900
},
{
"epoch": 20.51282051282051,
"grad_norm": 80.95457458496094,
"learning_rate": 9.600900000000002e-06,
"loss": 1.4076,
"step": 4000
},
{
"epoch": 20.51282051282051,
"eval_loss": 1.4769134521484375,
"eval_runtime": 36.4725,
"eval_samples_per_second": 10.775,
"eval_steps_per_second": 1.371,
"step": 4000
},
{
"epoch": 21.025641025641026,
"grad_norm": 32.276214599609375,
"learning_rate": 9.5909e-06,
"loss": 1.3868,
"step": 4100
},
{
"epoch": 21.025641025641026,
"eval_loss": 1.463292121887207,
"eval_runtime": 36.5192,
"eval_samples_per_second": 10.761,
"eval_steps_per_second": 1.369,
"step": 4100
},
{
"epoch": 21.53846153846154,
"grad_norm": 54.65959167480469,
"learning_rate": 9.5809e-06,
"loss": 1.3795,
"step": 4200
},
{
"epoch": 21.53846153846154,
"eval_loss": 1.4630039930343628,
"eval_runtime": 36.5288,
"eval_samples_per_second": 10.759,
"eval_steps_per_second": 1.369,
"step": 4200
},
{
"epoch": 22.05128205128205,
"grad_norm": 42.31818389892578,
"learning_rate": 9.5709e-06,
"loss": 1.3787,
"step": 4300
},
{
"epoch": 22.05128205128205,
"eval_loss": 1.4471133947372437,
"eval_runtime": 36.2949,
"eval_samples_per_second": 10.828,
"eval_steps_per_second": 1.378,
"step": 4300
},
{
"epoch": 22.564102564102566,
"grad_norm": 34.44257736206055,
"learning_rate": 9.5609e-06,
"loss": 1.4027,
"step": 4400
},
{
"epoch": 22.564102564102566,
"eval_loss": 1.4606964588165283,
"eval_runtime": 36.5672,
"eval_samples_per_second": 10.747,
"eval_steps_per_second": 1.367,
"step": 4400
},
{
"epoch": 23.076923076923077,
"grad_norm": 42.65989303588867,
"learning_rate": 9.5509e-06,
"loss": 1.3459,
"step": 4500
},
{
"epoch": 23.076923076923077,
"eval_loss": 1.454709768295288,
"eval_runtime": 37.2024,
"eval_samples_per_second": 10.564,
"eval_steps_per_second": 1.344,
"step": 4500
},
{
"epoch": 23.58974358974359,
"grad_norm": 35.11396789550781,
"learning_rate": 9.5409e-06,
"loss": 1.3367,
"step": 4600
},
{
"epoch": 23.58974358974359,
"eval_loss": 1.4562979936599731,
"eval_runtime": 36.4579,
"eval_samples_per_second": 10.78,
"eval_steps_per_second": 1.371,
"step": 4600
},
{
"epoch": 24.102564102564102,
"grad_norm": 32.71805953979492,
"learning_rate": 9.5309e-06,
"loss": 1.3575,
"step": 4700
},
{
"epoch": 24.102564102564102,
"eval_loss": 1.4620144367218018,
"eval_runtime": 36.4055,
"eval_samples_per_second": 10.795,
"eval_steps_per_second": 1.373,
"step": 4700
},
{
"epoch": 24.615384615384617,
"grad_norm": 28.839757919311523,
"learning_rate": 9.5209e-06,
"loss": 1.3549,
"step": 4800
},
{
"epoch": 24.615384615384617,
"eval_loss": 1.4431304931640625,
"eval_runtime": 36.5027,
"eval_samples_per_second": 10.766,
"eval_steps_per_second": 1.37,
"step": 4800
},
{
"epoch": 25.128205128205128,
"grad_norm": 45.3994140625,
"learning_rate": 9.5109e-06,
"loss": 1.3885,
"step": 4900
},
{
"epoch": 25.128205128205128,
"eval_loss": 1.4312200546264648,
"eval_runtime": 36.3039,
"eval_samples_per_second": 10.825,
"eval_steps_per_second": 1.377,
"step": 4900
},
{
"epoch": 25.641025641025642,
"grad_norm": 22.972829818725586,
"learning_rate": 9.5009e-06,
"loss": 1.3469,
"step": 5000
},
{
"epoch": 25.641025641025642,
"eval_loss": 1.416171669960022,
"eval_runtime": 36.6695,
"eval_samples_per_second": 10.717,
"eval_steps_per_second": 1.364,
"step": 5000
},
{
"epoch": 26.153846153846153,
"grad_norm": 77.40106964111328,
"learning_rate": 9.490900000000001e-06,
"loss": 1.3363,
"step": 5100
},
{
"epoch": 26.153846153846153,
"eval_loss": 1.4090278148651123,
"eval_runtime": 36.2573,
"eval_samples_per_second": 10.839,
"eval_steps_per_second": 1.379,
"step": 5100
},
{
"epoch": 26.666666666666668,
"grad_norm": 29.757932662963867,
"learning_rate": 9.480900000000001e-06,
"loss": 1.3183,
"step": 5200
},
{
"epoch": 26.666666666666668,
"eval_loss": 1.4073749780654907,
"eval_runtime": 36.2439,
"eval_samples_per_second": 10.843,
"eval_steps_per_second": 1.38,
"step": 5200
},
{
"epoch": 27.17948717948718,
"grad_norm": 56.78797149658203,
"learning_rate": 9.470900000000001e-06,
"loss": 1.3568,
"step": 5300
},
{
"epoch": 27.17948717948718,
"eval_loss": 1.4153741598129272,
"eval_runtime": 36.2756,
"eval_samples_per_second": 10.834,
"eval_steps_per_second": 1.378,
"step": 5300
},
{
"epoch": 27.692307692307693,
"grad_norm": 62.353477478027344,
"learning_rate": 9.460900000000001e-06,
"loss": 1.3304,
"step": 5400
},
{
"epoch": 27.692307692307693,
"eval_loss": 1.4334921836853027,
"eval_runtime": 36.2626,
"eval_samples_per_second": 10.838,
"eval_steps_per_second": 1.379,
"step": 5400
},
{
"epoch": 28.205128205128204,
"grad_norm": 100.7852554321289,
"learning_rate": 9.450900000000001e-06,
"loss": 1.2897,
"step": 5500
},
{
"epoch": 28.205128205128204,
"eval_loss": 1.4160270690917969,
"eval_runtime": 36.6139,
"eval_samples_per_second": 10.734,
"eval_steps_per_second": 1.366,
"step": 5500
},
{
"epoch": 28.71794871794872,
"grad_norm": 62.06657409667969,
"learning_rate": 9.440900000000001e-06,
"loss": 1.3233,
"step": 5600
},
{
"epoch": 28.71794871794872,
"eval_loss": 1.431317687034607,
"eval_runtime": 36.5341,
"eval_samples_per_second": 10.757,
"eval_steps_per_second": 1.369,
"step": 5600
},
{
"epoch": 29.23076923076923,
"grad_norm": 32.661346435546875,
"learning_rate": 9.4309e-06,
"loss": 1.305,
"step": 5700
},
{
"epoch": 29.23076923076923,
"eval_loss": 1.3954827785491943,
"eval_runtime": 36.5903,
"eval_samples_per_second": 10.741,
"eval_steps_per_second": 1.366,
"step": 5700
},
{
"epoch": 29.743589743589745,
"grad_norm": 25.690454483032227,
"learning_rate": 9.421000000000002e-06,
"loss": 1.2961,
"step": 5800
},
{
"epoch": 29.743589743589745,
"eval_loss": 1.4036046266555786,
"eval_runtime": 36.5935,
"eval_samples_per_second": 10.74,
"eval_steps_per_second": 1.366,
"step": 5800
},
{
"epoch": 30.256410256410255,
"grad_norm": 45.45426940917969,
"learning_rate": 9.411000000000002e-06,
"loss": 1.3175,
"step": 5900
},
{
"epoch": 30.256410256410255,
"eval_loss": 1.3845292329788208,
"eval_runtime": 36.5594,
"eval_samples_per_second": 10.75,
"eval_steps_per_second": 1.368,
"step": 5900
},
{
"epoch": 30.76923076923077,
"grad_norm": 41.62439727783203,
"learning_rate": 9.401000000000002e-06,
"loss": 1.3242,
"step": 6000
},
{
"epoch": 30.76923076923077,
"eval_loss": 1.3939634561538696,
"eval_runtime": 36.4895,
"eval_samples_per_second": 10.77,
"eval_steps_per_second": 1.37,
"step": 6000
},
{
"epoch": 31.28205128205128,
"grad_norm": 26.999319076538086,
"learning_rate": 9.391e-06,
"loss": 1.2886,
"step": 6100
},
{
"epoch": 31.28205128205128,
"eval_loss": 1.3804558515548706,
"eval_runtime": 36.2599,
"eval_samples_per_second": 10.838,
"eval_steps_per_second": 1.379,
"step": 6100
},
{
"epoch": 31.794871794871796,
"grad_norm": 24.70287322998047,
"learning_rate": 9.381e-06,
"loss": 1.2893,
"step": 6200
},
{
"epoch": 31.794871794871796,
"eval_loss": 1.3821990489959717,
"eval_runtime": 36.2613,
"eval_samples_per_second": 10.838,
"eval_steps_per_second": 1.379,
"step": 6200
},
{
"epoch": 32.30769230769231,
"grad_norm": 41.910606384277344,
"learning_rate": 9.371e-06,
"loss": 1.3093,
"step": 6300
},
{
"epoch": 32.30769230769231,
"eval_loss": 1.3875064849853516,
"eval_runtime": 36.4998,
"eval_samples_per_second": 10.767,
"eval_steps_per_second": 1.37,
"step": 6300
},
{
"epoch": 32.82051282051282,
"grad_norm": 82.20216369628906,
"learning_rate": 9.361e-06,
"loss": 1.3184,
"step": 6400
},
{
"epoch": 32.82051282051282,
"eval_loss": 1.3840612173080444,
"eval_runtime": 36.4787,
"eval_samples_per_second": 10.773,
"eval_steps_per_second": 1.371,
"step": 6400
},
{
"epoch": 33.333333333333336,
"grad_norm": 797.53271484375,
"learning_rate": 9.351e-06,
"loss": 1.2939,
"step": 6500
},
{
"epoch": 33.333333333333336,
"eval_loss": 1.3881950378417969,
"eval_runtime": 36.4166,
"eval_samples_per_second": 10.792,
"eval_steps_per_second": 1.373,
"step": 6500
},
{
"epoch": 33.84615384615385,
"grad_norm": 35.14693069458008,
"learning_rate": 9.341000000000001e-06,
"loss": 1.2881,
"step": 6600
},
{
"epoch": 33.84615384615385,
"eval_loss": 1.4039666652679443,
"eval_runtime": 36.4141,
"eval_samples_per_second": 10.793,
"eval_steps_per_second": 1.373,
"step": 6600
},
{
"epoch": 34.35897435897436,
"grad_norm": 38.676666259765625,
"learning_rate": 9.331000000000001e-06,
"loss": 1.2699,
"step": 6700
},
{
"epoch": 34.35897435897436,
"eval_loss": 1.3800386190414429,
"eval_runtime": 36.2367,
"eval_samples_per_second": 10.845,
"eval_steps_per_second": 1.38,
"step": 6700
},
{
"epoch": 34.87179487179487,
"grad_norm": 26.051170349121094,
"learning_rate": 9.321000000000001e-06,
"loss": 1.3079,
"step": 6800
},
{
"epoch": 34.87179487179487,
"eval_loss": 1.3784704208374023,
"eval_runtime": 36.3119,
"eval_samples_per_second": 10.823,
"eval_steps_per_second": 1.377,
"step": 6800
},
{
"epoch": 35.38461538461539,
"grad_norm": 42.667236328125,
"learning_rate": 9.311000000000001e-06,
"loss": 1.2622,
"step": 6900
},
{
"epoch": 35.38461538461539,
"eval_loss": 1.3637058734893799,
"eval_runtime": 36.1116,
"eval_samples_per_second": 10.883,
"eval_steps_per_second": 1.385,
"step": 6900
},
{
"epoch": 35.8974358974359,
"grad_norm": 38.78388977050781,
"learning_rate": 9.301000000000001e-06,
"loss": 1.2652,
"step": 7000
},
{
"epoch": 35.8974358974359,
"eval_loss": 1.3452589511871338,
"eval_runtime": 36.2593,
"eval_samples_per_second": 10.839,
"eval_steps_per_second": 1.379,
"step": 7000
},
{
"epoch": 36.41025641025641,
"grad_norm": 44.43967056274414,
"learning_rate": 9.291000000000001e-06,
"loss": 1.2378,
"step": 7100
},
{
"epoch": 36.41025641025641,
"eval_loss": 1.3494073152542114,
"eval_runtime": 35.4402,
"eval_samples_per_second": 11.089,
"eval_steps_per_second": 1.411,
"step": 7100
},
{
"epoch": 36.92307692307692,
"grad_norm": 49.02131271362305,
"learning_rate": 9.281000000000001e-06,
"loss": 1.2932,
"step": 7200
},
{
"epoch": 36.92307692307692,
"eval_loss": 1.3460158109664917,
"eval_runtime": 35.9361,
"eval_samples_per_second": 10.936,
"eval_steps_per_second": 1.391,
"step": 7200
},
{
"epoch": 37.43589743589744,
"grad_norm": 28.279098510742188,
"learning_rate": 9.271000000000002e-06,
"loss": 1.2598,
"step": 7300
},
{
"epoch": 37.43589743589744,
"eval_loss": 1.36253023147583,
"eval_runtime": 36.2944,
"eval_samples_per_second": 10.828,
"eval_steps_per_second": 1.378,
"step": 7300
},
{
"epoch": 37.94871794871795,
"grad_norm": 35.21017074584961,
"learning_rate": 9.261000000000002e-06,
"loss": 1.2703,
"step": 7400
},
{
"epoch": 37.94871794871795,
"eval_loss": 1.3509865999221802,
"eval_runtime": 36.2661,
"eval_samples_per_second": 10.837,
"eval_steps_per_second": 1.379,
"step": 7400
},
{
"epoch": 38.46153846153846,
"grad_norm": 51.673316955566406,
"learning_rate": 9.251000000000002e-06,
"loss": 1.2393,
"step": 7500
},
{
"epoch": 38.46153846153846,
"eval_loss": 1.3402855396270752,
"eval_runtime": 36.392,
"eval_samples_per_second": 10.799,
"eval_steps_per_second": 1.374,
"step": 7500
},
{
"epoch": 38.97435897435897,
"grad_norm": 53.73936462402344,
"learning_rate": 9.241000000000002e-06,
"loss": 1.2577,
"step": 7600
},
{
"epoch": 38.97435897435897,
"eval_loss": 1.3487578630447388,
"eval_runtime": 36.2119,
"eval_samples_per_second": 10.853,
"eval_steps_per_second": 1.381,
"step": 7600
},
{
"epoch": 39.48717948717949,
"grad_norm": 55.994686126708984,
"learning_rate": 9.231000000000002e-06,
"loss": 1.229,
"step": 7700
},
{
"epoch": 39.48717948717949,
"eval_loss": 1.340031623840332,
"eval_runtime": 36.2063,
"eval_samples_per_second": 10.854,
"eval_steps_per_second": 1.381,
"step": 7700
},
{
"epoch": 40.0,
"grad_norm": 86.7531509399414,
"learning_rate": 9.221e-06,
"loss": 1.2941,
"step": 7800
},
{
"epoch": 40.0,
"eval_loss": 1.3422337770462036,
"eval_runtime": 36.7462,
"eval_samples_per_second": 10.695,
"eval_steps_per_second": 1.361,
"step": 7800
},
{
"epoch": 40.51282051282051,
"grad_norm": 60.86371612548828,
"learning_rate": 9.211e-06,
"loss": 1.2423,
"step": 7900
},
{
"epoch": 40.51282051282051,
"eval_loss": 1.3336257934570312,
"eval_runtime": 36.2441,
"eval_samples_per_second": 10.843,
"eval_steps_per_second": 1.38,
"step": 7900
},
{
"epoch": 41.02564102564103,
"grad_norm": 28.535411834716797,
"learning_rate": 9.2011e-06,
"loss": 1.2676,
"step": 8000
},
{
"epoch": 41.02564102564103,
"eval_loss": 1.338461995124817,
"eval_runtime": 36.2212,
"eval_samples_per_second": 10.85,
"eval_steps_per_second": 1.38,
"step": 8000
},
{
"epoch": 41.53846153846154,
"grad_norm": 35.707183837890625,
"learning_rate": 9.1911e-06,
"loss": 1.2428,
"step": 8100
},
{
"epoch": 41.53846153846154,
"eval_loss": 1.3225666284561157,
"eval_runtime": 36.2043,
"eval_samples_per_second": 10.855,
"eval_steps_per_second": 1.381,
"step": 8100
},
{
"epoch": 42.05128205128205,
"grad_norm": 29.23111343383789,
"learning_rate": 9.181100000000001e-06,
"loss": 1.2269,
"step": 8200
},
{
"epoch": 42.05128205128205,
"eval_loss": 1.3405094146728516,
"eval_runtime": 36.2498,
"eval_samples_per_second": 10.841,
"eval_steps_per_second": 1.379,
"step": 8200
},
{
"epoch": 42.56410256410256,
"grad_norm": 20.379304885864258,
"learning_rate": 9.171100000000001e-06,
"loss": 1.2187,
"step": 8300
},
{
"epoch": 42.56410256410256,
"eval_loss": 1.3247309923171997,
"eval_runtime": 36.3237,
"eval_samples_per_second": 10.819,
"eval_steps_per_second": 1.377,
"step": 8300
},
{
"epoch": 43.07692307692308,
"grad_norm": 44.43791198730469,
"learning_rate": 9.161100000000001e-06,
"loss": 1.2321,
"step": 8400
},
{
"epoch": 43.07692307692308,
"eval_loss": 1.334086298942566,
"eval_runtime": 36.2039,
"eval_samples_per_second": 10.855,
"eval_steps_per_second": 1.381,
"step": 8400
},
{
"epoch": 43.58974358974359,
"grad_norm": 30.97890853881836,
"learning_rate": 9.151100000000001e-06,
"loss": 1.2071,
"step": 8500
},
{
"epoch": 43.58974358974359,
"eval_loss": 1.3306576013565063,
"eval_runtime": 36.3547,
"eval_samples_per_second": 10.81,
"eval_steps_per_second": 1.375,
"step": 8500
},
{
"epoch": 44.1025641025641,
"grad_norm": 40.07706832885742,
"learning_rate": 9.141100000000001e-06,
"loss": 1.25,
"step": 8600
},
{
"epoch": 44.1025641025641,
"eval_loss": 1.3270679712295532,
"eval_runtime": 36.1754,
"eval_samples_per_second": 10.864,
"eval_steps_per_second": 1.382,
"step": 8600
},
{
"epoch": 44.61538461538461,
"grad_norm": 27.011186599731445,
"learning_rate": 9.1311e-06,
"loss": 1.1968,
"step": 8700
},
{
"epoch": 44.61538461538461,
"eval_loss": 1.3117327690124512,
"eval_runtime": 36.1727,
"eval_samples_per_second": 10.865,
"eval_steps_per_second": 1.382,
"step": 8700
},
{
"epoch": 45.12820512820513,
"grad_norm": 25.976228713989258,
"learning_rate": 9.1211e-06,
"loss": 1.2492,
"step": 8800
},
{
"epoch": 45.12820512820513,
"eval_loss": 1.3281300067901611,
"eval_runtime": 36.2531,
"eval_samples_per_second": 10.84,
"eval_steps_per_second": 1.379,
"step": 8800
},
{
"epoch": 45.64102564102564,
"grad_norm": 21.215715408325195,
"learning_rate": 9.1111e-06,
"loss": 1.221,
"step": 8900
},
{
"epoch": 45.64102564102564,
"eval_loss": 1.3373734951019287,
"eval_runtime": 36.2242,
"eval_samples_per_second": 10.849,
"eval_steps_per_second": 1.38,
"step": 8900
},
{
"epoch": 46.15384615384615,
"grad_norm": 48.8258171081543,
"learning_rate": 9.1011e-06,
"loss": 1.2123,
"step": 9000
},
{
"epoch": 46.15384615384615,
"eval_loss": 1.3303853273391724,
"eval_runtime": 36.4019,
"eval_samples_per_second": 10.796,
"eval_steps_per_second": 1.374,
"step": 9000
},
{
"epoch": 46.666666666666664,
"grad_norm": 36.76605224609375,
"learning_rate": 9.0911e-06,
"loss": 1.1951,
"step": 9100
},
{
"epoch": 46.666666666666664,
"eval_loss": 1.3182373046875,
"eval_runtime": 36.387,
"eval_samples_per_second": 10.801,
"eval_steps_per_second": 1.374,
"step": 9100
},
{
"epoch": 47.17948717948718,
"grad_norm": 40.79771423339844,
"learning_rate": 9.0811e-06,
"loss": 1.2155,
"step": 9200
},
{
"epoch": 47.17948717948718,
"eval_loss": 1.3303160667419434,
"eval_runtime": 36.2265,
"eval_samples_per_second": 10.848,
"eval_steps_per_second": 1.38,
"step": 9200
},
{
"epoch": 47.69230769230769,
"grad_norm": 48.06431579589844,
"learning_rate": 9.0711e-06,
"loss": 1.2236,
"step": 9300
},
{
"epoch": 47.69230769230769,
"eval_loss": 1.3128286600112915,
"eval_runtime": 36.3665,
"eval_samples_per_second": 10.807,
"eval_steps_per_second": 1.375,
"step": 9300
},
{
"epoch": 48.205128205128204,
"grad_norm": 31.19647216796875,
"learning_rate": 9.0611e-06,
"loss": 1.2033,
"step": 9400
},
{
"epoch": 48.205128205128204,
"eval_loss": 1.3134888410568237,
"eval_runtime": 36.5129,
"eval_samples_per_second": 10.763,
"eval_steps_per_second": 1.369,
"step": 9400
},
{
"epoch": 48.717948717948715,
"grad_norm": 20.11866569519043,
"learning_rate": 9.0511e-06,
"loss": 1.1955,
"step": 9500
},
{
"epoch": 48.717948717948715,
"eval_loss": 1.3154560327529907,
"eval_runtime": 36.6023,
"eval_samples_per_second": 10.737,
"eval_steps_per_second": 1.366,
"step": 9500
},
{
"epoch": 49.23076923076923,
"grad_norm": 36.3143424987793,
"learning_rate": 9.0411e-06,
"loss": 1.2067,
"step": 9600
},
{
"epoch": 49.23076923076923,
"eval_loss": 1.3158589601516724,
"eval_runtime": 36.4851,
"eval_samples_per_second": 10.772,
"eval_steps_per_second": 1.37,
"step": 9600
},
{
"epoch": 49.743589743589745,
"grad_norm": 48.41688537597656,
"learning_rate": 9.0311e-06,
"loss": 1.2295,
"step": 9700
},
{
"epoch": 49.743589743589745,
"eval_loss": 1.306788682937622,
"eval_runtime": 36.291,
"eval_samples_per_second": 10.829,
"eval_steps_per_second": 1.378,
"step": 9700
},
{
"epoch": 50.256410256410255,
"grad_norm": 26.129995346069336,
"learning_rate": 9.0211e-06,
"loss": 1.1809,
"step": 9800
},
{
"epoch": 50.256410256410255,
"eval_loss": 1.3299375772476196,
"eval_runtime": 36.4651,
"eval_samples_per_second": 10.777,
"eval_steps_per_second": 1.371,
"step": 9800
},
{
"epoch": 50.76923076923077,
"grad_norm": 19.543821334838867,
"learning_rate": 9.011100000000001e-06,
"loss": 1.2179,
"step": 9900
},
{
"epoch": 50.76923076923077,
"eval_loss": 1.317457675933838,
"eval_runtime": 36.2864,
"eval_samples_per_second": 10.831,
"eval_steps_per_second": 1.378,
"step": 9900
},
{
"epoch": 51.282051282051285,
"grad_norm": 25.619775772094727,
"learning_rate": 9.001100000000001e-06,
"loss": 1.1653,
"step": 10000
},
{
"epoch": 51.282051282051285,
"eval_loss": 1.31196928024292,
"eval_runtime": 36.3263,
"eval_samples_per_second": 10.819,
"eval_steps_per_second": 1.376,
"step": 10000
},
{
"epoch": 51.794871794871796,
"grad_norm": 45.30315017700195,
"learning_rate": 8.991100000000001e-06,
"loss": 1.2391,
"step": 10100
},
{
"epoch": 51.794871794871796,
"eval_loss": 1.305709719657898,
"eval_runtime": 36.2103,
"eval_samples_per_second": 10.853,
"eval_steps_per_second": 1.381,
"step": 10100
},
{
"epoch": 52.30769230769231,
"grad_norm": 26.942337036132812,
"learning_rate": 8.981100000000001e-06,
"loss": 1.2195,
"step": 10200
},
{
"epoch": 52.30769230769231,
"eval_loss": 1.3068941831588745,
"eval_runtime": 36.2565,
"eval_samples_per_second": 10.839,
"eval_steps_per_second": 1.379,
"step": 10200
},
{
"epoch": 52.82051282051282,
"grad_norm": 26.20073890686035,
"learning_rate": 8.9711e-06,
"loss": 1.1639,
"step": 10300
},
{
"epoch": 52.82051282051282,
"eval_loss": 1.3013452291488647,
"eval_runtime": 36.2146,
"eval_samples_per_second": 10.852,
"eval_steps_per_second": 1.381,
"step": 10300
},
{
"epoch": 53.333333333333336,
"grad_norm": 41.40350341796875,
"learning_rate": 8.9611e-06,
"loss": 1.2033,
"step": 10400
},
{
"epoch": 53.333333333333336,
"eval_loss": 1.305737853050232,
"eval_runtime": 36.4865,
"eval_samples_per_second": 10.771,
"eval_steps_per_second": 1.37,
"step": 10400
},
{
"epoch": 53.84615384615385,
"grad_norm": 28.133567810058594,
"learning_rate": 8.9511e-06,
"loss": 1.1906,
"step": 10500
},
{
"epoch": 53.84615384615385,
"eval_loss": 1.2961195707321167,
"eval_runtime": 36.2734,
"eval_samples_per_second": 10.834,
"eval_steps_per_second": 1.378,
"step": 10500
}
],
"logging_steps": 100,
"max_steps": 100000,
"num_input_tokens_seen": 0,
"num_train_epochs": 513,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 4.013553245184e+19,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}