| { | |
| "best_metric": 0.22929497063159943, | |
| "best_model_checkpoint": "/scratch/skscla001/results/mms-1b-all-bem-genbed-combined-adapter-test/checkpoint-4800", | |
| "epoch": 7.142857142857143, | |
| "eval_steps": 100, | |
| "global_step": 5200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.13736263736263737, | |
| "grad_norm": 3.0789577960968018, | |
| "learning_rate": 0.00028799999999999995, | |
| "loss": 6.4783, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.13736263736263737, | |
| "eval_loss": 0.6402517557144165, | |
| "eval_runtime": 132.5796, | |
| "eval_samples_per_second": 14.625, | |
| "eval_steps_per_second": 1.833, | |
| "eval_wer": 0.69, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.27472527472527475, | |
| "grad_norm": 1.5154528617858887, | |
| "learning_rate": 0.0002986752529898804, | |
| "loss": 0.5413, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.27472527472527475, | |
| "eval_loss": 0.32597342133522034, | |
| "eval_runtime": 131.6167, | |
| "eval_samples_per_second": 14.732, | |
| "eval_steps_per_second": 1.846, | |
| "eval_wer": 0.4931351351351351, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.41208791208791207, | |
| "grad_norm": 1.3768789768218994, | |
| "learning_rate": 0.0002972953081876725, | |
| "loss": 0.437, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.41208791208791207, | |
| "eval_loss": 0.30144402384757996, | |
| "eval_runtime": 132.6277, | |
| "eval_samples_per_second": 14.62, | |
| "eval_steps_per_second": 1.832, | |
| "eval_wer": 0.46778378378378377, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5494505494505495, | |
| "grad_norm": 2.0690324306488037, | |
| "learning_rate": 0.00029591536338546457, | |
| "loss": 0.4402, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5494505494505495, | |
| "eval_loss": 0.29820021986961365, | |
| "eval_runtime": 132.486, | |
| "eval_samples_per_second": 14.636, | |
| "eval_steps_per_second": 1.834, | |
| "eval_wer": 0.4817837837837838, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6868131868131868, | |
| "grad_norm": 1.9618090391159058, | |
| "learning_rate": 0.0002945354185832566, | |
| "loss": 0.4153, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6868131868131868, | |
| "eval_loss": 0.2935521602630615, | |
| "eval_runtime": 133.0965, | |
| "eval_samples_per_second": 14.568, | |
| "eval_steps_per_second": 1.826, | |
| "eval_wer": 0.47021621621621623, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8241758241758241, | |
| "grad_norm": 1.3698606491088867, | |
| "learning_rate": 0.0002931554737810487, | |
| "loss": 0.4154, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8241758241758241, | |
| "eval_loss": 0.2883508503437042, | |
| "eval_runtime": 131.9274, | |
| "eval_samples_per_second": 14.697, | |
| "eval_steps_per_second": 1.842, | |
| "eval_wer": 0.4492972972972973, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9615384615384616, | |
| "grad_norm": 1.5283702611923218, | |
| "learning_rate": 0.0002917755289788408, | |
| "loss": 0.3789, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.9615384615384616, | |
| "eval_loss": 0.2806306481361389, | |
| "eval_runtime": 132.4493, | |
| "eval_samples_per_second": 14.64, | |
| "eval_steps_per_second": 1.835, | |
| "eval_wer": 0.4521081081081081, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.098901098901099, | |
| "grad_norm": 0.4610430598258972, | |
| "learning_rate": 0.0002903955841766329, | |
| "loss": 0.3667, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.098901098901099, | |
| "eval_loss": 0.2763822078704834, | |
| "eval_runtime": 133.2255, | |
| "eval_samples_per_second": 14.554, | |
| "eval_steps_per_second": 1.824, | |
| "eval_wer": 0.4351891891891892, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2362637362637363, | |
| "grad_norm": 0.7730151414871216, | |
| "learning_rate": 0.000289015639374425, | |
| "loss": 0.3929, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.2362637362637363, | |
| "eval_loss": 0.27625423669815063, | |
| "eval_runtime": 133.7811, | |
| "eval_samples_per_second": 14.494, | |
| "eval_steps_per_second": 1.816, | |
| "eval_wer": 0.46535135135135136, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.3736263736263736, | |
| "grad_norm": 0.9868985414505005, | |
| "learning_rate": 0.0002876356945722171, | |
| "loss": 0.3847, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.3736263736263736, | |
| "eval_loss": 0.27053016424179077, | |
| "eval_runtime": 133.1254, | |
| "eval_samples_per_second": 14.565, | |
| "eval_steps_per_second": 1.825, | |
| "eval_wer": 0.44064864864864867, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.510989010989011, | |
| "grad_norm": 0.7302649021148682, | |
| "learning_rate": 0.00028625574977000917, | |
| "loss": 0.3833, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.510989010989011, | |
| "eval_loss": 0.2696720063686371, | |
| "eval_runtime": 134.0459, | |
| "eval_samples_per_second": 14.465, | |
| "eval_steps_per_second": 1.813, | |
| "eval_wer": 0.42464864864864865, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6483516483516483, | |
| "grad_norm": 0.8602890372276306, | |
| "learning_rate": 0.00028487580496780126, | |
| "loss": 0.3742, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.6483516483516483, | |
| "eval_loss": 0.2667711675167084, | |
| "eval_runtime": 133.3019, | |
| "eval_samples_per_second": 14.546, | |
| "eval_steps_per_second": 1.823, | |
| "eval_wer": 0.42502702702702705, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "grad_norm": 1.5730996131896973, | |
| "learning_rate": 0.00028349586016559336, | |
| "loss": 0.3694, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.7857142857142856, | |
| "eval_loss": 0.2689613401889801, | |
| "eval_runtime": 133.41, | |
| "eval_samples_per_second": 14.534, | |
| "eval_steps_per_second": 1.821, | |
| "eval_wer": 0.4189189189189189, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "grad_norm": 1.534987449645996, | |
| "learning_rate": 0.00028211591536338545, | |
| "loss": 0.3494, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.9230769230769231, | |
| "eval_loss": 0.26353445649147034, | |
| "eval_runtime": 132.3792, | |
| "eval_samples_per_second": 14.647, | |
| "eval_steps_per_second": 1.836, | |
| "eval_wer": 0.416, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.0604395604395602, | |
| "grad_norm": 1.566632866859436, | |
| "learning_rate": 0.00028073597056117754, | |
| "loss": 0.3724, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.0604395604395602, | |
| "eval_loss": 0.2625681757926941, | |
| "eval_runtime": 134.0736, | |
| "eval_samples_per_second": 14.462, | |
| "eval_steps_per_second": 1.812, | |
| "eval_wer": 0.4323243243243243, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.197802197802198, | |
| "grad_norm": 1.1042734384536743, | |
| "learning_rate": 0.00027935602575896964, | |
| "loss": 0.3723, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.197802197802198, | |
| "eval_loss": 0.2598220705986023, | |
| "eval_runtime": 132.9896, | |
| "eval_samples_per_second": 14.58, | |
| "eval_steps_per_second": 1.827, | |
| "eval_wer": 0.4247027027027027, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.3351648351648353, | |
| "grad_norm": 1.444578766822815, | |
| "learning_rate": 0.00027797608095676173, | |
| "loss": 0.3505, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.3351648351648353, | |
| "eval_loss": 0.25828292965888977, | |
| "eval_runtime": 133.6765, | |
| "eval_samples_per_second": 14.505, | |
| "eval_steps_per_second": 1.818, | |
| "eval_wer": 0.412, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.4725274725274726, | |
| "grad_norm": 2.0142714977264404, | |
| "learning_rate": 0.00027659613615455377, | |
| "loss": 0.3393, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.4725274725274726, | |
| "eval_loss": 0.2563156187534332, | |
| "eval_runtime": 132.7739, | |
| "eval_samples_per_second": 14.604, | |
| "eval_steps_per_second": 1.83, | |
| "eval_wer": 0.41281081081081084, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.60989010989011, | |
| "grad_norm": 2.562643527984619, | |
| "learning_rate": 0.00027521619135234586, | |
| "loss": 0.3352, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.60989010989011, | |
| "eval_loss": 0.25453999638557434, | |
| "eval_runtime": 133.7423, | |
| "eval_samples_per_second": 14.498, | |
| "eval_steps_per_second": 1.817, | |
| "eval_wer": 0.4155675675675676, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.7472527472527473, | |
| "grad_norm": 3.292581796646118, | |
| "learning_rate": 0.00027383624655013796, | |
| "loss": 0.3516, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.7472527472527473, | |
| "eval_loss": 0.25513678789138794, | |
| "eval_runtime": 133.1058, | |
| "eval_samples_per_second": 14.567, | |
| "eval_steps_per_second": 1.826, | |
| "eval_wer": 0.4315135135135135, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.8846153846153846, | |
| "grad_norm": 0.9819747805595398, | |
| "learning_rate": 0.00027245630174793005, | |
| "loss": 0.3489, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.8846153846153846, | |
| "eval_loss": 0.256041020154953, | |
| "eval_runtime": 134.2008, | |
| "eval_samples_per_second": 14.448, | |
| "eval_steps_per_second": 1.811, | |
| "eval_wer": 0.426972972972973, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.021978021978022, | |
| "grad_norm": 0.77928626537323, | |
| "learning_rate": 0.00027107635694572214, | |
| "loss": 0.3512, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.021978021978022, | |
| "eval_loss": 0.25362125039100647, | |
| "eval_runtime": 132.6388, | |
| "eval_samples_per_second": 14.619, | |
| "eval_steps_per_second": 1.832, | |
| "eval_wer": 0.4038918918918919, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.159340659340659, | |
| "grad_norm": 1.5085543394088745, | |
| "learning_rate": 0.00026971021159153633, | |
| "loss": 0.339, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.159340659340659, | |
| "eval_loss": 0.24898956716060638, | |
| "eval_runtime": 133.1617, | |
| "eval_samples_per_second": 14.561, | |
| "eval_steps_per_second": 1.825, | |
| "eval_wer": 0.3988648648648649, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.2967032967032965, | |
| "grad_norm": 0.6166655421257019, | |
| "learning_rate": 0.0002683302667893284, | |
| "loss": 0.3374, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.2967032967032965, | |
| "eval_loss": 0.24945656955242157, | |
| "eval_runtime": 132.9861, | |
| "eval_samples_per_second": 14.58, | |
| "eval_steps_per_second": 1.827, | |
| "eval_wer": 0.39637837837837836, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.4340659340659343, | |
| "grad_norm": 0.9863089919090271, | |
| "learning_rate": 0.00026695032198712046, | |
| "loss": 0.3295, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.4340659340659343, | |
| "eval_loss": 0.2518012523651123, | |
| "eval_runtime": 133.4804, | |
| "eval_samples_per_second": 14.526, | |
| "eval_steps_per_second": 1.82, | |
| "eval_wer": 0.4037297297297297, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "grad_norm": 0.5508905649185181, | |
| "learning_rate": 0.00026557037718491256, | |
| "loss": 0.3391, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 3.571428571428571, | |
| "eval_loss": 0.24910376965999603, | |
| "eval_runtime": 133.409, | |
| "eval_samples_per_second": 14.534, | |
| "eval_steps_per_second": 1.821, | |
| "eval_wer": 0.4077297297297297, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 3.708791208791209, | |
| "grad_norm": 0.5989585518836975, | |
| "learning_rate": 0.00026419043238270465, | |
| "loss": 0.3373, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 3.708791208791209, | |
| "eval_loss": 0.24448369443416595, | |
| "eval_runtime": 133.5414, | |
| "eval_samples_per_second": 14.52, | |
| "eval_steps_per_second": 1.82, | |
| "eval_wer": 0.3989189189189189, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.696869969367981, | |
| "learning_rate": 0.00026281048758049674, | |
| "loss": 0.3097, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "eval_loss": 0.24623404443264008, | |
| "eval_runtime": 134.7186, | |
| "eval_samples_per_second": 14.393, | |
| "eval_steps_per_second": 1.804, | |
| "eval_wer": 0.4117837837837838, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 3.9835164835164836, | |
| "grad_norm": 0.8349074721336365, | |
| "learning_rate": 0.00026143054277828884, | |
| "loss": 0.3458, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 3.9835164835164836, | |
| "eval_loss": 0.24432998895645142, | |
| "eval_runtime": 133.9009, | |
| "eval_samples_per_second": 14.481, | |
| "eval_steps_per_second": 1.815, | |
| "eval_wer": 0.4034054054054054, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.1208791208791204, | |
| "grad_norm": 0.9415439963340759, | |
| "learning_rate": 0.00026005059797608093, | |
| "loss": 0.313, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.1208791208791204, | |
| "eval_loss": 0.24325776100158691, | |
| "eval_runtime": 133.6886, | |
| "eval_samples_per_second": 14.504, | |
| "eval_steps_per_second": 1.818, | |
| "eval_wer": 0.3882162162162162, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.258241758241758, | |
| "grad_norm": 1.5970408916473389, | |
| "learning_rate": 0.000258670653173873, | |
| "loss": 0.3171, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.258241758241758, | |
| "eval_loss": 0.24258361756801605, | |
| "eval_runtime": 135.0574, | |
| "eval_samples_per_second": 14.357, | |
| "eval_steps_per_second": 1.799, | |
| "eval_wer": 0.3968108108108108, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.395604395604396, | |
| "grad_norm": 1.8427647352218628, | |
| "learning_rate": 0.0002572907083716651, | |
| "loss": 0.3122, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.395604395604396, | |
| "eval_loss": 0.2430083155632019, | |
| "eval_runtime": 133.0857, | |
| "eval_samples_per_second": 14.57, | |
| "eval_steps_per_second": 1.826, | |
| "eval_wer": 0.39356756756756756, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.532967032967033, | |
| "grad_norm": 0.6071587800979614, | |
| "learning_rate": 0.0002559107635694572, | |
| "loss": 0.3255, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 4.532967032967033, | |
| "eval_loss": 0.24039362370967865, | |
| "eval_runtime": 133.7381, | |
| "eval_samples_per_second": 14.498, | |
| "eval_steps_per_second": 1.817, | |
| "eval_wer": 0.38216216216216214, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 4.670329670329671, | |
| "grad_norm": 0.9580681324005127, | |
| "learning_rate": 0.0002545308187672493, | |
| "loss": 0.3253, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 4.670329670329671, | |
| "eval_loss": 0.2355840504169464, | |
| "eval_runtime": 132.9989, | |
| "eval_samples_per_second": 14.579, | |
| "eval_steps_per_second": 1.827, | |
| "eval_wer": 0.3945945945945946, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 4.8076923076923075, | |
| "grad_norm": 1.4110567569732666, | |
| "learning_rate": 0.0002531508739650414, | |
| "loss": 0.3341, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 4.8076923076923075, | |
| "eval_loss": 0.23690326511859894, | |
| "eval_runtime": 133.2507, | |
| "eval_samples_per_second": 14.552, | |
| "eval_steps_per_second": 1.824, | |
| "eval_wer": 0.3872432432432432, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 4.945054945054945, | |
| "grad_norm": 1.9591115713119507, | |
| "learning_rate": 0.0002517709291628335, | |
| "loss": 0.3183, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 4.945054945054945, | |
| "eval_loss": 0.2345106154680252, | |
| "eval_runtime": 133.4011, | |
| "eval_samples_per_second": 14.535, | |
| "eval_steps_per_second": 1.822, | |
| "eval_wer": 0.38535135135135135, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 5.082417582417582, | |
| "grad_norm": 0.616919219493866, | |
| "learning_rate": 0.0002503909843606256, | |
| "loss": 0.3461, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 5.082417582417582, | |
| "eval_loss": 0.23949113488197327, | |
| "eval_runtime": 133.8985, | |
| "eval_samples_per_second": 14.481, | |
| "eval_steps_per_second": 1.815, | |
| "eval_wer": 0.3828108108108108, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 5.21978021978022, | |
| "grad_norm": 0.7555241584777832, | |
| "learning_rate": 0.0002490110395584176, | |
| "loss": 0.3147, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 5.21978021978022, | |
| "eval_loss": 0.23591694235801697, | |
| "eval_runtime": 133.3437, | |
| "eval_samples_per_second": 14.541, | |
| "eval_steps_per_second": 1.822, | |
| "eval_wer": 0.37751351351351353, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 5.357142857142857, | |
| "grad_norm": 0.6169973015785217, | |
| "learning_rate": 0.0002476310947562097, | |
| "loss": 0.317, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 5.357142857142857, | |
| "eval_loss": 0.2320287972688675, | |
| "eval_runtime": 134.295, | |
| "eval_samples_per_second": 14.438, | |
| "eval_steps_per_second": 1.809, | |
| "eval_wer": 0.38075675675675674, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 5.4945054945054945, | |
| "grad_norm": 0.44300732016563416, | |
| "learning_rate": 0.0002462511499540018, | |
| "loss": 0.3094, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 5.4945054945054945, | |
| "eval_loss": 0.23655937612056732, | |
| "eval_runtime": 134.695, | |
| "eval_samples_per_second": 14.395, | |
| "eval_steps_per_second": 1.804, | |
| "eval_wer": 0.37972972972972974, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 5.631868131868131, | |
| "grad_norm": 1.5346554517745972, | |
| "learning_rate": 0.0002448712051517939, | |
| "loss": 0.2913, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 5.631868131868131, | |
| "eval_loss": 0.2357349544763565, | |
| "eval_runtime": 134.0434, | |
| "eval_samples_per_second": 14.465, | |
| "eval_steps_per_second": 1.813, | |
| "eval_wer": 0.37491891891891893, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 5.769230769230769, | |
| "grad_norm": 0.7017680406570435, | |
| "learning_rate": 0.000243491260349586, | |
| "loss": 0.3195, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 5.769230769230769, | |
| "eval_loss": 0.23318681120872498, | |
| "eval_runtime": 133.1069, | |
| "eval_samples_per_second": 14.567, | |
| "eval_steps_per_second": 1.826, | |
| "eval_wer": 0.3694054054054054, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 5.906593406593407, | |
| "grad_norm": 0.46099045872688293, | |
| "learning_rate": 0.00024212511499540016, | |
| "loss": 0.3189, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 5.906593406593407, | |
| "eval_loss": 0.23127850890159607, | |
| "eval_runtime": 135.2118, | |
| "eval_samples_per_second": 14.34, | |
| "eval_steps_per_second": 1.797, | |
| "eval_wer": 0.38697297297297295, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 6.043956043956044, | |
| "grad_norm": 1.010475516319275, | |
| "learning_rate": 0.00024074517019319225, | |
| "loss": 0.3105, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 6.043956043956044, | |
| "eval_loss": 0.23260627686977386, | |
| "eval_runtime": 134.6368, | |
| "eval_samples_per_second": 14.402, | |
| "eval_steps_per_second": 1.805, | |
| "eval_wer": 0.38064864864864867, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 6.181318681318682, | |
| "grad_norm": 0.7957504391670227, | |
| "learning_rate": 0.00023936522539098434, | |
| "loss": 0.2937, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 6.181318681318682, | |
| "eval_loss": 0.23464266955852509, | |
| "eval_runtime": 134.508, | |
| "eval_samples_per_second": 14.415, | |
| "eval_steps_per_second": 1.807, | |
| "eval_wer": 0.3783783783783784, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 6.318681318681318, | |
| "grad_norm": 0.73511803150177, | |
| "learning_rate": 0.00023798528058877644, | |
| "loss": 0.3088, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 6.318681318681318, | |
| "eval_loss": 0.2313350886106491, | |
| "eval_runtime": 135.48, | |
| "eval_samples_per_second": 14.312, | |
| "eval_steps_per_second": 1.794, | |
| "eval_wer": 0.3725945945945946, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 6.456043956043956, | |
| "grad_norm": 0.587340772151947, | |
| "learning_rate": 0.00023660533578656853, | |
| "loss": 0.2852, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 6.456043956043956, | |
| "eval_loss": 0.2307436466217041, | |
| "eval_runtime": 135.4353, | |
| "eval_samples_per_second": 14.317, | |
| "eval_steps_per_second": 1.794, | |
| "eval_wer": 0.37086486486486486, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 6.593406593406593, | |
| "grad_norm": 0.5956315994262695, | |
| "learning_rate": 0.0002352253909843606, | |
| "loss": 0.3083, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 6.593406593406593, | |
| "eval_loss": 0.22929497063159943, | |
| "eval_runtime": 134.2045, | |
| "eval_samples_per_second": 14.448, | |
| "eval_steps_per_second": 1.811, | |
| "eval_wer": 0.3751891891891892, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 6.730769230769231, | |
| "grad_norm": 1.3765567541122437, | |
| "learning_rate": 0.0002338454461821527, | |
| "loss": 0.3194, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 6.730769230769231, | |
| "eval_loss": 0.22924266755580902, | |
| "eval_runtime": 135.6964, | |
| "eval_samples_per_second": 14.289, | |
| "eval_steps_per_second": 1.791, | |
| "eval_wer": 0.37108108108108107, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 6.868131868131869, | |
| "grad_norm": 0.7330417037010193, | |
| "learning_rate": 0.00023246550137994478, | |
| "loss": 0.297, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 6.868131868131869, | |
| "eval_loss": 0.2303166538476944, | |
| "eval_runtime": 134.3817, | |
| "eval_samples_per_second": 14.429, | |
| "eval_steps_per_second": 1.808, | |
| "eval_wer": 0.3715135135135135, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 7.0054945054945055, | |
| "grad_norm": 0.9262694716453552, | |
| "learning_rate": 0.00023108555657773688, | |
| "loss": 0.3086, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 7.0054945054945055, | |
| "eval_loss": 0.23395021259784698, | |
| "eval_runtime": 134.6683, | |
| "eval_samples_per_second": 14.398, | |
| "eval_steps_per_second": 1.804, | |
| "eval_wer": 0.4027027027027027, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "grad_norm": 0.5672425627708435, | |
| "learning_rate": 0.00022970561177552897, | |
| "loss": 0.3058, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "eval_loss": 0.22942779958248138, | |
| "eval_runtime": 133.8761, | |
| "eval_samples_per_second": 14.484, | |
| "eval_steps_per_second": 1.815, | |
| "eval_wer": 0.36648648648648646, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 7.142857142857143, | |
| "step": 5200, | |
| "total_flos": 2.4864489394936607e+19, | |
| "train_loss": 0.46319753646850587, | |
| "train_runtime": 13134.8195, | |
| "train_samples_per_second": 13.286, | |
| "train_steps_per_second": 1.663 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 21840, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 30, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 3 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.4864489394936607e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |