| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9619084263178146, | |
| "eval_steps": 500, | |
| "global_step": 15000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03206361421059382, | |
| "grad_norm": 4.760500431060791, | |
| "learning_rate": 9.859999999999998e-05, | |
| "loss": 5.7778, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.03206361421059382, | |
| "eval_loss": 2.885216236114502, | |
| "eval_runtime": 184.1165, | |
| "eval_samples_per_second": 38.139, | |
| "eval_steps_per_second": 0.597, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.06412722842118763, | |
| "grad_norm": 5.567607402801514, | |
| "learning_rate": 0.0001986, | |
| "loss": 1.4914, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.06412722842118763, | |
| "eval_loss": 1.2011666297912598, | |
| "eval_runtime": 186.6136, | |
| "eval_samples_per_second": 37.629, | |
| "eval_steps_per_second": 0.589, | |
| "eval_wer": 0.7805710944870381, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.09619084263178146, | |
| "grad_norm": 5.768416881561279, | |
| "learning_rate": 0.00029859999999999994, | |
| "loss": 0.8803, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.09619084263178146, | |
| "eval_loss": 1.1211999654769897, | |
| "eval_runtime": 186.8607, | |
| "eval_samples_per_second": 37.579, | |
| "eval_steps_per_second": 0.589, | |
| "eval_wer": 0.7589657050983936, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.12825445684237527, | |
| "grad_norm": 9.690132141113281, | |
| "learning_rate": 0.00028904444444444443, | |
| "loss": 0.7723, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.12825445684237527, | |
| "eval_loss": 0.9680694937705994, | |
| "eval_runtime": 188.1934, | |
| "eval_samples_per_second": 37.313, | |
| "eval_steps_per_second": 0.585, | |
| "eval_wer": 0.6770220173002565, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.16031807105296908, | |
| "grad_norm": 5.588994979858398, | |
| "learning_rate": 0.0002779333333333333, | |
| "loss": 0.6988, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.16031807105296908, | |
| "eval_loss": 0.9452723860740662, | |
| "eval_runtime": 187.5323, | |
| "eval_samples_per_second": 37.444, | |
| "eval_steps_per_second": 0.587, | |
| "eval_wer": 0.6598812102207045, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.19238168526356292, | |
| "grad_norm": 5.276751518249512, | |
| "learning_rate": 0.0002668222222222222, | |
| "loss": 0.6392, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.19238168526356292, | |
| "eval_loss": 0.8690649271011353, | |
| "eval_runtime": 187.6799, | |
| "eval_samples_per_second": 37.415, | |
| "eval_steps_per_second": 0.586, | |
| "eval_wer": 0.6200055807278864, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.22444529947415673, | |
| "grad_norm": 6.366265296936035, | |
| "learning_rate": 0.0002557111111111111, | |
| "loss": 0.6114, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.22444529947415673, | |
| "eval_loss": 0.8661066293716431, | |
| "eval_runtime": 188.4956, | |
| "eval_samples_per_second": 37.253, | |
| "eval_steps_per_second": 0.584, | |
| "eval_wer": 0.619155184097583, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.25650891368475054, | |
| "grad_norm": 11.82204818725586, | |
| "learning_rate": 0.0002446222222222222, | |
| "loss": 0.5807, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.25650891368475054, | |
| "eval_loss": 0.7884626984596252, | |
| "eval_runtime": 188.2678, | |
| "eval_samples_per_second": 37.298, | |
| "eval_steps_per_second": 0.584, | |
| "eval_wer": 0.5793592792888558, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.2885725278953444, | |
| "grad_norm": 14.343709945678711, | |
| "learning_rate": 0.00023355555555555553, | |
| "loss": 0.5534, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.2885725278953444, | |
| "eval_loss": 0.7738627791404724, | |
| "eval_runtime": 188.8344, | |
| "eval_samples_per_second": 37.186, | |
| "eval_steps_per_second": 0.583, | |
| "eval_wer": 0.549037324439602, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.32063614210593816, | |
| "grad_norm": 8.92720890045166, | |
| "learning_rate": 0.00022244444444444444, | |
| "loss": 0.5358, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.32063614210593816, | |
| "eval_loss": 0.7416187524795532, | |
| "eval_runtime": 189.1576, | |
| "eval_samples_per_second": 37.122, | |
| "eval_steps_per_second": 0.582, | |
| "eval_wer": 0.5414767668983111, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.352699756316532, | |
| "grad_norm": 7.230262279510498, | |
| "learning_rate": 0.0002113333333333333, | |
| "loss": 0.5189, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.352699756316532, | |
| "eval_loss": 0.7361556887626648, | |
| "eval_runtime": 188.5731, | |
| "eval_samples_per_second": 37.238, | |
| "eval_steps_per_second": 0.583, | |
| "eval_wer": 0.5303285985729281, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "grad_norm": 7.765280246734619, | |
| "learning_rate": 0.0002002444444444444, | |
| "loss": 0.4991, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.38476337052712584, | |
| "eval_loss": 0.7187824845314026, | |
| "eval_runtime": 188.8496, | |
| "eval_samples_per_second": 37.183, | |
| "eval_steps_per_second": 0.582, | |
| "eval_wer": 0.5065839301611768, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.4168269847377196, | |
| "grad_norm": 5.903895854949951, | |
| "learning_rate": 0.00018913333333333331, | |
| "loss": 0.48, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.4168269847377196, | |
| "eval_loss": 0.6984608173370361, | |
| "eval_runtime": 189.9958, | |
| "eval_samples_per_second": 36.959, | |
| "eval_steps_per_second": 0.579, | |
| "eval_wer": 0.5177719608286052, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.44889059894831346, | |
| "grad_norm": 3.641240358352661, | |
| "learning_rate": 0.00017804444444444444, | |
| "loss": 0.463, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.44889059894831346, | |
| "eval_loss": 0.6681538820266724, | |
| "eval_runtime": 189.2879, | |
| "eval_samples_per_second": 37.097, | |
| "eval_steps_per_second": 0.581, | |
| "eval_wer": 0.49330977026003536, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.4809542131589073, | |
| "grad_norm": 6.733245849609375, | |
| "learning_rate": 0.00016693333333333332, | |
| "loss": 0.4477, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.4809542131589073, | |
| "eval_loss": 0.6624513268470764, | |
| "eval_runtime": 189.8301, | |
| "eval_samples_per_second": 36.991, | |
| "eval_steps_per_second": 0.579, | |
| "eval_wer": 0.48671919637518435, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.5130178273695011, | |
| "grad_norm": 7.44530725479126, | |
| "learning_rate": 0.00015584444444444442, | |
| "loss": 0.4431, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.5130178273695011, | |
| "eval_loss": 0.6373856663703918, | |
| "eval_runtime": 189.4419, | |
| "eval_samples_per_second": 37.067, | |
| "eval_steps_per_second": 0.581, | |
| "eval_wer": 0.47356462350017936, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.5450814415800949, | |
| "grad_norm": 3.664278745651245, | |
| "learning_rate": 0.00014473333333333332, | |
| "loss": 0.4392, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.5450814415800949, | |
| "eval_loss": 0.6391619443893433, | |
| "eval_runtime": 189.4208, | |
| "eval_samples_per_second": 37.071, | |
| "eval_steps_per_second": 0.581, | |
| "eval_wer": 0.4772452464157111, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.5771450557906888, | |
| "grad_norm": 11.637319564819336, | |
| "learning_rate": 0.00013362222222222222, | |
| "loss": 0.4197, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.5771450557906888, | |
| "eval_loss": 0.6158761978149414, | |
| "eval_runtime": 188.8242, | |
| "eval_samples_per_second": 37.188, | |
| "eval_steps_per_second": 0.583, | |
| "eval_wer": 0.45473631060736924, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.6092086700012825, | |
| "grad_norm": 7.102973461151123, | |
| "learning_rate": 0.0001225111111111111, | |
| "loss": 0.4147, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.6092086700012825, | |
| "eval_loss": 0.5994922518730164, | |
| "eval_runtime": 188.4237, | |
| "eval_samples_per_second": 37.267, | |
| "eval_steps_per_second": 0.584, | |
| "eval_wer": 0.45217183326911065, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.6412722842118763, | |
| "grad_norm": 6.166309833526611, | |
| "learning_rate": 0.0001114, | |
| "loss": 0.3912, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.6412722842118763, | |
| "eval_loss": 0.5847700834274292, | |
| "eval_runtime": 188.8879, | |
| "eval_samples_per_second": 37.175, | |
| "eval_steps_per_second": 0.582, | |
| "eval_wer": 0.4285866142255411, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.6733358984224702, | |
| "grad_norm": 8.538312911987305, | |
| "learning_rate": 0.00010028888888888889, | |
| "loss": 0.3742, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.6733358984224702, | |
| "eval_loss": 0.585001528263092, | |
| "eval_runtime": 189.3338, | |
| "eval_samples_per_second": 37.088, | |
| "eval_steps_per_second": 0.581, | |
| "eval_wer": 0.4259025498611462, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.705399512633064, | |
| "grad_norm": 16.837343215942383, | |
| "learning_rate": 8.917777777777777e-05, | |
| "loss": 0.402, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.705399512633064, | |
| "eval_loss": 0.6351918578147888, | |
| "eval_runtime": 188.7345, | |
| "eval_samples_per_second": 37.206, | |
| "eval_steps_per_second": 0.583, | |
| "eval_wer": 0.44894298356342766, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.7374631268436578, | |
| "grad_norm": 4.569055557250977, | |
| "learning_rate": 7.806666666666666e-05, | |
| "loss": 0.5746, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.7374631268436578, | |
| "eval_loss": 0.7711716294288635, | |
| "eval_runtime": 188.4281, | |
| "eval_samples_per_second": 37.266, | |
| "eval_steps_per_second": 0.584, | |
| "eval_wer": 0.5170810135664837, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.275555555555556e-05, | |
| "loss": 0.5783, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.7695267410542517, | |
| "eval_loss": NaN, | |
| "eval_runtime": 184.9484, | |
| "eval_samples_per_second": 37.967, | |
| "eval_steps_per_second": 0.595, | |
| "eval_wer": 1.0, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.8015903552648455, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.275555555555556e-05, | |
| "loss": 0.0, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.8015903552648455, | |
| "eval_loss": NaN, | |
| "eval_runtime": 184.8998, | |
| "eval_samples_per_second": 37.977, | |
| "eval_steps_per_second": 0.595, | |
| "eval_wer": 1.0, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.8336539694754392, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.275555555555556e-05, | |
| "loss": 0.0, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.8336539694754392, | |
| "eval_loss": NaN, | |
| "eval_runtime": 184.8775, | |
| "eval_samples_per_second": 37.982, | |
| "eval_steps_per_second": 0.595, | |
| "eval_wer": 1.0, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.8657175836860331, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.275555555555556e-05, | |
| "loss": 0.0, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.8657175836860331, | |
| "eval_loss": NaN, | |
| "eval_runtime": 185.2742, | |
| "eval_samples_per_second": 37.901, | |
| "eval_steps_per_second": 0.594, | |
| "eval_wer": 1.0, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.8977811978966269, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.275555555555556e-05, | |
| "loss": 0.0, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.8977811978966269, | |
| "eval_loss": NaN, | |
| "eval_runtime": 185.0536, | |
| "eval_samples_per_second": 37.946, | |
| "eval_steps_per_second": 0.594, | |
| "eval_wer": 1.0, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.9298448121072207, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.275555555555556e-05, | |
| "loss": 0.0, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.9298448121072207, | |
| "eval_loss": NaN, | |
| "eval_runtime": 184.7301, | |
| "eval_samples_per_second": 38.012, | |
| "eval_steps_per_second": 0.595, | |
| "eval_wer": 1.0, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.9619084263178146, | |
| "grad_norm": NaN, | |
| "learning_rate": 7.275555555555556e-05, | |
| "loss": 0.0, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.9619084263178146, | |
| "eval_loss": NaN, | |
| "eval_runtime": 185.3079, | |
| "eval_samples_per_second": 37.894, | |
| "eval_steps_per_second": 0.594, | |
| "eval_wer": 1.0, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.9619084263178146, | |
| "step": 15000, | |
| "total_flos": 1.7109669148845115e+19, | |
| "train_loss": 0.6328924499511719, | |
| "train_runtime": 11517.021, | |
| "train_samples_per_second": 10.419, | |
| "train_steps_per_second": 1.302 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 15000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.7109669148845115e+19, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |