| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 50.0, | |
| "global_step": 7800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 7.275e-06, | |
| "loss": 16.819, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 1.4775e-05, | |
| "loss": 6.1908, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 2.2274999999999996e-05, | |
| "loss": 4.1987, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 2.9775e-05, | |
| "loss": 3.489, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_loss": 3.3590216636657715, | |
| "eval_runtime": 66.6533, | |
| "eval_samples_per_second": 32.226, | |
| "eval_steps_per_second": 2.025, | |
| "eval_wer": 1.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.21, | |
| "learning_rate": 3.7275e-05, | |
| "loss": 3.2231, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "learning_rate": 4.4775e-05, | |
| "loss": 3.1038, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.49, | |
| "learning_rate": 5.227499999999999e-05, | |
| "loss": 3.0217, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "learning_rate": 5.9774999999999996e-05, | |
| "loss": 2.9903, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "eval_loss": 2.970390796661377, | |
| "eval_runtime": 66.5925, | |
| "eval_samples_per_second": 32.256, | |
| "eval_steps_per_second": 2.027, | |
| "eval_wer": 1.0000977995110025, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "learning_rate": 6.7275e-05, | |
| "loss": 2.906, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 6.41, | |
| "learning_rate": 7.477499999999999e-05, | |
| "loss": 2.5346, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 7.05, | |
| "learning_rate": 7.393014705882353e-05, | |
| "loss": 1.8923, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 7.282720588235294e-05, | |
| "loss": 1.6712, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "eval_loss": 0.6178616285324097, | |
| "eval_runtime": 66.5699, | |
| "eval_samples_per_second": 32.267, | |
| "eval_steps_per_second": 2.028, | |
| "eval_wer": 0.6566259168704156, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "learning_rate": 7.172426470588234e-05, | |
| "loss": 1.5317, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 8.97, | |
| "learning_rate": 7.062132352941176e-05, | |
| "loss": 1.4151, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 9.62, | |
| "learning_rate": 6.951838235294117e-05, | |
| "loss": 1.3153, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 10.26, | |
| "learning_rate": 6.841544117647059e-05, | |
| "loss": 1.2635, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 10.26, | |
| "eval_loss": 0.31762251257896423, | |
| "eval_runtime": 67.311, | |
| "eval_samples_per_second": 31.912, | |
| "eval_steps_per_second": 2.006, | |
| "eval_wer": 0.45310513447432765, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 10.9, | |
| "learning_rate": 6.731249999999999e-05, | |
| "loss": 1.1929, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 11.54, | |
| "learning_rate": 6.62095588235294e-05, | |
| "loss": 1.166, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 12.18, | |
| "learning_rate": 6.510661764705882e-05, | |
| "loss": 1.0971, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 12.82, | |
| "learning_rate": 6.400367647058824e-05, | |
| "loss": 1.0819, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 12.82, | |
| "eval_loss": 0.2516830265522003, | |
| "eval_runtime": 66.7082, | |
| "eval_samples_per_second": 32.2, | |
| "eval_steps_per_second": 2.024, | |
| "eval_wer": 0.35080684596577016, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 13.46, | |
| "learning_rate": 6.290073529411764e-05, | |
| "loss": 1.0892, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 14.1, | |
| "learning_rate": 6.179779411764705e-05, | |
| "loss": 1.0534, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 14.74, | |
| "learning_rate": 6.069485294117646e-05, | |
| "loss": 1.012, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "learning_rate": 5.9591911764705876e-05, | |
| "loss": 1.0136, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "eval_loss": 0.22566433250904083, | |
| "eval_runtime": 66.3228, | |
| "eval_samples_per_second": 32.387, | |
| "eval_steps_per_second": 2.036, | |
| "eval_wer": 0.3123716381418093, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 16.03, | |
| "learning_rate": 5.848897058823529e-05, | |
| "loss": 0.9991, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 16.67, | |
| "learning_rate": 5.73860294117647e-05, | |
| "loss": 0.9722, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 17.31, | |
| "learning_rate": 5.6283088235294115e-05, | |
| "loss": 0.9794, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 17.95, | |
| "learning_rate": 5.5180147058823523e-05, | |
| "loss": 0.9625, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 17.95, | |
| "eval_loss": 0.19747723639011383, | |
| "eval_runtime": 67.1687, | |
| "eval_samples_per_second": 31.979, | |
| "eval_steps_per_second": 2.01, | |
| "eval_wer": 0.2311002444987775, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 18.59, | |
| "learning_rate": 5.407720588235294e-05, | |
| "loss": 0.9505, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 19.23, | |
| "learning_rate": 5.297426470588235e-05, | |
| "loss": 0.9337, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 19.87, | |
| "learning_rate": 5.187132352941176e-05, | |
| "loss": 0.936, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 20.51, | |
| "learning_rate": 5.076838235294117e-05, | |
| "loss": 0.901, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 20.51, | |
| "eval_loss": 0.19860759377479553, | |
| "eval_runtime": 66.128, | |
| "eval_samples_per_second": 32.482, | |
| "eval_steps_per_second": 2.041, | |
| "eval_wer": 0.20968215158924206, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 21.15, | |
| "learning_rate": 4.9665441176470586e-05, | |
| "loss": 0.9179, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 21.79, | |
| "learning_rate": 4.8562499999999995e-05, | |
| "loss": 0.8963, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 22.44, | |
| "learning_rate": 4.745955882352941e-05, | |
| "loss": 0.8826, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 23.08, | |
| "learning_rate": 4.635661764705882e-05, | |
| "loss": 0.8842, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 23.08, | |
| "eval_loss": 0.19037693738937378, | |
| "eval_runtime": 67.9694, | |
| "eval_samples_per_second": 31.602, | |
| "eval_steps_per_second": 1.986, | |
| "eval_wer": 0.2039119804400978, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 23.72, | |
| "learning_rate": 4.5253676470588234e-05, | |
| "loss": 0.8683, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 24.36, | |
| "learning_rate": 4.416176470588235e-05, | |
| "loss": 0.8599, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 4.305882352941176e-05, | |
| "loss": 0.8658, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 25.64, | |
| "learning_rate": 4.1955882352941173e-05, | |
| "loss": 0.8542, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 25.64, | |
| "eval_loss": 0.1846681386232376, | |
| "eval_runtime": 67.5681, | |
| "eval_samples_per_second": 31.79, | |
| "eval_steps_per_second": 1.998, | |
| "eval_wer": 0.19814180929095354, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 26.28, | |
| "learning_rate": 4.085294117647058e-05, | |
| "loss": 0.8468, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 26.92, | |
| "learning_rate": 3.975e-05, | |
| "loss": 0.8462, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 27.56, | |
| "learning_rate": 3.8647058823529406e-05, | |
| "loss": 0.8284, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 28.21, | |
| "learning_rate": 3.754411764705882e-05, | |
| "loss": 0.8244, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 28.21, | |
| "eval_loss": 0.1804967224597931, | |
| "eval_runtime": 66.5917, | |
| "eval_samples_per_second": 32.256, | |
| "eval_steps_per_second": 2.027, | |
| "eval_wer": 0.18474327628361858, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 28.85, | |
| "learning_rate": 3.6441176470588236e-05, | |
| "loss": 0.8288, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 29.49, | |
| "learning_rate": 3.5338235294117645e-05, | |
| "loss": 0.7904, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 30.13, | |
| "learning_rate": 3.423529411764706e-05, | |
| "loss": 0.8157, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 30.77, | |
| "learning_rate": 3.313235294117647e-05, | |
| "loss": 0.7689, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 30.77, | |
| "eval_loss": 0.17355979979038239, | |
| "eval_runtime": 69.3305, | |
| "eval_samples_per_second": 30.982, | |
| "eval_steps_per_second": 1.947, | |
| "eval_wer": 0.18317848410757948, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 31.41, | |
| "learning_rate": 3.202941176470588e-05, | |
| "loss": 0.7948, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 32.05, | |
| "learning_rate": 3.092647058823529e-05, | |
| "loss": 0.8035, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 32.69, | |
| "learning_rate": 2.9823529411764704e-05, | |
| "loss": 0.7754, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 2.8720588235294116e-05, | |
| "loss": 0.7825, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "eval_loss": 0.1698261797428131, | |
| "eval_runtime": 67.6625, | |
| "eval_samples_per_second": 31.746, | |
| "eval_steps_per_second": 1.995, | |
| "eval_wer": 0.18210268948655256, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 33.97, | |
| "learning_rate": 2.7617647058823528e-05, | |
| "loss": 0.7748, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 34.62, | |
| "learning_rate": 2.651470588235294e-05, | |
| "loss": 0.7704, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 35.26, | |
| "learning_rate": 2.5411764705882348e-05, | |
| "loss": 0.7597, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 35.9, | |
| "learning_rate": 2.430882352941176e-05, | |
| "loss": 0.7817, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 35.9, | |
| "eval_loss": 0.17581327259540558, | |
| "eval_runtime": 67.4489, | |
| "eval_samples_per_second": 31.846, | |
| "eval_steps_per_second": 2.002, | |
| "eval_wer": 0.18034229828850856, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 36.54, | |
| "learning_rate": 2.3205882352941172e-05, | |
| "loss": 0.744, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 37.18, | |
| "learning_rate": 2.2102941176470584e-05, | |
| "loss": 0.7508, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 37.82, | |
| "learning_rate": 2.1e-05, | |
| "loss": 0.746, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "learning_rate": 1.989705882352941e-05, | |
| "loss": 0.7488, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "eval_loss": 0.16634251177310944, | |
| "eval_runtime": 67.3054, | |
| "eval_samples_per_second": 31.914, | |
| "eval_steps_per_second": 2.006, | |
| "eval_wer": 0.17603911980440098, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 39.1, | |
| "learning_rate": 1.8794117647058823e-05, | |
| "loss": 0.7486, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 39.74, | |
| "learning_rate": 1.7691176470588234e-05, | |
| "loss": 0.7281, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 40.38, | |
| "learning_rate": 1.6588235294117646e-05, | |
| "loss": 0.7317, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 41.03, | |
| "learning_rate": 1.5485294117647058e-05, | |
| "loss": 0.7171, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 41.03, | |
| "eval_loss": 0.16363976895809174, | |
| "eval_runtime": 68.4211, | |
| "eval_samples_per_second": 31.394, | |
| "eval_steps_per_second": 1.973, | |
| "eval_wer": 0.17212713936430318, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 41.67, | |
| "learning_rate": 1.438235294117647e-05, | |
| "loss": 0.7196, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 42.31, | |
| "learning_rate": 1.3279411764705882e-05, | |
| "loss": 0.7292, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 42.95, | |
| "learning_rate": 1.2176470588235294e-05, | |
| "loss": 0.7319, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 43.59, | |
| "learning_rate": 1.1073529411764704e-05, | |
| "loss": 0.7222, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 43.59, | |
| "eval_loss": 0.16627496480941772, | |
| "eval_runtime": 67.2415, | |
| "eval_samples_per_second": 31.945, | |
| "eval_steps_per_second": 2.008, | |
| "eval_wer": 0.17290953545232274, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 44.23, | |
| "learning_rate": 9.970588235294116e-06, | |
| "loss": 0.709, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 44.87, | |
| "learning_rate": 8.86764705882353e-06, | |
| "loss": 0.7162, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 45.51, | |
| "learning_rate": 7.764705882352941e-06, | |
| "loss": 0.7282, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 46.15, | |
| "learning_rate": 6.6838235294117634e-06, | |
| "loss": 0.7156, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 46.15, | |
| "eval_loss": 0.16330638527870178, | |
| "eval_runtime": 67.815, | |
| "eval_samples_per_second": 31.674, | |
| "eval_steps_per_second": 1.991, | |
| "eval_wer": 0.17154034229828852, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 46.79, | |
| "learning_rate": 5.580882352941176e-06, | |
| "loss": 0.7001, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 47.44, | |
| "learning_rate": 4.477941176470588e-06, | |
| "loss": 0.7034, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 48.08, | |
| "learning_rate": 3.3749999999999995e-06, | |
| "loss": 0.6991, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 48.72, | |
| "learning_rate": 2.2720588235294113e-06, | |
| "loss": 0.7121, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 48.72, | |
| "eval_loss": 0.16656027734279633, | |
| "eval_runtime": 66.7215, | |
| "eval_samples_per_second": 32.194, | |
| "eval_steps_per_second": 2.023, | |
| "eval_wer": 0.17183374083129585, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 49.36, | |
| "learning_rate": 1.1691176470588234e-06, | |
| "loss": 0.7012, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 6.617647058823529e-08, | |
| "loss": 0.6972, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "step": 7800, | |
| "total_flos": 2.5302348778981876e+19, | |
| "train_loss": 1.4054935827010717, | |
| "train_runtime": 11681.5311, | |
| "train_samples_per_second": 21.243, | |
| "train_steps_per_second": 0.668 | |
| } | |
| ], | |
| "max_steps": 7800, | |
| "num_train_epochs": 50, | |
| "total_flos": 2.5302348778981876e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |