| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 50.0, | |
| "global_step": 10600, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.4299999999999998e-06, | |
| "loss": 11.8059, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 6.93e-06, | |
| "loss": 6.741, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 1.0429999999999998e-05, | |
| "loss": 4.6303, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 1.3929999999999999e-05, | |
| "loss": 4.0577, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 1.7429999999999997e-05, | |
| "loss": 3.6055, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 2.0929999999999998e-05, | |
| "loss": 3.2771, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.3, | |
| "learning_rate": 2.4429999999999995e-05, | |
| "loss": 3.0775, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.77, | |
| "learning_rate": 2.793e-05, | |
| "loss": 2.9768, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.25, | |
| "learning_rate": 3.1429999999999996e-05, | |
| "loss": 2.8612, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "learning_rate": 3.493e-05, | |
| "loss": 2.8112, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "eval_loss": 2.778607130050659, | |
| "eval_runtime": 101.5624, | |
| "eval_samples_per_second": 29.913, | |
| "eval_steps_per_second": 0.935, | |
| "eval_wer": 1.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 5.19, | |
| "learning_rate": 3.843e-05, | |
| "loss": 2.7937, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 5.66, | |
| "learning_rate": 4.192999999999999e-05, | |
| "loss": 2.7598, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 6.13, | |
| "learning_rate": 4.543e-05, | |
| "loss": 2.5782, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 6.6, | |
| "learning_rate": 4.8929999999999994e-05, | |
| "loss": 2.0779, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 7.08, | |
| "learning_rate": 5.243e-05, | |
| "loss": 1.7522, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 7.55, | |
| "learning_rate": 5.593e-05, | |
| "loss": 1.6003, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "learning_rate": 5.942999999999999e-05, | |
| "loss": 1.534, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 8.49, | |
| "learning_rate": 6.293e-05, | |
| "loss": 1.4587, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 8.96, | |
| "learning_rate": 6.642999999999999e-05, | |
| "loss": 1.4313, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 9.43, | |
| "learning_rate": 6.992999999999999e-05, | |
| "loss": 1.3803, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 9.43, | |
| "eval_loss": 0.49710744619369507, | |
| "eval_runtime": 101.1307, | |
| "eval_samples_per_second": 30.04, | |
| "eval_steps_per_second": 0.939, | |
| "eval_wer": 0.5787768779670484, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 9.91, | |
| "learning_rate": 6.920232558139535e-05, | |
| "loss": 1.3528, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 10.38, | |
| "learning_rate": 6.838837209302325e-05, | |
| "loss": 1.3382, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 10.85, | |
| "learning_rate": 6.757441860465116e-05, | |
| "loss": 1.3103, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 11.32, | |
| "learning_rate": 6.676046511627906e-05, | |
| "loss": 1.2818, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 11.79, | |
| "learning_rate": 6.594651162790697e-05, | |
| "loss": 1.265, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 12.26, | |
| "learning_rate": 6.513255813953489e-05, | |
| "loss": 1.2608, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 12.74, | |
| "learning_rate": 6.431860465116279e-05, | |
| "loss": 1.2263, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 13.21, | |
| "learning_rate": 6.350465116279069e-05, | |
| "loss": 1.2069, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 13.68, | |
| "learning_rate": 6.26906976744186e-05, | |
| "loss": 1.2102, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 14.15, | |
| "learning_rate": 6.18767441860465e-05, | |
| "loss": 1.1894, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 14.15, | |
| "eval_loss": 0.36050862073898315, | |
| "eval_runtime": 100.2529, | |
| "eval_samples_per_second": 30.303, | |
| "eval_steps_per_second": 0.948, | |
| "eval_wer": 0.44417760402122314, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 14.62, | |
| "learning_rate": 6.106279069767441e-05, | |
| "loss": 1.1845, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 15.09, | |
| "learning_rate": 6.024883720930232e-05, | |
| "loss": 1.1819, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 15.57, | |
| "learning_rate": 5.943488372093023e-05, | |
| "loss": 1.1565, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 16.04, | |
| "learning_rate": 5.8620930232558136e-05, | |
| "loss": 1.1506, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 16.51, | |
| "learning_rate": 5.7806976744186036e-05, | |
| "loss": 1.1254, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 16.98, | |
| "learning_rate": 5.699302325581395e-05, | |
| "loss": 1.1444, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 17.45, | |
| "learning_rate": 5.6179069767441856e-05, | |
| "loss": 1.1189, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 17.92, | |
| "learning_rate": 5.536511627906976e-05, | |
| "loss": 1.1219, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 18.4, | |
| "learning_rate": 5.455116279069767e-05, | |
| "loss": 1.1075, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 18.87, | |
| "learning_rate": 5.373720930232558e-05, | |
| "loss": 1.1019, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 18.87, | |
| "eval_loss": 0.3251776695251465, | |
| "eval_runtime": 100.5327, | |
| "eval_samples_per_second": 30.219, | |
| "eval_steps_per_second": 0.945, | |
| "eval_wer": 0.42356883552080427, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 19.34, | |
| "learning_rate": 5.292325581395348e-05, | |
| "loss": 1.1104, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 19.81, | |
| "learning_rate": 5.212558139534883e-05, | |
| "loss": 1.0985, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 20.28, | |
| "learning_rate": 5.1311627906976736e-05, | |
| "loss": 1.0878, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 20.75, | |
| "learning_rate": 5.049767441860465e-05, | |
| "loss": 1.0766, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 21.23, | |
| "learning_rate": 4.968372093023256e-05, | |
| "loss": 1.071, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 21.7, | |
| "learning_rate": 4.8869767441860464e-05, | |
| "loss": 1.0508, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 22.17, | |
| "learning_rate": 4.8055813953488364e-05, | |
| "loss": 1.0644, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 22.64, | |
| "learning_rate": 4.7249999999999997e-05, | |
| "loss": 1.0571, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 23.11, | |
| "learning_rate": 4.6436046511627903e-05, | |
| "loss": 1.0457, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 23.58, | |
| "learning_rate": 4.562209302325581e-05, | |
| "loss": 1.0366, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 23.58, | |
| "eval_loss": 0.31269633769989014, | |
| "eval_runtime": 101.0057, | |
| "eval_samples_per_second": 30.078, | |
| "eval_steps_per_second": 0.941, | |
| "eval_wer": 0.40234571348785253, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 24.06, | |
| "learning_rate": 4.480813953488372e-05, | |
| "loss": 1.0395, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 24.53, | |
| "learning_rate": 4.3994186046511624e-05, | |
| "loss": 1.0344, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 4.318023255813954e-05, | |
| "loss": 1.0318, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 25.47, | |
| "learning_rate": 4.236627906976744e-05, | |
| "loss": 1.0291, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 25.94, | |
| "learning_rate": 4.1552325581395345e-05, | |
| "loss": 1.0325, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 26.42, | |
| "learning_rate": 4.073837209302325e-05, | |
| "loss": 1.0205, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 26.89, | |
| "learning_rate": 3.992441860465116e-05, | |
| "loss": 1.0276, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 27.36, | |
| "learning_rate": 3.9110465116279065e-05, | |
| "loss": 1.0122, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 27.83, | |
| "learning_rate": 3.829651162790698e-05, | |
| "loss": 1.0054, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 28.3, | |
| "learning_rate": 3.748255813953488e-05, | |
| "loss": 1.0217, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 28.3, | |
| "eval_loss": 0.3026880621910095, | |
| "eval_runtime": 100.3358, | |
| "eval_samples_per_second": 30.278, | |
| "eval_steps_per_second": 0.947, | |
| "eval_wer": 0.39525272270315553, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 28.77, | |
| "learning_rate": 3.6668604651162786e-05, | |
| "loss": 1.0134, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 29.25, | |
| "learning_rate": 3.585465116279069e-05, | |
| "loss": 0.9899, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 29.72, | |
| "learning_rate": 3.50406976744186e-05, | |
| "loss": 0.9984, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 30.19, | |
| "learning_rate": 3.422674418604651e-05, | |
| "loss": 0.9952, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 30.66, | |
| "learning_rate": 3.341279069767442e-05, | |
| "loss": 0.9957, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 31.13, | |
| "learning_rate": 3.259883720930232e-05, | |
| "loss": 0.9922, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 31.6, | |
| "learning_rate": 3.178488372093023e-05, | |
| "loss": 0.9707, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 32.08, | |
| "learning_rate": 3.097093023255814e-05, | |
| "loss": 0.9876, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 32.55, | |
| "learning_rate": 3.0156976744186045e-05, | |
| "loss": 0.9701, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "learning_rate": 2.9343023255813948e-05, | |
| "loss": 0.9774, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 33.02, | |
| "eval_loss": 0.2924236059188843, | |
| "eval_runtime": 101.009, | |
| "eval_samples_per_second": 30.077, | |
| "eval_steps_per_second": 0.941, | |
| "eval_wer": 0.3965372800893605, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 33.49, | |
| "learning_rate": 2.852906976744186e-05, | |
| "loss": 0.9745, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 33.96, | |
| "learning_rate": 2.7715116279069765e-05, | |
| "loss": 0.9682, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 34.43, | |
| "learning_rate": 2.6901162790697676e-05, | |
| "loss": 0.9593, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 34.91, | |
| "learning_rate": 2.608720930232558e-05, | |
| "loss": 0.9591, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 35.38, | |
| "learning_rate": 2.5273255813953486e-05, | |
| "loss": 0.9582, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 35.85, | |
| "learning_rate": 2.4459302325581396e-05, | |
| "loss": 0.9567, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 36.32, | |
| "learning_rate": 2.36453488372093e-05, | |
| "loss": 0.9476, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 36.79, | |
| "learning_rate": 2.2831395348837207e-05, | |
| "loss": 0.9534, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 37.26, | |
| "learning_rate": 2.2017441860465117e-05, | |
| "loss": 0.9473, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 37.74, | |
| "learning_rate": 2.120348837209302e-05, | |
| "loss": 0.9485, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 37.74, | |
| "eval_loss": 0.2948923110961914, | |
| "eval_runtime": 99.2137, | |
| "eval_samples_per_second": 30.621, | |
| "eval_steps_per_second": 0.958, | |
| "eval_wer": 0.3755375593409662, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 38.21, | |
| "learning_rate": 2.0389534883720927e-05, | |
| "loss": 0.9344, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 38.68, | |
| "learning_rate": 1.9575581395348838e-05, | |
| "loss": 0.9357, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 39.15, | |
| "learning_rate": 1.876162790697674e-05, | |
| "loss": 0.9545, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 39.62, | |
| "learning_rate": 1.794767441860465e-05, | |
| "loss": 0.9289, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 40.09, | |
| "learning_rate": 1.7133720930232558e-05, | |
| "loss": 0.9434, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 40.57, | |
| "learning_rate": 1.6319767441860465e-05, | |
| "loss": 0.9319, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 41.04, | |
| "learning_rate": 1.551395348837209e-05, | |
| "loss": 0.9357, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 41.51, | |
| "learning_rate": 1.4699999999999998e-05, | |
| "loss": 0.9105, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 41.98, | |
| "learning_rate": 1.3886046511627905e-05, | |
| "loss": 0.9299, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 42.45, | |
| "learning_rate": 1.3072093023255814e-05, | |
| "loss": 0.9332, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 42.45, | |
| "eval_loss": 0.28152021765708923, | |
| "eval_runtime": 100.0017, | |
| "eval_samples_per_second": 30.379, | |
| "eval_steps_per_second": 0.95, | |
| "eval_wer": 0.3664898073163921, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 42.92, | |
| "learning_rate": 1.225813953488372e-05, | |
| "loss": 0.922, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 43.4, | |
| "learning_rate": 1.1444186046511626e-05, | |
| "loss": 0.9161, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 43.87, | |
| "learning_rate": 1.0630232558139534e-05, | |
| "loss": 0.9203, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 44.34, | |
| "learning_rate": 9.816279069767441e-06, | |
| "loss": 0.9221, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 44.81, | |
| "learning_rate": 9.002325581395346e-06, | |
| "loss": 0.8945, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 45.28, | |
| "learning_rate": 8.188372093023255e-06, | |
| "loss": 0.9347, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 45.75, | |
| "learning_rate": 7.374418604651162e-06, | |
| "loss": 0.8967, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 46.23, | |
| "learning_rate": 6.560465116279069e-06, | |
| "loss": 0.9211, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 46.7, | |
| "learning_rate": 5.7465116279069765e-06, | |
| "loss": 0.9007, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 47.17, | |
| "learning_rate": 4.9325581395348825e-06, | |
| "loss": 0.9093, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 47.17, | |
| "eval_loss": 0.2779529094696045, | |
| "eval_runtime": 102.291, | |
| "eval_samples_per_second": 29.7, | |
| "eval_steps_per_second": 0.929, | |
| "eval_wer": 0.3623010332309411, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 47.64, | |
| "learning_rate": 4.11860465116279e-06, | |
| "loss": 0.9066, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 48.11, | |
| "learning_rate": 3.3046511627906976e-06, | |
| "loss": 0.9154, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 48.58, | |
| "learning_rate": 2.4906976744186045e-06, | |
| "loss": 0.9112, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 49.06, | |
| "learning_rate": 1.6767441860465114e-06, | |
| "loss": 0.8968, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 49.53, | |
| "learning_rate": 8.627906976744186e-07, | |
| "loss": 0.9083, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 5.697674418604651e-08, | |
| "loss": 0.9076, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "step": 10600, | |
| "total_flos": 3.899540035775118e+19, | |
| "train_loss": 1.4489057677646853, | |
| "train_runtime": 15709.6404, | |
| "train_samples_per_second": 21.544, | |
| "train_steps_per_second": 0.675 | |
| } | |
| ], | |
| "max_steps": 10600, | |
| "num_train_epochs": 50, | |
| "total_flos": 3.899540035775118e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |