| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.5195263290501386, | |
| "eval_steps": 500, | |
| "global_step": 20000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 0.000975, | |
| "loss": 1.0238, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "eval_loss": 0.8929525017738342, | |
| "eval_runtime": 4.0089, | |
| "eval_samples_per_second": 74.833, | |
| "eval_steps_per_second": 37.417, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 0.00095, | |
| "loss": 0.9971, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "eval_loss": 0.8510345220565796, | |
| "eval_runtime": 4.0089, | |
| "eval_samples_per_second": 74.833, | |
| "eval_steps_per_second": 37.417, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 0.000925, | |
| "loss": 0.9522, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "eval_loss": 0.8565580248832703, | |
| "eval_runtime": 4.0278, | |
| "eval_samples_per_second": 74.483, | |
| "eval_steps_per_second": 37.241, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 0.0009000000000000001, | |
| "loss": 0.9143, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "eval_loss": 0.8810315132141113, | |
| "eval_runtime": 4.0097, | |
| "eval_samples_per_second": 74.819, | |
| "eval_steps_per_second": 37.409, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 0.000875, | |
| "loss": 0.9151, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "eval_loss": 0.8461073040962219, | |
| "eval_runtime": 4.0108, | |
| "eval_samples_per_second": 74.798, | |
| "eval_steps_per_second": 37.399, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 0.00085, | |
| "loss": 0.9263, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "eval_loss": 0.8521466255187988, | |
| "eval_runtime": 4.0059, | |
| "eval_samples_per_second": 74.89, | |
| "eval_steps_per_second": 37.445, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 0.000825, | |
| "loss": 0.9102, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "eval_loss": 0.8521981835365295, | |
| "eval_runtime": 4.0413, | |
| "eval_samples_per_second": 74.234, | |
| "eval_steps_per_second": 37.117, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 0.0008, | |
| "loss": 0.8866, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.8259674906730652, | |
| "eval_runtime": 4.0316, | |
| "eval_samples_per_second": 74.412, | |
| "eval_steps_per_second": 37.206, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 0.0007750000000000001, | |
| "loss": 0.911, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "eval_loss": 0.8008900284767151, | |
| "eval_runtime": 4.0224, | |
| "eval_samples_per_second": 74.583, | |
| "eval_steps_per_second": 37.292, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 0.00075, | |
| "loss": 0.8536, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_loss": 0.808663547039032, | |
| "eval_runtime": 5.0615, | |
| "eval_samples_per_second": 59.271, | |
| "eval_steps_per_second": 29.635, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 0.000725, | |
| "loss": 0.8277, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "eval_loss": 0.7903586626052856, | |
| "eval_runtime": 4.0081, | |
| "eval_samples_per_second": 74.849, | |
| "eval_steps_per_second": 37.425, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 0.0007, | |
| "loss": 0.8413, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "eval_loss": 0.8013813495635986, | |
| "eval_runtime": 3.994, | |
| "eval_samples_per_second": 75.113, | |
| "eval_steps_per_second": 37.556, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 0.000675, | |
| "loss": 0.8491, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "eval_loss": 0.7867017984390259, | |
| "eval_runtime": 4.0072, | |
| "eval_samples_per_second": 74.864, | |
| "eval_steps_per_second": 37.432, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 0.0006500000000000001, | |
| "loss": 0.8077, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "eval_loss": 0.759165346622467, | |
| "eval_runtime": 4.0166, | |
| "eval_samples_per_second": 74.691, | |
| "eval_steps_per_second": 37.345, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 0.000625, | |
| "loss": 0.796, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "eval_loss": 0.7527943849563599, | |
| "eval_runtime": 3.998, | |
| "eval_samples_per_second": 75.037, | |
| "eval_steps_per_second": 37.518, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 0.0006, | |
| "loss": 0.7952, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "eval_loss": 0.7407393455505371, | |
| "eval_runtime": 3.9421, | |
| "eval_samples_per_second": 76.102, | |
| "eval_steps_per_second": 38.051, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 0.000575, | |
| "loss": 0.5761, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_loss": 0.7574229836463928, | |
| "eval_runtime": 3.9855, | |
| "eval_samples_per_second": 75.273, | |
| "eval_steps_per_second": 37.636, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 0.00055, | |
| "loss": 0.6084, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "eval_loss": 0.7349967956542969, | |
| "eval_runtime": 3.995, | |
| "eval_samples_per_second": 75.093, | |
| "eval_steps_per_second": 37.547, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 0.0005250000000000001, | |
| "loss": 0.5863, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 0.7407773733139038, | |
| "eval_runtime": 3.9874, | |
| "eval_samples_per_second": 75.238, | |
| "eval_steps_per_second": 37.619, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 0.0005, | |
| "loss": 0.5835, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "eval_loss": 0.719527006149292, | |
| "eval_runtime": 3.9471, | |
| "eval_samples_per_second": 76.006, | |
| "eval_steps_per_second": 38.003, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 0.000475, | |
| "loss": 0.5751, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "eval_loss": 0.7423205971717834, | |
| "eval_runtime": 3.9978, | |
| "eval_samples_per_second": 75.041, | |
| "eval_steps_per_second": 37.52, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 0.00045000000000000004, | |
| "loss": 0.5746, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "eval_loss": 0.7284040451049805, | |
| "eval_runtime": 3.9657, | |
| "eval_samples_per_second": 75.649, | |
| "eval_steps_per_second": 37.824, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 0.000425, | |
| "loss": 0.5847, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "eval_loss": 0.7247716188430786, | |
| "eval_runtime": 3.962, | |
| "eval_samples_per_second": 75.719, | |
| "eval_steps_per_second": 37.86, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 0.0004, | |
| "loss": 0.5759, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "eval_loss": 0.7151039838790894, | |
| "eval_runtime": 3.9922, | |
| "eval_samples_per_second": 75.147, | |
| "eval_steps_per_second": 37.573, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 0.000375, | |
| "loss": 0.5922, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "eval_loss": 0.7031030058860779, | |
| "eval_runtime": 4.0386, | |
| "eval_samples_per_second": 74.282, | |
| "eval_steps_per_second": 37.141, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 0.00035, | |
| "loss": 0.5678, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "eval_loss": 0.710035502910614, | |
| "eval_runtime": 4.0808, | |
| "eval_samples_per_second": 73.515, | |
| "eval_steps_per_second": 36.758, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 0.00032500000000000004, | |
| "loss": 0.5627, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "eval_loss": 0.7093074321746826, | |
| "eval_runtime": 4.0326, | |
| "eval_samples_per_second": 74.393, | |
| "eval_steps_per_second": 37.197, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 0.0003, | |
| "loss": 0.5568, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "eval_loss": 0.6942310333251953, | |
| "eval_runtime": 4.0875, | |
| "eval_samples_per_second": 73.395, | |
| "eval_steps_per_second": 36.697, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 0.000275, | |
| "loss": 0.5589, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "eval_loss": 0.6947352290153503, | |
| "eval_runtime": 3.9856, | |
| "eval_samples_per_second": 75.27, | |
| "eval_steps_per_second": 37.635, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 0.00025, | |
| "loss": 0.5675, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "eval_loss": 0.6902616620063782, | |
| "eval_runtime": 4.0066, | |
| "eval_samples_per_second": 74.876, | |
| "eval_steps_per_second": 37.438, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 0.00022500000000000002, | |
| "loss": 0.5422, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "eval_loss": 0.6929380297660828, | |
| "eval_runtime": 4.107, | |
| "eval_samples_per_second": 73.047, | |
| "eval_steps_per_second": 36.523, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 0.0002, | |
| "loss": 0.5104, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "eval_loss": 0.7098422050476074, | |
| "eval_runtime": 4.0323, | |
| "eval_samples_per_second": 74.4, | |
| "eval_steps_per_second": 37.2, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 0.000175, | |
| "loss": 0.3835, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "eval_loss": 0.7105218768119812, | |
| "eval_runtime": 4.0594, | |
| "eval_samples_per_second": 73.903, | |
| "eval_steps_per_second": 36.952, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 0.00015, | |
| "loss": 0.3805, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "eval_loss": 0.7144222855567932, | |
| "eval_runtime": 4.0853, | |
| "eval_samples_per_second": 73.434, | |
| "eval_steps_per_second": 36.717, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 0.000125, | |
| "loss": 0.3718, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "eval_loss": 0.7210414409637451, | |
| "eval_runtime": 5.0511, | |
| "eval_samples_per_second": 59.393, | |
| "eval_steps_per_second": 29.697, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3688, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "eval_loss": 0.7145898342132568, | |
| "eval_runtime": 4.7793, | |
| "eval_samples_per_second": 62.77, | |
| "eval_steps_per_second": 31.385, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 7.5e-05, | |
| "loss": 0.3645, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "eval_loss": 0.7136221528053284, | |
| "eval_runtime": 4.0171, | |
| "eval_samples_per_second": 74.681, | |
| "eval_steps_per_second": 37.34, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "learning_rate": 5e-05, | |
| "loss": 0.3865, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "eval_loss": 0.7075753808021545, | |
| "eval_runtime": 3.9658, | |
| "eval_samples_per_second": 75.646, | |
| "eval_steps_per_second": 37.823, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.3633, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "eval_loss": 0.7097809314727783, | |
| "eval_runtime": 4.0163, | |
| "eval_samples_per_second": 74.696, | |
| "eval_steps_per_second": 37.348, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 0.0, | |
| "loss": 0.3674, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "eval_loss": 0.7079904079437256, | |
| "eval_runtime": 4.0804, | |
| "eval_samples_per_second": 73.522, | |
| "eval_steps_per_second": 36.761, | |
| "step": 20000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 20000, | |
| "num_train_epochs": 3, | |
| "save_steps": 5000, | |
| "total_flos": 3.549121832463667e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |