| { | |
| "best_metric": 108.23912626940026, | |
| "best_model_checkpoint": "./whisper-small-ha-v2/checkpoint-1000", | |
| "epoch": 12.738853503184714, | |
| "eval_steps": 500, | |
| "global_step": 2000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1592356687898089, | |
| "grad_norm": 19.899324417114258, | |
| "learning_rate": 0.00021, | |
| "loss": 4.2616, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3184713375796178, | |
| "grad_norm": 14.790164947509766, | |
| "learning_rate": 0.00046, | |
| "loss": 2.8345, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.47770700636942676, | |
| "grad_norm": 56.03984451293945, | |
| "learning_rate": 0.0005, | |
| "loss": 3.4138, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6369426751592356, | |
| "grad_norm": 10.28797721862793, | |
| "learning_rate": 0.0005, | |
| "loss": 3.5843, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.7961783439490446, | |
| "grad_norm": 8.962028503417969, | |
| "learning_rate": 0.0005, | |
| "loss": 3.5061, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.9554140127388535, | |
| "grad_norm": 7.9880499839782715, | |
| "learning_rate": 0.0005, | |
| "loss": 3.4731, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.1146496815286624, | |
| "grad_norm": 8.54383659362793, | |
| "learning_rate": 0.0005, | |
| "loss": 2.8313, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.2738853503184713, | |
| "grad_norm": 8.352940559387207, | |
| "learning_rate": 0.0005, | |
| "loss": 2.7241, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.4331210191082802, | |
| "grad_norm": 10.003473281860352, | |
| "learning_rate": 0.0005, | |
| "loss": 2.8259, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.5923566878980893, | |
| "grad_norm": 7.971159934997559, | |
| "learning_rate": 0.0005, | |
| "loss": 2.865, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.7515923566878981, | |
| "grad_norm": 7.780789375305176, | |
| "learning_rate": 0.0005, | |
| "loss": 2.813, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.910828025477707, | |
| "grad_norm": 7.645856857299805, | |
| "learning_rate": 0.0005, | |
| "loss": 2.911, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.070063694267516, | |
| "grad_norm": 7.892724514007568, | |
| "learning_rate": 0.0005, | |
| "loss": 2.4582, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.229299363057325, | |
| "grad_norm": 7.8490681648254395, | |
| "learning_rate": 0.0005, | |
| "loss": 1.9971, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.388535031847134, | |
| "grad_norm": 8.876250267028809, | |
| "learning_rate": 0.0005, | |
| "loss": 2.1415, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 2.5477707006369426, | |
| "grad_norm": 9.139019966125488, | |
| "learning_rate": 0.0005, | |
| "loss": 2.2764, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.7070063694267517, | |
| "grad_norm": 9.341145515441895, | |
| "learning_rate": 0.0005, | |
| "loss": 2.3182, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.8662420382165603, | |
| "grad_norm": 7.993480682373047, | |
| "learning_rate": 0.0005, | |
| "loss": 2.3467, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.0254777070063694, | |
| "grad_norm": 7.322677135467529, | |
| "learning_rate": 0.0005, | |
| "loss": 2.1176, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 3.1847133757961785, | |
| "grad_norm": 7.923095226287842, | |
| "learning_rate": 0.0005, | |
| "loss": 1.5509, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.1847133757961785, | |
| "eval_loss": 3.9122631549835205, | |
| "eval_runtime": 241.8174, | |
| "eval_samples_per_second": 2.729, | |
| "eval_steps_per_second": 0.174, | |
| "eval_wer": 134.7576164016095, | |
| "eval_wer_ortho": 136.11328125, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 3.343949044585987, | |
| "grad_norm": 7.589674472808838, | |
| "learning_rate": 0.0005, | |
| "loss": 1.6755, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 3.5031847133757963, | |
| "grad_norm": 7.347910404205322, | |
| "learning_rate": 0.0005, | |
| "loss": 1.7794, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 3.662420382165605, | |
| "grad_norm": 7.744343280792236, | |
| "learning_rate": 0.0005, | |
| "loss": 1.8251, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 3.821656050955414, | |
| "grad_norm": 7.262452125549316, | |
| "learning_rate": 0.0005, | |
| "loss": 1.8631, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.980891719745223, | |
| "grad_norm": 7.604316711425781, | |
| "learning_rate": 0.0005, | |
| "loss": 1.9072, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 4.140127388535032, | |
| "grad_norm": 6.618544578552246, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3487, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 4.2993630573248405, | |
| "grad_norm": 7.4200239181518555, | |
| "learning_rate": 0.0005, | |
| "loss": 1.347, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 4.45859872611465, | |
| "grad_norm": 7.097377777099609, | |
| "learning_rate": 0.0005, | |
| "loss": 1.4564, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 4.617834394904459, | |
| "grad_norm": 7.271367073059082, | |
| "learning_rate": 0.0005, | |
| "loss": 1.5359, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 4.777070063694268, | |
| "grad_norm": 7.738811492919922, | |
| "learning_rate": 0.0005, | |
| "loss": 1.564, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 4.936305732484076, | |
| "grad_norm": 7.1541361808776855, | |
| "learning_rate": 0.0005, | |
| "loss": 1.6548, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 5.095541401273885, | |
| "grad_norm": 6.448052883148193, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3242, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 5.254777070063694, | |
| "grad_norm": 6.672939300537109, | |
| "learning_rate": 0.0005, | |
| "loss": 1.2112, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 5.414012738853503, | |
| "grad_norm": 6.929521083831787, | |
| "learning_rate": 0.0005, | |
| "loss": 1.246, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 5.573248407643312, | |
| "grad_norm": 7.8832197189331055, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3477, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 5.732484076433121, | |
| "grad_norm": 6.741547107696533, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3434, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 5.89171974522293, | |
| "grad_norm": 6.631527423858643, | |
| "learning_rate": 0.0005, | |
| "loss": 1.4936, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 6.050955414012739, | |
| "grad_norm": 6.012191295623779, | |
| "learning_rate": 0.0005, | |
| "loss": 1.3185, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 6.210191082802548, | |
| "grad_norm": 6.420320987701416, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0981, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 6.369426751592357, | |
| "grad_norm": 6.100344181060791, | |
| "learning_rate": 0.0005, | |
| "loss": 1.1254, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.369426751592357, | |
| "eval_loss": 4.613819599151611, | |
| "eval_runtime": 232.7229, | |
| "eval_samples_per_second": 2.836, | |
| "eval_steps_per_second": 0.18, | |
| "eval_wer": 108.23912626940026, | |
| "eval_wer_ortho": 104.8828125, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 6.528662420382165, | |
| "grad_norm": 7.62544584274292, | |
| "learning_rate": 0.0005, | |
| "loss": 1.2221, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 6.687898089171974, | |
| "grad_norm": 6.837728977203369, | |
| "learning_rate": 0.0005, | |
| "loss": 1.2167, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 6.8471337579617835, | |
| "grad_norm": 7.663355827331543, | |
| "learning_rate": 0.0005, | |
| "loss": 1.2218, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 7.006369426751593, | |
| "grad_norm": 6.275888919830322, | |
| "learning_rate": 0.0005, | |
| "loss": 1.2511, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 7.165605095541402, | |
| "grad_norm": 5.376118183135986, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0101, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 7.32484076433121, | |
| "grad_norm": 12.314349174499512, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0479, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 7.484076433121019, | |
| "grad_norm": 6.386078834533691, | |
| "learning_rate": 0.0005, | |
| "loss": 1.075, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 7.643312101910828, | |
| "grad_norm": 6.5437912940979, | |
| "learning_rate": 0.0005, | |
| "loss": 1.1115, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 7.802547770700637, | |
| "grad_norm": 6.102975368499756, | |
| "learning_rate": 0.0005, | |
| "loss": 1.1557, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 7.961783439490446, | |
| "grad_norm": 5.722137451171875, | |
| "learning_rate": 0.0005, | |
| "loss": 1.1659, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 8.121019108280255, | |
| "grad_norm": 6.168821811676025, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0219, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 8.280254777070065, | |
| "grad_norm": 5.716577053070068, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9961, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 8.439490445859873, | |
| "grad_norm": 5.789382457733154, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0159, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 8.598726114649681, | |
| "grad_norm": 6.052881717681885, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0636, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 8.757961783439491, | |
| "grad_norm": 5.608780384063721, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0547, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 8.9171974522293, | |
| "grad_norm": 6.0210747718811035, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0525, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 9.07643312101911, | |
| "grad_norm": 5.303377151489258, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9966, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 9.235668789808917, | |
| "grad_norm": 5.898992538452148, | |
| "learning_rate": 0.0005, | |
| "loss": 0.913, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 9.394904458598726, | |
| "grad_norm": 5.273637771606445, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9564, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 9.554140127388536, | |
| "grad_norm": 5.149616718292236, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9898, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 9.554140127388536, | |
| "eval_loss": 5.00140905380249, | |
| "eval_runtime": 243.3495, | |
| "eval_samples_per_second": 2.712, | |
| "eval_steps_per_second": 0.173, | |
| "eval_wer": 123.87430542249473, | |
| "eval_wer_ortho": 125.99609375, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 9.713375796178344, | |
| "grad_norm": 5.37247371673584, | |
| "learning_rate": 0.0005, | |
| "loss": 1.043, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 9.872611464968152, | |
| "grad_norm": 5.714049816131592, | |
| "learning_rate": 0.0005, | |
| "loss": 1.0139, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 10.031847133757962, | |
| "grad_norm": 5.447414398193359, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9852, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 10.19108280254777, | |
| "grad_norm": 4.800140380859375, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8618, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 10.35031847133758, | |
| "grad_norm": 4.940687656402588, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9099, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 10.509554140127388, | |
| "grad_norm": 5.976550579071045, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9144, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 10.668789808917197, | |
| "grad_norm": 5.915425777435303, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9963, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 10.828025477707007, | |
| "grad_norm": 6.186098575592041, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9809, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 10.987261146496815, | |
| "grad_norm": 6.102351665496826, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9992, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 11.146496815286625, | |
| "grad_norm": 5.012561798095703, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8408, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 11.305732484076433, | |
| "grad_norm": 4.676059246063232, | |
| "learning_rate": 0.0005, | |
| "loss": 0.838, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 11.464968152866241, | |
| "grad_norm": 4.5027689933776855, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8774, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 11.624203821656051, | |
| "grad_norm": 5.583326816558838, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9229, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 11.78343949044586, | |
| "grad_norm": 5.260093688964844, | |
| "learning_rate": 0.0005, | |
| "loss": 0.9436, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 11.94267515923567, | |
| "grad_norm": 4.777396202087402, | |
| "learning_rate": 0.0005, | |
| "loss": 0.949, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 12.101910828025478, | |
| "grad_norm": 4.361011028289795, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8302, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 12.261146496815286, | |
| "grad_norm": 4.748295783996582, | |
| "learning_rate": 0.0005, | |
| "loss": 0.843, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 12.420382165605096, | |
| "grad_norm": 5.090404510498047, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8528, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 12.579617834394904, | |
| "grad_norm": 5.081445217132568, | |
| "learning_rate": 0.0005, | |
| "loss": 0.8623, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 12.738853503184714, | |
| "grad_norm": 5.345047950744629, | |
| "learning_rate": 0.0005, | |
| "loss": 0.892, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 12.738853503184714, | |
| "eval_loss": 4.75683069229126, | |
| "eval_runtime": 241.3599, | |
| "eval_samples_per_second": 2.735, | |
| "eval_steps_per_second": 0.174, | |
| "eval_wer": 114.79210576738839, | |
| "eval_wer_ortho": 118.00781250000001, | |
| "step": 2000 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 2355, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 15, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.22088071102464e+18, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |