| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 200.0, | |
| "global_step": 6400, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 3.12, | |
| "learning_rate": 1.3439999999999998e-05, | |
| "loss": 22.3442, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "learning_rate": 2.7439999999999998e-05, | |
| "loss": 5.9373, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 9.38, | |
| "learning_rate": 4.1439999999999996e-05, | |
| "loss": 3.7179, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 5.544e-05, | |
| "loss": 3.0789, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 15.62, | |
| "learning_rate": 6.944e-05, | |
| "loss": 2.9287, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 15.62, | |
| "eval_loss": 2.877366304397583, | |
| "eval_runtime": 13.8367, | |
| "eval_samples_per_second": 27.102, | |
| "eval_steps_per_second": 0.867, | |
| "eval_wer": 1.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 18.75, | |
| "learning_rate": 6.886101694915253e-05, | |
| "loss": 2.8183, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 21.88, | |
| "learning_rate": 6.767457627118643e-05, | |
| "loss": 2.4012, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 6.648813559322033e-05, | |
| "loss": 1.6359, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 28.12, | |
| "learning_rate": 6.530169491525422e-05, | |
| "loss": 1.2603, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 31.25, | |
| "learning_rate": 6.411525423728813e-05, | |
| "loss": 1.1182, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 31.25, | |
| "eval_loss": 0.6247776746749878, | |
| "eval_runtime": 13.8909, | |
| "eval_samples_per_second": 26.996, | |
| "eval_steps_per_second": 0.864, | |
| "eval_wer": 0.713119216113228, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 34.38, | |
| "learning_rate": 6.292881355932203e-05, | |
| "loss": 1.0279, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 37.5, | |
| "learning_rate": 6.174237288135593e-05, | |
| "loss": 0.9694, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 40.62, | |
| "learning_rate": 6.0555932203389824e-05, | |
| "loss": 0.8993, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 43.75, | |
| "learning_rate": 5.936949152542372e-05, | |
| "loss": 0.8663, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 46.88, | |
| "learning_rate": 5.818305084745762e-05, | |
| "loss": 0.8329, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 46.88, | |
| "eval_loss": 0.5573042631149292, | |
| "eval_runtime": 13.9992, | |
| "eval_samples_per_second": 26.787, | |
| "eval_steps_per_second": 0.857, | |
| "eval_wer": 0.5792052259118128, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 5.6996610169491526e-05, | |
| "loss": 0.8219, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 53.12, | |
| "learning_rate": 5.581016949152542e-05, | |
| "loss": 0.8078, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 56.25, | |
| "learning_rate": 5.4623728813559317e-05, | |
| "loss": 0.7718, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 59.38, | |
| "learning_rate": 5.3437288135593215e-05, | |
| "loss": 0.7365, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 62.5, | |
| "learning_rate": 5.2250847457627114e-05, | |
| "loss": 0.7109, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 62.5, | |
| "eval_loss": 0.54203200340271, | |
| "eval_runtime": 13.6769, | |
| "eval_samples_per_second": 27.419, | |
| "eval_steps_per_second": 0.877, | |
| "eval_wer": 0.568317909635275, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 65.62, | |
| "learning_rate": 5.106440677966102e-05, | |
| "loss": 0.691, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 68.75, | |
| "learning_rate": 4.987796610169491e-05, | |
| "loss": 0.69, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 71.88, | |
| "learning_rate": 4.870338983050847e-05, | |
| "loss": 0.6662, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "learning_rate": 4.7528813559322025e-05, | |
| "loss": 0.6387, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 78.12, | |
| "learning_rate": 4.634237288135593e-05, | |
| "loss": 0.6295, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 78.12, | |
| "eval_loss": 0.5166199803352356, | |
| "eval_runtime": 13.9794, | |
| "eval_samples_per_second": 26.825, | |
| "eval_steps_per_second": 0.858, | |
| "eval_wer": 0.5394665215024497, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 81.25, | |
| "learning_rate": 4.515593220338983e-05, | |
| "loss": 0.6157, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 84.38, | |
| "learning_rate": 4.396949152542372e-05, | |
| "loss": 0.596, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 87.5, | |
| "learning_rate": 4.2783050847457626e-05, | |
| "loss": 0.5883, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 90.62, | |
| "learning_rate": 4.159661016949152e-05, | |
| "loss": 0.5843, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 93.75, | |
| "learning_rate": 4.041016949152542e-05, | |
| "loss": 0.5715, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 93.75, | |
| "eval_loss": 0.5486962795257568, | |
| "eval_runtime": 13.6771, | |
| "eval_samples_per_second": 27.418, | |
| "eval_steps_per_second": 0.877, | |
| "eval_wer": 0.562874251497006, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 96.88, | |
| "learning_rate": 3.922372881355932e-05, | |
| "loss": 0.5581, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 3.8037288135593213e-05, | |
| "loss": 0.5262, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 103.12, | |
| "learning_rate": 3.685084745762712e-05, | |
| "loss": 0.5337, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 106.25, | |
| "learning_rate": 3.566440677966101e-05, | |
| "loss": 0.5229, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 109.38, | |
| "learning_rate": 3.4477966101694916e-05, | |
| "loss": 0.5016, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 109.38, | |
| "eval_loss": 0.5369759202003479, | |
| "eval_runtime": 14.0144, | |
| "eval_samples_per_second": 26.758, | |
| "eval_steps_per_second": 0.856, | |
| "eval_wer": 0.5470876428960262, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 112.5, | |
| "learning_rate": 3.329152542372881e-05, | |
| "loss": 0.4817, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 115.62, | |
| "learning_rate": 3.2105084745762706e-05, | |
| "loss": 0.4959, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 118.75, | |
| "learning_rate": 3.0918644067796605e-05, | |
| "loss": 0.4786, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 121.88, | |
| "learning_rate": 2.9732203389830507e-05, | |
| "loss": 0.4671, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 125.0, | |
| "learning_rate": 2.8545762711864405e-05, | |
| "loss": 0.4661, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 125.0, | |
| "eval_loss": 0.5621253252029419, | |
| "eval_runtime": 13.7208, | |
| "eval_samples_per_second": 27.331, | |
| "eval_steps_per_second": 0.875, | |
| "eval_wer": 0.5394665215024497, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 128.12, | |
| "learning_rate": 2.7359322033898304e-05, | |
| "loss": 0.4538, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 131.25, | |
| "learning_rate": 2.61728813559322e-05, | |
| "loss": 0.4482, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 134.38, | |
| "learning_rate": 2.4986440677966097e-05, | |
| "loss": 0.4518, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 137.5, | |
| "learning_rate": 2.38e-05, | |
| "loss": 0.4371, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 140.62, | |
| "learning_rate": 2.2613559322033898e-05, | |
| "loss": 0.423, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 140.62, | |
| "eval_loss": 0.5657923221588135, | |
| "eval_runtime": 13.7521, | |
| "eval_samples_per_second": 27.269, | |
| "eval_steps_per_second": 0.873, | |
| "eval_wer": 0.5247686445291235, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 143.75, | |
| "learning_rate": 2.1427118644067796e-05, | |
| "loss": 0.4134, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 146.88, | |
| "learning_rate": 2.024067796610169e-05, | |
| "loss": 0.4014, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 150.0, | |
| "learning_rate": 1.905423728813559e-05, | |
| "loss": 0.4001, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 153.12, | |
| "learning_rate": 1.786779661016949e-05, | |
| "loss": 0.3777, | |
| "step": 4900 | |
| }, | |
| { | |
| "epoch": 156.25, | |
| "learning_rate": 1.6681355932203387e-05, | |
| "loss": 0.3793, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 156.25, | |
| "eval_loss": 0.5920763611793518, | |
| "eval_runtime": 13.612, | |
| "eval_samples_per_second": 27.549, | |
| "eval_steps_per_second": 0.882, | |
| "eval_wer": 0.4980947196516059, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 159.38, | |
| "learning_rate": 1.549491525423729e-05, | |
| "loss": 0.3843, | |
| "step": 5100 | |
| }, | |
| { | |
| "epoch": 162.5, | |
| "learning_rate": 1.4308474576271186e-05, | |
| "loss": 0.3918, | |
| "step": 5200 | |
| }, | |
| { | |
| "epoch": 165.62, | |
| "learning_rate": 1.3122033898305082e-05, | |
| "loss": 0.3593, | |
| "step": 5300 | |
| }, | |
| { | |
| "epoch": 168.75, | |
| "learning_rate": 1.1935593220338983e-05, | |
| "loss": 0.3674, | |
| "step": 5400 | |
| }, | |
| { | |
| "epoch": 171.88, | |
| "learning_rate": 1.074915254237288e-05, | |
| "loss": 0.3651, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 171.88, | |
| "eval_loss": 0.5986579060554504, | |
| "eval_runtime": 13.8361, | |
| "eval_samples_per_second": 27.103, | |
| "eval_steps_per_second": 0.867, | |
| "eval_wer": 0.48884050081654873, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 175.0, | |
| "learning_rate": 9.56271186440678e-06, | |
| "loss": 0.345, | |
| "step": 5600 | |
| }, | |
| { | |
| "epoch": 178.12, | |
| "learning_rate": 8.376271186440677e-06, | |
| "loss": 0.3527, | |
| "step": 5700 | |
| }, | |
| { | |
| "epoch": 181.25, | |
| "learning_rate": 7.189830508474576e-06, | |
| "loss": 0.3446, | |
| "step": 5800 | |
| }, | |
| { | |
| "epoch": 184.38, | |
| "learning_rate": 6.003389830508474e-06, | |
| "loss": 0.3461, | |
| "step": 5900 | |
| }, | |
| { | |
| "epoch": 187.5, | |
| "learning_rate": 4.816949152542372e-06, | |
| "loss": 0.3351, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 187.5, | |
| "eval_loss": 0.6017099022865295, | |
| "eval_runtime": 13.6253, | |
| "eval_samples_per_second": 27.522, | |
| "eval_steps_per_second": 0.881, | |
| "eval_wer": 0.49482852476864453, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 190.62, | |
| "learning_rate": 3.630508474576271e-06, | |
| "loss": 0.3368, | |
| "step": 6100 | |
| }, | |
| { | |
| "epoch": 193.75, | |
| "learning_rate": 2.444067796610169e-06, | |
| "loss": 0.3405, | |
| "step": 6200 | |
| }, | |
| { | |
| "epoch": 196.88, | |
| "learning_rate": 1.2576271186440677e-06, | |
| "loss": 0.3271, | |
| "step": 6300 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "learning_rate": 7.118644067796609e-08, | |
| "loss": 0.334, | |
| "step": 6400 | |
| }, | |
| { | |
| "epoch": 200.0, | |
| "step": 6400, | |
| "total_flos": 2.9581900723233317e+19, | |
| "train_loss": 1.1891312944889068, | |
| "train_runtime": 10628.5861, | |
| "train_samples_per_second": 18.968, | |
| "train_steps_per_second": 0.602 | |
| } | |
| ], | |
| "max_steps": 6400, | |
| "num_train_epochs": 200, | |
| "total_flos": 2.9581900723233317e+19, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |