| { | |
| "best_metric": 12.206885082321635, | |
| "best_model_checkpoint": "./checkpoint-6000", | |
| "epoch": 1.0958333333333334, | |
| "global_step": 6000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 0.0, | |
| "loss": 2.996, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 9.4e-06, | |
| "loss": 1.9547, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.94e-05, | |
| "loss": 0.7554, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 2.94e-05, | |
| "loss": 0.5206, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 3.94e-05, | |
| "loss": 0.4737, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 4.94e-05, | |
| "loss": 0.3952, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 5.94e-05, | |
| "loss": 0.3662, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 6.939999999999999e-05, | |
| "loss": 0.3557, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 7.94e-05, | |
| "loss": 0.3687, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 8.94e-05, | |
| "loss": 0.3288, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 9.94e-05, | |
| "loss": 0.2986, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 9.914545454545455e-05, | |
| "loss": 0.3076, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 9.823636363636364e-05, | |
| "loss": 0.2977, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 9.732727272727273e-05, | |
| "loss": 0.3167, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 9.641818181818182e-05, | |
| "loss": 0.2799, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 9.550909090909092e-05, | |
| "loss": 0.2079, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 9.46e-05, | |
| "loss": 0.2246, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 9.36909090909091e-05, | |
| "loss": 0.2568, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 9.278181818181819e-05, | |
| "loss": 0.2734, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 9.187272727272727e-05, | |
| "loss": 0.183, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 9.096363636363638e-05, | |
| "loss": 0.2445, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "eval_loss": 0.3058685064315796, | |
| "eval_runtime": 2133.1044, | |
| "eval_samples_per_second": 7.441, | |
| "eval_steps_per_second": 0.233, | |
| "eval_wer": 32.41627267903669, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 9.005454545454545e-05, | |
| "loss": 0.1608, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 8.914545454545455e-05, | |
| "loss": 0.1718, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 8.823636363636364e-05, | |
| "loss": 0.1721, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 8.732727272727273e-05, | |
| "loss": 0.1762, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 8.641818181818182e-05, | |
| "loss": 0.1587, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 8.550909090909092e-05, | |
| "loss": 0.2133, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 8.46e-05, | |
| "loss": 0.2291, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 8.36909090909091e-05, | |
| "loss": 0.1172, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 8.278181818181819e-05, | |
| "loss": 0.1034, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 8.187272727272728e-05, | |
| "loss": 0.1746, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 8.096363636363636e-05, | |
| "loss": 0.1649, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 8.005454545454547e-05, | |
| "loss": 0.1153, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 7.914545454545454e-05, | |
| "loss": 0.1521, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 7.823636363636364e-05, | |
| "loss": 0.1684, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 7.732727272727273e-05, | |
| "loss": 0.2243, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 7.641818181818182e-05, | |
| "loss": 0.2149, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 7.550909090909092e-05, | |
| "loss": 0.1927, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 7.46e-05, | |
| "loss": 0.2019, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 7.36909090909091e-05, | |
| "loss": 0.2003, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 7.278181818181819e-05, | |
| "loss": 0.1823, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "eval_loss": 0.20035691559314728, | |
| "eval_runtime": 2105.4622, | |
| "eval_samples_per_second": 7.538, | |
| "eval_steps_per_second": 0.236, | |
| "eval_wer": 22.12587013850942, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 7.187272727272728e-05, | |
| "loss": 0.1707, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 7.096363636363636e-05, | |
| "loss": 0.1659, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 7.005454545454547e-05, | |
| "loss": 0.193, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 6.914545454545454e-05, | |
| "loss": 0.1882, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 6.823636363636364e-05, | |
| "loss": 0.163, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 6.732727272727273e-05, | |
| "loss": 0.1689, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 6.641818181818182e-05, | |
| "loss": 0.1585, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 6.550909090909091e-05, | |
| "loss": 0.1643, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 6.460000000000001e-05, | |
| "loss": 0.1634, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 6.369090909090909e-05, | |
| "loss": 0.1362, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 6.278181818181819e-05, | |
| "loss": 0.1446, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 6.187272727272728e-05, | |
| "loss": 0.1439, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 6.0963636363636366e-05, | |
| "loss": 0.1534, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 6.0054545454545454e-05, | |
| "loss": 0.1264, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 5.914545454545455e-05, | |
| "loss": 0.1373, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 5.8236363636363645e-05, | |
| "loss": 0.1406, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 5.732727272727273e-05, | |
| "loss": 0.1393, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 5.641818181818183e-05, | |
| "loss": 0.1377, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 5.550909090909091e-05, | |
| "loss": 0.1467, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 5.4600000000000006e-05, | |
| "loss": 0.1412, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "eval_loss": 0.17524316906929016, | |
| "eval_runtime": 2121.3824, | |
| "eval_samples_per_second": 7.482, | |
| "eval_steps_per_second": 0.234, | |
| "eval_wer": 20.070007048220916, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 5.3690909090909094e-05, | |
| "loss": 0.1186, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 5.278181818181819e-05, | |
| "loss": 0.1437, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 5.187272727272727e-05, | |
| "loss": 0.1287, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 5.096363636363637e-05, | |
| "loss": 0.1242, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 5.0054545454545455e-05, | |
| "loss": 0.1431, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 4.914545454545455e-05, | |
| "loss": 0.1316, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 4.823636363636364e-05, | |
| "loss": 0.1119, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 4.7327272727272734e-05, | |
| "loss": 0.1286, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 4.641818181818182e-05, | |
| "loss": 0.1479, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 4.550909090909091e-05, | |
| "loss": 0.1144, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 4.46e-05, | |
| "loss": 0.1071, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 4.3690909090909095e-05, | |
| "loss": 0.1525, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 4.2781818181818184e-05, | |
| "loss": 0.1158, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 4.187272727272727e-05, | |
| "loss": 0.1111, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 4.096363636363637e-05, | |
| "loss": 0.1237, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 4.0054545454545456e-05, | |
| "loss": 0.154, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 3.9145454545454545e-05, | |
| "loss": 0.1298, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 3.823636363636364e-05, | |
| "loss": 0.1273, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 3.732727272727273e-05, | |
| "loss": 0.121, | |
| "step": 3950 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 3.641818181818182e-05, | |
| "loss": 0.1093, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "eval_loss": 0.14134088158607483, | |
| "eval_runtime": 2102.7704, | |
| "eval_samples_per_second": 7.548, | |
| "eval_steps_per_second": 0.236, | |
| "eval_wer": 16.053313059799006, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 3.5509090909090905e-05, | |
| "loss": 0.1054, | |
| "step": 4050 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 3.46e-05, | |
| "loss": 0.1305, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 3.3690909090909096e-05, | |
| "loss": 0.1138, | |
| "step": 4150 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 3.2781818181818185e-05, | |
| "loss": 0.1275, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 3.187272727272727e-05, | |
| "loss": 0.0885, | |
| "step": 4250 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 3.096363636363637e-05, | |
| "loss": 0.1089, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 3.0054545454545457e-05, | |
| "loss": 0.0838, | |
| "step": 4350 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 2.914545454545455e-05, | |
| "loss": 0.0833, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 2.8236363636363637e-05, | |
| "loss": 0.0968, | |
| "step": 4450 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 2.732727272727273e-05, | |
| "loss": 0.0962, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 2.641818181818182e-05, | |
| "loss": 0.0779, | |
| "step": 4550 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 2.550909090909091e-05, | |
| "loss": 0.1033, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 2.46e-05, | |
| "loss": 0.0762, | |
| "step": 4650 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 2.369090909090909e-05, | |
| "loss": 0.1111, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 2.2781818181818182e-05, | |
| "loss": 0.0875, | |
| "step": 4750 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 2.1872727272727274e-05, | |
| "loss": 0.1337, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 2.0963636363636366e-05, | |
| "loss": 0.1232, | |
| "step": 4850 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 2.0054545454545458e-05, | |
| "loss": 0.1153, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 1.9145454545454546e-05, | |
| "loss": 0.1189, | |
| "step": 4950 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 1.8236363636363638e-05, | |
| "loss": 0.1137, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "eval_loss": 0.11552777886390686, | |
| "eval_runtime": 2104.6449, | |
| "eval_samples_per_second": 7.541, | |
| "eval_steps_per_second": 0.236, | |
| "eval_wer": 13.31084238118996, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 1.7327272727272727e-05, | |
| "loss": 0.1159, | |
| "step": 5050 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 1.641818181818182e-05, | |
| "loss": 0.1152, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 1.550909090909091e-05, | |
| "loss": 0.1051, | |
| "step": 5150 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 1.4599999999999999e-05, | |
| "loss": 0.1196, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 1.3690909090909091e-05, | |
| "loss": 0.117, | |
| "step": 5250 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 1.2781818181818183e-05, | |
| "loss": 0.0972, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.1872727272727273e-05, | |
| "loss": 0.0965, | |
| "step": 5350 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.0963636363636365e-05, | |
| "loss": 0.1008, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 1.0054545454545455e-05, | |
| "loss": 0.0843, | |
| "step": 5450 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 9.145454545454546e-06, | |
| "loss": 0.0583, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 8.236363636363636e-06, | |
| "loss": 0.06, | |
| "step": 5550 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 7.327272727272727e-06, | |
| "loss": 0.0601, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 6.418181818181819e-06, | |
| "loss": 0.0591, | |
| "step": 5650 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 5.50909090909091e-06, | |
| "loss": 0.0608, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 4.6e-06, | |
| "loss": 0.0628, | |
| "step": 5750 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 3.6909090909090915e-06, | |
| "loss": 0.0636, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 2.781818181818182e-06, | |
| "loss": 0.0654, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 1.8727272727272728e-06, | |
| "loss": 0.061, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 9.636363636363636e-07, | |
| "loss": 0.0559, | |
| "step": 5950 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 5.454545454545455e-08, | |
| "loss": 0.0585, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "eval_loss": 0.10796044021844864, | |
| "eval_runtime": 2137.698, | |
| "eval_samples_per_second": 7.425, | |
| "eval_steps_per_second": 0.232, | |
| "eval_wer": 12.206885082321635, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "step": 6000, | |
| "total_flos": 2.490351440560128e+19, | |
| "train_loss": 0.17721048017342886, | |
| "train_runtime": 61737.3829, | |
| "train_samples_per_second": 6.22, | |
| "train_steps_per_second": 0.097 | |
| } | |
| ], | |
| "max_steps": 6000, | |
| "num_train_epochs": 9223372036854775807, | |
| "total_flos": 2.490351440560128e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |