| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9995189995189995, | |
| "eval_steps": 500, | |
| "global_step": 1039, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.9230769230769234e-06, | |
| "loss": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.615384615384616e-06, | |
| "loss": 0.0, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.923076923076923e-05, | |
| "loss": 0.0, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.8846153846153845e-05, | |
| "loss": 0.0, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.846153846153846e-05, | |
| "loss": 0.0, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.8076923076923084e-05, | |
| "loss": 0.0, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.769230769230769e-05, | |
| "loss": 0.0, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.730769230769232e-05, | |
| "loss": 0.0, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.692307692307693e-05, | |
| "loss": 0.0, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.653846153846155e-05, | |
| "loss": 0.0, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.615384615384617e-05, | |
| "loss": 0.0, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00010576923076923077, | |
| "loss": 0.0, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00011538461538461538, | |
| "loss": 0.0, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.000125, | |
| "loss": 0.0, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00013461538461538464, | |
| "loss": 0.0, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00014423076923076924, | |
| "loss": 0.0, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00015384615384615385, | |
| "loss": 0.0, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00016346153846153846, | |
| "loss": 0.0, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001730769230769231, | |
| "loss": 0.0, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001826923076923077, | |
| "loss": 0.0, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019230769230769233, | |
| "loss": 0.0, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019999943552317104, | |
| "loss": 0.0, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019997967950328128, | |
| "loss": 0.0, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001999317060143023, | |
| "loss": 0.0, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001998555285958899, | |
| "loss": 0.0, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019975116874775242, | |
| "loss": 0.0, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019961865592358288, | |
| "loss": 0.0, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001994580275227462, | |
| "loss": 0.0, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019926932887972393, | |
| "loss": 0.0, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001990526132513194, | |
| "loss": 0.0, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019880794180162693, | |
| "loss": 0.0, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019853538358476932, | |
| "loss": 0.0, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019823501552540865, | |
| "loss": 0.0, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019790692239703557, | |
| "loss": 0.0, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019755119679804367, | |
| "loss": 0.0, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019716793912559507, | |
| "loss": 0.0, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019675725754728527, | |
| "loss": 0.0, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019631926797061456, | |
| "loss": 0.0, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019585409401027556, | |
| "loss": 0.0, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019536186695326486, | |
| "loss": 0.0, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019484272572182986, | |
| "loss": 0.0, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019429681683426022, | |
| "loss": 0.0, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019372429436353606, | |
| "loss": 0.0, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001931253198938432, | |
| "loss": 0.0, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019250006247496928, | |
| "loss": 0.0, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019184869857459232, | |
| "loss": 0.0, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019117141202847586, | |
| "loss": 0.0, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00019046839398858474, | |
| "loss": 0.0, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00018973984286913584, | |
| "loss": 0.0, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001889859642905992, | |
| "loss": 0.0, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00018820697102166526, | |
| "loss": 0.0, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00018740308291919497, | |
| "loss": 0.0, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001865745268661689, | |
| "loss": 0.0, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00018572153670765365, | |
| "loss": 0.0, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00018484435318480332, | |
| "loss": 0.0, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001839432238669147, | |
| "loss": 0.0, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00018301840308155507, | |
| "loss": 0.0, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00018207015184278305, | |
| "loss": 0.0, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.000181098737777482, | |
| "loss": 0.0, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00018010443504982694, | |
| "loss": 0.0, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.000179087524283907, | |
| "loss": 0.0, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00017804829248452395, | |
| "loss": 0.0, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00017698703295619052, | |
| "loss": 0.0, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00017590404522035028, | |
| "loss": 0.0, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00017479963493084329, | |
| "loss": 0.0, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001736741137876405, | |
| "loss": 0.0, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00017252779944887235, | |
| "loss": 0.0, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00017136101544117525, | |
| "loss": 0.0, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00017017409106838207, | |
| "loss": 0.0, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00016896736131858208, | |
| "loss": 0.0, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001677411667695765, | |
| "loss": 0.0, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00016649585349275662, | |
| "loss": 0.0, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001652317729554313, | |
| "loss": 0.0, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001639492819216316, | |
| "loss": 0.0, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001626487423514207, | |
| "loss": 0.0, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00016133052129873693, | |
| "loss": 0.0, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001599949908077996, | |
| "loss": 0.0, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00015864252780810616, | |
| "loss": 0.0, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00015727351400805052, | |
| "loss": 0.0, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001558883357871928, | |
| "loss": 0.0, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00015448738408721052, | |
| "loss": 0.0, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00015307105430156255, | |
| "loss": 0.0, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001516397461638962, | |
| "loss": 0.0, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001501938636352297, | |
| "loss": 0.0, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00014873381478994134, | |
| "loss": 0.0, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00014726001170059792, | |
| "loss": 0.0, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00014577287032165468, | |
| "loss": 0.0, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00014427281037205945, | |
| "loss": 0.0, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00014276025521679471, | |
| "loss": 0.0, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00014123563174739037, | |
| "loss": 0.0, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00013969937026144118, | |
| "loss": 0.0, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00013815190434116317, | |
| "loss": 0.0, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00013659367073102268, | |
| "loss": 0.0, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00013502510921447323, | |
| "loss": 0.0, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00013344666248983432, | |
| "loss": 0.0, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.000131858776045348, | |
| "loss": 0.0, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00013026189803344774, | |
| "loss": 0.0, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00012865647914427544, | |
| "loss": 0.0, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00012704297247848216, | |
| "loss": 0.0, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00012542183341934872, | |
| "loss": 0.0, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00012379351950426187, | |
| "loss": 0.0, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001221584902955827, | |
| "loss": 0.0, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00012051720725094324, | |
| "loss": 0.0, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00011887013359300837, | |
| "loss": 0.0, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00011721773417873965, | |
| "loss": 0.0, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00011556047536819777, | |
| "loss": 0.0, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00011389882489292061, | |
| "loss": 0.0, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.0001122332517239147, | |
| "loss": 0.0, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00011056422593929635, | |
| "loss": 0.0, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00010889221859162062, | |
| "loss": 0.0, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00010721770157493527, | |
| "loss": 0.0, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.000105541147491597, | |
| "loss": 0.0, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00010386302951888804, | |
| "loss": 0.0, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00010218382127547022, | |
| "loss": 0.0, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.0, | |
| "learning_rate": 0.00010050399668771479, | |
| "loss": 0.0, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.882402985594515e-05, | |
| "loss": 0.0, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.71443949206304e-05, | |
| "loss": 0.0, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.546556592856789e-05, | |
| "loss": 0.0, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.378801669909197e-05, | |
| "loss": 0.0, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.211222069034695e-05, | |
| "loss": 0.0, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.043865086566214e-05, | |
| "loss": 0.0, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.87677795600663e-05, | |
| "loss": 0.0, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.710007834697969e-05, | |
| "loss": 0.0, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.543601790512083e-05, | |
| "loss": 0.0, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.377606788566597e-05, | |
| "loss": 0.0, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.212069677969851e-05, | |
| "loss": 0.0, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.047037178598567e-05, | |
| "loss": 0.0, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.882555867912017e-05, | |
| "loss": 0.0, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.718672167806354e-05, | |
| "loss": 0.0, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.55543233151289e-05, | |
| "loss": 0.0, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.392882430543928e-05, | |
| "loss": 0.0, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.231068341689923e-05, | |
| "loss": 0.0, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.070035734071574e-05, | |
| "loss": 0.0, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.909830056250527e-05, | |
| "loss": 0.0, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.750496523402352e-05, | |
| "loss": 0.0, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.592080104555357e-05, | |
| "loss": 0.0, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.434625509898897e-05, | |
| "loss": 0.0, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.278177178164721e-05, | |
| "loss": 0.0, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.122779264084932e-05, | |
| "loss": 0.0, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.968475625930124e-05, | |
| "loss": 0.0, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.815309813131153e-05, | |
| "loss": 0.0, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.663325053988112e-05, | |
| "loss": 0.0, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.5125642434699044e-05, | |
| "loss": 0.0, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.363069931107902e-05, | |
| "loss": 0.0, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.214884308987136e-05, | |
| "loss": 0.0, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.068049199838307e-05, | |
| "loss": 0.0, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.9226060452340825e-05, | |
| "loss": 0.0, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.7785958938929644e-05, | |
| "loss": 0.0, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.6360593900940074e-05, | |
| "loss": 0.0, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.4950367622057173e-05, | |
| "loss": 0.0, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.355567811332311e-05, | |
| "loss": 0.0, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.21769190008056e-05, | |
| "loss": 0.0, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.081447941450428e-05, | |
| "loss": 0.0, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.946874387852545e-05, | |
| "loss": 0.0, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.8140092202557185e-05, | |
| "loss": 0.0, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.682889937467493e-05, | |
| "loss": 0.0, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.553553545550768e-05, | |
| "loss": 0.0, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.426036547379528e-05, | |
| "loss": 0.0, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.300374932336533e-05, | |
| "loss": 0.0, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.176604166155976e-05, | |
| "loss": 0.0, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.054759180913921e-05, | |
| "loss": 0.0, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.9348743651693357e-05, | |
| "loss": 0.0, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.8169835542585587e-05, | |
| "loss": 0.0, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.7011200207458677e-05, | |
| "loss": 0.0, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.5873164650328996e-05, | |
| "loss": 0.0, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.4756050061295534e-05, | |
| "loss": 0.0, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.36601717258897e-05, | |
| "loss": 0.0, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.2585838936091754e-05, | |
| "loss": 0.0, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.153335490303856e-05, | |
| "loss": 0.0, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.0503016671447785e-05, | |
| "loss": 0.0, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.9495115035782307e-05, | |
| "loss": 0.0, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.8509934458178712e-05, | |
| "loss": 0.0, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.754775298816307e-05, | |
| "loss": 0.0, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.6608842184176243e-05, | |
| "loss": 0.0, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.5693467036931576e-05, | |
| "loss": 0.0, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.48018858946259e-05, | |
| "loss": 0.0, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.3934350390025463e-05, | |
| "loss": 0.0, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.3091105369447165e-05, | |
| "loss": 0.0, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.22723888236549e-05, | |
| "loss": 0.0, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.1478431820691083e-05, | |
| "loss": 0.0, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.0709458440661801e-05, | |
| "loss": 0.0, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.965685712494199e-06, | |
| "loss": 0.0, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.247323552684051e-06, | |
| "loss": 0.0, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.554574706050488e-06, | |
| "loss": 0.0, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.887634688515e-06, | |
| "loss": 0.0, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.0, | |
| "learning_rate": 7.246691731920485e-06, | |
| "loss": 0.0, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.631926730906324e-06, | |
| "loss": 0.0, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.043513191853978e-06, | |
| "loss": 0.0, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 0.0, | |
| "learning_rate": 5.481617183918053e-06, | |
| "loss": 0.0, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.946397292156158e-06, | |
| "loss": 0.0, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.438004572771182e-06, | |
| "loss": 0.0, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.9565825104783685e-06, | |
| "loss": 0.0, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.5022669780093497e-06, | |
| "loss": 0.0, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.0751861977645125e-06, | |
| "loss": 0.0, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.6754607056244883e-06, | |
| "loss": 0.0, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.303203316931102e-06, | |
| "loss": 0.0, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.9585190946472488e-06, | |
| "loss": 0.0, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.6415053197047725e-06, | |
| "loss": 0.0, | |
| "step": 985 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.3522514635486816e-06, | |
| "loss": 0.0, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.0908391628854041e-06, | |
| "loss": 0.0, | |
| "step": 995 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.0, | |
| "learning_rate": 8.57342196642319e-07, | |
| "loss": 0.0, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.0, | |
| "learning_rate": 6.518264651449779e-07, | |
| "loss": 0.0, | |
| "step": 1005 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.743499715179067e-07, | |
| "loss": 0.0, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.0, | |
| "learning_rate": 3.249628053142884e-07, | |
| "loss": 0.0, | |
| "step": 1015 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 0.0, | |
| "learning_rate": 2.0370712837906037e-07, | |
| "loss": 0.0, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.0, | |
| "learning_rate": 1.1061716294951118e-07, | |
| "loss": 0.0, | |
| "step": 1025 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 0.0, | |
| "learning_rate": 4.5719181996650705e-08, | |
| "loss": 0.0, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.0, | |
| "learning_rate": 9.031501810174981e-09, | |
| "loss": 0.0, | |
| "step": 1035 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": NaN, | |
| "eval_runtime": 266.2722, | |
| "eval_samples_per_second": 8.675, | |
| "eval_steps_per_second": 0.545, | |
| "step": 1039 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1039, | |
| "total_flos": 5523816681832448.0, | |
| "train_loss": 0.0, | |
| "train_runtime": 11435.8659, | |
| "train_samples_per_second": 2.908, | |
| "train_steps_per_second": 0.091 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 1039, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 5523816681832448.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |