| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.6775106082036775, | |
| "eval_steps": 100, | |
| "global_step": 2600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.14144271570014144, | |
| "eval_loss": 3.626366138458252, | |
| "eval_runtime": 164.728, | |
| "eval_samples_per_second": 34.335, | |
| "eval_steps_per_second": 4.292, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2828854314002829, | |
| "eval_loss": 3.1079769134521484, | |
| "eval_runtime": 161.0986, | |
| "eval_samples_per_second": 35.109, | |
| "eval_steps_per_second": 4.389, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4243281471004243, | |
| "eval_loss": 2.5440056324005127, | |
| "eval_runtime": 161.942, | |
| "eval_samples_per_second": 34.926, | |
| "eval_steps_per_second": 4.366, | |
| "eval_wer": 0.9862761130617415, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5657708628005658, | |
| "eval_loss": 1.2381267547607422, | |
| "eval_runtime": 162.1273, | |
| "eval_samples_per_second": 34.886, | |
| "eval_steps_per_second": 4.361, | |
| "eval_wer": 0.8110079681913649, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7072135785007072, | |
| "grad_norm": 3.555830955505371, | |
| "learning_rate": 0.00029699999999999996, | |
| "loss": 3.4586, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7072135785007072, | |
| "eval_loss": 1.0035638809204102, | |
| "eval_runtime": 161.1941, | |
| "eval_samples_per_second": 35.088, | |
| "eval_steps_per_second": 4.386, | |
| "eval_wer": 0.7184358616709153, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8486562942008486, | |
| "eval_loss": 0.8524216413497925, | |
| "eval_runtime": 162.0442, | |
| "eval_samples_per_second": 34.904, | |
| "eval_steps_per_second": 4.363, | |
| "eval_wer": 0.6379042213778398, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9900990099009901, | |
| "eval_loss": 0.7411457300186157, | |
| "eval_runtime": 162.8526, | |
| "eval_samples_per_second": 34.731, | |
| "eval_steps_per_second": 4.341, | |
| "eval_wer": 0.5673288121462813, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.1315417256011315, | |
| "eval_loss": 0.6202757954597473, | |
| "eval_runtime": 162.5052, | |
| "eval_samples_per_second": 34.805, | |
| "eval_steps_per_second": 4.351, | |
| "eval_wer": 0.48383114488640916, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.272984441301273, | |
| "eval_loss": 0.6016646027565002, | |
| "eval_runtime": 162.6225, | |
| "eval_samples_per_second": 34.78, | |
| "eval_steps_per_second": 4.347, | |
| "eval_wer": 0.48198739839353566, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.4144271570014144, | |
| "grad_norm": 1.029418706893921, | |
| "learning_rate": 0.00022928571428571426, | |
| "loss": 0.8146, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.4144271570014144, | |
| "eval_loss": 0.5637466311454773, | |
| "eval_runtime": 163.5441, | |
| "eval_samples_per_second": 34.584, | |
| "eval_steps_per_second": 4.323, | |
| "eval_wer": 0.4461706186971927, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5558698727015559, | |
| "eval_loss": 0.5380699038505554, | |
| "eval_runtime": 163.2757, | |
| "eval_samples_per_second": 34.641, | |
| "eval_steps_per_second": 4.33, | |
| "eval_wer": 0.4207589822519359, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6973125884016973, | |
| "eval_loss": 0.5155439972877502, | |
| "eval_runtime": 163.2795, | |
| "eval_samples_per_second": 34.64, | |
| "eval_steps_per_second": 4.33, | |
| "eval_wer": 0.4043416221762622, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.8387553041018387, | |
| "eval_loss": 0.4857912063598633, | |
| "eval_runtime": 163.399, | |
| "eval_samples_per_second": 34.615, | |
| "eval_steps_per_second": 4.327, | |
| "eval_wer": 0.39034518140862234, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.9801980198019802, | |
| "eval_loss": 0.47579917311668396, | |
| "eval_runtime": 162.8853, | |
| "eval_samples_per_second": 34.724, | |
| "eval_steps_per_second": 4.34, | |
| "eval_wer": 0.38234492488737115, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.1216407355021216, | |
| "grad_norm": 1.043891429901123, | |
| "learning_rate": 0.0001582857142857143, | |
| "loss": 0.6294, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.1216407355021216, | |
| "eval_loss": 0.46035128831863403, | |
| "eval_runtime": 163.2756, | |
| "eval_samples_per_second": 34.641, | |
| "eval_steps_per_second": 4.33, | |
| "eval_wer": 0.36421207894441504, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.263083451202263, | |
| "eval_loss": 0.45305728912353516, | |
| "eval_runtime": 164.2898, | |
| "eval_samples_per_second": 34.427, | |
| "eval_steps_per_second": 4.303, | |
| "eval_wer": 0.35355041444214647, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.4045261669024045, | |
| "eval_loss": 0.44349414110183716, | |
| "eval_runtime": 164.2629, | |
| "eval_samples_per_second": 34.433, | |
| "eval_steps_per_second": 4.304, | |
| "eval_wer": 0.3510813973995158, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.545968882602546, | |
| "eval_loss": 0.4366327226161957, | |
| "eval_runtime": 164.0684, | |
| "eval_samples_per_second": 34.473, | |
| "eval_steps_per_second": 4.309, | |
| "eval_wer": 0.3498949866127972, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.6874115983026874, | |
| "eval_loss": 0.43093863129615784, | |
| "eval_runtime": 164.1859, | |
| "eval_samples_per_second": 34.449, | |
| "eval_steps_per_second": 4.306, | |
| "eval_wer": 0.34513331088772387, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.828854314002829, | |
| "grad_norm": 0.763507604598999, | |
| "learning_rate": 8.728571428571428e-05, | |
| "loss": 0.4914, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.828854314002829, | |
| "eval_loss": 0.4252130389213562, | |
| "eval_runtime": 164.7353, | |
| "eval_samples_per_second": 34.334, | |
| "eval_steps_per_second": 4.292, | |
| "eval_wer": 0.3367803376460969, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.9702970297029703, | |
| "eval_loss": 0.42176735401153564, | |
| "eval_runtime": 163.889, | |
| "eval_samples_per_second": 34.511, | |
| "eval_steps_per_second": 4.314, | |
| "eval_wer": 0.3285075272954644, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.1117397454031117, | |
| "eval_loss": 0.42082250118255615, | |
| "eval_runtime": 165.1043, | |
| "eval_samples_per_second": 34.257, | |
| "eval_steps_per_second": 4.282, | |
| "eval_wer": 0.32510862071729757, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.253182461103253, | |
| "eval_loss": 0.41444018483161926, | |
| "eval_runtime": 164.0969, | |
| "eval_samples_per_second": 34.467, | |
| "eval_steps_per_second": 4.308, | |
| "eval_wer": 0.32363362352299874, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.3946251768033946, | |
| "eval_loss": 0.41402822732925415, | |
| "eval_runtime": 164.4996, | |
| "eval_samples_per_second": 34.383, | |
| "eval_steps_per_second": 4.298, | |
| "eval_wer": 0.31654722395908486, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.536067892503536, | |
| "grad_norm": 0.8644358515739441, | |
| "learning_rate": 1.614285714285714e-05, | |
| "loss": 0.4011, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.536067892503536, | |
| "eval_loss": 0.4133088290691376, | |
| "eval_runtime": 164.6412, | |
| "eval_samples_per_second": 34.353, | |
| "eval_steps_per_second": 4.294, | |
| "eval_wer": 0.31574559504913985, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.6775106082036775, | |
| "eval_loss": 0.4086272120475769, | |
| "eval_runtime": 163.5841, | |
| "eval_samples_per_second": 34.575, | |
| "eval_steps_per_second": 4.322, | |
| "eval_wer": 0.3122344604235807, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 3.6775106082036775, | |
| "step": 2600, | |
| "total_flos": 9.93185228450562e+18, | |
| "train_loss": 1.1287707856985238, | |
| "train_runtime": 7180.9983, | |
| "train_samples_per_second": 11.586, | |
| "train_steps_per_second": 0.362 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2600, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 400, | |
| "total_flos": 9.93185228450562e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |