| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9994655264564404, | |
| "eval_steps": 500, | |
| "global_step": 935, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0010689470871191875, | |
| "grad_norm": 23.473513373152795, | |
| "learning_rate": 1.0638297872340426e-07, | |
| "loss": 1.353, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005344735435595938, | |
| "grad_norm": 21.711768425322266, | |
| "learning_rate": 5.319148936170213e-07, | |
| "loss": 1.3519, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.010689470871191877, | |
| "grad_norm": 8.713068756520672, | |
| "learning_rate": 1.0638297872340427e-06, | |
| "loss": 1.2088, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.016034206306787813, | |
| "grad_norm": 8.310866409645099, | |
| "learning_rate": 1.595744680851064e-06, | |
| "loss": 1.0635, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.021378941742383754, | |
| "grad_norm": 3.066210523084854, | |
| "learning_rate": 2.1276595744680853e-06, | |
| "loss": 0.9417, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02672367717797969, | |
| "grad_norm": 2.493239812049204, | |
| "learning_rate": 2.6595744680851065e-06, | |
| "loss": 0.8884, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.032068412613575625, | |
| "grad_norm": 2.403610159527061, | |
| "learning_rate": 3.191489361702128e-06, | |
| "loss": 0.8557, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03741314804917156, | |
| "grad_norm": 2.3974276447793694, | |
| "learning_rate": 3.723404255319149e-06, | |
| "loss": 0.8349, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.04275788348476751, | |
| "grad_norm": 2.2921630085655025, | |
| "learning_rate": 4.255319148936171e-06, | |
| "loss": 0.8163, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.048102618920363445, | |
| "grad_norm": 2.3116910060192817, | |
| "learning_rate": 4.787234042553192e-06, | |
| "loss": 0.798, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.05344735435595938, | |
| "grad_norm": 2.242005189567634, | |
| "learning_rate": 5.319148936170213e-06, | |
| "loss": 0.7863, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05879208979155532, | |
| "grad_norm": 2.536220362549579, | |
| "learning_rate": 5.851063829787235e-06, | |
| "loss": 0.7663, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.06413682522715125, | |
| "grad_norm": 2.492174958150876, | |
| "learning_rate": 6.382978723404256e-06, | |
| "loss": 0.7579, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06948156066274719, | |
| "grad_norm": 2.3311723974999095, | |
| "learning_rate": 6.914893617021278e-06, | |
| "loss": 0.7397, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07482629609834313, | |
| "grad_norm": 2.5040282086769317, | |
| "learning_rate": 7.446808510638298e-06, | |
| "loss": 0.7338, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08017103153393906, | |
| "grad_norm": 2.411086582278843, | |
| "learning_rate": 7.97872340425532e-06, | |
| "loss": 0.7215, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08551576696953501, | |
| "grad_norm": 2.3433794783119883, | |
| "learning_rate": 8.510638297872341e-06, | |
| "loss": 0.7124, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09086050240513095, | |
| "grad_norm": 2.354225850924038, | |
| "learning_rate": 9.042553191489362e-06, | |
| "loss": 0.7177, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.09620523784072689, | |
| "grad_norm": 2.4091177426727266, | |
| "learning_rate": 9.574468085106385e-06, | |
| "loss": 0.7163, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.10154997327632283, | |
| "grad_norm": 2.498651758504887, | |
| "learning_rate": 9.999965114314806e-06, | |
| "loss": 0.7031, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.10689470871191876, | |
| "grad_norm": 2.423896856776365, | |
| "learning_rate": 9.998744166446685e-06, | |
| "loss": 0.7005, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.1122394441475147, | |
| "grad_norm": 2.387782305227773, | |
| "learning_rate": 9.995779421092695e-06, | |
| "loss": 0.6846, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.11758417958311064, | |
| "grad_norm": 2.424890050569407, | |
| "learning_rate": 9.991071912495701e-06, | |
| "loss": 0.7027, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.12292891501870658, | |
| "grad_norm": 2.115700167815373, | |
| "learning_rate": 9.984623282856502e-06, | |
| "loss": 0.6923, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1282736504543025, | |
| "grad_norm": 2.0709758825284363, | |
| "learning_rate": 9.97643578176095e-06, | |
| "loss": 0.6859, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.13361838588989844, | |
| "grad_norm": 2.1305031798421186, | |
| "learning_rate": 9.966512265395188e-06, | |
| "loss": 0.6846, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.13896312132549438, | |
| "grad_norm": 2.0044100900913384, | |
| "learning_rate": 9.95485619554928e-06, | |
| "loss": 0.6799, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.14430785676109031, | |
| "grad_norm": 2.050948697307653, | |
| "learning_rate": 9.941471638409576e-06, | |
| "loss": 0.6767, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.14965259219668625, | |
| "grad_norm": 1.952049269951998, | |
| "learning_rate": 9.926363263140234e-06, | |
| "loss": 0.6669, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1549973276322822, | |
| "grad_norm": 2.099731647032975, | |
| "learning_rate": 9.90953634025439e-06, | |
| "loss": 0.6786, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.16034206306787813, | |
| "grad_norm": 1.993767696614697, | |
| "learning_rate": 9.890996739775562e-06, | |
| "loss": 0.6674, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.16568679850347406, | |
| "grad_norm": 2.0517367712519503, | |
| "learning_rate": 9.870750929189914e-06, | |
| "loss": 0.6748, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.17103153393907003, | |
| "grad_norm": 2.0491814339157486, | |
| "learning_rate": 9.848805971190074e-06, | |
| "loss": 0.6621, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.17637626937466597, | |
| "grad_norm": 2.0777487120565397, | |
| "learning_rate": 9.825169521211354e-06, | |
| "loss": 0.6543, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.1817210048102619, | |
| "grad_norm": 2.351975397829705, | |
| "learning_rate": 9.799849824761159e-06, | |
| "loss": 0.6552, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.18706574024585784, | |
| "grad_norm": 2.0302930695761034, | |
| "learning_rate": 9.772855714542569e-06, | |
| "loss": 0.6525, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.19241047568145378, | |
| "grad_norm": 2.188255761505827, | |
| "learning_rate": 9.744196607373086e-06, | |
| "loss": 0.6512, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.19775521111704972, | |
| "grad_norm": 2.1631601855326763, | |
| "learning_rate": 9.71388250089959e-06, | |
| "loss": 0.6506, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.20309994655264565, | |
| "grad_norm": 2.041485612804426, | |
| "learning_rate": 9.681923970110698e-06, | |
| "loss": 0.6402, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2084446819882416, | |
| "grad_norm": 2.1331611096395973, | |
| "learning_rate": 9.648332163647705e-06, | |
| "loss": 0.6234, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.21378941742383753, | |
| "grad_norm": 2.076285244795538, | |
| "learning_rate": 9.613118799915417e-06, | |
| "loss": 0.6422, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.21913415285943347, | |
| "grad_norm": 1.9966049947480653, | |
| "learning_rate": 9.576296162994214e-06, | |
| "loss": 0.62, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.2244788882950294, | |
| "grad_norm": 2.003867139711824, | |
| "learning_rate": 9.537877098354787e-06, | |
| "loss": 0.6431, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.22982362373062534, | |
| "grad_norm": 1.9196987285290612, | |
| "learning_rate": 9.497875008377033e-06, | |
| "loss": 0.6234, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.23516835916622128, | |
| "grad_norm": 2.1957665769389774, | |
| "learning_rate": 9.456303847674674e-06, | |
| "loss": 0.6256, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.24051309460181722, | |
| "grad_norm": 1.884124911991183, | |
| "learning_rate": 9.41317811822723e-06, | |
| "loss": 0.6052, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.24585783003741316, | |
| "grad_norm": 1.9253697770132223, | |
| "learning_rate": 9.36851286432104e-06, | |
| "loss": 0.6228, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.25120256547300907, | |
| "grad_norm": 1.8020291089671914, | |
| "learning_rate": 9.322323667301113e-06, | |
| "loss": 0.6191, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.256547300908605, | |
| "grad_norm": 1.9107218407751854, | |
| "learning_rate": 9.274626640135616e-06, | |
| "loss": 0.6121, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.26189203634420094, | |
| "grad_norm": 1.8979104758577288, | |
| "learning_rate": 9.225438421794919e-06, | |
| "loss": 0.616, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2672367717797969, | |
| "grad_norm": 1.9569646424906317, | |
| "learning_rate": 9.174776171447126e-06, | |
| "loss": 0.6047, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2725815072153928, | |
| "grad_norm": 2.0364213519267715, | |
| "learning_rate": 9.12265756247216e-06, | |
| "loss": 0.6093, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.27792624265098875, | |
| "grad_norm": 1.9725707409017998, | |
| "learning_rate": 9.06910077629645e-06, | |
| "loss": 0.599, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.2832709780865847, | |
| "grad_norm": 1.9492899906874035, | |
| "learning_rate": 9.014124496050391e-06, | |
| "loss": 0.5902, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.28861571352218063, | |
| "grad_norm": 2.174543196329182, | |
| "learning_rate": 8.957747900050797e-06, | |
| "loss": 0.5937, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.29396044895777657, | |
| "grad_norm": 2.4160108027006615, | |
| "learning_rate": 8.899990655110586e-06, | |
| "loss": 0.6018, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.2993051843933725, | |
| "grad_norm": 2.087978454933178, | |
| "learning_rate": 8.840872909678081e-06, | |
| "loss": 0.5832, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.30464991982896844, | |
| "grad_norm": 1.9762659778049954, | |
| "learning_rate": 8.780415286808284e-06, | |
| "loss": 0.5922, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.3099946552645644, | |
| "grad_norm": 2.269903050974625, | |
| "learning_rate": 8.718638876968564e-06, | |
| "loss": 0.587, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3153393907001603, | |
| "grad_norm": 2.0490480592368043, | |
| "learning_rate": 8.655565230681329e-06, | |
| "loss": 0.5748, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.32068412613575625, | |
| "grad_norm": 2.2764533528101167, | |
| "learning_rate": 8.591216351006181e-06, | |
| "loss": 0.575, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3260288615713522, | |
| "grad_norm": 2.0737769584580072, | |
| "learning_rate": 8.525614685864209e-06, | |
| "loss": 0.5716, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.33137359700694813, | |
| "grad_norm": 1.9102889564355032, | |
| "learning_rate": 8.458783120207099e-06, | |
| "loss": 0.5686, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3367183324425441, | |
| "grad_norm": 1.8800981653614064, | |
| "learning_rate": 8.390744968033785e-06, | |
| "loss": 0.5629, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.34206306787814006, | |
| "grad_norm": 1.96332588340258, | |
| "learning_rate": 8.321523964257431e-06, | |
| "loss": 0.5657, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.347407803313736, | |
| "grad_norm": 1.9485967093770373, | |
| "learning_rate": 8.251144256425562e-06, | |
| "loss": 0.575, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.35275253874933193, | |
| "grad_norm": 1.929135934757674, | |
| "learning_rate": 8.179630396296285e-06, | |
| "loss": 0.5632, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.35809727418492787, | |
| "grad_norm": 1.9559096020055298, | |
| "learning_rate": 8.107007331273449e-06, | |
| "loss": 0.5626, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.3634420096205238, | |
| "grad_norm": 2.012794105136495, | |
| "learning_rate": 8.033300395703845e-06, | |
| "loss": 0.5546, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.36878674505611975, | |
| "grad_norm": 1.9997480665852025, | |
| "learning_rate": 7.958535302039368e-06, | |
| "loss": 0.5458, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.3741314804917157, | |
| "grad_norm": 1.8451676891383395, | |
| "learning_rate": 7.88273813186732e-06, | |
| "loss": 0.5483, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3794762159273116, | |
| "grad_norm": 1.895313478955964, | |
| "learning_rate": 7.805935326811913e-06, | |
| "loss": 0.5517, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.38482095136290756, | |
| "grad_norm": 1.9301374165518745, | |
| "learning_rate": 7.728153679310186e-06, | |
| "loss": 0.5464, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.3901656867985035, | |
| "grad_norm": 2.014621013972388, | |
| "learning_rate": 7.649420323265547e-06, | |
| "loss": 0.5441, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.39551042223409943, | |
| "grad_norm": 1.9682365332357328, | |
| "learning_rate": 7.569762724582179e-06, | |
| "loss": 0.5247, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.40085515766969537, | |
| "grad_norm": 1.9950047285262882, | |
| "learning_rate": 7.48920867158365e-06, | |
| "loss": 0.5435, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.4061998931052913, | |
| "grad_norm": 1.8591367525423361, | |
| "learning_rate": 7.407786265319023e-06, | |
| "loss": 0.5363, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.41154462854088725, | |
| "grad_norm": 2.038203032181739, | |
| "learning_rate": 7.325523909759902e-06, | |
| "loss": 0.5312, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.4168893639764832, | |
| "grad_norm": 2.1300865254144856, | |
| "learning_rate": 7.242450301891772e-06, | |
| "loss": 0.5257, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.4222340994120791, | |
| "grad_norm": 1.882363933745456, | |
| "learning_rate": 7.158594421703152e-06, | |
| "loss": 0.5096, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.42757883484767506, | |
| "grad_norm": 1.8727951959035394, | |
| "learning_rate": 7.073985522076001e-06, | |
| "loss": 0.5184, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.432923570283271, | |
| "grad_norm": 1.9894963218262207, | |
| "learning_rate": 6.9886531185809385e-06, | |
| "loss": 0.5259, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.43826830571886694, | |
| "grad_norm": 2.0264734474019854, | |
| "learning_rate": 6.902626979180821e-06, | |
| "loss": 0.5287, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.4436130411544629, | |
| "grad_norm": 1.8722413198751242, | |
| "learning_rate": 6.8159371138462745e-06, | |
| "loss": 0.5203, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.4489577765900588, | |
| "grad_norm": 1.9496986261286822, | |
| "learning_rate": 6.728613764086806e-06, | |
| "loss": 0.5117, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.45430251202565475, | |
| "grad_norm": 1.9774605540105135, | |
| "learning_rate": 6.640687392401132e-06, | |
| "loss": 0.5031, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.4596472474612507, | |
| "grad_norm": 1.954818070964011, | |
| "learning_rate": 6.552188671650434e-06, | |
| "loss": 0.5031, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4649919828968466, | |
| "grad_norm": 1.9617332949799173, | |
| "learning_rate": 6.46314847435821e-06, | |
| "loss": 0.5066, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.47033671833244256, | |
| "grad_norm": 2.004929689013094, | |
| "learning_rate": 6.373597861940488e-06, | |
| "loss": 0.4988, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4756814537680385, | |
| "grad_norm": 1.9774178895562935, | |
| "learning_rate": 6.283568073870147e-06, | |
| "loss": 0.4975, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.48102618920363444, | |
| "grad_norm": 1.9264282288547034, | |
| "learning_rate": 6.1930905167791025e-06, | |
| "loss": 0.4995, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.4863709246392304, | |
| "grad_norm": 1.8764137691966425, | |
| "learning_rate": 6.102196753502202e-06, | |
| "loss": 0.4808, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.4917156600748263, | |
| "grad_norm": 2.050651493580463, | |
| "learning_rate": 6.010918492066628e-06, | |
| "loss": 0.482, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.49706039551042225, | |
| "grad_norm": 1.9228723322188161, | |
| "learning_rate": 5.919287574630628e-06, | |
| "loss": 0.4843, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.5024051309460181, | |
| "grad_norm": 1.9851687404052978, | |
| "learning_rate": 5.827335966375485e-06, | |
| "loss": 0.4894, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5077498663816141, | |
| "grad_norm": 1.9039931208460692, | |
| "learning_rate": 5.735095744354543e-06, | |
| "loss": 0.4673, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.51309460181721, | |
| "grad_norm": 1.9780097627545055, | |
| "learning_rate": 5.642599086303233e-06, | |
| "loss": 0.4788, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.518439337252806, | |
| "grad_norm": 2.0794416013824994, | |
| "learning_rate": 5.5498782594139476e-06, | |
| "loss": 0.4662, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.5237840726884019, | |
| "grad_norm": 1.9483487482401622, | |
| "learning_rate": 5.456965609079741e-06, | |
| "loss": 0.4763, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5291288081239979, | |
| "grad_norm": 1.8945000444925455, | |
| "learning_rate": 5.363893547610715e-06, | |
| "loss": 0.4807, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.5344735435595938, | |
| "grad_norm": 1.9768379335510178, | |
| "learning_rate": 5.270694542927089e-06, | |
| "loss": 0.4803, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5398182789951897, | |
| "grad_norm": 1.9758527545184905, | |
| "learning_rate": 5.1774011072328575e-06, | |
| "loss": 0.4627, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.5451630144307856, | |
| "grad_norm": 1.866392771665028, | |
| "learning_rate": 5.084045785674001e-06, | |
| "loss": 0.4608, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5505077498663816, | |
| "grad_norm": 1.957078999847815, | |
| "learning_rate": 4.9906611449852035e-06, | |
| "loss": 0.4542, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.5558524853019775, | |
| "grad_norm": 1.9342632236304118, | |
| "learning_rate": 4.897279762129044e-06, | |
| "loss": 0.4547, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5611972207375735, | |
| "grad_norm": 1.966938473585663, | |
| "learning_rate": 4.8039342129316175e-06, | |
| "loss": 0.4526, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5665419561731694, | |
| "grad_norm": 1.9823937668875962, | |
| "learning_rate": 4.710657060718547e-06, | |
| "loss": 0.4503, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.5718866916087654, | |
| "grad_norm": 1.931604909577632, | |
| "learning_rate": 4.617480844955367e-06, | |
| "loss": 0.4543, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5772314270443613, | |
| "grad_norm": 1.9170992207978994, | |
| "learning_rate": 4.52443806989622e-06, | |
| "loss": 0.4383, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5825761624799572, | |
| "grad_norm": 1.9043676805575804, | |
| "learning_rate": 4.431561193244852e-06, | |
| "loss": 0.4546, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.5879208979155531, | |
| "grad_norm": 1.9233137290689744, | |
| "learning_rate": 4.338882614831817e-06, | |
| "loss": 0.446, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5932656333511491, | |
| "grad_norm": 1.8515724013719734, | |
| "learning_rate": 4.246434665311907e-06, | |
| "loss": 0.4321, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.598610368786745, | |
| "grad_norm": 1.8828506758113701, | |
| "learning_rate": 4.154249594885687e-06, | |
| "loss": 0.4337, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.603955104222341, | |
| "grad_norm": 1.9687942950195187, | |
| "learning_rate": 4.062359562049109e-06, | |
| "loss": 0.431, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.6092998396579369, | |
| "grad_norm": 1.8995143280590803, | |
| "learning_rate": 3.970796622375116e-06, | |
| "loss": 0.4405, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.6146445750935329, | |
| "grad_norm": 1.8557131453687543, | |
| "learning_rate": 3.879592717331141e-06, | |
| "loss": 0.4283, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6199893105291288, | |
| "grad_norm": 1.9309890578975801, | |
| "learning_rate": 3.78877966313642e-06, | |
| "loss": 0.4367, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.6253340459647247, | |
| "grad_norm": 1.973985583949257, | |
| "learning_rate": 3.698389139663003e-06, | |
| "loss": 0.4324, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.6306787814003206, | |
| "grad_norm": 1.8679236635919254, | |
| "learning_rate": 3.608452679384311e-06, | |
| "loss": 0.4289, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6360235168359166, | |
| "grad_norm": 1.827491182434645, | |
| "learning_rate": 3.5190016563751316e-06, | |
| "loss": 0.427, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.6413682522715125, | |
| "grad_norm": 2.0739567283668583, | |
| "learning_rate": 3.4300672753668635e-06, | |
| "loss": 0.4163, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6467129877071085, | |
| "grad_norm": 1.8900939036994293, | |
| "learning_rate": 3.34168056086183e-06, | |
| "loss": 0.4322, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.6520577231427044, | |
| "grad_norm": 1.9421083071105274, | |
| "learning_rate": 3.2538723463104737e-06, | |
| "loss": 0.4139, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6574024585783004, | |
| "grad_norm": 1.9389760896779549, | |
| "learning_rate": 3.166673263355199e-06, | |
| "loss": 0.4238, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6627471940138963, | |
| "grad_norm": 1.874356566114327, | |
| "learning_rate": 3.0801137311446087e-06, | |
| "loss": 0.4165, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6680919294494923, | |
| "grad_norm": 1.9580525970654405, | |
| "learning_rate": 2.994223945721872e-06, | |
| "loss": 0.4082, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.6734366648850882, | |
| "grad_norm": 1.9147611470583412, | |
| "learning_rate": 2.9090338694909254e-06, | |
| "loss": 0.4219, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6787814003206841, | |
| "grad_norm": 1.86847720750241, | |
| "learning_rate": 2.8245732207641705e-06, | |
| "loss": 0.4132, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.6841261357562801, | |
| "grad_norm": 1.8982814741174805, | |
| "learning_rate": 2.740871463395325e-06, | |
| "loss": 0.4129, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.689470871191876, | |
| "grad_norm": 1.901393825227574, | |
| "learning_rate": 2.65795779650105e-06, | |
| "loss": 0.4033, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.694815606627472, | |
| "grad_norm": 1.8509358915404532, | |
| "learning_rate": 2.575861144274914e-06, | |
| "loss": 0.4013, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.7001603420630679, | |
| "grad_norm": 1.8958219311827695, | |
| "learning_rate": 2.4946101458972744e-06, | |
| "loss": 0.4027, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.7055050774986639, | |
| "grad_norm": 1.8426235286345023, | |
| "learning_rate": 2.414233145544585e-06, | |
| "loss": 0.3964, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.7108498129342598, | |
| "grad_norm": 1.9191012877180513, | |
| "learning_rate": 2.33475818250161e-06, | |
| "loss": 0.3954, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.7161945483698557, | |
| "grad_norm": 1.902716391588277, | |
| "learning_rate": 2.256212981379996e-06, | |
| "loss": 0.3957, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.7215392838054516, | |
| "grad_norm": 1.947059841210012, | |
| "learning_rate": 2.178624942446626e-06, | |
| "loss": 0.401, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7268840192410476, | |
| "grad_norm": 1.917543839246915, | |
| "learning_rate": 2.1020211320651135e-06, | |
| "loss": 0.3888, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.7322287546766435, | |
| "grad_norm": 1.8327154685147313, | |
| "learning_rate": 2.0264282732537827e-06, | |
| "loss": 0.4003, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.7375734901122395, | |
| "grad_norm": 1.7919182630342896, | |
| "learning_rate": 1.9518727363634187e-06, | |
| "loss": 0.3873, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7429182255478354, | |
| "grad_norm": 1.9169963711238958, | |
| "learning_rate": 1.8783805298780427e-06, | |
| "loss": 0.4018, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.7482629609834314, | |
| "grad_norm": 1.8380910216299624, | |
| "learning_rate": 1.8059772913419305e-06, | |
| "loss": 0.3946, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7536076964190273, | |
| "grad_norm": 1.9558297629437302, | |
| "learning_rate": 1.7346882784160346e-06, | |
| "loss": 0.3916, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.7589524318546232, | |
| "grad_norm": 1.897087729189459, | |
| "learning_rate": 1.6645383600669124e-06, | |
| "loss": 0.3919, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.7642971672902191, | |
| "grad_norm": 1.8755131544746726, | |
| "learning_rate": 1.5955520078912628e-06, | |
| "loss": 0.3846, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.7696419027258151, | |
| "grad_norm": 1.8807950794746984, | |
| "learning_rate": 1.527753287579084e-06, | |
| "loss": 0.3801, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.774986638161411, | |
| "grad_norm": 1.8519074792344092, | |
| "learning_rate": 1.461165850518424e-06, | |
| "loss": 0.3788, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.780331373597007, | |
| "grad_norm": 1.894015594982712, | |
| "learning_rate": 1.3958129255446585e-06, | |
| "loss": 0.383, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.7856761090326029, | |
| "grad_norm": 1.8830744382706812, | |
| "learning_rate": 1.3317173108371834e-06, | |
| "loss": 0.3803, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.7910208444681989, | |
| "grad_norm": 1.723519908836632, | |
| "learning_rate": 1.268901365966337e-06, | |
| "loss": 0.3822, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.7963655799037948, | |
| "grad_norm": 1.7952600980364406, | |
| "learning_rate": 1.2073870040933212e-06, | |
| "loss": 0.3715, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.8017103153393907, | |
| "grad_norm": 1.9566873596603427, | |
| "learning_rate": 1.1471956843258676e-06, | |
| "loss": 0.384, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.8070550507749866, | |
| "grad_norm": 1.8386259618144518, | |
| "learning_rate": 1.0883484042322796e-06, | |
| "loss": 0.3869, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.8123997862105826, | |
| "grad_norm": 1.8260297894288255, | |
| "learning_rate": 1.0308656925165033e-06, | |
| "loss": 0.3812, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.8177445216461785, | |
| "grad_norm": 1.8047190650634146, | |
| "learning_rate": 9.74767601856737e-07, | |
| "loss": 0.3806, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.8230892570817745, | |
| "grad_norm": 1.8221471376738447, | |
| "learning_rate": 9.200737019101169e-07, | |
| "loss": 0.3777, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.8284339925173704, | |
| "grad_norm": 1.963769411688169, | |
| "learning_rate": 8.668030724858984e-07, | |
| "loss": 0.3688, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.8337787279529664, | |
| "grad_norm": 1.7734438619458048, | |
| "learning_rate": 8.149742968895253e-07, | |
| "loss": 0.3756, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.8391234633885623, | |
| "grad_norm": 1.800103499706633, | |
| "learning_rate": 7.646054554398863e-07, | |
| "loss": 0.3773, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.8444681988241582, | |
| "grad_norm": 1.8625669508711113, | |
| "learning_rate": 7.157141191620548e-07, | |
| "loss": 0.3682, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.8498129342597541, | |
| "grad_norm": 1.8353500996941807, | |
| "learning_rate": 6.683173436576851e-07, | |
| "loss": 0.3646, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.8551576696953501, | |
| "grad_norm": 1.8832589125337102, | |
| "learning_rate": 6.224316631552207e-07, | |
| "loss": 0.376, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.860502405130946, | |
| "grad_norm": 1.7768361367435437, | |
| "learning_rate": 5.780730847419652e-07, | |
| "loss": 0.3657, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.865847140566542, | |
| "grad_norm": 1.7627566159259074, | |
| "learning_rate": 5.35257082780069e-07, | |
| "loss": 0.3674, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.8711918760021379, | |
| "grad_norm": 1.8420469952326894, | |
| "learning_rate": 4.939985935083314e-07, | |
| "loss": 0.3669, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.8765366114377339, | |
| "grad_norm": 1.7499771334622984, | |
| "learning_rate": 4.5431200983174493e-07, | |
| "loss": 0.3671, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.8818813468733298, | |
| "grad_norm": 1.7862835118779703, | |
| "learning_rate": 4.1621117630056606e-07, | |
| "loss": 0.3729, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.8872260823089257, | |
| "grad_norm": 1.7739216023779494, | |
| "learning_rate": 3.7970938428068813e-07, | |
| "loss": 0.3714, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8925708177445216, | |
| "grad_norm": 1.7066183325006379, | |
| "learning_rate": 3.4481936731698415e-07, | |
| "loss": 0.3655, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.8979155531801176, | |
| "grad_norm": 1.7916893607261628, | |
| "learning_rate": 3.1155329669124876e-07, | |
| "loss": 0.3702, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.9032602886157135, | |
| "grad_norm": 1.803470706275281, | |
| "learning_rate": 2.7992277717627856e-07, | |
| "loss": 0.3599, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.9086050240513095, | |
| "grad_norm": 1.7999593269005436, | |
| "learning_rate": 2.4993884298758097e-07, | |
| "loss": 0.3621, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9139497594869054, | |
| "grad_norm": 1.749439500356134, | |
| "learning_rate": 2.2161195393412493e-07, | |
| "loss": 0.361, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.9192944949225014, | |
| "grad_norm": 1.804710229821361, | |
| "learning_rate": 1.9495199176945977e-07, | |
| "loss": 0.3705, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.9246392303580973, | |
| "grad_norm": 1.7380753832775804, | |
| "learning_rate": 1.6996825674449768e-07, | |
| "loss": 0.3581, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.9299839657936932, | |
| "grad_norm": 1.9009106101736641, | |
| "learning_rate": 1.4666946436314832e-07, | |
| "loss": 0.3699, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.9353287012292891, | |
| "grad_norm": 1.7420463319700745, | |
| "learning_rate": 1.2506374234193985e-07, | |
| "loss": 0.36, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.9406734366648851, | |
| "grad_norm": 1.7367484197909535, | |
| "learning_rate": 1.0515862777468689e-07, | |
| "loss": 0.3611, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.946018172100481, | |
| "grad_norm": 1.8449072611969448, | |
| "learning_rate": 8.69610645031993e-08, | |
| "loss": 0.3585, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.951362907536077, | |
| "grad_norm": 1.7050020816812541, | |
| "learning_rate": 7.047740069494102e-08, | |
| "loss": 0.3663, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.9567076429716729, | |
| "grad_norm": 1.7491629746492106, | |
| "learning_rate": 5.571338662849257e-08, | |
| "loss": 0.3621, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.9620523784072689, | |
| "grad_norm": 1.8299154207593904, | |
| "learning_rate": 4.267417268758123e-08, | |
| "loss": 0.3657, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9673971138428648, | |
| "grad_norm": 1.7386431635341528, | |
| "learning_rate": 3.1364307564384357e-08, | |
| "loss": 0.3604, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.9727418492784607, | |
| "grad_norm": 1.851821490817842, | |
| "learning_rate": 2.178773667273204e-08, | |
| "loss": 0.3639, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.9780865847140566, | |
| "grad_norm": 1.7134927436590126, | |
| "learning_rate": 1.3947800771760278e-08, | |
| "loss": 0.3623, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.9834313201496526, | |
| "grad_norm": 1.7740180149347793, | |
| "learning_rate": 7.84723480049765e-09, | |
| "loss": 0.3727, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.9887760555852485, | |
| "grad_norm": 1.6783259441701572, | |
| "learning_rate": 3.4881669237890603e-09, | |
| "loss": 0.36, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.9941207910208445, | |
| "grad_norm": 1.6994301851701823, | |
| "learning_rate": 8.721177898912691e-10, | |
| "loss": 0.3739, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9994655264564404, | |
| "grad_norm": 1.7638563088423833, | |
| "learning_rate": 0.0, | |
| "loss": 0.3534, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.9994655264564404, | |
| "eval_loss": 0.3047424554824829, | |
| "eval_runtime": 0.9585, | |
| "eval_samples_per_second": 2.087, | |
| "eval_steps_per_second": 1.043, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 0.9994655264564404, | |
| "step": 935, | |
| "total_flos": 195717633146880.0, | |
| "train_loss": 0.5211284054791864, | |
| "train_runtime": 23277.081, | |
| "train_samples_per_second": 1.286, | |
| "train_steps_per_second": 0.04 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 935, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 195717633146880.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |