| { | |
| "best_metric": 99.44547134935306, | |
| "best_model_checkpoint": "./output_dir/checkpoint-1000", | |
| "epoch": 5.0761421319796955, | |
| "eval_steps": 1000, | |
| "global_step": 5000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.10152284263959391, | |
| "grad_norm": 45.86384582519531, | |
| "learning_rate": 1.9000000000000002e-06, | |
| "loss": 2.9237, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.20304568527918782, | |
| "grad_norm": 45.55335998535156, | |
| "learning_rate": 3.900000000000001e-06, | |
| "loss": 2.3206, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.30456852791878175, | |
| "grad_norm": 48.86751174926758, | |
| "learning_rate": 5.9e-06, | |
| "loss": 2.3718, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.40609137055837563, | |
| "grad_norm": 57.0309944152832, | |
| "learning_rate": 7.9e-06, | |
| "loss": 2.2004, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.5076142131979695, | |
| "grad_norm": 50.58991241455078, | |
| "learning_rate": 9.9e-06, | |
| "loss": 2.1753, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.6091370558375635, | |
| "grad_norm": 55.411651611328125, | |
| "learning_rate": 9.78888888888889e-06, | |
| "loss": 2.1806, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.7106598984771574, | |
| "grad_norm": 52.69567108154297, | |
| "learning_rate": 9.566666666666668e-06, | |
| "loss": 2.1058, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.8121827411167513, | |
| "grad_norm": 54.5142822265625, | |
| "learning_rate": 9.344444444444446e-06, | |
| "loss": 1.9591, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.9137055837563451, | |
| "grad_norm": 34.65861511230469, | |
| "learning_rate": 9.124444444444444e-06, | |
| "loss": 2.0829, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.015228426395939, | |
| "grad_norm": 47.46334457397461, | |
| "learning_rate": 8.902222222222224e-06, | |
| "loss": 1.9297, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.015228426395939, | |
| "eval_loss": 1.9972329139709473, | |
| "eval_runtime": 143.8629, | |
| "eval_samples_per_second": 3.476, | |
| "eval_steps_per_second": 1.738, | |
| "eval_wer": 99.44547134935306, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.116751269035533, | |
| "grad_norm": 35.19301986694336, | |
| "learning_rate": 8.68e-06, | |
| "loss": 1.3307, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.218274111675127, | |
| "grad_norm": 32.941097259521484, | |
| "learning_rate": 8.457777777777778e-06, | |
| "loss": 1.3622, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.3197969543147208, | |
| "grad_norm": 60.84815216064453, | |
| "learning_rate": 8.235555555555557e-06, | |
| "loss": 1.3763, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.4213197969543148, | |
| "grad_norm": 41.705345153808594, | |
| "learning_rate": 8.013333333333333e-06, | |
| "loss": 1.3411, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.5228426395939088, | |
| "grad_norm": 46.46513366699219, | |
| "learning_rate": 7.791111111111111e-06, | |
| "loss": 1.3745, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.6243654822335025, | |
| "grad_norm": 52.48996353149414, | |
| "learning_rate": 7.56888888888889e-06, | |
| "loss": 1.4233, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.7258883248730963, | |
| "grad_norm": 71.2730712890625, | |
| "learning_rate": 7.346666666666668e-06, | |
| "loss": 1.2765, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.8274111675126905, | |
| "grad_norm": 41.1866455078125, | |
| "learning_rate": 7.124444444444445e-06, | |
| "loss": 1.4133, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.9289340101522843, | |
| "grad_norm": 40.041481018066406, | |
| "learning_rate": 6.902222222222223e-06, | |
| "loss": 1.2927, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.030456852791878, | |
| "grad_norm": 29.876537322998047, | |
| "learning_rate": 6.680000000000001e-06, | |
| "loss": 1.1678, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.030456852791878, | |
| "eval_loss": 1.9988778829574585, | |
| "eval_runtime": 150.1664, | |
| "eval_samples_per_second": 3.33, | |
| "eval_steps_per_second": 1.665, | |
| "eval_wer": 99.44547134935306, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.1319796954314723, | |
| "grad_norm": 31.42411231994629, | |
| "learning_rate": 6.457777777777778e-06, | |
| "loss": 0.6302, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.233502538071066, | |
| "grad_norm": 29.325563430786133, | |
| "learning_rate": 6.235555555555556e-06, | |
| "loss": 0.7149, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.33502538071066, | |
| "grad_norm": 28.083486557006836, | |
| "learning_rate": 6.013333333333335e-06, | |
| "loss": 0.7032, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.436548223350254, | |
| "grad_norm": 59.06522750854492, | |
| "learning_rate": 5.791111111111112e-06, | |
| "loss": 0.7279, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.5380710659898478, | |
| "grad_norm": 25.313156127929688, | |
| "learning_rate": 5.56888888888889e-06, | |
| "loss": 0.7762, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.6395939086294415, | |
| "grad_norm": 42.13996887207031, | |
| "learning_rate": 5.346666666666667e-06, | |
| "loss": 0.6803, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.7411167512690353, | |
| "grad_norm": 40.32999038696289, | |
| "learning_rate": 5.124444444444445e-06, | |
| "loss": 0.7569, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.8426395939086295, | |
| "grad_norm": 26.812973022460938, | |
| "learning_rate": 4.902222222222222e-06, | |
| "loss": 0.7796, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.9441624365482233, | |
| "grad_norm": 35.564395904541016, | |
| "learning_rate": 4.680000000000001e-06, | |
| "loss": 0.7115, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 3.045685279187817, | |
| "grad_norm": 26.95085906982422, | |
| "learning_rate": 4.457777777777778e-06, | |
| "loss": 0.5123, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.045685279187817, | |
| "eval_loss": 2.0867390632629395, | |
| "eval_runtime": 147.2269, | |
| "eval_samples_per_second": 3.396, | |
| "eval_steps_per_second": 1.698, | |
| "eval_wer": 102.40295748613678, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.1472081218274113, | |
| "grad_norm": 10.427111625671387, | |
| "learning_rate": 4.235555555555556e-06, | |
| "loss": 0.3293, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 3.248730964467005, | |
| "grad_norm": 25.181922912597656, | |
| "learning_rate": 4.013333333333334e-06, | |
| "loss": 0.3875, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 3.350253807106599, | |
| "grad_norm": 25.681047439575195, | |
| "learning_rate": 3.7911111111111114e-06, | |
| "loss": 0.336, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 3.451776649746193, | |
| "grad_norm": 22.395158767700195, | |
| "learning_rate": 3.568888888888889e-06, | |
| "loss": 0.3491, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 3.553299492385787, | |
| "grad_norm": 24.35469627380371, | |
| "learning_rate": 3.346666666666667e-06, | |
| "loss": 0.3499, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.6548223350253806, | |
| "grad_norm": 43.59064865112305, | |
| "learning_rate": 3.124444444444445e-06, | |
| "loss": 0.3475, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 3.7563451776649748, | |
| "grad_norm": 3.5348989963531494, | |
| "learning_rate": 2.9022222222222223e-06, | |
| "loss": 0.3032, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 3.8578680203045685, | |
| "grad_norm": 41.54439926147461, | |
| "learning_rate": 2.68e-06, | |
| "loss": 0.3408, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 3.9593908629441623, | |
| "grad_norm": 20.793079376220703, | |
| "learning_rate": 2.457777777777778e-06, | |
| "loss": 0.2949, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 4.060913705583756, | |
| "grad_norm": 17.149005889892578, | |
| "learning_rate": 2.235555555555556e-06, | |
| "loss": 0.2025, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.060913705583756, | |
| "eval_loss": 2.169760227203369, | |
| "eval_runtime": 148.2971, | |
| "eval_samples_per_second": 3.372, | |
| "eval_steps_per_second": 1.686, | |
| "eval_wer": 103.32717190388169, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 4.16243654822335, | |
| "grad_norm": 1.9773640632629395, | |
| "learning_rate": 2.0133333333333337e-06, | |
| "loss": 0.1643, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 4.2639593908629445, | |
| "grad_norm": 10.312366485595703, | |
| "learning_rate": 1.7911111111111113e-06, | |
| "loss": 0.1461, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 4.365482233502538, | |
| "grad_norm": 1.339500069618225, | |
| "learning_rate": 1.568888888888889e-06, | |
| "loss": 0.1546, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 4.467005076142132, | |
| "grad_norm": 23.366968154907227, | |
| "learning_rate": 1.3466666666666668e-06, | |
| "loss": 0.1184, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 4.568527918781726, | |
| "grad_norm": 3.0241127014160156, | |
| "learning_rate": 1.1244444444444446e-06, | |
| "loss": 0.18, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 4.67005076142132, | |
| "grad_norm": 21.15237045288086, | |
| "learning_rate": 9.022222222222222e-07, | |
| "loss": 0.2008, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 4.771573604060913, | |
| "grad_norm": 4.779166221618652, | |
| "learning_rate": 6.800000000000001e-07, | |
| "loss": 0.1546, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 4.873096446700508, | |
| "grad_norm": 1.9334161281585693, | |
| "learning_rate": 4.5777777777777784e-07, | |
| "loss": 0.1429, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 4.974619289340102, | |
| "grad_norm": 18.270540237426758, | |
| "learning_rate": 2.3555555555555556e-07, | |
| "loss": 0.1373, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 5.0761421319796955, | |
| "grad_norm": 1.0400763750076294, | |
| "learning_rate": 1.3333333333333334e-08, | |
| "loss": 0.0843, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 5.0761421319796955, | |
| "eval_loss": 2.2213687896728516, | |
| "eval_runtime": 142.0959, | |
| "eval_samples_per_second": 3.519, | |
| "eval_steps_per_second": 1.759, | |
| "eval_wer": 104.25138632162663, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 5.0761421319796955, | |
| "step": 5000, | |
| "total_flos": 2.8844110761984e+18, | |
| "train_loss": 0.9465061870574951, | |
| "train_runtime": 2908.7621, | |
| "train_samples_per_second": 3.438, | |
| "train_steps_per_second": 1.719 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 5000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.8844110761984e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |