| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.0, | |
| "global_step": 10700, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 1.47e-05, | |
| "loss": 9.2459, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 2.97e-05, | |
| "loss": 4.0521, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 4.4699999999999996e-05, | |
| "loss": 3.4098, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.74, | |
| "learning_rate": 5.97e-05, | |
| "loss": 3.2533, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.67, | |
| "learning_rate": 7.47e-05, | |
| "loss": 3.1982, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.61, | |
| "learning_rate": 8.969999999999998e-05, | |
| "loss": 3.1301, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 6.54, | |
| "learning_rate": 0.00010469999999999998, | |
| "loss": 3.0863, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 7.48, | |
| "learning_rate": 0.0001197, | |
| "loss": 3.0034, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 8.41, | |
| "learning_rate": 0.0001347, | |
| "loss": 2.5449, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 9.35, | |
| "learning_rate": 0.00014969999999999998, | |
| "loss": 2.0192, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 10.28, | |
| "learning_rate": 0.0001647, | |
| "loss": 1.846, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 11.21, | |
| "learning_rate": 0.00017969999999999998, | |
| "loss": 1.7371, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 12.15, | |
| "learning_rate": 0.0001947, | |
| "loss": 1.6727, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 13.08, | |
| "learning_rate": 0.00020969999999999997, | |
| "loss": 1.6512, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 14.02, | |
| "learning_rate": 0.0002247, | |
| "loss": 1.6324, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 14.95, | |
| "learning_rate": 0.0002397, | |
| "loss": 1.5982, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 15.89, | |
| "learning_rate": 0.00025469999999999996, | |
| "loss": 1.564, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 16.82, | |
| "learning_rate": 0.0002697, | |
| "loss": 1.5777, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 17.76, | |
| "learning_rate": 0.0002847, | |
| "loss": 1.5636, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 18.69, | |
| "learning_rate": 0.00029969999999999997, | |
| "loss": 1.5412, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 18.69, | |
| "eval_loss": 0.6161174178123474, | |
| "eval_runtime": 82.7828, | |
| "eval_samples_per_second": 19.485, | |
| "eval_steps_per_second": 19.485, | |
| "eval_wer": 0.5747115912836477, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 19.63, | |
| "learning_rate": 0.00029662068965517237, | |
| "loss": 1.5532, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 20.56, | |
| "learning_rate": 0.0002931724137931034, | |
| "loss": 1.547, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 21.5, | |
| "learning_rate": 0.00028975862068965515, | |
| "loss": 1.5271, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 22.43, | |
| "learning_rate": 0.00028631034482758615, | |
| "loss": 1.5141, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 23.36, | |
| "learning_rate": 0.0002828620689655172, | |
| "loss": 1.492, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 24.3, | |
| "learning_rate": 0.00027941379310344826, | |
| "loss": 1.4687, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 25.23, | |
| "learning_rate": 0.00027596551724137926, | |
| "loss": 1.4519, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 26.17, | |
| "learning_rate": 0.0002725172413793103, | |
| "loss": 1.4215, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 27.1, | |
| "learning_rate": 0.0002690689655172414, | |
| "loss": 1.4192, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 28.04, | |
| "learning_rate": 0.00026562068965517243, | |
| "loss": 1.418, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 28.97, | |
| "learning_rate": 0.00026217241379310343, | |
| "loss": 1.3969, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 29.91, | |
| "learning_rate": 0.00025875862068965515, | |
| "loss": 1.3693, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 30.84, | |
| "learning_rate": 0.0002553448275862069, | |
| "loss": 1.3633, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 31.78, | |
| "learning_rate": 0.0002518965517241379, | |
| "loss": 1.3744, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 32.71, | |
| "learning_rate": 0.00024844827586206894, | |
| "loss": 1.3554, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 33.64, | |
| "learning_rate": 0.000245, | |
| "loss": 1.3262, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 34.58, | |
| "learning_rate": 0.00024155172413793102, | |
| "loss": 1.321, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 35.51, | |
| "learning_rate": 0.00023810344827586205, | |
| "loss": 1.3375, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 36.45, | |
| "learning_rate": 0.0002346551724137931, | |
| "loss": 1.3009, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 37.38, | |
| "learning_rate": 0.00023120689655172413, | |
| "loss": 1.311, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 37.38, | |
| "eval_loss": 0.5707210302352905, | |
| "eval_runtime": 86.6018, | |
| "eval_samples_per_second": 18.625, | |
| "eval_steps_per_second": 18.625, | |
| "eval_wer": 0.5070499908441677, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 38.32, | |
| "learning_rate": 0.00022775862068965516, | |
| "loss": 1.2889, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 39.25, | |
| "learning_rate": 0.00022431034482758616, | |
| "loss": 1.3253, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 40.19, | |
| "learning_rate": 0.00022086206896551722, | |
| "loss": 1.2888, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 41.12, | |
| "learning_rate": 0.00021741379310344825, | |
| "loss": 1.2567, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 42.06, | |
| "learning_rate": 0.0002139655172413793, | |
| "loss": 1.2445, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 42.99, | |
| "learning_rate": 0.00021051724137931033, | |
| "loss": 1.2608, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 43.93, | |
| "learning_rate": 0.00020706896551724136, | |
| "loss": 1.2272, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 44.86, | |
| "learning_rate": 0.0002036206896551724, | |
| "loss": 1.2186, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 45.79, | |
| "learning_rate": 0.0002001724137931034, | |
| "loss": 1.2086, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 46.73, | |
| "learning_rate": 0.00019672413793103444, | |
| "loss": 1.2042, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 47.66, | |
| "learning_rate": 0.0001932758620689655, | |
| "loss": 1.1856, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 48.6, | |
| "learning_rate": 0.00018982758620689653, | |
| "loss": 1.1965, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 49.53, | |
| "learning_rate": 0.00018637931034482758, | |
| "loss": 1.2104, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 50.47, | |
| "learning_rate": 0.0001829310344827586, | |
| "loss": 1.1631, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 51.4, | |
| "learning_rate": 0.00017948275862068964, | |
| "loss": 1.1535, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 52.34, | |
| "learning_rate": 0.0001760344827586207, | |
| "loss": 1.1615, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 53.27, | |
| "learning_rate": 0.0001725862068965517, | |
| "loss": 1.1632, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 54.21, | |
| "learning_rate": 0.00016913793103448272, | |
| "loss": 1.1431, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 55.14, | |
| "learning_rate": 0.00016568965517241378, | |
| "loss": 1.1245, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 56.07, | |
| "learning_rate": 0.0001622413793103448, | |
| "loss": 1.1367, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 56.07, | |
| "eval_loss": 0.5372028946876526, | |
| "eval_runtime": 83.9889, | |
| "eval_samples_per_second": 19.205, | |
| "eval_steps_per_second": 19.205, | |
| "eval_wer": 0.46639809558688883, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 57.01, | |
| "learning_rate": 0.00015879310344827586, | |
| "loss": 1.1242, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 57.94, | |
| "learning_rate": 0.0001553448275862069, | |
| "loss": 1.1016, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 58.88, | |
| "learning_rate": 0.00015189655172413792, | |
| "loss": 1.0946, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 59.81, | |
| "learning_rate": 0.00014844827586206895, | |
| "loss": 1.0953, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 60.75, | |
| "learning_rate": 0.000145, | |
| "loss": 1.1016, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 61.68, | |
| "learning_rate": 0.00014155172413793103, | |
| "loss": 1.0745, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 62.62, | |
| "learning_rate": 0.00013810344827586206, | |
| "loss": 1.072, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 63.55, | |
| "learning_rate": 0.00013465517241379309, | |
| "loss": 1.0621, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 64.49, | |
| "learning_rate": 0.00013120689655172414, | |
| "loss": 1.0659, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 65.42, | |
| "learning_rate": 0.00012775862068965517, | |
| "loss": 1.0271, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 66.36, | |
| "learning_rate": 0.0001243103448275862, | |
| "loss": 1.0381, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 67.29, | |
| "learning_rate": 0.00012086206896551723, | |
| "loss": 1.0332, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 68.22, | |
| "learning_rate": 0.00011741379310344827, | |
| "loss": 1.0279, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 69.16, | |
| "learning_rate": 0.00011396551724137931, | |
| "loss": 1.0174, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 70.09, | |
| "learning_rate": 0.00011051724137931032, | |
| "loss": 1.0124, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 71.03, | |
| "learning_rate": 0.00010706896551724137, | |
| "loss": 1.0142, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 71.96, | |
| "learning_rate": 0.00010362068965517241, | |
| "loss": 0.9809, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 72.9, | |
| "learning_rate": 0.00010017241379310345, | |
| "loss": 0.9609, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 73.83, | |
| "learning_rate": 9.672413793103446e-05, | |
| "loss": 0.9808, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 74.77, | |
| "learning_rate": 9.32758620689655e-05, | |
| "loss": 0.9696, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 74.77, | |
| "eval_loss": 0.5442632436752319, | |
| "eval_runtime": 84.0493, | |
| "eval_samples_per_second": 19.191, | |
| "eval_steps_per_second": 19.191, | |
| "eval_wer": 0.4327961911737777, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 75.7, | |
| "learning_rate": 8.982758620689655e-05, | |
| "loss": 0.9559, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 76.64, | |
| "learning_rate": 8.637931034482759e-05, | |
| "loss": 0.9507, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 77.57, | |
| "learning_rate": 8.29310344827586e-05, | |
| "loss": 0.9358, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 78.5, | |
| "learning_rate": 7.948275862068965e-05, | |
| "loss": 0.9367, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 79.44, | |
| "learning_rate": 7.603448275862069e-05, | |
| "loss": 0.9223, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 80.37, | |
| "learning_rate": 7.258620689655172e-05, | |
| "loss": 0.9258, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 81.31, | |
| "learning_rate": 6.913793103448276e-05, | |
| "loss": 0.8865, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 82.24, | |
| "learning_rate": 6.568965517241379e-05, | |
| "loss": 0.9062, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 83.18, | |
| "learning_rate": 6.224137931034483e-05, | |
| "loss": 0.88, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 84.11, | |
| "learning_rate": 5.8793103448275856e-05, | |
| "loss": 0.8794, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 85.05, | |
| "learning_rate": 5.534482758620689e-05, | |
| "loss": 0.8892, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 85.98, | |
| "learning_rate": 5.1896551724137926e-05, | |
| "loss": 0.8746, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 86.92, | |
| "learning_rate": 4.844827586206896e-05, | |
| "loss": 0.8465, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 87.85, | |
| "learning_rate": 4.5034482758620686e-05, | |
| "loss": 0.8453, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 88.79, | |
| "learning_rate": 4.162068965517241e-05, | |
| "loss": 0.8472, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 89.72, | |
| "learning_rate": 3.817241379310344e-05, | |
| "loss": 0.8638, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 90.65, | |
| "learning_rate": 3.4724137931034475e-05, | |
| "loss": 0.8354, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 91.59, | |
| "learning_rate": 3.127586206896551e-05, | |
| "loss": 0.825, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 92.52, | |
| "learning_rate": 2.7827586206896548e-05, | |
| "loss": 0.8395, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 93.46, | |
| "learning_rate": 2.4379310344827583e-05, | |
| "loss": 0.8163, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 93.46, | |
| "eval_loss": 0.5916153788566589, | |
| "eval_runtime": 83.5748, | |
| "eval_samples_per_second": 19.3, | |
| "eval_steps_per_second": 19.3, | |
| "eval_wer": 0.4123786852224867, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 94.39, | |
| "learning_rate": 2.093103448275862e-05, | |
| "loss": 0.8116, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 95.33, | |
| "learning_rate": 1.7482758620689653e-05, | |
| "loss": 0.8103, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 96.26, | |
| "learning_rate": 1.4034482758620688e-05, | |
| "loss": 0.8252, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 97.2, | |
| "learning_rate": 1.0586206896551723e-05, | |
| "loss": 0.805, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 98.13, | |
| "learning_rate": 7.1379310344827575e-06, | |
| "loss": 0.7983, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 99.07, | |
| "learning_rate": 3.689655172413793e-06, | |
| "loss": 0.7991, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 2.413793103448276e-07, | |
| "loss": 0.8191, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 10700, | |
| "total_flos": 4.070849231885728e+19, | |
| "train_loss": 1.4013342235244324, | |
| "train_runtime": 14378.2108, | |
| "train_samples_per_second": 23.647, | |
| "train_steps_per_second": 0.744 | |
| } | |
| ], | |
| "max_steps": 10700, | |
| "num_train_epochs": 100, | |
| "total_flos": 4.070849231885728e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |