| { | |
| "best_metric": 1.1292170286178589, | |
| "best_model_checkpoint": "/fs/startiger0/nmoghe/experiments/pbert/base/small/conv/en-de/mbert-tlm-chat-pe/checkpoint-237500", | |
| "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": 1.8404, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 4.9e-05, | |
| "loss": 1.6309, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 4.85e-05, | |
| "loss": 1.5295, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.4759, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 4.75e-05, | |
| "loss": 1.4358, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.3657480478286743, | |
| "eval_runtime": 8.9273, | |
| "eval_samples_per_second": 112.016, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 4.7e-05, | |
| "loss": 1.386, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 4.6500000000000005e-05, | |
| "loss": 1.3555, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 1.3255, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 4.55e-05, | |
| "loss": 1.3031, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.276, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.3058555126190186, | |
| "eval_runtime": 8.8749, | |
| "eval_samples_per_second": 112.677, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 4.4500000000000004e-05, | |
| "loss": 1.249, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 1.2339, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 4.35e-05, | |
| "loss": 1.2192, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 4.3e-05, | |
| "loss": 1.2013, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 4.25e-05, | |
| "loss": 1.1863, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.3136135339736938, | |
| "eval_runtime": 8.847, | |
| "eval_samples_per_second": 113.033, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 4.2e-05, | |
| "loss": 1.1564, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 4.15e-05, | |
| "loss": 1.1451, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 4.1e-05, | |
| "loss": 1.1454, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "learning_rate": 4.05e-05, | |
| "loss": 1.1322, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 4e-05, | |
| "loss": 1.1245, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.2940223217010498, | |
| "eval_runtime": 8.844, | |
| "eval_samples_per_second": 113.071, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "learning_rate": 3.9500000000000005e-05, | |
| "loss": 1.0988, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 3.9000000000000006e-05, | |
| "loss": 1.0768, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "learning_rate": 3.85e-05, | |
| "loss": 1.0807, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 3.8e-05, | |
| "loss": 1.0793, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 1.0549, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 1.267952561378479, | |
| "eval_runtime": 8.8578, | |
| "eval_samples_per_second": 112.895, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "learning_rate": 3.7e-05, | |
| "loss": 1.0449, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "learning_rate": 3.65e-05, | |
| "loss": 1.0326, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "learning_rate": 3.6e-05, | |
| "loss": 1.0234, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "learning_rate": 3.55e-05, | |
| "loss": 1.0157, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 3.5e-05, | |
| "loss": 1.0127, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 1.2752054929733276, | |
| "eval_runtime": 8.8449, | |
| "eval_samples_per_second": 113.059, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "learning_rate": 3.45e-05, | |
| "loss": 0.9872, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 0.9805, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "learning_rate": 3.35e-05, | |
| "loss": 0.9781, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "learning_rate": 3.3e-05, | |
| "loss": 0.9722, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.9763, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 1.2642838954925537, | |
| "eval_runtime": 8.8491, | |
| "eval_samples_per_second": 113.006, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 0.9378, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "learning_rate": 3.15e-05, | |
| "loss": 0.946, | |
| "step": 92500 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "learning_rate": 3.1e-05, | |
| "loss": 0.9478, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "learning_rate": 3.05e-05, | |
| "loss": 0.9375, | |
| "step": 97500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.9409, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.2444370985031128, | |
| "eval_runtime": 8.8563, | |
| "eval_samples_per_second": 112.914, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "learning_rate": 2.95e-05, | |
| "loss": 0.9093, | |
| "step": 102500 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "learning_rate": 2.9e-05, | |
| "loss": 0.9071, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "learning_rate": 2.8499999999999998e-05, | |
| "loss": 0.9018, | |
| "step": 107500 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 0.8963, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.8909, | |
| "step": 112500 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 1.2180256843566895, | |
| "eval_runtime": 8.851, | |
| "eval_samples_per_second": 112.982, | |
| "step": 112500 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "learning_rate": 2.7000000000000002e-05, | |
| "loss": 0.8717, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "learning_rate": 2.6500000000000004e-05, | |
| "loss": 0.8698, | |
| "step": 117500 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 0.8663, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "learning_rate": 2.5500000000000003e-05, | |
| "loss": 0.8672, | |
| "step": 122500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.8495, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 1.2572065591812134, | |
| "eval_runtime": 8.854, | |
| "eval_samples_per_second": 112.943, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "learning_rate": 2.45e-05, | |
| "loss": 0.8392, | |
| "step": 127500 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "learning_rate": 2.4e-05, | |
| "loss": 0.8377, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "learning_rate": 2.35e-05, | |
| "loss": 0.833, | |
| "step": 132500 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "learning_rate": 2.3000000000000003e-05, | |
| "loss": 0.8347, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.8226, | |
| "step": 137500 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 1.213756799697876, | |
| "eval_runtime": 8.8526, | |
| "eval_samples_per_second": 112.962, | |
| "step": 137500 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 0.8166, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "learning_rate": 2.15e-05, | |
| "loss": 0.8027, | |
| "step": 142500 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "learning_rate": 2.1e-05, | |
| "loss": 0.8011, | |
| "step": 145000 | |
| }, | |
| { | |
| "epoch": 11.8, | |
| "learning_rate": 2.05e-05, | |
| "loss": 0.7975, | |
| "step": 147500 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7952, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 1.2098476886749268, | |
| "eval_runtime": 8.8556, | |
| "eval_samples_per_second": 112.923, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "learning_rate": 1.9500000000000003e-05, | |
| "loss": 0.7738, | |
| "step": 152500 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "learning_rate": 1.9e-05, | |
| "loss": 0.7779, | |
| "step": 155000 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "learning_rate": 1.85e-05, | |
| "loss": 0.7733, | |
| "step": 157500 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "learning_rate": 1.8e-05, | |
| "loss": 0.7686, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.773, | |
| "step": 162500 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 1.1837059259414673, | |
| "eval_runtime": 8.8557, | |
| "eval_samples_per_second": 112.922, | |
| "step": 162500 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "learning_rate": 1.7000000000000003e-05, | |
| "loss": 0.7526, | |
| "step": 165000 | |
| }, | |
| { | |
| "epoch": 13.4, | |
| "learning_rate": 1.65e-05, | |
| "loss": 0.7529, | |
| "step": 167500 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.7508, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 13.8, | |
| "learning_rate": 1.55e-05, | |
| "loss": 0.7448, | |
| "step": 172500 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.7389, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 1.2173564434051514, | |
| "eval_runtime": 8.857, | |
| "eval_samples_per_second": 112.906, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 14.2, | |
| "learning_rate": 1.45e-05, | |
| "loss": 0.7323, | |
| "step": 177500 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 0.7212, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 14.6, | |
| "learning_rate": 1.3500000000000001e-05, | |
| "loss": 0.7293, | |
| "step": 182500 | |
| }, | |
| { | |
| "epoch": 14.8, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 0.7242, | |
| "step": 185000 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.7081, | |
| "step": 187500 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_loss": 1.1677768230438232, | |
| "eval_runtime": 8.8548, | |
| "eval_samples_per_second": 112.933, | |
| "step": 187500 | |
| }, | |
| { | |
| "epoch": 15.2, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.7082, | |
| "step": 190000 | |
| }, | |
| { | |
| "epoch": 15.4, | |
| "learning_rate": 1.1500000000000002e-05, | |
| "loss": 0.6987, | |
| "step": 192500 | |
| }, | |
| { | |
| "epoch": 15.6, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 0.7037, | |
| "step": 195000 | |
| }, | |
| { | |
| "epoch": 15.8, | |
| "learning_rate": 1.05e-05, | |
| "loss": 0.7017, | |
| "step": 197500 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6963, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 1.1409714221954346, | |
| "eval_runtime": 8.8534, | |
| "eval_samples_per_second": 112.95, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 16.2, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.69, | |
| "step": 202500 | |
| }, | |
| { | |
| "epoch": 16.4, | |
| "learning_rate": 9e-06, | |
| "loss": 0.6777, | |
| "step": 205000 | |
| }, | |
| { | |
| "epoch": 16.6, | |
| "learning_rate": 8.500000000000002e-06, | |
| "loss": 0.6786, | |
| "step": 207500 | |
| }, | |
| { | |
| "epoch": 16.8, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.6806, | |
| "step": 210000 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.6715, | |
| "step": 212500 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_loss": 1.1735785007476807, | |
| "eval_runtime": 8.8543, | |
| "eval_samples_per_second": 112.939, | |
| "step": 212500 | |
| }, | |
| { | |
| "epoch": 17.2, | |
| "learning_rate": 7.000000000000001e-06, | |
| "loss": 0.661, | |
| "step": 215000 | |
| }, | |
| { | |
| "epoch": 17.4, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 0.6583, | |
| "step": 217500 | |
| }, | |
| { | |
| "epoch": 17.6, | |
| "learning_rate": 6e-06, | |
| "loss": 0.6584, | |
| "step": 220000 | |
| }, | |
| { | |
| "epoch": 17.8, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 0.661, | |
| "step": 222500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6508, | |
| "step": 225000 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 1.148830533027649, | |
| "eval_runtime": 8.857, | |
| "eval_samples_per_second": 112.905, | |
| "step": 225000 | |
| }, | |
| { | |
| "epoch": 18.2, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.6502, | |
| "step": 227500 | |
| }, | |
| { | |
| "epoch": 18.4, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.6405, | |
| "step": 230000 | |
| }, | |
| { | |
| "epoch": 18.6, | |
| "learning_rate": 3.5000000000000004e-06, | |
| "loss": 0.6493, | |
| "step": 232500 | |
| }, | |
| { | |
| "epoch": 18.8, | |
| "learning_rate": 3e-06, | |
| "loss": 0.6369, | |
| "step": 235000 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.6367, | |
| "step": 237500 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_loss": 1.1292170286178589, | |
| "eval_runtime": 8.8689, | |
| "eval_samples_per_second": 112.754, | |
| "step": 237500 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.6379, | |
| "step": 240000 | |
| }, | |
| { | |
| "epoch": 19.4, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.6314, | |
| "step": 242500 | |
| }, | |
| { | |
| "epoch": 19.6, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.6362, | |
| "step": 245000 | |
| }, | |
| { | |
| "epoch": 19.8, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 0.629, | |
| "step": 247500 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.6271, | |
| "step": 250000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 1.1654033660888672, | |
| "eval_runtime": 8.8585, | |
| "eval_samples_per_second": 112.886, | |
| "step": 250000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 250000, | |
| "train_runtime": 68180.3676, | |
| "train_samples_per_second": 3.667 | |
| } | |
| ], | |
| "max_steps": 250000, | |
| "num_train_epochs": 20, | |
| "total_flos": 292133171565386244, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |