| { | |
| "best_metric": 0.8716280849435623, | |
| "best_model_checkpoint": "best_model_big/checkpoint-2968", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 3710, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1347708894878706, | |
| "grad_norm": 18.93534278869629, | |
| "learning_rate": 1.946091644204852e-05, | |
| "loss": 0.5038, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2695417789757412, | |
| "grad_norm": 7.889501571655273, | |
| "learning_rate": 1.8921832884097035e-05, | |
| "loss": 0.4015, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.40431266846361186, | |
| "grad_norm": 17.194637298583984, | |
| "learning_rate": 1.8382749326145554e-05, | |
| "loss": 0.359, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.5390835579514824, | |
| "grad_norm": 10.652689933776855, | |
| "learning_rate": 1.7843665768194072e-05, | |
| "loss": 0.3494, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6738544474393531, | |
| "grad_norm": 12.343999862670898, | |
| "learning_rate": 1.7304582210242588e-05, | |
| "loss": 0.357, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.8086253369272237, | |
| "grad_norm": 5.575014114379883, | |
| "learning_rate": 1.6765498652291106e-05, | |
| "loss": 0.3446, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9433962264150944, | |
| "grad_norm": 10.024672508239746, | |
| "learning_rate": 1.6226415094339625e-05, | |
| "loss": 0.3307, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8470863462356444, | |
| "eval_confusion_matrix": [ | |
| [ | |
| 1759, | |
| 519 | |
| ], | |
| [ | |
| 200, | |
| 2224 | |
| ] | |
| ], | |
| "eval_f1": 0.8608476872459842, | |
| "eval_loss": 0.3607315421104431, | |
| "eval_precision": 0.8107911046299672, | |
| "eval_recall": 0.9174917491749175, | |
| "eval_runtime": 22.8032, | |
| "eval_samples_per_second": 206.199, | |
| "eval_steps_per_second": 4.298, | |
| "step": 742 | |
| }, | |
| { | |
| "epoch": 1.0781671159029649, | |
| "grad_norm": 8.227458000183105, | |
| "learning_rate": 1.5687331536388143e-05, | |
| "loss": 0.301, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.2129380053908356, | |
| "grad_norm": 13.138360023498535, | |
| "learning_rate": 1.5148247978436658e-05, | |
| "loss": 0.261, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.3477088948787062, | |
| "grad_norm": 21.908050537109375, | |
| "learning_rate": 1.4609164420485175e-05, | |
| "loss": 0.2568, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.482479784366577, | |
| "grad_norm": 65.63346862792969, | |
| "learning_rate": 1.4070080862533696e-05, | |
| "loss": 0.2525, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.6172506738544474, | |
| "grad_norm": 10.492274284362793, | |
| "learning_rate": 1.3530997304582212e-05, | |
| "loss": 0.2738, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.7520215633423182, | |
| "grad_norm": 4.424431800842285, | |
| "learning_rate": 1.299191374663073e-05, | |
| "loss": 0.2612, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.8867924528301887, | |
| "grad_norm": 5.688779830932617, | |
| "learning_rate": 1.2452830188679246e-05, | |
| "loss": 0.2821, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8517652062951936, | |
| "eval_confusion_matrix": [ | |
| [ | |
| 1750, | |
| 528 | |
| ], | |
| [ | |
| 169, | |
| 2255 | |
| ] | |
| ], | |
| "eval_f1": 0.8661417322834646, | |
| "eval_loss": 0.3994266390800476, | |
| "eval_precision": 0.8102766798418972, | |
| "eval_recall": 0.9302805280528053, | |
| "eval_runtime": 22.8205, | |
| "eval_samples_per_second": 206.043, | |
| "eval_steps_per_second": 4.294, | |
| "step": 1484 | |
| }, | |
| { | |
| "epoch": 2.0215633423180592, | |
| "grad_norm": 17.876014709472656, | |
| "learning_rate": 1.1913746630727763e-05, | |
| "loss": 0.2494, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 2.1563342318059298, | |
| "grad_norm": 30.222442626953125, | |
| "learning_rate": 1.1374663072776282e-05, | |
| "loss": 0.2105, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.2911051212938007, | |
| "grad_norm": 8.314850807189941, | |
| "learning_rate": 1.0835579514824798e-05, | |
| "loss": 0.2062, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.4258760107816713, | |
| "grad_norm": 14.823955535888672, | |
| "learning_rate": 1.0296495956873315e-05, | |
| "loss": 0.1949, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.560646900269542, | |
| "grad_norm": 15.158774375915527, | |
| "learning_rate": 9.757412398921834e-06, | |
| "loss": 0.2077, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.6954177897574123, | |
| "grad_norm": 3.280972719192505, | |
| "learning_rate": 9.21832884097035e-06, | |
| "loss": 0.2029, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.830188679245283, | |
| "grad_norm": 23.08829116821289, | |
| "learning_rate": 8.67924528301887e-06, | |
| "loss": 0.2147, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.964959568733154, | |
| "grad_norm": 6.822205543518066, | |
| "learning_rate": 8.140161725067386e-06, | |
| "loss": 0.2226, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8555933645257338, | |
| "eval_confusion_matrix": [ | |
| [ | |
| 1752, | |
| 526 | |
| ], | |
| [ | |
| 153, | |
| 2271 | |
| ] | |
| ], | |
| "eval_f1": 0.8699482857690097, | |
| "eval_loss": 0.482412725687027, | |
| "eval_precision": 0.8119413657490168, | |
| "eval_recall": 0.9368811881188119, | |
| "eval_runtime": 22.8014, | |
| "eval_samples_per_second": 206.215, | |
| "eval_steps_per_second": 4.298, | |
| "step": 2226 | |
| }, | |
| { | |
| "epoch": 3.0997304582210243, | |
| "grad_norm": 21.149120330810547, | |
| "learning_rate": 7.601078167115904e-06, | |
| "loss": 0.1726, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.234501347708895, | |
| "grad_norm": 6.035734176635742, | |
| "learning_rate": 7.061994609164421e-06, | |
| "loss": 0.1614, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.3692722371967654, | |
| "grad_norm": 9.38839340209961, | |
| "learning_rate": 6.522911051212939e-06, | |
| "loss": 0.1648, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.5040431266846364, | |
| "grad_norm": 7.1731486320495605, | |
| "learning_rate": 5.983827493261456e-06, | |
| "loss": 0.1678, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 3.638814016172507, | |
| "grad_norm": 29.55657958984375, | |
| "learning_rate": 5.444743935309974e-06, | |
| "loss": 0.1492, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 3.7735849056603774, | |
| "grad_norm": 15.47530746459961, | |
| "learning_rate": 4.905660377358491e-06, | |
| "loss": 0.1707, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 3.908355795148248, | |
| "grad_norm": 8.083237648010254, | |
| "learning_rate": 4.366576819407008e-06, | |
| "loss": 0.1727, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8572947681837516, | |
| "eval_confusion_matrix": [ | |
| [ | |
| 1753, | |
| 525 | |
| ], | |
| [ | |
| 146, | |
| 2278 | |
| ] | |
| ], | |
| "eval_f1": 0.8716280849435623, | |
| "eval_loss": 0.5570098161697388, | |
| "eval_precision": 0.8127006778451659, | |
| "eval_recall": 0.9397689768976898, | |
| "eval_runtime": 22.7535, | |
| "eval_samples_per_second": 206.649, | |
| "eval_steps_per_second": 4.307, | |
| "step": 2968 | |
| }, | |
| { | |
| "epoch": 4.0431266846361185, | |
| "grad_norm": 8.28526782989502, | |
| "learning_rate": 3.827493261455526e-06, | |
| "loss": 0.155, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 4.177897574123989, | |
| "grad_norm": 8.550823211669922, | |
| "learning_rate": 3.2884097035040433e-06, | |
| "loss": 0.1385, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 4.3126684636118595, | |
| "grad_norm": 7.248845100402832, | |
| "learning_rate": 2.749326145552561e-06, | |
| "loss": 0.1245, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 4.44743935309973, | |
| "grad_norm": 4.223452091217041, | |
| "learning_rate": 2.2102425876010783e-06, | |
| "loss": 0.1316, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 4.5822102425876015, | |
| "grad_norm": 26.39322853088379, | |
| "learning_rate": 1.6711590296495958e-06, | |
| "loss": 0.1347, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 4.716981132075472, | |
| "grad_norm": 9.451475143432617, | |
| "learning_rate": 1.1320754716981133e-06, | |
| "loss": 0.1373, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 4.8517520215633425, | |
| "grad_norm": 7.989397048950195, | |
| "learning_rate": 5.929919137466308e-07, | |
| "loss": 0.127, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 4.986522911051213, | |
| "grad_norm": 8.661871910095215, | |
| "learning_rate": 5.3908355795148254e-08, | |
| "loss": 0.1288, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8538919608677159, | |
| "eval_confusion_matrix": [ | |
| [ | |
| 1721, | |
| 557 | |
| ], | |
| [ | |
| 130, | |
| 2294 | |
| ] | |
| ], | |
| "eval_f1": 0.8697630331753554, | |
| "eval_loss": 0.6569812893867493, | |
| "eval_precision": 0.8046299544019643, | |
| "eval_recall": 0.9463696369636964, | |
| "eval_runtime": 22.4622, | |
| "eval_samples_per_second": 209.329, | |
| "eval_steps_per_second": 4.363, | |
| "step": 3710 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 3710, | |
| "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": 2.8848225720991744e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |