| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 5.020080321285141, | |
| "eval_steps": 200, | |
| "global_step": 10000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.10040160642570281, | |
| "eval_loss": 0.3807084858417511, | |
| "eval_runtime": 206.6442, | |
| "eval_samples_per_second": 26.04, | |
| "eval_steps_per_second": 0.411, | |
| "eval_wer": 0.2514466384298753, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.20080321285140562, | |
| "eval_loss": 0.253967821598053, | |
| "eval_runtime": 206.4967, | |
| "eval_samples_per_second": 26.059, | |
| "eval_steps_per_second": 0.412, | |
| "eval_wer": 0.2642923899858816, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.25100401606425704, | |
| "grad_norm": 1.857898235321045, | |
| "learning_rate": 0.0001491, | |
| "loss": 2.4874, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.30120481927710846, | |
| "eval_loss": 0.2642447352409363, | |
| "eval_runtime": 205.9594, | |
| "eval_samples_per_second": 26.127, | |
| "eval_steps_per_second": 0.413, | |
| "eval_wer": 0.30376424267732505, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.40160642570281124, | |
| "eval_loss": 0.3125462234020233, | |
| "eval_runtime": 205.1387, | |
| "eval_samples_per_second": 26.231, | |
| "eval_steps_per_second": 0.414, | |
| "eval_wer": 0.39048300821253157, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.5020080321285141, | |
| "grad_norm": 3.634869337081909, | |
| "learning_rate": 0.00029909999999999995, | |
| "loss": 0.3991, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.5020080321285141, | |
| "eval_loss": 0.3531426191329956, | |
| "eval_runtime": 205.1455, | |
| "eval_samples_per_second": 26.23, | |
| "eval_steps_per_second": 0.414, | |
| "eval_wer": 0.3939430094056354, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.6024096385542169, | |
| "eval_loss": 0.3571958839893341, | |
| "eval_runtime": 209.3856, | |
| "eval_samples_per_second": 25.699, | |
| "eval_steps_per_second": 0.406, | |
| "eval_wer": 0.40390542663405515, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.7028112449799196, | |
| "eval_loss": 0.36791086196899414, | |
| "eval_runtime": 206.8164, | |
| "eval_samples_per_second": 26.018, | |
| "eval_steps_per_second": 0.411, | |
| "eval_wer": 0.4052576110083716, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.7530120481927711, | |
| "grad_norm": 3.5483558177948, | |
| "learning_rate": 0.0002834333333333333, | |
| "loss": 0.4512, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.8032128514056225, | |
| "eval_loss": 0.35897189378738403, | |
| "eval_runtime": 207.7252, | |
| "eval_samples_per_second": 25.904, | |
| "eval_steps_per_second": 0.409, | |
| "eval_wer": 0.38767921414225776, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.9036144578313253, | |
| "eval_loss": 0.3732704222202301, | |
| "eval_runtime": 205.6494, | |
| "eval_samples_per_second": 26.166, | |
| "eval_steps_per_second": 0.413, | |
| "eval_wer": 0.4006840462128895, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.0040160642570282, | |
| "grad_norm": 2.2905380725860596, | |
| "learning_rate": 0.00026676666666666663, | |
| "loss": 0.4333, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.0040160642570282, | |
| "eval_loss": 0.377088725566864, | |
| "eval_runtime": 204.613, | |
| "eval_samples_per_second": 26.298, | |
| "eval_steps_per_second": 0.415, | |
| "eval_wer": 0.4243273876990992, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.104417670682731, | |
| "eval_loss": 0.3604430556297302, | |
| "eval_runtime": 208.0965, | |
| "eval_samples_per_second": 25.858, | |
| "eval_steps_per_second": 0.408, | |
| "eval_wer": 0.3867048459901768, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 1.2048192771084336, | |
| "eval_loss": 0.3431110978126526, | |
| "eval_runtime": 206.2637, | |
| "eval_samples_per_second": 26.088, | |
| "eval_steps_per_second": 0.412, | |
| "eval_wer": 0.38137564875022373, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.2550200803212852, | |
| "grad_norm": 1.8699342012405396, | |
| "learning_rate": 0.00025009999999999995, | |
| "loss": 0.3468, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.3052208835341366, | |
| "eval_loss": 0.32902058959007263, | |
| "eval_runtime": 205.4777, | |
| "eval_samples_per_second": 26.188, | |
| "eval_steps_per_second": 0.414, | |
| "eval_wer": 0.3778559923641353, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.4056224899598393, | |
| "eval_loss": 0.33407700061798096, | |
| "eval_runtime": 205.6359, | |
| "eval_samples_per_second": 26.168, | |
| "eval_steps_per_second": 0.413, | |
| "eval_wer": 0.3647119648432063, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.5060240963855422, | |
| "grad_norm": 2.062389373779297, | |
| "learning_rate": 0.0002334333333333333, | |
| "loss": 0.3503, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.5060240963855422, | |
| "eval_loss": 0.3247535228729248, | |
| "eval_runtime": 206.3116, | |
| "eval_samples_per_second": 26.082, | |
| "eval_steps_per_second": 0.412, | |
| "eval_wer": 0.3614706993577124, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.606425702811245, | |
| "eval_loss": 0.33116209506988525, | |
| "eval_runtime": 203.9912, | |
| "eval_samples_per_second": 26.379, | |
| "eval_steps_per_second": 0.417, | |
| "eval_wer": 0.35512736383702204, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.7068273092369477, | |
| "eval_loss": 0.3410908281803131, | |
| "eval_runtime": 204.5054, | |
| "eval_samples_per_second": 26.312, | |
| "eval_steps_per_second": 0.416, | |
| "eval_wer": 0.3836226610193084, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.7570281124497993, | |
| "grad_norm": 0.9907544255256653, | |
| "learning_rate": 0.00021679999999999998, | |
| "loss": 0.3418, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.8072289156626506, | |
| "eval_loss": 0.3116574287414551, | |
| "eval_runtime": 205.0392, | |
| "eval_samples_per_second": 26.244, | |
| "eval_steps_per_second": 0.415, | |
| "eval_wer": 0.33752908190658, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.9076305220883534, | |
| "eval_loss": 0.3196774423122406, | |
| "eval_runtime": 206.2716, | |
| "eval_samples_per_second": 26.087, | |
| "eval_steps_per_second": 0.412, | |
| "eval_wer": 0.34317644017578397, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 2.0080321285140563, | |
| "grad_norm": 1.0384626388549805, | |
| "learning_rate": 0.0002001333333333333, | |
| "loss": 0.3181, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.0080321285140563, | |
| "eval_loss": 0.30675315856933594, | |
| "eval_runtime": 206.0737, | |
| "eval_samples_per_second": 26.112, | |
| "eval_steps_per_second": 0.412, | |
| "eval_wer": 0.3339696553918352, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.108433734939759, | |
| "eval_loss": 0.31376445293426514, | |
| "eval_runtime": 209.2791, | |
| "eval_samples_per_second": 25.712, | |
| "eval_steps_per_second": 0.406, | |
| "eval_wer": 0.3358388514386844, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 2.208835341365462, | |
| "eval_loss": 0.31388720870018005, | |
| "eval_runtime": 204.9118, | |
| "eval_samples_per_second": 26.26, | |
| "eval_steps_per_second": 0.415, | |
| "eval_wer": 0.3333731034619897, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 2.2590361445783134, | |
| "grad_norm": 0.5868389010429382, | |
| "learning_rate": 0.00018346666666666664, | |
| "loss": 0.2423, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.3092369477911645, | |
| "eval_loss": 0.3191888928413391, | |
| "eval_runtime": 204.834, | |
| "eval_samples_per_second": 26.27, | |
| "eval_steps_per_second": 0.415, | |
| "eval_wer": 0.32848137763725666, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 2.4096385542168672, | |
| "eval_loss": 0.2928995192050934, | |
| "eval_runtime": 204.43, | |
| "eval_samples_per_second": 26.322, | |
| "eval_steps_per_second": 0.416, | |
| "eval_wer": 0.31682872994094136, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.5100401606425704, | |
| "grad_norm": 1.3247759342193604, | |
| "learning_rate": 0.0001668, | |
| "loss": 0.2327, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.5100401606425704, | |
| "eval_loss": 0.29208171367645264, | |
| "eval_runtime": 206.3612, | |
| "eval_samples_per_second": 26.076, | |
| "eval_steps_per_second": 0.412, | |
| "eval_wer": 0.3103064288412973, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.610441767068273, | |
| "eval_loss": 0.2801830470561981, | |
| "eval_runtime": 204.3678, | |
| "eval_samples_per_second": 26.33, | |
| "eval_steps_per_second": 0.416, | |
| "eval_wer": 0.3037443576129969, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 2.710843373493976, | |
| "eval_loss": 0.2811721861362457, | |
| "eval_runtime": 204.4403, | |
| "eval_samples_per_second": 26.321, | |
| "eval_steps_per_second": 0.416, | |
| "eval_wer": 0.29624768836127185, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 2.7610441767068274, | |
| "grad_norm": 1.381541132926941, | |
| "learning_rate": 0.00015013333333333331, | |
| "loss": 0.2374, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.8112449799196786, | |
| "eval_loss": 0.28872984647750854, | |
| "eval_runtime": 204.3069, | |
| "eval_samples_per_second": 26.338, | |
| "eval_steps_per_second": 0.416, | |
| "eval_wer": 0.30422159915687325, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 2.9116465863453813, | |
| "eval_loss": 0.27397701144218445, | |
| "eval_runtime": 204.1464, | |
| "eval_samples_per_second": 26.359, | |
| "eval_steps_per_second": 0.416, | |
| "eval_wer": 0.2927081469108553, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 3.0120481927710845, | |
| "grad_norm": 1.4617916345596313, | |
| "learning_rate": 0.00013346666666666667, | |
| "loss": 0.2136, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.0120481927710845, | |
| "eval_loss": 0.2662462592124939, | |
| "eval_runtime": 203.8941, | |
| "eval_samples_per_second": 26.391, | |
| "eval_steps_per_second": 0.417, | |
| "eval_wer": 0.28296446539004555, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.112449799196787, | |
| "eval_loss": 0.28285130858421326, | |
| "eval_runtime": 206.1704, | |
| "eval_samples_per_second": 26.1, | |
| "eval_steps_per_second": 0.412, | |
| "eval_wer": 0.2890294100101414, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 3.21285140562249, | |
| "eval_loss": 0.2729070484638214, | |
| "eval_runtime": 206.2438, | |
| "eval_samples_per_second": 26.09, | |
| "eval_steps_per_second": 0.412, | |
| "eval_wer": 0.28692159319135396, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 3.2630522088353415, | |
| "grad_norm": 0.8870707750320435, | |
| "learning_rate": 0.00011679999999999998, | |
| "loss": 0.167, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.3132530120481927, | |
| "eval_loss": 0.2776893675327301, | |
| "eval_runtime": 204.2022, | |
| "eval_samples_per_second": 26.351, | |
| "eval_steps_per_second": 0.416, | |
| "eval_wer": 0.28892998468850045, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 3.4136546184738954, | |
| "eval_loss": 0.2711654603481293, | |
| "eval_runtime": 203.7376, | |
| "eval_samples_per_second": 26.411, | |
| "eval_steps_per_second": 0.417, | |
| "eval_wer": 0.28095607389289906, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 3.5140562248995986, | |
| "grad_norm": 1.0165985822677612, | |
| "learning_rate": 0.00010013333333333333, | |
| "loss": 0.1614, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.5140562248995986, | |
| "eval_loss": 0.2688385844230652, | |
| "eval_runtime": 204.1623, | |
| "eval_samples_per_second": 26.356, | |
| "eval_steps_per_second": 0.416, | |
| "eval_wer": 0.27091411640716656, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.6144578313253013, | |
| "eval_loss": 0.2589295208454132, | |
| "eval_runtime": 205.2749, | |
| "eval_samples_per_second": 26.214, | |
| "eval_steps_per_second": 0.414, | |
| "eval_wer": 0.26626101135437175, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 3.714859437751004, | |
| "eval_loss": 0.26514673233032227, | |
| "eval_runtime": 204.2135, | |
| "eval_samples_per_second": 26.35, | |
| "eval_steps_per_second": 0.416, | |
| "eval_wer": 0.2669768736701863, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 3.765060240963855, | |
| "grad_norm": 0.7397546172142029, | |
| "learning_rate": 8.346666666666666e-05, | |
| "loss": 0.1529, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.8152610441767068, | |
| "eval_loss": 0.25074735283851624, | |
| "eval_runtime": 204.2336, | |
| "eval_samples_per_second": 26.347, | |
| "eval_steps_per_second": 0.416, | |
| "eval_wer": 0.2637157231203643, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 3.9156626506024095, | |
| "eval_loss": 0.2493942528963089, | |
| "eval_runtime": 206.072, | |
| "eval_samples_per_second": 26.112, | |
| "eval_steps_per_second": 0.412, | |
| "eval_wer": 0.2567957207341566, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 4.016064257028113, | |
| "grad_norm": 0.785851776599884, | |
| "learning_rate": 6.68e-05, | |
| "loss": 0.1496, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.016064257028113, | |
| "eval_loss": 0.25821030139923096, | |
| "eval_runtime": 204.9558, | |
| "eval_samples_per_second": 26.254, | |
| "eval_steps_per_second": 0.415, | |
| "eval_wer": 0.2580484797868321, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 4.116465863453815, | |
| "eval_loss": 0.2650238871574402, | |
| "eval_runtime": 204.7852, | |
| "eval_samples_per_second": 26.276, | |
| "eval_steps_per_second": 0.415, | |
| "eval_wer": 0.25753146811429933, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 4.216867469879518, | |
| "eval_loss": 0.26561084389686584, | |
| "eval_runtime": 210.1403, | |
| "eval_samples_per_second": 25.607, | |
| "eval_steps_per_second": 0.404, | |
| "eval_wer": 0.25598043309670104, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 4.267068273092369, | |
| "grad_norm": 0.34119465947151184, | |
| "learning_rate": 5.013333333333332e-05, | |
| "loss": 0.1128, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 4.317269076305221, | |
| "eval_loss": 0.25430822372436523, | |
| "eval_runtime": 203.852, | |
| "eval_samples_per_second": 26.397, | |
| "eval_steps_per_second": 0.417, | |
| "eval_wer": 0.25118813259360895, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 4.417670682730924, | |
| "eval_loss": 0.2586837112903595, | |
| "eval_runtime": 202.7677, | |
| "eval_samples_per_second": 26.538, | |
| "eval_steps_per_second": 0.419, | |
| "eval_wer": 0.24987571834794886, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 4.518072289156627, | |
| "grad_norm": 0.5558347105979919, | |
| "learning_rate": 3.346666666666666e-05, | |
| "loss": 0.1109, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.518072289156627, | |
| "eval_loss": 0.2540307939052582, | |
| "eval_runtime": 202.8954, | |
| "eval_samples_per_second": 26.521, | |
| "eval_steps_per_second": 0.419, | |
| "eval_wer": 0.24599813080395316, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 4.618473895582329, | |
| "eval_loss": 0.2546459436416626, | |
| "eval_runtime": 208.4343, | |
| "eval_samples_per_second": 25.816, | |
| "eval_steps_per_second": 0.408, | |
| "eval_wer": 0.24245858935353656, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 4.718875502008032, | |
| "eval_loss": 0.25800344347953796, | |
| "eval_runtime": 205.7304, | |
| "eval_samples_per_second": 26.156, | |
| "eval_steps_per_second": 0.413, | |
| "eval_wer": 0.24198134780966016, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 4.769076305220883, | |
| "grad_norm": 0.9226210117340088, | |
| "learning_rate": 1.68e-05, | |
| "loss": 0.1028, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 4.8192771084337345, | |
| "eval_loss": 0.25135332345962524, | |
| "eval_runtime": 203.1799, | |
| "eval_samples_per_second": 26.484, | |
| "eval_steps_per_second": 0.418, | |
| "eval_wer": 0.24035077253474915, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 4.919678714859438, | |
| "eval_loss": 0.2509777843952179, | |
| "eval_runtime": 203.2596, | |
| "eval_samples_per_second": 26.474, | |
| "eval_steps_per_second": 0.418, | |
| "eval_wer": 0.24025134721310823, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 5.020080321285141, | |
| "grad_norm": 4.224822521209717, | |
| "learning_rate": 1.3333333333333334e-07, | |
| "loss": 0.1069, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 5.020080321285141, | |
| "eval_loss": 0.2515573501586914, | |
| "eval_runtime": 209.1262, | |
| "eval_samples_per_second": 25.731, | |
| "eval_steps_per_second": 0.406, | |
| "eval_wer": 0.24033088747042097, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 5.020080321285141, | |
| "step": 10000, | |
| "total_flos": 6.356146932571761e+18, | |
| "train_loss": 0.3559208065032959, | |
| "train_runtime": 12994.9936, | |
| "train_samples_per_second": 3.078, | |
| "train_steps_per_second": 0.77 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 10000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 6, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.356146932571761e+18, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |