fmm170/SLAM-MetaMath-7B
06
1{2 "best_metric": null,3 "best_model_checkpoint": null,4 "epoch": 3.0991735537190084,5 "eval_steps": 500,6 "global_step": 750,7 "is_hyper_param_search": false,8 "is_local_process_zero": true,9 "is_world_process_zero": true,10 "log_history": [11 {12 "epoch": 0.04,13 "grad_norm": 1.8158831755876264,14 "learning_rate": 1.9994733999271596e-05,15 "loss": 0.9462,16 "step": 1017 },18 {19 "epoch": 0.08,20 "grad_norm": 0.7720995032121265,21 "learning_rate": 1.997894154323911e-05,22 "loss": 0.6045,23 "step": 2024 },25 {26 "epoch": 0.12,27 "grad_norm": 1.806336281375529,28 "learning_rate": 1.995263926451954e-05,29 "loss": 0.5446,30 "step": 3031 },32 {33 "epoch": 0.17,34 "grad_norm": 0.6001969481128436,35 "learning_rate": 1.9915854864676665e-05,36 "loss": 0.5239,37 "step": 4038 },39 {40 "epoch": 0.21,41 "grad_norm": 0.6309057874111494,42 "learning_rate": 1.986862708504576e-05,43 "loss": 0.4963,44 "step": 5045 },46 {47 "epoch": 0.25,48 "grad_norm": 0.5486760047432362,49 "learning_rate": 1.9811005665931205e-05,50 "loss": 0.484,51 "step": 6052 },53 {54 "epoch": 0.29,55 "grad_norm": 0.7847931435043445,56 "learning_rate": 1.9743051294220015e-05,57 "loss": 0.4729,58 "step": 7059 },60 {61 "epoch": 0.33,62 "grad_norm": 0.6422404181588646,63 "learning_rate": 1.966483553946637e-05,64 "loss": 0.4631,65 "step": 8066 },67 {68 "epoch": 0.37,69 "grad_norm": 0.46980286610936084,70 "learning_rate": 1.9576440778514576e-05,71 "loss": 0.4496,72 "step": 9073 },74 {75 "epoch": 0.41,76 "grad_norm": 0.6748157285016159,77 "learning_rate": 1.947796010873974e-05,78 "loss": 0.4473,79 "step": 10080 },81 {82 "epoch": 0.45,83 "grad_norm": 0.6701373318229686,84 "learning_rate": 1.936949724999762e-05,85 "loss": 0.4491,86 "step": 11087 },88 {89 "epoch": 0.5,90 "grad_norm": 0.5733561102044787,91 "learning_rate": 1.9251166435386837e-05,92 "loss": 0.4374,93 "step": 12094 },95 {96 "epoch": 0.54,97 "grad_norm": 0.578554895730767,98 "learning_rate": 1.9123092290938587e-05,99 "loss": 0.4376,100 "step": 130101 },102 {103 "epoch": 0.58,104 "grad_norm": 0.5367282139195013,105 "learning_rate": 1.8985409704360457e-05,106 "loss": 0.4332,107 "step": 140108 },109 {110 "epoch": 0.62,111 "grad_norm": 0.5624688722686745,112 "learning_rate": 1.883826368297269e-05,113 "loss": 0.4277,114 "step": 150115 },116 {117 "epoch": 0.66,118 "grad_norm": 0.5839028700275709,119 "learning_rate": 1.868180920098644e-05,120 "loss": 0.4254,121 "step": 160122 },123 {124 "epoch": 0.7,125 "grad_norm": 0.4935136821032842,126 "learning_rate": 1.8516211036284937e-05,127 "loss": 0.4207,128 "step": 170129 },130 {131 "epoch": 0.74,132 "grad_norm": 0.5586132232899661,133 "learning_rate": 1.834164359687937e-05,134 "loss": 0.4161,135 "step": 180136 },137 {138 "epoch": 0.79,139 "grad_norm": 0.566512709511239,140 "learning_rate": 1.815829073722234e-05,141 "loss": 0.4116,142 "step": 190143 },144 {145 "epoch": 0.83,146 "grad_norm": 0.5431202464449885,147 "learning_rate": 1.796634556457236e-05,148 "loss": 0.4118,149 "step": 200150 },151 {152 "epoch": 0.87,153 "grad_norm": 0.550094582271266,154 "learning_rate": 1.7766010235613225e-05,155 "loss": 0.4112,156 "step": 210157 },158 {159 "epoch": 0.91,160 "grad_norm": 0.5212828857696244,161 "learning_rate": 1.7557495743542586e-05,162 "loss": 0.4098,163 "step": 220164 },165 {166 "epoch": 0.95,167 "grad_norm": 0.5034678718408173,168 "learning_rate": 1.7341021695853856e-05,169 "loss": 0.4016,170 "step": 230171 },172 {173 "epoch": 0.99,174 "grad_norm": 0.5377071037066575,175 "learning_rate": 1.7116816083045603e-05,176 "loss": 0.4023,177 "step": 240178 },179 {180 "epoch": 1.03,181 "grad_norm": 0.5358132554777233,182 "learning_rate": 1.6885115038501903e-05,183 "loss": 0.3877,184 "step": 250185 },186 {187 "epoch": 1.07,188 "grad_norm": 0.5373094775514344,189 "learning_rate": 1.6646162589796616e-05,190 "loss": 0.3792,191 "step": 260192 },193 {194 "epoch": 1.12,195 "grad_norm": 0.566594664706901,196 "learning_rate": 1.6400210401683535e-05,197 "loss": 0.3784,198 "step": 270199 },200 {201 "epoch": 1.16,202 "grad_norm": 0.6557675132012587,203 "learning_rate": 1.614751751104301e-05,204 "loss": 0.3735,205 "step": 280206 },207 {208 "epoch": 1.2,209 "grad_norm": 0.5857582835459575,210 "learning_rate": 1.5888350054064275e-05,211 "loss": 0.3735,212 "step": 290213 },214 {215 "epoch": 1.24,216 "grad_norm": 0.6287898080255828,217 "learning_rate": 1.562298098595078e-05,218 "loss": 0.3745,219 "step": 300220 },221 {222 "epoch": 1.28,223 "grad_norm": 0.5387884908676019,224 "learning_rate": 1.5351689793443714e-05,225 "loss": 0.3706,226 "step": 310227 },228 {229 "epoch": 1.32,230 "grad_norm": 0.6505705438520281,231 "learning_rate": 1.5074762200466557e-05,232 "loss": 0.3734,233 "step": 320234 },235 {236 "epoch": 1.36,237 "grad_norm": 0.5405019807593993,238 "learning_rate": 1.479248986720057e-05,239 "loss": 0.3668,240 "step": 330241 },242 {243 "epoch": 1.4,244 "grad_norm": 0.5205231569472534,245 "learning_rate": 1.4505170082908269e-05,246 "loss": 0.3664,247 "step": 340248 },249 {250 "epoch": 1.45,251 "grad_norm": 0.7025073411496863,252 "learning_rate": 1.4213105452828332e-05,253 "loss": 0.3659,254 "step": 350255 },256 {257 "epoch": 1.49,258 "grad_norm": 0.7368388542259531,259 "learning_rate": 1.3916603579471705e-05,260 "loss": 0.3645,261 "step": 360262 },263 {264 "epoch": 1.53,265 "grad_norm": 0.7151238148197194,266 "learning_rate": 1.36159767386546e-05,267 "loss": 0.3618,268 "step": 370269 },270 {271 "epoch": 1.57,272 "grad_norm": 0.604346829441288,273 "learning_rate": 1.3311541550609566e-05,274 "loss": 0.3603,275 "step": 380276 },277 {278 "epoch": 1.61,279 "grad_norm": 0.6661473924162761,280 "learning_rate": 1.3003618646520999e-05,281 "loss": 0.3606,282 "step": 390283 },284 {285 "epoch": 1.65,286 "grad_norm": 0.5466798434132922,287 "learning_rate": 1.2692532330836346e-05,288 "loss": 0.3583,289 "step": 400290 },291 {292 "epoch": 1.69,293 "grad_norm": 0.6779030801513176,294 "learning_rate": 1.2378610239708606e-05,295 "loss": 0.3573,296 "step": 410297 },298 {299 "epoch": 1.74,300 "grad_norm": 0.5083437268440109,301 "learning_rate": 1.2062182995929883e-05,302 "loss": 0.3547,303 "step": 420304 },305 {306 "epoch": 1.78,307 "grad_norm": 0.6067040915613267,308 "learning_rate": 1.1743583860719425e-05,309 "loss": 0.3532,310 "step": 430311 },312 {313 "epoch": 1.82,314 "grad_norm": 0.6645707597609398,315 "learning_rate": 1.1423148382732854e-05,316 "loss": 0.3526,317 "step": 440318 },319 {320 "epoch": 1.86,321 "grad_norm": 0.5862283620521659,322 "learning_rate": 1.1101214044662258e-05,323 "loss": 0.3483,324 "step": 450325 },326 {327 "epoch": 1.9,328 "grad_norm": 0.6152606534643925,329 "learning_rate": 1.0778119907799399e-05,330 "loss": 0.3517,331 "step": 460332 },333 {334 "epoch": 1.94,335 "grad_norm": 0.6300108077242406,336 "learning_rate": 1.0454206254936287e-05,337 "loss": 0.3478,338 "step": 470339 },340 {341 "epoch": 1.98,342 "grad_norm": 0.6581226831741345,343 "learning_rate": 1.012981423197931e-05,344 "loss": 0.3455,345 "step": 480346 },347 {348 "epoch": 2.02,349 "grad_norm": 0.7070055372083373,350 "learning_rate": 9.805285488654298e-06,351 "loss": 0.3286,352 "step": 490353 },354 {355 "epoch": 2.07,356 "grad_norm": 0.7303222672572833,357 "learning_rate": 9.480961818681004e-06,358 "loss": 0.3178,359 "step": 500360 },361 {362 "epoch": 2.11,363 "grad_norm": 0.6358667770163009,364 "learning_rate": 9.157184799795889e-06,365 "loss": 0.3203,366 "step": 510367 },368 {369 "epoch": 2.15,370 "grad_norm": 0.7022272949079124,371 "learning_rate": 8.83429543400241e-06,372 "loss": 0.3168,373 "step": 520374 },375 {376 "epoch": 2.19,377 "grad_norm": 0.6500559223400848,378 "learning_rate": 8.51263378842766e-06,379 "loss": 0.3152,380 "step": 530381 },382 {383 "epoch": 2.23,384 "grad_norm": 0.6968759254483634,385 "learning_rate": 8.19253863716362e-06,386 "loss": 0.318,387 "step": 540388 },389 {390 "epoch": 2.27,391 "grad_norm": 0.7646027572063252,392 "learning_rate": 7.874347104470234e-06,393 "loss": 0.3126,394 "step": 550395 },396 {397 "epoch": 2.31,398 "grad_norm": 0.6487425810329057,399 "learning_rate": 7.558394309716088e-06,400 "loss": 0.3165,401 "step": 560402 },403 {404 "epoch": 2.36,405 "grad_norm": 0.6910383894617135,406 "learning_rate": 7.245013014430645e-06,407 "loss": 0.3106,408 "step": 570409 },410 {411 "epoch": 2.4,412 "grad_norm": 0.7144330622054906,413 "learning_rate": 6.934533271839751e-06,414 "loss": 0.3168,415 "step": 580416 },417 {418 "epoch": 2.44,419 "grad_norm": 0.653595551359032,420 "learning_rate": 6.627282079253539e-06,421 "loss": 0.3117,422 "step": 590423 },424 {425 "epoch": 2.48,426 "grad_norm": 0.7309868347167029,427 "learning_rate": 6.323583033672799e-06,428 "loss": 0.3065,429 "step": 600430 },431 {432 "epoch": 2.52,433 "grad_norm": 0.7366115924981546,434 "learning_rate": 6.023755990976582e-06,435 "loss": 0.3085,436 "step": 610437 },438 {439 "epoch": 2.56,440 "grad_norm": 0.6638801997661135,441 "learning_rate": 5.728116729049929e-06,442 "loss": 0.3001,443 "step": 620444 },445 {446 "epoch": 2.6,447 "grad_norm": 0.680122959471132,448 "learning_rate": 5.436976615206577e-06,449 "loss": 0.3073,450 "step": 630451 },452 {453 "epoch": 2.64,454 "grad_norm": 0.7121214178744179,455 "learning_rate": 5.1506422782568345e-06,456 "loss": 0.3047,457 "step": 640458 },459 {460 "epoch": 2.69,461 "grad_norm": 0.7310044879071071,462 "learning_rate": 4.869415285566091e-06,463 "loss": 0.3021,464 "step": 650465 },466 {467 "epoch": 2.73,468 "grad_norm": 0.8413979985957987,469 "learning_rate": 4.593591825444028e-06,470 "loss": 0.3034,471 "step": 660472 },473 {474 "epoch": 2.77,475 "grad_norm": 0.7363824394289317,476 "learning_rate": 4.3234623951990115e-06,477 "loss": 0.3013,478 "step": 670479 },480 {481 "epoch": 2.81,482 "grad_norm": 0.6819313325698338,483 "learning_rate": 4.059311495186338e-06,484 "loss": 0.2994,485 "step": 680486 },487 {488 "epoch": 2.85,489 "grad_norm": 0.7194669299210141,490 "learning_rate": 3.8014173291723846e-06,491 "loss": 0.3051,492 "step": 690493 },494 {495 "epoch": 2.89,496 "grad_norm": 0.6888037587213363,497 "learning_rate": 3.550051511330361e-06,498 "loss": 0.2974,499 "step": 700500 },501 {502 "epoch": 2.93,503 "grad_norm": 0.7066646700469487,504 "learning_rate": 3.305478780176241e-06,505 "loss": 0.298,506 "step": 710507 },508 {509 "epoch": 2.98,510 "grad_norm": 0.6584460661936957,511 "learning_rate": 3.0679567197461135e-06,512 "loss": 0.2993,513 "step": 720514 },515 {516 "epoch": 3.02,517 "grad_norm": 0.7131090221570907,518 "learning_rate": 2.8377354883086105e-06,519 "loss": 0.2879,520 "step": 730521 },522 {523 "epoch": 3.06,524 "grad_norm": 0.8676596722156668,525 "learning_rate": 2.6150575548982295e-06,526 "loss": 0.2783,527 "step": 740528 },529 {530 "epoch": 3.1,531 "grad_norm": 0.7765116623139634,532 "learning_rate": 2.400157443946881e-06,533 "loss": 0.2765,534 "step": 750535 }536 ],537 "logging_steps": 10,538 "max_steps": 968,539 "num_input_tokens_seen": 0,540 "num_train_epochs": 4,541 "save_steps": 50,542 "total_flos": 1.1342029719789896e+19,543 "train_batch_size": 64,544 "trial_name": null,545 "trial_params": null546}547