| { | |
| "best_global_step": 2000, | |
| "best_metric": 0.8264462809917356, | |
| "best_model_checkpoint": "./models/checkpoints_v4/checkpoint-2000", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 2280, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.21929824561403508, | |
| "grad_norm": 3.511173725128174, | |
| "learning_rate": 4.298245614035088e-06, | |
| "loss": 0.8133, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.43859649122807015, | |
| "grad_norm": 5.231376647949219, | |
| "learning_rate": 8.68421052631579e-06, | |
| "loss": 0.5079, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6578947368421053, | |
| "grad_norm": 4.2739081382751465, | |
| "learning_rate": 1.3070175438596493e-05, | |
| "loss": 0.3543, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8771929824561403, | |
| "grad_norm": 4.163576602935791, | |
| "learning_rate": 1.7456140350877195e-05, | |
| "loss": 0.3143, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0964912280701755, | |
| "grad_norm": 3.981715202331543, | |
| "learning_rate": 1.9795321637426903e-05, | |
| "loss": 0.3005, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.3157894736842106, | |
| "grad_norm": 4.984482288360596, | |
| "learning_rate": 1.9307992202729045e-05, | |
| "loss": 0.2521, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.5350877192982457, | |
| "grad_norm": 7.214444637298584, | |
| "learning_rate": 1.882066276803119e-05, | |
| "loss": 0.2481, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.7543859649122808, | |
| "grad_norm": 3.6749589443206787, | |
| "learning_rate": 1.8333333333333333e-05, | |
| "loss": 0.2475, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.973684210526316, | |
| "grad_norm": 5.445246696472168, | |
| "learning_rate": 1.784600389863548e-05, | |
| "loss": 0.2289, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.192982456140351, | |
| "grad_norm": 12.260554313659668, | |
| "learning_rate": 1.7358674463937624e-05, | |
| "loss": 0.1811, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.192982456140351, | |
| "eval_accuracy": 0.8987154150197628, | |
| "eval_auc": 0.9507858176268903, | |
| "eval_f1": 0.7724750277469479, | |
| "eval_false_negatives": 147, | |
| "eval_false_positives": 58, | |
| "eval_loss": 0.29392391443252563, | |
| "eval_npv": 0.9091470951792336, | |
| "eval_precision": 0.8571428571428571, | |
| "eval_recall": 0.703030303030303, | |
| "eval_runtime": 7.4602, | |
| "eval_samples_per_second": 271.308, | |
| "eval_specificity": 0.9620667102681492, | |
| "eval_steps_per_second": 4.289, | |
| "eval_true_negatives": 1471, | |
| "eval_true_positives": 348, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.412280701754386, | |
| "grad_norm": 3.9946436882019043, | |
| "learning_rate": 1.6871345029239766e-05, | |
| "loss": 0.1934, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.6315789473684212, | |
| "grad_norm": 6.733097553253174, | |
| "learning_rate": 1.638401559454191e-05, | |
| "loss": 0.1759, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.8508771929824563, | |
| "grad_norm": 12.989212036132812, | |
| "learning_rate": 1.5896686159844057e-05, | |
| "loss": 0.1953, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.0701754385964914, | |
| "grad_norm": 4.068276882171631, | |
| "learning_rate": 1.5409356725146202e-05, | |
| "loss": 0.1248, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.2894736842105265, | |
| "grad_norm": 0.6126830577850342, | |
| "learning_rate": 1.4922027290448344e-05, | |
| "loss": 0.1289, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.5087719298245617, | |
| "grad_norm": 6.3761796951293945, | |
| "learning_rate": 1.4434697855750488e-05, | |
| "loss": 0.1117, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 3.7280701754385968, | |
| "grad_norm": 4.546126365661621, | |
| "learning_rate": 1.3947368421052631e-05, | |
| "loss": 0.1491, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 3.9473684210526314, | |
| "grad_norm": 4.006205081939697, | |
| "learning_rate": 1.3460038986354777e-05, | |
| "loss": 0.1345, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.166666666666667, | |
| "grad_norm": 2.2420225143432617, | |
| "learning_rate": 1.297270955165692e-05, | |
| "loss": 0.0904, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 4.385964912280702, | |
| "grad_norm": 3.2088565826416016, | |
| "learning_rate": 1.2485380116959064e-05, | |
| "loss": 0.0996, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.385964912280702, | |
| "eval_accuracy": 0.9115612648221344, | |
| "eval_auc": 0.9578519002979434, | |
| "eval_f1": 0.8182741116751269, | |
| "eval_false_negatives": 92, | |
| "eval_false_positives": 87, | |
| "eval_loss": 0.3153094947338104, | |
| "eval_npv": 0.940026075619296, | |
| "eval_precision": 0.8224489795918367, | |
| "eval_recall": 0.8141414141414142, | |
| "eval_runtime": 7.4628, | |
| "eval_samples_per_second": 271.212, | |
| "eval_specificity": 0.9431000654022237, | |
| "eval_steps_per_second": 4.288, | |
| "eval_true_negatives": 1442, | |
| "eval_true_positives": 403, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 4.605263157894737, | |
| "grad_norm": 10.743696212768555, | |
| "learning_rate": 1.1998050682261208e-05, | |
| "loss": 0.0909, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 4.824561403508772, | |
| "grad_norm": 1.6505861282348633, | |
| "learning_rate": 1.1510721247563355e-05, | |
| "loss": 0.088, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 5.043859649122807, | |
| "grad_norm": 6.567928314208984, | |
| "learning_rate": 1.1023391812865499e-05, | |
| "loss": 0.1129, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 5.2631578947368425, | |
| "grad_norm": 3.445005416870117, | |
| "learning_rate": 1.0536062378167643e-05, | |
| "loss": 0.0605, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 5.482456140350878, | |
| "grad_norm": 1.5516074895858765, | |
| "learning_rate": 1.0048732943469786e-05, | |
| "loss": 0.0642, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 5.701754385964913, | |
| "grad_norm": 9.6685209274292, | |
| "learning_rate": 9.56140350877193e-06, | |
| "loss": 0.0759, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 5.921052631578947, | |
| "grad_norm": 12.226219177246094, | |
| "learning_rate": 9.074074074074075e-06, | |
| "loss": 0.0806, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 6.140350877192983, | |
| "grad_norm": 1.530657172203064, | |
| "learning_rate": 8.586744639376219e-06, | |
| "loss": 0.0636, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 6.359649122807017, | |
| "grad_norm": 4.697056293487549, | |
| "learning_rate": 8.099415204678363e-06, | |
| "loss": 0.0402, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 6.578947368421053, | |
| "grad_norm": 0.09656322002410889, | |
| "learning_rate": 7.612085769980507e-06, | |
| "loss": 0.0629, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.578947368421053, | |
| "eval_accuracy": 0.9179841897233202, | |
| "eval_auc": 0.9618949468524354, | |
| "eval_f1": 0.8241525423728814, | |
| "eval_false_negatives": 106, | |
| "eval_false_positives": 60, | |
| "eval_loss": 0.38592642545700073, | |
| "eval_npv": 0.9326984126984127, | |
| "eval_precision": 0.8663697104677061, | |
| "eval_recall": 0.7858585858585858, | |
| "eval_runtime": 7.4828, | |
| "eval_samples_per_second": 270.486, | |
| "eval_specificity": 0.960758665794637, | |
| "eval_steps_per_second": 4.276, | |
| "eval_true_negatives": 1469, | |
| "eval_true_positives": 389, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 6.798245614035087, | |
| "grad_norm": 36.65667724609375, | |
| "learning_rate": 7.124756335282652e-06, | |
| "loss": 0.0606, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 7.017543859649122, | |
| "grad_norm": 2.3685500621795654, | |
| "learning_rate": 6.637426900584796e-06, | |
| "loss": 0.0513, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 7.2368421052631575, | |
| "grad_norm": 2.3388023376464844, | |
| "learning_rate": 6.15009746588694e-06, | |
| "loss": 0.0328, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 7.456140350877193, | |
| "grad_norm": 0.18014875054359436, | |
| "learning_rate": 5.662768031189084e-06, | |
| "loss": 0.0526, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 7.675438596491228, | |
| "grad_norm": 5.354085445404053, | |
| "learning_rate": 5.175438596491229e-06, | |
| "loss": 0.033, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 7.894736842105263, | |
| "grad_norm": 0.14219924807548523, | |
| "learning_rate": 4.688109161793373e-06, | |
| "loss": 0.0286, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 8.114035087719298, | |
| "grad_norm": 0.8848229646682739, | |
| "learning_rate": 4.200779727095517e-06, | |
| "loss": 0.023, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 8.333333333333334, | |
| "grad_norm": 0.09083296358585358, | |
| "learning_rate": 3.713450292397661e-06, | |
| "loss": 0.0342, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 8.552631578947368, | |
| "grad_norm": 5.818343162536621, | |
| "learning_rate": 3.2261208576998056e-06, | |
| "loss": 0.0317, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 8.771929824561404, | |
| "grad_norm": 0.027245348319411278, | |
| "learning_rate": 2.7387914230019497e-06, | |
| "loss": 0.0344, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.771929824561404, | |
| "eval_accuracy": 0.9169960474308301, | |
| "eval_auc": 0.9587609251441821, | |
| "eval_f1": 0.8264462809917356, | |
| "eval_false_negatives": 95, | |
| "eval_false_positives": 73, | |
| "eval_loss": 0.44914552569389343, | |
| "eval_npv": 0.938749194068343, | |
| "eval_precision": 0.8456659619450317, | |
| "eval_recall": 0.8080808080808081, | |
| "eval_runtime": 7.4877, | |
| "eval_samples_per_second": 270.31, | |
| "eval_specificity": 0.9522563767168084, | |
| "eval_steps_per_second": 4.274, | |
| "eval_true_negatives": 1456, | |
| "eval_true_positives": 400, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 8.991228070175438, | |
| "grad_norm": 6.446545124053955, | |
| "learning_rate": 2.2514619883040934e-06, | |
| "loss": 0.031, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 9.210526315789474, | |
| "grad_norm": 3.8243560791015625, | |
| "learning_rate": 1.7641325536062378e-06, | |
| "loss": 0.0322, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 9.429824561403509, | |
| "grad_norm": 0.14315390586853027, | |
| "learning_rate": 1.2768031189083821e-06, | |
| "loss": 0.025, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 9.649122807017545, | |
| "grad_norm": 0.08034974336624146, | |
| "learning_rate": 7.894736842105263e-07, | |
| "loss": 0.0143, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 9.868421052631579, | |
| "grad_norm": 0.1442982703447342, | |
| "learning_rate": 3.021442495126706e-07, | |
| "loss": 0.0121, | |
| "step": 2250 | |
| } | |
| ], | |
| "logging_steps": 50, | |
| "max_steps": 2280, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 3, | |
| "early_stopping_threshold": 0.001 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 0 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9579873525657600.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |