| { | |
| "best_metric": 0.7763074636459351, | |
| "best_model_checkpoint": "./t5_finetuned/checkpoint-24000", | |
| "epoch": 1.0, | |
| "eval_steps": 2000, | |
| "global_step": 25622, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.019514479743970026, | |
| "grad_norm": 50375.1171875, | |
| "learning_rate": 3.9219420810241204e-05, | |
| "loss": 1.6162, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.03902895948794005, | |
| "grad_norm": 40185.09765625, | |
| "learning_rate": 3.84388416204824e-05, | |
| "loss": 1.0952, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.05854343923191008, | |
| "grad_norm": 26678.357421875, | |
| "learning_rate": 3.76582624307236e-05, | |
| "loss": 1.0438, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.0780579189758801, | |
| "grad_norm": 75600.265625, | |
| "learning_rate": 3.68776832409648e-05, | |
| "loss": 1.0177, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.0780579189758801, | |
| "eval_loss": 0.9050301909446716, | |
| "eval_runtime": 247.5245, | |
| "eval_samples_per_second": 87.171, | |
| "eval_steps_per_second": 5.45, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.09757239871985013, | |
| "grad_norm": 48858.64453125, | |
| "learning_rate": 3.6097104051205996e-05, | |
| "loss": 0.994, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.11708687846382015, | |
| "grad_norm": 38273.890625, | |
| "learning_rate": 3.53165248614472e-05, | |
| "loss": 0.9774, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.13660135820779018, | |
| "grad_norm": 33092.80078125, | |
| "learning_rate": 3.45359456716884e-05, | |
| "loss": 0.9659, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.1561158379517602, | |
| "grad_norm": 42197.96484375, | |
| "learning_rate": 3.375536648192959e-05, | |
| "loss": 0.9396, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.1561158379517602, | |
| "eval_loss": 0.8554931879043579, | |
| "eval_runtime": 246.9587, | |
| "eval_samples_per_second": 87.371, | |
| "eval_steps_per_second": 5.462, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.17563031769573023, | |
| "grad_norm": 31069.32421875, | |
| "learning_rate": 3.2974787292170794e-05, | |
| "loss": 0.9339, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.19514479743970026, | |
| "grad_norm": 40504.2109375, | |
| "learning_rate": 3.2194208102411995e-05, | |
| "loss": 0.9271, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.21465927718367028, | |
| "grad_norm": 53872.625, | |
| "learning_rate": 3.141362891265319e-05, | |
| "loss": 0.9036, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.2341737569276403, | |
| "grad_norm": 40223.39453125, | |
| "learning_rate": 3.063304972289439e-05, | |
| "loss": 0.8948, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.2341737569276403, | |
| "eval_loss": 0.8311675190925598, | |
| "eval_runtime": 247.1162, | |
| "eval_samples_per_second": 87.315, | |
| "eval_steps_per_second": 5.459, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.25368823667161033, | |
| "grad_norm": 37229.52734375, | |
| "learning_rate": 2.985247053313559e-05, | |
| "loss": 0.9039, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.27320271641558036, | |
| "grad_norm": 42615.79296875, | |
| "learning_rate": 2.9071891343376786e-05, | |
| "loss": 0.9099, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.2927171961595504, | |
| "grad_norm": 45535.15234375, | |
| "learning_rate": 2.8291312153617988e-05, | |
| "loss": 0.9024, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.3122316759035204, | |
| "grad_norm": 44434.55078125, | |
| "learning_rate": 2.7510732963859185e-05, | |
| "loss": 0.9221, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.3122316759035204, | |
| "eval_loss": 0.8146435618400574, | |
| "eval_runtime": 247.4985, | |
| "eval_samples_per_second": 87.18, | |
| "eval_steps_per_second": 5.451, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.33174615564749044, | |
| "grad_norm": 106098.890625, | |
| "learning_rate": 2.6730153774100383e-05, | |
| "loss": 0.8931, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.35126063539146046, | |
| "grad_norm": 29800.005859375, | |
| "learning_rate": 2.5949574584341584e-05, | |
| "loss": 0.8991, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.3707751151354305, | |
| "grad_norm": 41034.53125, | |
| "learning_rate": 2.5168995394582782e-05, | |
| "loss": 0.8879, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.3902895948794005, | |
| "grad_norm": 36309.5625, | |
| "learning_rate": 2.438841620482398e-05, | |
| "loss": 0.873, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.3902895948794005, | |
| "eval_loss": 0.8043464422225952, | |
| "eval_runtime": 247.5255, | |
| "eval_samples_per_second": 87.171, | |
| "eval_steps_per_second": 5.45, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.40980407462337054, | |
| "grad_norm": 28724.736328125, | |
| "learning_rate": 2.360783701506518e-05, | |
| "loss": 0.8964, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.42931855436734057, | |
| "grad_norm": 32160.0625, | |
| "learning_rate": 2.282725782530638e-05, | |
| "loss": 0.8837, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.4488330341113106, | |
| "grad_norm": 33788.734375, | |
| "learning_rate": 2.2046678635547577e-05, | |
| "loss": 0.8498, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.4683475138552806, | |
| "grad_norm": 57106.34375, | |
| "learning_rate": 2.1266099445788778e-05, | |
| "loss": 0.8723, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.4683475138552806, | |
| "eval_loss": 0.7961086630821228, | |
| "eval_runtime": 246.9977, | |
| "eval_samples_per_second": 87.357, | |
| "eval_steps_per_second": 5.462, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.48786199359925064, | |
| "grad_norm": 42744.36328125, | |
| "learning_rate": 2.0485520256029976e-05, | |
| "loss": 0.8664, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.5073764733432207, | |
| "grad_norm": 59340.5859375, | |
| "learning_rate": 1.9704941066271174e-05, | |
| "loss": 0.8719, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.5268909530871907, | |
| "grad_norm": 76933.4453125, | |
| "learning_rate": 1.8924361876512375e-05, | |
| "loss": 0.8896, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.5464054328311607, | |
| "grad_norm": 38514.16015625, | |
| "learning_rate": 1.8143782686753573e-05, | |
| "loss": 0.848, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.5464054328311607, | |
| "eval_loss": 0.7900422811508179, | |
| "eval_runtime": 247.446, | |
| "eval_samples_per_second": 87.199, | |
| "eval_steps_per_second": 5.452, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.5659199125751307, | |
| "grad_norm": 37615.3515625, | |
| "learning_rate": 1.736320349699477e-05, | |
| "loss": 0.8632, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.5854343923191008, | |
| "grad_norm": 81223.0625, | |
| "learning_rate": 1.6582624307235972e-05, | |
| "loss": 0.8622, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.6049488720630708, | |
| "grad_norm": 54736.93359375, | |
| "learning_rate": 1.580204511747717e-05, | |
| "loss": 0.8655, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.6244633518070408, | |
| "grad_norm": 45588.96484375, | |
| "learning_rate": 1.5021465927718368e-05, | |
| "loss": 0.877, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.6244633518070408, | |
| "eval_loss": 0.7854538559913635, | |
| "eval_runtime": 247.3837, | |
| "eval_samples_per_second": 87.221, | |
| "eval_steps_per_second": 5.453, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.6439778315510108, | |
| "grad_norm": 43015.6015625, | |
| "learning_rate": 1.4240886737959567e-05, | |
| "loss": 0.8687, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.6634923112949809, | |
| "grad_norm": 35532.26171875, | |
| "learning_rate": 1.3460307548200767e-05, | |
| "loss": 0.8749, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.6830067910389509, | |
| "grad_norm": 32677.763671875, | |
| "learning_rate": 1.2679728358441965e-05, | |
| "loss": 0.8727, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.7025212707829209, | |
| "grad_norm": 42200.640625, | |
| "learning_rate": 1.1899149168683164e-05, | |
| "loss": 0.862, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.7025212707829209, | |
| "eval_loss": 0.7817208170890808, | |
| "eval_runtime": 247.398, | |
| "eval_samples_per_second": 87.216, | |
| "eval_steps_per_second": 5.453, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.722035750526891, | |
| "grad_norm": 30234.58203125, | |
| "learning_rate": 1.1118569978924364e-05, | |
| "loss": 0.8432, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.741550230270861, | |
| "grad_norm": 43296.27734375, | |
| "learning_rate": 1.0337990789165561e-05, | |
| "loss": 0.8711, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.761064710014831, | |
| "grad_norm": 20832.39453125, | |
| "learning_rate": 9.557411599406761e-06, | |
| "loss": 0.8716, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.780579189758801, | |
| "grad_norm": 33938.4140625, | |
| "learning_rate": 8.77683240964796e-06, | |
| "loss": 0.8584, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.780579189758801, | |
| "eval_loss": 0.7791029214859009, | |
| "eval_runtime": 247.0845, | |
| "eval_samples_per_second": 87.326, | |
| "eval_steps_per_second": 5.46, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.800093669502771, | |
| "grad_norm": 56116.62109375, | |
| "learning_rate": 7.996253219889158e-06, | |
| "loss": 0.8776, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.8196081492467411, | |
| "grad_norm": 33855.76953125, | |
| "learning_rate": 7.215674030130358e-06, | |
| "loss": 0.8571, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.8391226289907111, | |
| "grad_norm": 83128.3359375, | |
| "learning_rate": 6.4350948403715564e-06, | |
| "loss": 0.8606, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.8586371087346811, | |
| "grad_norm": 32873.78515625, | |
| "learning_rate": 5.654515650612755e-06, | |
| "loss": 0.8406, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.8586371087346811, | |
| "eval_loss": 0.777363121509552, | |
| "eval_runtime": 247.4058, | |
| "eval_samples_per_second": 87.213, | |
| "eval_steps_per_second": 5.453, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.8781515884786512, | |
| "grad_norm": 31377.396484375, | |
| "learning_rate": 4.873936460853954e-06, | |
| "loss": 0.8368, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.8976660682226212, | |
| "grad_norm": 22825.560546875, | |
| "learning_rate": 4.093357271095153e-06, | |
| "loss": 0.8419, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.9171805479665912, | |
| "grad_norm": 39448.79296875, | |
| "learning_rate": 3.312778081336352e-06, | |
| "loss": 0.8731, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.9366950277105612, | |
| "grad_norm": 67876.9296875, | |
| "learning_rate": 2.5321988915775507e-06, | |
| "loss": 0.8458, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.9366950277105612, | |
| "eval_loss": 0.7763074636459351, | |
| "eval_runtime": 247.0633, | |
| "eval_samples_per_second": 87.334, | |
| "eval_steps_per_second": 5.46, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.9562095074545313, | |
| "grad_norm": 32097.791015625, | |
| "learning_rate": 1.7516197018187498e-06, | |
| "loss": 0.8461, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 0.9757239871985013, | |
| "grad_norm": 34695.05859375, | |
| "learning_rate": 9.710405120599484e-07, | |
| "loss": 0.8384, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.9952384669424713, | |
| "grad_norm": 41189.7265625, | |
| "learning_rate": 1.9046132230114745e-07, | |
| "loss": 0.8632, | |
| "step": 25500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 25622, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 2000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.548269480443904e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |