| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.1577182486758373, | |
| "eval_steps": 500, | |
| "global_step": 226500, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.004763174941889266, | |
| "grad_norm": 23.845611572265625, | |
| "learning_rate": 4.976184125290554e-05, | |
| "loss": 1.6947, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.009526349883778531, | |
| "grad_norm": 11.257529258728027, | |
| "learning_rate": 4.952368250581108e-05, | |
| "loss": 1.2773, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.014289524825667798, | |
| "grad_norm": 19.1054630279541, | |
| "learning_rate": 4.9285523758716614e-05, | |
| "loss": 1.1464, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.019052699767557062, | |
| "grad_norm": 8.174609184265137, | |
| "learning_rate": 4.904736501162215e-05, | |
| "loss": 1.0952, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.02381587470944633, | |
| "grad_norm": 22.265609741210938, | |
| "learning_rate": 4.880920626452769e-05, | |
| "loss": 1.0419, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.028579049651335595, | |
| "grad_norm": 3.621635913848877, | |
| "learning_rate": 4.857104751743322e-05, | |
| "loss": 0.9673, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.03334222459322486, | |
| "grad_norm": 9.05103874206543, | |
| "learning_rate": 4.8332888770338755e-05, | |
| "loss": 0.9227, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.038105399535114125, | |
| "grad_norm": 6.0216474533081055, | |
| "learning_rate": 4.809473002324429e-05, | |
| "loss": 0.917, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.042868574477003395, | |
| "grad_norm": 12.011100769042969, | |
| "learning_rate": 4.785657127614983e-05, | |
| "loss": 0.8644, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.04763174941889266, | |
| "grad_norm": 10.062934875488281, | |
| "learning_rate": 4.761841252905537e-05, | |
| "loss": 0.8573, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.05239492436078192, | |
| "grad_norm": 4.252622127532959, | |
| "learning_rate": 4.7380253781960904e-05, | |
| "loss": 0.8198, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 0.05715809930267119, | |
| "grad_norm": 11.673833847045898, | |
| "learning_rate": 4.714209503486644e-05, | |
| "loss": 0.7742, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 0.061921274244560454, | |
| "grad_norm": 8.625577926635742, | |
| "learning_rate": 4.690393628777198e-05, | |
| "loss": 0.7513, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 0.06668444918644972, | |
| "grad_norm": 11.9224853515625, | |
| "learning_rate": 4.666577754067752e-05, | |
| "loss": 0.7376, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 0.07144762412833898, | |
| "grad_norm": 3.3534774780273438, | |
| "learning_rate": 4.642761879358305e-05, | |
| "loss": 0.733, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 0.07621079907022825, | |
| "grad_norm": 6.773929595947266, | |
| "learning_rate": 4.618946004648859e-05, | |
| "loss": 0.6743, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 0.08097397401211752, | |
| "grad_norm": 12.177743911743164, | |
| "learning_rate": 4.595130129939413e-05, | |
| "loss": 0.6837, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 0.08573714895400679, | |
| "grad_norm": 10.937847137451172, | |
| "learning_rate": 4.5713142552299664e-05, | |
| "loss": 0.6745, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 0.09050032389589605, | |
| "grad_norm": 9.08661937713623, | |
| "learning_rate": 4.54749838052052e-05, | |
| "loss": 0.6366, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 0.09526349883778532, | |
| "grad_norm": 14.092183113098145, | |
| "learning_rate": 4.523682505811074e-05, | |
| "loss": 0.6218, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.10002667377967459, | |
| "grad_norm": 9.430176734924316, | |
| "learning_rate": 4.4998666311016275e-05, | |
| "loss": 0.6286, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 0.10478984872156384, | |
| "grad_norm": 6.812849998474121, | |
| "learning_rate": 4.476050756392181e-05, | |
| "loss": 0.6121, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 0.10955302366345311, | |
| "grad_norm": 6.120726108551025, | |
| "learning_rate": 4.452234881682735e-05, | |
| "loss": 0.581, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 0.11431619860534238, | |
| "grad_norm": 13.171046257019043, | |
| "learning_rate": 4.4284190069732887e-05, | |
| "loss": 0.5797, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 0.11907937354723164, | |
| "grad_norm": 9.89405632019043, | |
| "learning_rate": 4.404603132263842e-05, | |
| "loss": 0.5461, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 0.12384254848912091, | |
| "grad_norm": 6.620090484619141, | |
| "learning_rate": 4.3807872575543954e-05, | |
| "loss": 0.5583, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 0.12860572343101018, | |
| "grad_norm": 17.443872451782227, | |
| "learning_rate": 4.356971382844949e-05, | |
| "loss": 0.562, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 0.13336889837289945, | |
| "grad_norm": 4.9741997718811035, | |
| "learning_rate": 4.333155508135503e-05, | |
| "loss": 0.5448, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 0.13813207331478872, | |
| "grad_norm": 13.39028549194336, | |
| "learning_rate": 4.3093396334260565e-05, | |
| "loss": 0.5166, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 0.14289524825667796, | |
| "grad_norm": 7.8550333976745605, | |
| "learning_rate": 4.28552375871661e-05, | |
| "loss": 0.5071, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.14765842319856723, | |
| "grad_norm": 13.109572410583496, | |
| "learning_rate": 4.261707884007164e-05, | |
| "loss": 0.489, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 0.1524215981404565, | |
| "grad_norm": 8.581804275512695, | |
| "learning_rate": 4.237892009297718e-05, | |
| "loss": 0.4992, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 0.15718477308234577, | |
| "grad_norm": 0.8759533166885376, | |
| "learning_rate": 4.2140761345882714e-05, | |
| "loss": 0.5041, | |
| "step": 16500 | |
| }, | |
| { | |
| "epoch": 0.16194794802423504, | |
| "grad_norm": 4.165910243988037, | |
| "learning_rate": 4.190260259878825e-05, | |
| "loss": 0.4923, | |
| "step": 17000 | |
| }, | |
| { | |
| "epoch": 0.1667111229661243, | |
| "grad_norm": 4.400519847869873, | |
| "learning_rate": 4.166444385169379e-05, | |
| "loss": 0.4895, | |
| "step": 17500 | |
| }, | |
| { | |
| "epoch": 0.17147429790801358, | |
| "grad_norm": 3.9758849143981934, | |
| "learning_rate": 4.142628510459932e-05, | |
| "loss": 0.4845, | |
| "step": 18000 | |
| }, | |
| { | |
| "epoch": 0.17623747284990282, | |
| "grad_norm": 11.77698802947998, | |
| "learning_rate": 4.1188126357504856e-05, | |
| "loss": 0.4523, | |
| "step": 18500 | |
| }, | |
| { | |
| "epoch": 0.1810006477917921, | |
| "grad_norm": 8.056093215942383, | |
| "learning_rate": 4.094996761041039e-05, | |
| "loss": 0.4815, | |
| "step": 19000 | |
| }, | |
| { | |
| "epoch": 0.18576382273368136, | |
| "grad_norm": 3.4172158241271973, | |
| "learning_rate": 4.071180886331593e-05, | |
| "loss": 0.478, | |
| "step": 19500 | |
| }, | |
| { | |
| "epoch": 0.19052699767557063, | |
| "grad_norm": 4.890318393707275, | |
| "learning_rate": 4.0473650116221474e-05, | |
| "loss": 0.4492, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 0.1952901726174599, | |
| "grad_norm": 9.595198631286621, | |
| "learning_rate": 4.023549136912701e-05, | |
| "loss": 0.4507, | |
| "step": 20500 | |
| }, | |
| { | |
| "epoch": 0.20005334755934917, | |
| "grad_norm": 8.482044219970703, | |
| "learning_rate": 3.999733262203255e-05, | |
| "loss": 0.4436, | |
| "step": 21000 | |
| }, | |
| { | |
| "epoch": 0.2048165225012384, | |
| "grad_norm": 5.411190509796143, | |
| "learning_rate": 3.9759173874938085e-05, | |
| "loss": 0.4406, | |
| "step": 21500 | |
| }, | |
| { | |
| "epoch": 0.20957969744312768, | |
| "grad_norm": 13.22948932647705, | |
| "learning_rate": 3.952101512784362e-05, | |
| "loss": 0.4222, | |
| "step": 22000 | |
| }, | |
| { | |
| "epoch": 0.21434287238501695, | |
| "grad_norm": 11.69010066986084, | |
| "learning_rate": 3.928285638074915e-05, | |
| "loss": 0.4119, | |
| "step": 22500 | |
| }, | |
| { | |
| "epoch": 0.21910604732690622, | |
| "grad_norm": 11.298418998718262, | |
| "learning_rate": 3.904469763365469e-05, | |
| "loss": 0.4058, | |
| "step": 23000 | |
| }, | |
| { | |
| "epoch": 0.2238692222687955, | |
| "grad_norm": 8.810882568359375, | |
| "learning_rate": 3.880653888656023e-05, | |
| "loss": 0.4201, | |
| "step": 23500 | |
| }, | |
| { | |
| "epoch": 0.22863239721068476, | |
| "grad_norm": 15.669658660888672, | |
| "learning_rate": 3.8568380139465764e-05, | |
| "loss": 0.4098, | |
| "step": 24000 | |
| }, | |
| { | |
| "epoch": 0.23339557215257403, | |
| "grad_norm": 3.749591112136841, | |
| "learning_rate": 3.83302213923713e-05, | |
| "loss": 0.3972, | |
| "step": 24500 | |
| }, | |
| { | |
| "epoch": 0.23815874709446327, | |
| "grad_norm": 8.824658393859863, | |
| "learning_rate": 3.809206264527684e-05, | |
| "loss": 0.3722, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 0.24292192203635254, | |
| "grad_norm": 13.041702270507812, | |
| "learning_rate": 3.7853903898182375e-05, | |
| "loss": 0.4254, | |
| "step": 25500 | |
| }, | |
| { | |
| "epoch": 0.24768509697824181, | |
| "grad_norm": 15.908291816711426, | |
| "learning_rate": 3.761574515108791e-05, | |
| "loss": 0.3736, | |
| "step": 26000 | |
| }, | |
| { | |
| "epoch": 0.2524482719201311, | |
| "grad_norm": 10.208578109741211, | |
| "learning_rate": 3.737758640399345e-05, | |
| "loss": 0.412, | |
| "step": 26500 | |
| }, | |
| { | |
| "epoch": 0.25721144686202035, | |
| "grad_norm": 5.080774784088135, | |
| "learning_rate": 3.713942765689899e-05, | |
| "loss": 0.3799, | |
| "step": 27000 | |
| }, | |
| { | |
| "epoch": 0.2619746218039096, | |
| "grad_norm": 8.108014106750488, | |
| "learning_rate": 3.6901268909804524e-05, | |
| "loss": 0.3557, | |
| "step": 27500 | |
| }, | |
| { | |
| "epoch": 0.2667377967457989, | |
| "grad_norm": 13.029576301574707, | |
| "learning_rate": 3.6663110162710054e-05, | |
| "loss": 0.3785, | |
| "step": 28000 | |
| }, | |
| { | |
| "epoch": 0.27150097168768816, | |
| "grad_norm": 9.865415573120117, | |
| "learning_rate": 3.642495141561559e-05, | |
| "loss": 0.4013, | |
| "step": 28500 | |
| }, | |
| { | |
| "epoch": 0.27626414662957743, | |
| "grad_norm": 4.568774700164795, | |
| "learning_rate": 3.618679266852113e-05, | |
| "loss": 0.3663, | |
| "step": 29000 | |
| }, | |
| { | |
| "epoch": 0.2810273215714667, | |
| "grad_norm": 3.7032947540283203, | |
| "learning_rate": 3.5948633921426665e-05, | |
| "loss": 0.3728, | |
| "step": 29500 | |
| }, | |
| { | |
| "epoch": 0.2857904965133559, | |
| "grad_norm": 5.560993194580078, | |
| "learning_rate": 3.57104751743322e-05, | |
| "loss": 0.364, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 0.2905536714552452, | |
| "grad_norm": 12.633008003234863, | |
| "learning_rate": 3.547231642723774e-05, | |
| "loss": 0.3738, | |
| "step": 30500 | |
| }, | |
| { | |
| "epoch": 0.29531684639713446, | |
| "grad_norm": 16.694622039794922, | |
| "learning_rate": 3.523415768014328e-05, | |
| "loss": 0.363, | |
| "step": 31000 | |
| }, | |
| { | |
| "epoch": 0.30008002133902373, | |
| "grad_norm": 6.716195583343506, | |
| "learning_rate": 3.4995998933048814e-05, | |
| "loss": 0.3547, | |
| "step": 31500 | |
| }, | |
| { | |
| "epoch": 0.304843196280913, | |
| "grad_norm": 5.371184349060059, | |
| "learning_rate": 3.475784018595435e-05, | |
| "loss": 0.3392, | |
| "step": 32000 | |
| }, | |
| { | |
| "epoch": 0.30960637122280227, | |
| "grad_norm": 15.001012802124023, | |
| "learning_rate": 3.451968143885989e-05, | |
| "loss": 0.3419, | |
| "step": 32500 | |
| }, | |
| { | |
| "epoch": 0.31436954616469154, | |
| "grad_norm": 14.817428588867188, | |
| "learning_rate": 3.4281522691765425e-05, | |
| "loss": 0.3454, | |
| "step": 33000 | |
| }, | |
| { | |
| "epoch": 0.3191327211065808, | |
| "grad_norm": 8.50633716583252, | |
| "learning_rate": 3.404336394467096e-05, | |
| "loss": 0.3322, | |
| "step": 33500 | |
| }, | |
| { | |
| "epoch": 0.3238958960484701, | |
| "grad_norm": 10.963394165039062, | |
| "learning_rate": 3.38052051975765e-05, | |
| "loss": 0.3507, | |
| "step": 34000 | |
| }, | |
| { | |
| "epoch": 0.32865907099035935, | |
| "grad_norm": 6.286830902099609, | |
| "learning_rate": 3.356704645048204e-05, | |
| "loss": 0.3662, | |
| "step": 34500 | |
| }, | |
| { | |
| "epoch": 0.3334222459322486, | |
| "grad_norm": 7.931427955627441, | |
| "learning_rate": 3.3328887703387574e-05, | |
| "loss": 0.3223, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 0.3381854208741379, | |
| "grad_norm": 0.015142062678933144, | |
| "learning_rate": 3.309072895629311e-05, | |
| "loss": 0.3492, | |
| "step": 35500 | |
| }, | |
| { | |
| "epoch": 0.34294859581602716, | |
| "grad_norm": 6.33823823928833, | |
| "learning_rate": 3.285257020919865e-05, | |
| "loss": 0.3443, | |
| "step": 36000 | |
| }, | |
| { | |
| "epoch": 0.34771177075791637, | |
| "grad_norm": 14.21090030670166, | |
| "learning_rate": 3.2614411462104185e-05, | |
| "loss": 0.3208, | |
| "step": 36500 | |
| }, | |
| { | |
| "epoch": 0.35247494569980564, | |
| "grad_norm": 6.757291793823242, | |
| "learning_rate": 3.237625271500972e-05, | |
| "loss": 0.3283, | |
| "step": 37000 | |
| }, | |
| { | |
| "epoch": 0.3572381206416949, | |
| "grad_norm": 25.391630172729492, | |
| "learning_rate": 3.213809396791525e-05, | |
| "loss": 0.3207, | |
| "step": 37500 | |
| }, | |
| { | |
| "epoch": 0.3620012955835842, | |
| "grad_norm": 4.002015113830566, | |
| "learning_rate": 3.189993522082079e-05, | |
| "loss": 0.3001, | |
| "step": 38000 | |
| }, | |
| { | |
| "epoch": 0.36676447052547345, | |
| "grad_norm": 8.16422176361084, | |
| "learning_rate": 3.166177647372633e-05, | |
| "loss": 0.2973, | |
| "step": 38500 | |
| }, | |
| { | |
| "epoch": 0.3715276454673627, | |
| "grad_norm": 4.861073017120361, | |
| "learning_rate": 3.1423617726631864e-05, | |
| "loss": 0.3104, | |
| "step": 39000 | |
| }, | |
| { | |
| "epoch": 0.376290820409252, | |
| "grad_norm": 9.573932647705078, | |
| "learning_rate": 3.11854589795374e-05, | |
| "loss": 0.2984, | |
| "step": 39500 | |
| }, | |
| { | |
| "epoch": 0.38105399535114126, | |
| "grad_norm": 10.641077041625977, | |
| "learning_rate": 3.094730023244294e-05, | |
| "loss": 0.314, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 0.38581717029303053, | |
| "grad_norm": 19.601299285888672, | |
| "learning_rate": 3.0709141485348475e-05, | |
| "loss": 0.292, | |
| "step": 40500 | |
| }, | |
| { | |
| "epoch": 0.3905803452349198, | |
| "grad_norm": 1.557869553565979, | |
| "learning_rate": 3.0470982738254012e-05, | |
| "loss": 0.3244, | |
| "step": 41000 | |
| }, | |
| { | |
| "epoch": 0.39534352017680907, | |
| "grad_norm": 9.880559921264648, | |
| "learning_rate": 3.0232823991159546e-05, | |
| "loss": 0.3053, | |
| "step": 41500 | |
| }, | |
| { | |
| "epoch": 0.40010669511869834, | |
| "grad_norm": 16.77007293701172, | |
| "learning_rate": 2.9994665244065083e-05, | |
| "loss": 0.2935, | |
| "step": 42000 | |
| }, | |
| { | |
| "epoch": 0.4048698700605876, | |
| "grad_norm": 15.44863510131836, | |
| "learning_rate": 2.975650649697062e-05, | |
| "loss": 0.2885, | |
| "step": 42500 | |
| }, | |
| { | |
| "epoch": 0.4096330450024768, | |
| "grad_norm": 2.127357006072998, | |
| "learning_rate": 2.9518347749876158e-05, | |
| "loss": 0.2896, | |
| "step": 43000 | |
| }, | |
| { | |
| "epoch": 0.4143962199443661, | |
| "grad_norm": 4.906562328338623, | |
| "learning_rate": 2.9280189002781695e-05, | |
| "loss": 0.2961, | |
| "step": 43500 | |
| }, | |
| { | |
| "epoch": 0.41915939488625537, | |
| "grad_norm": 5.996885299682617, | |
| "learning_rate": 2.9042030255687232e-05, | |
| "loss": 0.283, | |
| "step": 44000 | |
| }, | |
| { | |
| "epoch": 0.42392256982814464, | |
| "grad_norm": 4.349148750305176, | |
| "learning_rate": 2.8803871508592766e-05, | |
| "loss": 0.285, | |
| "step": 44500 | |
| }, | |
| { | |
| "epoch": 0.4286857447700339, | |
| "grad_norm": 11.325324058532715, | |
| "learning_rate": 2.8565712761498303e-05, | |
| "loss": 0.2991, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 0.4334489197119232, | |
| "grad_norm": 29.92687225341797, | |
| "learning_rate": 2.832755401440384e-05, | |
| "loss": 0.2758, | |
| "step": 45500 | |
| }, | |
| { | |
| "epoch": 0.43821209465381245, | |
| "grad_norm": 6.005437850952148, | |
| "learning_rate": 2.8089395267309377e-05, | |
| "loss": 0.285, | |
| "step": 46000 | |
| }, | |
| { | |
| "epoch": 0.4429752695957017, | |
| "grad_norm": 7.799108505249023, | |
| "learning_rate": 2.7851236520214917e-05, | |
| "loss": 0.2851, | |
| "step": 46500 | |
| }, | |
| { | |
| "epoch": 0.447738444537591, | |
| "grad_norm": 7.533286094665527, | |
| "learning_rate": 2.7613077773120455e-05, | |
| "loss": 0.2713, | |
| "step": 47000 | |
| }, | |
| { | |
| "epoch": 0.45250161947948025, | |
| "grad_norm": 16.269254684448242, | |
| "learning_rate": 2.737491902602599e-05, | |
| "loss": 0.2954, | |
| "step": 47500 | |
| }, | |
| { | |
| "epoch": 0.4572647944213695, | |
| "grad_norm": 12.095314025878906, | |
| "learning_rate": 2.713676027893153e-05, | |
| "loss": 0.2733, | |
| "step": 48000 | |
| }, | |
| { | |
| "epoch": 0.4620279693632588, | |
| "grad_norm": 9.375202178955078, | |
| "learning_rate": 2.6898601531837066e-05, | |
| "loss": 0.266, | |
| "step": 48500 | |
| }, | |
| { | |
| "epoch": 0.46679114430514806, | |
| "grad_norm": 18.472373962402344, | |
| "learning_rate": 2.66604427847426e-05, | |
| "loss": 0.2616, | |
| "step": 49000 | |
| }, | |
| { | |
| "epoch": 0.4715543192470373, | |
| "grad_norm": 10.298583984375, | |
| "learning_rate": 2.6422284037648137e-05, | |
| "loss": 0.273, | |
| "step": 49500 | |
| }, | |
| { | |
| "epoch": 0.47631749418892655, | |
| "grad_norm": 5.778076648712158, | |
| "learning_rate": 2.6184125290553674e-05, | |
| "loss": 0.2754, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 0.4810806691308158, | |
| "grad_norm": 3.6435136795043945, | |
| "learning_rate": 2.594596654345921e-05, | |
| "loss": 0.2664, | |
| "step": 50500 | |
| }, | |
| { | |
| "epoch": 0.4858438440727051, | |
| "grad_norm": 21.803163528442383, | |
| "learning_rate": 2.5707807796364748e-05, | |
| "loss": 0.2621, | |
| "step": 51000 | |
| }, | |
| { | |
| "epoch": 0.49060701901459436, | |
| "grad_norm": 11.081718444824219, | |
| "learning_rate": 2.5469649049270282e-05, | |
| "loss": 0.2806, | |
| "step": 51500 | |
| }, | |
| { | |
| "epoch": 0.49537019395648363, | |
| "grad_norm": 3.4980528354644775, | |
| "learning_rate": 2.523149030217582e-05, | |
| "loss": 0.278, | |
| "step": 52000 | |
| }, | |
| { | |
| "epoch": 0.5001333688983729, | |
| "grad_norm": 7.148010730743408, | |
| "learning_rate": 2.4993331555081356e-05, | |
| "loss": 0.2592, | |
| "step": 52500 | |
| }, | |
| { | |
| "epoch": 0.5048965438402622, | |
| "grad_norm": 3.838651657104492, | |
| "learning_rate": 2.4755172807986893e-05, | |
| "loss": 0.2636, | |
| "step": 53000 | |
| }, | |
| { | |
| "epoch": 0.5096597187821514, | |
| "grad_norm": 12.54211711883545, | |
| "learning_rate": 2.451701406089243e-05, | |
| "loss": 0.2483, | |
| "step": 53500 | |
| }, | |
| { | |
| "epoch": 0.5144228937240407, | |
| "grad_norm": 17.774137496948242, | |
| "learning_rate": 2.4278855313797964e-05, | |
| "loss": 0.2442, | |
| "step": 54000 | |
| }, | |
| { | |
| "epoch": 0.51918606866593, | |
| "grad_norm": 10.924212455749512, | |
| "learning_rate": 2.40406965667035e-05, | |
| "loss": 0.2367, | |
| "step": 54500 | |
| }, | |
| { | |
| "epoch": 0.5239492436078192, | |
| "grad_norm": 0.054358381778001785, | |
| "learning_rate": 2.380253781960904e-05, | |
| "loss": 0.2239, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 0.5287124185497085, | |
| "grad_norm": 1.8560116291046143, | |
| "learning_rate": 2.3564379072514575e-05, | |
| "loss": 0.2621, | |
| "step": 55500 | |
| }, | |
| { | |
| "epoch": 0.5334755934915978, | |
| "grad_norm": 0.23067010939121246, | |
| "learning_rate": 2.3326220325420116e-05, | |
| "loss": 0.245, | |
| "step": 56000 | |
| }, | |
| { | |
| "epoch": 0.5382387684334871, | |
| "grad_norm": 9.185586929321289, | |
| "learning_rate": 2.308806157832565e-05, | |
| "loss": 0.2578, | |
| "step": 56500 | |
| }, | |
| { | |
| "epoch": 0.5430019433753763, | |
| "grad_norm": 5.946952819824219, | |
| "learning_rate": 2.2849902831231187e-05, | |
| "loss": 0.2053, | |
| "step": 57000 | |
| }, | |
| { | |
| "epoch": 0.5477651183172656, | |
| "grad_norm": 21.640161514282227, | |
| "learning_rate": 2.2611744084136724e-05, | |
| "loss": 0.2419, | |
| "step": 57500 | |
| }, | |
| { | |
| "epoch": 0.5525282932591549, | |
| "grad_norm": 4.966391563415527, | |
| "learning_rate": 2.237358533704226e-05, | |
| "loss": 0.2472, | |
| "step": 58000 | |
| }, | |
| { | |
| "epoch": 0.5572914682010441, | |
| "grad_norm": 6.430874347686768, | |
| "learning_rate": 2.2135426589947798e-05, | |
| "loss": 0.2176, | |
| "step": 58500 | |
| }, | |
| { | |
| "epoch": 0.5620546431429334, | |
| "grad_norm": 1.1396760940551758, | |
| "learning_rate": 2.1897267842853332e-05, | |
| "loss": 0.2297, | |
| "step": 59000 | |
| }, | |
| { | |
| "epoch": 0.5668178180848226, | |
| "grad_norm": 0.023226283490657806, | |
| "learning_rate": 2.165910909575887e-05, | |
| "loss": 0.2286, | |
| "step": 59500 | |
| }, | |
| { | |
| "epoch": 0.5715809930267118, | |
| "grad_norm": 7.5434770584106445, | |
| "learning_rate": 2.1420950348664406e-05, | |
| "loss": 0.2453, | |
| "step": 60000 | |
| }, | |
| { | |
| "epoch": 0.5763441679686011, | |
| "grad_norm": 20.513235092163086, | |
| "learning_rate": 2.1182791601569943e-05, | |
| "loss": 0.2237, | |
| "step": 60500 | |
| }, | |
| { | |
| "epoch": 0.5811073429104904, | |
| "grad_norm": 0.2174549102783203, | |
| "learning_rate": 2.094463285447548e-05, | |
| "loss": 0.2206, | |
| "step": 61000 | |
| }, | |
| { | |
| "epoch": 0.5858705178523796, | |
| "grad_norm": 5.196407794952393, | |
| "learning_rate": 2.0706474107381018e-05, | |
| "loss": 0.2249, | |
| "step": 61500 | |
| }, | |
| { | |
| "epoch": 0.5906336927942689, | |
| "grad_norm": 6.904590129852295, | |
| "learning_rate": 2.046831536028655e-05, | |
| "loss": 0.2158, | |
| "step": 62000 | |
| }, | |
| { | |
| "epoch": 0.5953968677361582, | |
| "grad_norm": 4.865005970001221, | |
| "learning_rate": 2.0230156613192092e-05, | |
| "loss": 0.2369, | |
| "step": 62500 | |
| }, | |
| { | |
| "epoch": 0.6001600426780475, | |
| "grad_norm": 8.591822624206543, | |
| "learning_rate": 1.999199786609763e-05, | |
| "loss": 0.2186, | |
| "step": 63000 | |
| }, | |
| { | |
| "epoch": 0.6049232176199367, | |
| "grad_norm": 11.895712852478027, | |
| "learning_rate": 1.9753839119003166e-05, | |
| "loss": 0.2175, | |
| "step": 63500 | |
| }, | |
| { | |
| "epoch": 0.609686392561826, | |
| "grad_norm": 29.44073486328125, | |
| "learning_rate": 1.95156803719087e-05, | |
| "loss": 0.239, | |
| "step": 64000 | |
| }, | |
| { | |
| "epoch": 0.6144495675037153, | |
| "grad_norm": 0.14603924751281738, | |
| "learning_rate": 1.9277521624814237e-05, | |
| "loss": 0.2342, | |
| "step": 64500 | |
| }, | |
| { | |
| "epoch": 0.6192127424456045, | |
| "grad_norm": 8.721657752990723, | |
| "learning_rate": 1.9039362877719774e-05, | |
| "loss": 0.2072, | |
| "step": 65000 | |
| }, | |
| { | |
| "epoch": 0.6239759173874938, | |
| "grad_norm": 7.2579755783081055, | |
| "learning_rate": 1.880120413062531e-05, | |
| "loss": 0.2176, | |
| "step": 65500 | |
| }, | |
| { | |
| "epoch": 0.6287390923293831, | |
| "grad_norm": 10.267131805419922, | |
| "learning_rate": 1.8563045383530848e-05, | |
| "loss": 0.2118, | |
| "step": 66000 | |
| }, | |
| { | |
| "epoch": 0.6335022672712723, | |
| "grad_norm": 12.264985084533691, | |
| "learning_rate": 1.8324886636436382e-05, | |
| "loss": 0.2215, | |
| "step": 66500 | |
| }, | |
| { | |
| "epoch": 0.6382654422131616, | |
| "grad_norm": 15.631693840026855, | |
| "learning_rate": 1.808672788934192e-05, | |
| "loss": 0.2213, | |
| "step": 67000 | |
| }, | |
| { | |
| "epoch": 0.6430286171550509, | |
| "grad_norm": 9.132320404052734, | |
| "learning_rate": 1.7848569142247456e-05, | |
| "loss": 0.2007, | |
| "step": 67500 | |
| }, | |
| { | |
| "epoch": 0.6477917920969402, | |
| "grad_norm": 2.836796283721924, | |
| "learning_rate": 1.7610410395152993e-05, | |
| "loss": 0.2197, | |
| "step": 68000 | |
| }, | |
| { | |
| "epoch": 0.6525549670388294, | |
| "grad_norm": 1.20395028591156, | |
| "learning_rate": 1.737225164805853e-05, | |
| "loss": 0.2092, | |
| "step": 68500 | |
| }, | |
| { | |
| "epoch": 0.6573181419807187, | |
| "grad_norm": 4.42557954788208, | |
| "learning_rate": 1.7134092900964068e-05, | |
| "loss": 0.2228, | |
| "step": 69000 | |
| }, | |
| { | |
| "epoch": 0.662081316922608, | |
| "grad_norm": 7.136639595031738, | |
| "learning_rate": 1.6895934153869605e-05, | |
| "loss": 0.2159, | |
| "step": 69500 | |
| }, | |
| { | |
| "epoch": 0.6668444918644972, | |
| "grad_norm": 12.695110321044922, | |
| "learning_rate": 1.6657775406775142e-05, | |
| "loss": 0.1984, | |
| "step": 70000 | |
| }, | |
| { | |
| "epoch": 0.6716076668063865, | |
| "grad_norm": 11.841693878173828, | |
| "learning_rate": 1.641961665968068e-05, | |
| "loss": 0.1913, | |
| "step": 70500 | |
| }, | |
| { | |
| "epoch": 0.6763708417482758, | |
| "grad_norm": 0.7839029431343079, | |
| "learning_rate": 1.6181457912586216e-05, | |
| "loss": 0.1911, | |
| "step": 71000 | |
| }, | |
| { | |
| "epoch": 0.681134016690165, | |
| "grad_norm": 6.188957691192627, | |
| "learning_rate": 1.594329916549175e-05, | |
| "loss": 0.2164, | |
| "step": 71500 | |
| }, | |
| { | |
| "epoch": 0.6858971916320543, | |
| "grad_norm": 11.414396286010742, | |
| "learning_rate": 1.5705140418397287e-05, | |
| "loss": 0.2026, | |
| "step": 72000 | |
| }, | |
| { | |
| "epoch": 0.6906603665739435, | |
| "grad_norm": 21.34324836730957, | |
| "learning_rate": 1.5466981671302824e-05, | |
| "loss": 0.1877, | |
| "step": 72500 | |
| }, | |
| { | |
| "epoch": 0.6954235415158327, | |
| "grad_norm": 10.534087181091309, | |
| "learning_rate": 1.5228822924208361e-05, | |
| "loss": 0.1889, | |
| "step": 73000 | |
| }, | |
| { | |
| "epoch": 0.700186716457722, | |
| "grad_norm": 0.4472333788871765, | |
| "learning_rate": 1.4990664177113897e-05, | |
| "loss": 0.1902, | |
| "step": 73500 | |
| }, | |
| { | |
| "epoch": 0.7049498913996113, | |
| "grad_norm": 27.742639541625977, | |
| "learning_rate": 1.4752505430019434e-05, | |
| "loss": 0.1929, | |
| "step": 74000 | |
| }, | |
| { | |
| "epoch": 0.7097130663415006, | |
| "grad_norm": 12.095074653625488, | |
| "learning_rate": 1.451434668292497e-05, | |
| "loss": 0.1754, | |
| "step": 74500 | |
| }, | |
| { | |
| "epoch": 0.7144762412833898, | |
| "grad_norm": 10.153932571411133, | |
| "learning_rate": 1.4276187935830506e-05, | |
| "loss": 0.1866, | |
| "step": 75000 | |
| }, | |
| { | |
| "epoch": 0.7192394162252791, | |
| "grad_norm": 9.065908432006836, | |
| "learning_rate": 1.4038029188736043e-05, | |
| "loss": 0.2004, | |
| "step": 75500 | |
| }, | |
| { | |
| "epoch": 0.7240025911671684, | |
| "grad_norm": 1.7107542753219604, | |
| "learning_rate": 1.3799870441641582e-05, | |
| "loss": 0.175, | |
| "step": 76000 | |
| }, | |
| { | |
| "epoch": 0.7287657661090576, | |
| "grad_norm": 1.4907644987106323, | |
| "learning_rate": 1.356171169454712e-05, | |
| "loss": 0.1865, | |
| "step": 76500 | |
| }, | |
| { | |
| "epoch": 0.7335289410509469, | |
| "grad_norm": 0.11667291074991226, | |
| "learning_rate": 1.3323552947452655e-05, | |
| "loss": 0.1694, | |
| "step": 77000 | |
| }, | |
| { | |
| "epoch": 0.7382921159928362, | |
| "grad_norm": 29.062976837158203, | |
| "learning_rate": 1.3085394200358192e-05, | |
| "loss": 0.1821, | |
| "step": 77500 | |
| }, | |
| { | |
| "epoch": 0.7430552909347254, | |
| "grad_norm": 7.0561113357543945, | |
| "learning_rate": 1.2847235453263729e-05, | |
| "loss": 0.1692, | |
| "step": 78000 | |
| }, | |
| { | |
| "epoch": 0.7478184658766147, | |
| "grad_norm": 7.916496753692627, | |
| "learning_rate": 1.2609076706169264e-05, | |
| "loss": 0.1711, | |
| "step": 78500 | |
| }, | |
| { | |
| "epoch": 0.752581640818504, | |
| "grad_norm": 0.009064608253538609, | |
| "learning_rate": 1.2370917959074802e-05, | |
| "loss": 0.1795, | |
| "step": 79000 | |
| }, | |
| { | |
| "epoch": 0.7573448157603933, | |
| "grad_norm": 16.94624900817871, | |
| "learning_rate": 1.2132759211980339e-05, | |
| "loss": 0.1732, | |
| "step": 79500 | |
| }, | |
| { | |
| "epoch": 0.7621079907022825, | |
| "grad_norm": 24.09770965576172, | |
| "learning_rate": 1.1894600464885874e-05, | |
| "loss": 0.1599, | |
| "step": 80000 | |
| }, | |
| { | |
| "epoch": 0.7668711656441718, | |
| "grad_norm": 17.340219497680664, | |
| "learning_rate": 1.1656441717791411e-05, | |
| "loss": 0.1702, | |
| "step": 80500 | |
| }, | |
| { | |
| "epoch": 0.7716343405860611, | |
| "grad_norm": 0.015018216334283352, | |
| "learning_rate": 1.1418282970696948e-05, | |
| "loss": 0.1679, | |
| "step": 81000 | |
| }, | |
| { | |
| "epoch": 0.7763975155279503, | |
| "grad_norm": 16.005643844604492, | |
| "learning_rate": 1.1180124223602485e-05, | |
| "loss": 0.1832, | |
| "step": 81500 | |
| }, | |
| { | |
| "epoch": 0.7811606904698396, | |
| "grad_norm": 0.0013979446375742555, | |
| "learning_rate": 1.0941965476508023e-05, | |
| "loss": 0.1505, | |
| "step": 82000 | |
| }, | |
| { | |
| "epoch": 0.7859238654117289, | |
| "grad_norm": 1.049574375152588, | |
| "learning_rate": 1.0703806729413558e-05, | |
| "loss": 0.1663, | |
| "step": 82500 | |
| }, | |
| { | |
| "epoch": 0.7906870403536181, | |
| "grad_norm": 5.492803573608398, | |
| "learning_rate": 1.0465647982319095e-05, | |
| "loss": 0.1561, | |
| "step": 83000 | |
| }, | |
| { | |
| "epoch": 0.7954502152955074, | |
| "grad_norm": 14.530741691589355, | |
| "learning_rate": 1.022748923522463e-05, | |
| "loss": 0.1876, | |
| "step": 83500 | |
| }, | |
| { | |
| "epoch": 0.8002133902373967, | |
| "grad_norm": 0.0616171695291996, | |
| "learning_rate": 9.98933048813017e-06, | |
| "loss": 0.1524, | |
| "step": 84000 | |
| }, | |
| { | |
| "epoch": 0.804976565179286, | |
| "grad_norm": 15.610015869140625, | |
| "learning_rate": 9.751171741035706e-06, | |
| "loss": 0.1588, | |
| "step": 84500 | |
| }, | |
| { | |
| "epoch": 0.8097397401211752, | |
| "grad_norm": 7.8352885246276855, | |
| "learning_rate": 9.513012993941242e-06, | |
| "loss": 0.1779, | |
| "step": 85000 | |
| }, | |
| { | |
| "epoch": 0.8145029150630644, | |
| "grad_norm": 21.532123565673828, | |
| "learning_rate": 9.274854246846779e-06, | |
| "loss": 0.1436, | |
| "step": 85500 | |
| }, | |
| { | |
| "epoch": 0.8192660900049537, | |
| "grad_norm": 0.025519462302327156, | |
| "learning_rate": 9.036695499752314e-06, | |
| "loss": 0.1472, | |
| "step": 86000 | |
| }, | |
| { | |
| "epoch": 0.8240292649468429, | |
| "grad_norm": 0.09057486802339554, | |
| "learning_rate": 8.798536752657852e-06, | |
| "loss": 0.1525, | |
| "step": 86500 | |
| }, | |
| { | |
| "epoch": 0.8287924398887322, | |
| "grad_norm": 0.9066371917724609, | |
| "learning_rate": 8.560378005563389e-06, | |
| "loss": 0.1614, | |
| "step": 87000 | |
| }, | |
| { | |
| "epoch": 0.8335556148306215, | |
| "grad_norm": 2.612293004989624, | |
| "learning_rate": 8.322219258468926e-06, | |
| "loss": 0.1492, | |
| "step": 87500 | |
| }, | |
| { | |
| "epoch": 0.8383187897725107, | |
| "grad_norm": 6.420555114746094, | |
| "learning_rate": 8.084060511374463e-06, | |
| "loss": 0.1518, | |
| "step": 88000 | |
| }, | |
| { | |
| "epoch": 0.8430819647144, | |
| "grad_norm": 0.04058153182268143, | |
| "learning_rate": 7.845901764279998e-06, | |
| "loss": 0.1581, | |
| "step": 88500 | |
| }, | |
| { | |
| "epoch": 0.8478451396562893, | |
| "grad_norm": 17.142908096313477, | |
| "learning_rate": 7.6077430171855355e-06, | |
| "loss": 0.166, | |
| "step": 89000 | |
| }, | |
| { | |
| "epoch": 0.8526083145981785, | |
| "grad_norm": 17.988386154174805, | |
| "learning_rate": 7.369584270091072e-06, | |
| "loss": 0.1613, | |
| "step": 89500 | |
| }, | |
| { | |
| "epoch": 0.8573714895400678, | |
| "grad_norm": 0.28535395860671997, | |
| "learning_rate": 7.131425522996608e-06, | |
| "loss": 0.1448, | |
| "step": 90000 | |
| }, | |
| { | |
| "epoch": 0.8621346644819571, | |
| "grad_norm": 0.02261945605278015, | |
| "learning_rate": 6.893266775902146e-06, | |
| "loss": 0.1397, | |
| "step": 90500 | |
| }, | |
| { | |
| "epoch": 0.8668978394238464, | |
| "grad_norm": 15.836788177490234, | |
| "learning_rate": 6.655108028807683e-06, | |
| "loss": 0.146, | |
| "step": 91000 | |
| }, | |
| { | |
| "epoch": 0.8716610143657356, | |
| "grad_norm": 0.6049064993858337, | |
| "learning_rate": 6.416949281713219e-06, | |
| "loss": 0.1405, | |
| "step": 91500 | |
| }, | |
| { | |
| "epoch": 0.8764241893076249, | |
| "grad_norm": 0.08131851255893707, | |
| "learning_rate": 6.178790534618756e-06, | |
| "loss": 0.1466, | |
| "step": 92000 | |
| }, | |
| { | |
| "epoch": 0.8811873642495142, | |
| "grad_norm": 16.24254608154297, | |
| "learning_rate": 5.940631787524292e-06, | |
| "loss": 0.1398, | |
| "step": 92500 | |
| }, | |
| { | |
| "epoch": 0.8859505391914034, | |
| "grad_norm": 8.061177504714578e-05, | |
| "learning_rate": 5.702473040429829e-06, | |
| "loss": 0.124, | |
| "step": 93000 | |
| }, | |
| { | |
| "epoch": 0.8907137141332927, | |
| "grad_norm": 2.3429534435272217, | |
| "learning_rate": 5.464314293335366e-06, | |
| "loss": 0.1448, | |
| "step": 93500 | |
| }, | |
| { | |
| "epoch": 0.895476889075182, | |
| "grad_norm": 21.22195053100586, | |
| "learning_rate": 5.2261555462409025e-06, | |
| "loss": 0.1304, | |
| "step": 94000 | |
| }, | |
| { | |
| "epoch": 0.9002400640170712, | |
| "grad_norm": 22.701601028442383, | |
| "learning_rate": 4.9879967991464396e-06, | |
| "loss": 0.1211, | |
| "step": 94500 | |
| }, | |
| { | |
| "epoch": 0.9050032389589605, | |
| "grad_norm": 15.280872344970703, | |
| "learning_rate": 4.749838052051976e-06, | |
| "loss": 0.1514, | |
| "step": 95000 | |
| }, | |
| { | |
| "epoch": 0.9097664139008498, | |
| "grad_norm": 0.01960950717329979, | |
| "learning_rate": 4.511679304957513e-06, | |
| "loss": 0.1115, | |
| "step": 95500 | |
| }, | |
| { | |
| "epoch": 0.914529588842739, | |
| "grad_norm": 0.0017501560505479574, | |
| "learning_rate": 4.273520557863049e-06, | |
| "loss": 0.1571, | |
| "step": 96000 | |
| }, | |
| { | |
| "epoch": 0.9192927637846283, | |
| "grad_norm": 0.05184149742126465, | |
| "learning_rate": 4.035361810768586e-06, | |
| "loss": 0.1499, | |
| "step": 96500 | |
| }, | |
| { | |
| "epoch": 0.9240559387265176, | |
| "grad_norm": 6.8865180015563965, | |
| "learning_rate": 3.7972030636741226e-06, | |
| "loss": 0.1272, | |
| "step": 97000 | |
| }, | |
| { | |
| "epoch": 0.9288191136684069, | |
| "grad_norm": 0.10639504343271255, | |
| "learning_rate": 3.5590443165796593e-06, | |
| "loss": 0.1263, | |
| "step": 97500 | |
| }, | |
| { | |
| "epoch": 0.9335822886102961, | |
| "grad_norm": 0.9005939960479736, | |
| "learning_rate": 3.3208855694851965e-06, | |
| "loss": 0.133, | |
| "step": 98000 | |
| }, | |
| { | |
| "epoch": 0.9383454635521854, | |
| "grad_norm": 1.1748387813568115, | |
| "learning_rate": 3.082726822390733e-06, | |
| "loss": 0.1325, | |
| "step": 98500 | |
| }, | |
| { | |
| "epoch": 0.9431086384940746, | |
| "grad_norm": 28.960805892944336, | |
| "learning_rate": 2.8445680752962694e-06, | |
| "loss": 0.1319, | |
| "step": 99000 | |
| }, | |
| { | |
| "epoch": 0.9478718134359638, | |
| "grad_norm": 8.118181228637695, | |
| "learning_rate": 2.606409328201806e-06, | |
| "loss": 0.1531, | |
| "step": 99500 | |
| }, | |
| { | |
| "epoch": 0.9526349883778531, | |
| "grad_norm": 0.004353045951575041, | |
| "learning_rate": 2.3682505811073433e-06, | |
| "loss": 0.1275, | |
| "step": 100000 | |
| }, | |
| { | |
| "epoch": 0.9573981633197424, | |
| "grad_norm": 2.1794090270996094, | |
| "learning_rate": 2.1300918340128795e-06, | |
| "loss": 0.1428, | |
| "step": 100500 | |
| }, | |
| { | |
| "epoch": 0.9621613382616316, | |
| "grad_norm": 8.74696159362793, | |
| "learning_rate": 1.8919330869184164e-06, | |
| "loss": 0.1135, | |
| "step": 101000 | |
| }, | |
| { | |
| "epoch": 0.9669245132035209, | |
| "grad_norm": 31.47783851623535, | |
| "learning_rate": 1.6537743398239533e-06, | |
| "loss": 0.1391, | |
| "step": 101500 | |
| }, | |
| { | |
| "epoch": 0.9716876881454102, | |
| "grad_norm": 1.8234331607818604, | |
| "learning_rate": 1.4156155927294898e-06, | |
| "loss": 0.12, | |
| "step": 102000 | |
| }, | |
| { | |
| "epoch": 0.9764508630872994, | |
| "grad_norm": 5.736881732940674, | |
| "learning_rate": 1.1774568456350265e-06, | |
| "loss": 0.1253, | |
| "step": 102500 | |
| }, | |
| { | |
| "epoch": 0.9812140380291887, | |
| "grad_norm": 0.919281005859375, | |
| "learning_rate": 9.392980985405633e-07, | |
| "loss": 0.1216, | |
| "step": 103000 | |
| }, | |
| { | |
| "epoch": 0.985977212971078, | |
| "grad_norm": 6.265872001647949, | |
| "learning_rate": 7.011393514460999e-07, | |
| "loss": 0.1245, | |
| "step": 103500 | |
| }, | |
| { | |
| "epoch": 0.9907403879129673, | |
| "grad_norm": 19.050594329833984, | |
| "learning_rate": 4.629806043516366e-07, | |
| "loss": 0.1298, | |
| "step": 104000 | |
| }, | |
| { | |
| "epoch": 0.9955035628548565, | |
| "grad_norm": 9.125242233276367, | |
| "learning_rate": 2.2482185725717335e-07, | |
| "loss": 0.1177, | |
| "step": 104500 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 104972, | |
| "total_flos": 1.5105749507715635e+17, | |
| "train_loss": 0.05896880445613715, | |
| "train_runtime": 27901.1518, | |
| "train_samples_per_second": 30.098, | |
| "train_steps_per_second": 3.762 | |
| }, | |
| { | |
| "epoch": 1.0002667377967458, | |
| "grad_norm": 14.302581787109375, | |
| "learning_rate": 2.2480186776155165e-07, | |
| "loss": 0.0893, | |
| "step": 105000 | |
| }, | |
| { | |
| "epoch": 1.005029912738635, | |
| "grad_norm": 2.5177512168884277, | |
| "learning_rate": 2.2444491248259286e-07, | |
| "loss": 0.1055, | |
| "step": 105500 | |
| }, | |
| { | |
| "epoch": 1.0097930876805243, | |
| "grad_norm": 2.1406190395355225, | |
| "learning_rate": 2.2408795720363407e-07, | |
| "loss": 0.1254, | |
| "step": 106000 | |
| }, | |
| { | |
| "epoch": 1.0145562626224136, | |
| "grad_norm": 1.1113516092300415, | |
| "learning_rate": 2.2373100192467528e-07, | |
| "loss": 0.0951, | |
| "step": 106500 | |
| }, | |
| { | |
| "epoch": 1.0193194375643029, | |
| "grad_norm": 35.0811653137207, | |
| "learning_rate": 2.233740466457165e-07, | |
| "loss": 0.11, | |
| "step": 107000 | |
| }, | |
| { | |
| "epoch": 1.0240826125061921, | |
| "grad_norm": 11.133783340454102, | |
| "learning_rate": 2.230170913667577e-07, | |
| "loss": 0.1132, | |
| "step": 107500 | |
| }, | |
| { | |
| "epoch": 1.0288457874480814, | |
| "grad_norm": 4.060661315917969, | |
| "learning_rate": 2.226601360877989e-07, | |
| "loss": 0.1202, | |
| "step": 108000 | |
| }, | |
| { | |
| "epoch": 1.0336089623899707, | |
| "grad_norm": 0.13336028158664703, | |
| "learning_rate": 2.2230318080884012e-07, | |
| "loss": 0.1139, | |
| "step": 108500 | |
| }, | |
| { | |
| "epoch": 1.03837213733186, | |
| "grad_norm": 1.6817339658737183, | |
| "learning_rate": 2.2194622552988133e-07, | |
| "loss": 0.1093, | |
| "step": 109000 | |
| }, | |
| { | |
| "epoch": 1.0431353122737492, | |
| "grad_norm": 7.226846694946289, | |
| "learning_rate": 2.215892702509225e-07, | |
| "loss": 0.1019, | |
| "step": 109500 | |
| }, | |
| { | |
| "epoch": 1.0478984872156385, | |
| "grad_norm": 20.337854385375977, | |
| "learning_rate": 2.2123231497196372e-07, | |
| "loss": 0.1159, | |
| "step": 110000 | |
| }, | |
| { | |
| "epoch": 1.0526616621575278, | |
| "grad_norm": 12.429189682006836, | |
| "learning_rate": 2.2087535969300493e-07, | |
| "loss": 0.1058, | |
| "step": 110500 | |
| }, | |
| { | |
| "epoch": 1.057424837099417, | |
| "grad_norm": 2.344127655029297, | |
| "learning_rate": 2.2051840441404614e-07, | |
| "loss": 0.1252, | |
| "step": 111000 | |
| }, | |
| { | |
| "epoch": 1.0621880120413063, | |
| "grad_norm": 3.0213356018066406, | |
| "learning_rate": 2.2016144913508735e-07, | |
| "loss": 0.1091, | |
| "step": 111500 | |
| }, | |
| { | |
| "epoch": 1.0669511869831956, | |
| "grad_norm": 8.44406509399414, | |
| "learning_rate": 2.198044938561286e-07, | |
| "loss": 0.1084, | |
| "step": 112000 | |
| }, | |
| { | |
| "epoch": 1.0717143619250848, | |
| "grad_norm": 5.177192211151123, | |
| "learning_rate": 2.194475385771698e-07, | |
| "loss": 0.111, | |
| "step": 112500 | |
| }, | |
| { | |
| "epoch": 1.0764775368669741, | |
| "grad_norm": 0.7378814816474915, | |
| "learning_rate": 2.19090583298211e-07, | |
| "loss": 0.1123, | |
| "step": 113000 | |
| }, | |
| { | |
| "epoch": 1.0812407118088634, | |
| "grad_norm": 11.683990478515625, | |
| "learning_rate": 2.1873362801925222e-07, | |
| "loss": 0.1049, | |
| "step": 113500 | |
| }, | |
| { | |
| "epoch": 1.0860038867507527, | |
| "grad_norm": 1.845595121383667, | |
| "learning_rate": 2.1837667274029343e-07, | |
| "loss": 0.1359, | |
| "step": 114000 | |
| }, | |
| { | |
| "epoch": 1.090767061692642, | |
| "grad_norm": 18.56471824645996, | |
| "learning_rate": 2.1801971746133464e-07, | |
| "loss": 0.1101, | |
| "step": 114500 | |
| }, | |
| { | |
| "epoch": 1.0955302366345312, | |
| "grad_norm": 2.1949081420898438, | |
| "learning_rate": 2.1766276218237585e-07, | |
| "loss": 0.1213, | |
| "step": 115000 | |
| }, | |
| { | |
| "epoch": 1.1002934115764205, | |
| "grad_norm": 8.824433326721191, | |
| "learning_rate": 2.1730580690341706e-07, | |
| "loss": 0.0972, | |
| "step": 115500 | |
| }, | |
| { | |
| "epoch": 1.1050565865183097, | |
| "grad_norm": 8.95617389678955, | |
| "learning_rate": 2.1694885162445827e-07, | |
| "loss": 0.1088, | |
| "step": 116000 | |
| }, | |
| { | |
| "epoch": 1.109819761460199, | |
| "grad_norm": 16.236719131469727, | |
| "learning_rate": 2.1659189634549948e-07, | |
| "loss": 0.126, | |
| "step": 116500 | |
| }, | |
| { | |
| "epoch": 1.1145829364020883, | |
| "grad_norm": 0.8696116805076599, | |
| "learning_rate": 2.1623494106654069e-07, | |
| "loss": 0.111, | |
| "step": 117000 | |
| }, | |
| { | |
| "epoch": 1.1193461113439775, | |
| "grad_norm": 22.61142349243164, | |
| "learning_rate": 2.158779857875819e-07, | |
| "loss": 0.1128, | |
| "step": 117500 | |
| }, | |
| { | |
| "epoch": 1.1241092862858668, | |
| "grad_norm": 9.516005516052246, | |
| "learning_rate": 2.155210305086231e-07, | |
| "loss": 0.1262, | |
| "step": 118000 | |
| }, | |
| { | |
| "epoch": 1.1288724612277559, | |
| "grad_norm": 2.8408925533294678, | |
| "learning_rate": 2.1516407522966432e-07, | |
| "loss": 0.1254, | |
| "step": 118500 | |
| }, | |
| { | |
| "epoch": 1.1336356361696454, | |
| "grad_norm": 0.08307074010372162, | |
| "learning_rate": 2.148071199507055e-07, | |
| "loss": 0.1128, | |
| "step": 119000 | |
| }, | |
| { | |
| "epoch": 1.1383988111115344, | |
| "grad_norm": 0.07671812921762466, | |
| "learning_rate": 2.144501646717467e-07, | |
| "loss": 0.1105, | |
| "step": 119500 | |
| }, | |
| { | |
| "epoch": 1.1431619860534237, | |
| "grad_norm": 23.30893325805664, | |
| "learning_rate": 2.1409320939278792e-07, | |
| "loss": 0.1016, | |
| "step": 120000 | |
| }, | |
| { | |
| "epoch": 1.147925160995313, | |
| "grad_norm": 2.4784510135650635, | |
| "learning_rate": 2.1373625411382913e-07, | |
| "loss": 0.121, | |
| "step": 120500 | |
| }, | |
| { | |
| "epoch": 1.1526883359372022, | |
| "grad_norm": 6.876108169555664, | |
| "learning_rate": 2.1337929883487034e-07, | |
| "loss": 0.1191, | |
| "step": 121000 | |
| }, | |
| { | |
| "epoch": 1.1574515108790915, | |
| "grad_norm": 0.040010467171669006, | |
| "learning_rate": 2.1302234355591155e-07, | |
| "loss": 0.107, | |
| "step": 121500 | |
| }, | |
| { | |
| "epoch": 1.1622146858209808, | |
| "grad_norm": 0.3573683202266693, | |
| "learning_rate": 2.1266538827695276e-07, | |
| "loss": 0.1226, | |
| "step": 122000 | |
| }, | |
| { | |
| "epoch": 1.16697786076287, | |
| "grad_norm": 5.653687477111816, | |
| "learning_rate": 2.1230843299799397e-07, | |
| "loss": 0.1171, | |
| "step": 122500 | |
| }, | |
| { | |
| "epoch": 1.1717410357047593, | |
| "grad_norm": 0.8669099807739258, | |
| "learning_rate": 2.1195147771903518e-07, | |
| "loss": 0.1194, | |
| "step": 123000 | |
| }, | |
| { | |
| "epoch": 1.1765042106466486, | |
| "grad_norm": 4.056451320648193, | |
| "learning_rate": 2.115945224400764e-07, | |
| "loss": 0.111, | |
| "step": 123500 | |
| }, | |
| { | |
| "epoch": 1.1812673855885378, | |
| "grad_norm": 20.773231506347656, | |
| "learning_rate": 2.112375671611176e-07, | |
| "loss": 0.1009, | |
| "step": 124000 | |
| }, | |
| { | |
| "epoch": 1.186030560530427, | |
| "grad_norm": 0.029131252318620682, | |
| "learning_rate": 2.108806118821588e-07, | |
| "loss": 0.1076, | |
| "step": 124500 | |
| }, | |
| { | |
| "epoch": 1.1907937354723164, | |
| "grad_norm": 0.8366307020187378, | |
| "learning_rate": 2.1052365660320002e-07, | |
| "loss": 0.1112, | |
| "step": 125000 | |
| }, | |
| { | |
| "epoch": 1.1955569104142056, | |
| "grad_norm": 9.74376106262207, | |
| "learning_rate": 2.1016670132424123e-07, | |
| "loss": 0.1101, | |
| "step": 125500 | |
| }, | |
| { | |
| "epoch": 1.200320085356095, | |
| "grad_norm": 4.660519123077393, | |
| "learning_rate": 2.0980974604528244e-07, | |
| "loss": 0.1084, | |
| "step": 126000 | |
| }, | |
| { | |
| "epoch": 1.2050832602979842, | |
| "grad_norm": 14.676050186157227, | |
| "learning_rate": 2.0945279076632365e-07, | |
| "loss": 0.1186, | |
| "step": 126500 | |
| }, | |
| { | |
| "epoch": 1.2098464352398735, | |
| "grad_norm": 20.924720764160156, | |
| "learning_rate": 2.0909583548736486e-07, | |
| "loss": 0.1106, | |
| "step": 127000 | |
| }, | |
| { | |
| "epoch": 1.2146096101817627, | |
| "grad_norm": 0.0010393880074843764, | |
| "learning_rate": 2.0873888020840607e-07, | |
| "loss": 0.1225, | |
| "step": 127500 | |
| }, | |
| { | |
| "epoch": 1.219372785123652, | |
| "grad_norm": 11.950651168823242, | |
| "learning_rate": 2.0838192492944728e-07, | |
| "loss": 0.1122, | |
| "step": 128000 | |
| }, | |
| { | |
| "epoch": 1.2241359600655413, | |
| "grad_norm": 2.3444035053253174, | |
| "learning_rate": 2.0802496965048849e-07, | |
| "loss": 0.116, | |
| "step": 128500 | |
| }, | |
| { | |
| "epoch": 1.2288991350074305, | |
| "grad_norm": 18.723051071166992, | |
| "learning_rate": 2.0766801437152967e-07, | |
| "loss": 0.1091, | |
| "step": 129000 | |
| }, | |
| { | |
| "epoch": 1.2336623099493198, | |
| "grad_norm": 13.435932159423828, | |
| "learning_rate": 2.0731105909257088e-07, | |
| "loss": 0.1068, | |
| "step": 129500 | |
| }, | |
| { | |
| "epoch": 1.238425484891209, | |
| "grad_norm": 1.0864711999893188, | |
| "learning_rate": 2.069541038136121e-07, | |
| "loss": 0.1266, | |
| "step": 130000 | |
| }, | |
| { | |
| "epoch": 1.2431886598330983, | |
| "grad_norm": 8.090726852416992, | |
| "learning_rate": 2.065971485346533e-07, | |
| "loss": 0.1256, | |
| "step": 130500 | |
| }, | |
| { | |
| "epoch": 1.2479518347749876, | |
| "grad_norm": 3.7775135040283203, | |
| "learning_rate": 2.062401932556945e-07, | |
| "loss": 0.1112, | |
| "step": 131000 | |
| }, | |
| { | |
| "epoch": 1.2527150097168769, | |
| "grad_norm": 0.0022473677527159452, | |
| "learning_rate": 2.0588323797673572e-07, | |
| "loss": 0.12, | |
| "step": 131500 | |
| }, | |
| { | |
| "epoch": 1.2574781846587662, | |
| "grad_norm": 1.232242465019226, | |
| "learning_rate": 2.0552628269777693e-07, | |
| "loss": 0.1023, | |
| "step": 132000 | |
| }, | |
| { | |
| "epoch": 1.2622413596006554, | |
| "grad_norm": 0.16098114848136902, | |
| "learning_rate": 2.0516932741881814e-07, | |
| "loss": 0.1346, | |
| "step": 132500 | |
| }, | |
| { | |
| "epoch": 1.2670045345425447, | |
| "grad_norm": 12.452221870422363, | |
| "learning_rate": 2.0481237213985935e-07, | |
| "loss": 0.1215, | |
| "step": 133000 | |
| }, | |
| { | |
| "epoch": 1.271767709484434, | |
| "grad_norm": 10.727972984313965, | |
| "learning_rate": 2.0445541686090056e-07, | |
| "loss": 0.1062, | |
| "step": 133500 | |
| }, | |
| { | |
| "epoch": 1.2765308844263232, | |
| "grad_norm": 0.5605247020721436, | |
| "learning_rate": 2.0409846158194177e-07, | |
| "loss": 0.1082, | |
| "step": 134000 | |
| }, | |
| { | |
| "epoch": 1.2812940593682125, | |
| "grad_norm": 0.9301961660385132, | |
| "learning_rate": 2.0374150630298298e-07, | |
| "loss": 0.1273, | |
| "step": 134500 | |
| }, | |
| { | |
| "epoch": 1.2860572343101018, | |
| "grad_norm": 0.5364490747451782, | |
| "learning_rate": 2.033845510240242e-07, | |
| "loss": 0.117, | |
| "step": 135000 | |
| }, | |
| { | |
| "epoch": 1.290820409251991, | |
| "grad_norm": 11.421608924865723, | |
| "learning_rate": 2.030275957450654e-07, | |
| "loss": 0.119, | |
| "step": 135500 | |
| }, | |
| { | |
| "epoch": 1.2955835841938803, | |
| "grad_norm": 29.850894927978516, | |
| "learning_rate": 2.026706404661066e-07, | |
| "loss": 0.0931, | |
| "step": 136000 | |
| }, | |
| { | |
| "epoch": 1.3003467591357696, | |
| "grad_norm": 1.1624155044555664, | |
| "learning_rate": 2.0231368518714782e-07, | |
| "loss": 0.1055, | |
| "step": 136500 | |
| }, | |
| { | |
| "epoch": 1.3051099340776589, | |
| "grad_norm": 8.889257431030273, | |
| "learning_rate": 2.0195672990818903e-07, | |
| "loss": 0.1174, | |
| "step": 137000 | |
| }, | |
| { | |
| "epoch": 1.3098731090195481, | |
| "grad_norm": 0.10517348349094391, | |
| "learning_rate": 2.0159977462923024e-07, | |
| "loss": 0.0913, | |
| "step": 137500 | |
| }, | |
| { | |
| "epoch": 1.3146362839614374, | |
| "grad_norm": 3.1619289075024426e-05, | |
| "learning_rate": 2.0124281935027145e-07, | |
| "loss": 0.1053, | |
| "step": 138000 | |
| }, | |
| { | |
| "epoch": 1.3193994589033267, | |
| "grad_norm": 8.190518379211426, | |
| "learning_rate": 2.0088586407131266e-07, | |
| "loss": 0.1017, | |
| "step": 138500 | |
| }, | |
| { | |
| "epoch": 1.324162633845216, | |
| "grad_norm": 8.892984390258789, | |
| "learning_rate": 2.0052890879235387e-07, | |
| "loss": 0.1225, | |
| "step": 139000 | |
| }, | |
| { | |
| "epoch": 1.3289258087871052, | |
| "grad_norm": 0.019881388172507286, | |
| "learning_rate": 2.0017195351339505e-07, | |
| "loss": 0.1109, | |
| "step": 139500 | |
| }, | |
| { | |
| "epoch": 1.3336889837289945, | |
| "grad_norm": 0.39558929204940796, | |
| "learning_rate": 1.9981499823443626e-07, | |
| "loss": 0.1113, | |
| "step": 140000 | |
| }, | |
| { | |
| "epoch": 1.3384521586708837, | |
| "grad_norm": 0.0027440183330327272, | |
| "learning_rate": 1.9945804295547747e-07, | |
| "loss": 0.1111, | |
| "step": 140500 | |
| }, | |
| { | |
| "epoch": 1.343215333612773, | |
| "grad_norm": 0.051865462213754654, | |
| "learning_rate": 1.9910108767651868e-07, | |
| "loss": 0.1001, | |
| "step": 141000 | |
| }, | |
| { | |
| "epoch": 1.3479785085546623, | |
| "grad_norm": 17.542285919189453, | |
| "learning_rate": 1.987441323975599e-07, | |
| "loss": 0.1093, | |
| "step": 141500 | |
| }, | |
| { | |
| "epoch": 1.3527416834965513, | |
| "grad_norm": 1.5649853944778442, | |
| "learning_rate": 1.983871771186011e-07, | |
| "loss": 0.1108, | |
| "step": 142000 | |
| }, | |
| { | |
| "epoch": 1.3575048584384408, | |
| "grad_norm": 3.2729179859161377, | |
| "learning_rate": 1.980302218396423e-07, | |
| "loss": 0.1094, | |
| "step": 142500 | |
| }, | |
| { | |
| "epoch": 1.3622680333803299, | |
| "grad_norm": 19.61039161682129, | |
| "learning_rate": 1.9767326656068352e-07, | |
| "loss": 0.1232, | |
| "step": 143000 | |
| }, | |
| { | |
| "epoch": 1.3670312083222194, | |
| "grad_norm": 0.48026251792907715, | |
| "learning_rate": 1.9731631128172473e-07, | |
| "loss": 0.1179, | |
| "step": 143500 | |
| }, | |
| { | |
| "epoch": 1.3717943832641084, | |
| "grad_norm": 0.011149807833135128, | |
| "learning_rate": 1.9695935600276594e-07, | |
| "loss": 0.1205, | |
| "step": 144000 | |
| }, | |
| { | |
| "epoch": 1.376557558205998, | |
| "grad_norm": 0.02112853154540062, | |
| "learning_rate": 1.9660240072380715e-07, | |
| "loss": 0.1124, | |
| "step": 144500 | |
| }, | |
| { | |
| "epoch": 1.381320733147887, | |
| "grad_norm": 34.17774200439453, | |
| "learning_rate": 1.9624544544484836e-07, | |
| "loss": 0.096, | |
| "step": 145000 | |
| }, | |
| { | |
| "epoch": 1.3860839080897764, | |
| "grad_norm": 0.002524700714275241, | |
| "learning_rate": 1.9588849016588957e-07, | |
| "loss": 0.1279, | |
| "step": 145500 | |
| }, | |
| { | |
| "epoch": 1.3908470830316655, | |
| "grad_norm": 0.4319082498550415, | |
| "learning_rate": 1.9553153488693078e-07, | |
| "loss": 0.1096, | |
| "step": 146000 | |
| }, | |
| { | |
| "epoch": 1.395610257973555, | |
| "grad_norm": 17.499792098999023, | |
| "learning_rate": 1.95174579607972e-07, | |
| "loss": 0.1137, | |
| "step": 146500 | |
| }, | |
| { | |
| "epoch": 1.400373432915444, | |
| "grad_norm": 16.145008087158203, | |
| "learning_rate": 1.948176243290132e-07, | |
| "loss": 0.1051, | |
| "step": 147000 | |
| }, | |
| { | |
| "epoch": 1.4051366078573335, | |
| "grad_norm": 0.8132957816123962, | |
| "learning_rate": 1.944606690500544e-07, | |
| "loss": 0.1063, | |
| "step": 147500 | |
| }, | |
| { | |
| "epoch": 1.4098997827992226, | |
| "grad_norm": 0.05169007182121277, | |
| "learning_rate": 1.9410371377109562e-07, | |
| "loss": 0.1047, | |
| "step": 148000 | |
| }, | |
| { | |
| "epoch": 1.4146629577411118, | |
| "grad_norm": 3.716925859451294, | |
| "learning_rate": 1.9374675849213683e-07, | |
| "loss": 0.1081, | |
| "step": 148500 | |
| }, | |
| { | |
| "epoch": 1.419426132683001, | |
| "grad_norm": 16.134973526000977, | |
| "learning_rate": 1.9338980321317804e-07, | |
| "loss": 0.127, | |
| "step": 149000 | |
| }, | |
| { | |
| "epoch": 1.4241893076248904, | |
| "grad_norm": 0.8836007118225098, | |
| "learning_rate": 1.9303284793421922e-07, | |
| "loss": 0.1116, | |
| "step": 149500 | |
| }, | |
| { | |
| "epoch": 1.4289524825667796, | |
| "grad_norm": 4.612446308135986, | |
| "learning_rate": 1.9267589265526043e-07, | |
| "loss": 0.1263, | |
| "step": 150000 | |
| }, | |
| { | |
| "epoch": 1.433715657508669, | |
| "grad_norm": 0.24257691204547882, | |
| "learning_rate": 1.9231893737630164e-07, | |
| "loss": 0.1067, | |
| "step": 150500 | |
| }, | |
| { | |
| "epoch": 1.4384788324505582, | |
| "grad_norm": 18.72231101989746, | |
| "learning_rate": 1.9196198209734285e-07, | |
| "loss": 0.1174, | |
| "step": 151000 | |
| }, | |
| { | |
| "epoch": 1.4432420073924475, | |
| "grad_norm": 0.09871924668550491, | |
| "learning_rate": 1.9160502681838409e-07, | |
| "loss": 0.1096, | |
| "step": 151500 | |
| }, | |
| { | |
| "epoch": 1.4480051823343367, | |
| "grad_norm": 1.2534486055374146, | |
| "learning_rate": 1.912480715394253e-07, | |
| "loss": 0.1155, | |
| "step": 152000 | |
| }, | |
| { | |
| "epoch": 1.452768357276226, | |
| "grad_norm": 14.395082473754883, | |
| "learning_rate": 1.908911162604665e-07, | |
| "loss": 0.1047, | |
| "step": 152500 | |
| }, | |
| { | |
| "epoch": 1.4575315322181153, | |
| "grad_norm": 2.9728424549102783, | |
| "learning_rate": 1.9053416098150771e-07, | |
| "loss": 0.0942, | |
| "step": 153000 | |
| }, | |
| { | |
| "epoch": 1.4622947071600045, | |
| "grad_norm": 10.138772964477539, | |
| "learning_rate": 1.9017720570254892e-07, | |
| "loss": 0.1219, | |
| "step": 153500 | |
| }, | |
| { | |
| "epoch": 1.4670578821018938, | |
| "grad_norm": 16.113460540771484, | |
| "learning_rate": 1.8982025042359013e-07, | |
| "loss": 0.1094, | |
| "step": 154000 | |
| }, | |
| { | |
| "epoch": 1.471821057043783, | |
| "grad_norm": 1.4605754613876343, | |
| "learning_rate": 1.8946329514463134e-07, | |
| "loss": 0.1227, | |
| "step": 154500 | |
| }, | |
| { | |
| "epoch": 1.4765842319856723, | |
| "grad_norm": 19.34916877746582, | |
| "learning_rate": 1.8910633986567255e-07, | |
| "loss": 0.1032, | |
| "step": 155000 | |
| }, | |
| { | |
| "epoch": 1.4813474069275616, | |
| "grad_norm": 5.430410861968994, | |
| "learning_rate": 1.8874938458671376e-07, | |
| "loss": 0.1107, | |
| "step": 155500 | |
| }, | |
| { | |
| "epoch": 1.4861105818694509, | |
| "grad_norm": 0.01834196224808693, | |
| "learning_rate": 1.8839242930775497e-07, | |
| "loss": 0.1243, | |
| "step": 156000 | |
| }, | |
| { | |
| "epoch": 1.4908737568113402, | |
| "grad_norm": 22.170907974243164, | |
| "learning_rate": 1.8803547402879618e-07, | |
| "loss": 0.1019, | |
| "step": 156500 | |
| }, | |
| { | |
| "epoch": 1.4956369317532294, | |
| "grad_norm": 0.28510582447052, | |
| "learning_rate": 1.876785187498374e-07, | |
| "loss": 0.1354, | |
| "step": 157000 | |
| }, | |
| { | |
| "epoch": 1.5004001066951187, | |
| "grad_norm": 8.662985801696777, | |
| "learning_rate": 1.873215634708786e-07, | |
| "loss": 0.1043, | |
| "step": 157500 | |
| }, | |
| { | |
| "epoch": 1.505163281637008, | |
| "grad_norm": 5.042104721069336, | |
| "learning_rate": 1.869646081919198e-07, | |
| "loss": 0.1102, | |
| "step": 158000 | |
| }, | |
| { | |
| "epoch": 1.5099264565788972, | |
| "grad_norm": 0.07364173978567123, | |
| "learning_rate": 1.8660765291296102e-07, | |
| "loss": 0.129, | |
| "step": 158500 | |
| }, | |
| { | |
| "epoch": 1.5146896315207865, | |
| "grad_norm": 3.405261278152466, | |
| "learning_rate": 1.862506976340022e-07, | |
| "loss": 0.111, | |
| "step": 159000 | |
| }, | |
| { | |
| "epoch": 1.5194528064626758, | |
| "grad_norm": 3.3657171726226807, | |
| "learning_rate": 1.8589374235504342e-07, | |
| "loss": 0.1089, | |
| "step": 159500 | |
| }, | |
| { | |
| "epoch": 1.524215981404565, | |
| "grad_norm": 0.028935715556144714, | |
| "learning_rate": 1.8553678707608463e-07, | |
| "loss": 0.1038, | |
| "step": 160000 | |
| }, | |
| { | |
| "epoch": 1.5289791563464543, | |
| "grad_norm": 1.048443078994751, | |
| "learning_rate": 1.8517983179712584e-07, | |
| "loss": 0.0938, | |
| "step": 160500 | |
| }, | |
| { | |
| "epoch": 1.5337423312883436, | |
| "grad_norm": 15.945822715759277, | |
| "learning_rate": 1.8482287651816705e-07, | |
| "loss": 0.1138, | |
| "step": 161000 | |
| }, | |
| { | |
| "epoch": 1.5385055062302329, | |
| "grad_norm": 0.36497658491134644, | |
| "learning_rate": 1.8446592123920826e-07, | |
| "loss": 0.107, | |
| "step": 161500 | |
| }, | |
| { | |
| "epoch": 1.5432686811721221, | |
| "grad_norm": 0.7298380732536316, | |
| "learning_rate": 1.8410896596024947e-07, | |
| "loss": 0.1012, | |
| "step": 162000 | |
| }, | |
| { | |
| "epoch": 1.5480318561140114, | |
| "grad_norm": 0.041379645466804504, | |
| "learning_rate": 1.8375201068129068e-07, | |
| "loss": 0.1171, | |
| "step": 162500 | |
| }, | |
| { | |
| "epoch": 1.5527950310559007, | |
| "grad_norm": 1.361636996269226, | |
| "learning_rate": 1.8339505540233188e-07, | |
| "loss": 0.0978, | |
| "step": 163000 | |
| }, | |
| { | |
| "epoch": 1.55755820599779, | |
| "grad_norm": 0.0006419995916076005, | |
| "learning_rate": 1.830381001233731e-07, | |
| "loss": 0.1152, | |
| "step": 163500 | |
| }, | |
| { | |
| "epoch": 1.5623213809396792, | |
| "grad_norm": 29.9415340423584, | |
| "learning_rate": 1.826811448444143e-07, | |
| "loss": 0.1124, | |
| "step": 164000 | |
| }, | |
| { | |
| "epoch": 1.5670845558815683, | |
| "grad_norm": 14.25601577758789, | |
| "learning_rate": 1.8232418956545551e-07, | |
| "loss": 0.1096, | |
| "step": 164500 | |
| }, | |
| { | |
| "epoch": 1.5718477308234577, | |
| "grad_norm": 10.357718467712402, | |
| "learning_rate": 1.8196723428649672e-07, | |
| "loss": 0.1068, | |
| "step": 165000 | |
| }, | |
| { | |
| "epoch": 1.5766109057653468, | |
| "grad_norm": 2.6006975173950195, | |
| "learning_rate": 1.8161027900753793e-07, | |
| "loss": 0.112, | |
| "step": 165500 | |
| }, | |
| { | |
| "epoch": 1.5813740807072363, | |
| "grad_norm": 0.013517569750547409, | |
| "learning_rate": 1.8125332372857914e-07, | |
| "loss": 0.1046, | |
| "step": 166000 | |
| }, | |
| { | |
| "epoch": 1.5861372556491253, | |
| "grad_norm": 8.826921463012695, | |
| "learning_rate": 1.8089636844962035e-07, | |
| "loss": 0.1148, | |
| "step": 166500 | |
| }, | |
| { | |
| "epoch": 1.5909004305910148, | |
| "grad_norm": 30.500505447387695, | |
| "learning_rate": 1.8053941317066156e-07, | |
| "loss": 0.1022, | |
| "step": 167000 | |
| }, | |
| { | |
| "epoch": 1.5956636055329039, | |
| "grad_norm": 0.0015779563691467047, | |
| "learning_rate": 1.8018245789170277e-07, | |
| "loss": 0.0967, | |
| "step": 167500 | |
| }, | |
| { | |
| "epoch": 1.6004267804747934, | |
| "grad_norm": 5.478053092956543, | |
| "learning_rate": 1.7982550261274398e-07, | |
| "loss": 0.1134, | |
| "step": 168000 | |
| }, | |
| { | |
| "epoch": 1.6051899554166824, | |
| "grad_norm": 0.11672214418649673, | |
| "learning_rate": 1.794685473337852e-07, | |
| "loss": 0.1, | |
| "step": 168500 | |
| }, | |
| { | |
| "epoch": 1.609953130358572, | |
| "grad_norm": 1.3032927513122559, | |
| "learning_rate": 1.791115920548264e-07, | |
| "loss": 0.1165, | |
| "step": 169000 | |
| }, | |
| { | |
| "epoch": 1.614716305300461, | |
| "grad_norm": 0.20952224731445312, | |
| "learning_rate": 1.7875463677586759e-07, | |
| "loss": 0.112, | |
| "step": 169500 | |
| }, | |
| { | |
| "epoch": 1.6194794802423504, | |
| "grad_norm": 5.338772296905518, | |
| "learning_rate": 1.783976814969088e-07, | |
| "loss": 0.1151, | |
| "step": 170000 | |
| }, | |
| { | |
| "epoch": 1.6242426551842395, | |
| "grad_norm": 14.308218955993652, | |
| "learning_rate": 1.7804072621795e-07, | |
| "loss": 0.1188, | |
| "step": 170500 | |
| }, | |
| { | |
| "epoch": 1.629005830126129, | |
| "grad_norm": 2.5902771949768066, | |
| "learning_rate": 1.7768377093899122e-07, | |
| "loss": 0.1095, | |
| "step": 171000 | |
| }, | |
| { | |
| "epoch": 1.633769005068018, | |
| "grad_norm": 1.145963191986084, | |
| "learning_rate": 1.7732681566003243e-07, | |
| "loss": 0.1081, | |
| "step": 171500 | |
| }, | |
| { | |
| "epoch": 1.6385321800099075, | |
| "grad_norm": 0.019109368324279785, | |
| "learning_rate": 1.7696986038107364e-07, | |
| "loss": 0.1037, | |
| "step": 172000 | |
| }, | |
| { | |
| "epoch": 1.6432953549517966, | |
| "grad_norm": 0.381287544965744, | |
| "learning_rate": 1.7661290510211485e-07, | |
| "loss": 0.1024, | |
| "step": 172500 | |
| }, | |
| { | |
| "epoch": 1.648058529893686, | |
| "grad_norm": 0.28823208808898926, | |
| "learning_rate": 1.7625594982315606e-07, | |
| "loss": 0.1112, | |
| "step": 173000 | |
| }, | |
| { | |
| "epoch": 1.6528217048355751, | |
| "grad_norm": 7.33282995223999, | |
| "learning_rate": 1.7589899454419726e-07, | |
| "loss": 0.1037, | |
| "step": 173500 | |
| }, | |
| { | |
| "epoch": 1.6575848797774646, | |
| "grad_norm": 18.685562133789062, | |
| "learning_rate": 1.7554203926523847e-07, | |
| "loss": 0.1096, | |
| "step": 174000 | |
| }, | |
| { | |
| "epoch": 1.6623480547193537, | |
| "grad_norm": 5.129574298858643, | |
| "learning_rate": 1.7518508398627968e-07, | |
| "loss": 0.1028, | |
| "step": 174500 | |
| }, | |
| { | |
| "epoch": 1.6671112296612431, | |
| "grad_norm": 1.4392452239990234, | |
| "learning_rate": 1.748281287073209e-07, | |
| "loss": 0.1106, | |
| "step": 175000 | |
| }, | |
| { | |
| "epoch": 1.6718744046031322, | |
| "grad_norm": 3.51692795753479, | |
| "learning_rate": 1.744711734283621e-07, | |
| "loss": 0.1142, | |
| "step": 175500 | |
| }, | |
| { | |
| "epoch": 1.6766375795450217, | |
| "grad_norm": 0.026211030781269073, | |
| "learning_rate": 1.7411421814940331e-07, | |
| "loss": 0.1083, | |
| "step": 176000 | |
| }, | |
| { | |
| "epoch": 1.6814007544869107, | |
| "grad_norm": 4.911788463592529, | |
| "learning_rate": 1.7375726287044452e-07, | |
| "loss": 0.1053, | |
| "step": 176500 | |
| }, | |
| { | |
| "epoch": 1.6861639294288002, | |
| "grad_norm": 5.630278587341309, | |
| "learning_rate": 1.7340030759148573e-07, | |
| "loss": 0.1087, | |
| "step": 177000 | |
| }, | |
| { | |
| "epoch": 1.6909271043706893, | |
| "grad_norm": 1.8357000350952148, | |
| "learning_rate": 1.7304335231252694e-07, | |
| "loss": 0.104, | |
| "step": 177500 | |
| }, | |
| { | |
| "epoch": 1.6956902793125785, | |
| "grad_norm": 2.734142780303955, | |
| "learning_rate": 1.7268639703356815e-07, | |
| "loss": 0.1339, | |
| "step": 178000 | |
| }, | |
| { | |
| "epoch": 1.7004534542544678, | |
| "grad_norm": 27.017423629760742, | |
| "learning_rate": 1.7232944175460936e-07, | |
| "loss": 0.1085, | |
| "step": 178500 | |
| }, | |
| { | |
| "epoch": 1.705216629196357, | |
| "grad_norm": 0.3949691653251648, | |
| "learning_rate": 1.7197248647565057e-07, | |
| "loss": 0.1074, | |
| "step": 179000 | |
| }, | |
| { | |
| "epoch": 1.7099798041382464, | |
| "grad_norm": 14.20468807220459, | |
| "learning_rate": 1.7161553119669176e-07, | |
| "loss": 0.099, | |
| "step": 179500 | |
| }, | |
| { | |
| "epoch": 1.7147429790801356, | |
| "grad_norm": 0.0762014091014862, | |
| "learning_rate": 1.7125857591773297e-07, | |
| "loss": 0.1108, | |
| "step": 180000 | |
| }, | |
| { | |
| "epoch": 1.719506154022025, | |
| "grad_norm": 19.510847091674805, | |
| "learning_rate": 1.7090162063877418e-07, | |
| "loss": 0.1225, | |
| "step": 180500 | |
| }, | |
| { | |
| "epoch": 1.7242693289639142, | |
| "grad_norm": 0.08469968289136887, | |
| "learning_rate": 1.7054466535981539e-07, | |
| "loss": 0.1015, | |
| "step": 181000 | |
| }, | |
| { | |
| "epoch": 1.7290325039058034, | |
| "grad_norm": 30.91697120666504, | |
| "learning_rate": 1.701877100808566e-07, | |
| "loss": 0.108, | |
| "step": 181500 | |
| }, | |
| { | |
| "epoch": 1.7337956788476927, | |
| "grad_norm": 5.243447780609131, | |
| "learning_rate": 1.698307548018978e-07, | |
| "loss": 0.1279, | |
| "step": 182000 | |
| }, | |
| { | |
| "epoch": 1.738558853789582, | |
| "grad_norm": 19.558759689331055, | |
| "learning_rate": 1.6947379952293902e-07, | |
| "loss": 0.1043, | |
| "step": 182500 | |
| }, | |
| { | |
| "epoch": 1.7433220287314712, | |
| "grad_norm": 18.18353271484375, | |
| "learning_rate": 1.6911684424398023e-07, | |
| "loss": 0.1005, | |
| "step": 183000 | |
| }, | |
| { | |
| "epoch": 1.7480852036733605, | |
| "grad_norm": 0.06688349694013596, | |
| "learning_rate": 1.6875988896502144e-07, | |
| "loss": 0.1021, | |
| "step": 183500 | |
| }, | |
| { | |
| "epoch": 1.7528483786152498, | |
| "grad_norm": 3.162032127380371, | |
| "learning_rate": 1.6840293368606264e-07, | |
| "loss": 0.1244, | |
| "step": 184000 | |
| }, | |
| { | |
| "epoch": 1.757611553557139, | |
| "grad_norm": 0.6052823662757874, | |
| "learning_rate": 1.6804597840710385e-07, | |
| "loss": 0.1145, | |
| "step": 184500 | |
| }, | |
| { | |
| "epoch": 1.7623747284990283, | |
| "grad_norm": 12.853279113769531, | |
| "learning_rate": 1.6768902312814506e-07, | |
| "loss": 0.1004, | |
| "step": 185000 | |
| }, | |
| { | |
| "epoch": 1.7671379034409176, | |
| "grad_norm": 0.010185870341956615, | |
| "learning_rate": 1.6733206784918627e-07, | |
| "loss": 0.1048, | |
| "step": 185500 | |
| }, | |
| { | |
| "epoch": 1.7719010783828069, | |
| "grad_norm": 7.452764839399606e-05, | |
| "learning_rate": 1.6697511257022748e-07, | |
| "loss": 0.1, | |
| "step": 186000 | |
| }, | |
| { | |
| "epoch": 1.7766642533246961, | |
| "grad_norm": 17.169069290161133, | |
| "learning_rate": 1.666181572912687e-07, | |
| "loss": 0.1176, | |
| "step": 186500 | |
| }, | |
| { | |
| "epoch": 1.7814274282665854, | |
| "grad_norm": 12.43346118927002, | |
| "learning_rate": 1.662612020123099e-07, | |
| "loss": 0.111, | |
| "step": 187000 | |
| }, | |
| { | |
| "epoch": 1.7861906032084747, | |
| "grad_norm": 13.447138786315918, | |
| "learning_rate": 1.6590424673335111e-07, | |
| "loss": 0.1198, | |
| "step": 187500 | |
| }, | |
| { | |
| "epoch": 1.790953778150364, | |
| "grad_norm": 2.679323673248291, | |
| "learning_rate": 1.6554729145439232e-07, | |
| "loss": 0.1083, | |
| "step": 188000 | |
| }, | |
| { | |
| "epoch": 1.7957169530922532, | |
| "grad_norm": 12.469381332397461, | |
| "learning_rate": 1.6519033617543353e-07, | |
| "loss": 0.0948, | |
| "step": 188500 | |
| }, | |
| { | |
| "epoch": 1.8004801280341425, | |
| "grad_norm": 5.650728225708008, | |
| "learning_rate": 1.6483338089647474e-07, | |
| "loss": 0.1329, | |
| "step": 189000 | |
| }, | |
| { | |
| "epoch": 1.8052433029760317, | |
| "grad_norm": 0.8955023884773254, | |
| "learning_rate": 1.6447642561751593e-07, | |
| "loss": 0.0933, | |
| "step": 189500 | |
| }, | |
| { | |
| "epoch": 1.810006477917921, | |
| "grad_norm": 26.87681770324707, | |
| "learning_rate": 1.6411947033855714e-07, | |
| "loss": 0.1015, | |
| "step": 190000 | |
| }, | |
| { | |
| "epoch": 1.8147696528598103, | |
| "grad_norm": 4.95468282699585, | |
| "learning_rate": 1.6376251505959835e-07, | |
| "loss": 0.1067, | |
| "step": 190500 | |
| }, | |
| { | |
| "epoch": 1.8195328278016993, | |
| "grad_norm": 0.931701123714447, | |
| "learning_rate": 1.6340555978063958e-07, | |
| "loss": 0.1282, | |
| "step": 191000 | |
| }, | |
| { | |
| "epoch": 1.8242960027435888, | |
| "grad_norm": 7.3879780769348145, | |
| "learning_rate": 1.630486045016808e-07, | |
| "loss": 0.0935, | |
| "step": 191500 | |
| }, | |
| { | |
| "epoch": 1.8290591776854779, | |
| "grad_norm": 0.022659489884972572, | |
| "learning_rate": 1.62691649222722e-07, | |
| "loss": 0.1086, | |
| "step": 192000 | |
| }, | |
| { | |
| "epoch": 1.8338223526273674, | |
| "grad_norm": 0.1140231192111969, | |
| "learning_rate": 1.623346939437632e-07, | |
| "loss": 0.1005, | |
| "step": 192500 | |
| }, | |
| { | |
| "epoch": 1.8385855275692564, | |
| "grad_norm": 0.6795700192451477, | |
| "learning_rate": 1.6197773866480442e-07, | |
| "loss": 0.11, | |
| "step": 193000 | |
| }, | |
| { | |
| "epoch": 1.843348702511146, | |
| "grad_norm": 9.37308406829834, | |
| "learning_rate": 1.6162078338584563e-07, | |
| "loss": 0.1224, | |
| "step": 193500 | |
| }, | |
| { | |
| "epoch": 1.848111877453035, | |
| "grad_norm": 33.167701721191406, | |
| "learning_rate": 1.6126382810688684e-07, | |
| "loss": 0.1122, | |
| "step": 194000 | |
| }, | |
| { | |
| "epoch": 1.8528750523949244, | |
| "grad_norm": 0.8446443676948547, | |
| "learning_rate": 1.6090687282792805e-07, | |
| "loss": 0.1098, | |
| "step": 194500 | |
| }, | |
| { | |
| "epoch": 1.8576382273368135, | |
| "grad_norm": 30.448610305786133, | |
| "learning_rate": 1.6054991754896926e-07, | |
| "loss": 0.1041, | |
| "step": 195000 | |
| }, | |
| { | |
| "epoch": 1.862401402278703, | |
| "grad_norm": 21.807527542114258, | |
| "learning_rate": 1.6019296227001047e-07, | |
| "loss": 0.097, | |
| "step": 195500 | |
| }, | |
| { | |
| "epoch": 1.867164577220592, | |
| "grad_norm": 0.030451184138655663, | |
| "learning_rate": 1.5983600699105168e-07, | |
| "loss": 0.1073, | |
| "step": 196000 | |
| }, | |
| { | |
| "epoch": 1.8719277521624815, | |
| "grad_norm": 0.03133384510874748, | |
| "learning_rate": 1.594790517120929e-07, | |
| "loss": 0.1123, | |
| "step": 196500 | |
| }, | |
| { | |
| "epoch": 1.8766909271043706, | |
| "grad_norm": 2.491389513015747, | |
| "learning_rate": 1.591220964331341e-07, | |
| "loss": 0.0996, | |
| "step": 197000 | |
| }, | |
| { | |
| "epoch": 1.88145410204626, | |
| "grad_norm": 29.67350959777832, | |
| "learning_rate": 1.587651411541753e-07, | |
| "loss": 0.0885, | |
| "step": 197500 | |
| }, | |
| { | |
| "epoch": 1.8862172769881491, | |
| "grad_norm": 0.28274720907211304, | |
| "learning_rate": 1.5840818587521652e-07, | |
| "loss": 0.1085, | |
| "step": 198000 | |
| }, | |
| { | |
| "epoch": 1.8909804519300386, | |
| "grad_norm": 0.04003256559371948, | |
| "learning_rate": 1.5805123059625773e-07, | |
| "loss": 0.1044, | |
| "step": 198500 | |
| }, | |
| { | |
| "epoch": 1.8957436268719277, | |
| "grad_norm": 44.06182861328125, | |
| "learning_rate": 1.5769427531729894e-07, | |
| "loss": 0.1161, | |
| "step": 199000 | |
| }, | |
| { | |
| "epoch": 1.9005068018138171, | |
| "grad_norm": 3.635102621046826e-05, | |
| "learning_rate": 1.5733732003834012e-07, | |
| "loss": 0.112, | |
| "step": 199500 | |
| }, | |
| { | |
| "epoch": 1.9052699767557062, | |
| "grad_norm": 0.03733281418681145, | |
| "learning_rate": 1.5698036475938133e-07, | |
| "loss": 0.1241, | |
| "step": 200000 | |
| }, | |
| { | |
| "epoch": 1.9100331516975957, | |
| "grad_norm": 0.009642825461924076, | |
| "learning_rate": 1.5662340948042254e-07, | |
| "loss": 0.1116, | |
| "step": 200500 | |
| }, | |
| { | |
| "epoch": 1.9147963266394847, | |
| "grad_norm": 0.04384785518050194, | |
| "learning_rate": 1.5626645420146375e-07, | |
| "loss": 0.0975, | |
| "step": 201000 | |
| }, | |
| { | |
| "epoch": 1.9195595015813742, | |
| "grad_norm": 28.704349517822266, | |
| "learning_rate": 1.5590949892250496e-07, | |
| "loss": 0.1376, | |
| "step": 201500 | |
| }, | |
| { | |
| "epoch": 1.9243226765232633, | |
| "grad_norm": 0.0023307004012167454, | |
| "learning_rate": 1.5555254364354617e-07, | |
| "loss": 0.1216, | |
| "step": 202000 | |
| }, | |
| { | |
| "epoch": 1.9290858514651528, | |
| "grad_norm": 3.073564291000366, | |
| "learning_rate": 1.5519558836458738e-07, | |
| "loss": 0.1139, | |
| "step": 202500 | |
| }, | |
| { | |
| "epoch": 1.9338490264070418, | |
| "grad_norm": 0.6808690428733826, | |
| "learning_rate": 1.548386330856286e-07, | |
| "loss": 0.0976, | |
| "step": 203000 | |
| }, | |
| { | |
| "epoch": 1.9386122013489313, | |
| "grad_norm": 5.560125350952148, | |
| "learning_rate": 1.544816778066698e-07, | |
| "loss": 0.1022, | |
| "step": 203500 | |
| }, | |
| { | |
| "epoch": 1.9433753762908204, | |
| "grad_norm": 0.019307024776935577, | |
| "learning_rate": 1.54124722527711e-07, | |
| "loss": 0.1042, | |
| "step": 204000 | |
| }, | |
| { | |
| "epoch": 1.9481385512327096, | |
| "grad_norm": 0.23428542912006378, | |
| "learning_rate": 1.5376776724875222e-07, | |
| "loss": 0.1173, | |
| "step": 204500 | |
| }, | |
| { | |
| "epoch": 1.952901726174599, | |
| "grad_norm": 5.58413553237915, | |
| "learning_rate": 1.5341081196979343e-07, | |
| "loss": 0.1216, | |
| "step": 205000 | |
| }, | |
| { | |
| "epoch": 1.9576649011164882, | |
| "grad_norm": 2.8649184703826904, | |
| "learning_rate": 1.5305385669083464e-07, | |
| "loss": 0.1034, | |
| "step": 205500 | |
| }, | |
| { | |
| "epoch": 1.9624280760583774, | |
| "grad_norm": 0.037286777049303055, | |
| "learning_rate": 1.5269690141187585e-07, | |
| "loss": 0.1011, | |
| "step": 206000 | |
| }, | |
| { | |
| "epoch": 1.9671912510002667, | |
| "grad_norm": 0.007163808681070805, | |
| "learning_rate": 1.5233994613291706e-07, | |
| "loss": 0.1087, | |
| "step": 206500 | |
| }, | |
| { | |
| "epoch": 1.971954425942156, | |
| "grad_norm": 24.465059280395508, | |
| "learning_rate": 1.5198299085395827e-07, | |
| "loss": 0.1122, | |
| "step": 207000 | |
| }, | |
| { | |
| "epoch": 1.9767176008840452, | |
| "grad_norm": 6.201001167297363, | |
| "learning_rate": 1.5162603557499948e-07, | |
| "loss": 0.0882, | |
| "step": 207500 | |
| }, | |
| { | |
| "epoch": 1.9814807758259345, | |
| "grad_norm": 0.19351300597190857, | |
| "learning_rate": 1.512690802960407e-07, | |
| "loss": 0.1109, | |
| "step": 208000 | |
| }, | |
| { | |
| "epoch": 1.9862439507678238, | |
| "grad_norm": 0.0022748825140297413, | |
| "learning_rate": 1.509121250170819e-07, | |
| "loss": 0.1045, | |
| "step": 208500 | |
| }, | |
| { | |
| "epoch": 1.991007125709713, | |
| "grad_norm": 0.00041236350079998374, | |
| "learning_rate": 1.505551697381231e-07, | |
| "loss": 0.1095, | |
| "step": 209000 | |
| }, | |
| { | |
| "epoch": 1.9957703006516023, | |
| "grad_norm": 11.131386756896973, | |
| "learning_rate": 1.501982144591643e-07, | |
| "loss": 0.1293, | |
| "step": 209500 | |
| }, | |
| { | |
| "epoch": 2.0005334755934916, | |
| "grad_norm": 0.002583843655884266, | |
| "learning_rate": 1.498412591802055e-07, | |
| "loss": 0.1142, | |
| "step": 210000 | |
| }, | |
| { | |
| "epoch": 2.0052966505353806, | |
| "grad_norm": 14.562299728393555, | |
| "learning_rate": 1.494843039012467e-07, | |
| "loss": 0.1048, | |
| "step": 210500 | |
| }, | |
| { | |
| "epoch": 2.01005982547727, | |
| "grad_norm": 4.4846510887146, | |
| "learning_rate": 1.4912734862228792e-07, | |
| "loss": 0.1019, | |
| "step": 211000 | |
| }, | |
| { | |
| "epoch": 2.014823000419159, | |
| "grad_norm": 0.0017532992642372847, | |
| "learning_rate": 1.4877039334332913e-07, | |
| "loss": 0.1092, | |
| "step": 211500 | |
| }, | |
| { | |
| "epoch": 2.0195861753610487, | |
| "grad_norm": 0.005973345600068569, | |
| "learning_rate": 1.4841343806437034e-07, | |
| "loss": 0.099, | |
| "step": 212000 | |
| }, | |
| { | |
| "epoch": 2.0243493503029377, | |
| "grad_norm": 25.91645622253418, | |
| "learning_rate": 1.4805648278541155e-07, | |
| "loss": 0.1197, | |
| "step": 212500 | |
| }, | |
| { | |
| "epoch": 2.029112525244827, | |
| "grad_norm": 0.009553452022373676, | |
| "learning_rate": 1.4769952750645276e-07, | |
| "loss": 0.1058, | |
| "step": 213000 | |
| }, | |
| { | |
| "epoch": 2.0338757001867163, | |
| "grad_norm": 2.3231234550476074, | |
| "learning_rate": 1.4734257222749397e-07, | |
| "loss": 0.1098, | |
| "step": 213500 | |
| }, | |
| { | |
| "epoch": 2.0386388751286058, | |
| "grad_norm": 7.085731506347656, | |
| "learning_rate": 1.4698561694853518e-07, | |
| "loss": 0.0992, | |
| "step": 214000 | |
| }, | |
| { | |
| "epoch": 2.043402050070495, | |
| "grad_norm": 0.2953038215637207, | |
| "learning_rate": 1.466286616695764e-07, | |
| "loss": 0.1126, | |
| "step": 214500 | |
| }, | |
| { | |
| "epoch": 2.0481652250123843, | |
| "grad_norm": 0.0010392339900135994, | |
| "learning_rate": 1.462717063906176e-07, | |
| "loss": 0.1034, | |
| "step": 215000 | |
| }, | |
| { | |
| "epoch": 2.0529283999542733, | |
| "grad_norm": 0.07149825990200043, | |
| "learning_rate": 1.459147511116588e-07, | |
| "loss": 0.1008, | |
| "step": 215500 | |
| }, | |
| { | |
| "epoch": 2.057691574896163, | |
| "grad_norm": 0.594185471534729, | |
| "learning_rate": 1.4555779583270002e-07, | |
| "loss": 0.1004, | |
| "step": 216000 | |
| }, | |
| { | |
| "epoch": 2.062454749838052, | |
| "grad_norm": 2.347266912460327, | |
| "learning_rate": 1.4520084055374123e-07, | |
| "loss": 0.1087, | |
| "step": 216500 | |
| }, | |
| { | |
| "epoch": 2.0672179247799414, | |
| "grad_norm": 0.16203026473522186, | |
| "learning_rate": 1.4484388527478244e-07, | |
| "loss": 0.1065, | |
| "step": 217000 | |
| }, | |
| { | |
| "epoch": 2.0719810997218304, | |
| "grad_norm": 0.03992275521159172, | |
| "learning_rate": 1.4448692999582365e-07, | |
| "loss": 0.1247, | |
| "step": 217500 | |
| }, | |
| { | |
| "epoch": 2.07674427466372, | |
| "grad_norm": 0.5468718409538269, | |
| "learning_rate": 1.4412997471686486e-07, | |
| "loss": 0.1078, | |
| "step": 218000 | |
| }, | |
| { | |
| "epoch": 2.081507449605609, | |
| "grad_norm": 7.58856201171875, | |
| "learning_rate": 1.4377301943790607e-07, | |
| "loss": 0.0921, | |
| "step": 218500 | |
| }, | |
| { | |
| "epoch": 2.0862706245474985, | |
| "grad_norm": 17.591888427734375, | |
| "learning_rate": 1.4341606415894728e-07, | |
| "loss": 0.1105, | |
| "step": 219000 | |
| }, | |
| { | |
| "epoch": 2.0910337994893875, | |
| "grad_norm": 9.395910263061523, | |
| "learning_rate": 1.4305910887998846e-07, | |
| "loss": 0.1021, | |
| "step": 219500 | |
| }, | |
| { | |
| "epoch": 2.095796974431277, | |
| "grad_norm": 0.0014383163070306182, | |
| "learning_rate": 1.4270215360102967e-07, | |
| "loss": 0.1089, | |
| "step": 220000 | |
| }, | |
| { | |
| "epoch": 2.100560149373166, | |
| "grad_norm": 0.11439737677574158, | |
| "learning_rate": 1.4234519832207088e-07, | |
| "loss": 0.1108, | |
| "step": 220500 | |
| }, | |
| { | |
| "epoch": 2.1053233243150555, | |
| "grad_norm": 1.5022146701812744, | |
| "learning_rate": 1.419882430431121e-07, | |
| "loss": 0.1098, | |
| "step": 221000 | |
| }, | |
| { | |
| "epoch": 2.1100864992569446, | |
| "grad_norm": 0.004490518942475319, | |
| "learning_rate": 1.416312877641533e-07, | |
| "loss": 0.1063, | |
| "step": 221500 | |
| }, | |
| { | |
| "epoch": 2.114849674198834, | |
| "grad_norm": 5.6877360343933105, | |
| "learning_rate": 1.412743324851945e-07, | |
| "loss": 0.1107, | |
| "step": 222000 | |
| }, | |
| { | |
| "epoch": 2.119612849140723, | |
| "grad_norm": 1.135617733001709, | |
| "learning_rate": 1.4091737720623572e-07, | |
| "loss": 0.115, | |
| "step": 222500 | |
| }, | |
| { | |
| "epoch": 2.1243760240826126, | |
| "grad_norm": 2.0897531509399414, | |
| "learning_rate": 1.4056042192727693e-07, | |
| "loss": 0.1242, | |
| "step": 223000 | |
| }, | |
| { | |
| "epoch": 2.1291391990245017, | |
| "grad_norm": 12.82266902923584, | |
| "learning_rate": 1.4020346664831814e-07, | |
| "loss": 0.1066, | |
| "step": 223500 | |
| }, | |
| { | |
| "epoch": 2.133902373966391, | |
| "grad_norm": 0.15470212697982788, | |
| "learning_rate": 1.3984651136935935e-07, | |
| "loss": 0.0952, | |
| "step": 224000 | |
| }, | |
| { | |
| "epoch": 2.13866554890828, | |
| "grad_norm": 7.721962174400687e-05, | |
| "learning_rate": 1.3948955609040056e-07, | |
| "loss": 0.1237, | |
| "step": 224500 | |
| }, | |
| { | |
| "epoch": 2.1434287238501697, | |
| "grad_norm": 0.023135853931307793, | |
| "learning_rate": 1.3913260081144177e-07, | |
| "loss": 0.1118, | |
| "step": 225000 | |
| }, | |
| { | |
| "epoch": 2.1481918987920587, | |
| "grad_norm": 1.2089338302612305, | |
| "learning_rate": 1.3877564553248298e-07, | |
| "loss": 0.0944, | |
| "step": 225500 | |
| }, | |
| { | |
| "epoch": 2.1529550737339482, | |
| "grad_norm": 7.6081109046936035, | |
| "learning_rate": 1.384186902535242e-07, | |
| "loss": 0.1025, | |
| "step": 226000 | |
| }, | |
| { | |
| "epoch": 2.1577182486758373, | |
| "grad_norm": 11.101881980895996, | |
| "learning_rate": 1.380617349745654e-07, | |
| "loss": 0.081, | |
| "step": 226500 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 314916, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.2593962720815494e+17, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |