| { | |
| "best_metric": 99.32937344318835, | |
| "best_model_checkpoint": "./whisper-tiny-ha/checkpoint-3000", | |
| "epoch": 25.0, | |
| "eval_steps": 500, | |
| "global_step": 3925, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1592356687898089, | |
| "grad_norm": 30.159414291381836, | |
| "learning_rate": 4.2000000000000004e-06, | |
| "loss": 3.418, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3184713375796178, | |
| "grad_norm": 21.70500373840332, | |
| "learning_rate": 9.200000000000002e-06, | |
| "loss": 2.553, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.47770700636942676, | |
| "grad_norm": 18.039278030395508, | |
| "learning_rate": 1e-05, | |
| "loss": 2.3832, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6369426751592356, | |
| "grad_norm": 18.557083129882812, | |
| "learning_rate": 1e-05, | |
| "loss": 2.3115, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7961783439490446, | |
| "grad_norm": 16.37763023376465, | |
| "learning_rate": 1e-05, | |
| "loss": 2.2126, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.9554140127388535, | |
| "grad_norm": 18.549989700317383, | |
| "learning_rate": 1e-05, | |
| "loss": 2.1729, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.1146496815286624, | |
| "grad_norm": 16.92412757873535, | |
| "learning_rate": 1e-05, | |
| "loss": 1.9976, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.2738853503184713, | |
| "grad_norm": 17.25154685974121, | |
| "learning_rate": 1e-05, | |
| "loss": 1.8783, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.4331210191082802, | |
| "grad_norm": 16.648714065551758, | |
| "learning_rate": 1e-05, | |
| "loss": 1.902, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.5923566878980893, | |
| "grad_norm": 16.841991424560547, | |
| "learning_rate": 1e-05, | |
| "loss": 1.925, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.7515923566878981, | |
| "grad_norm": 19.01357650756836, | |
| "learning_rate": 1e-05, | |
| "loss": 1.8552, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.910828025477707, | |
| "grad_norm": 16.73857879638672, | |
| "learning_rate": 1e-05, | |
| "loss": 1.8734, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.070063694267516, | |
| "grad_norm": 16.232080459594727, | |
| "learning_rate": 1e-05, | |
| "loss": 1.756, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.229299363057325, | |
| "grad_norm": 16.035682678222656, | |
| "learning_rate": 1e-05, | |
| "loss": 1.584, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.388535031847134, | |
| "grad_norm": 15.77041244506836, | |
| "learning_rate": 1e-05, | |
| "loss": 1.5446, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.5477707006369426, | |
| "grad_norm": 17.873939514160156, | |
| "learning_rate": 1e-05, | |
| "loss": 1.5419, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.7070063694267517, | |
| "grad_norm": 16.445964813232422, | |
| "learning_rate": 1e-05, | |
| "loss": 1.6085, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.8662420382165603, | |
| "grad_norm": 16.27969741821289, | |
| "learning_rate": 1e-05, | |
| "loss": 1.6325, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.0254777070063694, | |
| "grad_norm": 15.765183448791504, | |
| "learning_rate": 1e-05, | |
| "loss": 1.4205, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 3.1847133757961785, | |
| "grad_norm": 14.563206672668457, | |
| "learning_rate": 1e-05, | |
| "loss": 1.3176, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.1847133757961785, | |
| "eval_loss": 2.107278823852539, | |
| "eval_runtime": 154.4492, | |
| "eval_samples_per_second": 4.273, | |
| "eval_steps_per_second": 0.272, | |
| "eval_wer": 132.4391645909178, | |
| "eval_wer_ortho": 133.80859375, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.343949044585987, | |
| "grad_norm": 18.346054077148438, | |
| "learning_rate": 1e-05, | |
| "loss": 1.3377, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 3.5031847133757963, | |
| "grad_norm": 16.357585906982422, | |
| "learning_rate": 1e-05, | |
| "loss": 1.3085, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.662420382165605, | |
| "grad_norm": 16.26805305480957, | |
| "learning_rate": 1e-05, | |
| "loss": 1.2712, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 3.821656050955414, | |
| "grad_norm": 16.663305282592773, | |
| "learning_rate": 1e-05, | |
| "loss": 1.2537, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.980891719745223, | |
| "grad_norm": 15.779778480529785, | |
| "learning_rate": 1e-05, | |
| "loss": 1.2877, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 4.140127388535032, | |
| "grad_norm": 14.90344524383545, | |
| "learning_rate": 1e-05, | |
| "loss": 1.1309, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 4.2993630573248405, | |
| "grad_norm": 15.136467933654785, | |
| "learning_rate": 1e-05, | |
| "loss": 1.0086, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 4.45859872611465, | |
| "grad_norm": 16.454153060913086, | |
| "learning_rate": 1e-05, | |
| "loss": 1.0473, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.617834394904459, | |
| "grad_norm": 17.21575927734375, | |
| "learning_rate": 1e-05, | |
| "loss": 1.0229, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 4.777070063694268, | |
| "grad_norm": 18.05030059814453, | |
| "learning_rate": 1e-05, | |
| "loss": 1.0785, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.936305732484076, | |
| "grad_norm": 17.032936096191406, | |
| "learning_rate": 1e-05, | |
| "loss": 1.0569, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 5.095541401273885, | |
| "grad_norm": 16.049070358276367, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9171, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.254777070063694, | |
| "grad_norm": 15.811676025390625, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8208, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 5.414012738853503, | |
| "grad_norm": 15.85254955291748, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8272, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 5.573248407643312, | |
| "grad_norm": 17.296329498291016, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8312, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 5.732484076433121, | |
| "grad_norm": 16.399415969848633, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8444, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.89171974522293, | |
| "grad_norm": 16.311349868774414, | |
| "learning_rate": 1e-05, | |
| "loss": 0.8216, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 6.050955414012739, | |
| "grad_norm": 14.990595817565918, | |
| "learning_rate": 1e-05, | |
| "loss": 0.7495, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 6.210191082802548, | |
| "grad_norm": 13.810046195983887, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5933, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 6.369426751592357, | |
| "grad_norm": 12.30604362487793, | |
| "learning_rate": 1e-05, | |
| "loss": 0.624, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.369426751592357, | |
| "eval_loss": 2.2332684993743896, | |
| "eval_runtime": 146.206, | |
| "eval_samples_per_second": 4.514, | |
| "eval_steps_per_second": 0.287, | |
| "eval_wer": 111.13240084307338, | |
| "eval_wer_ortho": 110.44921875, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.528662420382165, | |
| "grad_norm": 17.28453826904297, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6305, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 6.687898089171974, | |
| "grad_norm": 13.922021865844727, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6513, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 6.8471337579617835, | |
| "grad_norm": 17.438993453979492, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6385, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 7.006369426751593, | |
| "grad_norm": 12.033106803894043, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6485, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 7.165605095541402, | |
| "grad_norm": 13.63040542602539, | |
| "learning_rate": 1e-05, | |
| "loss": 0.443, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 7.32484076433121, | |
| "grad_norm": 14.44725513458252, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4627, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 7.484076433121019, | |
| "grad_norm": 15.006793975830078, | |
| "learning_rate": 1e-05, | |
| "loss": 0.448, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 7.643312101910828, | |
| "grad_norm": 13.251697540283203, | |
| "learning_rate": 1e-05, | |
| "loss": 0.4475, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 7.802547770700637, | |
| "grad_norm": 13.863065719604492, | |
| "learning_rate": 1e-05, | |
| "loss": 0.5043, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 7.961783439490446, | |
| "grad_norm": 14.142547607421875, | |
| "learning_rate": 1e-05, | |
| "loss": 0.479, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 8.121019108280255, | |
| "grad_norm": 11.161627769470215, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3498, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 8.280254777070065, | |
| "grad_norm": 11.146525382995605, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3167, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 8.439490445859873, | |
| "grad_norm": 12.954503059387207, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3085, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 8.598726114649681, | |
| "grad_norm": 12.451419830322266, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3417, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 8.757961783439491, | |
| "grad_norm": 12.595786094665527, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3114, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 8.9171974522293, | |
| "grad_norm": 12.925387382507324, | |
| "learning_rate": 1e-05, | |
| "loss": 0.3434, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 9.07643312101911, | |
| "grad_norm": 11.76626205444336, | |
| "learning_rate": 1e-05, | |
| "loss": 0.2683, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 9.235668789808917, | |
| "grad_norm": 14.959568977355957, | |
| "learning_rate": 1e-05, | |
| "loss": 0.2238, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 9.394904458598726, | |
| "grad_norm": 9.14432144165039, | |
| "learning_rate": 1e-05, | |
| "loss": 0.2047, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 9.554140127388536, | |
| "grad_norm": 11.008196830749512, | |
| "learning_rate": 1e-05, | |
| "loss": 0.2135, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 9.554140127388536, | |
| "eval_loss": 2.437509775161743, | |
| "eval_runtime": 137.3182, | |
| "eval_samples_per_second": 4.806, | |
| "eval_steps_per_second": 0.306, | |
| "eval_wer": 100.44069745161907, | |
| "eval_wer_ortho": 101.62109374999999, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 9.713375796178344, | |
| "grad_norm": 13.016201972961426, | |
| "learning_rate": 1e-05, | |
| "loss": 0.2261, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 9.872611464968152, | |
| "grad_norm": 8.497207641601562, | |
| "learning_rate": 1e-05, | |
| "loss": 0.2192, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 10.031847133757962, | |
| "grad_norm": 7.207610130310059, | |
| "learning_rate": 1e-05, | |
| "loss": 0.2127, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 10.19108280254777, | |
| "grad_norm": 8.204691886901855, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1344, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 10.35031847133758, | |
| "grad_norm": 8.206758499145508, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1364, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 10.509554140127388, | |
| "grad_norm": 13.253020286560059, | |
| "learning_rate": 1e-05, | |
| "loss": 0.134, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 10.668789808917197, | |
| "grad_norm": 7.275868892669678, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1378, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 10.828025477707007, | |
| "grad_norm": 12.992132186889648, | |
| "learning_rate": 1e-05, | |
| "loss": 0.148, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 10.987261146496815, | |
| "grad_norm": 7.762783050537109, | |
| "learning_rate": 1e-05, | |
| "loss": 0.1472, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 11.146496815286625, | |
| "grad_norm": 7.062361717224121, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0902, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 11.305732484076433, | |
| "grad_norm": 8.7595853805542, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0882, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 11.464968152866241, | |
| "grad_norm": 6.6487603187561035, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0955, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 11.624203821656051, | |
| "grad_norm": 7.425637245178223, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0897, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 11.78343949044586, | |
| "grad_norm": 6.008634567260742, | |
| "learning_rate": 1e-05, | |
| "loss": 0.092, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 11.94267515923567, | |
| "grad_norm": 5.510532855987549, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0872, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 12.101910828025478, | |
| "grad_norm": 5.332154750823975, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0675, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 12.261146496815286, | |
| "grad_norm": 3.963823080062866, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0536, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 12.420382165605096, | |
| "grad_norm": 4.313061237335205, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0581, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 12.579617834394904, | |
| "grad_norm": 5.253906726837158, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0617, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 12.738853503184714, | |
| "grad_norm": 4.258543968200684, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0593, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 12.738853503184714, | |
| "eval_loss": 2.585127830505371, | |
| "eval_runtime": 142.1585, | |
| "eval_samples_per_second": 4.643, | |
| "eval_steps_per_second": 0.295, | |
| "eval_wer": 107.2810883310979, | |
| "eval_wer_ortho": 108.41796875, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 12.898089171974522, | |
| "grad_norm": 3.7591681480407715, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0569, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 13.05732484076433, | |
| "grad_norm": 3.663628578186035, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0531, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 13.21656050955414, | |
| "grad_norm": 4.044254779815674, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0377, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 13.375796178343949, | |
| "grad_norm": 2.5736799240112305, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0372, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 13.535031847133759, | |
| "grad_norm": 4.766959190368652, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0367, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 13.694267515923567, | |
| "grad_norm": 4.401877403259277, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0371, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 13.853503184713375, | |
| "grad_norm": 12.848492622375488, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0421, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 14.012738853503185, | |
| "grad_norm": 4.98983097076416, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0394, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 14.171974522292993, | |
| "grad_norm": 2.3211112022399902, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0268, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 14.331210191082803, | |
| "grad_norm": 3.7861273288726807, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0244, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 14.490445859872612, | |
| "grad_norm": 3.650113105773926, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0264, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 14.64968152866242, | |
| "grad_norm": 3.045107364654541, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0265, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 14.80891719745223, | |
| "grad_norm": 2.0573339462280273, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0262, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 14.968152866242038, | |
| "grad_norm": 2.7981200218200684, | |
| "learning_rate": 1e-05, | |
| "loss": 0.028, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 15.127388535031848, | |
| "grad_norm": 2.967073678970337, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0186, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 15.286624203821656, | |
| "grad_norm": 1.6514441967010498, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0173, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 15.445859872611464, | |
| "grad_norm": 1.3541009426116943, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0171, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 15.605095541401274, | |
| "grad_norm": 2.451819658279419, | |
| "learning_rate": 1e-05, | |
| "loss": 0.017, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 15.764331210191083, | |
| "grad_norm": 2.10622501373291, | |
| "learning_rate": 1e-05, | |
| "loss": 0.018, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 15.923566878980893, | |
| "grad_norm": 1.4680453538894653, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0169, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 15.923566878980893, | |
| "eval_loss": 2.752995252609253, | |
| "eval_runtime": 142.6116, | |
| "eval_samples_per_second": 4.628, | |
| "eval_steps_per_second": 0.295, | |
| "eval_wer": 102.24180877562752, | |
| "eval_wer_ortho": 103.02734375, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 16.0828025477707, | |
| "grad_norm": 8.074922561645508, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0176, | |
| "step": 2525 | |
| }, | |
| { | |
| "epoch": 16.24203821656051, | |
| "grad_norm": 1.393212914466858, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0129, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 16.401273885350317, | |
| "grad_norm": 0.9443471431732178, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0141, | |
| "step": 2575 | |
| }, | |
| { | |
| "epoch": 16.56050955414013, | |
| "grad_norm": 3.502882719039917, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0135, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 16.719745222929937, | |
| "grad_norm": 0.8746881484985352, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0116, | |
| "step": 2625 | |
| }, | |
| { | |
| "epoch": 16.878980891719745, | |
| "grad_norm": 2.561609983444214, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0121, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 17.038216560509554, | |
| "grad_norm": 0.8592103123664856, | |
| "learning_rate": 1e-05, | |
| "loss": 0.012, | |
| "step": 2675 | |
| }, | |
| { | |
| "epoch": 17.197452229299362, | |
| "grad_norm": 0.8304956555366516, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0088, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 17.356687898089174, | |
| "grad_norm": 0.9471040368080139, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0089, | |
| "step": 2725 | |
| }, | |
| { | |
| "epoch": 17.515923566878982, | |
| "grad_norm": 1.9444098472595215, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0093, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 17.67515923566879, | |
| "grad_norm": 0.5337225794792175, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0099, | |
| "step": 2775 | |
| }, | |
| { | |
| "epoch": 17.8343949044586, | |
| "grad_norm": 2.9927666187286377, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0091, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 17.993630573248407, | |
| "grad_norm": 0.6209370493888855, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0091, | |
| "step": 2825 | |
| }, | |
| { | |
| "epoch": 18.15286624203822, | |
| "grad_norm": 0.6852763891220093, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0064, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 18.312101910828027, | |
| "grad_norm": 0.5594916343688965, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0068, | |
| "step": 2875 | |
| }, | |
| { | |
| "epoch": 18.471337579617835, | |
| "grad_norm": 1.6789747476577759, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0064, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 18.630573248407643, | |
| "grad_norm": 3.211310625076294, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0067, | |
| "step": 2925 | |
| }, | |
| { | |
| "epoch": 18.78980891719745, | |
| "grad_norm": 0.5374391078948975, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0074, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 18.94904458598726, | |
| "grad_norm": 0.6938732266426086, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0062, | |
| "step": 2975 | |
| }, | |
| { | |
| "epoch": 19.10828025477707, | |
| "grad_norm": 0.3173889219760895, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0046, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 19.10828025477707, | |
| "eval_loss": 2.8555185794830322, | |
| "eval_runtime": 143.4984, | |
| "eval_samples_per_second": 4.599, | |
| "eval_steps_per_second": 0.293, | |
| "eval_wer": 99.32937344318835, | |
| "eval_wer_ortho": 100.60546875, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 19.26751592356688, | |
| "grad_norm": 1.3399635553359985, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0049, | |
| "step": 3025 | |
| }, | |
| { | |
| "epoch": 19.426751592356688, | |
| "grad_norm": 0.486775666475296, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0047, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 19.585987261146496, | |
| "grad_norm": 0.2713943421840668, | |
| "learning_rate": 1e-05, | |
| "loss": 0.004, | |
| "step": 3075 | |
| }, | |
| { | |
| "epoch": 19.745222929936304, | |
| "grad_norm": 0.41901519894599915, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0054, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 19.904458598726116, | |
| "grad_norm": 0.5328053832054138, | |
| "learning_rate": 1e-05, | |
| "loss": 0.008, | |
| "step": 3125 | |
| }, | |
| { | |
| "epoch": 20.063694267515924, | |
| "grad_norm": 0.2040318101644516, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0053, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 20.222929936305732, | |
| "grad_norm": 0.23346512019634247, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0037, | |
| "step": 3175 | |
| }, | |
| { | |
| "epoch": 20.38216560509554, | |
| "grad_norm": 0.3862437307834625, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0031, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 20.54140127388535, | |
| "grad_norm": 0.32829225063323975, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0037, | |
| "step": 3225 | |
| }, | |
| { | |
| "epoch": 20.70063694267516, | |
| "grad_norm": 0.22343598306179047, | |
| "learning_rate": 1e-05, | |
| "loss": 0.003, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 20.85987261146497, | |
| "grad_norm": 0.18372109532356262, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0029, | |
| "step": 3275 | |
| }, | |
| { | |
| "epoch": 21.019108280254777, | |
| "grad_norm": 0.1415814310312271, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0028, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 21.178343949044585, | |
| "grad_norm": 0.12073400616645813, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0021, | |
| "step": 3325 | |
| }, | |
| { | |
| "epoch": 21.337579617834393, | |
| "grad_norm": 0.14868922531604767, | |
| "learning_rate": 1e-05, | |
| "loss": 0.003, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 21.496815286624205, | |
| "grad_norm": 0.1244274452328682, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0021, | |
| "step": 3375 | |
| }, | |
| { | |
| "epoch": 21.656050955414013, | |
| "grad_norm": 0.18004997074604034, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0028, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 21.81528662420382, | |
| "grad_norm": 0.13512195646762848, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0023, | |
| "step": 3425 | |
| }, | |
| { | |
| "epoch": 21.97452229299363, | |
| "grad_norm": 0.1442134827375412, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0021, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 22.133757961783438, | |
| "grad_norm": 0.11067620664834976, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0017, | |
| "step": 3475 | |
| }, | |
| { | |
| "epoch": 22.29299363057325, | |
| "grad_norm": 0.49698173999786377, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0016, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 22.29299363057325, | |
| "eval_loss": 2.9374513626098633, | |
| "eval_runtime": 141.705, | |
| "eval_samples_per_second": 4.658, | |
| "eval_steps_per_second": 0.296, | |
| "eval_wer": 100.30657214025676, | |
| "eval_wer_ortho": 101.46484375, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 22.452229299363058, | |
| "grad_norm": 0.12976475059986115, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0018, | |
| "step": 3525 | |
| }, | |
| { | |
| "epoch": 22.611464968152866, | |
| "grad_norm": 0.13230012357234955, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0019, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 22.770700636942674, | |
| "grad_norm": 0.12926578521728516, | |
| "learning_rate": 1e-05, | |
| "loss": 0.003, | |
| "step": 3575 | |
| }, | |
| { | |
| "epoch": 22.929936305732483, | |
| "grad_norm": 0.11274802684783936, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0017, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 23.089171974522294, | |
| "grad_norm": 0.07794158905744553, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0015, | |
| "step": 3625 | |
| }, | |
| { | |
| "epoch": 23.248407643312103, | |
| "grad_norm": 0.08103267848491669, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0014, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 23.40764331210191, | |
| "grad_norm": 0.08904325217008591, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0014, | |
| "step": 3675 | |
| }, | |
| { | |
| "epoch": 23.56687898089172, | |
| "grad_norm": 0.08841922879219055, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0029, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 23.726114649681527, | |
| "grad_norm": 0.09364067763090134, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0015, | |
| "step": 3725 | |
| }, | |
| { | |
| "epoch": 23.88535031847134, | |
| "grad_norm": 0.09942711889743805, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0014, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 24.044585987261147, | |
| "grad_norm": 0.07535886764526367, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0013, | |
| "step": 3775 | |
| }, | |
| { | |
| "epoch": 24.203821656050955, | |
| "grad_norm": 0.09200573712587357, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0012, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 24.363057324840764, | |
| "grad_norm": 0.07769662886857986, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0012, | |
| "step": 3825 | |
| }, | |
| { | |
| "epoch": 24.522292993630572, | |
| "grad_norm": 0.06928899139165878, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0012, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 24.681528662420384, | |
| "grad_norm": 0.07362724840641022, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0012, | |
| "step": 3875 | |
| }, | |
| { | |
| "epoch": 24.840764331210192, | |
| "grad_norm": 0.06966041028499603, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0012, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "grad_norm": 0.09524974972009659, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0013, | |
| "step": 3925 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 3925, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 25, | |
| "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": 1.543604097024e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |