| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 50.0, | |
| "global_step": 46900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 1.74475e-05, | |
| "loss": 7.5225, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 3.49475e-05, | |
| "loss": 2.9185, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 5.2447499999999995e-05, | |
| "loss": 1.5385, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "learning_rate": 6.99475e-05, | |
| "loss": 1.2881, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "eval_loss": 0.37639039754867554, | |
| "eval_runtime": 433.5071, | |
| "eval_samples_per_second": 21.185, | |
| "eval_steps_per_second": 1.324, | |
| "eval_wer": 0.6460503785612609, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 5.33, | |
| "learning_rate": 6.837482517482516e-05, | |
| "loss": 1.2081, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "learning_rate": 6.674475524475524e-05, | |
| "loss": 1.1501, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 7.46, | |
| "learning_rate": 6.511305361305361e-05, | |
| "loss": 1.1115, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 8.53, | |
| "learning_rate": 6.348135198135198e-05, | |
| "loss": 1.0767, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 8.53, | |
| "eval_loss": 0.26572874188423157, | |
| "eval_runtime": 423.456, | |
| "eval_samples_per_second": 21.688, | |
| "eval_steps_per_second": 1.356, | |
| "eval_wer": 0.5164111979865079, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 9.59, | |
| "learning_rate": 6.185128205128204e-05, | |
| "loss": 1.0506, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 10.66, | |
| "learning_rate": 6.021958041958042e-05, | |
| "loss": 1.0261, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 11.73, | |
| "learning_rate": 5.858787878787878e-05, | |
| "loss": 1.0029, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 12.79, | |
| "learning_rate": 5.695617715617715e-05, | |
| "loss": 0.9841, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 12.79, | |
| "eval_loss": 0.23304273188114166, | |
| "eval_runtime": 424.6539, | |
| "eval_samples_per_second": 21.627, | |
| "eval_steps_per_second": 1.352, | |
| "eval_wer": 0.4444742433932292, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 13.86, | |
| "learning_rate": 5.532447552447552e-05, | |
| "loss": 0.9694, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 14.93, | |
| "learning_rate": 5.369603729603729e-05, | |
| "loss": 0.9593, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 15.99, | |
| "learning_rate": 5.206433566433566e-05, | |
| "loss": 0.9428, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 17.06, | |
| "learning_rate": 5.0432634032634027e-05, | |
| "loss": 0.9274, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 17.06, | |
| "eval_loss": 0.213353231549263, | |
| "eval_runtime": 424.1743, | |
| "eval_samples_per_second": 21.651, | |
| "eval_steps_per_second": 1.353, | |
| "eval_wer": 0.39289913972727086, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 18.12, | |
| "learning_rate": 4.8800932400932396e-05, | |
| "loss": 0.9171, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 19.19, | |
| "learning_rate": 4.7170862470862466e-05, | |
| "loss": 0.9057, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 20.26, | |
| "learning_rate": 4.5539160839160836e-05, | |
| "loss": 0.8881, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 21.32, | |
| "learning_rate": 4.3909090909090906e-05, | |
| "loss": 0.8781, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 21.32, | |
| "eval_loss": 0.19452312588691711, | |
| "eval_runtime": 418.381, | |
| "eval_samples_per_second": 21.951, | |
| "eval_steps_per_second": 1.372, | |
| "eval_wer": 0.3886287211437295, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 22.39, | |
| "learning_rate": 4.2277389277389276e-05, | |
| "loss": 0.8633, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 23.45, | |
| "learning_rate": 4.064731934731934e-05, | |
| "loss": 0.8537, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 24.52, | |
| "learning_rate": 3.901561771561771e-05, | |
| "loss": 0.8462, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 25.59, | |
| "learning_rate": 3.738391608391608e-05, | |
| "loss": 0.8381, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 25.59, | |
| "eval_loss": 0.184038907289505, | |
| "eval_runtime": 425.8794, | |
| "eval_samples_per_second": 21.565, | |
| "eval_steps_per_second": 1.348, | |
| "eval_wer": 0.3736719410806016, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 26.65, | |
| "learning_rate": 3.575221445221445e-05, | |
| "loss": 0.8237, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 27.72, | |
| "learning_rate": 3.412214452214452e-05, | |
| "loss": 0.819, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 28.78, | |
| "learning_rate": 3.249207459207459e-05, | |
| "loss": 0.8114, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 29.85, | |
| "learning_rate": 3.086037296037296e-05, | |
| "loss": 0.8054, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 29.85, | |
| "eval_loss": 0.17562612891197205, | |
| "eval_runtime": 420.9127, | |
| "eval_samples_per_second": 21.819, | |
| "eval_steps_per_second": 1.364, | |
| "eval_wer": 0.3523198481628948, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 30.92, | |
| "learning_rate": 2.9228671328671326e-05, | |
| "loss": 0.7997, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 31.98, | |
| "learning_rate": 2.7598601398601396e-05, | |
| "loss": 0.7867, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 33.05, | |
| "learning_rate": 2.5966899766899766e-05, | |
| "loss": 0.7764, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 34.12, | |
| "learning_rate": 2.4336829836829836e-05, | |
| "loss": 0.7763, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 34.12, | |
| "eval_loss": 0.17449165880680084, | |
| "eval_runtime": 413.6966, | |
| "eval_samples_per_second": 22.2, | |
| "eval_steps_per_second": 1.387, | |
| "eval_wer": 0.3299156231304025, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 35.18, | |
| "learning_rate": 2.2705128205128205e-05, | |
| "loss": 0.7676, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 36.25, | |
| "learning_rate": 2.1075058275058275e-05, | |
| "loss": 0.7571, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 37.31, | |
| "learning_rate": 1.944335664335664e-05, | |
| "loss": 0.7556, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 38.38, | |
| "learning_rate": 1.7814918414918412e-05, | |
| "loss": 0.7474, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 38.38, | |
| "eval_loss": 0.1677420288324356, | |
| "eval_runtime": 414.2329, | |
| "eval_samples_per_second": 22.171, | |
| "eval_steps_per_second": 1.386, | |
| "eval_wer": 0.30740824789057825, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 39.45, | |
| "learning_rate": 1.6183216783216782e-05, | |
| "loss": 0.7411, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 40.51, | |
| "learning_rate": 1.455151515151515e-05, | |
| "loss": 0.7369, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 41.58, | |
| "learning_rate": 1.2919813519813518e-05, | |
| "loss": 0.7309, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 42.64, | |
| "learning_rate": 1.1288111888111887e-05, | |
| "loss": 0.7298, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 42.64, | |
| "eval_loss": 0.1649060845375061, | |
| "eval_runtime": 413.307, | |
| "eval_samples_per_second": 22.221, | |
| "eval_steps_per_second": 1.389, | |
| "eval_wer": 0.2962886555401976, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 43.71, | |
| "learning_rate": 9.65967365967366e-06, | |
| "loss": 0.7201, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 44.78, | |
| "learning_rate": 8.027972027972028e-06, | |
| "loss": 0.72, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 45.84, | |
| "learning_rate": 6.396270396270396e-06, | |
| "loss": 0.7092, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 46.91, | |
| "learning_rate": 4.764568764568764e-06, | |
| "loss": 0.7125, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 46.91, | |
| "eval_loss": 0.161702960729599, | |
| "eval_runtime": 419.2465, | |
| "eval_samples_per_second": 21.906, | |
| "eval_steps_per_second": 1.369, | |
| "eval_wer": 0.29307036907144185, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 47.97, | |
| "learning_rate": 3.1344988344988342e-06, | |
| "loss": 0.7093, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 49.04, | |
| "learning_rate": 1.5044289044289044e-06, | |
| "loss": 0.7049, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "step": 46900, | |
| "total_flos": 2.4630771835403823e+20, | |
| "train_loss": 1.0648517145022656, | |
| "train_runtime": 95151.9589, | |
| "train_samples_per_second": 15.765, | |
| "train_steps_per_second": 0.493 | |
| } | |
| ], | |
| "max_steps": 46900, | |
| "num_train_epochs": 50, | |
| "total_flos": 2.4630771835403823e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |