|
{ |
|
"best_metric": 0.39623910188674927, |
|
"best_model_checkpoint": "mikhail-panzo/ceb_b64_le5_s4000/checkpoint-2500", |
|
"epoch": 99.00990099009901, |
|
"eval_steps": 500, |
|
"global_step": 2500, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 1.9801980198019802, |
|
"grad_norm": 2.883972644805908, |
|
"learning_rate": 2.5000000000000004e-07, |
|
"loss": 0.8004, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 3.9603960396039604, |
|
"grad_norm": 2.5420947074890137, |
|
"learning_rate": 5.000000000000001e-07, |
|
"loss": 0.77, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 5.9405940594059405, |
|
"grad_norm": 1.3930779695510864, |
|
"learning_rate": 7.5e-07, |
|
"loss": 0.7525, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 7.920792079207921, |
|
"grad_norm": 1.8414520025253296, |
|
"learning_rate": 1.0000000000000002e-06, |
|
"loss": 0.722, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 9.900990099009901, |
|
"grad_norm": 2.0665407180786133, |
|
"learning_rate": 1.25e-06, |
|
"loss": 0.7076, |
|
"step": 250 |
|
}, |
|
{ |
|
"epoch": 11.881188118811881, |
|
"grad_norm": 1.2595587968826294, |
|
"learning_rate": 1.5e-06, |
|
"loss": 0.6757, |
|
"step": 300 |
|
}, |
|
{ |
|
"epoch": 13.861386138613861, |
|
"grad_norm": 1.4587587118148804, |
|
"learning_rate": 1.75e-06, |
|
"loss": 0.6562, |
|
"step": 350 |
|
}, |
|
{ |
|
"epoch": 15.841584158415841, |
|
"grad_norm": 2.241546154022217, |
|
"learning_rate": 2.0000000000000003e-06, |
|
"loss": 0.6016, |
|
"step": 400 |
|
}, |
|
{ |
|
"epoch": 17.821782178217823, |
|
"grad_norm": 1.4243369102478027, |
|
"learning_rate": 2.25e-06, |
|
"loss": 0.5593, |
|
"step": 450 |
|
}, |
|
{ |
|
"epoch": 19.801980198019802, |
|
"grad_norm": 1.6889928579330444, |
|
"learning_rate": 2.5e-06, |
|
"loss": 0.536, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 19.801980198019802, |
|
"eval_loss": 0.4640245735645294, |
|
"eval_runtime": 6.7454, |
|
"eval_samples_per_second": 26.685, |
|
"eval_steps_per_second": 3.41, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 21.782178217821784, |
|
"grad_norm": 1.1601781845092773, |
|
"learning_rate": 2.7500000000000004e-06, |
|
"loss": 0.5198, |
|
"step": 550 |
|
}, |
|
{ |
|
"epoch": 23.762376237623762, |
|
"grad_norm": 1.1969332695007324, |
|
"learning_rate": 3e-06, |
|
"loss": 0.5114, |
|
"step": 600 |
|
}, |
|
{ |
|
"epoch": 25.742574257425744, |
|
"grad_norm": 1.1022064685821533, |
|
"learning_rate": 3.2500000000000002e-06, |
|
"loss": 0.5092, |
|
"step": 650 |
|
}, |
|
{ |
|
"epoch": 27.722772277227723, |
|
"grad_norm": 1.2418344020843506, |
|
"learning_rate": 3.5e-06, |
|
"loss": 0.4951, |
|
"step": 700 |
|
}, |
|
{ |
|
"epoch": 29.702970297029704, |
|
"grad_norm": 1.1083300113677979, |
|
"learning_rate": 3.7500000000000005e-06, |
|
"loss": 0.4911, |
|
"step": 750 |
|
}, |
|
{ |
|
"epoch": 31.683168316831683, |
|
"grad_norm": 1.2230323553085327, |
|
"learning_rate": 4.000000000000001e-06, |
|
"loss": 0.4866, |
|
"step": 800 |
|
}, |
|
{ |
|
"epoch": 33.663366336633665, |
|
"grad_norm": 1.1232423782348633, |
|
"learning_rate": 4.25e-06, |
|
"loss": 0.4876, |
|
"step": 850 |
|
}, |
|
{ |
|
"epoch": 35.64356435643565, |
|
"grad_norm": 0.9751402139663696, |
|
"learning_rate": 4.5e-06, |
|
"loss": 0.4772, |
|
"step": 900 |
|
}, |
|
{ |
|
"epoch": 37.62376237623762, |
|
"grad_norm": 1.0373197793960571, |
|
"learning_rate": 4.75e-06, |
|
"loss": 0.4791, |
|
"step": 950 |
|
}, |
|
{ |
|
"epoch": 39.603960396039604, |
|
"grad_norm": 1.0164889097213745, |
|
"learning_rate": 5e-06, |
|
"loss": 0.4674, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 39.603960396039604, |
|
"eval_loss": 0.4247289001941681, |
|
"eval_runtime": 6.5527, |
|
"eval_samples_per_second": 27.47, |
|
"eval_steps_per_second": 3.51, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 41.584158415841586, |
|
"grad_norm": 1.087668538093567, |
|
"learning_rate": 5.2500000000000006e-06, |
|
"loss": 0.4754, |
|
"step": 1050 |
|
}, |
|
{ |
|
"epoch": 43.56435643564357, |
|
"grad_norm": 0.891859769821167, |
|
"learning_rate": 5.500000000000001e-06, |
|
"loss": 0.4677, |
|
"step": 1100 |
|
}, |
|
{ |
|
"epoch": 45.54455445544554, |
|
"grad_norm": 1.018160343170166, |
|
"learning_rate": 5.75e-06, |
|
"loss": 0.4642, |
|
"step": 1150 |
|
}, |
|
{ |
|
"epoch": 47.524752475247524, |
|
"grad_norm": 1.4765816926956177, |
|
"learning_rate": 6e-06, |
|
"loss": 0.4602, |
|
"step": 1200 |
|
}, |
|
{ |
|
"epoch": 49.504950495049506, |
|
"grad_norm": 1.3534387350082397, |
|
"learning_rate": 6.25e-06, |
|
"loss": 0.4634, |
|
"step": 1250 |
|
}, |
|
{ |
|
"epoch": 51.48514851485149, |
|
"grad_norm": 1.031537413597107, |
|
"learning_rate": 6.5000000000000004e-06, |
|
"loss": 0.4567, |
|
"step": 1300 |
|
}, |
|
{ |
|
"epoch": 53.46534653465346, |
|
"grad_norm": 1.0172531604766846, |
|
"learning_rate": 6.750000000000001e-06, |
|
"loss": 0.4592, |
|
"step": 1350 |
|
}, |
|
{ |
|
"epoch": 55.445544554455445, |
|
"grad_norm": 0.8680336475372314, |
|
"learning_rate": 7e-06, |
|
"loss": 0.4583, |
|
"step": 1400 |
|
}, |
|
{ |
|
"epoch": 57.42574257425743, |
|
"grad_norm": 1.2397912740707397, |
|
"learning_rate": 7.25e-06, |
|
"loss": 0.4459, |
|
"step": 1450 |
|
}, |
|
{ |
|
"epoch": 59.40594059405941, |
|
"grad_norm": 1.021185040473938, |
|
"learning_rate": 7.500000000000001e-06, |
|
"loss": 0.4479, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 59.40594059405941, |
|
"eval_loss": 0.4102235734462738, |
|
"eval_runtime": 7.6974, |
|
"eval_samples_per_second": 23.384, |
|
"eval_steps_per_second": 2.988, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 61.386138613861384, |
|
"grad_norm": 1.005591630935669, |
|
"learning_rate": 7.75e-06, |
|
"loss": 0.4469, |
|
"step": 1550 |
|
}, |
|
{ |
|
"epoch": 63.366336633663366, |
|
"grad_norm": 0.8193312287330627, |
|
"learning_rate": 8.000000000000001e-06, |
|
"loss": 0.4425, |
|
"step": 1600 |
|
}, |
|
{ |
|
"epoch": 65.34653465346534, |
|
"grad_norm": 0.9143276810646057, |
|
"learning_rate": 8.25e-06, |
|
"loss": 0.4424, |
|
"step": 1650 |
|
}, |
|
{ |
|
"epoch": 67.32673267326733, |
|
"grad_norm": 1.0127445459365845, |
|
"learning_rate": 8.5e-06, |
|
"loss": 0.4446, |
|
"step": 1700 |
|
}, |
|
{ |
|
"epoch": 69.3069306930693, |
|
"grad_norm": 1.045551061630249, |
|
"learning_rate": 8.750000000000001e-06, |
|
"loss": 0.4417, |
|
"step": 1750 |
|
}, |
|
{ |
|
"epoch": 71.2871287128713, |
|
"grad_norm": 0.9629473090171814, |
|
"learning_rate": 9e-06, |
|
"loss": 0.4362, |
|
"step": 1800 |
|
}, |
|
{ |
|
"epoch": 73.26732673267327, |
|
"grad_norm": 0.9655708074569702, |
|
"learning_rate": 9.250000000000001e-06, |
|
"loss": 0.4393, |
|
"step": 1850 |
|
}, |
|
{ |
|
"epoch": 75.24752475247524, |
|
"grad_norm": 1.167992115020752, |
|
"learning_rate": 9.5e-06, |
|
"loss": 0.4373, |
|
"step": 1900 |
|
}, |
|
{ |
|
"epoch": 77.22772277227723, |
|
"grad_norm": 0.9866817593574524, |
|
"learning_rate": 9.75e-06, |
|
"loss": 0.4369, |
|
"step": 1950 |
|
}, |
|
{ |
|
"epoch": 79.20792079207921, |
|
"grad_norm": 0.8923723101615906, |
|
"learning_rate": 1e-05, |
|
"loss": 0.4356, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 79.20792079207921, |
|
"eval_loss": 0.4031977355480194, |
|
"eval_runtime": 7.0669, |
|
"eval_samples_per_second": 25.471, |
|
"eval_steps_per_second": 3.255, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 81.18811881188118, |
|
"grad_norm": 1.0076066255569458, |
|
"learning_rate": 9.755e-06, |
|
"loss": 0.4365, |
|
"step": 2050 |
|
}, |
|
{ |
|
"epoch": 83.16831683168317, |
|
"grad_norm": 0.8251845836639404, |
|
"learning_rate": 9.505000000000001e-06, |
|
"loss": 0.4342, |
|
"step": 2100 |
|
}, |
|
{ |
|
"epoch": 85.14851485148515, |
|
"grad_norm": 1.029945731163025, |
|
"learning_rate": 9.255e-06, |
|
"loss": 0.4325, |
|
"step": 2150 |
|
}, |
|
{ |
|
"epoch": 87.12871287128714, |
|
"grad_norm": 1.0486702919006348, |
|
"learning_rate": 9.005000000000001e-06, |
|
"loss": 0.4223, |
|
"step": 2200 |
|
}, |
|
{ |
|
"epoch": 89.10891089108911, |
|
"grad_norm": 0.9815123677253723, |
|
"learning_rate": 8.755e-06, |
|
"loss": 0.427, |
|
"step": 2250 |
|
}, |
|
{ |
|
"epoch": 91.08910891089108, |
|
"grad_norm": 1.0457566976547241, |
|
"learning_rate": 8.505e-06, |
|
"loss": 0.4297, |
|
"step": 2300 |
|
}, |
|
{ |
|
"epoch": 93.06930693069307, |
|
"grad_norm": 1.0108842849731445, |
|
"learning_rate": 8.255000000000001e-06, |
|
"loss": 0.4262, |
|
"step": 2350 |
|
}, |
|
{ |
|
"epoch": 95.04950495049505, |
|
"grad_norm": 1.0584673881530762, |
|
"learning_rate": 8.005e-06, |
|
"loss": 0.4238, |
|
"step": 2400 |
|
}, |
|
{ |
|
"epoch": 97.02970297029702, |
|
"grad_norm": 1.346966028213501, |
|
"learning_rate": 7.755000000000001e-06, |
|
"loss": 0.4217, |
|
"step": 2450 |
|
}, |
|
{ |
|
"epoch": 99.00990099009901, |
|
"grad_norm": 0.9860787987709045, |
|
"learning_rate": 7.505e-06, |
|
"loss": 0.4277, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 99.00990099009901, |
|
"eval_loss": 0.39623910188674927, |
|
"eval_runtime": 7.3733, |
|
"eval_samples_per_second": 24.412, |
|
"eval_steps_per_second": 3.119, |
|
"step": 2500 |
|
} |
|
], |
|
"logging_steps": 50, |
|
"max_steps": 4000, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 160, |
|
"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": 2.701326498107371e+16, |
|
"train_batch_size": 16, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|