| { | |
| "best_global_step": 20000, | |
| "best_metric": 0.06115037298782882, | |
| "best_model_checkpoint": "/home/cluster-dgxa100/slp01/bagas-fine-tune-whisper/whisper-large-v2-javanese-openslr-v2/checkpoint-20000", | |
| "epoch": 8.0, | |
| "eval_steps": 1000, | |
| "global_step": 20000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 19.06084442138672, | |
| "learning_rate": 1.97e-06, | |
| "loss": 1.8907, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 16.973052978515625, | |
| "learning_rate": 3.97e-06, | |
| "loss": 0.6026, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 8.261700630187988, | |
| "learning_rate": 5.9700000000000004e-06, | |
| "loss": 0.5317, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 7.87921667098999, | |
| "learning_rate": 7.970000000000002e-06, | |
| "loss": 0.491, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 7.55085563659668, | |
| "learning_rate": 9.970000000000001e-06, | |
| "loss": 0.4821, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "eval_loss": 0.3569938838481903, | |
| "eval_runtime": 880.4774, | |
| "eval_samples_per_second": 1.42, | |
| "eval_steps_per_second": 0.71, | |
| "eval_wer": 0.22683549273655282, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 5.795778751373291, | |
| "learning_rate": 1.1970000000000002e-05, | |
| "loss": 0.4608, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 8.08930492401123, | |
| "learning_rate": 1.3970000000000002e-05, | |
| "loss": 0.4761, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 7.31475305557251, | |
| "learning_rate": 1.597e-05, | |
| "loss": 0.4942, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 7.746321201324463, | |
| "learning_rate": 1.796e-05, | |
| "loss": 0.4758, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 5.482341289520264, | |
| "learning_rate": 1.9960000000000002e-05, | |
| "loss": 0.4761, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_loss": 0.34740468859672546, | |
| "eval_runtime": 901.2547, | |
| "eval_samples_per_second": 1.387, | |
| "eval_steps_per_second": 0.693, | |
| "eval_wer": 0.2396937573616019, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 7.196951866149902, | |
| "learning_rate": 1.9782222222222226e-05, | |
| "loss": 0.4924, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 6.221120834350586, | |
| "learning_rate": 1.9560000000000002e-05, | |
| "loss": 0.4779, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 6.380108833312988, | |
| "learning_rate": 1.933777777777778e-05, | |
| "loss": 0.3806, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 5.491684913635254, | |
| "learning_rate": 1.9115555555555555e-05, | |
| "loss": 0.3465, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 4.4581379890441895, | |
| "learning_rate": 1.8893333333333334e-05, | |
| "loss": 0.3099, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "eval_loss": 0.30754217505455017, | |
| "eval_runtime": 891.4205, | |
| "eval_samples_per_second": 1.402, | |
| "eval_steps_per_second": 0.701, | |
| "eval_wer": 0.20259128386336867, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 7.791854381561279, | |
| "learning_rate": 1.8671111111111114e-05, | |
| "loss": 0.3275, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.3599999999999999, | |
| "grad_norm": 5.212815761566162, | |
| "learning_rate": 1.844888888888889e-05, | |
| "loss": 0.3296, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 9.706353187561035, | |
| "learning_rate": 1.822666666666667e-05, | |
| "loss": 0.3142, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 5.222635269165039, | |
| "learning_rate": 1.8004444444444446e-05, | |
| "loss": 0.3097, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 3.670520067214966, | |
| "learning_rate": 1.7782222222222222e-05, | |
| "loss": 0.3108, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_loss": 0.26660874485969543, | |
| "eval_runtime": 896.6083, | |
| "eval_samples_per_second": 1.394, | |
| "eval_steps_per_second": 0.697, | |
| "eval_wer": 0.19709462112288967, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 5.417144298553467, | |
| "learning_rate": 1.756e-05, | |
| "loss": 0.3096, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 5.865799427032471, | |
| "learning_rate": 1.733777777777778e-05, | |
| "loss": 0.3266, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.8399999999999999, | |
| "grad_norm": 3.2357375621795654, | |
| "learning_rate": 1.7115555555555557e-05, | |
| "loss": 0.3134, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 6.072234630584717, | |
| "learning_rate": 1.6893333333333336e-05, | |
| "loss": 0.289, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 3.143366813659668, | |
| "learning_rate": 1.6671111111111113e-05, | |
| "loss": 0.3131, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.21725231409072876, | |
| "eval_runtime": 885.5009, | |
| "eval_samples_per_second": 1.412, | |
| "eval_steps_per_second": 0.706, | |
| "eval_wer": 0.15312131919905772, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 4.550586700439453, | |
| "learning_rate": 1.644888888888889e-05, | |
| "loss": 0.1525, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 3.604379892349243, | |
| "learning_rate": 1.6226666666666668e-05, | |
| "loss": 0.1656, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 3.664323568344116, | |
| "learning_rate": 1.6004444444444444e-05, | |
| "loss": 0.1743, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 2.5190961360931396, | |
| "learning_rate": 1.5782222222222224e-05, | |
| "loss": 0.175, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 3.1211793422698975, | |
| "learning_rate": 1.556e-05, | |
| "loss": 0.1796, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_loss": 0.2026291936635971, | |
| "eval_runtime": 894.7957, | |
| "eval_samples_per_second": 1.397, | |
| "eval_steps_per_second": 0.698, | |
| "eval_wer": 0.13839811542991756, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 3.203092336654663, | |
| "learning_rate": 1.533777777777778e-05, | |
| "loss": 0.1721, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 1.1959407329559326, | |
| "learning_rate": 1.5115555555555557e-05, | |
| "loss": 0.1722, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 4.8345160484313965, | |
| "learning_rate": 1.4893333333333335e-05, | |
| "loss": 0.17, | |
| "step": 6600 | |
| }, | |
| { | |
| "epoch": 2.7199999999999998, | |
| "grad_norm": 5.333383560180664, | |
| "learning_rate": 1.4671111111111111e-05, | |
| "loss": 0.1751, | |
| "step": 6800 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 1.5021018981933594, | |
| "learning_rate": 1.444888888888889e-05, | |
| "loss": 0.17, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "eval_loss": 0.19221611320972443, | |
| "eval_runtime": 891.3109, | |
| "eval_samples_per_second": 1.402, | |
| "eval_steps_per_second": 0.701, | |
| "eval_wer": 0.13780918727915195, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 3.0855863094329834, | |
| "learning_rate": 1.4226666666666669e-05, | |
| "loss": 0.1672, | |
| "step": 7200 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 4.673288822174072, | |
| "learning_rate": 1.4004444444444445e-05, | |
| "loss": 0.1724, | |
| "step": 7400 | |
| }, | |
| { | |
| "epoch": 3.04, | |
| "grad_norm": 2.8060922622680664, | |
| "learning_rate": 1.3782222222222223e-05, | |
| "loss": 0.1262, | |
| "step": 7600 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 1.5213590860366821, | |
| "learning_rate": 1.3560000000000002e-05, | |
| "loss": 0.094, | |
| "step": 7800 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 5.938966751098633, | |
| "learning_rate": 1.3337777777777778e-05, | |
| "loss": 0.0995, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "eval_loss": 0.17917467653751373, | |
| "eval_runtime": 900.0369, | |
| "eval_samples_per_second": 1.389, | |
| "eval_steps_per_second": 0.694, | |
| "eval_wer": 0.12063211621515509, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 3.2800000000000002, | |
| "grad_norm": 3.5296432971954346, | |
| "learning_rate": 1.3115555555555556e-05, | |
| "loss": 0.1002, | |
| "step": 8200 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 5.334218502044678, | |
| "learning_rate": 1.2893333333333336e-05, | |
| "loss": 0.1061, | |
| "step": 8400 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 5.724509239196777, | |
| "learning_rate": 1.2671111111111112e-05, | |
| "loss": 0.1012, | |
| "step": 8600 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 2.5498409271240234, | |
| "learning_rate": 1.244888888888889e-05, | |
| "loss": 0.099, | |
| "step": 8800 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "grad_norm": 6.009134769439697, | |
| "learning_rate": 1.2226666666666666e-05, | |
| "loss": 0.0972, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.6, | |
| "eval_loss": 0.16701579093933105, | |
| "eval_runtime": 899.6835, | |
| "eval_samples_per_second": 1.389, | |
| "eval_steps_per_second": 0.695, | |
| "eval_wer": 0.11493914409108755, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 3.68, | |
| "grad_norm": 2.472468137741089, | |
| "learning_rate": 1.2004444444444445e-05, | |
| "loss": 0.0933, | |
| "step": 9200 | |
| }, | |
| { | |
| "epoch": 3.76, | |
| "grad_norm": 2.989001750946045, | |
| "learning_rate": 1.1782222222222223e-05, | |
| "loss": 0.0999, | |
| "step": 9400 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 3.3118479251861572, | |
| "learning_rate": 1.156e-05, | |
| "loss": 0.0949, | |
| "step": 9600 | |
| }, | |
| { | |
| "epoch": 3.92, | |
| "grad_norm": 4.284938812255859, | |
| "learning_rate": 1.1337777777777779e-05, | |
| "loss": 0.0964, | |
| "step": 9800 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.4380695819854736, | |
| "learning_rate": 1.1115555555555557e-05, | |
| "loss": 0.097, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.1545259952545166, | |
| "eval_runtime": 903.5859, | |
| "eval_samples_per_second": 1.383, | |
| "eval_steps_per_second": 0.692, | |
| "eval_wer": 0.10963879073419709, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 4.08, | |
| "grad_norm": 1.92518949508667, | |
| "learning_rate": 1.0893333333333333e-05, | |
| "loss": 0.0524, | |
| "step": 10200 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 2.2278802394866943, | |
| "learning_rate": 1.0671111111111112e-05, | |
| "loss": 0.0523, | |
| "step": 10400 | |
| }, | |
| { | |
| "epoch": 4.24, | |
| "grad_norm": 0.5611541271209717, | |
| "learning_rate": 1.044888888888889e-05, | |
| "loss": 0.0521, | |
| "step": 10600 | |
| }, | |
| { | |
| "epoch": 4.32, | |
| "grad_norm": 1.7205854654312134, | |
| "learning_rate": 1.0226666666666666e-05, | |
| "loss": 0.0499, | |
| "step": 10800 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "grad_norm": 2.563293218612671, | |
| "learning_rate": 1.0004444444444446e-05, | |
| "loss": 0.0553, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.4, | |
| "eval_loss": 0.1574954092502594, | |
| "eval_runtime": 893.7485, | |
| "eval_samples_per_second": 1.399, | |
| "eval_steps_per_second": 0.699, | |
| "eval_wer": 0.10296427169218689, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 2.2215042114257812, | |
| "learning_rate": 9.783333333333335e-06, | |
| "loss": 0.0537, | |
| "step": 11200 | |
| }, | |
| { | |
| "epoch": 4.5600000000000005, | |
| "grad_norm": 1.360498070716858, | |
| "learning_rate": 9.561111111111113e-06, | |
| "loss": 0.0588, | |
| "step": 11400 | |
| }, | |
| { | |
| "epoch": 4.64, | |
| "grad_norm": 4.086482048034668, | |
| "learning_rate": 9.340000000000002e-06, | |
| "loss": 0.0537, | |
| "step": 11600 | |
| }, | |
| { | |
| "epoch": 4.72, | |
| "grad_norm": 2.1945817470550537, | |
| "learning_rate": 9.117777777777778e-06, | |
| "loss": 0.0599, | |
| "step": 11800 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 2.7617948055267334, | |
| "learning_rate": 8.895555555555556e-06, | |
| "loss": 0.0526, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "eval_loss": 0.14308780431747437, | |
| "eval_runtime": 905.6641, | |
| "eval_samples_per_second": 1.38, | |
| "eval_steps_per_second": 0.69, | |
| "eval_wer": 0.08902630545740087, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 4.88, | |
| "grad_norm": 0.3309612572193146, | |
| "learning_rate": 8.673333333333334e-06, | |
| "loss": 0.0527, | |
| "step": 12200 | |
| }, | |
| { | |
| "epoch": 4.96, | |
| "grad_norm": 0.3702127933502197, | |
| "learning_rate": 8.451111111111112e-06, | |
| "loss": 0.0574, | |
| "step": 12400 | |
| }, | |
| { | |
| "epoch": 5.04, | |
| "grad_norm": 1.9310526847839355, | |
| "learning_rate": 8.22888888888889e-06, | |
| "loss": 0.0385, | |
| "step": 12600 | |
| }, | |
| { | |
| "epoch": 5.12, | |
| "grad_norm": 0.8932788372039795, | |
| "learning_rate": 8.006666666666667e-06, | |
| "loss": 0.0369, | |
| "step": 12800 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "grad_norm": 2.246431589126587, | |
| "learning_rate": 7.784444444444445e-06, | |
| "loss": 0.0299, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.2, | |
| "eval_loss": 0.13719478249549866, | |
| "eval_runtime": 896.1487, | |
| "eval_samples_per_second": 1.395, | |
| "eval_steps_per_second": 0.697, | |
| "eval_wer": 0.08333333333333333, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 5.28, | |
| "grad_norm": 2.8669216632843018, | |
| "learning_rate": 7.562222222222223e-06, | |
| "loss": 0.0276, | |
| "step": 13200 | |
| }, | |
| { | |
| "epoch": 5.36, | |
| "grad_norm": 0.38674354553222656, | |
| "learning_rate": 7.341111111111112e-06, | |
| "loss": 0.0334, | |
| "step": 13400 | |
| }, | |
| { | |
| "epoch": 5.44, | |
| "grad_norm": 4.218255996704102, | |
| "learning_rate": 7.11888888888889e-06, | |
| "loss": 0.0253, | |
| "step": 13600 | |
| }, | |
| { | |
| "epoch": 5.52, | |
| "grad_norm": 2.424956798553467, | |
| "learning_rate": 6.896666666666667e-06, | |
| "loss": 0.0287, | |
| "step": 13800 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 1.3083245754241943, | |
| "learning_rate": 6.674444444444445e-06, | |
| "loss": 0.0311, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "eval_loss": 0.12584801018238068, | |
| "eval_runtime": 892.859, | |
| "eval_samples_per_second": 1.4, | |
| "eval_steps_per_second": 0.7, | |
| "eval_wer": 0.07803297997644287, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 5.68, | |
| "grad_norm": 1.7961013317108154, | |
| "learning_rate": 6.452222222222223e-06, | |
| "loss": 0.0291, | |
| "step": 14200 | |
| }, | |
| { | |
| "epoch": 5.76, | |
| "grad_norm": 0.16184474527835846, | |
| "learning_rate": 6.2300000000000005e-06, | |
| "loss": 0.0287, | |
| "step": 14400 | |
| }, | |
| { | |
| "epoch": 5.84, | |
| "grad_norm": 0.13447080552577972, | |
| "learning_rate": 6.007777777777778e-06, | |
| "loss": 0.0303, | |
| "step": 14600 | |
| }, | |
| { | |
| "epoch": 5.92, | |
| "grad_norm": 2.1226162910461426, | |
| "learning_rate": 5.785555555555556e-06, | |
| "loss": 0.0261, | |
| "step": 14800 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.8055452108383179, | |
| "learning_rate": 5.563333333333334e-06, | |
| "loss": 0.0295, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.12012948095798492, | |
| "eval_runtime": 888.9428, | |
| "eval_samples_per_second": 1.406, | |
| "eval_steps_per_second": 0.703, | |
| "eval_wer": 0.07253631723596388, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 6.08, | |
| "grad_norm": 0.1165083572268486, | |
| "learning_rate": 5.341111111111111e-06, | |
| "loss": 0.0143, | |
| "step": 15200 | |
| }, | |
| { | |
| "epoch": 6.16, | |
| "grad_norm": 0.12827950716018677, | |
| "learning_rate": 5.118888888888889e-06, | |
| "loss": 0.0169, | |
| "step": 15400 | |
| }, | |
| { | |
| "epoch": 6.24, | |
| "grad_norm": 3.075695276260376, | |
| "learning_rate": 4.896666666666667e-06, | |
| "loss": 0.0144, | |
| "step": 15600 | |
| }, | |
| { | |
| "epoch": 6.32, | |
| "grad_norm": 0.4469108581542969, | |
| "learning_rate": 4.6744444444444445e-06, | |
| "loss": 0.0124, | |
| "step": 15800 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "grad_norm": 0.264642596244812, | |
| "learning_rate": 4.452222222222223e-06, | |
| "loss": 0.0151, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.4, | |
| "eval_loss": 0.12285350263118744, | |
| "eval_runtime": 883.7262, | |
| "eval_samples_per_second": 1.414, | |
| "eval_steps_per_second": 0.707, | |
| "eval_wer": 0.07420494699646643, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 6.48, | |
| "grad_norm": 3.0525002479553223, | |
| "learning_rate": 4.23e-06, | |
| "loss": 0.0122, | |
| "step": 16200 | |
| }, | |
| { | |
| "epoch": 6.5600000000000005, | |
| "grad_norm": 1.7435777187347412, | |
| "learning_rate": 4.007777777777778e-06, | |
| "loss": 0.0134, | |
| "step": 16400 | |
| }, | |
| { | |
| "epoch": 6.64, | |
| "grad_norm": 0.03412042185664177, | |
| "learning_rate": 3.785555555555556e-06, | |
| "loss": 0.0146, | |
| "step": 16600 | |
| }, | |
| { | |
| "epoch": 6.72, | |
| "grad_norm": 4.467521667480469, | |
| "learning_rate": 3.5633333333333337e-06, | |
| "loss": 0.0139, | |
| "step": 16800 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "grad_norm": 1.6740899085998535, | |
| "learning_rate": 3.3411111111111115e-06, | |
| "loss": 0.0163, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 6.8, | |
| "eval_loss": 0.1136574074625969, | |
| "eval_runtime": 890.0494, | |
| "eval_samples_per_second": 1.404, | |
| "eval_steps_per_second": 0.702, | |
| "eval_wer": 0.06576364350215941, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 6.88, | |
| "grad_norm": 0.25668439269065857, | |
| "learning_rate": 3.118888888888889e-06, | |
| "loss": 0.0143, | |
| "step": 17200 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "grad_norm": 1.1925643682479858, | |
| "learning_rate": 2.896666666666667e-06, | |
| "loss": 0.0132, | |
| "step": 17400 | |
| }, | |
| { | |
| "epoch": 7.04, | |
| "grad_norm": 0.45678913593292236, | |
| "learning_rate": 2.6744444444444446e-06, | |
| "loss": 0.0103, | |
| "step": 17600 | |
| }, | |
| { | |
| "epoch": 7.12, | |
| "grad_norm": 0.25028079748153687, | |
| "learning_rate": 2.4522222222222224e-06, | |
| "loss": 0.0089, | |
| "step": 17800 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "grad_norm": 0.04460795596241951, | |
| "learning_rate": 2.2300000000000002e-06, | |
| "loss": 0.0082, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.2, | |
| "eval_loss": 0.1142345443367958, | |
| "eval_runtime": 885.5644, | |
| "eval_samples_per_second": 1.412, | |
| "eval_steps_per_second": 0.706, | |
| "eval_wer": 0.06389870435806831, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 7.28, | |
| "grad_norm": 6.106534957885742, | |
| "learning_rate": 2.007777777777778e-06, | |
| "loss": 0.0063, | |
| "step": 18200 | |
| }, | |
| { | |
| "epoch": 7.36, | |
| "grad_norm": 0.5640923380851746, | |
| "learning_rate": 1.7855555555555557e-06, | |
| "loss": 0.0088, | |
| "step": 18400 | |
| }, | |
| { | |
| "epoch": 7.44, | |
| "grad_norm": 0.09705500304698944, | |
| "learning_rate": 1.5633333333333333e-06, | |
| "loss": 0.0091, | |
| "step": 18600 | |
| }, | |
| { | |
| "epoch": 7.52, | |
| "grad_norm": 0.042120128870010376, | |
| "learning_rate": 1.3411111111111112e-06, | |
| "loss": 0.0066, | |
| "step": 18800 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 0.024996856227517128, | |
| "learning_rate": 1.118888888888889e-06, | |
| "loss": 0.0092, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "eval_loss": 0.11208222806453705, | |
| "eval_runtime": 886.2816, | |
| "eval_samples_per_second": 1.41, | |
| "eval_steps_per_second": 0.705, | |
| "eval_wer": 0.0627208480565371, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 7.68, | |
| "grad_norm": 0.38041412830352783, | |
| "learning_rate": 8.966666666666668e-07, | |
| "loss": 0.0068, | |
| "step": 19200 | |
| }, | |
| { | |
| "epoch": 7.76, | |
| "grad_norm": 3.13415265083313, | |
| "learning_rate": 6.744444444444446e-07, | |
| "loss": 0.0062, | |
| "step": 19400 | |
| }, | |
| { | |
| "epoch": 7.84, | |
| "grad_norm": 0.7191887497901917, | |
| "learning_rate": 4.5222222222222224e-07, | |
| "loss": 0.0067, | |
| "step": 19600 | |
| }, | |
| { | |
| "epoch": 7.92, | |
| "grad_norm": 0.09878556430339813, | |
| "learning_rate": 2.3000000000000002e-07, | |
| "loss": 0.0077, | |
| "step": 19800 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.007985732518136501, | |
| "learning_rate": 7.777777777777778e-09, | |
| "loss": 0.006, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.11124598234891891, | |
| "eval_runtime": 889.6592, | |
| "eval_samples_per_second": 1.405, | |
| "eval_steps_per_second": 0.703, | |
| "eval_wer": 0.06115037298782882, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 20000, | |
| "total_flos": 1.698366962958336e+20, | |
| "train_loss": 0.16200368287563324, | |
| "train_runtime": 40923.5162, | |
| "train_samples_per_second": 1.955, | |
| "train_steps_per_second": 0.489 | |
| } | |
| ], | |
| "logging_steps": 200, | |
| "max_steps": 20000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 8, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.698366962958336e+20, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |