| { | |
| "best_metric": 1.0504688024520874, | |
| "best_model_checkpoint": "/fs/startiger0/nmoghe/experiments/pbert/base/small/conv/en-it/mbert-tlm-chat-pe/checkpoint-225000", | |
| "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": 1.8222, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 4.9e-05, | |
| "loss": 1.6035, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 4.85e-05, | |
| "loss": 1.4988, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 4.8e-05, | |
| "loss": 1.4465, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 4.75e-05, | |
| "loss": 1.3893, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.4548144340515137, | |
| "eval_runtime": 9.5125, | |
| "eval_samples_per_second": 105.125, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 4.7e-05, | |
| "loss": 1.3425, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 4.6500000000000005e-05, | |
| "loss": 1.3139, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 1.2891, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 4.55e-05, | |
| "loss": 1.2596, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 4.5e-05, | |
| "loss": 1.2512, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 1.3658138513565063, | |
| "eval_runtime": 9.3984, | |
| "eval_samples_per_second": 106.401, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 4.4500000000000004e-05, | |
| "loss": 1.2094, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 1.187, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 4.35e-05, | |
| "loss": 1.1791, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 4.3e-05, | |
| "loss": 1.1635, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 4.25e-05, | |
| "loss": 1.1478, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 1.2877916097640991, | |
| "eval_runtime": 9.4132, | |
| "eval_samples_per_second": 106.233, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 4.2e-05, | |
| "loss": 1.1201, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 3.4, | |
| "learning_rate": 4.15e-05, | |
| "loss": 1.1054, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 4.1e-05, | |
| "loss": 1.1036, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "learning_rate": 4.05e-05, | |
| "loss": 1.0881, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "learning_rate": 4e-05, | |
| "loss": 1.0839, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 1.2922987937927246, | |
| "eval_runtime": 9.404, | |
| "eval_samples_per_second": 106.338, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "learning_rate": 3.9500000000000005e-05, | |
| "loss": 1.0512, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "learning_rate": 3.9000000000000006e-05, | |
| "loss": 1.0497, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 4.6, | |
| "learning_rate": 3.85e-05, | |
| "loss": 1.045, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "learning_rate": 3.8e-05, | |
| "loss": 1.0245, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 1.0315, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 1.2622771263122559, | |
| "eval_runtime": 9.3953, | |
| "eval_samples_per_second": 106.436, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "learning_rate": 3.7e-05, | |
| "loss": 0.994, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 5.4, | |
| "learning_rate": 3.65e-05, | |
| "loss": 0.9962, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "learning_rate": 3.6e-05, | |
| "loss": 0.9982, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 5.8, | |
| "learning_rate": 3.55e-05, | |
| "loss": 0.9833, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.9806, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 1.2437635660171509, | |
| "eval_runtime": 9.3976, | |
| "eval_samples_per_second": 106.41, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 6.2, | |
| "learning_rate": 3.45e-05, | |
| "loss": 0.9458, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 3.4000000000000007e-05, | |
| "loss": 0.9548, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "learning_rate": 3.35e-05, | |
| "loss": 0.9523, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "learning_rate": 3.3e-05, | |
| "loss": 0.9439, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "learning_rate": 3.2500000000000004e-05, | |
| "loss": 0.9384, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 1.1912026405334473, | |
| "eval_runtime": 9.4045, | |
| "eval_samples_per_second": 106.332, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "learning_rate": 3.2000000000000005e-05, | |
| "loss": 0.9196, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "learning_rate": 3.15e-05, | |
| "loss": 0.9098, | |
| "step": 92500 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "learning_rate": 3.1e-05, | |
| "loss": 0.9152, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 7.8, | |
| "learning_rate": 3.05e-05, | |
| "loss": 0.9073, | |
| "step": 97500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "learning_rate": 3e-05, | |
| "loss": 0.8991, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 1.1714017391204834, | |
| "eval_runtime": 9.4059, | |
| "eval_samples_per_second": 106.316, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 8.2, | |
| "learning_rate": 2.95e-05, | |
| "loss": 0.8676, | |
| "step": 102500 | |
| }, | |
| { | |
| "epoch": 8.4, | |
| "learning_rate": 2.9e-05, | |
| "loss": 0.8792, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 8.6, | |
| "learning_rate": 2.8499999999999998e-05, | |
| "loss": 0.872, | |
| "step": 107500 | |
| }, | |
| { | |
| "epoch": 8.8, | |
| "learning_rate": 2.8000000000000003e-05, | |
| "loss": 0.8728, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "learning_rate": 2.7500000000000004e-05, | |
| "loss": 0.8617, | |
| "step": 112500 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 1.1749364137649536, | |
| "eval_runtime": 9.4187, | |
| "eval_samples_per_second": 106.171, | |
| "step": 112500 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "learning_rate": 2.7000000000000002e-05, | |
| "loss": 0.8444, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 9.4, | |
| "learning_rate": 2.6500000000000004e-05, | |
| "loss": 0.8463, | |
| "step": 117500 | |
| }, | |
| { | |
| "epoch": 9.6, | |
| "learning_rate": 2.6000000000000002e-05, | |
| "loss": 0.8334, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 9.8, | |
| "learning_rate": 2.5500000000000003e-05, | |
| "loss": 0.8304, | |
| "step": 122500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.8291, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 1.1757138967514038, | |
| "eval_runtime": 9.3955, | |
| "eval_samples_per_second": 106.433, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 10.2, | |
| "learning_rate": 2.45e-05, | |
| "loss": 0.8166, | |
| "step": 127500 | |
| }, | |
| { | |
| "epoch": 10.4, | |
| "learning_rate": 2.4e-05, | |
| "loss": 0.8126, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 10.6, | |
| "learning_rate": 2.35e-05, | |
| "loss": 0.8102, | |
| "step": 132500 | |
| }, | |
| { | |
| "epoch": 10.8, | |
| "learning_rate": 2.3000000000000003e-05, | |
| "loss": 0.8029, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "learning_rate": 2.25e-05, | |
| "loss": 0.8031, | |
| "step": 137500 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_loss": 1.1473666429519653, | |
| "eval_runtime": 9.3894, | |
| "eval_samples_per_second": 106.503, | |
| "step": 137500 | |
| }, | |
| { | |
| "epoch": 11.2, | |
| "learning_rate": 2.2000000000000003e-05, | |
| "loss": 0.7793, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 11.4, | |
| "learning_rate": 2.15e-05, | |
| "loss": 0.7842, | |
| "step": 142500 | |
| }, | |
| { | |
| "epoch": 11.6, | |
| "learning_rate": 2.1e-05, | |
| "loss": 0.7734, | |
| "step": 145000 | |
| }, | |
| { | |
| "epoch": 11.8, | |
| "learning_rate": 2.05e-05, | |
| "loss": 0.7817, | |
| "step": 147500 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "learning_rate": 2e-05, | |
| "loss": 0.7701, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_loss": 1.1467667818069458, | |
| "eval_runtime": 9.4066, | |
| "eval_samples_per_second": 106.308, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 12.2, | |
| "learning_rate": 1.9500000000000003e-05, | |
| "loss": 0.7573, | |
| "step": 152500 | |
| }, | |
| { | |
| "epoch": 12.4, | |
| "learning_rate": 1.9e-05, | |
| "loss": 0.7483, | |
| "step": 155000 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "learning_rate": 1.85e-05, | |
| "loss": 0.753, | |
| "step": 157500 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "learning_rate": 1.8e-05, | |
| "loss": 0.7475, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "learning_rate": 1.75e-05, | |
| "loss": 0.7395, | |
| "step": 162500 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_loss": 1.1321425437927246, | |
| "eval_runtime": 9.3951, | |
| "eval_samples_per_second": 106.438, | |
| "step": 162500 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "learning_rate": 1.7000000000000003e-05, | |
| "loss": 0.7292, | |
| "step": 165000 | |
| }, | |
| { | |
| "epoch": 13.4, | |
| "learning_rate": 1.65e-05, | |
| "loss": 0.7296, | |
| "step": 167500 | |
| }, | |
| { | |
| "epoch": 13.6, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 0.719, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 13.8, | |
| "learning_rate": 1.55e-05, | |
| "loss": 0.7197, | |
| "step": 172500 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.7122, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_loss": 1.1006277799606323, | |
| "eval_runtime": 9.3983, | |
| "eval_samples_per_second": 106.402, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 14.2, | |
| "learning_rate": 1.45e-05, | |
| "loss": 0.7011, | |
| "step": 177500 | |
| }, | |
| { | |
| "epoch": 14.4, | |
| "learning_rate": 1.4000000000000001e-05, | |
| "loss": 0.7032, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 14.6, | |
| "learning_rate": 1.3500000000000001e-05, | |
| "loss": 0.6977, | |
| "step": 182500 | |
| }, | |
| { | |
| "epoch": 14.8, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 0.7004, | |
| "step": 185000 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.6871, | |
| "step": 187500 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_loss": 1.1152338981628418, | |
| "eval_runtime": 9.3979, | |
| "eval_samples_per_second": 106.407, | |
| "step": 187500 | |
| }, | |
| { | |
| "epoch": 15.2, | |
| "learning_rate": 1.2e-05, | |
| "loss": 0.6793, | |
| "step": 190000 | |
| }, | |
| { | |
| "epoch": 15.4, | |
| "learning_rate": 1.1500000000000002e-05, | |
| "loss": 0.6775, | |
| "step": 192500 | |
| }, | |
| { | |
| "epoch": 15.6, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 0.6764, | |
| "step": 195000 | |
| }, | |
| { | |
| "epoch": 15.8, | |
| "learning_rate": 1.05e-05, | |
| "loss": 0.6731, | |
| "step": 197500 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6773, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 1.076183795928955, | |
| "eval_runtime": 9.4003, | |
| "eval_samples_per_second": 106.38, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 16.2, | |
| "learning_rate": 9.5e-06, | |
| "loss": 0.6625, | |
| "step": 202500 | |
| }, | |
| { | |
| "epoch": 16.4, | |
| "learning_rate": 9e-06, | |
| "loss": 0.6525, | |
| "step": 205000 | |
| }, | |
| { | |
| "epoch": 16.6, | |
| "learning_rate": 8.500000000000002e-06, | |
| "loss": 0.6566, | |
| "step": 207500 | |
| }, | |
| { | |
| "epoch": 16.8, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 0.6557, | |
| "step": 210000 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "learning_rate": 7.5e-06, | |
| "loss": 0.653, | |
| "step": 212500 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_loss": 1.0961084365844727, | |
| "eval_runtime": 9.3914, | |
| "eval_samples_per_second": 106.48, | |
| "step": 212500 | |
| }, | |
| { | |
| "epoch": 17.2, | |
| "learning_rate": 7.000000000000001e-06, | |
| "loss": 0.6431, | |
| "step": 215000 | |
| }, | |
| { | |
| "epoch": 17.4, | |
| "learning_rate": 6.5000000000000004e-06, | |
| "loss": 0.6416, | |
| "step": 217500 | |
| }, | |
| { | |
| "epoch": 17.6, | |
| "learning_rate": 6e-06, | |
| "loss": 0.6391, | |
| "step": 220000 | |
| }, | |
| { | |
| "epoch": 17.8, | |
| "learning_rate": 5.500000000000001e-06, | |
| "loss": 0.6345, | |
| "step": 222500 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6365, | |
| "step": 225000 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_loss": 1.0504688024520874, | |
| "eval_runtime": 9.4106, | |
| "eval_samples_per_second": 106.263, | |
| "step": 225000 | |
| }, | |
| { | |
| "epoch": 18.2, | |
| "learning_rate": 4.5e-06, | |
| "loss": 0.623, | |
| "step": 227500 | |
| }, | |
| { | |
| "epoch": 18.4, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 0.6205, | |
| "step": 230000 | |
| }, | |
| { | |
| "epoch": 18.6, | |
| "learning_rate": 3.5000000000000004e-06, | |
| "loss": 0.6231, | |
| "step": 232500 | |
| }, | |
| { | |
| "epoch": 18.8, | |
| "learning_rate": 3e-06, | |
| "loss": 0.6235, | |
| "step": 235000 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "learning_rate": 2.5e-06, | |
| "loss": 0.6104, | |
| "step": 237500 | |
| }, | |
| { | |
| "epoch": 19.0, | |
| "eval_loss": 1.0633997917175293, | |
| "eval_runtime": 9.3934, | |
| "eval_samples_per_second": 106.458, | |
| "step": 237500 | |
| }, | |
| { | |
| "epoch": 19.2, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 0.6154, | |
| "step": 240000 | |
| }, | |
| { | |
| "epoch": 19.4, | |
| "learning_rate": 1.5e-06, | |
| "loss": 0.6103, | |
| "step": 242500 | |
| }, | |
| { | |
| "epoch": 19.6, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 0.5998, | |
| "step": 245000 | |
| }, | |
| { | |
| "epoch": 19.8, | |
| "learning_rate": 5.000000000000001e-07, | |
| "loss": 0.6125, | |
| "step": 247500 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "learning_rate": 0.0, | |
| "loss": 0.6093, | |
| "step": 250000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 1.0744267702102661, | |
| "eval_runtime": 9.4035, | |
| "eval_samples_per_second": 106.344, | |
| "step": 250000 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "step": 250000, | |
| "train_runtime": 70464.5842, | |
| "train_samples_per_second": 3.548 | |
| } | |
| ], | |
| "max_steps": 250000, | |
| "num_train_epochs": 20, | |
| "total_flos": 305681719606891488, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |