| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.0, | |
| "global_step": 7900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 1.3719999999999999e-05, | |
| "loss": 10.7972, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "learning_rate": 2.772e-05, | |
| "loss": 4.2661, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 3.8, | |
| "learning_rate": 4.1719999999999994e-05, | |
| "loss": 3.3145, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 5.06, | |
| "learning_rate": 5.5719999999999995e-05, | |
| "loss": 3.039, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 6.33, | |
| "learning_rate": 6.971999999999999e-05, | |
| "loss": 2.9774, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 6.33, | |
| "eval_loss": 2.9768528938293457, | |
| "eval_runtime": 42.649, | |
| "eval_samples_per_second": 22.345, | |
| "eval_steps_per_second": 0.703, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 7.59, | |
| "learning_rate": 6.907297297297297e-05, | |
| "loss": 2.9216, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 8.86, | |
| "learning_rate": 6.812702702702703e-05, | |
| "loss": 2.453, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 10.13, | |
| "learning_rate": 6.718108108108108e-05, | |
| "loss": 1.7073, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 11.39, | |
| "learning_rate": 6.623513513513513e-05, | |
| "loss": 1.4457, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 12.66, | |
| "learning_rate": 6.528918918918918e-05, | |
| "loss": 1.3453, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 12.66, | |
| "eval_loss": 0.6522914171218872, | |
| "eval_runtime": 42.9729, | |
| "eval_samples_per_second": 22.177, | |
| "eval_steps_per_second": 0.698, | |
| "eval_wer": 0.6979675258317247, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 13.92, | |
| "learning_rate": 6.434324324324325e-05, | |
| "loss": 1.2867, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 15.19, | |
| "learning_rate": 6.339729729729728e-05, | |
| "loss": 1.2507, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 16.46, | |
| "learning_rate": 6.245135135135135e-05, | |
| "loss": 1.21, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 17.72, | |
| "learning_rate": 6.15054054054054e-05, | |
| "loss": 1.1863, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 18.99, | |
| "learning_rate": 6.055945945945946e-05, | |
| "loss": 1.1658, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 18.99, | |
| "eval_loss": 0.5635793805122375, | |
| "eval_runtime": 43.1337, | |
| "eval_samples_per_second": 22.094, | |
| "eval_steps_per_second": 0.696, | |
| "eval_wer": 0.6358578403542636, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 20.25, | |
| "learning_rate": 5.9613513513513504e-05, | |
| "loss": 1.1476, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 21.52, | |
| "learning_rate": 5.866756756756756e-05, | |
| "loss": 1.1346, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 22.78, | |
| "learning_rate": 5.772162162162162e-05, | |
| "loss": 1.1074, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 24.05, | |
| "learning_rate": 5.677567567567567e-05, | |
| "loss": 1.0892, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 25.32, | |
| "learning_rate": 5.5829729729729724e-05, | |
| "loss": 1.0797, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 25.32, | |
| "eval_loss": 0.5003558993339539, | |
| "eval_runtime": 43.4309, | |
| "eval_samples_per_second": 21.943, | |
| "eval_steps_per_second": 0.691, | |
| "eval_wer": 0.5759055296922903, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 26.58, | |
| "learning_rate": 5.489324324324324e-05, | |
| "loss": 1.0695, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 27.85, | |
| "learning_rate": 5.394729729729729e-05, | |
| "loss": 1.065, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 29.11, | |
| "learning_rate": 5.300135135135134e-05, | |
| "loss": 1.0534, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 30.38, | |
| "learning_rate": 5.20554054054054e-05, | |
| "loss": 1.0323, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 31.65, | |
| "learning_rate": 5.110945945945946e-05, | |
| "loss": 1.044, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 31.65, | |
| "eval_loss": 0.49577832221984863, | |
| "eval_runtime": 42.5512, | |
| "eval_samples_per_second": 22.397, | |
| "eval_steps_per_second": 0.705, | |
| "eval_wer": 0.5569433405245827, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 32.91, | |
| "learning_rate": 5.0163513513513505e-05, | |
| "loss": 1.0187, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 34.18, | |
| "learning_rate": 4.921756756756756e-05, | |
| "loss": 1.0094, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 35.44, | |
| "learning_rate": 4.8271621621621615e-05, | |
| "loss": 1.0002, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 36.71, | |
| "learning_rate": 4.732567567567567e-05, | |
| "loss": 0.9978, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 37.97, | |
| "learning_rate": 4.6379729729729725e-05, | |
| "loss": 0.9915, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 37.97, | |
| "eval_loss": 0.49712273478507996, | |
| "eval_runtime": 41.1197, | |
| "eval_samples_per_second": 23.176, | |
| "eval_steps_per_second": 0.73, | |
| "eval_wer": 0.5350289542409447, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 39.24, | |
| "learning_rate": 4.5433783783783776e-05, | |
| "loss": 0.9786, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 40.51, | |
| "learning_rate": 4.4487837837837834e-05, | |
| "loss": 0.9819, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 41.77, | |
| "learning_rate": 4.354189189189189e-05, | |
| "loss": 0.9735, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 43.04, | |
| "learning_rate": 4.259594594594594e-05, | |
| "loss": 0.9624, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 44.3, | |
| "learning_rate": 4.1649999999999996e-05, | |
| "loss": 0.9429, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 44.3, | |
| "eval_loss": 0.48288407921791077, | |
| "eval_runtime": 41.6865, | |
| "eval_samples_per_second": 22.861, | |
| "eval_steps_per_second": 0.72, | |
| "eval_wer": 0.5228795276484615, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 45.57, | |
| "learning_rate": 4.070405405405405e-05, | |
| "loss": 0.9448, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 46.84, | |
| "learning_rate": 3.9758108108108106e-05, | |
| "loss": 0.9383, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 48.1, | |
| "learning_rate": 3.8812162162162164e-05, | |
| "loss": 0.939, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 49.37, | |
| "learning_rate": 3.786621621621621e-05, | |
| "loss": 0.9322, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 50.63, | |
| "learning_rate": 3.692027027027027e-05, | |
| "loss": 0.9266, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 50.63, | |
| "eval_loss": 0.4515039622783661, | |
| "eval_runtime": 41.6834, | |
| "eval_samples_per_second": 22.863, | |
| "eval_steps_per_second": 0.72, | |
| "eval_wer": 0.5074372658112865, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 51.9, | |
| "learning_rate": 3.5974324324324326e-05, | |
| "loss": 0.9147, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 53.16, | |
| "learning_rate": 3.502837837837838e-05, | |
| "loss": 0.9129, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 54.43, | |
| "learning_rate": 3.4091891891891893e-05, | |
| "loss": 0.9071, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 55.7, | |
| "learning_rate": 3.3145945945945945e-05, | |
| "loss": 0.9098, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 56.96, | |
| "learning_rate": 3.220945945945946e-05, | |
| "loss": 0.8965, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 56.96, | |
| "eval_loss": 0.45992422103881836, | |
| "eval_runtime": 41.9829, | |
| "eval_samples_per_second": 22.7, | |
| "eval_steps_per_second": 0.715, | |
| "eval_wer": 0.5039173384807539, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 58.23, | |
| "learning_rate": 3.126351351351351e-05, | |
| "loss": 0.8992, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 59.49, | |
| "learning_rate": 3.0327027027027026e-05, | |
| "loss": 0.9006, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 60.76, | |
| "learning_rate": 2.9381081081081077e-05, | |
| "loss": 0.8931, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 62.03, | |
| "learning_rate": 2.8435135135135132e-05, | |
| "loss": 0.8846, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 63.29, | |
| "learning_rate": 2.7489189189189187e-05, | |
| "loss": 0.878, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 63.29, | |
| "eval_loss": 0.4735390543937683, | |
| "eval_runtime": 41.3368, | |
| "eval_samples_per_second": 23.054, | |
| "eval_steps_per_second": 0.726, | |
| "eval_wer": 0.4954013852617236, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 64.56, | |
| "learning_rate": 2.6543243243243242e-05, | |
| "loss": 0.8696, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 65.82, | |
| "learning_rate": 2.5597297297297294e-05, | |
| "loss": 0.8756, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 67.09, | |
| "learning_rate": 2.465135135135135e-05, | |
| "loss": 0.8588, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 68.35, | |
| "learning_rate": 2.3705405405405404e-05, | |
| "loss": 0.8616, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 69.62, | |
| "learning_rate": 2.275945945945946e-05, | |
| "loss": 0.8494, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 69.62, | |
| "eval_loss": 0.44596442580223083, | |
| "eval_runtime": 41.8864, | |
| "eval_samples_per_second": 22.752, | |
| "eval_steps_per_second": 0.716, | |
| "eval_wer": 0.4877938003860565, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 70.89, | |
| "learning_rate": 2.181351351351351e-05, | |
| "loss": 0.8535, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 72.15, | |
| "learning_rate": 2.0867567567567565e-05, | |
| "loss": 0.8511, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 73.42, | |
| "learning_rate": 1.992162162162162e-05, | |
| "loss": 0.8436, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 74.68, | |
| "learning_rate": 1.8975675675675675e-05, | |
| "loss": 0.8527, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 75.95, | |
| "learning_rate": 1.8029729729729727e-05, | |
| "loss": 0.8343, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 75.95, | |
| "eval_loss": 0.4510088264942169, | |
| "eval_runtime": 41.5079, | |
| "eval_samples_per_second": 22.959, | |
| "eval_steps_per_second": 0.723, | |
| "eval_wer": 0.479504939252867, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 77.22, | |
| "learning_rate": 1.708378378378378e-05, | |
| "loss": 0.8474, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 78.48, | |
| "learning_rate": 1.6137837837837837e-05, | |
| "loss": 0.83, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 79.75, | |
| "learning_rate": 1.5191891891891891e-05, | |
| "loss": 0.8168, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 81.01, | |
| "learning_rate": 1.4245945945945945e-05, | |
| "loss": 0.8363, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 82.28, | |
| "learning_rate": 1.33e-05, | |
| "loss": 0.8236, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 82.28, | |
| "eval_loss": 0.4537811577320099, | |
| "eval_runtime": 40.81, | |
| "eval_samples_per_second": 23.352, | |
| "eval_steps_per_second": 0.735, | |
| "eval_wer": 0.47893720903826503, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 83.54, | |
| "learning_rate": 1.2354054054054053e-05, | |
| "loss": 0.8189, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 84.81, | |
| "learning_rate": 1.1408108108108108e-05, | |
| "loss": 0.8121, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 86.08, | |
| "learning_rate": 1.0462162162162161e-05, | |
| "loss": 0.8098, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 87.34, | |
| "learning_rate": 9.516216216216216e-06, | |
| "loss": 0.8091, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 88.61, | |
| "learning_rate": 8.57027027027027e-06, | |
| "loss": 0.8069, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 88.61, | |
| "eval_loss": 0.45264118909835815, | |
| "eval_runtime": 41.4122, | |
| "eval_samples_per_second": 23.013, | |
| "eval_steps_per_second": 0.724, | |
| "eval_wer": 0.47484955149313046, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 89.87, | |
| "learning_rate": 7.6243243243243236e-06, | |
| "loss": 0.8042, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 91.14, | |
| "learning_rate": 6.678378378378378e-06, | |
| "loss": 0.8042, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 92.41, | |
| "learning_rate": 5.732432432432432e-06, | |
| "loss": 0.7968, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 93.67, | |
| "learning_rate": 4.786486486486486e-06, | |
| "loss": 0.798, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 94.94, | |
| "learning_rate": 3.84054054054054e-06, | |
| "loss": 0.7958, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 94.94, | |
| "eval_loss": 0.449596643447876, | |
| "eval_runtime": 41.0608, | |
| "eval_samples_per_second": 23.209, | |
| "eval_steps_per_second": 0.731, | |
| "eval_wer": 0.4699670716475531, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 96.2, | |
| "learning_rate": 2.8945945945945945e-06, | |
| "loss": 0.7953, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 97.47, | |
| "learning_rate": 1.9486486486486487e-06, | |
| "loss": 0.7917, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 98.73, | |
| "learning_rate": 1.0027027027027026e-06, | |
| "loss": 0.7881, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 5.675675675675675e-08, | |
| "loss": 0.795, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 7900, | |
| "total_flos": 3.986125981994791e+19, | |
| "train_loss": 1.2487838832034341, | |
| "train_runtime": 15247.9542, | |
| "train_samples_per_second": 16.383, | |
| "train_steps_per_second": 0.518 | |
| } | |
| ], | |
| "max_steps": 7900, | |
| "num_train_epochs": 100, | |
| "total_flos": 3.986125981994791e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |