| { | |
| "best_metric": 0.3741886615753174, | |
| "best_model_checkpoint": "/scratch/skscla001/speech/results/mms-1b-toigen-balanced-model/checkpoint-2400", | |
| "epoch": 12.053571428571429, | |
| "eval_steps": 100, | |
| "global_step": 2700, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.44642857142857145, | |
| "grad_norm": 3.282467842102051, | |
| "learning_rate": 0.00028199999999999997, | |
| "loss": 7.7726, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.44642857142857145, | |
| "eval_loss": 3.8109493255615234, | |
| "eval_runtime": 17.8235, | |
| "eval_samples_per_second": 11.446, | |
| "eval_steps_per_second": 2.861, | |
| "eval_wer": 0.9938476100331283, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "grad_norm": 2.164923667907715, | |
| "learning_rate": 0.00029574018126888213, | |
| "loss": 2.5726, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.8928571428571429, | |
| "eval_loss": 0.8106288313865662, | |
| "eval_runtime": 17.6933, | |
| "eval_samples_per_second": 11.53, | |
| "eval_steps_per_second": 2.882, | |
| "eval_wer": 0.616658778987222, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.3392857142857144, | |
| "grad_norm": 1.3534202575683594, | |
| "learning_rate": 0.0002912084592145015, | |
| "loss": 0.7986, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.3392857142857144, | |
| "eval_loss": 0.5409455299377441, | |
| "eval_runtime": 17.7815, | |
| "eval_samples_per_second": 11.473, | |
| "eval_steps_per_second": 2.868, | |
| "eval_wer": 0.5257927117841931, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 0.9854668378829956, | |
| "learning_rate": 0.00028667673716012085, | |
| "loss": 0.6324, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "eval_loss": 0.5256258845329285, | |
| "eval_runtime": 17.7084, | |
| "eval_samples_per_second": 11.52, | |
| "eval_steps_per_second": 2.88, | |
| "eval_wer": 0.5054424988168481, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.232142857142857, | |
| "grad_norm": 18.835981369018555, | |
| "learning_rate": 0.00028214501510574015, | |
| "loss": 0.603, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.232142857142857, | |
| "eval_loss": 0.4854464828968048, | |
| "eval_runtime": 17.7046, | |
| "eval_samples_per_second": 11.522, | |
| "eval_steps_per_second": 2.881, | |
| "eval_wer": 0.4831992427827733, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.678571428571429, | |
| "grad_norm": 32.54256820678711, | |
| "learning_rate": 0.0002776132930513595, | |
| "loss": 0.59, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.678571428571429, | |
| "eval_loss": 0.47332894802093506, | |
| "eval_runtime": 17.6431, | |
| "eval_samples_per_second": 11.563, | |
| "eval_steps_per_second": 2.891, | |
| "eval_wer": 0.4846190250828206, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "grad_norm": 1.3252086639404297, | |
| "learning_rate": 0.0002730815709969788, | |
| "loss": 0.5489, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "eval_loss": 0.4439888894557953, | |
| "eval_runtime": 17.8297, | |
| "eval_samples_per_second": 11.442, | |
| "eval_steps_per_second": 2.86, | |
| "eval_wer": 0.46568859441552296, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 0.5452375411987305, | |
| "learning_rate": 0.00026854984894259817, | |
| "loss": 0.5173, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "eval_loss": 0.43219566345214844, | |
| "eval_runtime": 17.6382, | |
| "eval_samples_per_second": 11.566, | |
| "eval_steps_per_second": 2.891, | |
| "eval_wer": 0.45764316138192146, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.017857142857143, | |
| "grad_norm": 0.7151035070419312, | |
| "learning_rate": 0.0002640181268882175, | |
| "loss": 0.5315, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.017857142857143, | |
| "eval_loss": 0.4285721480846405, | |
| "eval_runtime": 17.7542, | |
| "eval_samples_per_second": 11.49, | |
| "eval_steps_per_second": 2.873, | |
| "eval_wer": 0.44533838144817794, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.464285714285714, | |
| "grad_norm": 1.8268319368362427, | |
| "learning_rate": 0.0002594864048338368, | |
| "loss": 0.4912, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.464285714285714, | |
| "eval_loss": 0.42536306381225586, | |
| "eval_runtime": 17.765, | |
| "eval_samples_per_second": 11.483, | |
| "eval_steps_per_second": 2.871, | |
| "eval_wer": 0.4458116422148604, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.910714285714286, | |
| "grad_norm": 0.850709080696106, | |
| "learning_rate": 0.0002549546827794562, | |
| "loss": 0.4728, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 4.910714285714286, | |
| "eval_loss": 0.43455594778060913, | |
| "eval_runtime": 17.7563, | |
| "eval_samples_per_second": 11.489, | |
| "eval_steps_per_second": 2.872, | |
| "eval_wer": 0.44297207761476576, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 5.357142857142857, | |
| "grad_norm": 0.7361202836036682, | |
| "learning_rate": 0.00025042296072507554, | |
| "loss": 0.4989, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.357142857142857, | |
| "eval_loss": 0.40502411127090454, | |
| "eval_runtime": 17.6139, | |
| "eval_samples_per_second": 11.582, | |
| "eval_steps_per_second": 2.895, | |
| "eval_wer": 0.42924751538097494, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.803571428571429, | |
| "grad_norm": 1.305498719215393, | |
| "learning_rate": 0.00024589123867069484, | |
| "loss": 0.4661, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.803571428571429, | |
| "eval_loss": 0.4019148647785187, | |
| "eval_runtime": 17.792, | |
| "eval_samples_per_second": 11.466, | |
| "eval_steps_per_second": 2.866, | |
| "eval_wer": 0.4254614292475154, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 1.875386357307434, | |
| "learning_rate": 0.00024135951661631417, | |
| "loss": 0.4755, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "eval_loss": 0.4128676652908325, | |
| "eval_runtime": 17.7386, | |
| "eval_samples_per_second": 11.5, | |
| "eval_steps_per_second": 2.875, | |
| "eval_wer": 0.44486512068149553, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.696428571428571, | |
| "grad_norm": 1.3318761587142944, | |
| "learning_rate": 0.0002368277945619335, | |
| "loss": 0.4603, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.696428571428571, | |
| "eval_loss": 0.40455254912376404, | |
| "eval_runtime": 17.8115, | |
| "eval_samples_per_second": 11.453, | |
| "eval_steps_per_second": 2.863, | |
| "eval_wer": 0.4254614292475154, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "grad_norm": 1.7303593158721924, | |
| "learning_rate": 0.00023229607250755283, | |
| "loss": 0.4229, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "eval_loss": 0.3939039707183838, | |
| "eval_runtime": 17.7505, | |
| "eval_samples_per_second": 11.493, | |
| "eval_steps_per_second": 2.873, | |
| "eval_wer": 0.41504969238050166, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 7.589285714285714, | |
| "grad_norm": 0.9812105894088745, | |
| "learning_rate": 0.0002277643504531722, | |
| "loss": 0.455, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 7.589285714285714, | |
| "eval_loss": 0.41328728199005127, | |
| "eval_runtime": 17.7719, | |
| "eval_samples_per_second": 11.479, | |
| "eval_steps_per_second": 2.87, | |
| "eval_wer": 0.41552295314718407, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 8.035714285714286, | |
| "grad_norm": 0.6829022765159607, | |
| "learning_rate": 0.00022323262839879152, | |
| "loss": 0.4501, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 8.035714285714286, | |
| "eval_loss": 0.3978167176246643, | |
| "eval_runtime": 17.743, | |
| "eval_samples_per_second": 11.497, | |
| "eval_steps_per_second": 2.874, | |
| "eval_wer": 0.4065309985802177, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 8.482142857142858, | |
| "grad_norm": 1.3150678873062134, | |
| "learning_rate": 0.00021870090634441088, | |
| "loss": 0.45, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 8.482142857142858, | |
| "eval_loss": 0.3925248682498932, | |
| "eval_runtime": 17.6795, | |
| "eval_samples_per_second": 11.539, | |
| "eval_steps_per_second": 2.885, | |
| "eval_wer": 0.42309512541410316, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 8.928571428571429, | |
| "grad_norm": 1.0976217985153198, | |
| "learning_rate": 0.0002141691842900302, | |
| "loss": 0.4226, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.928571428571429, | |
| "eval_loss": 0.3901020586490631, | |
| "eval_runtime": 17.7714, | |
| "eval_samples_per_second": 11.479, | |
| "eval_steps_per_second": 2.87, | |
| "eval_wer": 0.4098438239469948, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 9.375, | |
| "grad_norm": 0.727407693862915, | |
| "learning_rate": 0.00020963746223564954, | |
| "loss": 0.3973, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 9.375, | |
| "eval_loss": 0.38098010420799255, | |
| "eval_runtime": 17.825, | |
| "eval_samples_per_second": 11.445, | |
| "eval_steps_per_second": 2.861, | |
| "eval_wer": 0.4055844770468528, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 9.821428571428571, | |
| "grad_norm": 2.031233072280884, | |
| "learning_rate": 0.00020510574018126884, | |
| "loss": 0.4038, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 9.821428571428571, | |
| "eval_loss": 0.41775575280189514, | |
| "eval_runtime": 17.8559, | |
| "eval_samples_per_second": 11.425, | |
| "eval_steps_per_second": 2.856, | |
| "eval_wer": 0.4117368670137246, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 10.267857142857142, | |
| "grad_norm": 1.3557627201080322, | |
| "learning_rate": 0.0002005740181268882, | |
| "loss": 0.4559, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 10.267857142857142, | |
| "eval_loss": 0.38752201199531555, | |
| "eval_runtime": 17.7454, | |
| "eval_samples_per_second": 11.496, | |
| "eval_steps_per_second": 2.874, | |
| "eval_wer": 0.40747752011358257, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 10.714285714285714, | |
| "grad_norm": 0.7480702996253967, | |
| "learning_rate": 0.00019604229607250753, | |
| "loss": 0.4399, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 10.714285714285714, | |
| "eval_loss": 0.3741886615753174, | |
| "eval_runtime": 17.7283, | |
| "eval_samples_per_second": 11.507, | |
| "eval_steps_per_second": 2.877, | |
| "eval_wer": 0.39895882631329865, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 11.160714285714286, | |
| "grad_norm": 1.065514087677002, | |
| "learning_rate": 0.00019151057401812688, | |
| "loss": 0.3545, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 11.160714285714286, | |
| "eval_loss": 0.38181087374687195, | |
| "eval_runtime": 17.8745, | |
| "eval_samples_per_second": 11.413, | |
| "eval_steps_per_second": 2.853, | |
| "eval_wer": 0.40132513014671084, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 11.607142857142858, | |
| "grad_norm": 1.19502854347229, | |
| "learning_rate": 0.0001869788519637462, | |
| "loss": 0.4452, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 11.607142857142858, | |
| "eval_loss": 0.3905617594718933, | |
| "eval_runtime": 17.8748, | |
| "eval_samples_per_second": 11.413, | |
| "eval_steps_per_second": 2.853, | |
| "eval_wer": 0.39801230477993377, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 12.053571428571429, | |
| "grad_norm": 0.8653120994567871, | |
| "learning_rate": 0.00018244712990936554, | |
| "loss": 0.4014, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 12.053571428571429, | |
| "eval_loss": 0.3751629889011383, | |
| "eval_runtime": 17.7431, | |
| "eval_samples_per_second": 11.497, | |
| "eval_steps_per_second": 2.874, | |
| "eval_wer": 0.39990534784666354, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 12.053571428571429, | |
| "step": 2700, | |
| "total_flos": 9.163482510982138e+18, | |
| "train_loss": 0.8344545293737341, | |
| "train_runtime": 2390.0152, | |
| "train_samples_per_second": 11.222, | |
| "train_steps_per_second": 2.812 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 6720, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.163482510982138e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |