| { | |
| "best_metric": 3.5733706206789404, | |
| "best_model_checkpoint": "./final-whisper-for-initial-publish/checkpoint-2000", | |
| "epoch": 3.2948929159802307, | |
| "eval_steps": 500, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04118616144975288, | |
| "grad_norm": 67.08393096923828, | |
| "learning_rate": 4.2000000000000006e-07, | |
| "loss": 2.8367, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08237232289950576, | |
| "grad_norm": 17.447174072265625, | |
| "learning_rate": 9.200000000000001e-07, | |
| "loss": 2.0282, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12355848434925865, | |
| "grad_norm": 12.4807710647583, | |
| "learning_rate": 1.42e-06, | |
| "loss": 1.3152, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16474464579901152, | |
| "grad_norm": 8.900789260864258, | |
| "learning_rate": 1.9200000000000003e-06, | |
| "loss": 0.7248, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.20593080724876442, | |
| "grad_norm": 7.897016525268555, | |
| "learning_rate": 2.42e-06, | |
| "loss": 0.5731, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2471169686985173, | |
| "grad_norm": 9.260579109191895, | |
| "learning_rate": 2.92e-06, | |
| "loss": 0.474, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2883031301482702, | |
| "grad_norm": 6.941812515258789, | |
| "learning_rate": 3.4200000000000007e-06, | |
| "loss": 0.4304, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.32948929159802304, | |
| "grad_norm": 6.919547080993652, | |
| "learning_rate": 3.920000000000001e-06, | |
| "loss": 0.3514, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.37067545304777594, | |
| "grad_norm": 6.444908618927002, | |
| "learning_rate": 4.42e-06, | |
| "loss": 0.273, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.41186161449752884, | |
| "grad_norm": 5.828249454498291, | |
| "learning_rate": 4.92e-06, | |
| "loss": 0.1655, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.45304777594728174, | |
| "grad_norm": 2.492833137512207, | |
| "learning_rate": 5.420000000000001e-06, | |
| "loss": 0.0894, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.4942339373970346, | |
| "grad_norm": 3.24698805809021, | |
| "learning_rate": 5.92e-06, | |
| "loss": 0.076, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5354200988467874, | |
| "grad_norm": 4.378988265991211, | |
| "learning_rate": 6.42e-06, | |
| "loss": 0.0763, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.5766062602965404, | |
| "grad_norm": 3.146955728530884, | |
| "learning_rate": 6.92e-06, | |
| "loss": 0.0679, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.6177924217462932, | |
| "grad_norm": 3.8782553672790527, | |
| "learning_rate": 7.420000000000001e-06, | |
| "loss": 0.0524, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.6589785831960461, | |
| "grad_norm": 6.15730094909668, | |
| "learning_rate": 7.92e-06, | |
| "loss": 0.0554, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.700164744645799, | |
| "grad_norm": 6.1213812828063965, | |
| "learning_rate": 8.42e-06, | |
| "loss": 0.0535, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.7413509060955519, | |
| "grad_norm": 4.010768890380859, | |
| "learning_rate": 8.920000000000001e-06, | |
| "loss": 0.0465, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7825370675453048, | |
| "grad_norm": 3.509321451187134, | |
| "learning_rate": 9.42e-06, | |
| "loss": 0.0436, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.8237232289950577, | |
| "grad_norm": 2.744082450866699, | |
| "learning_rate": 9.920000000000002e-06, | |
| "loss": 0.0468, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8237232289950577, | |
| "eval_loss": 0.04186907038092613, | |
| "eval_runtime": 1739.4189, | |
| "eval_samples_per_second": 1.395, | |
| "eval_steps_per_second": 0.175, | |
| "eval_wer": 10.859939408063388, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8649093904448105, | |
| "grad_norm": 2.581739664077759, | |
| "learning_rate": 9.953333333333333e-06, | |
| "loss": 0.0387, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.9060955518945635, | |
| "grad_norm": 1.7439745664596558, | |
| "learning_rate": 9.89777777777778e-06, | |
| "loss": 0.0372, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.9472817133443163, | |
| "grad_norm": 3.0741066932678223, | |
| "learning_rate": 9.842222222222223e-06, | |
| "loss": 0.0344, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.9884678747940692, | |
| "grad_norm": 2.640840530395508, | |
| "learning_rate": 9.786666666666667e-06, | |
| "loss": 0.0331, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.029654036243822, | |
| "grad_norm": 0.908171534538269, | |
| "learning_rate": 9.731111111111113e-06, | |
| "loss": 0.0267, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 1.0708401976935749, | |
| "grad_norm": 1.0449379682540894, | |
| "learning_rate": 9.675555555555555e-06, | |
| "loss": 0.0203, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 1.112026359143328, | |
| "grad_norm": 2.090158700942993, | |
| "learning_rate": 9.620000000000001e-06, | |
| "loss": 0.0185, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 1.1532125205930808, | |
| "grad_norm": 3.4184322357177734, | |
| "learning_rate": 9.564444444444445e-06, | |
| "loss": 0.0226, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.1943986820428336, | |
| "grad_norm": 2.6463236808776855, | |
| "learning_rate": 9.508888888888889e-06, | |
| "loss": 0.0155, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 1.2355848434925865, | |
| "grad_norm": 0.7438070774078369, | |
| "learning_rate": 9.453333333333335e-06, | |
| "loss": 0.0184, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.2767710049423393, | |
| "grad_norm": 1.7319138050079346, | |
| "learning_rate": 9.397777777777779e-06, | |
| "loss": 0.0161, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.3179571663920924, | |
| "grad_norm": 1.9704327583312988, | |
| "learning_rate": 9.342222222222223e-06, | |
| "loss": 0.0137, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.3591433278418452, | |
| "grad_norm": 1.808190107345581, | |
| "learning_rate": 9.286666666666667e-06, | |
| "loss": 0.0169, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.400329489291598, | |
| "grad_norm": 2.1106882095336914, | |
| "learning_rate": 9.231111111111111e-06, | |
| "loss": 0.0141, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.441515650741351, | |
| "grad_norm": 1.5317939519882202, | |
| "learning_rate": 9.175555555555557e-06, | |
| "loss": 0.0114, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.4827018121911038, | |
| "grad_norm": 2.758979082107544, | |
| "learning_rate": 9.12e-06, | |
| "loss": 0.0123, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.5238879736408566, | |
| "grad_norm": 2.9624338150024414, | |
| "learning_rate": 9.064444444444447e-06, | |
| "loss": 0.014, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.5650741350906094, | |
| "grad_norm": 2.520521402359009, | |
| "learning_rate": 9.008888888888889e-06, | |
| "loss": 0.0107, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.6062602965403623, | |
| "grad_norm": 1.0015959739685059, | |
| "learning_rate": 8.953333333333335e-06, | |
| "loss": 0.0111, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.6474464579901154, | |
| "grad_norm": 2.2675178050994873, | |
| "learning_rate": 8.897777777777779e-06, | |
| "loss": 0.0124, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.6474464579901154, | |
| "eval_loss": 0.02125450409948826, | |
| "eval_runtime": 1730.2385, | |
| "eval_samples_per_second": 1.403, | |
| "eval_steps_per_second": 0.176, | |
| "eval_wer": 5.243532975996271, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.6886326194398682, | |
| "grad_norm": 1.1241651773452759, | |
| "learning_rate": 8.842222222222223e-06, | |
| "loss": 0.0108, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.729818780889621, | |
| "grad_norm": 0.9443243145942688, | |
| "learning_rate": 8.786666666666668e-06, | |
| "loss": 0.01, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.771004942339374, | |
| "grad_norm": 1.7507250308990479, | |
| "learning_rate": 8.73111111111111e-06, | |
| "loss": 0.0108, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.812191103789127, | |
| "grad_norm": 0.8334024548530579, | |
| "learning_rate": 8.675555555555556e-06, | |
| "loss": 0.0108, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.8533772652388798, | |
| "grad_norm": 2.0746381282806396, | |
| "learning_rate": 8.62e-06, | |
| "loss": 0.009, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.8945634266886326, | |
| "grad_norm": 1.5471493005752563, | |
| "learning_rate": 8.564444444444445e-06, | |
| "loss": 0.0124, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.9357495881383855, | |
| "grad_norm": 1.4362578392028809, | |
| "learning_rate": 8.50888888888889e-06, | |
| "loss": 0.009, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.9769357495881383, | |
| "grad_norm": 1.5219734907150269, | |
| "learning_rate": 8.453333333333334e-06, | |
| "loss": 0.0108, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.018121911037891, | |
| "grad_norm": 0.7635548710823059, | |
| "learning_rate": 8.397777777777778e-06, | |
| "loss": 0.0069, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 2.059308072487644, | |
| "grad_norm": 0.622107982635498, | |
| "learning_rate": 8.342222222222222e-06, | |
| "loss": 0.0045, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 2.100494233937397, | |
| "grad_norm": 3.157203197479248, | |
| "learning_rate": 8.286666666666668e-06, | |
| "loss": 0.0055, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 2.1416803953871497, | |
| "grad_norm": 0.8712319731712341, | |
| "learning_rate": 8.231111111111112e-06, | |
| "loss": 0.0025, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.182866556836903, | |
| "grad_norm": 0.7438367605209351, | |
| "learning_rate": 8.175555555555556e-06, | |
| "loss": 0.0038, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 2.224052718286656, | |
| "grad_norm": 0.4088328778743744, | |
| "learning_rate": 8.120000000000002e-06, | |
| "loss": 0.004, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 2.2652388797364087, | |
| "grad_norm": 0.34847044944763184, | |
| "learning_rate": 8.064444444444444e-06, | |
| "loss": 0.003, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 2.3064250411861615, | |
| "grad_norm": 0.08477868884801865, | |
| "learning_rate": 8.00888888888889e-06, | |
| "loss": 0.0024, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.3476112026359144, | |
| "grad_norm": 0.5055556297302246, | |
| "learning_rate": 7.953333333333334e-06, | |
| "loss": 0.004, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 2.3887973640856672, | |
| "grad_norm": 0.27648621797561646, | |
| "learning_rate": 7.897777777777778e-06, | |
| "loss": 0.0037, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 2.42998352553542, | |
| "grad_norm": 2.3948631286621094, | |
| "learning_rate": 7.842222222222224e-06, | |
| "loss": 0.0022, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 2.471169686985173, | |
| "grad_norm": 0.3471105694770813, | |
| "learning_rate": 7.786666666666666e-06, | |
| "loss": 0.0034, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.471169686985173, | |
| "eval_loss": 0.017010504379868507, | |
| "eval_runtime": 1743.3725, | |
| "eval_samples_per_second": 1.392, | |
| "eval_steps_per_second": 0.174, | |
| "eval_wer": 3.821952924726171, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.5123558484349258, | |
| "grad_norm": 0.14169706404209137, | |
| "learning_rate": 7.731111111111112e-06, | |
| "loss": 0.0043, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 2.5535420098846786, | |
| "grad_norm": 0.4921598732471466, | |
| "learning_rate": 7.675555555555556e-06, | |
| "loss": 0.0024, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.594728171334432, | |
| "grad_norm": 0.20753732323646545, | |
| "learning_rate": 7.620000000000001e-06, | |
| "loss": 0.004, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 2.6359143327841847, | |
| "grad_norm": 0.10961499810218811, | |
| "learning_rate": 7.564444444444446e-06, | |
| "loss": 0.004, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.6771004942339376, | |
| "grad_norm": 1.0841516256332397, | |
| "learning_rate": 7.50888888888889e-06, | |
| "loss": 0.0046, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 2.7182866556836904, | |
| "grad_norm": 0.12003188580274582, | |
| "learning_rate": 7.453333333333334e-06, | |
| "loss": 0.0023, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.7594728171334433, | |
| "grad_norm": 0.055499833077192307, | |
| "learning_rate": 7.3977777777777786e-06, | |
| "loss": 0.0035, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 2.800658978583196, | |
| "grad_norm": 2.9141526222229004, | |
| "learning_rate": 7.342222222222223e-06, | |
| "loss": 0.0023, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.841845140032949, | |
| "grad_norm": 0.025740938261151314, | |
| "learning_rate": 7.2866666666666675e-06, | |
| "loss": 0.0024, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 2.883031301482702, | |
| "grad_norm": 0.6345125436782837, | |
| "learning_rate": 7.231111111111112e-06, | |
| "loss": 0.004, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.9242174629324547, | |
| "grad_norm": 0.5243808627128601, | |
| "learning_rate": 7.1755555555555556e-06, | |
| "loss": 0.0024, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 2.9654036243822075, | |
| "grad_norm": 0.44930708408355713, | |
| "learning_rate": 7.1200000000000004e-06, | |
| "loss": 0.0025, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 3.0065897858319603, | |
| "grad_norm": 1.6802186965942383, | |
| "learning_rate": 7.0644444444444445e-06, | |
| "loss": 0.0044, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 3.047775947281713, | |
| "grad_norm": 0.2793377637863159, | |
| "learning_rate": 7.008888888888889e-06, | |
| "loss": 0.0015, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 3.088962108731466, | |
| "grad_norm": 0.07968737185001373, | |
| "learning_rate": 6.953333333333334e-06, | |
| "loss": 0.0015, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 3.130148270181219, | |
| "grad_norm": 0.04705784469842911, | |
| "learning_rate": 6.897777777777779e-06, | |
| "loss": 0.001, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.171334431630972, | |
| "grad_norm": 0.1313919574022293, | |
| "learning_rate": 6.842222222222222e-06, | |
| "loss": 0.0008, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 3.212520593080725, | |
| "grad_norm": 0.4405129551887512, | |
| "learning_rate": 6.786666666666667e-06, | |
| "loss": 0.0011, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 3.253706754530478, | |
| "grad_norm": 1.2785202264785767, | |
| "learning_rate": 6.731111111111111e-06, | |
| "loss": 0.0017, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 3.2948929159802307, | |
| "grad_norm": 0.04460221529006958, | |
| "learning_rate": 6.675555555555556e-06, | |
| "loss": 0.0008, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.2948929159802307, | |
| "eval_loss": 0.015330803580582142, | |
| "eval_runtime": 1719.9199, | |
| "eval_samples_per_second": 1.411, | |
| "eval_steps_per_second": 0.177, | |
| "eval_wer": 3.5733706206789404, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 9, | |
| "save_steps": 500, | |
| "total_flos": 9.23126978543616e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |