| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.4722536806342017, | |
| "eval_steps": 100, | |
| "global_step": 2600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.056625141562853906, | |
| "eval_loss": 3.599447011947632, | |
| "eval_runtime": 153.6415, | |
| "eval_samples_per_second": 36.813, | |
| "eval_steps_per_second": 4.602, | |
| "eval_wer": 1.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11325028312570781, | |
| "eval_loss": 3.0223705768585205, | |
| "eval_runtime": 151.1357, | |
| "eval_samples_per_second": 37.423, | |
| "eval_steps_per_second": 4.678, | |
| "eval_wer": 1.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.16987542468856173, | |
| "eval_loss": 1.9626648426055908, | |
| "eval_runtime": 153.3369, | |
| "eval_samples_per_second": 36.886, | |
| "eval_steps_per_second": 4.611, | |
| "eval_wer": 0.9060519009484682, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.22650056625141562, | |
| "eval_loss": 0.9921229481697083, | |
| "eval_runtime": 154.0126, | |
| "eval_samples_per_second": 36.724, | |
| "eval_steps_per_second": 4.591, | |
| "eval_wer": 0.6951581582706103, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.28312570781426954, | |
| "grad_norm": 1.7500211000442505, | |
| "learning_rate": 0.0002982, | |
| "loss": 3.6184, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.28312570781426954, | |
| "eval_loss": 0.8504024147987366, | |
| "eval_runtime": 153.4805, | |
| "eval_samples_per_second": 36.852, | |
| "eval_steps_per_second": 4.606, | |
| "eval_wer": 0.6157981736771998, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.33975084937712347, | |
| "eval_loss": 0.7935531735420227, | |
| "eval_runtime": 152.4296, | |
| "eval_samples_per_second": 37.106, | |
| "eval_steps_per_second": 4.638, | |
| "eval_wer": 0.5963634029304618, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.39637599093997733, | |
| "eval_loss": 0.7600815892219543, | |
| "eval_runtime": 156.4822, | |
| "eval_samples_per_second": 36.145, | |
| "eval_steps_per_second": 4.518, | |
| "eval_wer": 0.5606072764038452, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.45300113250283125, | |
| "eval_loss": 0.6909334063529968, | |
| "eval_runtime": 152.429, | |
| "eval_samples_per_second": 37.106, | |
| "eval_steps_per_second": 4.638, | |
| "eval_wer": 0.5127666062172008, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.5096262740656852, | |
| "eval_loss": 0.6616554856300354, | |
| "eval_runtime": 152.3205, | |
| "eval_samples_per_second": 37.132, | |
| "eval_steps_per_second": 4.642, | |
| "eval_wer": 0.4916146426794627, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.5662514156285391, | |
| "grad_norm": 3.9567878246307373, | |
| "learning_rate": 0.00022914285714285712, | |
| "loss": 0.4662, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5662514156285391, | |
| "eval_loss": 0.6466770172119141, | |
| "eval_runtime": 152.3467, | |
| "eval_samples_per_second": 37.126, | |
| "eval_steps_per_second": 4.641, | |
| "eval_wer": 0.4811510006258927, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.622876557191393, | |
| "eval_loss": 0.6144490838050842, | |
| "eval_runtime": 152.8109, | |
| "eval_samples_per_second": 37.013, | |
| "eval_steps_per_second": 4.627, | |
| "eval_wer": 0.46394697565437887, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.6795016987542469, | |
| "eval_loss": 0.5942133069038391, | |
| "eval_runtime": 157.9769, | |
| "eval_samples_per_second": 35.803, | |
| "eval_steps_per_second": 4.475, | |
| "eval_wer": 0.45369196450064997, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.7361268403171007, | |
| "eval_loss": 0.5675162076950073, | |
| "eval_runtime": 165.0294, | |
| "eval_samples_per_second": 34.273, | |
| "eval_steps_per_second": 4.284, | |
| "eval_wer": 0.4351077658840333, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.7927519818799547, | |
| "eval_loss": 0.5538555383682251, | |
| "eval_runtime": 154.4584, | |
| "eval_samples_per_second": 36.618, | |
| "eval_steps_per_second": 4.577, | |
| "eval_wer": 0.42295902810097735, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.8493771234428086, | |
| "grad_norm": 1.9575612545013428, | |
| "learning_rate": 0.00015785714285714285, | |
| "loss": 0.3508, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8493771234428086, | |
| "eval_loss": 0.5447892546653748, | |
| "eval_runtime": 152.2149, | |
| "eval_samples_per_second": 37.158, | |
| "eval_steps_per_second": 4.645, | |
| "eval_wer": 0.41440516120749143, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.9060022650056625, | |
| "eval_loss": 0.5326293110847473, | |
| "eval_runtime": 157.3458, | |
| "eval_samples_per_second": 35.946, | |
| "eval_steps_per_second": 4.493, | |
| "eval_wer": 0.406621623790342, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.9626274065685164, | |
| "eval_loss": 0.5154709219932556, | |
| "eval_runtime": 154.1323, | |
| "eval_samples_per_second": 36.696, | |
| "eval_steps_per_second": 4.587, | |
| "eval_wer": 0.3989343775577346, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.0192525481313703, | |
| "eval_loss": 0.5067195296287537, | |
| "eval_runtime": 153.6675, | |
| "eval_samples_per_second": 36.807, | |
| "eval_steps_per_second": 4.601, | |
| "eval_wer": 0.38559804849866, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.0758776896942241, | |
| "eval_loss": 0.49161842465400696, | |
| "eval_runtime": 153.2473, | |
| "eval_samples_per_second": 36.908, | |
| "eval_steps_per_second": 4.613, | |
| "eval_wer": 0.3724061562163984, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.1325028312570782, | |
| "grad_norm": 0.9259862303733826, | |
| "learning_rate": 8.671428571428571e-05, | |
| "loss": 0.2774, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.1325028312570782, | |
| "eval_loss": 0.48551619052886963, | |
| "eval_runtime": 153.8121, | |
| "eval_samples_per_second": 36.772, | |
| "eval_steps_per_second": 4.597, | |
| "eval_wer": 0.36971000304922086, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.189127972819932, | |
| "eval_loss": 0.4797590672969818, | |
| "eval_runtime": 152.9797, | |
| "eval_samples_per_second": 36.972, | |
| "eval_steps_per_second": 4.622, | |
| "eval_wer": 0.3661311806904078, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 1.245753114382786, | |
| "eval_loss": 0.47735241055488586, | |
| "eval_runtime": 152.7086, | |
| "eval_samples_per_second": 37.038, | |
| "eval_steps_per_second": 4.63, | |
| "eval_wer": 0.36462261879924895, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.3023782559456398, | |
| "eval_loss": 0.46992629766464233, | |
| "eval_runtime": 152.5925, | |
| "eval_samples_per_second": 37.066, | |
| "eval_steps_per_second": 4.633, | |
| "eval_wer": 0.3584760315193144, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 1.3590033975084936, | |
| "eval_loss": 0.4651219844818115, | |
| "eval_runtime": 153.5802, | |
| "eval_samples_per_second": 36.828, | |
| "eval_steps_per_second": 4.603, | |
| "eval_wer": 0.3550255974065574, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.4156285390713477, | |
| "grad_norm": 0.43387308716773987, | |
| "learning_rate": 1.5428571428571428e-05, | |
| "loss": 0.2328, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.4156285390713477, | |
| "eval_loss": 0.4611109495162964, | |
| "eval_runtime": 153.4779, | |
| "eval_samples_per_second": 36.852, | |
| "eval_steps_per_second": 4.607, | |
| "eval_wer": 0.3570156152204266, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.4722536806342017, | |
| "eval_loss": 0.4593363106250763, | |
| "eval_runtime": 153.5126, | |
| "eval_samples_per_second": 36.844, | |
| "eval_steps_per_second": 4.605, | |
| "eval_wer": 0.3553465680216976, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.4722536806342017, | |
| "step": 2600, | |
| "total_flos": 2.3804122051094954e+19, | |
| "train_loss": 0.9595056893275334, | |
| "train_runtime": 10902.6713, | |
| "train_samples_per_second": 15.262, | |
| "train_steps_per_second": 0.238 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2600, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 400, | |
| "total_flos": 2.3804122051094954e+19, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |