| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 29.992926196651734, | |
| "eval_steps": 2000, | |
| "global_step": 63600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 0.00014849999999999998, | |
| "loss": 15.6695, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 0.0002982, | |
| "loss": 4.4987, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 0.00029763258785942486, | |
| "loss": 2.3147, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 0.0002952364217252396, | |
| "loss": 1.7721, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "eval_cer": 0.29026945911872865, | |
| "eval_loss": 1.1368173360824585, | |
| "eval_runtime": 2718.6813, | |
| "eval_samples_per_second": 6.221, | |
| "eval_steps_per_second": 0.389, | |
| "eval_wer": 0.6589364856615053, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 0.0002928402555910543, | |
| "loss": 1.5735, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 0.000290444089456869, | |
| "loss": 1.4611, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 0.00028805271565495203, | |
| "loss": 1.4056, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 0.00028565654952076674, | |
| "loss": 1.3501, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "eval_cer": 0.2240432390931656, | |
| "eval_loss": 0.8561204671859741, | |
| "eval_runtime": 2698.4922, | |
| "eval_samples_per_second": 6.268, | |
| "eval_steps_per_second": 0.392, | |
| "eval_wer": 0.5451323644839807, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 0.00028326038338658146, | |
| "loss": 1.2967, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 0.00028086421725239617, | |
| "loss": 1.244, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "learning_rate": 0.0002784728434504792, | |
| "loss": 1.2361, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 0.0002760766773162939, | |
| "loss": 1.2133, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "eval_cer": 0.2002927101645895, | |
| "eval_loss": 0.7505359053611755, | |
| "eval_runtime": 2721.6856, | |
| "eval_samples_per_second": 6.215, | |
| "eval_steps_per_second": 0.389, | |
| "eval_wer": 0.49743516558444134, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.07, | |
| "learning_rate": 0.0002736805111821086, | |
| "loss": 1.1755, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "learning_rate": 0.0002712843450479233, | |
| "loss": 1.1403, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "learning_rate": 0.000268888178913738, | |
| "loss": 1.1253, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "learning_rate": 0.00026649680511182103, | |
| "loss": 1.0981, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "eval_cer": 0.18418492651709803, | |
| "eval_loss": 0.6767656803131104, | |
| "eval_runtime": 2740.3957, | |
| "eval_samples_per_second": 6.172, | |
| "eval_steps_per_second": 0.386, | |
| "eval_wer": 0.46858375155594667, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.01, | |
| "learning_rate": 0.00026410063897763575, | |
| "loss": 1.0984, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "learning_rate": 0.00026170447284345046, | |
| "loss": 1.0614, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "learning_rate": 0.00025930830670926517, | |
| "loss": 1.0553, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "learning_rate": 0.0002569169329073482, | |
| "loss": 1.0375, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "eval_cer": 0.17071197058066542, | |
| "eval_loss": 0.64134681224823, | |
| "eval_runtime": 2703.0312, | |
| "eval_samples_per_second": 6.257, | |
| "eval_steps_per_second": 0.391, | |
| "eval_wer": 0.44041417913406117, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.95, | |
| "learning_rate": 0.0002545207667731629, | |
| "loss": 1.0351, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 5.19, | |
| "learning_rate": 0.00025212460063897763, | |
| "loss": 1.0087, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 5.42, | |
| "learning_rate": 0.00024973322683706067, | |
| "loss": 0.9934, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 5.66, | |
| "learning_rate": 0.0002473370607028754, | |
| "loss": 0.9927, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.66, | |
| "eval_cer": 0.16338767301962598, | |
| "eval_loss": 0.6106029152870178, | |
| "eval_runtime": 2826.5991, | |
| "eval_samples_per_second": 5.984, | |
| "eval_steps_per_second": 0.374, | |
| "eval_wer": 0.42461289632043386, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 5.89, | |
| "learning_rate": 0.0002449408945686901, | |
| "loss": 0.9821, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 6.13, | |
| "learning_rate": 0.0002425447284345048, | |
| "loss": 0.9637, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 6.37, | |
| "learning_rate": 0.00024014856230031946, | |
| "loss": 0.9497, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "learning_rate": 0.00023775239616613414, | |
| "loss": 0.9439, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "eval_cer": 0.1613024947979602, | |
| "eval_loss": 0.5999171733856201, | |
| "eval_runtime": 2741.759, | |
| "eval_samples_per_second": 6.169, | |
| "eval_steps_per_second": 0.386, | |
| "eval_wer": 0.41588373807768236, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 6.84, | |
| "learning_rate": 0.00023536102236421723, | |
| "loss": 0.948, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 7.07, | |
| "learning_rate": 0.0002329696485623003, | |
| "loss": 0.9367, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 7.31, | |
| "learning_rate": 0.000230573482428115, | |
| "loss": 0.9087, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 7.55, | |
| "learning_rate": 0.0002281773162939297, | |
| "loss": 0.9059, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 7.55, | |
| "eval_cer": 0.15351797484699242, | |
| "eval_loss": 0.5740103721618652, | |
| "eval_runtime": 2702.3695, | |
| "eval_samples_per_second": 6.259, | |
| "eval_steps_per_second": 0.392, | |
| "eval_wer": 0.39852056228147376, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 7.78, | |
| "learning_rate": 0.00022578115015974438, | |
| "loss": 0.9075, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "learning_rate": 0.0002233849840255591, | |
| "loss": 0.8999, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 8.25, | |
| "learning_rate": 0.00022098881789137377, | |
| "loss": 0.8597, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 8.49, | |
| "learning_rate": 0.00021859265175718849, | |
| "loss": 0.8772, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 8.49, | |
| "eval_cer": 0.14781427082015555, | |
| "eval_loss": 0.5568912625312805, | |
| "eval_runtime": 2727.4118, | |
| "eval_samples_per_second": 6.201, | |
| "eval_steps_per_second": 0.388, | |
| "eval_wer": 0.3954364182701837, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 8.72, | |
| "learning_rate": 0.00021619648562300317, | |
| "loss": 0.8785, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 8.96, | |
| "learning_rate": 0.00021380511182108623, | |
| "loss": 0.8702, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 9.2, | |
| "learning_rate": 0.00021140894568690095, | |
| "loss": 0.8447, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 9.43, | |
| "learning_rate": 0.00020901277955271563, | |
| "loss": 0.8483, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 9.43, | |
| "eval_cer": 0.14274528108464166, | |
| "eval_loss": 0.5406663417816162, | |
| "eval_runtime": 2824.1098, | |
| "eval_samples_per_second": 5.989, | |
| "eval_steps_per_second": 0.375, | |
| "eval_wer": 0.3784141632772796, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 9.67, | |
| "learning_rate": 0.00020661661341853031, | |
| "loss": 0.8358, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 9.9, | |
| "learning_rate": 0.00020422523961661338, | |
| "loss": 0.8391, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 10.14, | |
| "learning_rate": 0.00020182907348242812, | |
| "loss": 0.8215, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 10.37, | |
| "learning_rate": 0.00019943769968051118, | |
| "loss": 0.81, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 10.37, | |
| "eval_cer": 0.1415456492625536, | |
| "eval_loss": 0.5282983779907227, | |
| "eval_runtime": 2742.622, | |
| "eval_samples_per_second": 6.167, | |
| "eval_steps_per_second": 0.386, | |
| "eval_wer": 0.37441032593614476, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 10.61, | |
| "learning_rate": 0.00019704153354632587, | |
| "loss": 0.8108, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 10.85, | |
| "learning_rate": 0.00019464536741214058, | |
| "loss": 0.8175, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 11.08, | |
| "learning_rate": 0.00019224920127795526, | |
| "loss": 0.8001, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 11.32, | |
| "learning_rate": 0.00018985782747603833, | |
| "loss": 0.793, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 11.32, | |
| "eval_cer": 0.13662715879199255, | |
| "eval_loss": 0.517921507358551, | |
| "eval_runtime": 2729.8979, | |
| "eval_samples_per_second": 6.196, | |
| "eval_steps_per_second": 0.388, | |
| "eval_wer": 0.36633129573690426, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 11.55, | |
| "learning_rate": 0.000187461661341853, | |
| "loss": 0.7827, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 11.79, | |
| "learning_rate": 0.00018506549520766772, | |
| "loss": 0.7899, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 12.03, | |
| "learning_rate": 0.0001826693290734824, | |
| "loss": 0.7806, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 12.26, | |
| "learning_rate": 0.00018027316293929712, | |
| "loss": 0.7577, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 12.26, | |
| "eval_cer": 0.1359182854425769, | |
| "eval_loss": 0.5058821439743042, | |
| "eval_runtime": 2722.7634, | |
| "eval_samples_per_second": 6.212, | |
| "eval_steps_per_second": 0.389, | |
| "eval_wer": 0.35946530932616605, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 0.00017788178913738016, | |
| "loss": 0.762, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 12.73, | |
| "learning_rate": 0.00017548562300319487, | |
| "loss": 0.7595, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 12.97, | |
| "learning_rate": 0.00017308945686900955, | |
| "loss": 0.7629, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "learning_rate": 0.00017069329073482426, | |
| "loss": 0.7379, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 13.2, | |
| "eval_cer": 0.13330526921919236, | |
| "eval_loss": 0.4969228506088257, | |
| "eval_runtime": 2824.8712, | |
| "eval_samples_per_second": 5.988, | |
| "eval_steps_per_second": 0.375, | |
| "eval_wer": 0.35324945095893884, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 13.44, | |
| "learning_rate": 0.00016830191693290736, | |
| "loss": 0.737, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 13.68, | |
| "learning_rate": 0.000165905750798722, | |
| "loss": 0.7444, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 13.91, | |
| "learning_rate": 0.00016350958466453675, | |
| "loss": 0.7372, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 14.15, | |
| "learning_rate": 0.00016111341853035144, | |
| "loss": 0.7328, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 14.15, | |
| "eval_cer": 0.13079476698787718, | |
| "eval_loss": 0.4908413589000702, | |
| "eval_runtime": 2825.8542, | |
| "eval_samples_per_second": 5.985, | |
| "eval_steps_per_second": 0.374, | |
| "eval_wer": 0.3475251528197322, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 14.38, | |
| "learning_rate": 0.0001587172523961661, | |
| "loss": 0.7184, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 14.62, | |
| "learning_rate": 0.00015632108626198083, | |
| "loss": 0.7216, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 14.85, | |
| "learning_rate": 0.0001539297124600639, | |
| "loss": 0.7238, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 15.09, | |
| "learning_rate": 0.00015153354632587858, | |
| "loss": 0.7119, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 15.09, | |
| "eval_cer": 0.12864851660072327, | |
| "eval_loss": 0.4887321889400482, | |
| "eval_runtime": 2731.0336, | |
| "eval_samples_per_second": 6.193, | |
| "eval_steps_per_second": 0.387, | |
| "eval_wer": 0.34784228845071313, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 15.33, | |
| "learning_rate": 0.00014914217252396165, | |
| "loss": 0.7124, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 15.56, | |
| "learning_rate": 0.00014674600638977636, | |
| "loss": 0.7294, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 15.8, | |
| "learning_rate": 0.00014434984025559104, | |
| "loss": 0.7545, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 16.03, | |
| "learning_rate": 0.00014195367412140575, | |
| "loss": 0.7572, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 16.03, | |
| "eval_cer": 0.13271854020075294, | |
| "eval_loss": 0.5169993042945862, | |
| "eval_runtime": 2729.0002, | |
| "eval_samples_per_second": 6.198, | |
| "eval_steps_per_second": 0.388, | |
| "eval_wer": 0.3576893497926726, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 16.27, | |
| "learning_rate": 0.00013956230031948882, | |
| "loss": 0.7687, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 16.51, | |
| "learning_rate": 0.0001371661341853035, | |
| "loss": 0.7884, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 16.74, | |
| "learning_rate": 0.00013476996805111819, | |
| "loss": 0.8156, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 16.98, | |
| "learning_rate": 0.0001323738019169329, | |
| "loss": 0.8198, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 16.98, | |
| "eval_cer": 0.1431662427967562, | |
| "eval_loss": 0.5838645696640015, | |
| "eval_runtime": 2730.4526, | |
| "eval_samples_per_second": 6.195, | |
| "eval_steps_per_second": 0.387, | |
| "eval_wer": 0.38254485487080686, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 17.21, | |
| "learning_rate": 0.00012997763578274758, | |
| "loss": 0.819, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 17.45, | |
| "learning_rate": 0.00012758626198083067, | |
| "loss": 0.8411, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 17.68, | |
| "learning_rate": 0.00012519009584664536, | |
| "loss": 0.8366, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 17.92, | |
| "learning_rate": 0.00012279392971246007, | |
| "loss": 0.8008, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 17.92, | |
| "eval_cer": 0.13762394377870937, | |
| "eval_loss": 0.5447062253952026, | |
| "eval_runtime": 2738.2931, | |
| "eval_samples_per_second": 6.177, | |
| "eval_steps_per_second": 0.386, | |
| "eval_wer": 0.36609344401366856, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 18.16, | |
| "learning_rate": 0.00012039776357827474, | |
| "loss": 0.8032, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 18.39, | |
| "learning_rate": 0.00011800159744408944, | |
| "loss": 0.7753, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 18.63, | |
| "learning_rate": 0.00011560543130990414, | |
| "loss": 0.7608, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 18.86, | |
| "learning_rate": 0.00011321405750798721, | |
| "loss": 0.759, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 18.86, | |
| "eval_cer": 0.1336804268071908, | |
| "eval_loss": 0.49982598423957825, | |
| "eval_runtime": 2725.5181, | |
| "eval_samples_per_second": 6.206, | |
| "eval_steps_per_second": 0.388, | |
| "eval_wer": 0.3533921619928803, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 19.1, | |
| "learning_rate": 0.00011081789137380191, | |
| "loss": 0.7285, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 19.34, | |
| "learning_rate": 0.00010842172523961661, | |
| "loss": 0.7036, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 19.57, | |
| "learning_rate": 0.00010602555910543131, | |
| "loss": 0.6953, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 19.81, | |
| "learning_rate": 0.00010363418530351436, | |
| "loss": 0.6907, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 19.81, | |
| "eval_cer": 0.12877502322923437, | |
| "eval_loss": 0.47100237011909485, | |
| "eval_runtime": 2667.1801, | |
| "eval_samples_per_second": 6.342, | |
| "eval_steps_per_second": 0.397, | |
| "eval_wer": 0.34119829698166165, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 20.04, | |
| "learning_rate": 0.00010123801916932906, | |
| "loss": 0.6858, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 20.28, | |
| "learning_rate": 9.884664536741213e-05, | |
| "loss": 0.6603, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 20.51, | |
| "learning_rate": 9.645047923322683e-05, | |
| "loss": 0.6609, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 20.75, | |
| "learning_rate": 9.405431309904153e-05, | |
| "loss": 0.659, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 20.75, | |
| "eval_cer": 0.12423387149543921, | |
| "eval_loss": 0.4578304886817932, | |
| "eval_runtime": 2665.8908, | |
| "eval_samples_per_second": 6.345, | |
| "eval_steps_per_second": 0.397, | |
| "eval_wer": 0.3324532819573611, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 20.99, | |
| "learning_rate": 9.165814696485623e-05, | |
| "loss": 0.6567, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 21.22, | |
| "learning_rate": 8.926198083067093e-05, | |
| "loss": 0.6437, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 21.46, | |
| "learning_rate": 8.686581469648561e-05, | |
| "loss": 0.6371, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 21.69, | |
| "learning_rate": 8.447444089456868e-05, | |
| "loss": 0.6345, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 21.69, | |
| "eval_cer": 0.12205708502554125, | |
| "eval_loss": 0.45305466651916504, | |
| "eval_runtime": 2667.1819, | |
| "eval_samples_per_second": 6.342, | |
| "eval_steps_per_second": 0.397, | |
| "eval_wer": 0.3256982930174662, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 21.93, | |
| "learning_rate": 8.207827476038337e-05, | |
| "loss": 0.6418, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 22.16, | |
| "learning_rate": 7.968210862619807e-05, | |
| "loss": 0.6306, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 22.4, | |
| "learning_rate": 7.728594249201278e-05, | |
| "loss": 0.6213, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 22.64, | |
| "learning_rate": 7.489456869009583e-05, | |
| "loss": 0.6242, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 22.64, | |
| "eval_cer": 0.12094251800538308, | |
| "eval_loss": 0.4497627019882202, | |
| "eval_runtime": 2727.8154, | |
| "eval_samples_per_second": 6.201, | |
| "eval_steps_per_second": 0.388, | |
| "eval_wer": 0.32180545314717474, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 22.87, | |
| "learning_rate": 7.249840255591053e-05, | |
| "loss": 0.6294, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 23.11, | |
| "learning_rate": 7.010223642172524e-05, | |
| "loss": 0.6141, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 23.34, | |
| "learning_rate": 6.770607028753993e-05, | |
| "loss": 0.6155, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 23.58, | |
| "learning_rate": 6.530990415335462e-05, | |
| "loss": 0.6163, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 23.58, | |
| "eval_cer": 0.11941571386818009, | |
| "eval_loss": 0.45521289110183716, | |
| "eval_runtime": 2664.8843, | |
| "eval_samples_per_second": 6.347, | |
| "eval_steps_per_second": 0.397, | |
| "eval_wer": 0.3188402349975026, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 23.82, | |
| "learning_rate": 6.291373801916932e-05, | |
| "loss": 0.6167, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 24.05, | |
| "learning_rate": 6.0522364217252394e-05, | |
| "loss": 0.6179, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 24.29, | |
| "learning_rate": 5.8126198083067085e-05, | |
| "loss": 0.6154, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 24.52, | |
| "learning_rate": 5.573482428115016e-05, | |
| "loss": 0.6121, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 24.52, | |
| "eval_cer": 0.1153500525656853, | |
| "eval_loss": 0.46334853768348694, | |
| "eval_runtime": 2666.3375, | |
| "eval_samples_per_second": 6.344, | |
| "eval_steps_per_second": 0.397, | |
| "eval_wer": 0.3136947093848362, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 24.76, | |
| "learning_rate": 5.3338658146964855e-05, | |
| "loss": 0.6227, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 24.99, | |
| "learning_rate": 5.0942492012779546e-05, | |
| "loss": 0.6156, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 25.23, | |
| "learning_rate": 4.854632587859424e-05, | |
| "loss": 0.6159, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 25.47, | |
| "learning_rate": 4.615015974440894e-05, | |
| "loss": 0.6054, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 25.47, | |
| "eval_cer": 0.11759009234983882, | |
| "eval_loss": 0.46227386593818665, | |
| "eval_runtime": 2666.9277, | |
| "eval_samples_per_second": 6.342, | |
| "eval_steps_per_second": 0.397, | |
| "eval_wer": 0.3171356309809798, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 25.7, | |
| "learning_rate": 4.375399361022364e-05, | |
| "loss": 0.6051, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 25.94, | |
| "learning_rate": 4.1362619808306704e-05, | |
| "loss": 0.5986, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 26.17, | |
| "learning_rate": 3.89664536741214e-05, | |
| "loss": 0.5916, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 26.41, | |
| "learning_rate": 3.65702875399361e-05, | |
| "loss": 0.591, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 26.41, | |
| "eval_cer": 0.11455829556310718, | |
| "eval_loss": 0.4413212835788727, | |
| "eval_runtime": 2669.4551, | |
| "eval_samples_per_second": 6.336, | |
| "eval_steps_per_second": 0.396, | |
| "eval_wer": 0.31158575743881267, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 26.64, | |
| "learning_rate": 3.41741214057508e-05, | |
| "loss": 0.5904, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 26.88, | |
| "learning_rate": 3.178274760383386e-05, | |
| "loss": 0.5887, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 27.12, | |
| "learning_rate": 2.9386581469648557e-05, | |
| "loss": 0.5768, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 27.35, | |
| "learning_rate": 2.6990415335463258e-05, | |
| "loss": 0.5713, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 27.35, | |
| "eval_cer": 0.11345245313801873, | |
| "eval_loss": 0.4338010549545288, | |
| "eval_runtime": 2668.5066, | |
| "eval_samples_per_second": 6.338, | |
| "eval_steps_per_second": 0.396, | |
| "eval_wer": 0.3092706673326515, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 27.59, | |
| "learning_rate": 2.4594249201277952e-05, | |
| "loss": 0.5653, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 27.82, | |
| "learning_rate": 2.219808306709265e-05, | |
| "loss": 0.569, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 28.06, | |
| "learning_rate": 1.980670926517572e-05, | |
| "loss": 0.5748, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 28.3, | |
| "learning_rate": 1.7410543130990413e-05, | |
| "loss": 0.5703, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 28.3, | |
| "eval_cer": 0.11209795975344294, | |
| "eval_loss": 0.42797738313674927, | |
| "eval_runtime": 2667.5384, | |
| "eval_samples_per_second": 6.341, | |
| "eval_steps_per_second": 0.397, | |
| "eval_wer": 0.30612309619516526, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 28.53, | |
| "learning_rate": 1.501437699680511e-05, | |
| "loss": 0.5606, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 28.77, | |
| "learning_rate": 1.2623003194888177e-05, | |
| "loss": 0.5647, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 29.0, | |
| "learning_rate": 1.0226837060702875e-05, | |
| "loss": 0.5567, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 29.24, | |
| "learning_rate": 7.830670926517571e-06, | |
| "loss": 0.5576, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 29.24, | |
| "eval_cer": 0.11193437359588548, | |
| "eval_loss": 0.42482054233551025, | |
| "eval_runtime": 2665.5981, | |
| "eval_samples_per_second": 6.345, | |
| "eval_steps_per_second": 0.397, | |
| "eval_wer": 0.30466427229265275, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 29.47, | |
| "learning_rate": 5.434504792332268e-06, | |
| "loss": 0.5596, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 29.71, | |
| "learning_rate": 3.0431309904153355e-06, | |
| "loss": 0.5567, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 29.95, | |
| "learning_rate": 6.469648562300319e-07, | |
| "loss": 0.5581, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 29.99, | |
| "step": 63600, | |
| "total_flos": 4.271260053717039e+20, | |
| "train_loss": 0.9622837933354408, | |
| "train_runtime": 156348.4692, | |
| "train_samples_per_second": 26.04, | |
| "train_steps_per_second": 0.407 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 63600, | |
| "num_train_epochs": 30, | |
| "save_steps": 2000, | |
| "total_flos": 4.271260053717039e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |