| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.0, | |
| "global_step": 11100, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 3.772222222222222e-06, | |
| "loss": 14.8289, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 7.66111111111111e-06, | |
| "loss": 7.3013, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.7, | |
| "learning_rate": 1.155e-05, | |
| "loss": 4.522, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "learning_rate": 1.5438888888888887e-05, | |
| "loss": 3.9098, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "learning_rate": 1.9327777777777774e-05, | |
| "loss": 3.511, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.41, | |
| "learning_rate": 2.3216666666666664e-05, | |
| "loss": 3.2448, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 6.31, | |
| "learning_rate": 2.7105555555555554e-05, | |
| "loss": 3.1281, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 7.21, | |
| "learning_rate": 3.099444444444444e-05, | |
| "loss": 3.0628, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 8.11, | |
| "learning_rate": 3.488333333333333e-05, | |
| "loss": 3.038, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "learning_rate": 3.877222222222222e-05, | |
| "loss": 3.0174, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 9.01, | |
| "eval_loss": 3.055245876312256, | |
| "eval_runtime": 67.6307, | |
| "eval_samples_per_second": 24.028, | |
| "eval_steps_per_second": 1.508, | |
| "eval_wer": 1.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 9.91, | |
| "learning_rate": 4.2661111111111106e-05, | |
| "loss": 2.9754, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 10.81, | |
| "learning_rate": 4.655e-05, | |
| "loss": 2.6398, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 11.71, | |
| "learning_rate": 5.043888888888888e-05, | |
| "loss": 2.0198, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 12.61, | |
| "learning_rate": 5.4327777777777774e-05, | |
| "loss": 1.6983, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 13.51, | |
| "learning_rate": 5.821666666666666e-05, | |
| "loss": 1.5587, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 14.41, | |
| "learning_rate": 6.210555555555555e-05, | |
| "loss": 1.4169, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 15.32, | |
| "learning_rate": 6.599444444444444e-05, | |
| "loss": 1.2776, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 16.22, | |
| "learning_rate": 6.988333333333333e-05, | |
| "loss": 1.161, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 17.12, | |
| "learning_rate": 6.926989247311828e-05, | |
| "loss": 1.0941, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "learning_rate": 6.851720430107527e-05, | |
| "loss": 1.0446, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 18.02, | |
| "eval_loss": 0.6708454489707947, | |
| "eval_runtime": 67.0128, | |
| "eval_samples_per_second": 24.249, | |
| "eval_steps_per_second": 1.522, | |
| "eval_wer": 0.7576906505816459, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 18.92, | |
| "learning_rate": 6.776451612903225e-05, | |
| "loss": 0.9991, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 19.82, | |
| "learning_rate": 6.701182795698924e-05, | |
| "loss": 0.9554, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 20.72, | |
| "learning_rate": 6.625913978494624e-05, | |
| "loss": 0.9412, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 21.62, | |
| "learning_rate": 6.550645161290322e-05, | |
| "loss": 0.8972, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 22.52, | |
| "learning_rate": 6.475376344086021e-05, | |
| "loss": 0.8861, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 23.42, | |
| "learning_rate": 6.400107526881719e-05, | |
| "loss": 0.8724, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 24.32, | |
| "learning_rate": 6.324838709677419e-05, | |
| "loss": 0.8638, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 25.23, | |
| "learning_rate": 6.249569892473118e-05, | |
| "loss": 0.8306, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 26.13, | |
| "learning_rate": 6.174301075268816e-05, | |
| "loss": 0.8184, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 27.03, | |
| "learning_rate": 6.0990322580645156e-05, | |
| "loss": 0.7995, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 27.03, | |
| "eval_loss": 0.42021921277046204, | |
| "eval_runtime": 67.5681, | |
| "eval_samples_per_second": 24.05, | |
| "eval_steps_per_second": 1.51, | |
| "eval_wer": 0.4770357604480827, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 27.93, | |
| "learning_rate": 6.023763440860215e-05, | |
| "loss": 0.7849, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 28.83, | |
| "learning_rate": 5.9484946236559134e-05, | |
| "loss": 0.7679, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 29.73, | |
| "learning_rate": 5.873225806451612e-05, | |
| "loss": 0.78, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 30.63, | |
| "learning_rate": 5.797956989247311e-05, | |
| "loss": 0.7505, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 31.53, | |
| "learning_rate": 5.72268817204301e-05, | |
| "loss": 0.7414, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 32.43, | |
| "learning_rate": 5.648172043010752e-05, | |
| "loss": 0.7269, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 33.33, | |
| "learning_rate": 5.5729032258064514e-05, | |
| "loss": 0.7023, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 34.23, | |
| "learning_rate": 5.49763440860215e-05, | |
| "loss": 0.7223, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 35.14, | |
| "learning_rate": 5.422365591397849e-05, | |
| "loss": 0.6899, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 36.04, | |
| "learning_rate": 5.3470967741935484e-05, | |
| "loss": 0.6978, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 36.04, | |
| "eval_loss": 0.3053520619869232, | |
| "eval_runtime": 66.0235, | |
| "eval_samples_per_second": 24.612, | |
| "eval_steps_per_second": 1.545, | |
| "eval_wer": 0.3494183541576906, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 36.94, | |
| "learning_rate": 5.271827956989247e-05, | |
| "loss": 0.683, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 37.84, | |
| "learning_rate": 5.1965591397849456e-05, | |
| "loss": 0.652, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 38.74, | |
| "learning_rate": 5.121290322580644e-05, | |
| "loss": 0.6738, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 39.64, | |
| "learning_rate": 5.0460215053763434e-05, | |
| "loss": 0.661, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 40.54, | |
| "learning_rate": 4.9707526881720426e-05, | |
| "loss": 0.6532, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 41.44, | |
| "learning_rate": 4.895483870967741e-05, | |
| "loss": 0.6375, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 42.34, | |
| "learning_rate": 4.8202150537634404e-05, | |
| "loss": 0.6339, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 43.24, | |
| "learning_rate": 4.74494623655914e-05, | |
| "loss": 0.624, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 44.14, | |
| "learning_rate": 4.669677419354838e-05, | |
| "loss": 0.6087, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 45.05, | |
| "learning_rate": 4.5944086021505375e-05, | |
| "loss": 0.6189, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 45.05, | |
| "eval_loss": 0.2878226935863495, | |
| "eval_runtime": 65.3432, | |
| "eval_samples_per_second": 24.869, | |
| "eval_steps_per_second": 1.561, | |
| "eval_wer": 0.31538130116329166, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 45.95, | |
| "learning_rate": 4.519139784946237e-05, | |
| "loss": 0.6067, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 46.85, | |
| "learning_rate": 4.443870967741935e-05, | |
| "loss": 0.6064, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 47.75, | |
| "learning_rate": 4.3686021505376345e-05, | |
| "loss": 0.5911, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 48.65, | |
| "learning_rate": 4.2933333333333324e-05, | |
| "loss": 0.6038, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 49.55, | |
| "learning_rate": 4.218064516129032e-05, | |
| "loss": 0.5886, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 50.45, | |
| "learning_rate": 4.142795698924731e-05, | |
| "loss": 0.5667, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 51.35, | |
| "learning_rate": 4.0675268817204295e-05, | |
| "loss": 0.5942, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 52.25, | |
| "learning_rate": 3.992258064516129e-05, | |
| "loss": 0.5785, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 53.15, | |
| "learning_rate": 3.916989247311827e-05, | |
| "loss": 0.5771, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 54.05, | |
| "learning_rate": 3.8417204301075265e-05, | |
| "loss": 0.5667, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 54.05, | |
| "eval_loss": 0.31139013171195984, | |
| "eval_runtime": 65.1359, | |
| "eval_samples_per_second": 24.948, | |
| "eval_steps_per_second": 1.566, | |
| "eval_wer": 0.32856527358897025, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 54.95, | |
| "learning_rate": 3.766451612903226e-05, | |
| "loss": 0.5553, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 55.86, | |
| "learning_rate": 3.6911827956989244e-05, | |
| "loss": 0.5524, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 56.76, | |
| "learning_rate": 3.616666666666667e-05, | |
| "loss": 0.5389, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 57.66, | |
| "learning_rate": 3.5413978494623646e-05, | |
| "loss": 0.5376, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 58.56, | |
| "learning_rate": 3.4661290322580645e-05, | |
| "loss": 0.5379, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 59.46, | |
| "learning_rate": 3.390860215053763e-05, | |
| "loss": 0.5398, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 60.36, | |
| "learning_rate": 3.3155913978494617e-05, | |
| "loss": 0.5388, | |
| "step": 6700 | |
| }, | |
| { | |
| "epoch": 61.26, | |
| "learning_rate": 3.240322580645161e-05, | |
| "loss": 0.5181, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 62.16, | |
| "learning_rate": 3.16505376344086e-05, | |
| "loss": 0.516, | |
| "step": 6900 | |
| }, | |
| { | |
| "epoch": 63.06, | |
| "learning_rate": 3.089784946236559e-05, | |
| "loss": 0.5173, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 63.06, | |
| "eval_loss": 0.3085043430328369, | |
| "eval_runtime": 65.8426, | |
| "eval_samples_per_second": 24.68, | |
| "eval_steps_per_second": 1.549, | |
| "eval_wer": 0.3021111589831969, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 63.96, | |
| "learning_rate": 3.014516129032258e-05, | |
| "loss": 0.515, | |
| "step": 7100 | |
| }, | |
| { | |
| "epoch": 64.86, | |
| "learning_rate": 2.939247311827957e-05, | |
| "loss": 0.5092, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 65.77, | |
| "learning_rate": 2.8639784946236554e-05, | |
| "loss": 0.5202, | |
| "step": 7300 | |
| }, | |
| { | |
| "epoch": 66.67, | |
| "learning_rate": 2.7887096774193547e-05, | |
| "loss": 0.5008, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 67.57, | |
| "learning_rate": 2.7134408602150536e-05, | |
| "loss": 0.4936, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 68.47, | |
| "learning_rate": 2.6381720430107525e-05, | |
| "loss": 0.4843, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 69.37, | |
| "learning_rate": 2.5629032258064514e-05, | |
| "loss": 0.4771, | |
| "step": 7700 | |
| }, | |
| { | |
| "epoch": 70.27, | |
| "learning_rate": 2.4876344086021503e-05, | |
| "loss": 0.4865, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 71.17, | |
| "learning_rate": 2.4123655913978492e-05, | |
| "loss": 0.4812, | |
| "step": 7900 | |
| }, | |
| { | |
| "epoch": 72.07, | |
| "learning_rate": 2.337096774193548e-05, | |
| "loss": 0.4682, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 72.07, | |
| "eval_loss": 0.305768221616745, | |
| "eval_runtime": 65.7283, | |
| "eval_samples_per_second": 24.723, | |
| "eval_steps_per_second": 1.552, | |
| "eval_wer": 0.2969409737182249, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 72.97, | |
| "learning_rate": 2.261827956989247e-05, | |
| "loss": 0.4835, | |
| "step": 8100 | |
| }, | |
| { | |
| "epoch": 73.87, | |
| "learning_rate": 2.1865591397849463e-05, | |
| "loss": 0.4734, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 74.77, | |
| "learning_rate": 2.111290322580645e-05, | |
| "loss": 0.4685, | |
| "step": 8300 | |
| }, | |
| { | |
| "epoch": 75.68, | |
| "learning_rate": 2.0360215053763437e-05, | |
| "loss": 0.4556, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 76.58, | |
| "learning_rate": 1.9607526881720426e-05, | |
| "loss": 0.4655, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 77.48, | |
| "learning_rate": 1.885483870967742e-05, | |
| "loss": 0.4623, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 78.38, | |
| "learning_rate": 1.8102150537634408e-05, | |
| "loss": 0.4661, | |
| "step": 8700 | |
| }, | |
| { | |
| "epoch": 79.28, | |
| "learning_rate": 1.7356989247311825e-05, | |
| "loss": 0.451, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 80.18, | |
| "learning_rate": 1.6604301075268817e-05, | |
| "loss": 0.4482, | |
| "step": 8900 | |
| }, | |
| { | |
| "epoch": 81.08, | |
| "learning_rate": 1.5851612903225803e-05, | |
| "loss": 0.451, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 81.08, | |
| "eval_loss": 0.31458327174186707, | |
| "eval_runtime": 66.1124, | |
| "eval_samples_per_second": 24.579, | |
| "eval_steps_per_second": 1.543, | |
| "eval_wer": 0.2907367514002585, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 81.98, | |
| "learning_rate": 1.5098924731182794e-05, | |
| "loss": 0.4504, | |
| "step": 9100 | |
| }, | |
| { | |
| "epoch": 82.88, | |
| "learning_rate": 1.4346236559139784e-05, | |
| "loss": 0.4373, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 83.78, | |
| "learning_rate": 1.3593548387096772e-05, | |
| "loss": 0.4257, | |
| "step": 9300 | |
| }, | |
| { | |
| "epoch": 84.68, | |
| "learning_rate": 1.2840860215053762e-05, | |
| "loss": 0.4454, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 85.59, | |
| "learning_rate": 1.2088172043010753e-05, | |
| "loss": 0.4303, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 86.49, | |
| "learning_rate": 1.133548387096774e-05, | |
| "loss": 0.4348, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 87.39, | |
| "learning_rate": 1.058279569892473e-05, | |
| "loss": 0.4411, | |
| "step": 9700 | |
| }, | |
| { | |
| "epoch": 88.29, | |
| "learning_rate": 9.83010752688172e-06, | |
| "loss": 0.4285, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 89.19, | |
| "learning_rate": 9.077419354838708e-06, | |
| "loss": 0.4235, | |
| "step": 9900 | |
| }, | |
| { | |
| "epoch": 90.09, | |
| "learning_rate": 8.324731182795698e-06, | |
| "loss": 0.4213, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 90.09, | |
| "eval_loss": 0.3030133545398712, | |
| "eval_runtime": 66.0396, | |
| "eval_samples_per_second": 24.606, | |
| "eval_steps_per_second": 1.545, | |
| "eval_wer": 0.2880654890133563, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 90.99, | |
| "learning_rate": 7.5720430107526875e-06, | |
| "loss": 0.4148, | |
| "step": 10100 | |
| }, | |
| { | |
| "epoch": 91.89, | |
| "learning_rate": 6.8193548387096766e-06, | |
| "loss": 0.4157, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 92.79, | |
| "learning_rate": 6.0666666666666665e-06, | |
| "loss": 0.4146, | |
| "step": 10300 | |
| }, | |
| { | |
| "epoch": 93.69, | |
| "learning_rate": 5.3139784946236555e-06, | |
| "loss": 0.4101, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 94.59, | |
| "learning_rate": 4.561290322580645e-06, | |
| "loss": 0.428, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 95.5, | |
| "learning_rate": 3.808602150537634e-06, | |
| "loss": 0.4138, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 96.4, | |
| "learning_rate": 3.0559139784946235e-06, | |
| "loss": 0.4012, | |
| "step": 10700 | |
| }, | |
| { | |
| "epoch": 97.3, | |
| "learning_rate": 2.303225806451613e-06, | |
| "loss": 0.3957, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 98.2, | |
| "learning_rate": 1.550537634408602e-06, | |
| "loss": 0.4195, | |
| "step": 10900 | |
| }, | |
| { | |
| "epoch": 99.1, | |
| "learning_rate": 8.053763440860214e-07, | |
| "loss": 0.4005, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 99.1, | |
| "eval_loss": 0.300111323595047, | |
| "eval_runtime": 66.7539, | |
| "eval_samples_per_second": 24.343, | |
| "eval_steps_per_second": 1.528, | |
| "eval_wer": 0.2789314950452391, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 5.268817204301075e-08, | |
| "loss": 0.4141, | |
| "step": 11100 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 11100, | |
| "total_flos": 5.044434281896765e+19, | |
| "train_loss": 1.0701171730660104, | |
| "train_runtime": 18720.6378, | |
| "train_samples_per_second": 18.952, | |
| "train_steps_per_second": 0.593 | |
| } | |
| ], | |
| "max_steps": 11100, | |
| "num_train_epochs": 100, | |
| "total_flos": 5.044434281896765e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |