|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 18.0, |
|
"eval_steps": 500, |
|
"global_step": 4500, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.4, |
|
"grad_norm": 0.6251392960548401, |
|
"learning_rate": 5e-05, |
|
"loss": 2.427, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"grad_norm": 0.6590011715888977, |
|
"learning_rate": 5e-05, |
|
"loss": 2.2748, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy": 0.5142539682539683, |
|
"eval_loss": 2.336690902709961, |
|
"eval_runtime": 5.4733, |
|
"eval_samples_per_second": 91.353, |
|
"eval_steps_per_second": 11.51, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_exact_match": 16.8, |
|
"eval_f1": 24.850246583481884, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 1.2, |
|
"grad_norm": 0.8194239139556885, |
|
"learning_rate": 5e-05, |
|
"loss": 2.2322, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 1.6, |
|
"grad_norm": 0.9867294430732727, |
|
"learning_rate": 5e-05, |
|
"loss": 2.1904, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"grad_norm": 0.9660400152206421, |
|
"learning_rate": 5e-05, |
|
"loss": 2.1657, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.5168253968253969, |
|
"eval_loss": 2.323688507080078, |
|
"eval_runtime": 5.1683, |
|
"eval_samples_per_second": 96.744, |
|
"eval_steps_per_second": 12.19, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_exact_match": 17.2, |
|
"eval_f1": 24.620036075036083, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 2.4, |
|
"grad_norm": 1.281128168106079, |
|
"learning_rate": 5e-05, |
|
"loss": 2.0251, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 2.8, |
|
"grad_norm": 1.3365116119384766, |
|
"learning_rate": 5e-05, |
|
"loss": 2.0158, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy": 0.5156507936507937, |
|
"eval_loss": 2.3560824394226074, |
|
"eval_runtime": 5.4675, |
|
"eval_samples_per_second": 91.449, |
|
"eval_steps_per_second": 11.523, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_exact_match": 19.0, |
|
"eval_f1": 27.242790688843332, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 3.2, |
|
"grad_norm": 1.7840828895568848, |
|
"learning_rate": 5e-05, |
|
"loss": 1.9394, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 3.6, |
|
"grad_norm": 1.8030190467834473, |
|
"learning_rate": 5e-05, |
|
"loss": 1.8502, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"grad_norm": 1.9682222604751587, |
|
"learning_rate": 5e-05, |
|
"loss": 1.8682, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.5131746031746032, |
|
"eval_loss": 2.4071946144104004, |
|
"eval_runtime": 5.2825, |
|
"eval_samples_per_second": 94.653, |
|
"eval_steps_per_second": 11.926, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_exact_match": 17.6, |
|
"eval_f1": 26.010988455988468, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 4.4, |
|
"grad_norm": 2.138505220413208, |
|
"learning_rate": 5e-05, |
|
"loss": 1.6795, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 4.8, |
|
"grad_norm": 2.3212926387786865, |
|
"learning_rate": 5e-05, |
|
"loss": 1.7036, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_accuracy": 0.5099365079365079, |
|
"eval_loss": 2.494770050048828, |
|
"eval_runtime": 5.1799, |
|
"eval_samples_per_second": 96.528, |
|
"eval_steps_per_second": 12.162, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_exact_match": 17.2, |
|
"eval_f1": 25.84752525252526, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 5.2, |
|
"grad_norm": 2.4878485202789307, |
|
"learning_rate": 5e-05, |
|
"loss": 1.6305, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 5.6, |
|
"grad_norm": 2.599329710006714, |
|
"learning_rate": 5e-05, |
|
"loss": 1.5477, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"grad_norm": 2.5542759895324707, |
|
"learning_rate": 5e-05, |
|
"loss": 1.5708, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.5065079365079365, |
|
"eval_loss": 2.5797038078308105, |
|
"eval_runtime": 5.6827, |
|
"eval_samples_per_second": 87.986, |
|
"eval_steps_per_second": 11.086, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_exact_match": 17.0, |
|
"eval_f1": 25.310858585858586, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 6.4, |
|
"grad_norm": 3.1968846321105957, |
|
"learning_rate": 5e-05, |
|
"loss": 1.3896, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 6.8, |
|
"grad_norm": 3.8637375831604004, |
|
"learning_rate": 5e-05, |
|
"loss": 1.421, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"eval_accuracy": 0.5038412698412699, |
|
"eval_loss": 2.6977317333221436, |
|
"eval_runtime": 5.3312, |
|
"eval_samples_per_second": 93.787, |
|
"eval_steps_per_second": 11.817, |
|
"step": 1750 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"eval_exact_match": 16.0, |
|
"eval_f1": 23.99441558441559, |
|
"step": 1750 |
|
}, |
|
{ |
|
"epoch": 7.2, |
|
"grad_norm": 3.468459367752075, |
|
"learning_rate": 5e-05, |
|
"loss": 1.3366, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 7.6, |
|
"grad_norm": 3.5727667808532715, |
|
"learning_rate": 5e-05, |
|
"loss": 1.2585, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"grad_norm": 3.508258819580078, |
|
"learning_rate": 5e-05, |
|
"loss": 1.296, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.5013333333333333, |
|
"eval_loss": 2.786928653717041, |
|
"eval_runtime": 5.8398, |
|
"eval_samples_per_second": 85.62, |
|
"eval_steps_per_second": 10.788, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_exact_match": 15.4, |
|
"eval_f1": 22.542871572871572, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 8.4, |
|
"grad_norm": 4.042112350463867, |
|
"learning_rate": 5e-05, |
|
"loss": 1.1084, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 8.8, |
|
"grad_norm": 3.7548155784606934, |
|
"learning_rate": 5e-05, |
|
"loss": 1.1457, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"eval_accuracy": 0.49692063492063493, |
|
"eval_loss": 2.9215197563171387, |
|
"eval_runtime": 5.1412, |
|
"eval_samples_per_second": 97.254, |
|
"eval_steps_per_second": 12.254, |
|
"step": 2250 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"eval_exact_match": 15.0, |
|
"eval_f1": 21.59087301587302, |
|
"step": 2250 |
|
}, |
|
{ |
|
"epoch": 9.2, |
|
"grad_norm": 4.0875773429870605, |
|
"learning_rate": 5e-05, |
|
"loss": 1.0677, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 9.6, |
|
"grad_norm": 4.143901348114014, |
|
"learning_rate": 5e-05, |
|
"loss": 1.0028, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"grad_norm": 5.0514445304870605, |
|
"learning_rate": 5e-05, |
|
"loss": 1.0401, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_accuracy": 0.4953015873015873, |
|
"eval_loss": 3.054328203201294, |
|
"eval_runtime": 5.4195, |
|
"eval_samples_per_second": 92.259, |
|
"eval_steps_per_second": 11.625, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_exact_match": 15.4, |
|
"eval_f1": 21.505119880119878, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 10.4, |
|
"grad_norm": 3.8740394115448, |
|
"learning_rate": 5e-05, |
|
"loss": 0.8595, |
|
"step": 2600 |
|
}, |
|
{ |
|
"epoch": 10.8, |
|
"grad_norm": 4.170004844665527, |
|
"learning_rate": 5e-05, |
|
"loss": 0.9186, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"eval_accuracy": 0.49298412698412697, |
|
"eval_loss": 3.177030324935913, |
|
"eval_runtime": 5.9501, |
|
"eval_samples_per_second": 84.033, |
|
"eval_steps_per_second": 10.588, |
|
"step": 2750 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"eval_exact_match": 15.6, |
|
"eval_f1": 20.912753357753356, |
|
"step": 2750 |
|
}, |
|
{ |
|
"epoch": 11.2, |
|
"grad_norm": 4.2533392906188965, |
|
"learning_rate": 5e-05, |
|
"loss": 0.8411, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 11.6, |
|
"grad_norm": 4.165424346923828, |
|
"learning_rate": 5e-05, |
|
"loss": 0.7934, |
|
"step": 2900 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"grad_norm": 4.517414093017578, |
|
"learning_rate": 5e-05, |
|
"loss": 0.8386, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_accuracy": 0.4916190476190476, |
|
"eval_loss": 3.2663722038269043, |
|
"eval_runtime": 5.1518, |
|
"eval_samples_per_second": 97.053, |
|
"eval_steps_per_second": 12.229, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_exact_match": 15.2, |
|
"eval_f1": 20.465463425463433, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 12.4, |
|
"grad_norm": 4.059298515319824, |
|
"learning_rate": 5e-05, |
|
"loss": 0.6908, |
|
"step": 3100 |
|
}, |
|
{ |
|
"epoch": 12.8, |
|
"grad_norm": 3.4847490787506104, |
|
"learning_rate": 5e-05, |
|
"loss": 0.7326, |
|
"step": 3200 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"eval_accuracy": 0.49047619047619045, |
|
"eval_loss": 3.38826584815979, |
|
"eval_runtime": 5.2085, |
|
"eval_samples_per_second": 95.997, |
|
"eval_steps_per_second": 12.096, |
|
"step": 3250 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"eval_exact_match": 14.0, |
|
"eval_f1": 20.144243697478988, |
|
"step": 3250 |
|
}, |
|
{ |
|
"epoch": 13.2, |
|
"grad_norm": 4.07555627822876, |
|
"learning_rate": 5e-05, |
|
"loss": 0.6901, |
|
"step": 3300 |
|
}, |
|
{ |
|
"epoch": 13.6, |
|
"grad_norm": 3.918684482574463, |
|
"learning_rate": 5e-05, |
|
"loss": 0.6432, |
|
"step": 3400 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"grad_norm": 4.3187384605407715, |
|
"learning_rate": 5e-05, |
|
"loss": 0.6762, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"eval_accuracy": 0.4887936507936508, |
|
"eval_loss": 3.466782808303833, |
|
"eval_runtime": 5.5309, |
|
"eval_samples_per_second": 90.402, |
|
"eval_steps_per_second": 11.391, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"eval_exact_match": 14.6, |
|
"eval_f1": 20.548629148629157, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 14.4, |
|
"grad_norm": 3.270817995071411, |
|
"learning_rate": 5e-05, |
|
"loss": 0.5785, |
|
"step": 3600 |
|
}, |
|
{ |
|
"epoch": 14.8, |
|
"grad_norm": 3.882906436920166, |
|
"learning_rate": 5e-05, |
|
"loss": 0.6117, |
|
"step": 3700 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"eval_accuracy": 0.48904761904761906, |
|
"eval_loss": 3.5602962970733643, |
|
"eval_runtime": 5.4424, |
|
"eval_samples_per_second": 91.871, |
|
"eval_steps_per_second": 11.576, |
|
"step": 3750 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"eval_exact_match": 14.8, |
|
"eval_f1": 20.941592997645635, |
|
"step": 3750 |
|
}, |
|
{ |
|
"epoch": 15.2, |
|
"grad_norm": 4.002102851867676, |
|
"learning_rate": 5e-05, |
|
"loss": 0.5725, |
|
"step": 3800 |
|
}, |
|
{ |
|
"epoch": 15.6, |
|
"grad_norm": 2.8693177700042725, |
|
"learning_rate": 5e-05, |
|
"loss": 0.5581, |
|
"step": 3900 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"grad_norm": 4.3405256271362305, |
|
"learning_rate": 5e-05, |
|
"loss": 0.5729, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"eval_accuracy": 0.48742857142857143, |
|
"eval_loss": 3.6583478450775146, |
|
"eval_runtime": 5.1711, |
|
"eval_samples_per_second": 96.692, |
|
"eval_steps_per_second": 12.183, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"eval_exact_match": 14.8, |
|
"eval_f1": 20.59385852709383, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 16.4, |
|
"grad_norm": 4.715936183929443, |
|
"learning_rate": 5e-05, |
|
"loss": 0.5028, |
|
"step": 4100 |
|
}, |
|
{ |
|
"epoch": 16.8, |
|
"grad_norm": 3.7355644702911377, |
|
"learning_rate": 5e-05, |
|
"loss": 0.5317, |
|
"step": 4200 |
|
}, |
|
{ |
|
"epoch": 17.0, |
|
"eval_accuracy": 0.4873968253968254, |
|
"eval_loss": 3.6984667778015137, |
|
"eval_runtime": 5.4593, |
|
"eval_samples_per_second": 91.587, |
|
"eval_steps_per_second": 11.54, |
|
"step": 4250 |
|
}, |
|
{ |
|
"epoch": 17.0, |
|
"eval_exact_match": 13.8, |
|
"eval_f1": 20.517505228159344, |
|
"step": 4250 |
|
}, |
|
{ |
|
"epoch": 17.2, |
|
"grad_norm": 2.7771668434143066, |
|
"learning_rate": 5e-05, |
|
"loss": 0.505, |
|
"step": 4300 |
|
}, |
|
{ |
|
"epoch": 17.6, |
|
"grad_norm": 3.6721012592315674, |
|
"learning_rate": 5e-05, |
|
"loss": 0.496, |
|
"step": 4400 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"grad_norm": 4.477765083312988, |
|
"learning_rate": 5e-05, |
|
"loss": 0.5109, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"eval_accuracy": 0.4871111111111111, |
|
"eval_loss": 3.7672173976898193, |
|
"eval_runtime": 5.1624, |
|
"eval_samples_per_second": 96.855, |
|
"eval_steps_per_second": 12.204, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"eval_exact_match": 13.2, |
|
"eval_f1": 19.935008488243795, |
|
"step": 4500 |
|
} |
|
], |
|
"logging_steps": 100, |
|
"max_steps": 12500, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 50, |
|
"save_steps": 500, |
|
"stateful_callbacks": { |
|
"TrainerControl": { |
|
"args": { |
|
"should_epoch_stop": false, |
|
"should_evaluate": false, |
|
"should_log": false, |
|
"should_save": true, |
|
"should_training_stop": false |
|
}, |
|
"attributes": {} |
|
} |
|
}, |
|
"total_flos": 1.1641847245543834e+17, |
|
"train_batch_size": 1, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|