| { |
| "best_global_step": 600, |
| "best_metric": 7.83868408203125, |
| "best_model_checkpoint": "out/checkpoint-600", |
| "epoch": 10.52980132450331, |
| "eval_steps": 150, |
| "global_step": 600, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.08830022075055188, |
| "grad_norm": 8.688450813293457, |
| "learning_rate": 9.976608187134502e-05, |
| "loss": 13.340193176269532, |
| "step": 5 |
| }, |
| { |
| "epoch": 0.17660044150110377, |
| "grad_norm": 4.834760665893555, |
| "learning_rate": 9.947368421052632e-05, |
| "loss": 10.211357879638673, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.26490066225165565, |
| "grad_norm": 1.3143389225006104, |
| "learning_rate": 9.918128654970761e-05, |
| "loss": 8.949986267089844, |
| "step": 15 |
| }, |
| { |
| "epoch": 0.35320088300220753, |
| "grad_norm": 0.37686967849731445, |
| "learning_rate": 9.888888888888889e-05, |
| "loss": 8.484906005859376, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.44150110375275936, |
| "grad_norm": 0.2962671220302582, |
| "learning_rate": 9.859649122807017e-05, |
| "loss": 8.364928436279296, |
| "step": 25 |
| }, |
| { |
| "epoch": 0.5298013245033113, |
| "grad_norm": 0.2862487733364105, |
| "learning_rate": 9.830409356725147e-05, |
| "loss": 8.333544921875, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.6181015452538632, |
| "grad_norm": 0.20227763056755066, |
| "learning_rate": 9.801169590643276e-05, |
| "loss": 8.327781677246094, |
| "step": 35 |
| }, |
| { |
| "epoch": 0.7064017660044151, |
| "grad_norm": 0.39236980676651, |
| "learning_rate": 9.771929824561404e-05, |
| "loss": 8.322337341308593, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.7947019867549668, |
| "grad_norm": 0.21722739934921265, |
| "learning_rate": 9.742690058479532e-05, |
| "loss": 8.316014099121094, |
| "step": 45 |
| }, |
| { |
| "epoch": 0.8830022075055187, |
| "grad_norm": 0.2812463641166687, |
| "learning_rate": 9.713450292397662e-05, |
| "loss": 8.31099853515625, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.9713024282560706, |
| "grad_norm": 0.2608415186405182, |
| "learning_rate": 9.68421052631579e-05, |
| "loss": 8.309403991699218, |
| "step": 55 |
| }, |
| { |
| "epoch": 1.0529801324503312, |
| "grad_norm": 0.25310298800468445, |
| "learning_rate": 9.654970760233919e-05, |
| "loss": 8.305143737792969, |
| "step": 60 |
| }, |
| { |
| "epoch": 1.141280353200883, |
| "grad_norm": 0.20996594429016113, |
| "learning_rate": 9.625730994152047e-05, |
| "loss": 8.303257751464844, |
| "step": 65 |
| }, |
| { |
| "epoch": 1.229580573951435, |
| "grad_norm": 0.22537994384765625, |
| "learning_rate": 9.596491228070175e-05, |
| "loss": 8.29981689453125, |
| "step": 70 |
| }, |
| { |
| "epoch": 1.3178807947019868, |
| "grad_norm": 0.17918814718723297, |
| "learning_rate": 9.567251461988305e-05, |
| "loss": 8.295522308349609, |
| "step": 75 |
| }, |
| { |
| "epoch": 1.4061810154525387, |
| "grad_norm": 0.207343190908432, |
| "learning_rate": 9.538011695906434e-05, |
| "loss": 8.290687561035156, |
| "step": 80 |
| }, |
| { |
| "epoch": 1.4944812362030906, |
| "grad_norm": 0.32762715220451355, |
| "learning_rate": 9.508771929824562e-05, |
| "loss": 8.281892395019531, |
| "step": 85 |
| }, |
| { |
| "epoch": 1.5827814569536423, |
| "grad_norm": 0.2938079535961151, |
| "learning_rate": 9.47953216374269e-05, |
| "loss": 8.285610961914063, |
| "step": 90 |
| }, |
| { |
| "epoch": 1.6710816777041941, |
| "grad_norm": 0.2513900399208069, |
| "learning_rate": 9.45029239766082e-05, |
| "loss": 8.284735107421875, |
| "step": 95 |
| }, |
| { |
| "epoch": 1.759381898454746, |
| "grad_norm": 0.24608662724494934, |
| "learning_rate": 9.421052631578949e-05, |
| "loss": 8.28292236328125, |
| "step": 100 |
| }, |
| { |
| "epoch": 1.847682119205298, |
| "grad_norm": 0.3305456340312958, |
| "learning_rate": 9.391812865497076e-05, |
| "loss": 8.274761962890626, |
| "step": 105 |
| }, |
| { |
| "epoch": 1.9359823399558498, |
| "grad_norm": 0.2837885022163391, |
| "learning_rate": 9.362573099415205e-05, |
| "loss": 8.268502807617187, |
| "step": 110 |
| }, |
| { |
| "epoch": 2.0176600441501105, |
| "grad_norm": 0.37587037682533264, |
| "learning_rate": 9.333333333333334e-05, |
| "loss": 8.26836929321289, |
| "step": 115 |
| }, |
| { |
| "epoch": 2.1059602649006623, |
| "grad_norm": 0.42066144943237305, |
| "learning_rate": 9.304093567251462e-05, |
| "loss": 8.240390014648437, |
| "step": 120 |
| }, |
| { |
| "epoch": 2.1942604856512142, |
| "grad_norm": 0.4040590822696686, |
| "learning_rate": 9.27485380116959e-05, |
| "loss": 8.256266021728516, |
| "step": 125 |
| }, |
| { |
| "epoch": 2.282560706401766, |
| "grad_norm": 0.3695836365222931, |
| "learning_rate": 9.24561403508772e-05, |
| "loss": 8.237049102783203, |
| "step": 130 |
| }, |
| { |
| "epoch": 2.370860927152318, |
| "grad_norm": 0.4006362557411194, |
| "learning_rate": 9.216374269005849e-05, |
| "loss": 8.241561889648438, |
| "step": 135 |
| }, |
| { |
| "epoch": 2.45916114790287, |
| "grad_norm": 0.38214218616485596, |
| "learning_rate": 9.187134502923977e-05, |
| "loss": 8.228585815429687, |
| "step": 140 |
| }, |
| { |
| "epoch": 2.5474613686534218, |
| "grad_norm": 0.40786632895469666, |
| "learning_rate": 9.157894736842105e-05, |
| "loss": 8.215204620361328, |
| "step": 145 |
| }, |
| { |
| "epoch": 2.6357615894039736, |
| "grad_norm": 0.349801242351532, |
| "learning_rate": 9.128654970760235e-05, |
| "loss": 8.210234069824219, |
| "step": 150 |
| }, |
| { |
| "epoch": 2.6357615894039736, |
| "eval_loss": 8.208977699279785, |
| "eval_runtime": 6.4293, |
| "eval_samples_per_second": 15.554, |
| "eval_steps_per_second": 7.777, |
| "step": 150 |
| }, |
| { |
| "epoch": 2.7240618101545255, |
| "grad_norm": 0.37537169456481934, |
| "learning_rate": 9.099415204678363e-05, |
| "loss": 8.202523803710937, |
| "step": 155 |
| }, |
| { |
| "epoch": 2.8123620309050774, |
| "grad_norm": 0.3929419219493866, |
| "learning_rate": 9.070175438596492e-05, |
| "loss": 8.18493423461914, |
| "step": 160 |
| }, |
| { |
| "epoch": 2.9006622516556293, |
| "grad_norm": 0.3615664541721344, |
| "learning_rate": 9.04093567251462e-05, |
| "loss": 8.17913818359375, |
| "step": 165 |
| }, |
| { |
| "epoch": 2.988962472406181, |
| "grad_norm": 0.28156760334968567, |
| "learning_rate": 9.011695906432748e-05, |
| "loss": 8.187295532226562, |
| "step": 170 |
| }, |
| { |
| "epoch": 3.0706401766004414, |
| "grad_norm": 0.4112038016319275, |
| "learning_rate": 8.982456140350878e-05, |
| "loss": 8.141925048828124, |
| "step": 175 |
| }, |
| { |
| "epoch": 3.1589403973509933, |
| "grad_norm": 0.46913009881973267, |
| "learning_rate": 8.953216374269007e-05, |
| "loss": 8.151603698730469, |
| "step": 180 |
| }, |
| { |
| "epoch": 3.247240618101545, |
| "grad_norm": 0.7114327549934387, |
| "learning_rate": 8.923976608187135e-05, |
| "loss": 8.168936157226563, |
| "step": 185 |
| }, |
| { |
| "epoch": 3.335540838852097, |
| "grad_norm": 0.6770473122596741, |
| "learning_rate": 8.894736842105263e-05, |
| "loss": 8.124656677246094, |
| "step": 190 |
| }, |
| { |
| "epoch": 3.423841059602649, |
| "grad_norm": 0.4436897933483124, |
| "learning_rate": 8.865497076023393e-05, |
| "loss": 8.115914154052735, |
| "step": 195 |
| }, |
| { |
| "epoch": 3.512141280353201, |
| "grad_norm": 0.5368486642837524, |
| "learning_rate": 8.836257309941522e-05, |
| "loss": 8.126858520507813, |
| "step": 200 |
| }, |
| { |
| "epoch": 3.6004415011037527, |
| "grad_norm": 0.4201599955558777, |
| "learning_rate": 8.807017543859649e-05, |
| "loss": 8.098291015625, |
| "step": 205 |
| }, |
| { |
| "epoch": 3.6887417218543046, |
| "grad_norm": 0.478750616312027, |
| "learning_rate": 8.777777777777778e-05, |
| "loss": 8.094570922851563, |
| "step": 210 |
| }, |
| { |
| "epoch": 3.7770419426048565, |
| "grad_norm": 0.3878437280654907, |
| "learning_rate": 8.748538011695907e-05, |
| "loss": 8.102629852294921, |
| "step": 215 |
| }, |
| { |
| "epoch": 3.8653421633554084, |
| "grad_norm": 0.64486163854599, |
| "learning_rate": 8.719298245614036e-05, |
| "loss": 8.064454650878906, |
| "step": 220 |
| }, |
| { |
| "epoch": 3.9536423841059603, |
| "grad_norm": 0.6539655327796936, |
| "learning_rate": 8.690058479532164e-05, |
| "loss": 8.068508911132813, |
| "step": 225 |
| }, |
| { |
| "epoch": 4.035320088300221, |
| "grad_norm": 0.5649951100349426, |
| "learning_rate": 8.660818713450293e-05, |
| "loss": 8.070999908447266, |
| "step": 230 |
| }, |
| { |
| "epoch": 4.123620309050772, |
| "grad_norm": 0.506738543510437, |
| "learning_rate": 8.631578947368421e-05, |
| "loss": 8.052703857421875, |
| "step": 235 |
| }, |
| { |
| "epoch": 4.211920529801325, |
| "grad_norm": 0.6467848420143127, |
| "learning_rate": 8.60233918128655e-05, |
| "loss": 8.04999237060547, |
| "step": 240 |
| }, |
| { |
| "epoch": 4.300220750551876, |
| "grad_norm": 0.6093457937240601, |
| "learning_rate": 8.573099415204678e-05, |
| "loss": 8.068696594238281, |
| "step": 245 |
| }, |
| { |
| "epoch": 4.3885209713024285, |
| "grad_norm": 0.6376983523368835, |
| "learning_rate": 8.543859649122808e-05, |
| "loss": 8.007734680175782, |
| "step": 250 |
| }, |
| { |
| "epoch": 4.47682119205298, |
| "grad_norm": 0.549198567867279, |
| "learning_rate": 8.514619883040936e-05, |
| "loss": 8.022071838378906, |
| "step": 255 |
| }, |
| { |
| "epoch": 4.565121412803532, |
| "grad_norm": 0.6188969612121582, |
| "learning_rate": 8.485380116959065e-05, |
| "loss": 8.03642578125, |
| "step": 260 |
| }, |
| { |
| "epoch": 4.653421633554084, |
| "grad_norm": 0.8166013360023499, |
| "learning_rate": 8.456140350877193e-05, |
| "loss": 7.964942169189453, |
| "step": 265 |
| }, |
| { |
| "epoch": 4.741721854304636, |
| "grad_norm": 0.6450521349906921, |
| "learning_rate": 8.426900584795321e-05, |
| "loss": 8.015605163574218, |
| "step": 270 |
| }, |
| { |
| "epoch": 4.830022075055187, |
| "grad_norm": 0.5845851302146912, |
| "learning_rate": 8.397660818713451e-05, |
| "loss": 8.016954803466797, |
| "step": 275 |
| }, |
| { |
| "epoch": 4.91832229580574, |
| "grad_norm": 0.6162020564079285, |
| "learning_rate": 8.36842105263158e-05, |
| "loss": 7.971800231933594, |
| "step": 280 |
| }, |
| { |
| "epoch": 5.0, |
| "grad_norm": 0.6822549104690552, |
| "learning_rate": 8.339181286549708e-05, |
| "loss": 7.994195556640625, |
| "step": 285 |
| }, |
| { |
| "epoch": 5.088300220750552, |
| "grad_norm": 0.8076459169387817, |
| "learning_rate": 8.309941520467836e-05, |
| "loss": 7.982322692871094, |
| "step": 290 |
| }, |
| { |
| "epoch": 5.176600441501104, |
| "grad_norm": 0.7110167741775513, |
| "learning_rate": 8.280701754385966e-05, |
| "loss": 7.952334594726563, |
| "step": 295 |
| }, |
| { |
| "epoch": 5.264900662251655, |
| "grad_norm": 0.7477928996086121, |
| "learning_rate": 8.251461988304095e-05, |
| "loss": 7.962551879882812, |
| "step": 300 |
| }, |
| { |
| "epoch": 5.264900662251655, |
| "eval_loss": 7.9789137840271, |
| "eval_runtime": 6.4308, |
| "eval_samples_per_second": 15.55, |
| "eval_steps_per_second": 7.775, |
| "step": 300 |
| }, |
| { |
| "epoch": 5.3532008830022075, |
| "grad_norm": 0.6691218614578247, |
| "learning_rate": 8.222222222222222e-05, |
| "loss": 7.981855773925782, |
| "step": 305 |
| }, |
| { |
| "epoch": 5.441501103752759, |
| "grad_norm": 0.590973973274231, |
| "learning_rate": 8.192982456140351e-05, |
| "loss": 7.941262817382812, |
| "step": 310 |
| }, |
| { |
| "epoch": 5.529801324503311, |
| "grad_norm": 0.6093570590019226, |
| "learning_rate": 8.16374269005848e-05, |
| "loss": 7.952169036865234, |
| "step": 315 |
| }, |
| { |
| "epoch": 5.618101545253863, |
| "grad_norm": 0.44569653272628784, |
| "learning_rate": 8.134502923976609e-05, |
| "loss": 7.9652587890625, |
| "step": 320 |
| }, |
| { |
| "epoch": 5.706401766004415, |
| "grad_norm": 0.7363209128379822, |
| "learning_rate": 8.105263157894737e-05, |
| "loss": 7.954380798339844, |
| "step": 325 |
| }, |
| { |
| "epoch": 5.7947019867549665, |
| "grad_norm": 0.8245615363121033, |
| "learning_rate": 8.076023391812866e-05, |
| "loss": 7.940582275390625, |
| "step": 330 |
| }, |
| { |
| "epoch": 5.883002207505519, |
| "grad_norm": 0.9126182198524475, |
| "learning_rate": 8.046783625730994e-05, |
| "loss": 7.918997192382813, |
| "step": 335 |
| }, |
| { |
| "epoch": 5.97130242825607, |
| "grad_norm": 0.7115766406059265, |
| "learning_rate": 8.017543859649124e-05, |
| "loss": 7.887815856933594, |
| "step": 340 |
| }, |
| { |
| "epoch": 6.052980132450331, |
| "grad_norm": 0.5615514516830444, |
| "learning_rate": 7.988304093567252e-05, |
| "loss": 7.894483184814453, |
| "step": 345 |
| }, |
| { |
| "epoch": 6.141280353200883, |
| "grad_norm": 0.764122486114502, |
| "learning_rate": 7.959064327485381e-05, |
| "loss": 7.8913932800292965, |
| "step": 350 |
| }, |
| { |
| "epoch": 6.229580573951435, |
| "grad_norm": 0.9715237617492676, |
| "learning_rate": 7.929824561403509e-05, |
| "loss": 7.890602111816406, |
| "step": 355 |
| }, |
| { |
| "epoch": 6.317880794701987, |
| "grad_norm": 0.8448056578636169, |
| "learning_rate": 7.900584795321638e-05, |
| "loss": 7.986685180664063, |
| "step": 360 |
| }, |
| { |
| "epoch": 6.406181015452539, |
| "grad_norm": 0.9147307872772217, |
| "learning_rate": 7.871345029239767e-05, |
| "loss": 7.92308349609375, |
| "step": 365 |
| }, |
| { |
| "epoch": 6.49448123620309, |
| "grad_norm": 1.0160162448883057, |
| "learning_rate": 7.842105263157895e-05, |
| "loss": 7.9209846496582035, |
| "step": 370 |
| }, |
| { |
| "epoch": 6.582781456953643, |
| "grad_norm": 0.6162037253379822, |
| "learning_rate": 7.812865497076024e-05, |
| "loss": 7.8999687194824215, |
| "step": 375 |
| }, |
| { |
| "epoch": 6.671081677704194, |
| "grad_norm": 0.7081504464149475, |
| "learning_rate": 7.783625730994153e-05, |
| "loss": 7.872428894042969, |
| "step": 380 |
| }, |
| { |
| "epoch": 6.7593818984547465, |
| "grad_norm": 0.6675496697425842, |
| "learning_rate": 7.754385964912281e-05, |
| "loss": 7.877265930175781, |
| "step": 385 |
| }, |
| { |
| "epoch": 6.847682119205298, |
| "grad_norm": 0.8263344764709473, |
| "learning_rate": 7.72514619883041e-05, |
| "loss": 7.908570861816406, |
| "step": 390 |
| }, |
| { |
| "epoch": 6.93598233995585, |
| "grad_norm": 0.7107247710227966, |
| "learning_rate": 7.695906432748539e-05, |
| "loss": 7.816087341308593, |
| "step": 395 |
| }, |
| { |
| "epoch": 7.0176600441501105, |
| "grad_norm": 0.8045026063919067, |
| "learning_rate": 7.666666666666667e-05, |
| "loss": 7.903431701660156, |
| "step": 400 |
| }, |
| { |
| "epoch": 7.105960264900662, |
| "grad_norm": 0.860241174697876, |
| "learning_rate": 7.637426900584795e-05, |
| "loss": 7.839199829101562, |
| "step": 405 |
| }, |
| { |
| "epoch": 7.194260485651214, |
| "grad_norm": 0.9684602618217468, |
| "learning_rate": 7.608187134502924e-05, |
| "loss": 7.877064514160156, |
| "step": 410 |
| }, |
| { |
| "epoch": 7.282560706401766, |
| "grad_norm": 0.904358983039856, |
| "learning_rate": 7.578947368421054e-05, |
| "loss": 7.8798271179199215, |
| "step": 415 |
| }, |
| { |
| "epoch": 7.370860927152318, |
| "grad_norm": 0.8752867579460144, |
| "learning_rate": 7.549707602339182e-05, |
| "loss": 7.820896911621094, |
| "step": 420 |
| }, |
| { |
| "epoch": 7.459161147902869, |
| "grad_norm": 0.8255132436752319, |
| "learning_rate": 7.52046783625731e-05, |
| "loss": 7.889141845703125, |
| "step": 425 |
| }, |
| { |
| "epoch": 7.547461368653422, |
| "grad_norm": 0.7994998097419739, |
| "learning_rate": 7.491228070175439e-05, |
| "loss": 7.799693298339844, |
| "step": 430 |
| }, |
| { |
| "epoch": 7.635761589403973, |
| "grad_norm": 1.000003695487976, |
| "learning_rate": 7.461988304093567e-05, |
| "loss": 7.874913024902344, |
| "step": 435 |
| }, |
| { |
| "epoch": 7.7240618101545255, |
| "grad_norm": 0.6996250152587891, |
| "learning_rate": 7.432748538011697e-05, |
| "loss": 7.898649597167969, |
| "step": 440 |
| }, |
| { |
| "epoch": 7.812362030905077, |
| "grad_norm": 0.7075074911117554, |
| "learning_rate": 7.403508771929825e-05, |
| "loss": 7.874346923828125, |
| "step": 445 |
| }, |
| { |
| "epoch": 7.900662251655629, |
| "grad_norm": 0.855798065662384, |
| "learning_rate": 7.374269005847954e-05, |
| "loss": 7.894575500488282, |
| "step": 450 |
| }, |
| { |
| "epoch": 7.900662251655629, |
| "eval_loss": 7.881262302398682, |
| "eval_runtime": 6.4538, |
| "eval_samples_per_second": 15.495, |
| "eval_steps_per_second": 7.747, |
| "step": 450 |
| }, |
| { |
| "epoch": 7.988962472406181, |
| "grad_norm": 1.0835974216461182, |
| "learning_rate": 7.345029239766082e-05, |
| "loss": 7.863267517089843, |
| "step": 455 |
| }, |
| { |
| "epoch": 8.070640176600442, |
| "grad_norm": 0.7638969421386719, |
| "learning_rate": 7.315789473684212e-05, |
| "loss": 7.783184814453125, |
| "step": 460 |
| }, |
| { |
| "epoch": 8.158940397350994, |
| "grad_norm": 0.9937992095947266, |
| "learning_rate": 7.28654970760234e-05, |
| "loss": 7.839315795898438, |
| "step": 465 |
| }, |
| { |
| "epoch": 8.247240618101545, |
| "grad_norm": 0.8237105011940002, |
| "learning_rate": 7.257309941520468e-05, |
| "loss": 7.841758728027344, |
| "step": 470 |
| }, |
| { |
| "epoch": 8.335540838852097, |
| "grad_norm": 0.7182467579841614, |
| "learning_rate": 7.228070175438597e-05, |
| "loss": 7.85023193359375, |
| "step": 475 |
| }, |
| { |
| "epoch": 8.42384105960265, |
| "grad_norm": 0.8129780888557434, |
| "learning_rate": 7.198830409356726e-05, |
| "loss": 7.865559387207031, |
| "step": 480 |
| }, |
| { |
| "epoch": 8.512141280353202, |
| "grad_norm": 0.6443418264389038, |
| "learning_rate": 7.169590643274853e-05, |
| "loss": 7.831302642822266, |
| "step": 485 |
| }, |
| { |
| "epoch": 8.600441501103752, |
| "grad_norm": 0.7900895476341248, |
| "learning_rate": 7.140350877192983e-05, |
| "loss": 7.88115234375, |
| "step": 490 |
| }, |
| { |
| "epoch": 8.688741721854305, |
| "grad_norm": 0.6896668076515198, |
| "learning_rate": 7.111111111111112e-05, |
| "loss": 7.8469383239746096, |
| "step": 495 |
| }, |
| { |
| "epoch": 8.777041942604857, |
| "grad_norm": 0.8964319229125977, |
| "learning_rate": 7.08187134502924e-05, |
| "loss": 7.783286285400391, |
| "step": 500 |
| }, |
| { |
| "epoch": 8.865342163355407, |
| "grad_norm": 1.1212209463119507, |
| "learning_rate": 7.052631578947368e-05, |
| "loss": 7.807972717285156, |
| "step": 505 |
| }, |
| { |
| "epoch": 8.95364238410596, |
| "grad_norm": 1.048243522644043, |
| "learning_rate": 7.023391812865497e-05, |
| "loss": 7.847767639160156, |
| "step": 510 |
| }, |
| { |
| "epoch": 9.035320088300221, |
| "grad_norm": 1.1085704565048218, |
| "learning_rate": 6.994152046783627e-05, |
| "loss": 7.852699279785156, |
| "step": 515 |
| }, |
| { |
| "epoch": 9.123620309050773, |
| "grad_norm": 0.9384048581123352, |
| "learning_rate": 6.964912280701754e-05, |
| "loss": 7.753205871582031, |
| "step": 520 |
| }, |
| { |
| "epoch": 9.211920529801324, |
| "grad_norm": 0.9842119216918945, |
| "learning_rate": 6.935672514619883e-05, |
| "loss": 7.836062622070313, |
| "step": 525 |
| }, |
| { |
| "epoch": 9.300220750551876, |
| "grad_norm": 1.29048490524292, |
| "learning_rate": 6.906432748538012e-05, |
| "loss": 7.795281982421875, |
| "step": 530 |
| }, |
| { |
| "epoch": 9.388520971302428, |
| "grad_norm": 0.804202139377594, |
| "learning_rate": 6.87719298245614e-05, |
| "loss": 7.802140808105468, |
| "step": 535 |
| }, |
| { |
| "epoch": 9.47682119205298, |
| "grad_norm": 0.8415213823318481, |
| "learning_rate": 6.847953216374268e-05, |
| "loss": 7.84085693359375, |
| "step": 540 |
| }, |
| { |
| "epoch": 9.565121412803531, |
| "grad_norm": 0.5611467957496643, |
| "learning_rate": 6.818713450292398e-05, |
| "loss": 7.810952758789062, |
| "step": 545 |
| }, |
| { |
| "epoch": 9.653421633554084, |
| "grad_norm": 0.730029284954071, |
| "learning_rate": 6.789473684210527e-05, |
| "loss": 7.858808898925782, |
| "step": 550 |
| }, |
| { |
| "epoch": 9.741721854304636, |
| "grad_norm": 0.7788743376731873, |
| "learning_rate": 6.760233918128655e-05, |
| "loss": 7.807586669921875, |
| "step": 555 |
| }, |
| { |
| "epoch": 9.830022075055188, |
| "grad_norm": 1.4132566452026367, |
| "learning_rate": 6.730994152046783e-05, |
| "loss": 7.82841796875, |
| "step": 560 |
| }, |
| { |
| "epoch": 9.918322295805739, |
| "grad_norm": 1.0049456357955933, |
| "learning_rate": 6.701754385964913e-05, |
| "loss": 7.835298156738281, |
| "step": 565 |
| }, |
| { |
| "epoch": 10.0, |
| "grad_norm": 0.8505070209503174, |
| "learning_rate": 6.672514619883041e-05, |
| "loss": 7.798960876464844, |
| "step": 570 |
| }, |
| { |
| "epoch": 10.088300220750552, |
| "grad_norm": 0.7744261622428894, |
| "learning_rate": 6.64327485380117e-05, |
| "loss": 7.813462829589843, |
| "step": 575 |
| }, |
| { |
| "epoch": 10.176600441501105, |
| "grad_norm": 0.9087972640991211, |
| "learning_rate": 6.614035087719298e-05, |
| "loss": 7.7771759033203125, |
| "step": 580 |
| }, |
| { |
| "epoch": 10.264900662251655, |
| "grad_norm": 0.759032666683197, |
| "learning_rate": 6.584795321637426e-05, |
| "loss": 7.78880615234375, |
| "step": 585 |
| }, |
| { |
| "epoch": 10.353200883002208, |
| "grad_norm": 0.7629870772361755, |
| "learning_rate": 6.555555555555556e-05, |
| "loss": 7.788232421875, |
| "step": 590 |
| }, |
| { |
| "epoch": 10.44150110375276, |
| "grad_norm": 0.8606480956077576, |
| "learning_rate": 6.526315789473685e-05, |
| "loss": 7.791365051269532, |
| "step": 595 |
| }, |
| { |
| "epoch": 10.52980132450331, |
| "grad_norm": 0.9902874827384949, |
| "learning_rate": 6.497076023391813e-05, |
| "loss": 7.858397674560547, |
| "step": 600 |
| }, |
| { |
| "epoch": 10.52980132450331, |
| "eval_loss": 7.83868408203125, |
| "eval_runtime": 6.4428, |
| "eval_samples_per_second": 15.521, |
| "eval_steps_per_second": 7.761, |
| "step": 600 |
| } |
| ], |
| "logging_steps": 5, |
| "max_steps": 1710, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 30, |
| "save_steps": 150, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 3.191666111204557e+17, |
| "train_batch_size": 1, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
