| { | |
| "best_metric": 0.0727909728884697, | |
| "best_model_checkpoint": "/data/wheld3/mt5-base-pointer-adv-cstop_artificial/checkpoint-400", | |
| "epoch": 187.49805447470817, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 12.5, | |
| "learning_rate": 0.0009333333333333333, | |
| "loss": 1.7423, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 12.5, | |
| "eval_exact_match": 0.23971377459749552, | |
| "eval_loss": 0.11730749905109406, | |
| "eval_runtime": 27.8131, | |
| "eval_samples_per_second": 20.098, | |
| "eval_steps_per_second": 2.517, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "learning_rate": 0.0008666666666666667, | |
| "loss": 0.3678, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 25.0, | |
| "eval_exact_match": 0.3363148479427549, | |
| "eval_loss": 0.0727909728884697, | |
| "eval_runtime": 29.0466, | |
| "eval_samples_per_second": 19.245, | |
| "eval_steps_per_second": 2.41, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 37.5, | |
| "learning_rate": 0.0008, | |
| "loss": 0.3202, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 37.5, | |
| "eval_exact_match": 0.33810375670840787, | |
| "eval_loss": 0.08794570714235306, | |
| "eval_runtime": 29.1974, | |
| "eval_samples_per_second": 19.146, | |
| "eval_steps_per_second": 2.397, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "learning_rate": 0.0007333333333333333, | |
| "loss": 0.3452, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_exact_match": 0.3363148479427549, | |
| "eval_loss": 0.09075574576854706, | |
| "eval_runtime": 28.6647, | |
| "eval_samples_per_second": 19.501, | |
| "eval_steps_per_second": 2.442, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 62.5, | |
| "learning_rate": 0.0006666666666666666, | |
| "loss": 0.3099, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 62.5, | |
| "eval_exact_match": 0.3434704830053667, | |
| "eval_loss": 0.10556001961231232, | |
| "eval_runtime": 28.9715, | |
| "eval_samples_per_second": 19.295, | |
| "eval_steps_per_second": 2.416, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "learning_rate": 0.0006, | |
| "loss": 0.3057, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 75.0, | |
| "eval_exact_match": 0.3470483005366726, | |
| "eval_loss": 0.11086518317461014, | |
| "eval_runtime": 29.2047, | |
| "eval_samples_per_second": 19.141, | |
| "eval_steps_per_second": 2.397, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 87.5, | |
| "learning_rate": 0.0005333333333333334, | |
| "loss": 0.3045, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 87.5, | |
| "eval_exact_match": 0.34525939177101966, | |
| "eval_loss": 0.1273432970046997, | |
| "eval_runtime": 29.0031, | |
| "eval_samples_per_second": 19.274, | |
| "eval_steps_per_second": 2.414, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "learning_rate": 0.00046666666666666666, | |
| "loss": 0.3052, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "eval_exact_match": 0.3416815742397138, | |
| "eval_loss": 0.10654404759407043, | |
| "eval_runtime": 28.7998, | |
| "eval_samples_per_second": 19.41, | |
| "eval_steps_per_second": 2.431, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 112.5, | |
| "learning_rate": 0.0004, | |
| "loss": 0.3037, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 112.5, | |
| "eval_exact_match": 0.33810375670840787, | |
| "eval_loss": 0.13873372972011566, | |
| "eval_runtime": 29.2765, | |
| "eval_samples_per_second": 19.094, | |
| "eval_steps_per_second": 2.391, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 125.0, | |
| "learning_rate": 0.0003333333333333333, | |
| "loss": 0.3036, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 125.0, | |
| "eval_exact_match": 0.34525939177101966, | |
| "eval_loss": 0.1421414017677307, | |
| "eval_runtime": 28.8117, | |
| "eval_samples_per_second": 19.402, | |
| "eval_steps_per_second": 2.43, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 137.5, | |
| "learning_rate": 0.0002666666666666667, | |
| "loss": 0.3023, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 137.5, | |
| "eval_exact_match": 0.33989266547406083, | |
| "eval_loss": 0.16489343345165253, | |
| "eval_runtime": 30.9662, | |
| "eval_samples_per_second": 18.052, | |
| "eval_steps_per_second": 2.261, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 150.0, | |
| "learning_rate": 0.0002, | |
| "loss": 0.3028, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 150.0, | |
| "eval_exact_match": 0.33989266547406083, | |
| "eval_loss": 0.1573849618434906, | |
| "eval_runtime": 29.0042, | |
| "eval_samples_per_second": 19.273, | |
| "eval_steps_per_second": 2.413, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 162.5, | |
| "learning_rate": 0.00013333333333333334, | |
| "loss": 0.3025, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 162.5, | |
| "eval_exact_match": 0.33989266547406083, | |
| "eval_loss": 0.15625949203968048, | |
| "eval_runtime": 28.8424, | |
| "eval_samples_per_second": 19.381, | |
| "eval_steps_per_second": 2.427, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 175.0, | |
| "learning_rate": 6.666666666666667e-05, | |
| "loss": 0.3017, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 175.0, | |
| "eval_exact_match": 0.33989266547406083, | |
| "eval_loss": 0.1589040458202362, | |
| "eval_runtime": 28.7706, | |
| "eval_samples_per_second": 19.43, | |
| "eval_steps_per_second": 2.433, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 187.5, | |
| "learning_rate": 0.0, | |
| "loss": 0.302, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 187.5, | |
| "eval_exact_match": 0.3416815742397138, | |
| "eval_loss": 0.15874968469142914, | |
| "eval_runtime": 33.2641, | |
| "eval_samples_per_second": 16.805, | |
| "eval_steps_per_second": 2.104, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 187.5, | |
| "step": 3000, | |
| "total_flos": 6.474719775139762e+16, | |
| "train_loss": 0.4079610900878906, | |
| "train_runtime": 34544.7098, | |
| "train_samples_per_second": 44.464, | |
| "train_steps_per_second": 0.087 | |
| } | |
| ], | |
| "max_steps": 3000, | |
| "num_train_epochs": 188, | |
| "total_flos": 6.474719775139762e+16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |