{ | |
"best_metric": null, | |
"best_model_checkpoint": null, | |
"epoch": 1.0, | |
"global_step": 14299, | |
"is_hyper_param_search": false, | |
"is_local_process_zero": true, | |
"is_world_process_zero": true, | |
"log_history": [ | |
{ | |
"epoch": 0.01, | |
"learning_rate": 9.488135456452206e-05, | |
"loss": 3.5169, | |
"step": 100 | |
}, | |
{ | |
"epoch": 0.01, | |
"learning_rate": 0.0001, | |
"loss": 1.7174, | |
"step": 200 | |
}, | |
{ | |
"epoch": 0.02, | |
"learning_rate": 0.0001, | |
"loss": 1.7028, | |
"step": 300 | |
}, | |
{ | |
"epoch": 0.03, | |
"learning_rate": 0.0001, | |
"loss": 1.7626, | |
"step": 400 | |
}, | |
{ | |
"epoch": 0.03, | |
"learning_rate": 0.0001, | |
"loss": 1.7534, | |
"step": 500 | |
}, | |
{ | |
"epoch": 0.04, | |
"learning_rate": 0.0001, | |
"loss": 1.7371, | |
"step": 600 | |
}, | |
{ | |
"epoch": 0.05, | |
"learning_rate": 0.0001, | |
"loss": 1.7427, | |
"step": 700 | |
}, | |
{ | |
"epoch": 0.06, | |
"learning_rate": 0.0001, | |
"loss": 1.7329, | |
"step": 800 | |
}, | |
{ | |
"epoch": 0.06, | |
"learning_rate": 0.0001, | |
"loss": 1.7513, | |
"step": 900 | |
}, | |
{ | |
"epoch": 0.07, | |
"learning_rate": 0.0001, | |
"loss": 1.7183, | |
"step": 1000 | |
}, | |
{ | |
"epoch": 0.08, | |
"learning_rate": 0.0001, | |
"loss": 1.6911, | |
"step": 1100 | |
}, | |
{ | |
"epoch": 0.08, | |
"learning_rate": 0.0001, | |
"loss": 1.7379, | |
"step": 1200 | |
}, | |
{ | |
"epoch": 0.09, | |
"learning_rate": 0.0001, | |
"loss": 1.7167, | |
"step": 1300 | |
}, | |
{ | |
"epoch": 0.1, | |
"learning_rate": 0.0001, | |
"loss": 1.739, | |
"step": 1400 | |
}, | |
{ | |
"epoch": 0.1, | |
"learning_rate": 0.0001, | |
"loss": 1.7262, | |
"step": 1500 | |
}, | |
{ | |
"epoch": 0.11, | |
"learning_rate": 0.0001, | |
"loss": 1.7185, | |
"step": 1600 | |
}, | |
{ | |
"epoch": 0.12, | |
"learning_rate": 0.0001, | |
"loss": 1.7117, | |
"step": 1700 | |
}, | |
{ | |
"epoch": 0.13, | |
"learning_rate": 0.0001, | |
"loss": 1.7502, | |
"step": 1800 | |
}, | |
{ | |
"epoch": 0.13, | |
"learning_rate": 0.0001, | |
"loss": 1.7643, | |
"step": 1900 | |
}, | |
{ | |
"epoch": 0.14, | |
"learning_rate": 0.0001, | |
"loss": 1.7477, | |
"step": 2000 | |
}, | |
{ | |
"epoch": 0.15, | |
"learning_rate": 0.0001, | |
"loss": 1.7249, | |
"step": 2100 | |
}, | |
{ | |
"epoch": 0.15, | |
"learning_rate": 0.0001, | |
"loss": 1.7645, | |
"step": 2200 | |
}, | |
{ | |
"epoch": 0.16, | |
"learning_rate": 0.0001, | |
"loss": 1.7791, | |
"step": 2300 | |
}, | |
{ | |
"epoch": 0.17, | |
"learning_rate": 0.0001, | |
"loss": 1.7606, | |
"step": 2400 | |
}, | |
{ | |
"epoch": 0.17, | |
"learning_rate": 0.0001, | |
"loss": 1.7265, | |
"step": 2500 | |
}, | |
{ | |
"epoch": 0.18, | |
"learning_rate": 0.0001, | |
"loss": 1.7175, | |
"step": 2600 | |
}, | |
{ | |
"epoch": 0.19, | |
"learning_rate": 0.0001, | |
"loss": 1.7584, | |
"step": 2700 | |
}, | |
{ | |
"epoch": 0.2, | |
"learning_rate": 0.0001, | |
"loss": 1.7327, | |
"step": 2800 | |
}, | |
{ | |
"epoch": 0.2, | |
"learning_rate": 0.0001, | |
"loss": 1.758, | |
"step": 2900 | |
}, | |
{ | |
"epoch": 0.21, | |
"learning_rate": 0.0001, | |
"loss": 1.7684, | |
"step": 3000 | |
}, | |
{ | |
"epoch": 0.22, | |
"learning_rate": 0.0001, | |
"loss": 1.7797, | |
"step": 3100 | |
}, | |
{ | |
"epoch": 0.22, | |
"learning_rate": 0.0001, | |
"loss": 1.7749, | |
"step": 3200 | |
}, | |
{ | |
"epoch": 0.23, | |
"learning_rate": 0.0001, | |
"loss": 1.7465, | |
"step": 3300 | |
}, | |
{ | |
"epoch": 0.24, | |
"learning_rate": 0.0001, | |
"loss": 1.7424, | |
"step": 3400 | |
}, | |
{ | |
"epoch": 0.24, | |
"learning_rate": 0.0001, | |
"loss": 1.7694, | |
"step": 3500 | |
}, | |
{ | |
"epoch": 0.25, | |
"learning_rate": 0.0001, | |
"loss": 1.7623, | |
"step": 3600 | |
}, | |
{ | |
"epoch": 0.26, | |
"learning_rate": 0.0001, | |
"loss": 1.7446, | |
"step": 3700 | |
}, | |
{ | |
"epoch": 0.27, | |
"learning_rate": 0.0001, | |
"loss": 1.777, | |
"step": 3800 | |
}, | |
{ | |
"epoch": 0.27, | |
"learning_rate": 0.0001, | |
"loss": 1.7267, | |
"step": 3900 | |
}, | |
{ | |
"epoch": 0.28, | |
"learning_rate": 0.0001, | |
"loss": 1.7354, | |
"step": 4000 | |
}, | |
{ | |
"epoch": 0.29, | |
"learning_rate": 0.0001, | |
"loss": 1.7529, | |
"step": 4100 | |
}, | |
{ | |
"epoch": 0.29, | |
"learning_rate": 0.0001, | |
"loss": 1.7626, | |
"step": 4200 | |
}, | |
{ | |
"epoch": 0.3, | |
"learning_rate": 0.0001, | |
"loss": 1.7977, | |
"step": 4300 | |
}, | |
{ | |
"epoch": 0.31, | |
"learning_rate": 0.0001, | |
"loss": 1.7783, | |
"step": 4400 | |
}, | |
{ | |
"epoch": 0.31, | |
"learning_rate": 0.0001, | |
"loss": 1.7496, | |
"step": 4500 | |
}, | |
{ | |
"epoch": 0.32, | |
"learning_rate": 0.0001, | |
"loss": 1.7424, | |
"step": 4600 | |
}, | |
{ | |
"epoch": 0.33, | |
"learning_rate": 0.0001, | |
"loss": 1.772, | |
"step": 4700 | |
}, | |
{ | |
"epoch": 0.34, | |
"learning_rate": 0.0001, | |
"loss": 1.7112, | |
"step": 4800 | |
}, | |
{ | |
"epoch": 0.34, | |
"learning_rate": 0.0001, | |
"loss": 1.7185, | |
"step": 4900 | |
}, | |
{ | |
"epoch": 0.35, | |
"learning_rate": 0.0001, | |
"loss": 1.7654, | |
"step": 5000 | |
}, | |
{ | |
"epoch": 0.36, | |
"learning_rate": 0.0001, | |
"loss": 1.7578, | |
"step": 5100 | |
}, | |
{ | |
"epoch": 0.36, | |
"learning_rate": 0.0001, | |
"loss": 1.7418, | |
"step": 5200 | |
}, | |
{ | |
"epoch": 0.37, | |
"learning_rate": 0.0001, | |
"loss": 1.7847, | |
"step": 5300 | |
}, | |
{ | |
"epoch": 0.38, | |
"learning_rate": 0.0001, | |
"loss": 1.7315, | |
"step": 5400 | |
}, | |
{ | |
"epoch": 0.38, | |
"learning_rate": 0.0001, | |
"loss": 1.8105, | |
"step": 5500 | |
}, | |
{ | |
"epoch": 0.39, | |
"learning_rate": 0.0001, | |
"loss": 1.7626, | |
"step": 5600 | |
}, | |
{ | |
"epoch": 0.4, | |
"learning_rate": 0.0001, | |
"loss": 1.7525, | |
"step": 5700 | |
}, | |
{ | |
"epoch": 0.41, | |
"learning_rate": 0.0001, | |
"loss": 1.7412, | |
"step": 5800 | |
}, | |
{ | |
"epoch": 0.41, | |
"learning_rate": 0.0001, | |
"loss": 1.7909, | |
"step": 5900 | |
}, | |
{ | |
"epoch": 0.42, | |
"learning_rate": 0.0001, | |
"loss": 1.7921, | |
"step": 6000 | |
}, | |
{ | |
"epoch": 0.43, | |
"learning_rate": 0.0001, | |
"loss": 1.7768, | |
"step": 6100 | |
}, | |
{ | |
"epoch": 0.43, | |
"learning_rate": 0.0001, | |
"loss": 1.7578, | |
"step": 6200 | |
}, | |
{ | |
"epoch": 0.44, | |
"learning_rate": 0.0001, | |
"loss": 1.774, | |
"step": 6300 | |
}, | |
{ | |
"epoch": 0.45, | |
"learning_rate": 0.0001, | |
"loss": 1.7622, | |
"step": 6400 | |
}, | |
{ | |
"epoch": 0.45, | |
"learning_rate": 0.0001, | |
"loss": 1.7623, | |
"step": 6500 | |
}, | |
{ | |
"epoch": 0.46, | |
"learning_rate": 0.0001, | |
"loss": 1.7699, | |
"step": 6600 | |
}, | |
{ | |
"epoch": 0.47, | |
"learning_rate": 0.0001, | |
"loss": 1.7684, | |
"step": 6700 | |
}, | |
{ | |
"epoch": 0.48, | |
"learning_rate": 0.0001, | |
"loss": 1.7562, | |
"step": 6800 | |
}, | |
{ | |
"epoch": 0.48, | |
"learning_rate": 0.0001, | |
"loss": 1.7519, | |
"step": 6900 | |
}, | |
{ | |
"epoch": 0.49, | |
"learning_rate": 0.0001, | |
"loss": 1.7722, | |
"step": 7000 | |
}, | |
{ | |
"epoch": 0.5, | |
"learning_rate": 0.0001, | |
"loss": 1.7586, | |
"step": 7100 | |
}, | |
{ | |
"epoch": 0.5, | |
"learning_rate": 0.0001, | |
"loss": 1.7702, | |
"step": 7200 | |
}, | |
{ | |
"epoch": 0.51, | |
"learning_rate": 0.0001, | |
"loss": 1.7498, | |
"step": 7300 | |
}, | |
{ | |
"epoch": 0.52, | |
"learning_rate": 0.0001, | |
"loss": 1.741, | |
"step": 7400 | |
}, | |
{ | |
"epoch": 0.52, | |
"learning_rate": 0.0001, | |
"loss": 1.731, | |
"step": 7500 | |
}, | |
{ | |
"epoch": 0.53, | |
"learning_rate": 0.0001, | |
"loss": 1.7366, | |
"step": 7600 | |
}, | |
{ | |
"epoch": 0.54, | |
"learning_rate": 0.0001, | |
"loss": 1.7384, | |
"step": 7700 | |
}, | |
{ | |
"epoch": 0.55, | |
"learning_rate": 0.0001, | |
"loss": 1.8058, | |
"step": 7800 | |
}, | |
{ | |
"epoch": 0.55, | |
"learning_rate": 0.0001, | |
"loss": 1.7583, | |
"step": 7900 | |
}, | |
{ | |
"epoch": 0.56, | |
"learning_rate": 0.0001, | |
"loss": 1.7529, | |
"step": 8000 | |
}, | |
{ | |
"epoch": 0.57, | |
"learning_rate": 0.0001, | |
"loss": 1.7159, | |
"step": 8100 | |
}, | |
{ | |
"epoch": 0.57, | |
"learning_rate": 0.0001, | |
"loss": 1.7357, | |
"step": 8200 | |
}, | |
{ | |
"epoch": 0.58, | |
"learning_rate": 0.0001, | |
"loss": 1.7377, | |
"step": 8300 | |
}, | |
{ | |
"epoch": 0.59, | |
"learning_rate": 0.0001, | |
"loss": 1.7135, | |
"step": 8400 | |
}, | |
{ | |
"epoch": 0.59, | |
"learning_rate": 0.0001, | |
"loss": 1.768, | |
"step": 8500 | |
}, | |
{ | |
"epoch": 0.6, | |
"learning_rate": 0.0001, | |
"loss": 1.7218, | |
"step": 8600 | |
}, | |
{ | |
"epoch": 0.61, | |
"learning_rate": 0.0001, | |
"loss": 1.7191, | |
"step": 8700 | |
}, | |
{ | |
"epoch": 0.62, | |
"learning_rate": 0.0001, | |
"loss": 1.7583, | |
"step": 8800 | |
}, | |
{ | |
"epoch": 0.62, | |
"learning_rate": 0.0001, | |
"loss": 1.7273, | |
"step": 8900 | |
}, | |
{ | |
"epoch": 0.63, | |
"learning_rate": 0.0001, | |
"loss": 1.7674, | |
"step": 9000 | |
}, | |
{ | |
"epoch": 0.64, | |
"learning_rate": 0.0001, | |
"loss": 1.7564, | |
"step": 9100 | |
}, | |
{ | |
"epoch": 0.64, | |
"learning_rate": 0.0001, | |
"loss": 1.7216, | |
"step": 9200 | |
}, | |
{ | |
"epoch": 0.65, | |
"learning_rate": 0.0001, | |
"loss": 1.754, | |
"step": 9300 | |
}, | |
{ | |
"epoch": 0.66, | |
"learning_rate": 0.0001, | |
"loss": 1.6841, | |
"step": 9400 | |
}, | |
{ | |
"epoch": 0.66, | |
"learning_rate": 0.0001, | |
"loss": 1.7266, | |
"step": 9500 | |
}, | |
{ | |
"epoch": 0.67, | |
"learning_rate": 0.0001, | |
"loss": 1.7307, | |
"step": 9600 | |
}, | |
{ | |
"epoch": 0.68, | |
"learning_rate": 0.0001, | |
"loss": 1.7247, | |
"step": 9700 | |
}, | |
{ | |
"epoch": 0.69, | |
"learning_rate": 0.0001, | |
"loss": 1.718, | |
"step": 9800 | |
}, | |
{ | |
"epoch": 0.69, | |
"learning_rate": 0.0001, | |
"loss": 1.6868, | |
"step": 9900 | |
}, | |
{ | |
"epoch": 0.7, | |
"learning_rate": 0.0001, | |
"loss": 1.6977, | |
"step": 10000 | |
}, | |
{ | |
"epoch": 0.71, | |
"learning_rate": 0.0001, | |
"loss": 1.7112, | |
"step": 10100 | |
}, | |
{ | |
"epoch": 0.71, | |
"learning_rate": 0.0001, | |
"loss": 1.7188, | |
"step": 10200 | |
}, | |
{ | |
"epoch": 0.72, | |
"learning_rate": 0.0001, | |
"loss": 1.7266, | |
"step": 10300 | |
}, | |
{ | |
"epoch": 0.73, | |
"learning_rate": 0.0001, | |
"loss": 1.7335, | |
"step": 10400 | |
}, | |
{ | |
"epoch": 0.73, | |
"learning_rate": 0.0001, | |
"loss": 1.7232, | |
"step": 10500 | |
}, | |
{ | |
"epoch": 0.74, | |
"learning_rate": 0.0001, | |
"loss": 1.7528, | |
"step": 10600 | |
}, | |
{ | |
"epoch": 0.75, | |
"learning_rate": 0.0001, | |
"loss": 1.736, | |
"step": 10700 | |
}, | |
{ | |
"epoch": 0.76, | |
"learning_rate": 0.0001, | |
"loss": 1.7049, | |
"step": 10800 | |
}, | |
{ | |
"epoch": 0.76, | |
"learning_rate": 0.0001, | |
"loss": 1.7306, | |
"step": 10900 | |
}, | |
{ | |
"epoch": 0.77, | |
"learning_rate": 0.0001, | |
"loss": 1.7341, | |
"step": 11000 | |
}, | |
{ | |
"epoch": 0.78, | |
"learning_rate": 0.0001, | |
"loss": 1.7264, | |
"step": 11100 | |
}, | |
{ | |
"epoch": 0.78, | |
"learning_rate": 0.0001, | |
"loss": 1.6726, | |
"step": 11200 | |
}, | |
{ | |
"epoch": 0.79, | |
"learning_rate": 0.0001, | |
"loss": 1.7051, | |
"step": 11300 | |
}, | |
{ | |
"epoch": 0.8, | |
"learning_rate": 0.0001, | |
"loss": 1.6784, | |
"step": 11400 | |
}, | |
{ | |
"epoch": 0.8, | |
"learning_rate": 0.0001, | |
"loss": 1.6752, | |
"step": 11500 | |
}, | |
{ | |
"epoch": 0.81, | |
"learning_rate": 0.0001, | |
"loss": 1.6932, | |
"step": 11600 | |
}, | |
{ | |
"epoch": 0.82, | |
"learning_rate": 0.0001, | |
"loss": 1.6927, | |
"step": 11700 | |
}, | |
{ | |
"epoch": 0.83, | |
"learning_rate": 0.0001, | |
"loss": 1.6961, | |
"step": 11800 | |
}, | |
{ | |
"epoch": 0.83, | |
"learning_rate": 0.0001, | |
"loss": 1.7069, | |
"step": 11900 | |
}, | |
{ | |
"epoch": 0.84, | |
"learning_rate": 0.0001, | |
"loss": 1.7093, | |
"step": 12000 | |
}, | |
{ | |
"epoch": 0.85, | |
"learning_rate": 0.0001, | |
"loss": 1.704, | |
"step": 12100 | |
}, | |
{ | |
"epoch": 0.85, | |
"learning_rate": 0.0001, | |
"loss": 1.7012, | |
"step": 12200 | |
}, | |
{ | |
"epoch": 0.86, | |
"learning_rate": 0.0001, | |
"loss": 1.7215, | |
"step": 12300 | |
}, | |
{ | |
"epoch": 0.87, | |
"learning_rate": 0.0001, | |
"loss": 1.7031, | |
"step": 12400 | |
}, | |
{ | |
"epoch": 0.87, | |
"learning_rate": 0.0001, | |
"loss": 1.7203, | |
"step": 12500 | |
}, | |
{ | |
"epoch": 0.88, | |
"learning_rate": 0.0001, | |
"loss": 1.6664, | |
"step": 12600 | |
}, | |
{ | |
"epoch": 0.89, | |
"learning_rate": 0.0001, | |
"loss": 1.6581, | |
"step": 12700 | |
}, | |
{ | |
"epoch": 0.9, | |
"learning_rate": 0.0001, | |
"loss": 1.6727, | |
"step": 12800 | |
}, | |
{ | |
"epoch": 0.9, | |
"learning_rate": 0.0001, | |
"loss": 1.7033, | |
"step": 12900 | |
}, | |
{ | |
"epoch": 0.91, | |
"learning_rate": 0.0001, | |
"loss": 1.6928, | |
"step": 13000 | |
}, | |
{ | |
"epoch": 0.92, | |
"learning_rate": 0.0001, | |
"loss": 1.6719, | |
"step": 13100 | |
}, | |
{ | |
"epoch": 0.92, | |
"learning_rate": 0.0001, | |
"loss": 1.7131, | |
"step": 13200 | |
}, | |
{ | |
"epoch": 0.93, | |
"learning_rate": 0.0001, | |
"loss": 1.6699, | |
"step": 13300 | |
}, | |
{ | |
"epoch": 0.94, | |
"learning_rate": 0.0001, | |
"loss": 1.666, | |
"step": 13400 | |
}, | |
{ | |
"epoch": 0.94, | |
"learning_rate": 0.0001, | |
"loss": 1.6937, | |
"step": 13500 | |
}, | |
{ | |
"epoch": 0.95, | |
"learning_rate": 0.0001, | |
"loss": 1.7058, | |
"step": 13600 | |
}, | |
{ | |
"epoch": 0.96, | |
"learning_rate": 0.0001, | |
"loss": 1.7061, | |
"step": 13700 | |
}, | |
{ | |
"epoch": 0.97, | |
"learning_rate": 0.0001, | |
"loss": 1.651, | |
"step": 13800 | |
}, | |
{ | |
"epoch": 0.97, | |
"learning_rate": 0.0001, | |
"loss": 1.653, | |
"step": 13900 | |
}, | |
{ | |
"epoch": 0.98, | |
"learning_rate": 0.0001, | |
"loss": 1.6569, | |
"step": 14000 | |
}, | |
{ | |
"epoch": 0.99, | |
"learning_rate": 0.0001, | |
"loss": 1.6767, | |
"step": 14100 | |
}, | |
{ | |
"epoch": 0.99, | |
"learning_rate": 0.0001, | |
"loss": 1.6741, | |
"step": 14200 | |
}, | |
{ | |
"epoch": 1.0, | |
"eval_loss": 1.6552734375, | |
"eval_runtime": 7.7504, | |
"eval_samples_per_second": 301.275, | |
"eval_steps_per_second": 37.676, | |
"step": 14299 | |
} | |
], | |
"max_steps": 14299, | |
"num_train_epochs": 1, | |
"total_flos": 5.1403509272569446e+17, | |
"trial_name": null, | |
"trial_params": null | |
} | |