| { | |
| "best_global_step": 11700, | |
| "best_metric": 0.5111122131347656, | |
| "best_model_checkpoint": "./output/gpt-medmentions/checkpoint-11700", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 29250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08547008547008547, | |
| "grad_norm": 7.1186957359313965, | |
| "learning_rate": 4.9153846153846157e-05, | |
| "loss": 0.9374, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.17094017094017094, | |
| "grad_norm": 2.6842756271362305, | |
| "learning_rate": 4.829914529914531e-05, | |
| "loss": 0.6708, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2564102564102564, | |
| "grad_norm": 4.874898433685303, | |
| "learning_rate": 4.7444444444444445e-05, | |
| "loss": 0.6453, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.3418803418803419, | |
| "grad_norm": 5.255520820617676, | |
| "learning_rate": 4.658974358974359e-05, | |
| "loss": 0.6121, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.42735042735042733, | |
| "grad_norm": 2.5840187072753906, | |
| "learning_rate": 4.573504273504274e-05, | |
| "loss": 0.5936, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.5128205128205128, | |
| "grad_norm": 5.246805667877197, | |
| "learning_rate": 4.488547008547009e-05, | |
| "loss": 0.5943, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.5982905982905983, | |
| "grad_norm": 3.491377592086792, | |
| "learning_rate": 4.4032478632478637e-05, | |
| "loss": 0.5663, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.6837606837606838, | |
| "grad_norm": 4.0636210441589355, | |
| "learning_rate": 4.317777777777778e-05, | |
| "loss": 0.5611, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.7692307692307693, | |
| "grad_norm": 4.316960334777832, | |
| "learning_rate": 4.2323076923076925e-05, | |
| "loss": 0.5449, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.8547008547008547, | |
| "grad_norm": 4.260451316833496, | |
| "learning_rate": 4.146837606837607e-05, | |
| "loss": 0.5481, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.9401709401709402, | |
| "grad_norm": 5.917566299438477, | |
| "learning_rate": 4.061367521367522e-05, | |
| "loss": 0.5307, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8341444876242783, | |
| "eval_f1": 0.44006334125098967, | |
| "eval_loss": 0.5368949770927429, | |
| "eval_precision": 0.41286584459961373, | |
| "eval_recall": 0.47109679606713, | |
| "eval_runtime": 20.612, | |
| "eval_samples_per_second": 141.859, | |
| "eval_steps_per_second": 17.757, | |
| "step": 5850 | |
| }, | |
| { | |
| "epoch": 1.0256410256410255, | |
| "grad_norm": 3.265369415283203, | |
| "learning_rate": 3.975897435897436e-05, | |
| "loss": 0.4748, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 3.2452337741851807, | |
| "learning_rate": 3.890427350427351e-05, | |
| "loss": 0.3767, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.1965811965811965, | |
| "grad_norm": 2.8009512424468994, | |
| "learning_rate": 3.804957264957265e-05, | |
| "loss": 0.3761, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.282051282051282, | |
| "grad_norm": 5.858109951019287, | |
| "learning_rate": 3.71948717948718e-05, | |
| "loss": 0.3922, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.3675213675213675, | |
| "grad_norm": 1.7275584936141968, | |
| "learning_rate": 3.634017094017094e-05, | |
| "loss": 0.3677, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.452991452991453, | |
| "grad_norm": 4.6104044914245605, | |
| "learning_rate": 3.5485470085470085e-05, | |
| "loss": 0.3746, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.5384615384615383, | |
| "grad_norm": 2.6036839485168457, | |
| "learning_rate": 3.4630769230769236e-05, | |
| "loss": 0.3661, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.623931623931624, | |
| "grad_norm": 2.7406065464019775, | |
| "learning_rate": 3.3776068376068374e-05, | |
| "loss": 0.3712, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.7094017094017095, | |
| "grad_norm": 5.254650115966797, | |
| "learning_rate": 3.2921367521367525e-05, | |
| "loss": 0.3774, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.7948717948717947, | |
| "grad_norm": 2.274414539337158, | |
| "learning_rate": 3.206666666666667e-05, | |
| "loss": 0.3541, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 1.8803418803418803, | |
| "grad_norm": 3.5981504917144775, | |
| "learning_rate": 3.121196581196581e-05, | |
| "loss": 0.3694, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 1.965811965811966, | |
| "grad_norm": 3.6442151069641113, | |
| "learning_rate": 3.0357264957264958e-05, | |
| "loss": 0.3585, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8454107464662687, | |
| "eval_f1": 0.48178988326848243, | |
| "eval_loss": 0.5111122131347656, | |
| "eval_precision": 0.4453316069630269, | |
| "eval_recall": 0.5247499576199356, | |
| "eval_runtime": 20.6467, | |
| "eval_samples_per_second": 141.621, | |
| "eval_steps_per_second": 17.727, | |
| "step": 11700 | |
| }, | |
| { | |
| "epoch": 2.051282051282051, | |
| "grad_norm": 3.4292666912078857, | |
| "learning_rate": 2.9502564102564105e-05, | |
| "loss": 0.2559, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.1367521367521367, | |
| "grad_norm": 6.138054370880127, | |
| "learning_rate": 2.864786324786325e-05, | |
| "loss": 0.1817, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 3.659104347229004, | |
| "learning_rate": 2.7793162393162394e-05, | |
| "loss": 0.1759, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.3076923076923075, | |
| "grad_norm": 5.470319747924805, | |
| "learning_rate": 2.693846153846154e-05, | |
| "loss": 0.184, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.393162393162393, | |
| "grad_norm": 2.9865291118621826, | |
| "learning_rate": 2.6083760683760682e-05, | |
| "loss": 0.1761, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.4786324786324787, | |
| "grad_norm": 6.152403831481934, | |
| "learning_rate": 2.522905982905983e-05, | |
| "loss": 0.1798, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.564102564102564, | |
| "grad_norm": 4.3192338943481445, | |
| "learning_rate": 2.4374358974358977e-05, | |
| "loss": 0.1757, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.6495726495726495, | |
| "grad_norm": 3.217804193496704, | |
| "learning_rate": 2.3521367521367523e-05, | |
| "loss": 0.1738, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.735042735042735, | |
| "grad_norm": 3.670557737350464, | |
| "learning_rate": 2.2666666666666668e-05, | |
| "loss": 0.1861, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.8205128205128203, | |
| "grad_norm": 2.3006069660186768, | |
| "learning_rate": 2.1811965811965812e-05, | |
| "loss": 0.1705, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 2.905982905982906, | |
| "grad_norm": 1.9008346796035767, | |
| "learning_rate": 2.0958974358974358e-05, | |
| "loss": 0.1672, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 2.9914529914529915, | |
| "grad_norm": 1.8553671836853027, | |
| "learning_rate": 2.0104273504273506e-05, | |
| "loss": 0.1758, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8497137463068983, | |
| "eval_f1": 0.48074844074844075, | |
| "eval_loss": 0.6349462270736694, | |
| "eval_precision": 0.4718413320274241, | |
| "eval_recall": 0.4899983047974233, | |
| "eval_runtime": 20.6084, | |
| "eval_samples_per_second": 141.884, | |
| "eval_steps_per_second": 17.76, | |
| "step": 17550 | |
| }, | |
| { | |
| "epoch": 3.076923076923077, | |
| "grad_norm": 1.145456314086914, | |
| "learning_rate": 1.924957264957265e-05, | |
| "loss": 0.0777, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 3.1623931623931623, | |
| "grad_norm": 2.4131710529327393, | |
| "learning_rate": 1.8394871794871797e-05, | |
| "loss": 0.0769, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 3.247863247863248, | |
| "grad_norm": 2.5216588973999023, | |
| "learning_rate": 1.754017094017094e-05, | |
| "loss": 0.0779, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 2.811354160308838, | |
| "learning_rate": 1.6685470085470086e-05, | |
| "loss": 0.0804, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 3.4188034188034186, | |
| "grad_norm": 0.5833438634872437, | |
| "learning_rate": 1.5830769230769233e-05, | |
| "loss": 0.0772, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 3.5042735042735043, | |
| "grad_norm": 3.458584785461426, | |
| "learning_rate": 1.4976068376068378e-05, | |
| "loss": 0.0751, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 3.58974358974359, | |
| "grad_norm": 0.8929054141044617, | |
| "learning_rate": 1.4121367521367524e-05, | |
| "loss": 0.0761, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 3.6752136752136755, | |
| "grad_norm": 5.908766269683838, | |
| "learning_rate": 1.3268376068376068e-05, | |
| "loss": 0.0736, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 3.7606837606837606, | |
| "grad_norm": 1.0228583812713623, | |
| "learning_rate": 1.2413675213675214e-05, | |
| "loss": 0.0716, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 3.8461538461538463, | |
| "grad_norm": 0.626966118812561, | |
| "learning_rate": 1.1560683760683762e-05, | |
| "loss": 0.0718, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 3.931623931623932, | |
| "grad_norm": 1.9417258501052856, | |
| "learning_rate": 1.0705982905982906e-05, | |
| "loss": 0.0751, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8496892277892879, | |
| "eval_f1": 0.49008894029434047, | |
| "eval_loss": 0.9264360070228577, | |
| "eval_precision": 0.46282485875706214, | |
| "eval_recall": 0.5207662315646719, | |
| "eval_runtime": 20.6775, | |
| "eval_samples_per_second": 141.41, | |
| "eval_steps_per_second": 17.7, | |
| "step": 23400 | |
| }, | |
| { | |
| "epoch": 4.017094017094017, | |
| "grad_norm": 0.14389610290527344, | |
| "learning_rate": 9.851282051282052e-06, | |
| "loss": 0.0641, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 4.102564102564102, | |
| "grad_norm": 0.6148084402084351, | |
| "learning_rate": 8.996581196581196e-06, | |
| "loss": 0.0352, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 4.188034188034188, | |
| "grad_norm": 1.3859856128692627, | |
| "learning_rate": 8.141880341880342e-06, | |
| "loss": 0.0399, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 4.273504273504273, | |
| "grad_norm": 1.6096197366714478, | |
| "learning_rate": 7.287179487179488e-06, | |
| "loss": 0.0385, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 4.358974358974359, | |
| "grad_norm": 0.26734989881515503, | |
| "learning_rate": 6.432478632478633e-06, | |
| "loss": 0.0411, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 4.444444444444445, | |
| "grad_norm": 1.3472919464111328, | |
| "learning_rate": 5.577777777777778e-06, | |
| "loss": 0.0395, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 4.52991452991453, | |
| "grad_norm": 0.8369725942611694, | |
| "learning_rate": 4.723076923076923e-06, | |
| "loss": 0.0432, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 4.615384615384615, | |
| "grad_norm": 0.7225199341773987, | |
| "learning_rate": 3.87008547008547e-06, | |
| "loss": 0.0412, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 4.700854700854701, | |
| "grad_norm": 0.6592767834663391, | |
| "learning_rate": 3.0153846153846154e-06, | |
| "loss": 0.0407, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 4.786324786324786, | |
| "grad_norm": 0.9876635670661926, | |
| "learning_rate": 2.160683760683761e-06, | |
| "loss": 0.0372, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 4.871794871794872, | |
| "grad_norm": 0.7053186297416687, | |
| "learning_rate": 1.3059829059829061e-06, | |
| "loss": 0.0384, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 4.957264957264957, | |
| "grad_norm": 0.24596278369426727, | |
| "learning_rate": 4.52991452991453e-07, | |
| "loss": 0.0387, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8518223388213949, | |
| "eval_f1": 0.4960441433034446, | |
| "eval_loss": 1.0903491973876953, | |
| "eval_precision": 0.47575686823877344, | |
| "eval_recall": 0.5181386675707748, | |
| "eval_runtime": 20.8036, | |
| "eval_samples_per_second": 140.552, | |
| "eval_steps_per_second": 17.593, | |
| "step": 29250 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 29250, | |
| "total_flos": 5.182622875540416e+16, | |
| "train_loss": 0.2549698821499816, | |
| "train_runtime": 7489.8753, | |
| "train_samples_per_second": 15.62, | |
| "train_steps_per_second": 3.905 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 29250, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.182622875540416e+16, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |