|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 20.0, |
|
"global_step": 250000, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.2, |
|
"learning_rate": 4.9500000000000004e-05, |
|
"loss": 5.7977, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"learning_rate": 4.9e-05, |
|
"loss": 5.7044, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 0.6, |
|
"learning_rate": 4.85e-05, |
|
"loss": 5.6674, |
|
"step": 7500 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"learning_rate": 4.8e-05, |
|
"loss": 5.6295, |
|
"step": 10000 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"learning_rate": 4.75e-05, |
|
"loss": 5.6137, |
|
"step": 12500 |
|
}, |
|
{ |
|
"epoch": 1.2, |
|
"learning_rate": 4.7e-05, |
|
"loss": 5.526, |
|
"step": 15000 |
|
}, |
|
{ |
|
"epoch": 1.4, |
|
"learning_rate": 4.6500000000000005e-05, |
|
"loss": 5.5301, |
|
"step": 17500 |
|
}, |
|
{ |
|
"epoch": 1.6, |
|
"learning_rate": 4.600000000000001e-05, |
|
"loss": 5.5289, |
|
"step": 20000 |
|
}, |
|
{ |
|
"epoch": 1.8, |
|
"learning_rate": 4.55e-05, |
|
"loss": 5.5168, |
|
"step": 22500 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"learning_rate": 4.5e-05, |
|
"loss": 5.5139, |
|
"step": 25000 |
|
}, |
|
{ |
|
"epoch": 2.2, |
|
"learning_rate": 4.4500000000000004e-05, |
|
"loss": 5.3875, |
|
"step": 27500 |
|
}, |
|
{ |
|
"epoch": 2.4, |
|
"learning_rate": 4.4000000000000006e-05, |
|
"loss": 5.4007, |
|
"step": 30000 |
|
}, |
|
{ |
|
"epoch": 2.6, |
|
"learning_rate": 4.35e-05, |
|
"loss": 5.4018, |
|
"step": 32500 |
|
}, |
|
{ |
|
"epoch": 2.8, |
|
"learning_rate": 4.3e-05, |
|
"loss": 5.4068, |
|
"step": 35000 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"learning_rate": 4.25e-05, |
|
"loss": 5.416, |
|
"step": 37500 |
|
}, |
|
{ |
|
"epoch": 3.2, |
|
"learning_rate": 4.2e-05, |
|
"loss": 5.2336, |
|
"step": 40000 |
|
}, |
|
{ |
|
"epoch": 3.4, |
|
"learning_rate": 4.15e-05, |
|
"loss": 5.2493, |
|
"step": 42500 |
|
}, |
|
{ |
|
"epoch": 3.6, |
|
"learning_rate": 4.1e-05, |
|
"loss": 5.2625, |
|
"step": 45000 |
|
}, |
|
{ |
|
"epoch": 3.8, |
|
"learning_rate": 4.05e-05, |
|
"loss": 5.266, |
|
"step": 47500 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"learning_rate": 4e-05, |
|
"loss": 5.2859, |
|
"step": 50000 |
|
}, |
|
{ |
|
"epoch": 4.2, |
|
"learning_rate": 3.9500000000000005e-05, |
|
"loss": 5.0473, |
|
"step": 52500 |
|
}, |
|
{ |
|
"epoch": 4.4, |
|
"learning_rate": 3.9000000000000006e-05, |
|
"loss": 5.0821, |
|
"step": 55000 |
|
}, |
|
{ |
|
"epoch": 4.6, |
|
"learning_rate": 3.85e-05, |
|
"loss": 5.1005, |
|
"step": 57500 |
|
}, |
|
{ |
|
"epoch": 4.8, |
|
"learning_rate": 3.8e-05, |
|
"loss": 5.1137, |
|
"step": 60000 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"learning_rate": 3.7500000000000003e-05, |
|
"loss": 5.1128, |
|
"step": 62500 |
|
}, |
|
{ |
|
"epoch": 5.2, |
|
"learning_rate": 3.7e-05, |
|
"loss": 4.8492, |
|
"step": 65000 |
|
}, |
|
{ |
|
"epoch": 5.4, |
|
"learning_rate": 3.65e-05, |
|
"loss": 4.8748, |
|
"step": 67500 |
|
}, |
|
{ |
|
"epoch": 5.6, |
|
"learning_rate": 3.6e-05, |
|
"loss": 4.9024, |
|
"step": 70000 |
|
}, |
|
{ |
|
"epoch": 5.8, |
|
"learning_rate": 3.55e-05, |
|
"loss": 4.8889, |
|
"step": 72500 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"learning_rate": 3.5e-05, |
|
"loss": 4.9244, |
|
"step": 75000 |
|
}, |
|
{ |
|
"epoch": 6.2, |
|
"learning_rate": 3.45e-05, |
|
"loss": 4.588, |
|
"step": 77500 |
|
}, |
|
{ |
|
"epoch": 6.4, |
|
"learning_rate": 3.4000000000000007e-05, |
|
"loss": 4.6426, |
|
"step": 80000 |
|
}, |
|
{ |
|
"epoch": 6.6, |
|
"learning_rate": 3.35e-05, |
|
"loss": 4.6667, |
|
"step": 82500 |
|
}, |
|
{ |
|
"epoch": 6.8, |
|
"learning_rate": 3.3e-05, |
|
"loss": 4.6899, |
|
"step": 85000 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"learning_rate": 3.2500000000000004e-05, |
|
"loss": 4.6906, |
|
"step": 87500 |
|
}, |
|
{ |
|
"epoch": 7.2, |
|
"learning_rate": 3.2000000000000005e-05, |
|
"loss": 4.3279, |
|
"step": 90000 |
|
}, |
|
{ |
|
"epoch": 7.4, |
|
"learning_rate": 3.15e-05, |
|
"loss": 4.3742, |
|
"step": 92500 |
|
}, |
|
{ |
|
"epoch": 7.6, |
|
"learning_rate": 3.1e-05, |
|
"loss": 4.4195, |
|
"step": 95000 |
|
}, |
|
{ |
|
"epoch": 7.8, |
|
"learning_rate": 3.05e-05, |
|
"loss": 4.4611, |
|
"step": 97500 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"learning_rate": 3e-05, |
|
"loss": 4.4487, |
|
"step": 100000 |
|
}, |
|
{ |
|
"epoch": 8.2, |
|
"learning_rate": 2.95e-05, |
|
"loss": 4.0717, |
|
"step": 102500 |
|
}, |
|
{ |
|
"epoch": 8.4, |
|
"learning_rate": 2.9e-05, |
|
"loss": 4.1147, |
|
"step": 105000 |
|
}, |
|
{ |
|
"epoch": 8.6, |
|
"learning_rate": 2.8499999999999998e-05, |
|
"loss": 4.1526, |
|
"step": 107500 |
|
}, |
|
{ |
|
"epoch": 8.8, |
|
"learning_rate": 2.8000000000000003e-05, |
|
"loss": 4.1648, |
|
"step": 110000 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"learning_rate": 2.7500000000000004e-05, |
|
"loss": 4.2034, |
|
"step": 112500 |
|
}, |
|
{ |
|
"epoch": 9.2, |
|
"learning_rate": 2.7000000000000002e-05, |
|
"loss": 3.8116, |
|
"step": 115000 |
|
}, |
|
{ |
|
"epoch": 9.4, |
|
"learning_rate": 2.6500000000000004e-05, |
|
"loss": 3.8477, |
|
"step": 117500 |
|
}, |
|
{ |
|
"epoch": 9.6, |
|
"learning_rate": 2.6000000000000002e-05, |
|
"loss": 3.8755, |
|
"step": 120000 |
|
}, |
|
{ |
|
"epoch": 9.8, |
|
"learning_rate": 2.5500000000000003e-05, |
|
"loss": 3.9126, |
|
"step": 122500 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"learning_rate": 2.5e-05, |
|
"loss": 3.9229, |
|
"step": 125000 |
|
}, |
|
{ |
|
"epoch": 10.2, |
|
"learning_rate": 2.45e-05, |
|
"loss": 3.5519, |
|
"step": 127500 |
|
}, |
|
{ |
|
"epoch": 10.4, |
|
"learning_rate": 2.4e-05, |
|
"loss": 3.5838, |
|
"step": 130000 |
|
}, |
|
{ |
|
"epoch": 10.6, |
|
"learning_rate": 2.35e-05, |
|
"loss": 3.6086, |
|
"step": 132500 |
|
}, |
|
{ |
|
"epoch": 10.8, |
|
"learning_rate": 2.3000000000000003e-05, |
|
"loss": 3.6591, |
|
"step": 135000 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"learning_rate": 2.25e-05, |
|
"loss": 3.6712, |
|
"step": 137500 |
|
}, |
|
{ |
|
"epoch": 11.2, |
|
"learning_rate": 2.2000000000000003e-05, |
|
"loss": 3.2807, |
|
"step": 140000 |
|
}, |
|
{ |
|
"epoch": 11.4, |
|
"learning_rate": 2.15e-05, |
|
"loss": 3.3334, |
|
"step": 142500 |
|
}, |
|
{ |
|
"epoch": 11.6, |
|
"learning_rate": 2.1e-05, |
|
"loss": 3.3916, |
|
"step": 145000 |
|
}, |
|
{ |
|
"epoch": 11.8, |
|
"learning_rate": 2.05e-05, |
|
"loss": 3.4028, |
|
"step": 147500 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"learning_rate": 2e-05, |
|
"loss": 3.4255, |
|
"step": 150000 |
|
}, |
|
{ |
|
"epoch": 12.2, |
|
"learning_rate": 1.9500000000000003e-05, |
|
"loss": 3.0687, |
|
"step": 152500 |
|
}, |
|
{ |
|
"epoch": 12.4, |
|
"learning_rate": 1.9e-05, |
|
"loss": 3.1065, |
|
"step": 155000 |
|
}, |
|
{ |
|
"epoch": 12.6, |
|
"learning_rate": 1.85e-05, |
|
"loss": 3.1308, |
|
"step": 157500 |
|
}, |
|
{ |
|
"epoch": 12.8, |
|
"learning_rate": 1.8e-05, |
|
"loss": 3.1547, |
|
"step": 160000 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"learning_rate": 1.75e-05, |
|
"loss": 3.1876, |
|
"step": 162500 |
|
}, |
|
{ |
|
"epoch": 13.2, |
|
"learning_rate": 1.7000000000000003e-05, |
|
"loss": 2.8464, |
|
"step": 165000 |
|
}, |
|
{ |
|
"epoch": 13.4, |
|
"learning_rate": 1.65e-05, |
|
"loss": 2.8833, |
|
"step": 167500 |
|
}, |
|
{ |
|
"epoch": 13.6, |
|
"learning_rate": 1.6000000000000003e-05, |
|
"loss": 2.9187, |
|
"step": 170000 |
|
}, |
|
{ |
|
"epoch": 13.8, |
|
"learning_rate": 1.55e-05, |
|
"loss": 2.9404, |
|
"step": 172500 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"learning_rate": 1.5e-05, |
|
"loss": 2.9635, |
|
"step": 175000 |
|
}, |
|
{ |
|
"epoch": 14.2, |
|
"learning_rate": 1.45e-05, |
|
"loss": 2.636, |
|
"step": 177500 |
|
}, |
|
{ |
|
"epoch": 14.4, |
|
"learning_rate": 1.4000000000000001e-05, |
|
"loss": 2.7018, |
|
"step": 180000 |
|
}, |
|
{ |
|
"epoch": 14.6, |
|
"learning_rate": 1.3500000000000001e-05, |
|
"loss": 2.7054, |
|
"step": 182500 |
|
}, |
|
{ |
|
"epoch": 14.8, |
|
"learning_rate": 1.3000000000000001e-05, |
|
"loss": 2.7442, |
|
"step": 185000 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"learning_rate": 1.25e-05, |
|
"loss": 2.7594, |
|
"step": 187500 |
|
}, |
|
{ |
|
"epoch": 15.2, |
|
"learning_rate": 1.2e-05, |
|
"loss": 2.4725, |
|
"step": 190000 |
|
}, |
|
{ |
|
"epoch": 15.4, |
|
"learning_rate": 1.1500000000000002e-05, |
|
"loss": 2.5238, |
|
"step": 192500 |
|
}, |
|
{ |
|
"epoch": 15.6, |
|
"learning_rate": 1.1000000000000001e-05, |
|
"loss": 2.5331, |
|
"step": 195000 |
|
}, |
|
{ |
|
"epoch": 15.8, |
|
"learning_rate": 1.05e-05, |
|
"loss": 2.5536, |
|
"step": 197500 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"learning_rate": 1e-05, |
|
"loss": 2.5667, |
|
"step": 200000 |
|
}, |
|
{ |
|
"epoch": 16.2, |
|
"learning_rate": 9.5e-06, |
|
"loss": 2.3507, |
|
"step": 202500 |
|
}, |
|
{ |
|
"epoch": 16.4, |
|
"learning_rate": 9e-06, |
|
"loss": 2.3498, |
|
"step": 205000 |
|
}, |
|
{ |
|
"epoch": 16.6, |
|
"learning_rate": 8.500000000000002e-06, |
|
"loss": 2.3682, |
|
"step": 207500 |
|
}, |
|
{ |
|
"epoch": 16.8, |
|
"learning_rate": 8.000000000000001e-06, |
|
"loss": 2.3927, |
|
"step": 210000 |
|
}, |
|
{ |
|
"epoch": 17.0, |
|
"learning_rate": 7.5e-06, |
|
"loss": 2.3977, |
|
"step": 212500 |
|
}, |
|
{ |
|
"epoch": 17.2, |
|
"learning_rate": 7.000000000000001e-06, |
|
"loss": 2.2321, |
|
"step": 215000 |
|
}, |
|
{ |
|
"epoch": 17.4, |
|
"learning_rate": 6.5000000000000004e-06, |
|
"loss": 2.2189, |
|
"step": 217500 |
|
}, |
|
{ |
|
"epoch": 17.6, |
|
"learning_rate": 6e-06, |
|
"loss": 2.2531, |
|
"step": 220000 |
|
}, |
|
{ |
|
"epoch": 17.8, |
|
"learning_rate": 5.500000000000001e-06, |
|
"loss": 2.239, |
|
"step": 222500 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"learning_rate": 5e-06, |
|
"loss": 2.2382, |
|
"step": 225000 |
|
}, |
|
{ |
|
"epoch": 18.2, |
|
"learning_rate": 4.5e-06, |
|
"loss": 2.1317, |
|
"step": 227500 |
|
}, |
|
{ |
|
"epoch": 18.4, |
|
"learning_rate": 4.000000000000001e-06, |
|
"loss": 2.1052, |
|
"step": 230000 |
|
}, |
|
{ |
|
"epoch": 18.6, |
|
"learning_rate": 3.5000000000000004e-06, |
|
"loss": 2.1229, |
|
"step": 232500 |
|
}, |
|
{ |
|
"epoch": 18.8, |
|
"learning_rate": 3e-06, |
|
"loss": 2.1331, |
|
"step": 235000 |
|
}, |
|
{ |
|
"epoch": 19.0, |
|
"learning_rate": 2.5e-06, |
|
"loss": 2.1464, |
|
"step": 237500 |
|
}, |
|
{ |
|
"epoch": 19.2, |
|
"learning_rate": 2.0000000000000003e-06, |
|
"loss": 2.0346, |
|
"step": 240000 |
|
}, |
|
{ |
|
"epoch": 19.4, |
|
"learning_rate": 1.5e-06, |
|
"loss": 2.0326, |
|
"step": 242500 |
|
}, |
|
{ |
|
"epoch": 19.6, |
|
"learning_rate": 1.0000000000000002e-06, |
|
"loss": 2.0521, |
|
"step": 245000 |
|
}, |
|
{ |
|
"epoch": 19.8, |
|
"learning_rate": 5.000000000000001e-07, |
|
"loss": 2.0473, |
|
"step": 247500 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"learning_rate": 0.0, |
|
"loss": 2.0588, |
|
"step": 250000 |
|
}, |
|
{ |
|
"epoch": 20.0, |
|
"step": 250000, |
|
"train_runtime": 46998.5214, |
|
"train_samples_per_second": 5.319 |
|
} |
|
], |
|
"max_steps": 250000, |
|
"num_train_epochs": 20, |
|
"total_flos": 316206548084269296, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|