| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 7.739938080495356, | |
| "eval_steps": 100, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15479876160990713, | |
| "eval_loss": 3.5586698055267334, | |
| "eval_runtime": 163.4718, | |
| "eval_samples_per_second": 34.599, | |
| "eval_steps_per_second": 4.325, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.30959752321981426, | |
| "eval_loss": 3.2505505084991455, | |
| "eval_runtime": 161.1897, | |
| "eval_samples_per_second": 35.089, | |
| "eval_steps_per_second": 4.386, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.46439628482972134, | |
| "eval_loss": 2.774010181427002, | |
| "eval_runtime": 161.9005, | |
| "eval_samples_per_second": 34.935, | |
| "eval_steps_per_second": 4.367, | |
| "eval_wer": 0.9999518544077289, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6191950464396285, | |
| "eval_loss": 1.11959707736969, | |
| "eval_runtime": 162.1425, | |
| "eval_samples_per_second": 34.883, | |
| "eval_steps_per_second": 4.36, | |
| "eval_wer": 0.7807449727977404, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.7739938080495357, | |
| "grad_norm": 2.676631212234497, | |
| "learning_rate": 0.00029699999999999996, | |
| "loss": 3.6484, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7739938080495357, | |
| "eval_loss": 0.9134386777877808, | |
| "eval_runtime": 162.7792, | |
| "eval_samples_per_second": 34.746, | |
| "eval_steps_per_second": 4.343, | |
| "eval_wer": 0.6538813371635827, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.9287925696594427, | |
| "eval_loss": 0.7674785256385803, | |
| "eval_runtime": 162.7057, | |
| "eval_samples_per_second": 34.762, | |
| "eval_steps_per_second": 4.345, | |
| "eval_wer": 0.5923191731796954, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 1.08359133126935, | |
| "eval_loss": 0.7207810282707214, | |
| "eval_runtime": 163.9828, | |
| "eval_samples_per_second": 34.491, | |
| "eval_steps_per_second": 4.311, | |
| "eval_wer": 0.5289595737510231, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.238390092879257, | |
| "eval_loss": 0.6209472417831421, | |
| "eval_runtime": 163.6218, | |
| "eval_samples_per_second": 34.568, | |
| "eval_steps_per_second": 4.321, | |
| "eval_wer": 0.4744748118309769, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.3931888544891642, | |
| "eval_loss": 0.6220189332962036, | |
| "eval_runtime": 162.6081, | |
| "eval_samples_per_second": 34.783, | |
| "eval_steps_per_second": 4.348, | |
| "eval_wer": 0.47879186660461237, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.5479876160990713, | |
| "grad_norm": 0.3950090706348419, | |
| "learning_rate": 0.0002672, | |
| "loss": 0.6286, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.5479876160990713, | |
| "eval_loss": 0.5738953351974487, | |
| "eval_runtime": 162.8483, | |
| "eval_samples_per_second": 34.732, | |
| "eval_steps_per_second": 4.341, | |
| "eval_wer": 0.4588114458121359, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.7027863777089784, | |
| "eval_loss": 0.564153790473938, | |
| "eval_runtime": 164.0752, | |
| "eval_samples_per_second": 34.472, | |
| "eval_steps_per_second": 4.309, | |
| "eval_wer": 0.4262008313138932, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.8575851393188856, | |
| "eval_loss": 0.5511888265609741, | |
| "eval_runtime": 164.5846, | |
| "eval_samples_per_second": 34.365, | |
| "eval_steps_per_second": 4.296, | |
| "eval_wer": 0.42080852497953813, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 2.0123839009287927, | |
| "eval_loss": 0.527522623538971, | |
| "eval_runtime": 162.1572, | |
| "eval_samples_per_second": 34.88, | |
| "eval_steps_per_second": 4.36, | |
| "eval_wer": 0.38652886328256647, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 2.1671826625387, | |
| "eval_loss": 0.4955059587955475, | |
| "eval_runtime": 160.6015, | |
| "eval_samples_per_second": 35.218, | |
| "eval_steps_per_second": 4.402, | |
| "eval_wer": 0.37545537706023013, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.321981424148607, | |
| "grad_norm": 0.4153783619403839, | |
| "learning_rate": 0.00023393333333333332, | |
| "loss": 0.4816, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.321981424148607, | |
| "eval_loss": 0.4908938407897949, | |
| "eval_runtime": 160.2948, | |
| "eval_samples_per_second": 35.285, | |
| "eval_steps_per_second": 4.411, | |
| "eval_wer": 0.37325672834651985, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.476780185758514, | |
| "eval_loss": 0.4982919991016388, | |
| "eval_runtime": 162.1379, | |
| "eval_samples_per_second": 34.884, | |
| "eval_steps_per_second": 4.36, | |
| "eval_wer": 0.37279132095456663, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "eval_loss": 0.48909762501716614, | |
| "eval_runtime": 161.7559, | |
| "eval_samples_per_second": 34.966, | |
| "eval_steps_per_second": 4.371, | |
| "eval_wer": 0.36550528799088444, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.7863777089783284, | |
| "eval_loss": 0.47961312532424927, | |
| "eval_runtime": 160.2886, | |
| "eval_samples_per_second": 35.286, | |
| "eval_steps_per_second": 4.411, | |
| "eval_wer": 0.3570958578742116, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.9411764705882355, | |
| "eval_loss": 0.46432051062583923, | |
| "eval_runtime": 162.1185, | |
| "eval_samples_per_second": 34.888, | |
| "eval_steps_per_second": 4.361, | |
| "eval_wer": 0.3591982154033798, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 3.0959752321981426, | |
| "grad_norm": 0.8931769728660583, | |
| "learning_rate": 0.00020079999999999997, | |
| "loss": 0.4017, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.0959752321981426, | |
| "eval_loss": 0.5084750652313232, | |
| "eval_runtime": 162.3398, | |
| "eval_samples_per_second": 34.841, | |
| "eval_steps_per_second": 4.355, | |
| "eval_wer": 0.3697902457030059, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 3.2507739938080498, | |
| "eval_loss": 0.6755269169807434, | |
| "eval_runtime": 163.0508, | |
| "eval_samples_per_second": 34.689, | |
| "eval_steps_per_second": 4.336, | |
| "eval_wer": 0.4530018776780986, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 3.405572755417957, | |
| "eval_loss": 0.710012674331665, | |
| "eval_runtime": 161.8251, | |
| "eval_samples_per_second": 34.951, | |
| "eval_steps_per_second": 4.369, | |
| "eval_wer": 0.5108247339956027, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.560371517027864, | |
| "eval_loss": 0.8310704231262207, | |
| "eval_runtime": 162.1264, | |
| "eval_samples_per_second": 34.886, | |
| "eval_steps_per_second": 4.361, | |
| "eval_wer": 0.5642984384779574, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.715170278637771, | |
| "eval_loss": 0.7031980156898499, | |
| "eval_runtime": 166.0393, | |
| "eval_samples_per_second": 34.064, | |
| "eval_steps_per_second": 4.258, | |
| "eval_wer": 0.5028807112708832, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.8699690402476783, | |
| "grad_norm": 4.817399024963379, | |
| "learning_rate": 0.00016766666666666666, | |
| "loss": 0.6839, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.8699690402476783, | |
| "eval_loss": 0.7070674896240234, | |
| "eval_runtime": 164.0683, | |
| "eval_samples_per_second": 34.473, | |
| "eval_steps_per_second": 4.309, | |
| "eval_wer": 0.5006660140264159, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 4.024767801857585, | |
| "eval_loss": 0.8223607540130615, | |
| "eval_runtime": 164.3114, | |
| "eval_samples_per_second": 34.422, | |
| "eval_steps_per_second": 4.303, | |
| "eval_wer": 0.5069409895524065, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 4.179566563467493, | |
| "eval_loss": 0.8343736529350281, | |
| "eval_runtime": 163.8342, | |
| "eval_samples_per_second": 34.523, | |
| "eval_steps_per_second": 4.315, | |
| "eval_wer": 0.5162491373914718, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 4.3343653250774, | |
| "eval_loss": 0.9089197516441345, | |
| "eval_runtime": 162.2031, | |
| "eval_samples_per_second": 34.87, | |
| "eval_steps_per_second": 4.359, | |
| "eval_wer": 0.5620195471104621, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 4.489164086687307, | |
| "eval_loss": 0.9664539098739624, | |
| "eval_runtime": 163.3849, | |
| "eval_samples_per_second": 34.618, | |
| "eval_steps_per_second": 4.327, | |
| "eval_wer": 0.5640095649243312, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.643962848297214, | |
| "grad_norm": 7.185929775238037, | |
| "learning_rate": 0.00013446666666666666, | |
| "loss": 0.8292, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.643962848297214, | |
| "eval_loss": 0.9127740859985352, | |
| "eval_runtime": 162.7094, | |
| "eval_samples_per_second": 34.761, | |
| "eval_steps_per_second": 4.345, | |
| "eval_wer": 0.5414774277414902, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.798761609907121, | |
| "eval_loss": 1.1924536228179932, | |
| "eval_runtime": 161.8545, | |
| "eval_samples_per_second": 34.945, | |
| "eval_steps_per_second": 4.368, | |
| "eval_wer": 0.5938598321323683, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.953560371517028, | |
| "eval_loss": 1.4327375888824463, | |
| "eval_runtime": 167.5589, | |
| "eval_samples_per_second": 33.755, | |
| "eval_steps_per_second": 4.219, | |
| "eval_wer": 0.6999406204361991, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 5.108359133126935, | |
| "eval_loss": 1.2741221189498901, | |
| "eval_runtime": 166.0465, | |
| "eval_samples_per_second": 34.063, | |
| "eval_steps_per_second": 4.258, | |
| "eval_wer": 0.7826707964885815, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 5.2631578947368425, | |
| "eval_loss": 1.9348175525665283, | |
| "eval_runtime": 163.4154, | |
| "eval_samples_per_second": 34.611, | |
| "eval_steps_per_second": 4.326, | |
| "eval_wer": 0.8741795188650479, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 5.41795665634675, | |
| "grad_norm": 0.6057500839233398, | |
| "learning_rate": 0.00010126666666666666, | |
| "loss": 1.4131, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.41795665634675, | |
| "eval_loss": 1.9216177463531494, | |
| "eval_runtime": 161.6687, | |
| "eval_samples_per_second": 34.985, | |
| "eval_steps_per_second": 4.373, | |
| "eval_wer": 0.9870167386175795, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 5.572755417956657, | |
| "eval_loss": 1.8565247058868408, | |
| "eval_runtime": 164.0672, | |
| "eval_samples_per_second": 34.474, | |
| "eval_steps_per_second": 4.309, | |
| "eval_wer": 0.9367045946943557, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 5.727554179566564, | |
| "eval_loss": 1.7827845811843872, | |
| "eval_runtime": 163.0587, | |
| "eval_samples_per_second": 34.687, | |
| "eval_steps_per_second": 4.336, | |
| "eval_wer": 0.8240278602493941, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 5.882352941176471, | |
| "eval_loss": 1.6846531629562378, | |
| "eval_runtime": 162.8654, | |
| "eval_samples_per_second": 34.728, | |
| "eval_steps_per_second": 4.341, | |
| "eval_wer": 0.8059090690247308, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 6.037151702786378, | |
| "eval_loss": 1.6440324783325195, | |
| "eval_runtime": 163.0292, | |
| "eval_samples_per_second": 34.693, | |
| "eval_steps_per_second": 4.337, | |
| "eval_wer": 0.7983823080996935, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 6.191950464396285, | |
| "grad_norm": 0.5905019640922546, | |
| "learning_rate": 6.806666666666666e-05, | |
| "loss": 1.7728, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.191950464396285, | |
| "eval_loss": 1.6765395402908325, | |
| "eval_runtime": 166.4929, | |
| "eval_samples_per_second": 33.971, | |
| "eval_steps_per_second": 4.246, | |
| "eval_wer": 0.8053313219174785, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 6.346749226006192, | |
| "eval_loss": 1.6733070611953735, | |
| "eval_runtime": 163.1125, | |
| "eval_samples_per_second": 34.675, | |
| "eval_steps_per_second": 4.334, | |
| "eval_wer": 0.8024265378504598, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 6.5015479876160995, | |
| "eval_loss": 1.6601324081420898, | |
| "eval_runtime": 163.6995, | |
| "eval_samples_per_second": 34.551, | |
| "eval_steps_per_second": 4.319, | |
| "eval_wer": 0.7899568294522636, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 6.656346749226007, | |
| "eval_loss": 1.6604827642440796, | |
| "eval_runtime": 162.8495, | |
| "eval_samples_per_second": 34.731, | |
| "eval_steps_per_second": 4.341, | |
| "eval_wer": 0.7972910080082168, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 6.811145510835914, | |
| "eval_loss": 1.6598834991455078, | |
| "eval_runtime": 163.8109, | |
| "eval_samples_per_second": 34.528, | |
| "eval_steps_per_second": 4.316, | |
| "eval_wer": 0.7804560992441142, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 6.965944272445821, | |
| "grad_norm": 0.9840993881225586, | |
| "learning_rate": 3.493333333333333e-05, | |
| "loss": 1.6777, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 6.965944272445821, | |
| "eval_loss": 1.635949730873108, | |
| "eval_runtime": 163.8566, | |
| "eval_samples_per_second": 34.518, | |
| "eval_steps_per_second": 4.315, | |
| "eval_wer": 0.7693184188987499, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 7.120743034055727, | |
| "eval_loss": 1.6399564743041992, | |
| "eval_runtime": 164.6612, | |
| "eval_samples_per_second": 34.349, | |
| "eval_steps_per_second": 4.294, | |
| "eval_wer": 0.7651137038404134, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 7.275541795665634, | |
| "eval_loss": 1.6759086847305298, | |
| "eval_runtime": 161.5503, | |
| "eval_samples_per_second": 35.011, | |
| "eval_steps_per_second": 4.376, | |
| "eval_wer": 0.7672481584310956, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 7.430340557275541, | |
| "eval_loss": 1.6848710775375366, | |
| "eval_runtime": 162.0136, | |
| "eval_samples_per_second": 34.911, | |
| "eval_steps_per_second": 4.364, | |
| "eval_wer": 0.7686283320761984, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 7.585139318885449, | |
| "eval_loss": 1.6857717037200928, | |
| "eval_runtime": 161.3723, | |
| "eval_samples_per_second": 35.049, | |
| "eval_steps_per_second": 4.381, | |
| "eval_wer": 0.7689974482836096, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 7.739938080495356, | |
| "grad_norm": 1.753746747970581, | |
| "learning_rate": 1.8e-06, | |
| "loss": 1.683, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 7.739938080495356, | |
| "eval_loss": 1.6955511569976807, | |
| "eval_runtime": 160.8529, | |
| "eval_samples_per_second": 35.163, | |
| "eval_steps_per_second": 4.395, | |
| "eval_wer": 0.7701529424981143, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 7.739938080495356, | |
| "step": 5000, | |
| "total_flos": 4.528229347801944e+19, | |
| "train_loss": 1.3220068420410156, | |
| "train_runtime": 20543.6538, | |
| "train_samples_per_second": 15.577, | |
| "train_steps_per_second": 0.243 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "save_steps": 400, | |
| "total_flos": 4.528229347801944e+19, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |