| { | |
| "best_metric": 0.8353737113402062, | |
| "best_model_checkpoint": "bert_trainer/run-6/checkpoint-2500", | |
| "epoch": 3.2216494845360826, | |
| "eval_steps": 250, | |
| "global_step": 2500, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5.985683583902487e-07, | |
| "loss": 1.498, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 1.4082, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 1.2917, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 1.2083, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 1.1513, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 1.1182, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 1.0869, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 1.0574, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.9843, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 1.0196, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "eval_accuracy": 0.6401417525773195, | |
| "eval_loss": 0.9714874029159546, | |
| "eval_runtime": 53.062, | |
| "eval_samples_per_second": 58.498, | |
| "eval_steps_per_second": 3.656, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.9886, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.9861, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.9746, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.9386, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.9046, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.9213, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.892, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.8545, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.87, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.8463, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "eval_accuracy": 0.6978092783505154, | |
| "eval_loss": 0.810293436050415, | |
| "eval_runtime": 53.0534, | |
| "eval_samples_per_second": 58.507, | |
| "eval_steps_per_second": 3.657, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.8516, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.8086, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.7894, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.7754, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.7558, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.7155, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.7151, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.758, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.6958, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.68, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "eval_accuracy": 0.7503221649484536, | |
| "eval_loss": 0.6983952522277832, | |
| "eval_runtime": 57.8438, | |
| "eval_samples_per_second": 53.662, | |
| "eval_steps_per_second": 3.354, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.6553, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.6829, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.7116, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.6283, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.6559, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.6118, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.6675, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.6426, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.6344, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.6754, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "eval_accuracy": 0.7970360824742269, | |
| "eval_loss": 0.6213943362236023, | |
| "eval_runtime": 57.838, | |
| "eval_samples_per_second": 53.667, | |
| "eval_steps_per_second": 3.354, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.6535, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.64, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.6442, | |
| "step": 1075 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5873, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.587, | |
| "step": 1125 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5789, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.605, | |
| "step": 1175 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.6078, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.6431, | |
| "step": 1225 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5824, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "eval_accuracy": 0.803479381443299, | |
| "eval_loss": 0.5681772232055664, | |
| "eval_runtime": 57.809, | |
| "eval_samples_per_second": 53.694, | |
| "eval_steps_per_second": 3.356, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.648, | |
| "step": 1275 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5709, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5504, | |
| "step": 1325 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5699, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5575, | |
| "step": 1375 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5644, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.587, | |
| "step": 1425 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5094, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5724, | |
| "step": 1475 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5652, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "eval_accuracy": 0.8128221649484536, | |
| "eval_loss": 0.5369997024536133, | |
| "eval_runtime": 57.8152, | |
| "eval_samples_per_second": 53.688, | |
| "eval_steps_per_second": 3.356, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5197, | |
| "step": 1525 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5499, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.58, | |
| "step": 1575 | |
| }, | |
| { | |
| "epoch": 2.06, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5348, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4741, | |
| "step": 1625 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5582, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5219, | |
| "step": 1675 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4802, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5147, | |
| "step": 1725 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4912, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "eval_accuracy": 0.8234536082474226, | |
| "eval_loss": 0.5059388279914856, | |
| "eval_runtime": 57.8239, | |
| "eval_samples_per_second": 53.68, | |
| "eval_steps_per_second": 3.355, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5133, | |
| "step": 1775 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4629, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4696, | |
| "step": 1825 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4623, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4649, | |
| "step": 1875 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5314, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4974, | |
| "step": 1925 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.464, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5332, | |
| "step": 1975 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4694, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "eval_accuracy": 0.8286082474226805, | |
| "eval_loss": 0.48415493965148926, | |
| "eval_runtime": 57.8531, | |
| "eval_samples_per_second": 53.653, | |
| "eval_steps_per_second": 3.353, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5191, | |
| "step": 2025 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4969, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4499, | |
| "step": 2075 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5144, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.482, | |
| "step": 2125 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4193, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4928, | |
| "step": 2175 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4515, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4775, | |
| "step": 2225 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5456, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "eval_accuracy": 0.833118556701031, | |
| "eval_loss": 0.4647773802280426, | |
| "eval_runtime": 57.8252, | |
| "eval_samples_per_second": 53.679, | |
| "eval_steps_per_second": 3.355, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.5226, | |
| "step": 2275 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4159, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4486, | |
| "step": 2325 | |
| }, | |
| { | |
| "epoch": 3.03, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.3866, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 3.06, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4634, | |
| "step": 2375 | |
| }, | |
| { | |
| "epoch": 3.09, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4179, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4446, | |
| "step": 2425 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4606, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 3.19, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.4261, | |
| "step": 2475 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "learning_rate": 1.1971367167804974e-06, | |
| "loss": 0.3907, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 3.22, | |
| "eval_accuracy": 0.8353737113402062, | |
| "eval_loss": 0.45589184761047363, | |
| "eval_runtime": 57.8334, | |
| "eval_samples_per_second": 53.671, | |
| "eval_steps_per_second": 3.354, | |
| "step": 2500 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 3000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 250, | |
| "total_flos": 1.051683215428608e+16, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 1.1971367167804974e-06, | |
| "per_device_eval_batch_size": 16, | |
| "per_device_train_batch_size": 16 | |
| } | |
| } | |