| { | |
| "best_metric": 98.0078125, | |
| "best_model_checkpoint": "./whisper-small-ha-v9/checkpoint-1000", | |
| "epoch": 12.738853503184714, | |
| "eval_steps": 500, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1592356687898089, | |
| "grad_norm": 14.99692440032959, | |
| "learning_rate": 0.00021, | |
| "loss": 2.5992, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3184713375796178, | |
| "grad_norm": 17.429601669311523, | |
| "learning_rate": 0.00045000000000000004, | |
| "loss": 2.8909, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.47770700636942676, | |
| "grad_norm": 18.359071731567383, | |
| "learning_rate": 0.0005, | |
| "loss": 3.5945, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6369426751592356, | |
| "grad_norm": 11.03101634979248, | |
| "learning_rate": 0.0005, | |
| "loss": 3.7546, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7961783439490446, | |
| "grad_norm": 9.160442352294922, | |
| "learning_rate": 0.0005, | |
| "loss": 3.53, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.9554140127388535, | |
| "grad_norm": 8.277521133422852, | |
| "learning_rate": 0.0005, | |
| "loss": 3.516, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.1146496815286624, | |
| "grad_norm": 8.703987121582031, | |
| "learning_rate": 0.0005, | |
| "loss": 2.8911, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.2738853503184713, | |
| "grad_norm": 8.430550575256348, | |
| "learning_rate": 0.0005, | |
| "loss": 2.7461, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.4331210191082802, | |
| "grad_norm": 8.455479621887207, | |
| "learning_rate": 0.0005, | |
| "loss": 2.8827, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.5923566878980893, | |
| "grad_norm": 10.029267311096191, | |
| "learning_rate": 0.0005, | |
| "loss": 2.9323, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.7515923566878981, | |
| "grad_norm": 7.152008533477783, | |
| "learning_rate": 0.0005, | |
| "loss": 2.8411, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.910828025477707, | |
| "grad_norm": 7.405454635620117, | |
| "learning_rate": 0.0005, | |
| "loss": 2.9588, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.070063694267516, | |
| "grad_norm": 7.2899675369262695, | |
| "learning_rate": 0.0005, | |
| "loss": 2.4946, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.229299363057325, | |
| "grad_norm": 8.068443298339844, | |
| "learning_rate": 0.0005, | |
| "loss": 2.0882, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.388535031847134, | |
| "grad_norm": 8.143331527709961, | |
| "learning_rate": 0.0005, | |
| "loss": 2.1759, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.5477707006369426, | |
| "grad_norm": 8.333662033081055, | |
| "learning_rate": 0.0005, | |
| "loss": 2.2816, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.7070063694267517, | |
| "grad_norm": 7.931263446807861, | |
| "learning_rate": 0.0005, | |
| "loss": 2.3499, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.8662420382165603, | |
| "grad_norm": 8.108153343200684, | |
| "learning_rate": 0.0005, | |
| "loss": 2.3946, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.0254777070063694, | |
| "grad_norm": 7.962264060974121, | |
| "learning_rate": 0.0005, | |
| "loss": 2.1774, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 3.1847133757961785, | |
| "grad_norm": 8.008493423461914, | |
| "learning_rate": 0.0005, | |
| "loss": 1.5785, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.1847133757961785, | |
| "eval_loss": 3.8839268684387207, | |
| "eval_runtime": 239.0145, | |
| "eval_samples_per_second": 2.761, | |
| "eval_steps_per_second": 0.176, | |
| "eval_wer": 99.84375, | |
| "eval_wer_ortho": 99.55930254838091, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.343949044585987, | |
| "grad_norm": 7.7201032638549805, | |
| "learning_rate": 0.0005, | |
| "loss": 1.7432, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 3.5031847133757963, | |
| "grad_norm": 7.00828742980957, | |
| "learning_rate": 0.0005, | |
| "loss": 1.8165, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.662420382165605, | |
| "grad_norm": 7.801667213439941, | |
| "learning_rate": 0.0005, | |
| "loss": 1.8769, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 3.821656050955414, | |
| "grad_norm": 6.696052074432373, | |
| "learning_rate": 0.0005, | |
| "loss": 1.9189, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.980891719745223, | |
| "grad_norm": 7.849315643310547, | |
| "learning_rate": 0.0005, | |
| "loss": 1.9248, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 4.140127388535032, | |
| "grad_norm": 6.837538719177246, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3935, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 4.2993630573248405, | |
| "grad_norm": 6.533350944519043, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3859, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 4.45859872611465, | |
| "grad_norm": 7.38162899017334, | |
| "learning_rate": 0.0005, | |
| "loss": 1.4676, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.617834394904459, | |
| "grad_norm": 7.072102069854736, | |
| "learning_rate": 0.0005, | |
| "loss": 1.5653, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 4.777070063694268, | |
| "grad_norm": 7.383370399475098, | |
| "learning_rate": 0.0005, | |
| "loss": 1.6075, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.936305732484076, | |
| "grad_norm": 6.835177898406982, | |
| "learning_rate": 0.0005, | |
| "loss": 1.6546, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 5.095541401273885, | |
| "grad_norm": 6.938526153564453, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3298, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.254777070063694, | |
| "grad_norm": 7.478129863739014, | |
| "learning_rate": 0.0005, | |
| "loss": 1.2336, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 5.414012738853503, | |
| "grad_norm": 6.950467109680176, | |
| "learning_rate": 0.0005, | |
| "loss": 1.2652, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 5.573248407643312, | |
| "grad_norm": 8.092499732971191, | |
| "learning_rate": 0.0005, | |
| "loss": 1.4009, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 5.732484076433121, | |
| "grad_norm": 6.488431930541992, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3538, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.89171974522293, | |
| "grad_norm": 6.798085689544678, | |
| "learning_rate": 0.0005, | |
| "loss": 1.4269, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 6.050955414012739, | |
| "grad_norm": 6.3589043617248535, | |
| "learning_rate": 0.0005, | |
| "loss": 1.2951, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 6.210191082802548, | |
| "grad_norm": 6.020321369171143, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0916, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 6.369426751592357, | |
| "grad_norm": 6.381227493286133, | |
| "learning_rate": 0.0005, | |
| "loss": 1.1623, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.369426751592357, | |
| "eval_loss": 4.484735488891602, | |
| "eval_runtime": 243.1604, | |
| "eval_samples_per_second": 2.714, | |
| "eval_steps_per_second": 0.173, | |
| "eval_wer": 98.0078125, | |
| "eval_wer_ortho": 97.75819122437248, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.528662420382165, | |
| "grad_norm": 6.898542404174805, | |
| "learning_rate": 0.0005, | |
| "loss": 1.2127, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 6.687898089171974, | |
| "grad_norm": 6.266862392425537, | |
| "learning_rate": 0.0005, | |
| "loss": 1.2291, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 6.8471337579617835, | |
| "grad_norm": 6.253779888153076, | |
| "learning_rate": 0.0005, | |
| "loss": 1.2292, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 7.006369426751593, | |
| "grad_norm": 5.176928520202637, | |
| "learning_rate": 0.0005, | |
| "loss": 1.2593, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 7.165605095541402, | |
| "grad_norm": 4.826560020446777, | |
| "learning_rate": 0.0005, | |
| "loss": 1.008, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 7.32484076433121, | |
| "grad_norm": 5.738828659057617, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0366, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 7.484076433121019, | |
| "grad_norm": 5.866888523101807, | |
| "learning_rate": 0.0005, | |
| "loss": 1.1405, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 7.643312101910828, | |
| "grad_norm": 6.350291728973389, | |
| "learning_rate": 0.0005, | |
| "loss": 1.1344, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 7.802547770700637, | |
| "grad_norm": 6.462769508361816, | |
| "learning_rate": 0.0005, | |
| "loss": 1.1296, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 7.961783439490446, | |
| "grad_norm": 6.252405166625977, | |
| "learning_rate": 0.0005, | |
| "loss": 1.1576, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 8.121019108280255, | |
| "grad_norm": 6.970676898956299, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9843, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 8.280254777070065, | |
| "grad_norm": 5.744741916656494, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9896, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 8.439490445859873, | |
| "grad_norm": 6.012212753295898, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0226, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 8.598726114649681, | |
| "grad_norm": 5.396963596343994, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0337, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 8.757961783439491, | |
| "grad_norm": 6.041420936584473, | |
| "learning_rate": 0.0005, | |
| "loss": 1.1026, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 8.9171974522293, | |
| "grad_norm": 7.039705276489258, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0957, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 9.07643312101911, | |
| "grad_norm": 5.3199028968811035, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0184, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 9.235668789808917, | |
| "grad_norm": 5.058565616607666, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9048, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 9.394904458598726, | |
| "grad_norm": 5.007670879364014, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9577, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 9.554140127388536, | |
| "grad_norm": 5.517505168914795, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9893, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 9.554140127388536, | |
| "eval_loss": 4.792229652404785, | |
| "eval_runtime": 245.7983, | |
| "eval_samples_per_second": 2.685, | |
| "eval_steps_per_second": 0.171, | |
| "eval_wer": 107.6171875, | |
| "eval_wer_ortho": 108.7373059973175, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 9.713375796178344, | |
| "grad_norm": 5.215830326080322, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0183, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 9.872611464968152, | |
| "grad_norm": 4.925009250640869, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0158, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 10.031847133757962, | |
| "grad_norm": 5.344029903411865, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0167, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 10.19108280254777, | |
| "grad_norm": 5.701788902282715, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8659, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 10.35031847133758, | |
| "grad_norm": 5.5644707679748535, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9147, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 10.509554140127388, | |
| "grad_norm": 5.331553936004639, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9256, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 10.668789808917197, | |
| "grad_norm": 5.594724655151367, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9863, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 10.828025477707007, | |
| "grad_norm": 5.949446201324463, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9587, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 10.987261146496815, | |
| "grad_norm": 4.683850288391113, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9825, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 11.146496815286625, | |
| "grad_norm": 4.596772193908691, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8346, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 11.305732484076433, | |
| "grad_norm": 4.516598701477051, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8364, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 11.464968152866241, | |
| "grad_norm": 4.428603649139404, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8787, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 11.624203821656051, | |
| "grad_norm": 5.087082862854004, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9147, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 11.78343949044586, | |
| "grad_norm": 5.706089496612549, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9533, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 11.94267515923567, | |
| "grad_norm": 5.720147132873535, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9424, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 12.101910828025478, | |
| "grad_norm": 4.992929458618164, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8299, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 12.261146496815286, | |
| "grad_norm": 4.936596393585205, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8422, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 12.420382165605096, | |
| "grad_norm": 4.403645038604736, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8743, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 12.579617834394904, | |
| "grad_norm": 4.9296793937683105, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8704, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 12.738853503184714, | |
| "grad_norm": 5.615696430206299, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8816, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 12.738853503184714, | |
| "eval_loss": 4.957520008087158, | |
| "eval_runtime": 246.0161, | |
| "eval_samples_per_second": 2.683, | |
| "eval_steps_per_second": 0.171, | |
| "eval_wer": 108.10546875, | |
| "eval_wer_ortho": 110.34680973366547, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 2041, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 13, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.22088071102464e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |