| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 99.99521531100478, | |
| "global_step": 5200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 1.6166666666666665e-07, | |
| "loss": 24.0656, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "learning_rate": 3.283333333333333e-07, | |
| "loss": 19.0268, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 5.77, | |
| "learning_rate": 4.95e-07, | |
| "loss": 6.1995, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 7.69, | |
| "learning_rate": 4.901020408163265e-07, | |
| "loss": 3.6484, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 9.61, | |
| "learning_rate": 4.798979591836734e-07, | |
| "loss": 3.5131, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 9.61, | |
| "eval_loss": 3.5294487476348877, | |
| "eval_runtime": 64.5425, | |
| "eval_samples_per_second": 17.229, | |
| "eval_steps_per_second": 4.307, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 11.54, | |
| "learning_rate": 4.696938775510204e-07, | |
| "loss": 3.4699, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 13.46, | |
| "learning_rate": 4.5948979591836735e-07, | |
| "loss": 3.3973, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 15.38, | |
| "learning_rate": 4.4928571428571426e-07, | |
| "loss": 3.2876, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 17.31, | |
| "learning_rate": 4.390816326530612e-07, | |
| "loss": 3.0902, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 19.23, | |
| "learning_rate": 4.288775510204081e-07, | |
| "loss": 2.8596, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 19.23, | |
| "eval_loss": 3.5708465576171875, | |
| "eval_runtime": 66.2433, | |
| "eval_samples_per_second": 16.787, | |
| "eval_steps_per_second": 4.197, | |
| "eval_wer": 1.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 21.15, | |
| "learning_rate": 4.1867346938775513e-07, | |
| "loss": 2.6397, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 23.08, | |
| "learning_rate": 4.0846938775510203e-07, | |
| "loss": 2.4302, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 3.98265306122449e-07, | |
| "loss": 2.2341, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 26.92, | |
| "learning_rate": 3.880612244897959e-07, | |
| "loss": 2.071, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 28.84, | |
| "learning_rate": 3.778571428571428e-07, | |
| "loss": 1.9055, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 28.84, | |
| "eval_loss": 3.6432793140411377, | |
| "eval_runtime": 65.1639, | |
| "eval_samples_per_second": 17.065, | |
| "eval_steps_per_second": 4.266, | |
| "eval_wer": 1.0007174887892376, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 30.77, | |
| "learning_rate": 3.676530612244898e-07, | |
| "loss": 1.7787, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 32.69, | |
| "learning_rate": 3.574489795918367e-07, | |
| "loss": 1.6655, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 34.61, | |
| "learning_rate": 3.4724489795918366e-07, | |
| "loss": 1.5696, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 36.54, | |
| "learning_rate": 3.3704081632653057e-07, | |
| "loss": 1.4902, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "learning_rate": 3.268367346938775e-07, | |
| "loss": 1.4239, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 38.46, | |
| "eval_loss": 3.6568963527679443, | |
| "eval_runtime": 66.0028, | |
| "eval_samples_per_second": 16.848, | |
| "eval_steps_per_second": 4.212, | |
| "eval_wer": 0.9994618834080717, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 40.38, | |
| "learning_rate": 3.166326530612245e-07, | |
| "loss": 1.3735, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 42.31, | |
| "learning_rate": 3.0642857142857144e-07, | |
| "loss": 1.3228, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 44.23, | |
| "learning_rate": 2.9622448979591834e-07, | |
| "loss": 1.2834, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 46.15, | |
| "learning_rate": 2.860204081632653e-07, | |
| "loss": 1.2438, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 48.08, | |
| "learning_rate": 2.758163265306122e-07, | |
| "loss": 1.2168, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 48.08, | |
| "eval_loss": 3.6079351902008057, | |
| "eval_runtime": 65.3166, | |
| "eval_samples_per_second": 17.025, | |
| "eval_steps_per_second": 4.256, | |
| "eval_wer": 0.995695067264574, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 2.656122448979592e-07, | |
| "loss": 1.1792, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 51.92, | |
| "learning_rate": 2.554081632653061e-07, | |
| "loss": 1.1706, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 53.84, | |
| "learning_rate": 2.4520408163265307e-07, | |
| "loss": 1.1429, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 55.77, | |
| "learning_rate": 2.3499999999999997e-07, | |
| "loss": 1.1318, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 57.69, | |
| "learning_rate": 2.2479591836734693e-07, | |
| "loss": 1.1063, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 57.69, | |
| "eval_loss": 3.5737504959106445, | |
| "eval_runtime": 64.9336, | |
| "eval_samples_per_second": 17.125, | |
| "eval_steps_per_second": 4.281, | |
| "eval_wer": 0.9924663677130044, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 59.61, | |
| "learning_rate": 2.1459183673469386e-07, | |
| "loss": 1.0893, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 61.54, | |
| "learning_rate": 2.0438775510204082e-07, | |
| "loss": 1.0744, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 63.46, | |
| "learning_rate": 1.9418367346938775e-07, | |
| "loss": 1.0591, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 65.38, | |
| "learning_rate": 1.839795918367347e-07, | |
| "loss": 1.05, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 67.31, | |
| "learning_rate": 1.7377551020408163e-07, | |
| "loss": 1.0404, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 67.31, | |
| "eval_loss": 3.4857470989227295, | |
| "eval_runtime": 64.9602, | |
| "eval_samples_per_second": 17.118, | |
| "eval_steps_per_second": 4.28, | |
| "eval_wer": 0.9888789237668162, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 69.23, | |
| "learning_rate": 1.6357142857142856e-07, | |
| "loss": 1.0325, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 71.15, | |
| "learning_rate": 1.5336734693877552e-07, | |
| "loss": 1.0266, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 73.08, | |
| "learning_rate": 1.4316326530612245e-07, | |
| "loss": 1.0164, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "learning_rate": 1.329591836734694e-07, | |
| "loss": 1.0002, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 76.92, | |
| "learning_rate": 1.227551020408163e-07, | |
| "loss": 1.001, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 76.92, | |
| "eval_loss": 3.4881510734558105, | |
| "eval_runtime": 64.7811, | |
| "eval_samples_per_second": 17.165, | |
| "eval_steps_per_second": 4.291, | |
| "eval_wer": 0.9858295964125561, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 78.84, | |
| "learning_rate": 1.1255102040816327e-07, | |
| "loss": 0.997, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 80.77, | |
| "learning_rate": 1.0234693877551021e-07, | |
| "loss": 0.9896, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 82.69, | |
| "learning_rate": 9.214285714285714e-08, | |
| "loss": 0.986, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 84.61, | |
| "learning_rate": 8.193877551020407e-08, | |
| "loss": 0.9823, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 86.54, | |
| "learning_rate": 7.173469387755101e-08, | |
| "loss": 0.982, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 86.54, | |
| "eval_loss": 3.3850555419921875, | |
| "eval_runtime": 65.2436, | |
| "eval_samples_per_second": 17.044, | |
| "eval_steps_per_second": 4.261, | |
| "eval_wer": 0.987085201793722, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 88.46, | |
| "learning_rate": 6.153061224489796e-08, | |
| "loss": 0.9675, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 90.38, | |
| "learning_rate": 5.132653061224489e-08, | |
| "loss": 0.9774, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 92.31, | |
| "learning_rate": 4.1122448979591836e-08, | |
| "loss": 0.9687, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 94.23, | |
| "learning_rate": 3.091836734693877e-08, | |
| "loss": 0.9771, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 96.15, | |
| "learning_rate": 2.0714285714285713e-08, | |
| "loss": 0.9612, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 96.15, | |
| "eval_loss": 3.386908531188965, | |
| "eval_runtime": 65.5128, | |
| "eval_samples_per_second": 16.974, | |
| "eval_steps_per_second": 4.243, | |
| "eval_wer": 0.9872645739910314, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 98.08, | |
| "learning_rate": 1.0510204081632651e-08, | |
| "loss": 0.9667, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 3.061224489795918e-10, | |
| "loss": 0.9664, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 5200, | |
| "total_flos": 1.4653197993726655e+20, | |
| "train_loss": 2.424026096050556, | |
| "train_runtime": 30676.4279, | |
| "train_samples_per_second": 10.875, | |
| "train_steps_per_second": 0.17 | |
| } | |
| ], | |
| "max_steps": 5200, | |
| "num_train_epochs": 100, | |
| "total_flos": 1.4653197993726655e+20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |