| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 15.0, | |
| "eval_steps": 300, | |
| "global_step": 16425, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.27, | |
| "eval_loss": 3.293024778366089, | |
| "eval_runtime": 329.2436, | |
| "eval_samples_per_second": 33.516, | |
| "eval_steps_per_second": 1.048, | |
| "eval_wer": 1.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 0.0007389237223065434, | |
| "loss": 5.6462, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "eval_loss": 3.415907859802246, | |
| "eval_runtime": 328.5335, | |
| "eval_samples_per_second": 33.589, | |
| "eval_steps_per_second": 1.05, | |
| "eval_wer": 1.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "eval_loss": 3.4422173500061035, | |
| "eval_runtime": 338.1217, | |
| "eval_samples_per_second": 32.636, | |
| "eval_steps_per_second": 1.02, | |
| "eval_wer": 1.0, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 0.0014838065068897525, | |
| "loss": 3.3522, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "eval_loss": 3.3719358444213867, | |
| "eval_runtime": 340.7912, | |
| "eval_samples_per_second": 32.381, | |
| "eval_steps_per_second": 1.012, | |
| "eval_wer": 1.0, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 0.002228689291472962, | |
| "loss": 3.2605, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "eval_loss": 3.4025769233703613, | |
| "eval_runtime": 335.1987, | |
| "eval_samples_per_second": 32.921, | |
| "eval_steps_per_second": 1.029, | |
| "eval_wer": 1.0, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "eval_loss": 3.444835662841797, | |
| "eval_runtime": 332.4199, | |
| "eval_samples_per_second": 33.196, | |
| "eval_steps_per_second": 1.038, | |
| "eval_wer": 1.0, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 0.0029735720760561708, | |
| "loss": 3.2766, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "eval_loss": 3.473637104034424, | |
| "eval_runtime": 334.5697, | |
| "eval_samples_per_second": 32.983, | |
| "eval_steps_per_second": 1.031, | |
| "eval_wer": 0.9999093792478477, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "eval_loss": 3.982806444168091, | |
| "eval_runtime": 342.0917, | |
| "eval_samples_per_second": 32.257, | |
| "eval_steps_per_second": 1.009, | |
| "eval_wer": 1.0, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 0.0037184548606393796, | |
| "loss": 3.2853, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "eval_loss": 3.553187370300293, | |
| "eval_runtime": 329.7168, | |
| "eval_samples_per_second": 33.468, | |
| "eval_steps_per_second": 1.046, | |
| "eval_wer": 1.0, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 0.004460358114084256, | |
| "loss": 3.3389, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "eval_loss": 3.781858444213867, | |
| "eval_runtime": 323.3731, | |
| "eval_samples_per_second": 34.125, | |
| "eval_steps_per_second": 1.067, | |
| "eval_wer": 1.0, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 3.01, | |
| "eval_loss": 3.2249505519866943, | |
| "eval_runtime": 323.988, | |
| "eval_samples_per_second": 34.06, | |
| "eval_steps_per_second": 1.065, | |
| "eval_wer": 1.0, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "learning_rate": 0.005205240898667465, | |
| "loss": 3.2186, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 3.29, | |
| "eval_loss": 3.2372846603393555, | |
| "eval_runtime": 324.034, | |
| "eval_samples_per_second": 34.055, | |
| "eval_steps_per_second": 1.065, | |
| "eval_wer": 1.0, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "eval_loss": 3.2161905765533447, | |
| "eval_runtime": 323.7754, | |
| "eval_samples_per_second": 34.082, | |
| "eval_steps_per_second": 1.066, | |
| "eval_wer": 1.0, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "learning_rate": 0.005950123683250674, | |
| "loss": 3.1916, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "eval_loss": 3.2367777824401855, | |
| "eval_runtime": 323.3817, | |
| "eval_samples_per_second": 34.124, | |
| "eval_steps_per_second": 1.067, | |
| "eval_wer": 1.0, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "learning_rate": 0.0066950064678338835, | |
| "loss": 3.2188, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 4.11, | |
| "eval_loss": 3.2376551628112793, | |
| "eval_runtime": 323.9231, | |
| "eval_samples_per_second": 34.067, | |
| "eval_steps_per_second": 1.065, | |
| "eval_wer": 1.0, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "eval_loss": 3.4206786155700684, | |
| "eval_runtime": 324.1643, | |
| "eval_samples_per_second": 34.041, | |
| "eval_steps_per_second": 1.064, | |
| "eval_wer": 1.0, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 4.57, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 5.3067, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 4.66, | |
| "eval_loss": NaN, | |
| "eval_runtime": 324.7836, | |
| "eval_samples_per_second": 33.976, | |
| "eval_steps_per_second": 1.062, | |
| "eval_wer": 1.0, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 4.93, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.844, | |
| "eval_samples_per_second": 34.075, | |
| "eval_steps_per_second": 1.065, | |
| "eval_wer": 1.0, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 5.02, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.0895, | |
| "eval_samples_per_second": 34.155, | |
| "eval_steps_per_second": 1.068, | |
| "eval_wer": 1.0, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 5.48, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 5.48, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.9944, | |
| "eval_samples_per_second": 34.059, | |
| "eval_steps_per_second": 1.065, | |
| "eval_wer": 1.0, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 5.75, | |
| "eval_loss": NaN, | |
| "eval_runtime": 324.3089, | |
| "eval_samples_per_second": 34.026, | |
| "eval_steps_per_second": 1.064, | |
| "eval_wer": 1.0, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 6.03, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.6949, | |
| "eval_samples_per_second": 34.091, | |
| "eval_steps_per_second": 1.066, | |
| "eval_wer": 1.0, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 6.3, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.4351, | |
| "eval_samples_per_second": 34.118, | |
| "eval_steps_per_second": 1.067, | |
| "eval_wer": 1.0, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 6.39, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 6.58, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.8786, | |
| "eval_samples_per_second": 34.071, | |
| "eval_steps_per_second": 1.065, | |
| "eval_wer": 1.0, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 6.85, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 6.85, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.8341, | |
| "eval_samples_per_second": 34.076, | |
| "eval_steps_per_second": 1.065, | |
| "eval_wer": 1.0, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 7.12, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.3556, | |
| "eval_samples_per_second": 34.127, | |
| "eval_steps_per_second": 1.067, | |
| "eval_wer": 1.0, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 7.31, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.4333, | |
| "eval_samples_per_second": 34.118, | |
| "eval_steps_per_second": 1.067, | |
| "eval_wer": 1.0, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 7.67, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.7546, | |
| "eval_samples_per_second": 34.084, | |
| "eval_steps_per_second": 1.066, | |
| "eval_wer": 1.0, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 7.76, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 7.95, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.2121, | |
| "eval_samples_per_second": 34.142, | |
| "eval_steps_per_second": 1.067, | |
| "eval_wer": 1.0, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 8.22, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 8.22, | |
| "eval_loss": NaN, | |
| "eval_runtime": 322.4572, | |
| "eval_samples_per_second": 34.222, | |
| "eval_steps_per_second": 1.07, | |
| "eval_wer": 1.0, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 8.49, | |
| "eval_loss": NaN, | |
| "eval_runtime": 318.5677, | |
| "eval_samples_per_second": 34.639, | |
| "eval_steps_per_second": 1.083, | |
| "eval_wer": 1.0, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 8.68, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 8.77, | |
| "eval_loss": NaN, | |
| "eval_runtime": 319.654, | |
| "eval_samples_per_second": 34.522, | |
| "eval_steps_per_second": 1.079, | |
| "eval_wer": 1.0, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 9.04, | |
| "eval_loss": NaN, | |
| "eval_runtime": 320.9193, | |
| "eval_samples_per_second": 34.386, | |
| "eval_steps_per_second": 1.075, | |
| "eval_wer": 1.0, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 9.13, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 9.32, | |
| "eval_loss": NaN, | |
| "eval_runtime": 322.8537, | |
| "eval_samples_per_second": 34.18, | |
| "eval_steps_per_second": 1.069, | |
| "eval_wer": 1.0, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 9.59, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 9.59, | |
| "eval_loss": NaN, | |
| "eval_runtime": 321.9757, | |
| "eval_samples_per_second": 34.273, | |
| "eval_steps_per_second": 1.072, | |
| "eval_wer": 1.0, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 9.86, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.2439, | |
| "eval_samples_per_second": 34.138, | |
| "eval_steps_per_second": 1.067, | |
| "eval_wer": 1.0, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 10.05, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 10.14, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.9537, | |
| "eval_samples_per_second": 34.064, | |
| "eval_steps_per_second": 1.065, | |
| "eval_wer": 1.0, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 10.41, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.8438, | |
| "eval_samples_per_second": 34.075, | |
| "eval_steps_per_second": 1.065, | |
| "eval_wer": 1.0, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 10.5, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 10.68, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.9188, | |
| "eval_samples_per_second": 34.067, | |
| "eval_steps_per_second": 1.065, | |
| "eval_wer": 1.0, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 10.96, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 10.96, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.8826, | |
| "eval_samples_per_second": 34.071, | |
| "eval_steps_per_second": 1.065, | |
| "eval_wer": 1.0, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 11.23, | |
| "eval_loss": NaN, | |
| "eval_runtime": 325.2969, | |
| "eval_samples_per_second": 33.923, | |
| "eval_steps_per_second": 1.061, | |
| "eval_wer": 1.0, | |
| "step": 12300 | |
| }, | |
| { | |
| "epoch": 11.42, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 11.51, | |
| "eval_loss": NaN, | |
| "eval_runtime": 324.7333, | |
| "eval_samples_per_second": 33.982, | |
| "eval_steps_per_second": 1.062, | |
| "eval_wer": 1.0, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 11.78, | |
| "eval_loss": NaN, | |
| "eval_runtime": 324.3204, | |
| "eval_samples_per_second": 34.025, | |
| "eval_steps_per_second": 1.064, | |
| "eval_wer": 1.0, | |
| "step": 12900 | |
| }, | |
| { | |
| "epoch": 11.87, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 12.05, | |
| "eval_loss": NaN, | |
| "eval_runtime": 324.8823, | |
| "eval_samples_per_second": 33.966, | |
| "eval_steps_per_second": 1.062, | |
| "eval_wer": 1.0, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 12.33, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 12.33, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.9396, | |
| "eval_samples_per_second": 34.065, | |
| "eval_steps_per_second": 1.065, | |
| "eval_wer": 1.0, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 12.6, | |
| "eval_loss": NaN, | |
| "eval_runtime": 326.2533, | |
| "eval_samples_per_second": 33.823, | |
| "eval_steps_per_second": 1.057, | |
| "eval_wer": 1.0, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 12.79, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 12.88, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.7094, | |
| "eval_samples_per_second": 34.089, | |
| "eval_steps_per_second": 1.066, | |
| "eval_wer": 1.0, | |
| "step": 14100 | |
| }, | |
| { | |
| "epoch": 13.15, | |
| "eval_loss": NaN, | |
| "eval_runtime": 325.3499, | |
| "eval_samples_per_second": 33.917, | |
| "eval_steps_per_second": 1.06, | |
| "eval_wer": 1.0, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 13.24, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 13.42, | |
| "eval_loss": NaN, | |
| "eval_runtime": 325.7039, | |
| "eval_samples_per_second": 33.88, | |
| "eval_steps_per_second": 1.059, | |
| "eval_wer": 1.0, | |
| "step": 14700 | |
| }, | |
| { | |
| "epoch": 13.7, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 13.7, | |
| "eval_loss": NaN, | |
| "eval_runtime": 323.516, | |
| "eval_samples_per_second": 34.11, | |
| "eval_steps_per_second": 1.066, | |
| "eval_wer": 1.0, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 13.97, | |
| "eval_loss": NaN, | |
| "eval_runtime": 471.9655, | |
| "eval_samples_per_second": 23.381, | |
| "eval_steps_per_second": 0.731, | |
| "eval_wer": 1.0, | |
| "step": 15300 | |
| }, | |
| { | |
| "epoch": 14.16, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 14.25, | |
| "eval_loss": NaN, | |
| "eval_runtime": 324.2368, | |
| "eval_samples_per_second": 34.034, | |
| "eval_steps_per_second": 1.064, | |
| "eval_wer": 1.0, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 14.52, | |
| "eval_loss": NaN, | |
| "eval_runtime": 325.6971, | |
| "eval_samples_per_second": 33.881, | |
| "eval_steps_per_second": 1.059, | |
| "eval_wer": 1.0, | |
| "step": 15900 | |
| }, | |
| { | |
| "epoch": 14.61, | |
| "learning_rate": 0.007152364497567974, | |
| "loss": 0.0, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 14.79, | |
| "eval_loss": NaN, | |
| "eval_runtime": 325.4219, | |
| "eval_samples_per_second": 33.91, | |
| "eval_steps_per_second": 1.06, | |
| "eval_wer": 1.0, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "step": 16425, | |
| "total_flos": 6.442470243808035e+19, | |
| "train_loss": 1.129231213434646, | |
| "train_runtime": 45647.7738, | |
| "train_samples_per_second": 14.392, | |
| "train_steps_per_second": 0.36 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 16425, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 15, | |
| "save_steps": 400, | |
| "total_flos": 6.442470243808035e+19, | |
| "train_batch_size": 20, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |