|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 19.0, |
|
"eval_steps": 500, |
|
"global_step": 4750, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.4, |
|
"learning_rate": 0.0001, |
|
"loss": 2.0446, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"learning_rate": 0.0001, |
|
"loss": 1.7752, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_accuracy": 0.6076455696202532, |
|
"eval_loss": 1.8067126274108887, |
|
"eval_runtime": 5.0018, |
|
"eval_samples_per_second": 99.965, |
|
"eval_steps_per_second": 12.596, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"eval_exact_match": 24.4, |
|
"eval_f1": 35.27913510466144, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 1.2, |
|
"learning_rate": 0.0001, |
|
"loss": 1.7141, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 1.6, |
|
"learning_rate": 0.0001, |
|
"loss": 1.6576, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"learning_rate": 0.0001, |
|
"loss": 1.6469, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_accuracy": 0.607746835443038, |
|
"eval_loss": 1.810449481010437, |
|
"eval_runtime": 4.5227, |
|
"eval_samples_per_second": 110.552, |
|
"eval_steps_per_second": 13.93, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 2.0, |
|
"eval_exact_match": 24.2, |
|
"eval_f1": 33.990575470281364, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 2.4, |
|
"learning_rate": 0.0001, |
|
"loss": 1.4453, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 2.8, |
|
"learning_rate": 0.0001, |
|
"loss": 1.474, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_accuracy": 0.6055696202531645, |
|
"eval_loss": 1.8687865734100342, |
|
"eval_runtime": 4.7138, |
|
"eval_samples_per_second": 106.072, |
|
"eval_steps_per_second": 13.365, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"eval_exact_match": 23.0, |
|
"eval_f1": 33.30356143856145, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 3.2, |
|
"learning_rate": 0.0001, |
|
"loss": 1.3572, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 3.6, |
|
"learning_rate": 0.0001, |
|
"loss": 1.2472, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"learning_rate": 0.0001, |
|
"loss": 1.2742, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_accuracy": 0.6030886075949367, |
|
"eval_loss": 1.9776239395141602, |
|
"eval_runtime": 4.5553, |
|
"eval_samples_per_second": 109.762, |
|
"eval_steps_per_second": 13.83, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 4.0, |
|
"eval_exact_match": 22.6, |
|
"eval_f1": 32.07458596958597, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 4.4, |
|
"learning_rate": 0.0001, |
|
"loss": 1.0303, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 4.8, |
|
"learning_rate": 0.0001, |
|
"loss": 1.0851, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_accuracy": 0.599873417721519, |
|
"eval_loss": 2.081591844558716, |
|
"eval_runtime": 5.0717, |
|
"eval_samples_per_second": 98.586, |
|
"eval_steps_per_second": 12.422, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 5.0, |
|
"eval_exact_match": 21.8, |
|
"eval_f1": 30.650287424993316, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 5.2, |
|
"learning_rate": 0.0001, |
|
"loss": 0.9745, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 5.6, |
|
"learning_rate": 0.0001, |
|
"loss": 0.8784, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"learning_rate": 0.0001, |
|
"loss": 0.9085, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_accuracy": 0.5969873417721518, |
|
"eval_loss": 2.2579739093780518, |
|
"eval_runtime": 4.9122, |
|
"eval_samples_per_second": 101.788, |
|
"eval_steps_per_second": 12.825, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 6.0, |
|
"eval_exact_match": 21.8, |
|
"eval_f1": 30.116953765189066, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 6.4, |
|
"learning_rate": 0.0001, |
|
"loss": 0.6992, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 6.8, |
|
"learning_rate": 0.0001, |
|
"loss": 0.7383, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"eval_accuracy": 0.5939240506329114, |
|
"eval_loss": 2.3878135681152344, |
|
"eval_runtime": 4.4125, |
|
"eval_samples_per_second": 113.315, |
|
"eval_steps_per_second": 14.278, |
|
"step": 1750 |
|
}, |
|
{ |
|
"epoch": 7.0, |
|
"eval_exact_match": 22.2, |
|
"eval_f1": 30.4995441160147, |
|
"step": 1750 |
|
}, |
|
{ |
|
"epoch": 7.2, |
|
"learning_rate": 0.0001, |
|
"loss": 0.6582, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 7.6, |
|
"learning_rate": 0.0001, |
|
"loss": 0.5881, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"learning_rate": 0.0001, |
|
"loss": 0.6143, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_accuracy": 0.5923544303797469, |
|
"eval_loss": 2.5656919479370117, |
|
"eval_runtime": 4.3988, |
|
"eval_samples_per_second": 113.667, |
|
"eval_steps_per_second": 14.322, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 8.0, |
|
"eval_exact_match": 21.8, |
|
"eval_f1": 30.237513923984515, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 8.4, |
|
"learning_rate": 0.0001, |
|
"loss": 0.4727, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 8.8, |
|
"learning_rate": 0.0001, |
|
"loss": 0.5047, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"eval_accuracy": 0.5898987341772152, |
|
"eval_loss": 2.6870768070220947, |
|
"eval_runtime": 4.5648, |
|
"eval_samples_per_second": 109.534, |
|
"eval_steps_per_second": 13.801, |
|
"step": 2250 |
|
}, |
|
{ |
|
"epoch": 9.0, |
|
"eval_exact_match": 21.0, |
|
"eval_f1": 29.401002657473246, |
|
"step": 2250 |
|
}, |
|
{ |
|
"epoch": 9.2, |
|
"learning_rate": 0.0001, |
|
"loss": 0.4595, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 9.6, |
|
"learning_rate": 0.0001, |
|
"loss": 0.4269, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"learning_rate": 0.0001, |
|
"loss": 0.4457, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_accuracy": 0.5886835443037974, |
|
"eval_loss": 2.8029491901397705, |
|
"eval_runtime": 4.7023, |
|
"eval_samples_per_second": 106.332, |
|
"eval_steps_per_second": 13.398, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 10.0, |
|
"eval_exact_match": 20.8, |
|
"eval_f1": 29.73924219571279, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 10.4, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3725, |
|
"step": 2600 |
|
}, |
|
{ |
|
"epoch": 10.8, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3942, |
|
"step": 2700 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"eval_accuracy": 0.5877215189873418, |
|
"eval_loss": 2.9129838943481445, |
|
"eval_runtime": 5.1228, |
|
"eval_samples_per_second": 97.603, |
|
"eval_steps_per_second": 12.298, |
|
"step": 2750 |
|
}, |
|
{ |
|
"epoch": 11.0, |
|
"eval_exact_match": 17.8, |
|
"eval_f1": 27.10153509378855, |
|
"step": 2750 |
|
}, |
|
{ |
|
"epoch": 11.2, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3736, |
|
"step": 2800 |
|
}, |
|
{ |
|
"epoch": 11.6, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3612, |
|
"step": 2900 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"learning_rate": 0.0001, |
|
"loss": 0.375, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_accuracy": 0.5900759493670886, |
|
"eval_loss": 2.9960711002349854, |
|
"eval_runtime": 4.5558, |
|
"eval_samples_per_second": 109.749, |
|
"eval_steps_per_second": 13.828, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 12.0, |
|
"eval_exact_match": 15.8, |
|
"eval_f1": 23.735175071397414, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 12.4, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3326, |
|
"step": 3100 |
|
}, |
|
{ |
|
"epoch": 12.8, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3503, |
|
"step": 3200 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"eval_accuracy": 0.5899493670886076, |
|
"eval_loss": 3.0212697982788086, |
|
"eval_runtime": 4.4053, |
|
"eval_samples_per_second": 113.5, |
|
"eval_steps_per_second": 14.301, |
|
"step": 3250 |
|
}, |
|
{ |
|
"epoch": 13.0, |
|
"eval_exact_match": 4.0, |
|
"eval_f1": 9.841706398176983, |
|
"step": 3250 |
|
}, |
|
{ |
|
"epoch": 13.2, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3386, |
|
"step": 3300 |
|
}, |
|
{ |
|
"epoch": 13.6, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3305, |
|
"step": 3400 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3442, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"eval_accuracy": 0.5883291139240506, |
|
"eval_loss": 3.077633857727051, |
|
"eval_runtime": 4.4554, |
|
"eval_samples_per_second": 112.224, |
|
"eval_steps_per_second": 14.14, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 14.0, |
|
"eval_exact_match": 5.0, |
|
"eval_f1": 11.835427535427533, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 14.4, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3131, |
|
"step": 3600 |
|
}, |
|
{ |
|
"epoch": 14.8, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3334, |
|
"step": 3700 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"eval_accuracy": 0.5892911392405064, |
|
"eval_loss": 3.1445634365081787, |
|
"eval_runtime": 4.5617, |
|
"eval_samples_per_second": 109.609, |
|
"eval_steps_per_second": 13.811, |
|
"step": 3750 |
|
}, |
|
{ |
|
"epoch": 15.0, |
|
"eval_exact_match": 1.2, |
|
"eval_f1": 4.139209256423578, |
|
"step": 3750 |
|
}, |
|
{ |
|
"epoch": 15.2, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3191, |
|
"step": 3800 |
|
}, |
|
{ |
|
"epoch": 15.6, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3173, |
|
"step": 3900 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3252, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"eval_accuracy": 0.5877721518987342, |
|
"eval_loss": 3.169813871383667, |
|
"eval_runtime": 4.4332, |
|
"eval_samples_per_second": 112.785, |
|
"eval_steps_per_second": 14.211, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 16.0, |
|
"eval_exact_match": 1.6, |
|
"eval_f1": 5.750371546798473, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 16.4, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3007, |
|
"step": 4100 |
|
}, |
|
{ |
|
"epoch": 16.8, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3185, |
|
"step": 4200 |
|
}, |
|
{ |
|
"epoch": 17.0, |
|
"eval_accuracy": 0.5890379746835444, |
|
"eval_loss": 3.1473915576934814, |
|
"eval_runtime": 4.7115, |
|
"eval_samples_per_second": 106.124, |
|
"eval_steps_per_second": 13.372, |
|
"step": 4250 |
|
}, |
|
{ |
|
"epoch": 17.0, |
|
"eval_exact_match": 1.2, |
|
"eval_f1": 4.06053598873457, |
|
"step": 4250 |
|
}, |
|
{ |
|
"epoch": 17.2, |
|
"learning_rate": 0.0001, |
|
"loss": 0.304, |
|
"step": 4300 |
|
}, |
|
{ |
|
"epoch": 17.6, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3042, |
|
"step": 4400 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"learning_rate": 0.0001, |
|
"loss": 0.315, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"eval_accuracy": 0.5886835443037974, |
|
"eval_loss": 3.2847023010253906, |
|
"eval_runtime": 4.3996, |
|
"eval_samples_per_second": 113.647, |
|
"eval_steps_per_second": 14.32, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 18.0, |
|
"eval_exact_match": 1.4, |
|
"eval_f1": 3.727712915289645, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 18.4, |
|
"learning_rate": 0.0001, |
|
"loss": 0.2938, |
|
"step": 4600 |
|
}, |
|
{ |
|
"epoch": 18.8, |
|
"learning_rate": 0.0001, |
|
"loss": 0.3047, |
|
"step": 4700 |
|
}, |
|
{ |
|
"epoch": 19.0, |
|
"eval_accuracy": 0.589493670886076, |
|
"eval_loss": 3.2327334880828857, |
|
"eval_runtime": 4.4051, |
|
"eval_samples_per_second": 113.505, |
|
"eval_steps_per_second": 14.302, |
|
"step": 4750 |
|
}, |
|
{ |
|
"epoch": 19.0, |
|
"eval_exact_match": 0.2, |
|
"eval_f1": 2.292747940989559, |
|
"step": 4750 |
|
} |
|
], |
|
"logging_steps": 100, |
|
"max_steps": 12500, |
|
"num_train_epochs": 50, |
|
"save_steps": 500, |
|
"total_flos": 3.0373712498786304e+17, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|