| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 501, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.059880239520958084, | |
| "grad_norm": 6.566332162525977, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 4.3864, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11976047904191617, | |
| "grad_norm": 2.0437379261117132, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 3.4794, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.17964071856287425, | |
| "grad_norm": 0.8566260346858398, | |
| "learning_rate": 1.9996500732179695e-05, | |
| "loss": 2.5375, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.23952095808383234, | |
| "grad_norm": 0.6117118904297301, | |
| "learning_rate": 1.995716208873644e-05, | |
| "loss": 1.8752, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2994011976047904, | |
| "grad_norm": 0.3923875723199884, | |
| "learning_rate": 1.9874283308955058e-05, | |
| "loss": 1.4434, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3592814371257485, | |
| "grad_norm": 0.2726465509781607, | |
| "learning_rate": 1.9748226800652062e-05, | |
| "loss": 1.2232, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.41916167664670656, | |
| "grad_norm": 0.22048277219015358, | |
| "learning_rate": 1.957954377686475e-05, | |
| "loss": 1.1031, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.47904191616766467, | |
| "grad_norm": 0.19131964365908752, | |
| "learning_rate": 1.9368971845536844e-05, | |
| "loss": 1.0318, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.5389221556886228, | |
| "grad_norm": 0.19325047947069812, | |
| "learning_rate": 1.911743178414665e-05, | |
| "loss": 0.9724, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5988023952095808, | |
| "grad_norm": 0.18220999359482376, | |
| "learning_rate": 1.8826023513381372e-05, | |
| "loss": 0.942, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.6586826347305389, | |
| "grad_norm": 0.17750272843534337, | |
| "learning_rate": 1.849602128746387e-05, | |
| "loss": 0.9106, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.718562874251497, | |
| "grad_norm": 0.16658108783370879, | |
| "learning_rate": 1.8128868122163125e-05, | |
| "loss": 0.8876, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.7784431137724551, | |
| "grad_norm": 0.1461861518555333, | |
| "learning_rate": 1.7726169484853438e-05, | |
| "loss": 0.872, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.8383233532934131, | |
| "grad_norm": 0.14457472093324725, | |
| "learning_rate": 1.7289686274214116e-05, | |
| "loss": 0.8576, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.8982035928143712, | |
| "grad_norm": 0.1544690749108873, | |
| "learning_rate": 1.6821327120267567e-05, | |
| "loss": 0.8467, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.9580838323353293, | |
| "grad_norm": 0.1429893393640351, | |
| "learning_rate": 1.6323140038425842e-05, | |
| "loss": 0.8252, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.0179640718562875, | |
| "grad_norm": 0.11444623635062835, | |
| "learning_rate": 1.5797303474040332e-05, | |
| "loss": 0.8209, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.0778443113772456, | |
| "grad_norm": 0.12105393067348605, | |
| "learning_rate": 1.524611677661454e-05, | |
| "loss": 0.8064, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.1377245508982037, | |
| "grad_norm": 0.11410840909819214, | |
| "learning_rate": 1.4671990145333697e-05, | |
| "loss": 0.7997, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.1976047904191618, | |
| "grad_norm": 0.1145244272492734, | |
| "learning_rate": 1.4077434089877038e-05, | |
| "loss": 0.7938, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.2574850299401197, | |
| "grad_norm": 0.12645762812759187, | |
| "learning_rate": 1.3465048452597682e-05, | |
| "loss": 0.7775, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.3173652694610778, | |
| "grad_norm": 0.11488140079430814, | |
| "learning_rate": 1.283751104007355e-05, | |
| "loss": 0.7767, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.377245508982036, | |
| "grad_norm": 0.13144847588925945, | |
| "learning_rate": 1.2197565913740531e-05, | |
| "loss": 0.7754, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.437125748502994, | |
| "grad_norm": 0.10603932803573335, | |
| "learning_rate": 1.1548011390810016e-05, | |
| "loss": 0.7644, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.4970059880239521, | |
| "grad_norm": 0.12250644964069578, | |
| "learning_rate": 1.0891687807939707e-05, | |
| "loss": 0.7735, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.55688622754491, | |
| "grad_norm": 0.1092716735846818, | |
| "learning_rate": 1.023146510116414e-05, | |
| "loss": 0.7646, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.6167664670658684, | |
| "grad_norm": 0.10437029986936892, | |
| "learning_rate": 9.570230256394595e-06, | |
| "loss": 0.7589, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.6766467065868262, | |
| "grad_norm": 0.10262505227638254, | |
| "learning_rate": 8.910874685364275e-06, | |
| "loss": 0.7579, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.7365269461077846, | |
| "grad_norm": 0.10634904902257739, | |
| "learning_rate": 8.256281582220486e-06, | |
| "loss": 0.7563, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.7964071856287425, | |
| "grad_norm": 0.10308159473170038, | |
| "learning_rate": 7.6093133160502e-06, | |
| "loss": 0.7541, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.8562874251497006, | |
| "grad_norm": 0.11967986382799979, | |
| "learning_rate": 6.972798914468369e-06, | |
| "loss": 0.7535, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.9161676646706587, | |
| "grad_norm": 0.09790252223897214, | |
| "learning_rate": 6.349521692999945e-06, | |
| "loss": 0.7521, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.9760479041916168, | |
| "grad_norm": 0.09917905729981547, | |
| "learning_rate": 5.742207084349274e-06, | |
| "loss": 0.748, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.035928143712575, | |
| "grad_norm": 0.09219293703087832, | |
| "learning_rate": 5.153510720776354e-06, | |
| "loss": 0.7458, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.095808383233533, | |
| "grad_norm": 0.09531637506680916, | |
| "learning_rate": 4.58600682169262e-06, | |
| "loss": 0.7417, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.155688622754491, | |
| "grad_norm": 0.099401468944997, | |
| "learning_rate": 4.042176937254474e-06, | |
| "loss": 0.7381, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.215568862275449, | |
| "grad_norm": 0.09333850738653798, | |
| "learning_rate": 3.5243990971758124e-06, | |
| "loss": 0.7368, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.2754491017964074, | |
| "grad_norm": 0.10118930739249864, | |
| "learning_rate": 3.034937412209178e-06, | |
| "loss": 0.7407, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.3353293413173652, | |
| "grad_norm": 0.09991557691778238, | |
| "learning_rate": 2.5759321737655017e-06, | |
| "loss": 0.7348, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.3952095808383236, | |
| "grad_norm": 0.08992918338455534, | |
| "learning_rate": 2.149390494964323e-06, | |
| "loss": 0.73, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.4550898203592815, | |
| "grad_norm": 0.0976318233532428, | |
| "learning_rate": 1.7571775340388275e-06, | |
| "loss": 0.7327, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.5149700598802394, | |
| "grad_norm": 0.0925071697104646, | |
| "learning_rate": 1.4010083384734308e-06, | |
| "loss": 0.7249, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.5748502994011977, | |
| "grad_norm": 0.09013756739384593, | |
| "learning_rate": 1.0824403455375287e-06, | |
| "loss": 0.7361, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.6347305389221556, | |
| "grad_norm": 0.10044956018733328, | |
| "learning_rate": 8.02866572008566e-07, | |
| "loss": 0.7403, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.694610778443114, | |
| "grad_norm": 0.09704814807171967, | |
| "learning_rate": 5.63509522864123e-07, | |
| "loss": 0.7332, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.754491017964072, | |
| "grad_norm": 0.11375533603468758, | |
| "learning_rate": 3.6541584557868604e-07, | |
| "loss": 0.7329, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.81437125748503, | |
| "grad_norm": 0.10095526024482018, | |
| "learning_rate": 2.0945175340055356e-07, | |
| "loss": 0.7348, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.874251497005988, | |
| "grad_norm": 0.0869826588136221, | |
| "learning_rate": 9.629923762170091e-08, | |
| "loss": 0.7303, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.934131736526946, | |
| "grad_norm": 0.08752359244376111, | |
| "learning_rate": 2.645308540337843e-08, | |
| "loss": 0.7358, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.9940119760479043, | |
| "grad_norm": 0.09120038510133631, | |
| "learning_rate": 2.1871619775404308e-10, | |
| "loss": 0.7293, | |
| "step": 500 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 501, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 56, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 913430022193152.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |